article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
electromagnetic waves in free space constitute an important part of the calculus based introductory physics course .
in addition to other properties , all the textbooks we have surveyed include a discussion of the transverse nature of the waves . specifically , that ( plane ) electromagnetic waves have the electric field strength @xmath0 and the magnetic induction @xmath1 perpendicular to the propagation direction of the wave and to each other so that @xmath2 is an even permutation of @xmath3 , where @xmath4 is the poynting vector .
the extent of the coverage and the arguments brought in the textbooks vary widely .
while we have not checked all textbooks , none of the ones we have surveyed include a full and detailed theoretical argument for these important properties , even though all textbooks include the infrastructure needed for that purpose , i.e. , the ( integral ) maxwell equations .
most textbooks derive the wave equations for the transverse components of the @xmath0 and @xmath1 fields , but do not show that these wave equations are unique , i.e. , that a longitudinal component violates the maxwell equations .
that is , they show that transverse waves are _ consistent _ with the maxwell equations , but not that they are required .
we make the case here that the theoretical argument is important for students of the calculus based introductory course .
much research has shown that understanding and retention of physics is enhanced if important concepts are not brought just as factoids , but rather derived from theoretical and physical argumentation . in the case of transversality of ( plane ) electromagnetic waves our case is even strengthened , as the argument is certainly at the right level for students of such courses . indeed ,
a general demonstration without assuming planar symmetry is clearly beyond the scope of the introductory course , and rightfully belongs to more advanced courses .
however , the physics and mathematics needed for the demonstration for plane electromagnetic waves are readily available for students of the introductory course , and in fact are already included in most textbooks we surveyed .
therefore , no new physics or mathematics are needed to demonstrate the transversality of ( plane ) electromagnetic waves in the introductory course .
all that is needed is to arrange the arguments that are already in the student s toolkit to achieve deeper level of understanding .
the textbooks we have surveyed ( and again , we emphasize our literature search was not exhaustive , as the number of textbooks is indeed very large ; we do believe we have found the tendencies of existing textbooks ) fall under the following categories .
a : textbooks that postulate the transversality ( but do nt derive it ) , and then show consistency with the ( either all four or just the dynamical ) maxwell equations @xcite ; category b : textbooks that only use the two dynamical equations ( faraday s law and the ampre
maxwell law ) to derive the wave equations for the ( transverse ) components of the @xmath0 and @xmath1 fields , but do nt show that _ only _ transverse fields are allowed @xcite ; and category c : textbooks that add to the treatment of category b also a demonstration that a longitudinal component of the @xmath0 field is disallowed ( but not that a transverse component is consistent with the maxwell equations ) @xcite . interestingly ,
even introductory books that discuss the differential maxwell equations do not include a demonstration of the transversality of electromagnetic waves @xcite ( we have not surveyed more advanced books , appropriate for upper division courses on electromagnetism ) .
none of the textbooks we have surveyed include physical arguments for the transversality of electromagnetic waves , even though all of them discuss polarization . on the other hand ,
none of the textbooks we have surveyed include a discussion of the impossibility of genuine spherical electromagnetic waves ( either monopole of a combination of multipoles ) @xcite .
either polarization or hypothetical spherical electromagnetic waves may be used in physical arguments for the transversality of electromagnetic waves .
this omission is surprising , as building strong physical intuition is widely considered a primary goal of such courses .
many of the properties of electromagnetic waves can be demonstrated also for the algebra based introductory course @xcite , but examination of these arguments and their use in textbooks are beyond the scope of this paper .
our theoretical argument is by no means original , and surely has been made numerous times .
it is however absent in its entirety from all textbooks we have surveyed .
even if it does appear in textbooks we have not surveyed , we believe it is fair to state that most textbooks do not include the full argument , and that similarly most calculus based introductory physics courses do not discuss it . in advanced courses
the transversality of electromagnetic waves is easily shown @xcite from the maxwell equations and the wave equation for the vector potential @xmath5 in the lorenz gauge , and the skew symmetry of the maxwell faraday tensor @xmath6 .
consider , say , a vector potential @xmath7 .
the only nonvanishing independent components of @xmath8 are @xmath9 and @xmath10 , and @xmath11 .
these @xmath0 and @xmath1 fields manifestly satisfy the aforementioned properties , although the formality of the demonstration in addition to being inappropriate for the introductory course is devoid of physical insight .
( one may of course try to impose longitudinality : take @xmath12 .
one then finds that @xmath13 is non - zero , and one might be tempted to identify it with @xmath14 .
however , the postulated vector potential is not a solution for the maxwell equations in the lorenz gauge , as can be readily verified . )
one can also show the transversality without invoking tensor analysis , using only vector analysis methods @xcite .
the demonstration using elementary methods is important , as it involves more physical insight than the advanced , more formal proof .
notably , electromagnetic waves in matter do not have to be transverse @xcite . longitudinal electromagnetic waves ( or combinations of longitudinal and transverse waves )
indeed exist in inhomogeneous media and in cavities ( because of the cavity s boundary conditions ) @xcite .
one may also have in vacuum electromagnetic waves with @xmath15 , but those are standing waves ( with vanishing poynting vector ) @xcite .
these interesting cases are typically beyond the scope of the introductory course . in this paper
we present a full argument for the transversality of ( plane ) electromagnetic waves in vacuum .
we are hopeful that instructors of calculus based introductory physics courses present a full argument in their classes .
we use si units despite their awkwardness ( for the description of electrodynamics ) , as this appears to be the nearly universal practice in textbooks .
we use only the integral form of the maxwell equations , as nearly all introductory textbooks ( for _ calculus_based physics ) refrain from introducing differential operators . in section [ phys_arg ]
we describe physical arguments for the transversality of electromagnetic waves , based on polarization and on ( lack of ) monopole radiation . in section [ theor_arg ]
we present a full argument at a level appropriate for the introductory calculus based physics course , as follows : first , we show that @xmath0 can not have a longitudinal component , and that @xmath0 may have a transversal component .
we then repeat the arguments for the @xmath1 field .
we then show that the @xmath0 and @xmath1 fields are orthogonal to each other and to the direction of propagation , and finally , we show that the @xmath0 and @xmath1 fields satisfy the wave equation .
we emphasize that the only assumption we make is that of planar symmetry . except for this assumption ( which makes the derivation possible for the introductory course )
our argument is general .
many introductory courses include a demonstration of polarization that includes two polaroid films , so that one can be rotated relative to the other .
when the polarization directions of the two polaroids are orthogonal , no light is going through the polaroids , and the image on the screen is dark .
assume there were a longitudinal component to the @xmath0 field .
this component may be transmitted through either polaroid , so that a longitudinal component would arrive to the screen through both polaroids even when they are orthogonal .
therefore , with a longitudinal component the screen will never get totally dark when the two polaroids are in a plane perpendicular to the direction of propagation of the light .
a variant of this argument was given by schutz @xcite .
schutz considers the two polarizers rotating relative to each other in the plane orthogonal to the direction of propagation . as
the brightness on the screen oscillates with the polarizers , one must conclude the electric field acts across its motion , i.e. , transversally . in order to make our argument compelling ,
students need to have at least a qualitative understanding of how simple dichroic devices such as a wire
grid polaroid work ( most introductory texts do not discuss in depth other polarizers , such as beam splitting polarizers , birefringent polarizers , or polarization by reflection ) . such understanding may be expected from students of the relevant course : there are three microscopic mechanism that filter out the component of the @xmath0 field in the plane of the polaroid parallel to the direction of the wires ( or polyvinyl alcohol polymers doped with an iodine solution in a commercial polaroid ) : first , by the lorentz force law , conduction electrons are accelerated along the wires , and oscillate with the oscillating @xmath0 field .
when colliding with lattice atoms , kinetic energy is transferred to the lattice , thereby transferring energy from the parallel component of the @xmath0 field to the lattice ( joule heating " ) .
second , the oscillating conduction electrons reradiate in all directions , so that half the reradiation is backward ( reflection ) and the remainder forward .
the backward reradiation takes away more energy from the parallel component of the @xmath0 field .
the forward reradiation is not entirely in the direction of the original incident plane wave , so that it scatters and little of it arrives at the detector .
most importantly , the forward reradiation is generally out of phase with the incident @xmath0 field ( phase difference of @xmath16 radians ) , thereby reducing it further by destructive interference @xcite .
some textbooks explain the polarization effect by stretching a mechanical model too far . in such texts the electric field
is modeled by a tension wave in a string that is passing through a fence with parallel beams @xcite .
the mechanical wave naturally is transmitted only if the string oscillates parallel to the beams .
this argument might imply the wrong component of the @xmath0 field is transmitted .
for this reason , those texts refer to the polarization direction of the polarizer , stating that perpendicular components are filtered out . but this argument may then persuade some students to believe a longitudinal component of the @xmath0 field is also filtered out , such that this sort of argumentation , in addition to not empowering the students with an understanding of simple dichroic devices work , also might prevent them from understanding a simple physical argument for why electromagnetic waves are transverse , and give the students a misleading physical picture of electromagnetic fields oscillating in the space between the wires . consider a hypothetical monopole electromagnetic wave .
then , a radially pulsating spherical charge distribution would emit such spherical outgoing waves .
as the fundamental equations of physics are symmetrical under time reversal ( @xmath17 ) , an incoming imploding spherical monopole wave would make an otherwise static spherical charge distribution pulsate radially .
each charge pulsates radially because of a radial force acting on it .
the form of the lorentz force law then implies there were a radial @xmath0 field acting on the charges .
but a radial @xmath0 field is perpendicular to the imploding spherical wave front and therefore in the same direction of its motion , so that the @xmath0 field would have a longitudinal component .
this qualitative argument implies that a monopole electromagnetic wave is necessarily longitudinal .
( while students of the introductory course generally have not studied multipole expansions of the electromagnetic field , the argument may still be used in reference to a truly spherical wave front . )
we do know , however , that outside a spherical charge distribution the electromagnetic field is static , as the monopole piece of the electromagnetic field is non - radiative .
students are exposed to this argument also in newtonian gravity : because of the inverse square force law ( common to both newton s and coulomb s laws ) the field strength of any spherical charge ( or mass ) distribution is the same as if all the charge ( or mass ) were concentrated at the center .
specifically , the electric field outside a radially pulsating spherical charge distribution is static .
therefore , the longitudinal component of the @xmath0 field can not exist , as such a longitudinal component would have to be radiated by a pulsating spherical charge distribution @xcite .
this physical argument may be cast in a more mathematical form , using a theorem from vector calculus that states that no unit two dimensional continuous vector field @xmath18 on a sphere may exist on the entire sphere ( @xmath19 , @xmath20 ) @xcite .
this theorem is normally beyond the scope of the introductory course .
in addition , it requires the assumption of transversality ( namely , that @xmath18 is a two - dimensional vector field on the sphere , or @xmath21 ) .
however , it may be used to argue that any such spherical electromagnetic wave would have to be longitudinal .
consider a plane electromagnetic wave traveling in vacuum in the direction of @xmath22 .
( our coordinates may be rotated to this orientation . ) because of the planar symmetry the most general electric field strength is @xmath23 . in what follows
we omit the explicit time dependence of fields .
we first show there can be no component @xmath24 : construct a gaussian surface enclosing a volume @xmath25 as in fig .
[ gauss1 ] . according to gauss s law , @xmath26 where @xmath27 is the total charge inside the gaussian surface , and @xmath28 is the permittivity of vacuum . here
, @xmath29 is a surface element normal to the ( orientable ) surface @xmath30 , defined conventionally such that it is positive when pointing outward .
calculate the integral on the lhs : take @xmath31
. then @xmath32\,dy\,dz=0\ ] ] as there are no charges .
the components @xmath33 and @xmath14 identically do not contribute to the lhs integral , as these components are not functions of @xmath34 or @xmath35 , respectively , and as they are in the plane of the faces of the surface whose normals are in the direction of @xmath22 .
notice that there is no flux of @xmath24 through the faces of the gaussian surfaces in the @xmath36 or @xmath37 directions because of the orthogonality of @xmath24 to the surface s normals .
this immediately implies @xmath38 , or @xmath24 does not obey a non - trivial wave equation .
therefore , the most general electric field is @xmath39 . without loss of generality
, we may rotate our coordinate system so that @xmath40 .
our demonstration depends on the lack of charges .
presence of charges requires a non - vanishing gradient of @xmath24 .
indeed , gauss s law would not be violated for longitudinal waves without assuming vacuum , as there are charges and inhomogeneities in matter .
therefore , this demonstration does not rule out longitudinal waves in matter or inhomogeneous media .
components of either @xmath0 or @xmath1 have to be independent of @xmath41 .
the arrows indicate a case in which they are not .
then , the surface integral must be nonzero , and proportional to their @xmath41 derivatives . in the case of @xmath0
this implies free charges inside the box , i.e. , a continuous charge distribution , which contradicts the vacuum assumption . in the case of @xmath1
that distribution would be of magnetic monopoles.,width=326 ] consider next an electric field @xmath40 , as in fig .
[ gauss2 ] . applying gauss s law
does not yield a restriction on @xmath33 , because the lhs of the integral vanishes identically , as @xmath42 : @xmath43\,dx\,dz=0\ , .\ ] ] this step may of course be combined with the preceding one by taking @xmath31 , and showing that gauss s law becomes @xmath38 . in conclusion ,
a transverse component for @xmath0 is consistent with gauss s law .
of course , were there not a transverse component , the solution would become trivial ( no electromagnetic wave ) .
therefore , for any electromagnetic wave a transverse @xmath0 field is necessary . or @xmath1 leads to no contradiction : as the fields are functions of only the spatial coordinate @xmath41 ( in addition to the time @xmath44 ) , the surface integral of the fields vanish along all faces , either because of the field being parallel to the face ( faces parallel to the @xmath45 and @xmath46 planes ) or because the field exiting the surface completely balances the field entering ( faces parallel to the @xmath47 plane ) .
it is therefore important that the fields are only functions of @xmath41 , otherwise the cancellation would not be exact . ,
width=326 ] we next use similarly the magnetic gauss law , @xmath48 the same argument as in step 1 implies that @xmath49 , so that also @xmath1 does not have a longitudinal component .
therefore , @xmath50 .
but we have no more freedom to rotate the coordinate system , a freedom which we have already exhausted in rotating it so that @xmath51 . we may therefore not set any of the components of @xmath1 equal to zero . as there are no magnetic monopoles , the @xmath1 field has to be transverse also in matter .
consider next a field @xmath50 .
applying gauss s law does not yield a restriction on @xmath52 or on @xmath53 , because the lhs of the integral vanishes identically , as @xmath54 and @xmath55 : @xmath56\,dx\,dz\\ & + & [ b_z(x , z+\,dz)-b_z(x , z)]\,dx\,dy=0\ , .\end{aligned}\ ] ] consider now an ampre loop enclosing an area @xmath57 as in fig .
[ ampere1 ] . the ampre
maxwell law is @xmath58 where @xmath59 is the permeability of free space , @xmath60 is the total conduction current through the loop , and @xmath61 is the flux of the @xmath0 field through it . here , @xmath62 is an element of the curve @xmath63 along the loop , conventionally taken to be positive counterclockwise . doing the integral on the lhs , @xmath64\,dy=\frac{\,\partial b_y}{\,\partial x}\,dx\,dy=0\ , , \ ] ] as the @xmath53 component is perpendicular to the entire loop , and as there are no electric field flux through the loop ( because the @xmath0 field in in the direction of @xmath36 ) and also no free currents .
therefore , @xmath65 .
therefore , we find that @xmath66 .
we therefore find that @xmath67 , i.e. , the fields are orthogonal to each other and also to the direction of propagation @xmath22 .
in addition , @xmath68 , i.e. , the cross product of the fields is in the direction of propagation of the wave .
component of the @xmath1 field is perpendicular to the entire loop , so that it does not contribute to the line integral around it . when applied to the ampre
maxwell law , an electric field in the direction of @xmath36 has zero flux through the loop .
the line integral around the loop is proportional to the @xmath41 derivative of the @xmath36 component of @xmath1 .
when applied to the faraday law , the @xmath41 derivative of the @xmath36 component of @xmath0 is proportional to the time derivative of the @xmath37 component of @xmath1 .
the arrow indicates a @xmath36 component to either @xmath0 or @xmath1.,width=326 ] this part of the argument appears in most textbooks we have surveyed @xcite .
however , without the preceding parts of the argument it is merely a demonstration of consistency of the transverse wave equation with the maxwell equations , not their unavoidability .
we now use the loop in fig .
[ ampere2 ] to evaluate the ampre
maxwell law . on the lhs , @xmath70\,dz=\frac{\,\partial b_z}{\,\partial x}\,dx\,dz\ , .\ ] ] the flux of the @xmath0 field is simply @xmath71 , so that the rhs is evaluated to equal @xmath72 so that the ampre maxwell law yields @xmath73 component of the @xmath1 field is in the plane of the loop .
the line integral around the loop is proportional to the @xmath41 derivative of the @xmath37 component of @xmath1 , which by faraday s law is proportional to the time derivative f the electric field s flux through it .
the arrow indicates the @xmath37 component of the @xmath1 field . the @xmath0 field ( which is in the direction of @xmath36 ) is not shown.,width=326 ]
we next consider again the ampre loop in fig .
[ ampere1 ] , but this time apply it to the faraday law , @xmath74 here , @xmath75 is the flux of the @xmath1 field through the loop .
the lhs is evaluated as @xmath76\,dy=\frac{\,\partial e_y}{\,\partial x}\ , dx\,dy\ , .\ ] ] the rhs is @xmath77 so that @xmath78 many introductory texts @xcite show that by differentiating eqs .
( [ eq1 ] ) and ( [ eq2 ] ) and eliminating mixed second derivative terms , one can obtain wave equations . with
the identification @xmath79 these equations lead to @xmath80 and @xmath81 to show the relation between the magnitudes of the @xmath0 and @xmath1 fields we present an adaptation of an argument originally made by feynman @xcite .
consider an infinite planar current sheet , carrying current density @xmath82 . a current sheet is a two dimensional surface carrying current , and can be approximated by a large number of wires , all carrying current in the same direction , so that the total current per unit length ( perpendicularly to the wires ) is @xmath83 , where @xmath60 is the total current included in the length @xmath84 .
current sheets are a very useful model in magnetohydrodynamics ( mhd ) and in heliophysics .
we first recall the @xmath1 field of a stationary current sheet @xcite .
consider a current sheet aligned as in fig .
[ current_sheet ] . to find the @xmath1 field outside the current sheet we construct an ampre loop
whose plane is perpendicular to the plane of the current sheet
. the @xmath1 field must be parallel to the plane of the current sheet , and perpendicular to the direction of the current .
( one may be convinced of that result by considering the elementary problem of current wire , and considering the current sheet as the limiting case of infinitely many such current wires . )
as this problem is stationary , we may use the original form of the ampre law ( no time changing flux of an electric field ) , i.e. , @xmath85 where @xmath60 is the total current going through the loop , whose length in the @xmath35direction is @xmath86 . because the current sheet is infinite , the magnitude of the field strength @xmath87 is constant along the legs of the loop in the @xmath35direction .
the lhs of the ampre law then is @xmath88 , and the rhs is simply @xmath89 .
putting the two expressions together , one finds that @xmath90 .
notably , @xmath91 is independent of the distance from the current sheet .
plane , carrying current density @xmath92 .
the direction of the @xmath1 field is as indicated .
the ampre loop is in the @xmath47 plane .
, width=326 ] consider next a current sheet that is abruptly turned on at time @xmath93 .
that is , @xmath94 , where @xmath95 is the heaviside step function .
before the current is turned on , the @xmath1 field is zero everywhere . following the abrupt turning on of the current , a non zero @xmath1 field is starting to fill up space , the wavefront propagating at the speed @xmath69 .
therefore , we have a propagating wavefront , so that in front of the wavefront the @xmath1 field vanishes , and behind it is uniform as shown above . to find the accompanying @xmath0 field we use the faraday law for the ampre loop in fig .
[ wave_front ] .
only the part of the ampre loop behind the wavefront has non - vanishing @xmath1field flux . as the wavefront is propagating with speed @xmath69 , the area of the loop where there is a @xmath1 field is increasing , so that there is a time
changing @xmath1field flux through the loop .
the @xmath0 field between the wavefront and the current sheet is parallel to the current sheet , and counter - parallel to the current .
that is , @xmath96 .
( see fig .
[ wave_front ] . ) using the faraday law , @xmath97 one finds that the rhs equals @xmath98 , and the flux of the @xmath1 field through the loop at the time @xmath99 is @xmath100 , so that the rhs of the faraday law equals @xmath101 .
setting the two hand sides equal to zero , we find @xmath102 , or @xmath103 although our argument was made only for the configuration of a current sheet that is abruptly turned on , the result that @xmath104 for an electromagnetic wave is general .
the general result , however , is beyond the scope of the introductory course .
an important consequence is that in an electromagnetic wave the @xmath0 and @xmath1 fields are in phase .
that is , as at any given time @xmath104 , when @xmath105 is maximal so is @xmath91 , when @xmath105 vanishes @xmath91 is zero too , etc .
the result that @xmath104 appears in many introductory texts ( e.g. , in @xcite ) .
however , in such books an extra assumption is typically made , namely that the fields are harmonic and are in phase .
students of the introductory course , who normally are not familiar with fourier theory , often find it unconvincing to base an important general physical result on the mathematical properties of sinusoidal waves , in addition to having to memorize yet another factoid , namely that the fields are in phase . at this point one
is in a position to make the following discussion .
we found that @xmath104 .
this means that the ratio @xmath106 for an electromagnetic wave equals @xmath69 , which depends in magnitude on the system of units .
e.g. , in si units this ratio is @xmath107 , and in cgs units is it @xmath108 .
it is therefore natural to use units that put the @xmath0 and @xmath1 fields on equal footing , that is use units in which distance is measured in light - seconds . in such units the speed of light is @xmath109 ( light-)second / second , or simply @xmath109 .
that is , in these units @xmath110 .
this choice of units becomes further motivated physically when one considers the energy density @xmath111 in the electromagnetic field ( not necessarily that of an electromagnetic wave ) , @xmath112 .
for an electromagnetic wave @xmath104 , so that the energy densities stored in the @xmath0 and @xmath1 fields are equal .
it is therefore suggestive to make the two fields symmetrical by using units in which @xmath109 .
it is often instructive to remind students that si or cgs units were developed because of their convenience in describing everyday phenomena involving humans , and while all unit systems are in principle equivalent , these are not necessarily the units in which physical phenomena take their simplest or most natural form .
plane , carrying current density @xmath92 as in fig .
[ current_sheet ] .
the current sheet is turned on abruptly , so that plane waves propagate with speed @xmath69 to the right of the current sheet in the @xmath22 direction , and to its left in the @xmath113 direction .
the wavefront ( wf ) is shown only to the right of the current sheet . at time
@xmath114 the wave front touched the dashed segment in the @xmath115 direction of the ampre loop .
the space between the wave front and the current sheet has uniform @xmath1 field , and ahead of the wavefront the @xmath1 ( and @xmath0 ) field vanishes .
the only part of the loop with non - vanishing @xmath1 field is behind the wave front , shown shaded.,width=326 ] we therefore showed that electromagnetic waves are transverse , and that the electric and magnetic fields are perpendicular to each other .
also the directions are so that @xmath116 is in the direction of propagation .
we have also shown they have the same speed @xmath69 , that they are in phase , and that the magnitudes of the @xmath0 and @xmath1 is related by @xmath104 .
this demonstration makes use of only the integral maxwell equations , and is appropriate for the level of most calculus
based physics courses , and includes only arguments already available for the perspective students .
we believe its main strength is that it avoids giving the students factoids without deep understanding , and instead empowers the student to gain deeper insight .
the author wishes to thank richard price for discussions .
this work has been supported by nasa / gsfc grant no .
ncc5580 and by nasa / ssc grant no .
nnx07al52a , and by nsf grant no .
phy0757344 .
physics for scientists and engineers with modern physics , a strategic approach _ ( pearson addison wesley , 2004 ) ; h.d .
young and r.a .
freedman , _ sears and zemansky s university physics _
, 12@xmath117 ed .
( pearson addison wesley , 2007 ) .
r. wolfson and j.m .
pasachoff , _ physics with modern physics for scientists and engineers _ , 3@xmath118 ed .
( addison wesley , 1999 ) ; p.a .
tipler and g.p .
mosca , _ physics for scientists and engineers _ , 5@xmath117 ed .
freeman , 2004 ) ; s.m .
lea and j.r .
burke , _ physics : the nature of things _ ( brooks / cole , 1997 ) ; p.m. fishbane , s. gasiorowicz , and s.t .
thornton , _ physics for scientists and engineers _ , 3@xmath118 ed .
( addison wesley , 2005 ) ; r.l .
reese , _ university physics _ ( brroks / cole , 2000 ) ; r.a .
serway and r.j .
beichner , _ physics for scientists and engineers _ , 5@xmath117 ed .
( saunders , 2000 ) .
purcell , _ electricity and magnetism : berkeley physics course _ , volume two , 2@xmath119 ed .
( mcgraw hill , 1985 ) .
feynman , r.b .
leighton , and m.l .
sands _ the feynman lectures on physics _ ( addison wesley , 1989 ) . | introductory calculus based physics textbooks state that electromagnetic waves are transverse and list many of their properties , but most such textbooks do not bring forth arguments why this is so .
both physical and theoretical arguments are at a level appropriate for students of courses based on such books , and could be readily used by instructors of such courses . here ,
we discuss two physical arguments ( based on polarization experiments and on lack of monopole electromagnetic radiation ) , and the full argument for the transversality of ( plane ) electromagnetic waves based on the integral maxwell equations .
we also show , at a level appropriate for the introductory course , why the electric and magnetic fields in a wave are in phase and the relation of their magnitudes . |
low surface brightness galaxies ( lsbgs ) , defined as those with @xmath1-band central surface brightnesses , @xmath2 , fainter than 23.0 mag arcsec@xmath3 , appear to have followed a very different evolutionary history from high surface brightness galaxies ( e.g. , mcgaugh 1992 ; knezek 1993 ) .
their stellar populations , stellar masses , current star formation rates , and other properties appear to differ significantly from their high surface brightness counterparts , and little is known about their corresponding dust properties
. a better understanding of these differences is needed to understand how they evolved to their present state .
it has been generally assumed , even though a population of red lsbgs has been discovered ( oneil et al .
1997a ) , that they contain little dust , and dust reddening has not been considered an important effect ( e.g. , bothun et al .
1997 ; bell et al .
lsbgs have low star formation rates , with suspected modest bursts in the range 10@xmath4 - 10@xmath3m@xmath5yr@xmath6 ( e.g. , vallenari et al .
2005 ) and low metallicities , with almost all lsbgs at or less than about one - third solar ( mcgaugh 1994 ; ronnback & bergvall 1995 ) .
their low metallicities imply that the dust to gas ratios should be systematically lower than in their high surface brightness galaxy counterparts , and the fact that the _ infrared astronomical satellite _ ( _ iras _ ; neugebauer et al . 1984 ) and the _ infrared space observatory _ ( _ iso _ ; kessler et al .
1996 ) were only able to detect two of these galaxies seems to indicate that dust is much less important in lsbgs .
furthermore , any data acquired with these observatories would not have been adequate to characterize dust temperatures or spatial distributions .
that dust plays a relatively minor role in the evolution of lsbgs is further reinforced by observations of highly transparent galaxies that appear to have multiple distant galaxies seen through their disks ( oneil et al .
1997b ; p.
knezek , private communication ) .
this has been confirmed by a more detailed analysis of the dust opacity of lsbg disks in comparison to those of high surface brightness galaxies , where lsbgs appear effectively transparent ( holwerda et al .
2005 ) .
additionally , pickering & van der hulst ( 1999 ) attempted to detect dust in lsbgs using submillimeter observations from the jcmt with scuba .
ten galaxies were observed , two of which were detected at 850@xmath0 with only one detected at 450@xmath0 .
they combined their submillimeter data with existing _
iras _ data , finding dust temperatures in the range 15 - 20k .
none of their very lsb ( @xmath7 ) galaxies were detected , and they concluded that the lack of detection in the lowest surface brightness galaxies was consistent with previous lines of evidence that only modest amounts of dust could exist . _ spitzer _ opens the opportunity to study any dust that may exist at low levels in lsbgs .
the increased sensitivity relative to previous observations gives a higher likelihood of the detection of diffuse dust emission that echoes the lsbgs diffuse optical appearance .
in addition , the resolution of the _ spitzer _ imaging instruments allows for analysis of dust temperature , mass , and spatial distribution in lsbgs not possible before , making it feasible to address crucial issues regarding chemical evolution and dust production .
we present here the infrared properties of a small sample of lsbgs with the full suite of _ spitzer _ instruments .
all _ spitzer _ observations presented here are part of the dust in low surface brightness galaxies guaranteed time observation program ( p.i.d .
62 ) whose principal investigator is m. j. rieke .
the galaxies in this sample were selected to be some of the brightest and closest known lsbgs , many taken from the work of pickering ( 1998 ) .
two galaxies are high surface brightness , low metallicity isolated late - type spirals that are included for the purpose of comparison with the lsbgs .
table 1 summarizes general information for each galaxy , with the lsbgs and high surface brightness galaxies separated by a line .
the objects were chosen such that the infrared background cirrus was low , improving the chances of detection of the faintest emission associated with each galaxy , and with sufficiently large angular diameters for resolution with the _ spitzer _ instruments . here
we briefly outline their main properties and unique characteristics . _
malin1 _ is one of the best known lsbgs ( bothun et al .
1987 ; see barth 2007 for a more recent view ) and the largest gas - rich galaxy found to date .
its optical disk is six times bigger than that of the milky way .
its spectrum is dominated by its old , metal - rich stellar population , with a smaller contribution from hot , young stars ( impey & bothun 1989 ) .
this lsbg is the most distant object in our sample . _
ugc5675 _ : this magellanic dwarf spiral ( e.g. , schneider et al .
1990 ; mcgaugh et al .
1995 ) has a very low surface brightness disk ( schombert & bothun 1988 ) and has the faintest @xmath8 of our sample .
it does not have any regions of distinct star formation or an extended hi envelope ( van zee et al .
1995 ) . _
ugc6151 _ is another magellanic spiral with a small core of optical emission surrounded by faint diffuse emission with little structure .
mcgaugh et al .
( 1995 ) note that the galaxy contains `` quite a few faint hii regions '' . _
ugc6614 _ has a star - forming distinctive ring and has a particularly extended disk that can be traced to at least 130@xmath9 ( van der hulst et al .
it is the most metal - rich lsbg known and is considered to be a `` cousin '' to malin 1 due to their similar properties .
the hi data show that there is a central depression in the gas distribution , which has led to discordant flux measurements in the literature ( bothun et al . 1985 ; giovanelli & haynes 1989 ; pickering et al .
1997 ; matthews et al . 2001 ) , although the most recent of these works agree . _
ugc9024 _ : this galaxy has a very low surface brightness disk coupled with a normal bulge .
it has been suggested that the large disk plus bulge indicates that it is a transition object between average sized lsbgs with no bulge and giant malin 1 type galaxies ( mcgaugh et al .
1995 ) . _
ugc6879 _ : this object , while in the impey et al .
( 1996 ) catalogue of low surface brightness galaxies , does not qualify as an lsbg , due to its bright central surface brightness .
( it is possible that it was either a candidate lsbg , later observed and found to be too bright , or that it qualified as a type of lsb disk using a `` diffuseness '' index involving the disk scale length ; see sprayberry et al .
1995 for details ) .
this galaxy is one of the few in the program to have been detected by _
iras _ , with photometry measurements at 60 and 100@xmath0 and upper limits at 12 and 25@xmath0 .
it also has a detection at 850@xmath0 by scuba ( pickering & van der hulst 1999 ) , making it an excellent candidate for further study in the mid- and far - ir at the higher sensitivity and resolution of _
_ ugc10445 _ is a nearby starbursting dwarf spiral .
the _ spitzer _
images of this object have been examined in detail in hinz et al .
it also has available _ iras _ and _ iso _ measurements .
each galaxy was observed with the infrared array camera ( irac ; fazio et al .
2004 ) at all four wavelengths ( 3.6 , 4.5 , 5.8 , and 8.0@xmath0 ) , in one 5@xmath10 field with a frame time of 30s and with a dither of five positions gaussian , for a total integration time of 150s per object per wavelength .
irac data were reduced at the _ spitzer _ science center ( ssc ) with the s14.0.0 pipeline .
the irac spatial resolution is @xmath112@xmath9 in all bands .
the multiband imaging photometer for _ spitzer _ ( mips ; rieke et al .
2004 ) data for the sample were observed in photometry mode at 24 , 70 , and 160@xmath0 .
the integration times on all galaxies were 690s at 24@xmath0 , 252s at 70@xmath0 , and 42s at 160@xmath0 .
the mips data were reduced using the data analysis tool version 3.00 ( dat ; gordon et al . 2005 ) , the same techniques that are used to calibrate the instrument itself .
two independent reductions were carried out to test for systematic errors on these faint sources ; the results agreed closely for both reductions .
the mips spatial resolutions are 6@xmath9 , 18@xmath9 , and 40@xmath9 at 24 , 70 , and 160@xmath0 , respectively .
dates for the irac and mips observations are given in table 1 .
the mips 24@xmath0 observations for ugc5675 appear to be corrupted .
the observations performed by _ spitzer _ just prior to this object were deep , high - redshift observations and unlikely to introduce saturated objects that might have effected our data set .
however , the southwestern corner of almost all the 24@xmath0 data collection events have highly saturated pixels marked as nan .
further processing and mosaicing of the image was impossible , and no information could be recovered .
infrared spectrograph ( irs ; houck et al .
2004 ) staring - mode observations were scheduled for only three of the targets in this program : malin1 , ugc6879 , and ugc10445 .
galaxies were observed in the low resolution 5.2 - 8.7@xmath0 ( sl2 ) and 7.4 - 14.5@xmath0 ( sl1 ) modes and in the low resolution 14.0 - 21.3@xmath0 ( ll2 ) and 19.5 - 38@xmath0 channels ( ll1 ) .
details of the observations , including dates , integration times and slit position angles , are given in table 2 .
the faint , diffuse , and complex nature of these sources makes such observations difficult .
_ spitzer s _ onboard peak - up algorithm centroided on a source considered far from the intended target position for ugc10445 and centroided on a bad pixel rather than malin 1 .
data were reduced with version s13.2.0 of the ssc data pipeline , with extractions using the pipeline developed by the formation and evolution of planetary systems ( feps ) legacy team ( hines et al .
additional data are available for ugc6879 .
these include the far and near - ultraviolet ( 1350 - 1750 , @xmath12 , and 1750 - 2750 , @xmath13 , respectively ) images from _ galaxy evolution explorer _
( _ galex _ ; martin et al .
2005 ) made available through the nasa extragalactic database ( ned ) .
figure 1 shows an image of each galaxy in the sample at all the _ spitzer _ wavelengths , with the exception of ugc10445 , which was presented by hinz et al .
( 2006 ) .
irac detects all galaxies in the sample at 3.6 and 4.5@xmath0 and all except for ugc5675 at 5.8 and 8.0@xmath0 . in general , the lsbgs are easily detected at the two shorter wavelength bands , representing the old stellar population , and difficult to image at 5.8 and 8.0@xmath0 , with many having only point - like detections at the longer wavelengths .
the mips observations of the lsbgs are varied .
there are clear detections of all lsbgs at 24@xmath0 ( with the exception of the corrupted data of ugc5675 , as explained above ) , three detections at 70@xmath0 , and two at 160@xmath0 . of these detections ,
two have extended emission at 24@xmath0 while two have point - like morphologies . at 70@xmath0 ,
two are extended , with one point - like , and at 160@xmath0 two are extended , with no point - like detections .
a summary of this rough classification is given in table 3 , with lsbgs listed above the solid line and hsbgs listed below the solid line .
the difficulty of detecting emission at the longer wavelengths for the lsbgs is not simply a result of the decreased resolution .
figure 2 shows the three mips images of ugc6614 , with both the 24 and 70@xmath0 images convolved with a kernel that transforms them to the resolution of the 160@xmath0 data .
the kernel was created using a fourier technique on the mips psfs generated by stinytim ( gordon et al . in preparation ) .
the galaxy seems to be more extended at 70@xmath0 than at 24@xmath0 , consistent with the star forming ring becoming more prominent as shown in figure 1 .
the signal to noise is too low at 160@xmath0 to confirm this trend , but the image does show that the peak remains on the galaxy center , as defined at 24 and 70@xmath0 . the changing brightness and morphology of the galaxy
does not seem to be related to the resolution differences .
aperture photometry was conducted on all _ spitzer _ images .
additional image processing in the form of background subtraction was first completed by subtracting a constant value from each image .
the value of this constant was determined by masking stars in the foreground , then taking an average of all pixel values outside the aperture used for the galaxy photometry . for galaxies where the foreground contamination was particularly high , we used a large region surrounding the galaxy to determine a background value that included a sampling of foreground stars .
table 3 shows the mips flux density values and their associated errors , along with the radii of apertures used .
table 4 shows the corresponding irac photometry values .
galaxies that are undetected at the various wavelengths have 3@xmath14 upper limits listed in tables 3 and 4 .
upper limits were calculated from the images themselves , using the mean value of the sky counts and adding three times the value of the standard deviation of the sky counts .
the photometric uncertainties are estimated to be 10% at 3.6 and 4.5@xmath0 and 15% at 5.8 and 8.0@xmath0 .
these values include a 3% absolute calibration uncertainty ( reach et al . 2005 ) , a contribution for scattered light in an extended source ( w. reach , private communication ) , and an uncertainty due to the sensitivity of the measurements to the background subtraction . the contribution of the scattered light is higher at 5.8 and 8.0@xmath0 .
we do not perform aperture corrections on the irac photometry , which in certain limiting cases can be up to 25 - 30% for the 5.8 and 8.0@xmath0 bands . in our case
, it is difficult to determine this correction for the mixture of point and extended sources seen in the irac images .
we mainly use the 8.0@xmath0 images to establish the presence of aromatic feature emission in our galaxies , so uncertainties of this magnitude have no effect on our conclusions . the mips flux calibration uncertainties are 4% at 24@xmath0 , 7% at 70@xmath0 , and 12% at 160@xmath0 ( engelbracht et al . 2007 ;
gordon et al .
2007 ; stansberry et al . 2007 ) .
photometric uncertainties bring these values to total errors of 10% at 24@xmath0 , 20% at 70@xmath0 , and 20% at 160@xmath0 .
ugc6879 , with its bright detections at all ir wavelengths , can be analyzed in detail based on the mid - infrared and submillimeter photometry .
figure 3 shows the spectral energy distribution ( sed ) for this galaxy , including _
galex _ , 2mass , irac , _ iras _ , _ spitzer _ , and scuba data points .
the emission by dust at the longer wavelengths can be modeled by an equation of the form @xmath15 where @xmath16 ( @xmath1732mpc ) , @xmath18 is the mass absorption coefficient , @xmath19 is the planck function , @xmath20 is the dust mass , and the sum extends over the number of dust components .
we adopt a two - component dust model consisting of warm and cool silicate grains ( @xmath21 ) .
further details regarding model assumptions and the fitting process can be found in hinz et al .
the data set is best fitted by a model consisting of a warm silicate component at t@xmath22k and a cool silicate component at @xmath23k , shown in figure 3 , where the quoted error bars are 1 @xmath14 .
we estimate the dust masses of ugc6879 to be @xmath24m@xmath5 for the warm component and @xmath25m@xmath5 for the cool dust , where the quoted error bars are again 1 @xmath14 . as shown in hinz et al .
( 2006 ) , choosing carbonaceous grains in place of silicate grains only modestly affects these values .
figure 4 shows the full irs spectra for ugc6879 , malin1 , ugc10445 .
we identify emission lines clearly detected in ugc6879 and ugc10445 : [ siv ] , [ neii ] , and [ siii ] ( see , e.g. , smith et al . 2004 ) .
additionally , we see the broad emission features usually attributed to polycyclic aromatic hydrocarbons ( pahs ) .
the data show the four main aromatic bands at 6.2 , 7.7 , 8.6 , and 11.3@xmath0 .
the aromatic feature at 12.7@xmath0 is likely to be contaminated by the [ ne ii ] 12.8@xmath0 line .
there are no spectral features detected in the malin 1 spectrum . despite the fact that _
spitzer _ was not aligned on the coordinates given as the central nucleus of the galaxy , the large extent of malin 1 ensures that irs took data on some portion of the disk , and the exposure time is long .
also , our photometry ( tables 3 and 4 ) shows the mid - ir excess to be very weak .
it is likely that no aromatic features in malin 1 are detectable with irs in reasonable exposure times .
the spectra for ugc6879 and ugc10445 were fitted with the publicly available idl tool pahfit , which was developed to decompose irs spectra of pah emission sources , with a special emphasis on the careful recovery of ambiguous silicate absorption , and weak , blended dust emission features ( smith et al .
the spectra were first prepared for pahfit by eliminating points with negative flux or with low ratios ( @xmath26 ) of signal - to - noise .
table 5 shows the fluxes or equivalent widths ( ew ) for the various features as given by pahfit .
the 7.7@xmath0 complex is a sum of the 7.4 , 7.6 and 7.9@xmath0 features .
pahfit does not calculate uncertainties on equivalent widths because it is difficult to compute uncertainities on the continuum of the spectrum .
the errors given in table 5 assume that the fractional errors on the equivalent widths are the same as on the integrated features , and thus are lower limits .
the temperature of the cool dust , t@xmath27k , found for ugc6879 , a high surface brightness spiral , coincides with that found for the starbursting dwarf galaxy ugc10445 ( hinz et al .
2006 ) using similar data and modeling techniques .
it is also in agreement with the submillimeter temperature estimates of such dust in lsbgs by pickering & van der hulst ( 1999 ) and with infrared and submillimeter estimates of the temperatures of other low metallicity environments such as dwarf galaxies ( popescu et al .
2002 ; lisenfeld et al .
2002 ; bottner et al .
additionally , observations of normal - sized high surface brightness galaxies , including the milky way ( reach et al . 1995 ;
lagache et al . 1998 ) , show that cool dust components exist , and it is becoming apparent that such a cool component is fairly ubiquitous across galaxy types ( see review by tuffs & popescu 2005 ) .
the total calculated dust mass of ugc6879 of @xmath28m@xmath5 falls within the range found for normal high surface brightness spiral galaxies of @xmath29m@xmath5 ( e.g. , sodroski et al . 1997 ; bendo et al .
2003 ) and is a factor of ten higher than the mass values for ugc10445 ( hinz et al .
the cool dust mass value for ugc10445 is considered a lower limit due to the fact that mips data are insensitive to dust colder than t=15 - 20k .
the scuba 850@xmath0 detection of ugc6879 allows us to better estimate the turnover of the sed .
the better constrained fit puts somewhat tighter constraints on the dust mass .
the hi gas mass to dust mass ratio found for ugc10445 was @xmath11500 ( hinz et al .
this was found to be inconsistent with the mean value of the ratio for normal spiral galaxies ( @xmath30 ; stevens et al.2005 ) , although the uncertainty in the dust mass value was large .
the total hi mass for ugc6879 is @xmath31@xmath32 ( sauty et al .
2003 ) , giving a hi gas mass to dust mass ratio of @xmath33 .
this value is consistent with the stevens et al .
( 2005 ) mean value .
popescu et al .
( 2002 ) propose that cool dust in galaxies is heated by the diffuse non - ionizing ultraviolet radiation produced by young stars , with a small contribution from the optical radiation produced by old stars .
this appears to be borne out for the high surface brightness galaxies , ugc6879 and ugc10445 , where the _ galex _ and 24@xmath0 images pinpoint the active star formation sites , and the corresponding 160@xmath0 emission traces the detectable cool dust .
figure 5 shows the central @xmath1-band surface brightnesses for the sample versus the ratio of 24@xmath0 to 160@xmath0 flux density .
the lower the central optical surface brightness for each object , the lower the this ratio appears to be .
this implies that there are not large amounts of dust extinction ; no highly obscured star formation is uncovered at 24@xmath0 , and those galaxies with the lowest surface brightnesses , i.e. , without bright star - forming regions , are not detected at 160@xmath0 .
the appearance of broad aromatic emission spectral features in the isolated star - bursting galaxies confirms the presence of dust grains indicated by the irac , mips , _ iras _ and scuba images and photometry and our dust modeling .
aromatic emission is believed to originate mostly from photodissociation envelopes at the edges of star - forming regions that are bathed in ultraviolet photons , with some suggestion that b stars no longer associated with hii regions can also contribute to the heating ( spoon 2003 ; calzetti et al .
2005 ) . in the cases of ugc6879 and ugc10445 , the high surface brightness galaxies for which we have irs spectra
, we clearly see star formation regions indicated by bright clumpy regions in the 24@xmath0 images and the corresponding 8@xmath0 emission that presumably accounts for the aromatic features .
in contrast , we see that the lsbg malin1 does not have dust emission at far - ir wavelengths , nor aromatic emission , which is not surprising , given the popescu et al .
( 2002 ) model and explanation .
malin 1 exhibits no active star - forming regions detectable at any of the wavelengths that indicate such activity . without those regions , uv photons
can not heat any existing dust to emit at long wavelengths , nor can the envelopes believed to be the site of aromatics exist .
that is not to say that dust can not exist in such an object , but simply that any such dust will not be heated and will not be detectable in far - ir images .
this appears to be consistent with results for irregular dwarf galaxies , where aromatic emission is found only in the brightest hii regions or where there is widespread , intense star formation ( hunter et al .
braine et al .
( 2000 ) calculated an average star formation rate over a lifetime of @xmath34yr for malin1 of @xmath35 yr@xmath6 based on its @xmath36-band luminosity . from this value they used scoville & young ( 1983 ) to derive a far - ir luminosity and translated this to an expected flux density of @xmath11100mjy at the _ iras _
100@xmath0 band .
this was below the detection limit of _ iras _ and indeed was not detected .
the longer integration times with mips now place that one - sigma upper limit at 160@xmath0 of @xmath1110mjy , with malin 1 still invisible .
one explanation for this low ir luminosity is that the current star formation rate is far below the average over the life of the galaxy .
additionally , simply scaling the two - component dust model for high surface brightness galaxy ugc6879 down to the 24@xmath0 flux density values for the lsbgs does not appear to fit their seds . a scaled model that fits , for instance ,
a 24@xmath0 flux density of 0.018jy , would predict a 70@xmath0 flux density of @xmath110.4jy and a 160@xmath0 flux density of @xmath111.5jy .
comparing with the measurements of ugc6614 shows 70 and 160@xmath0 outputs only @xmath1125% of these predictions .
the 160@xmath0 output of ugc6151 also appears to be somewhat below the expected value .
therefore , it appears that the emission at the longer wavelengths for at least some lsbgs is fundamentally different from that of high surface brightness galaxies and that they are not simply low - luminosity versions of normal galaxies .
either lsbgs do not produce or maintain dust in the same quantities as other galaxies , or the dust is much colder and , therefore , undetectable in the far - ir .
comparisons of the _ spitzer _ data for different lsbgs may also reveal differences in evolutionary history .
figure 6 shows the ir seds of all the galaxies in the sample .
while the two high surface brightness galaxies have similar seds from 3.6 to 160@xmath0 , the lsbgs show a variety of steepnesses between wavelengths .
some have steepnesses from 24 to 160@xmath0 that are similar to the high surface brightness galaxies , while others are shallower ( ugc6614 ) , and some appear to turn over after 70@xmath0 ( ugc9024 ) .
for instance , ugc6151 and ugc6614 have very similar far - ir flux densities , yet ugc6614 is much brighter in red giant light , as represented by the 3.6@xmath0 flux density , compared with ugc6151 .
this may imply that ugc6614 formed stars at a much greater rate in the past , accumulating an old stellar population , while ugc6151 may have formed stars at a more constant rate over its lifetime .
the relatively high metallicity of ugc6614 supports the hypothesis that its star formation was more vigorous in the past .
lsbgs are generally metal - poor , consistent with the well - known luminosity - metallicity ( @xmath37 ) relation for other galaxies ( e.g. , de naray et al . 2004 ) . in figure 7
we show the metallicities of the entire hsbg plus lsbg sample versus the absolute magnitude at 24@xmath0 .
absolute magnitudes are calculated using the mips 24@xmath0 magnitude zero point of @xmath38 calculated by engelbracht et al .
average metallicities are taken from a variety of sources in the literature ( de naray et al .
2004 ; mcgaugh 1994 ) .
others are calculated using the sloan digital sky survey - sky server .
equivalent widths of optical emission lines such as [ n ii ] and [ o iii ] are available online , and we use those values in conjunction with the rough metallicity formulations of wegner et al .
( 2003 ) and salzer et al .
( 2005 ) to obtain metallicities .
metallicities are notoriously difficult to determine for lsbgs , and the variety of sources used to obtain them for this sample may inflate errors
. however , figure 7 shows a weak correlation in the expected direction that higher metallicity galaxies have brighter absolute magnitudes at 24@xmath0 . to probe the physical properties of galaxies that may contribute to the lack of aromatic emission features , we calculate @xmath39 , a comparison of the contribution of 8@xmath0 flux with the shorter irac wavelengths defined as @xmath40 , and @xmath41 , which is the ratio of the 8 and 24@xmath0 flux densities , for all galaxies that are detected at those wavelengths ( see engelbracht et al .
2005 and their figures 1 and 2 ) .
we show @xmath39 versus @xmath41 and @xmath41 versus the metallicity of each galaxy in figure 8 .
the data points have large error bars associated with the photometry so that trends are difficult to determine .
we see that the values for the lsbgs are consistent with those found for normal galaxies by engelbracht et al .
( 2005 ) , occupying similar parameter space as their high surface brightness counterparts in both plots .
most of our sample have relatively high 8-to-24@xmath0 flux ratios , so that the correlation of increasing @xmath41 with decreasing @xmath39 is not sampled by our galaxies .
in fact , all the galaxies in our sample that are detected at both 8 and 24@xmath0 have @xmath41 larger than 0.2 , and all of the galaxies in the engelbracht et al .
( 2005 ) with @xmath41 greater than this value have detected aromatic features .
we see the same general metallicity trend as engelbracht et al .
( 2005 ) , with lower metallicity galaxies displaying weak aromatic emission , that is , diminishing 8@xmath0 flux density relative to 24@xmath0 flux density .
one explanation for this trend is that harsh radiation fields in low - metallicity galaxies destroy pah molecules ( galliano et al .
2003 , 2005 ; madden et al .
this is unlikely to be the case for lsbgs , where the radiation fields are presumably not strong enough to destroy aromatics .
another explanation is that there are not enough carbon - rich asymptotic red - giant branch stars necessary to create large amounts of aromatic molecules in low - metallicity galaxies .
_ spitzer _ data on five low surface brightness galaxies indicate that a fraction of these objects contains modest amounts of dust , despite their low metallicities and apparent transparency .
the lsbgs are detected at all irac wavelengths , and two are detected at all of the mips wavelengths . those lsbgs and late - type
high surface brightness counterparts that have detectable dust appear to be the same galaxies that have the largest amounts of star formation , while those that do not have detectable dust are the most diffuse , least star - forming galaxies ( e.g. , malin 1 ) .
one explanation for this is that any dust existing in galaxies has to be heated to temperatures in the range 15 - 20k by ultraviolet photons escaping from star - forming regions before being detectable at far - ir and submillimeter wavelengths .
the gathering evidence shows that modest amounts of dust can be created and maintained in a variety of environments and in galaxies of widely varying apparent formation histories .
we also find that lsbgs exhibit less far - ir emission and greater variety in far - ir properties than is predicted by scaling related but higher surface brightness galaxy seds .
we thank dean hines and jeroen bouwman for allowing us to use the feps data reduction pipeline .
this work is based on observations made with the spitzer space telescope , which is operated by the jet propulsion laboratory , california institute of technology under a contract with nasa .
support for this work was provided by nasa through an award issued by jpl / caltech .
this research has made use of the nasa / ipac extragalactic database ( ned ) which is operated by the jet propulsion laboratory , california institute of technology , under contract with the national aeronautics and space administration .
barth , a. 2007 , accepted to , astro - ph/0701018 bell , e. f. , barnaby , d. , bower , r. g. , de jong , r. s. , harper , d. a. , hereld , m. , loewenstein , r. f. , & rauscher , b. j. 2000 , , 312 , 470 bendo , g. j. , et al .
2003 , , 125 , 2361 bothun g. d. , impey , c. d. , malin , d. f. , & mould , j. r. 1987 , , 94 , 23 bttner , c. , klein , u. , & heithausen , a. 2003 , , 408 , 493 braine , j. , herpin , f. , & radford , s. j. e. 2000 , , 358 , 494 calzetti , d. , et al .
2005 , , 633 , 871 de naray r. k. , mcgaugh , s. s. , & de blok , w. j. g. 2004 , , 355 , 887 engelbracht , c. w. , gordon , k. d. , rieke , g. h. , werner , m. w. , dale , d. a. , & latter , w. b. 2005 , , 628 , l29 engelbracht , c. w. , et al . 2007 ,
accepted to fazio , g. g. , et al .
2004 , , 154 , 10 galliano , f. , madden , s. c. , jones , a. p. , wilson , d. d. , & le peintre , f. 2003 , , 407 , 159 galliano , f. , madden , s. c. , jones , a. p. , wilson , c. d. , & bernard , j .-
2005 , , 434 , 867 giovanelli , r. , & haynes , m. p. 1989
, , 97 , 633 gordon , k. d. , et al .
2005 , , 117 , 503 gordon , k. d. , et al .
2007 , submitted to hines , d. c. , et al . 2006 , , 638 , 1070 hinz , j. l. , misselt , k. , rieke , m. j. , rieke , g. h. , smith , p. s. , blaylock , m. , & gordon , k. d. 2006 , , 651 , 874 holwerda , b. w. , gonzalez , r. a. , allen , r. j. , & van der kruit , p. c. 2005 , , 129 ,
1396 houck , j. r. , et al .
2004 , , 154 , 18 hunter , d. a , elmegreen , b. g. , & martin , e. 2006 , , 132 , 801 impey , c. , & bothun , g. 1989 , , 341 , 89 impey , c. d. , sprayberry , d. , irwin , m. j. , & bothun , g. d. 1996 , , 105 , 209 kessler , m. f. , et al .
1996 , , 315 , l27 knezek , p. m. 1993 , ph.d .
thesis lagache , g. , abergel , a. , boulanger , f. , & puget , j .- l .
1998 , , 333 , 709 lisenfeld , u. , israel , f. p. , stil , j. m. , & siever , a. 2002 , , 382 , 860 madden , s. c. , galliano , f. , jones , a. p. , & sauvage , m. 2006 , , 446 , 877 martin , d. c. , et al . 2005 , , 619 , l1 matthews l. d. , van driel , w. , & monnier - ragaigne , d. 2001 , , 365 , 1 mcgaugh , s. s. 1992 , ph.d .
thesis mcgaugh , s. s. 1994 , , 426 , 135 mcgaugh , s. s. & bothun , g. d. 1994 , , 107 , 530 mcgaugh , s. s. , schombert , j. m. , & bothun , g. d. 1995 , , 109 , 2019 neugebauer , g. , et al .
1984 , , 278 , l1 oneil , k. , bothun , g. d. , schombert , j. , cornell , m. e. , & impey , c. d. 1997a , , 114 , 2448 oneil , k. , bothun , g. d. , & impey , c. d. 1997b , bulletin of the american astronomical society , 29 , 1398 patterson , r. j. & thuan , t. x. 1996 , , 107 , 103 pickering , t. e. , impey , c. d. , van gorkom , j. h. , & bothun , g. d. 1997 , , 114 , 1858 pickering , t. e. 1998 , ph.d .
thesis pickering , t. e. , & van der hulst , j. m. 1999 , astronomische gesellschaft meeting abstracts , 15 , 2 popescu , c. c. , tuff , r. j. , vlk , h. j. , pierini , d. , & madore , b. f. 2002 , , 567 , 221 reach , w. t. , etal .
1995 , , 451 , 188 rieke , g. h. , et al .
2004 , , 154 , 25 roennback , j. , & bergvall , n. 1995 , , 302 , 353 salzer , j. j. , lee , j. c. , melbourne , j. , hinz , j. l. , alonso - herrero , a. , & jangren , a. 2005 , , 624 , 661 sauty , s. , et al .
2003 , , 411 , 381 scoville , n. , & young , j. s. 1983 , , 265 , 148 smith , j. d. t. , et al .
2004 , , 154 , 199 smith , j. d. t. , et al .
2006 , , accepted , astro - ph/0610913 sodroski , t. j. , odegard , n. , arendt , r. g. , dwek , e. , weiland , j. l. , hauser , m. g. , & kelsall , t. 1997 , , 480 , 173 sprayberry , d. , impey , c. d. , bothun , g. d. , & irwin , m. j. 1995 , , 109 , 558 spoon , h. w. w. 2003 , ph.d .
thesis , rijksuniversiteit groningen stansberry , j. , et al .
2007 , submitted tostevens , j. a. , amure , m. , & gear , w. k. 2005 , , 357 , 361 tuffs , r. j. & popescu , c. c. 2005 , air conf .
proc . 761 : the spectral energy distributions of gas - rich galaxies , 761 , 344 vallenari , a. , schmidtobreick , l. , & bomans , d. j. 2005 , , 435 , 821 van der hulst , j. m. , skillman , e. d. , smith , t. r. , bothun , g. d. , mcgaugh , s. s. , & de blok , w. j. g. 1993 , , 106 , 548 van zee , l. , haynes , m. p. , & giovanelli , r. 1995 , , 109 , 990 van zee , l. 2000 , , 199 , 2757 wegner , g. , salzer , j. j. , jangren , a. , gronwall , c. , & melbourne , j. 2003 , , 125 , 2373 lllcccc malin1 & s & 24750 & 25.50 & -22.50 & 2004 jun 9 & 2005 jan 30 + ugc5675 & sm : & 1102 & 23.70 & -12.95 & 2004 apr 26 & 2004 jun 2 + ugc6151 & sm : & 1331 & 23.51 & -17.21 & 2004 may 18 & 2004 jun 4 + ugc6614 & ( r)sa(r)a & 6351 & 24.30 & -20.00 & 2003 dec 19 & 2004 jun 4 + ugc9024 & s & 2323 & 24.71 & -16.58 & 2004 jan 20 & 2004 jul 10 + ugc6879 & sab(r)d & 2383 & 20.40 & -18.20 & 2004 jun 9 & 2004 jun 4 + ugc10445 & sbc & 963 & 21.79 & -17.53 & 2004 mar 8 & 2004 mar 21 + lcccccc malin1 & 2005 jan 4 & sl & 1707 & -160.92 + & & ll & 377.5 & 115.55 + ugc6879 & 2004 jun 27 & sl & 1707 & 19.01 + & & ll & 377.5 & -64.50 + ugc10445 & 2004 jul 14 & sl & 1707 & 42.37 + & & ll & 377.5 & -41.15 + lcccccccccc malin 1 & point - like & no detection & no detection & @[email protected] & 24.90 & @xmath44 & & @xmath45 & + ugc5675 & & no detection & no detection & & & @xmath46 & & @xmath47 & + ugc6151 & extended & extended & extended & @xmath48e-4 & 62.25 & @xmath49 & 49.25 & @xmath50 & 80 + ugc6614 & extended & extended & extended & @xmath51e-3 & 62.25 & @xmath49 & 54.18 & @xmath52 & 56 + ugc9024 & point - like & point - like & no detection & @xmath53e-4 & 24.90 & @xmath54 & 24.63 & @xmath47 & + ugc6879 & extended & extended & extended & @xmath55e-3 & 62.25 & @xmath56 & 54.18 & @xmath57 & 56 + ugc10445 & extended & extended & extended & @xmath58e-3 & 105.8 & @xmath59 & 98.50 & @xmath60 & 120 + lcccccccccc malin 1 & [email protected] & [email protected] & 18 & [email protected] & [email protected] & 18 + ugc5675 & [email protected] & [email protected] & 30 & @xmath61e-5 & @xmath62e-4 & + ugc6151 & [email protected] & [email protected] & 60 & [email protected] & [email protected] & 60 + ugc6614 & [email protected] & [email protected] & 108 & [email protected] & [email protected] & 60 + ugc9024 & [email protected] & [email protected] & 60 & [email protected] & [email protected] & 24 + ugc6879 & [email protected] & [email protected] & 108 & [email protected] & [email protected] & 60 + ugc10445 & [email protected] & [email protected] & 78 & [email protected] & [email protected] & 78 + lcccccccccc ugc6879 & & + & [email protected] + @xmath63ne @xmath64 $ ] & [email protected] + @xmath63s @xmath64 $ ] & [email protected] + 7.7@xmath0 complex ew & [email protected]@xmath0 + & & + ugc10445 & & + & [email protected] + @xmath63ne @xmath64 $ ] & [email protected] + @xmath63s iii@xmath64 $ ] & [email protected]
+ 7.7@xmath0 complex ew & [email protected]@xmath0 + | we examine the infrared properties of five low surface brightness galaxies ( lsbgs ) and compare them with related but higher surface brightness galaxies , using _ spitzer space telescope _ images and spectra .
all the lsbgs are detected in the 3.6 and 4.5@xmath0 bands , representing the stellar population .
all but one are detected at 5.8 and 8.0@xmath0 , revealing emission from hot dust and aromatic molecules , though many are faint or point - like at these wavelengths .
detections of lsbgs at the far - infrared wavelengths , 24 , 70 , and 160@xmath0 , are varied in morphology and brightness , with only two detections at 160@xmath0 , resulting in highly varied spectral energy distributions .
consistent with previous expectations for these galaxies , we find that detectable dust components exist for only some lsbgs , with the strength of dust emission dependent on the existence of bright star forming regions .
however , the far - infrared emission may be relatively weak compared with normal star - forming galaxies . |
time - dependent density functional theory ( tddft ) is a remarkably successful theory of many - body systems in time - dependent external potentials @xcite .
the analog of static density functional theory @xcite , tddft is based on the runge - gross proof of a one - to - one mapping between the external potential and the time - dependent density of electrons evolving under it , for a specified initial wavefunction @xcite . because knowledge of the external potential specifies the many - body hamiltonian ,
all properties of the interacting electronic system can be extracted from just its one - body density and the initial - state . in practise
, tddft utilizes the kohn - sham ( ks ) scheme , where the time - dependent density of an interacting system evolving in an external potential @xmath0 is calculated from a fictitious noninteracting system of fermions moving in an effective potential , the ks potential @xmath1 , that is defined to reproduce the density of the interacting system .
the ks potential is written as the sum : v(,t)= v(,t ) + v(,t ) + v(,t ) [ eq : vs ] where @xmath2 denotes the time - dependent density , and @xmath3(@xmath4 ) is the initial interacting(noninteracting ) wavefunction . the second term , @xmath5(\br , t ) = \int \frac{n(\br',t)}{\vert\br - \br'\vert}d^3r'$ ] , is the classical hartree potential , while the third is the exchange - correlation ( xc ) potential @xmath6 .
this is unknown as a functional of the time - dependent density and initial - states , and must be approximated in practise .
armed with an approximation for @xmath7(\br , t)$ ] , one then propagates the time - dependent ks equation : i_t _ i(,t ) = ( -^2/2 + v(,t))_i(,t ) finding the @xmath8 single - particle orbitals @xmath9 , that initially made up the initial ks determinant @xmath4 .
( atomic units , @xmath10 are used throughout this paper ) .
the density of the interacting system is obtained from @xmath11 , and the runge - gross theorem then assures us that all properties of the true system can be extracted from the orbitals . in the linear response regime , in which lie the majority of tddft applications so far , excitation energies and oscillator strengths are obtained from a perturbative formulation of these equations @xcite , leading to the matrix formulation that operates in most of the quantum chemistry codes today .
clearly the accuracy of tddft then depends on the approximation used for the xc kernel . despite being known generally to have `` memory - dependence '' that is
, @xmath12 depends on the history @xmath13 and on the initial - states , @xmath3 and @xmath4 @xcite almost all approximations used today neglect this , and bootstrap a ground - state functional : that is , the instantaneous density is input into an approximate ground - state functional , @xmath14(\br , t ) = v\xc^{\rm gs}[n(\br t)](\br t)$ ] .
the superscript a denotes this `` adiabatic approximation '' . in the linear response regime
@xcite , adiabatic tddft has achieved an unprecedented balance between accuracy and efficiency for the calculations of excitations and response properties .
for the first time , one can compute the quantum mechanical spectrum of systems as large as biomolecules ( see eg .
@xcite ) , run coupled electron - ion dynamics on interesting chemical reactions @xcite , and study molecular transport through nanostructures @xcite
. excitations for which the adiabatic approximation fails have been recognized , and , in various stages of being corrected @xcite .
sometimes the spatial long - ranged - ness of the true functional is important , and approximations more sophisticated than the usual semi - local ones are required @xcite .
but tddft applies also for strong field dynamics , and this is the regime it is particularly promising for , given that correlated wavefunction calculations for more than two or three electrons in strong - fields become prohibitively expensive @xcite .
compared to calculations of spectra , in addition to an approximation for the xc potential , now a new ingredient needs to be considered in the calculation : _ observable functionals _ ,
i.e. the observables of interest must be expressed as functionals of the ks wavefunction .
the rg theorem guarantees that all properties of the true system can be obtained from the ks orbitals , as they themselves are implicit density - initial - state - functionals , but how ? if the property is related directly to the density , then no additional observable functional is needed .
for example , in high - harmonic - generation it is the dipole moment , @xmath15 that is of interest , and indeed tddft calculations have been successful for high - harmonic generation spectra of a range of interesting systems @xcite .
but if we are interested in , for example , measuring double - ionization probabilities , we require the interacting pair - density as a functional of the density ( or ks orbitals ) , and this is highly non - trivial @xcite .
non - sequential double - ionization is a fascinating problem that dogged tddft for several years @xcite : tddft s promise in capturing electron correlation made it attractive for this problem , but it was soon realized that not only does one need to go beyond the usual semilocal gga s for the xc functional in order to obtain the knee structure @xcite , but using an uncorrelated ks expression for the pair - density yielded a knee that was too high .
( in ref .
@xcite , an adiabatic correlated expression , based on the ground - state perdew - wang pair density @xcite was shown to lower the knee appropriately ) .
approximating a general observable simply by the appropriate operator evaluated on the ks wavefunction can lead to gross inaccuracies , due to the lack of correlation in the ks wavefunction itself .
it is important to remember that the ks wavefunction is designed to reproduce the exact interacting one - body density , but is _ not _ supposed to be an approximation to the true wavefunction .
indeed another aspect of the same double - ionization problem recently demonstrated this : wilken and bauer s calculations of ion - momentum - recoil distributions in double - ionization @xcite showed that the ks momentum - distributions were drastically wrong , displaying a single maximum instead of the characteristic two - hump structure , and with a significantly overestimated magnitude .
typically ks momentum distributions are not those of the true system , even if the exact ks orbitals were used , that is if the exact @xmath6 was known for the problem , and used to generate the ks orbitals .
another example of this will be given in sec .
[ sec : momentum ] .
returning to the question of the approximate @xmath7(\br , t)$ ] : the accuracy of the adiabatic approximation for real - time dynamics gets mixed reviews . in strong - field double - ionization , the xc potential appears not to be significantly non - local in time in a wide range of cases , although it depends on how the field is ramped on @xcite .
dynamics of two electrons in parabolic wells , on the other hand , yield an exact correlation potential that appear to depend strongly on the history @xcite .
we shall revisit ref .
@xcite in sec .
[ sec : hooke ] .
although significant advances have been made in understanding @xcite and modelling @xcite memory - dependence , it remains today a difficult problem .
one situation in which the usual approximations ( spatially semi - local and adiabatic density dependence ) do poorly , is when the true wavefunction fundamentally can not be described by a single slater - determinant .
this is also true for the ground state where molecular dissociation curves are a notorious problem @xcite : when an electron - pair bond dissociates , the true wavefunction develops a heitler - london character , which in a minimal description requires two slater - determinants .
the ks description however operates via a single slater - determinant .
the exact xc potential allows for the fundamental difference in the ks and true wavefunctions by developing rather stark and unusual peak and step features , that are difficult to capture in approximations . in time - dependent problems , coefficients of the interacting wavefunction expanded in a basis of single - slater - determinants
may change in time dramatically : whereas at one time a single - slater determinant may dominate , at a later time , two single - slater determinants are essential .
such situations may happen in electronic quantum control problems ( now becoming experimentally accessible with the advent of attosecond lasers ) @xcite .
consider for example evolving the ground - state of the he atom ( 1s@xmath16 ) to its first accessible excited state ( 1s2p ) .
the ks ground state is a single - slater determinant composed of a doubly - occupied spatial orbital .
this evolves under the ks hamiltonian , which is a one - body evolution operator .
the target single excitation is however a _
double _ slater determinant , and therefore can not be reached by any one - body operator . the exact ks system attains the density of the true target state using a single orbital , not two .
this yields a very unnatural description of the true system ; consequently , the ks potential develops rather unnatural features , difficult to model @xcite .
the inability of ks methods to change occupation numbers leads to challenges in developing approximations .
the two ingredients needed in a tddft calculation - approximations for the xc potential and the relevant observable functional - determine the accuracy of the results . as discussed above ,
sometimes the simplest approximations work well , but in other cases , it is more challenging to derive suitable approximations . in this paper
, we explore a generalization of tddft based on the phase - space density . that is , instead of taking the coordinate - space density , @xmath2 as our basic variable , we consider taking the density in phase - space , @xmath17 : a quasi - probability distribution which is somewhat related to the probability of finding an electron at position @xmath18 with momentum @xmath19 ( see sec . [ sec : wigner ] ) .
the idea here is two - fold : first , as the basic variable has more information , functionals of them may be simpler to approximate , or equally , simple functionals of them may be more accurate than simple functionals of the coordinate - density alone , while retaining the favorable system - size scaling property of density functional theories .
second , more observables are directly obtained without needing additional observable - functionals , in particular , those pertaining to one - body operators , such as momentum , kinetic energy .
as the coordinate and momentum operators do not commute , there is no unique phase - space density function , and we begin in sec . [
sec : wigner ] by discussing the wigner function , which turns out to be a particularly useful choice for our purposes .
we discuss the 1 - 1 mapping between wigner functions and potentials , for a given initial - state .
we discuss also the formal equivalence of the wigner function with time - dependent density - matrix functional theory , which has seen a recent resurgence of interest @xcite .
next , we return to tddft and explore the true and ks phase - spaces , and the implications of a phase - space based theory , for two examples pertaining to the challenges in tddft described above : momentum distributions ( sec .
[ sec : momentum ] ) and history - dependence in the xc functional ( sec [ sec : hooke ] ) .
quantum phase - space distributions have played an important role in the development of quantum mechanics . although a precise joint probability distribution in position and momentum is impossible due to the heisenberg uncertainty principle , various quasi - probability distributions
have been defined , which contain position and momentum distribution information in a way consistent with the uncertainty principle .
such distributions are particularly useful in a semiclassical context , relating the quantum states to the underlying classical trajectories , and have been exploited extensively , for example in quantum chaos and quantum optics .
phase - space approaches have however been largely , although not entirely @xcite neglected in density functional theories , where position plays a preferred role over momentum .
most recently @xcite , gill and co - workers , have developed and tested models for the ground - state correlation energy based on various `` phase - space intracules '' which are essentially different contractions of the two - body reduced wigner function .
for example , there is the omega intracule , which is a function of three variables ; the distance between two electrons , the magnitude of the difference in their momentum vectors , and the angle between their position and momentum @xcite .
this is motivated by the very physical idea that the correlation between two electrons depends on their relative momentum as well as on their relation distance @xcite .
wigner functions were first introduced in 1932 @xcite , with the intent of application to many - body systems .
the wigner function quasi - probability distributions are real but not positive semi - definite ; indeed the negative areas have been interpreted as an indication of non - classical behaviour , from interference to tunneling processes , and research in how to classify these areas continues today ( see for example refs .
although initially introduced to treat many - particle dynamics , the majority of applications of the wigner function actually involve one - particle systems , especially for making quantum - classical correspondence .
it has been argued that the wigner function is a particularly suitable approach to studying transport @xcite due to its setting within a semiclassical picture while yet being rigorously quantum mechanical .
@xcite develops a formalism based on reduced - wigner - functions for indistinguishable fermionic systems accounting for pauli - exchange ( but not correlation ) . for a system of @xmath8 identical particles
, the reduced @xmath20-body wigner phase space density is defined as the fourier transform of the @xmath20th - order density matrix @xmath21 : w_k(_1 .. _k,_1 .. _k , t)=()^3k d^3y_1 .. d^3y_k e^i_k*p_k * _ k(*r_1 * -*y_1*/2 .. *r_k*-*y_k*/2 , * r_1*+*y_1*/2 .. *r_k*+*y_k*/2;t ) where the @xmath20-th order density matrix is @xmath22 .
we consider here a density - functional theory based on the one - body wigner function @xmath23 ( henceforth referred to as simply wigner function ) : w(,,t ) = ( ) ^3d^3y_1(-*y*/2,+ * y*/2,t)e^i [ eq : wigner ] the one - body densities in coordinate - space and momentum - space are directly obtained by integration : n(,t ) = w(,,t ) d^3p + n(,t ) = w(,,t ) d^3r [ eq : nrnp ] several observations can immediately be made about such a phase - space density - functional theory ( psdft ) whose basic variable is the wigner phase - space density ( taking the role of the coordinate - space density in usual density functional theories ) .
first , the theory is formally equivalent to the recently developed time - dependent one - body density - matrix functional theory ( 1dmft ) @xcite : @xmath24 and @xmath25 contain exactly the same information , in different forms related via a fourier transform .
second , as in the 1dmft , there is no kohn - sham counterpart : the wigner - function of an interacting system can not be reproduced by any non - interacting system in a local potential . in a non - interacting system , the one - body reduced density matrix is idempotent , while in an interacting system , it can not be .
it is interesting that generally while a non - interacting system may be found that reproduces the coordinate - space density of an interacting system , and a different non - interacting system may be found reproducing its momentum - space density , one can not find a non - interacting system that reproduces its phase - space density .
third , the wigner function directly gives the expectation value of any one - body operator : no additional observable functional would be needed , for example , for kinetic energy or momentum distributions .
fourth , the one - to - one mapping between the potential and the phase - space density , for a given initial - state , follows directly from the runge - gross theorem : we have w(r , p , t ) n(r , t ) + ( n(,t),_0 ) ( _ 1 .. _n , t ) where both arrows indicate a unique mapping , the first following from @xmath26-integration ( eq . [ eq : nrnp ] ) , and the second from rg , while also ( _ 1 .. _n , t ) w(r , p , t ) uniquely for a specified initial - state .
hence , for local external potentials , there is a 1 - 1 mapping between the one - body wigner function and the many - body wavefunction for a given initial state , i.e. that all observables may be extracted from the one - body wigner function and initial - state alone .
we note that this 1 - 1 mapping holds for local external potentials ( i.e. purely multiplicative in coordinate - space .
gilbert @xcite has shown that for the _ ground state _ , the extension of the hohenberg - kohn therem to one - body reduced density matrices applies also to the wider class of _ non - local _ potentials , @xmath27 .
formally , one may be led to search for a one - to - one @xmath28 mapping in the time - dependent case also , given that these are conjugate variables that couple together in an energy functional , ( c.f .
the @xmath29 mapping of runge - gross or hohenberg - kohn @xcite , and the @xmath30 mapping of ghosh and dhara in td - current - density functional theory @xcite .
although a proof for time - dependent non - local potentials is yet to be found , the mapping may be generalized to vector potentials @xmath31 , extending the realm of application to include magnetic fields : w ( , t ) + ( ( , t),_0 ) ( _ 1 .. _n , t ) where the first arrow follows from ( , t ) = d^3 * p * w(,*p*,t ) and the second from the ghosh - dhara proof @xcite of the 1 - 1 mapping between currents and vector - potentials .
that is , any observable of any time - dependent interacting electronic system evolving under an external vector potential , is a functional of its initial wavefunction and time - dependent one - body phase - space density .
the runge - gross proof proceeded in two steps , where the first proved a one - to - one mapping between currents and ( scalar ) potentials for a given initial state.the second step , to go from currents to densities , holds under the condition that a boundary term , that involves the density and the gradient of the potential , vanishes .
the implications of this boundary condition have been discussed in ref .
@xcite ; it means that tddft can not be applied to periodic systems in uniform electric fields @xcite .
we point out that here there is no requirement on the boundary conditions for the @xmath32 ( or @xmath33 mappings to hold .
time - dependent phase - space functional theory involves solving the equation of motion for the wigner function : ( * r , p*,t)&=&w(*r , p*,t ) + & -&id^3yd^3p_1d^3p_2d^3r_2 e^i(-_1 ) ( - ) w_2(,_2,_1,_2,t ) where the last term must be approximated as a functional of the one - body @xmath24 and the initial - state @xmath3 .
this equation follows directly from that for the density - matrix , which may be more familiar : i(*r*,*r*,t)=(-^2/2 + v(*r*,t)+ ^2/2 - v(*r*,t))_1(*r*,*r*,t)+ d^3r_2(-)_2(,_2,,_2,t ) [ eq : rho1dot ] where @xmath34 is the second - order density - matrix , to be approximated as a functional of @xmath25 .
this equation is first in the bbgky ( bogoliubov - born - green - kirkwood - yvon ) heirarchy of reduced density matrix evolutions , which gives the equation of motion for the @xmath20th reduced dm in terms of the @xmath35st and lower dm s .
[ eq : rho1dot ] is the crux of 1dmft , which has only very recently begun to be explored .
all the development and results so far have been in the linear response regime @xcite , using functionals for @xmath34 which have been adiabatically bootstrapped from ground - state ones ( eg . refs.@xcite ) .
the accuracy of the time - dependent psdft(1dmft ) depends on approximate functionals for the term involving the second - order wigner function @xmath36(@xmath34 ) in terms of @xmath24(@xmath25 ) .
as has been noted above , and earlier in the literature , the kinetic term is treated exactly , in contrast to dft , where the correlation potential / energy has a kinetic component .
the simplest approximation would be the uncorrelated one , where @xmath34 is written as a product of antisymmetrized one - body @xmath25 s .
this truncates the bbgky heirarchy at the first level , and is equivalent to time - dependent hartree - fock .
its inability to change occupation numbers makes it unable to treat many of the challenges that tddft faces , that the phase - space dft would hope to treat .
for example , the electronic quantum control problem mentioned in the introduction ( see also @xcite ) .
section [ sec : momentum ] discusses another example where a change in occupation number is crucial for an accurate description .
natural orbitals @xmath37 with occupations @xmath38 are defined by diagonalizing @xmath39 : _
1(*r*,*r*,t)=_if_i(t ) _
i(*r*,t ) ^*_i(*r*,t ) we may define `` natural wigner orbitals '' via the analogous expansion : w(,,t ) = _ if_i(t)w_i(,,t ) where @xmath40 . from eq .
[ eq : rho1dot ] the time evolution of the occupation numbers can be derived : i(t)=_2(*r , r_2,r,r_2*,t ) + ^*_i(*r*,t ) _
i(*r*,t ) d*r*d*r*d*r_2 * -c.c the right - hand - side of this equation is only nonzero if the contraction of @xmath34 on to the natural orbitals , @xmath41 , is imaginary .
therefore occupation numbers can not change with using an uncorrelated product expression for @xmath34 .
it was also shown that bootstrapping any adiabatic functional from ground - state 1dmft can not change occupation numbers @xcite .
a systematic approach might consider then truncating the bbgky heirarchy at the next order ; that is , solving the equation of motion for @xmath34 consistently with eq .
[ eq : rho1dot ] , but using the uncorrelated ( antisymmetrized product form ) for @xmath42 .
however , it has been shown that this violates fundamental trace relations between density matrices of different order @xcite ; ways to get around this become rapidly complicated .
moreover , one must deal with a four - point function ( second - order wigner function ) rather than the two - point one ( first - order ) .
the problem of an appropriate @xmath36-functional in terms of @xmath24 , which is able to change occupation numbers while retaining fundamental physical properties , is an important area for future research .
we now describe in a little more detail two problems that are particularly challenging for tddft approximations , and discuss how a psdft can ameliorate them . in general ,
the momentum - distribution of the ks system is not the same as that of the true system : although the sum of the squares of the ks orbitals in coordinate - space yield the exact density of the true system , the sum - square of their fourier transforms to momentum - space does not yield the exact momentum - density . in cases
where the correlation component to the kinetic energy , @xmath43 is relatively small , calculating the momentum distribution simply from the fourier transforms of the relevant ks orbitals can sometimes be quite accurate , as recently illustrated by the compton profiles obtained in ref .
@xcite , provided accurate enough ground - state xc potentials are used .
earlier , for the the ground state case , lam and platzman derived a local - density approximation for the correlation correction to ks momentum - densities for compton profiles @xcite .
the ks orbital momentum distribution has also been successful in electron momentum spectroscopy @xcite : this is essentially a triple differential cross - section measurement , where one electron is scattered from an atom , molecule , or surface , causing the ejection of another electron .
the experiment obtains a map of the electron momentum distribution before ejection , thus imaging the momentum distribution of the dyson orbital .
@xcite show that the ks orbital yields a good approximation to the dyson orbital in their momentum distributions , better than , for example , hartree - fock .
away from the ground - state , wilken and bauer @xcite have shown that ks momentum - distributions of tddft are miserably poor for non - sequential double - ionization processes ; not only their shape is wrong , lacking the characteristic double - hump structure , but also their magnitude is significantly underestimated . ref .
@xcite develops a `` product - phase '' approximation to extract the true momentum distribution from the ks system , i.e. an observable functional for the momentum - distribution .
the possibility of working directly in momentum - space is hindered by the lack of a one - to - one mapping theorem for momentum - densities ( even for the ground - state case ) @xcite .
in fact the earlier interest in developing ways to associate phase - space distributions to a given ground - state coordinate - space density @xcite was motivated by trying to get good momentum - space properties derived from the coordinate - space density . one may view the density - matrix developments as a rigorous approach to this problem ; as discussed in sec .
[ sec : wigner ] , the exact one - body density matrix , or equivalently , the one - body phase - space density , directly yields all one - body information , i.e. the expectation value of any one - body operator , such as the momentum - density operator , is exact .
even a simple model consisting of two electrons in one - dimension illustrates the discrepancy between the ks and true momentum - densities in ionization or scattering problems . with an exactly solvable model
, one can pinpoint the origin of the differences between the momentum distributions of the true and ks systems . in this particular example
, we will see it is due to the single - slater determinant description being a poor description of a state in which two electrons are doing completely different things .
one electon is at rest ( modelling a `` bound '' electron ) , while the other is moving away from it ( modelling a `` scattered '' or `` ionized''electron ) , and we consider that enough time has elapsed that the true wavefunction describes independent electrons : ( x , x,t ) = ( _ 0(x)_p(x,t ) + _ 0(x)_p(x , t))/ the orbital for the bound electron is simply chosen as a gaussian @xmath44 while that for the ionized electron is , for simplicity , chosen to be a free - particle dispersing gaussian , which was initially ejected by some laser field ( turned off thereafter ) , such that its initial spread in position is @xmath45 , and average momentum is @xmath46 : @xmath47 the coordinate - density , @xmath48 , and momentum - density , @xmath49 of this state at a snapshot in time is shown in fig .
[ fig : mom - dist ] : neither are remarkable , as both are what might be expected quasiclassically from the sum of two probability distributions centered around the average position and momentum for each electron
. now we consider the kohn - sham description , which , for two electrons in a singlet state , operates via the doubly - occupied orbital , reproducing the time - evolving density @xmath50 of the true system : @xmath51 where @xmath52 is the current - density of the true system .
we compute the ks momentum distribution , @xmath53 from the fourier transform @xmath54 of the orbital @xmath55 . from figure
[ fig : mom - dist ] , we see that while the overall envelope of the ks momentum distribution follows that of the true system , there are distinct oscillations . in figure
[ fig : wigner_ion ] we plot the corresponding wigner phase - space distributions , @xmath56 and @xmath57 ; oscillations are evident in the latter also , and dip into negative values .
this is a typical signature of non - classical behavior : the ks orbital describes an electron spatially delocalized in two regions , impossible for a classical particle .
the coherence between the two separated parts of the wavepacket gives rise to the oscillations in the phase - space density : integrating over momentum washes out these oscillations , yielding the sum of the two gaussians in the left - hand - side of fig .
[ fig : mom - dist ] ( i.e the true interacting density ) whereas the oscillations persist in the ks momentum distribution on the right . on the other hand
, the true wigner function looks much like a classical joint phase - space distribution would be : a peak at the position and momentum of each electron , each locally smeared out . aside from the antisymmetrizing ,
the true wavefunction has classical features , with each electron in their independent gaussian wavepackets , quite in contrast to the ks description which describes both electrons in the same delocalized orbital . although yielding the exact coordinate density of the true system , this exact ks wavefunction is clearly far from the true wavefunction .
fundamentally , it associates to the true two - determinant state , a _
single _ determinant ; in a sense , this is a time - dependent analog of the static correlation problems that haunt ground - state density functional theory , for example in dissociation problems @xcite .
we can understand the discrepancy in the momentum distributions as arising from this fundamental difference in the nature of the true and ks states in this particular case .
( in the next section , we shall see an example where the momentum distributions are not so different ) .
we stress that this is the ks system with the _ exact _ xc functional : what is required in tddft to get the correct exact momentum distribution from this , is the ( unknown ) momentum observable functional . what are the implications of this for a psdft ?
as explained in the earlier section , in a phase - space dft we deal directly with the interacting system ; there is no non - interacting system that can reproduce an interacting phase - space density .
momentum distributions are immediately obtained without the need for an additional observable functional . and
, had we the exact functional for the @xmath36-term in terms of @xmath24 , these would be exact . in practise , approximations for this term are needed , and for this example , it is crucial for the approximation to be able to change occupation number .
the state starts in a ground - state , both electrons bound , which , in a first approximation would be in the same doubly - occupied orbital.in the process of ionization , one electron leaves , and a two - orbital description is necessary , each with occupation number 0.5 .
any approximation unable to change occupation numbers will once again yield wiggles in the momentum and phase - space distribution .
these should not be there in such a state which is of classical nature .
we comment here that a similar problem occurs for certain quantum control theory type problems , as discussed in the introduction .
the time - dependent exchange - correlation potential functionally depends on the initial state ( both the true and ks ) as well the history of the density @xcite .
this dependence on the past is referred to as `` memory - dependence '' .
any adiabatic approximation ( e.g. , alda ) utilizes only the instantaneous density , so ignore this temporal nonlocality of the xc potential . yet
, adiabatic approximations work remarkably well for the calculation of most , but not all , excitations and their oscillator strengths . for dynamics in strong fields ,
there are both parameter regimes and phenomena in which the memory - dependence is minimal , but also examples where memory - dependence is vital @xcite .
memory - dependence is not easy to capture in approximations although there has been intense progress recently @xcite
. ref .
@xcite used a numerically exactly - solvable example to demonstrate explicitly a case where the correlation potential depends on the density ultra - nonlocally in time .
there , time slices were found in which the density was practically identical , semi - locally in time , whereas the correlation potential differed significantly .
we shall revisit this example here , with a view to asking whether functionals in a psdft could be less non - local in time , and thus easier to approximate . specifically
, we shall ask whether the _ phase - space _ density can distinguish those time slices mentioned above , that the coordinate - space density could not .
we shall find that indeed it can , and moreover , that the size of the differences in the phase - space density tracks the difference in the correlation potential .
this suggests that the memory - dependence in a psdft could be milder than in the usual dft .
time - dependent hooke s atom is a system of two coulombically interacting electrons in a time - dependent parabolic well : the hamiltonian is @xmath58 for our investigations , we chose @xmath59 , as in ref .
the hamiltonian decouples into relative ( @xmath60 ) and center of mass ( @xmath61 ) co - ordinates , so one may solve for the exact time - dependent 2-electron wave function as a product of @xmath61 and @xmath60 wavefunctions , @xmath62 and thereby obtain time - dependent density @xmath63 .
knowledge of the time - evolution of the density yields the doubly - occupied time - dependent ks orbital , and then , via inversion of the time - dependent ks equation , the exchange - correlation potential @xmath64 .
a detailed derivation can be found in refs .
@xcite , along with computational details .
we mention here only that the time - propagation method used is the crank - nicholson scheme , and a grid chosen to sample points near the origin more densely than further away , since the density decays exponentially . with chosen parameters @xmath65 , @xmath66 , and @xmath67 , the density retains a near - gaussian shape at all times , with a time - varying width ; thus it may be parametrized by its variance @xmath68 @xcite .
this is shown in the top panel of the fig .
[ fig : nonloc ] .
the lower panel shows a density - weighted correlation potential , @xmath69 , which was also studied in the earlier paper @xcite , as a simple way to track the time - evolution of the correlation potential .
a pair of time slices is indicated within which @xmath70 is practically identical in each time slice , while the density - weighted correlation potential @xmath71 differs quite dramatically .
other such pairs of time slices may be found .
this indicates that local , or even semi - local in time density - dependence is not enough to specify the correlation potential , i.e. that the correlation potential depends on a significant history of the density @xcite .
any adiabatic approximation that uses only instantaneous density information would erroneously predict the same correlation potential for each time slice .
( lower panel ) and @xmath72 ( top panel ) .
in the two time slices indicated , the density is almost identical locally in time , whereas the density - weighted correlation potential in the lower panel is quite distinct.,width=340,height=226 ] we now consider the phase - space distribution at these time slices : if these are also nearly identical at these pairs of time slices , then it would suggest the memory - dependence in a psdft is just as non - local as in dft and just as challenging to model .
we shall instead find that the phase - space distributions distinguish the system at these different pairs of time slices , with the differences greater with greater differences in the density - weighted correlation potential @xmath73 .
this suggests that perhaps the memory - dependence would be less severe in psdft , and therefore perhaps easier to approximate .
knowledge of the true and ks wavefunctions , enables us to derive expressions for wigner phase - space distribution for the interacting and ks cases : @xmath74 @xmath75 where @xmath76 = @xmath77 , @xmath78 and @xmath79 is the angle between @xmath80 and @xmath81 and @xmath82 is the bessel function of zeroth order . in eq .
[ eq : ws ] , @xmath55 is the ks orbital . both distributions @xmath24 and @xmath83 depend only on the magnitudes of the coordinate and momentum , @xmath84 and @xmath26 , and the phase - space angle @xmath79 ; for simplicity , we look at the angle - averaged quantities : @xmath85 + further , to highlight the component of the wigner function that is due to correlation , we take the difference between the angle - averaged true and ks wigner functions , defining w(r , p , t)= w_av(r , p , t)- w_s , av(r , p , t ) to simplify the analysis even further , we consider only the momentum - distribution of this correlation component , defined as : @xmath86 we now turn to the plots in figure [ fig : nc_hooke ] .
the top left panel shows the true and ks momentum probability distributions at @xmath87 , which has one of the larger values for the density - weighted correlation potential in this run ( fig .
[ fig : nonloc ] ) .
we include in all these plots the jacobian factor , so that we are actually plotting @xmath88 .
although the momentum distributions are not identical , the differences are relatively small , in contrast to the ionization example in the previous section . .
top right and lower two panels : correlation components of the momentum distributions @xmath89 at three different time - pairs indicated . in each time - pair , the coordinate density , @xmath90 , is identical , while the momentum - density is not , with the difference in its correlation component @xmath91 growing as the difference in the density - weighted correlation potential @xmath92 at those times.,width=302,height=226 ] we now study their difference @xmath91 .
the other three panels show @xmath91 at pairs of times at which the coordinate - density are practically identical ( i.e. essentially indistinguishable to the eye on the scale of plots similar to the top left panel ) . the first time - pair , ( 4.8,29.8 ) , is chosen near the minimum of the @xmath70 at each of the two time slices indicated in figure [ fig : nonloc ] ; here the difference in @xmath73 is relatively small ( 0.0005 au ) , while it is bigger ( 0.011 au ) at the second time pair ( 5.1,29.3 ) , and bigger still ( 0.0264 ) at the third pair ( 5.5,29.3 ) .
in contrast to the coordinate - density , the momentum - densities are not identical at these pairs of times . moreover
, their difference increases with the difference in the value of the density - weighted correlation potential at those times .
the momentum distribution appears to somewhat track the correlation potential .
what are the implications of this for memory in a psdft ?
the results suggest that having momentum information in the basic variable may reduce their memory - dependence .
of course this is not conclusive , but the fact that momentum distributions distinguish the system at pairs of times when the coordinate - density does not , yet the correlation potential is different , does suggest that functionals of the phase - space density may be less non - local in time .
a psdft approach could be more successful than the usual tddft for some applications for which the usual functional approximations in tddft perform poorly .
in many strong - field applications , memory - dependence is required in the tddft xc potential , challenging to model , and additional observable functionals are needed , depending on the measured quantity of interest .
the examples worked through in section [ sec : examples ] demonstrate each of these aspects clearly .
being exactly - numerically solvable two - electron problems , one can perform a detailed analysis that helps to understand these challenges in their broader context .
the first example , a model of ionization , illustrated , via a phase - space exploration , that even though an exact ks treatment gets the exact coordinate - space density correct , it may do so with a wavefunction that is fundamentally different than the true wavefunction ; therefore approximating observables not directly related to the coordinate density by their value on the ks orbitals does not work well .
we considered the momentum distribution ( eg modelling that as measured by an ion - momentum recoil experiment ) , obtained from the ks orbitals and found spurious oscillations as a consequence of the single - determinantal nature of the ks wavefunction : a single delocalized orbital is describing the true two - orbital state .
the exact coordinate - space density is recovered ( by definition ) , but the momentum distribution displays oscillations , present also in the ks phase - space distribution .
the latter goes negative , a sign of non - classical behavior : the ks system describes the exact density putting both electrons in a doubly - occupied orbital , which must therefore be delocalized in two regions .
the true system is however a rather classical one , aside from the antisymmetry , and this is reflected in its phase - space density profile .
the second example explored the memory - dependence of the correlation potential in a time - dependent hooke s atom . solving directly for ( a density - weighted measure of ) the correlation potential ,
time slices were found in which the coordinate - density evolved identically , while the correlation potential varied dramatically , illustrating the fact that the correlation potential is a highly non - local in time functional of the coordinate - density .
we then asked whether , if momentum - distribution information was included in the basic variable , the correlation potential would be just as non - local in time as a functional of the phase - space density , or whether the phase - space density could distinguish the state of the system at these times .
we found the latter was true , suggesting that correlation functionals of the phase - space density are less non - local in time , and therefore easier to approximate than correlation functionals of the density alone .
in psdft , the basic variable is the one - body wigner phase - space density , from which momentum - space and position - space densities can be obtained via integration over the conjugate variable . formally equivalent to density - matrix functional theory ,
all one - body observables are directly obtained from the basic variable ( including the momentum distribution ) , without the need for additional observable functionals . in its equation of motion
the action of the kinetic energy operator is exact as a phase - space functional ( equally , density - matrix functional ) , while approximations are needed for the second - order wigner function as a functional of the first - order one .
an important future direction is to develop functional approximations for this that lead to changing occupation numbers , in order to treat many of the problems that are challenging in the usual tddft , such as electronic quantum control problems , and ionization dynamics . in light of the studies in this paper ,
this would be well worth pursuing . | we discuss two problems which are particularly challenging for approximations in time - dependent density functional theory ( tddft ) to capture : momentum - distributions in ionization processes , and memory - dependence in real - time dynamics .
we propose an extension of tddft to phase - space densities , discuss some formal aspects of such a `` phase - space density functional theory '' and explain why it could ameliorate the problems in both cases .
for each problem , a two - electron model system is exactly numerically solved and analysed in phase - space via the wigner function distribution . |
professor belyaev ( with v. g. zelevinskii ) ( bz ) pioneered the mapping of fermion systems onto bosons more than thirty years ago @xcite .
these original attempts at bosonization of fermion systems was motivated by collective particle - hole modes in nuclei . since that time the interacting boson model @xcite ( ibm ) has been phenomenologically very successful in explaining low energy nuclear spectroscopy for heavy nuclei .
the bosons in this model are thought to represent monopole ( j=0 ) , quadrupole ( j=2 ) , and sometimes hexadecapole ( j=4 ) correlated pairs of valence nucleons in the shell model .
the ibm hamiltonian is hermitian , usually has at most two - boson interactions , and conserves boson number , reflecting the particle - particle , rather than particle - hole , nature of the underlying fermion pairs .
while one can numerically diagonalize the general ibm hamiltonian , one of the strengths of ibm are the algebraic limits corresponding to the subgroups su(3 ) , u(5 ) , or o(6 ) , with analytic expressions for excitation bands and transition strengths , which encompass an enormous amount of nuclear data .
the microscopic reasons for the success of such a simple model are elusive .
otsuka , arima , and iachello , along with talmi , have used a mapping of the shell model hamiltonian to the ibm hamiltonian @xcite based on the seniority model @xcite , but these attempts have not done well for well - deformed nuclei .
for this reason we have revisited boson mappings to see if we can understand the success of the ibm starting from the shell model . in the next section
we sketch out various historic approaches to boson mappings @xcite .
we follow marumori @xcite and otsuka et al .
@xcite ( oai ) in our mapping procedure which maps fermion states into boson states and construct boson operators that reproduce fermion matrix elements .
we give the boson representation of the hamiltonian and review the result that , in the full space , it factorizes into a boson image , which is the same as the bz hamiltonian , times a normalization operator which projects out the spurious states .
however , since our goal is to understand the ibm which only deals with a few of the enormous degrees of freedom of the shell model , we go on to discuss boson images in truncated spaces .
this , we shall see , gives rigorous insight into the oai mapping and shows how to go systematically beyond it .
the fundamental goal is to solve the many - fermion schrdinger equation @xmath0 and find transition matrix elements between eigenstates , @xmath1 .
as the fermion fock space may be so large as to make direct solution intractable , the idea of a boson mapping is to replace the fermion operators with boson operators , _ using only a minimal number of boson degrees of freedom _ , that approximate the spectrum and transition matrix elements of the original fermion problem .
there are two approaches to boson mappings which we now review .
the first approach , epitomized by belyaev and zelevinskii@xcite , is to map fermion operators to boson operators so as to preserve the original algebra .
specifically , consider a space with @xmath2 single - fermion states ; @xmath3 signify fermion creation and annihilation operators .
the set of all bilinear femion operators , @xmath4 , form the lie algebra of @xmath5 , as embodied by the commutation relations @xmath6 = 0 \\
% \left [ a_i a_j , a^\dagger_k a^\dagger_l \right ] = \delta_{il } \delta_{jk } + \delta_{ik } a^\dagger_l a_j + \delta_{jl } a^\dagger_k a_i - ( i \leftrightarrow j ) \\ \left [ a_i a_j , a_k^\dagger a_l \right ] = \delta_{jk } a_i a_l - ( i \leftrightarrow j ) \\ \left [ a^\dagger_i a_j , a^\dagger_k a_l \right ] = \delta_{jk } a^\dagger_i a_l -\delta_{il } a^\dagger_k a_j\end{aligned}\ ] ] at this point it is convenient to introduce collective fermion pair operators @xmath7 we always choose the @xmath8 matrices @xmath9 to be antisymmetric to preserve the underlying fermion statistics , thus eliminating the need later on to distinguish between ` ideal ' and ` physical ' bosons .
we also assume the following normalization and completeness relations for the matrices : @xmath10 generic one- and two - body fermion operators we represent by @xmath11 , @xmath12 where @xmath13 ; from such operators one can construct a fermion hamiltonian @xmath14 .
now one has the following commutation relations : @xmath15 = \left [ \hat{a}^\dagger_\alpha , \hat{a}^\dagger_\beta \right ] = 0 ; \label{commute1 } \\ \left [ \hat{a}_\alpha , \hat{a}^\dagger_\beta \right ] = \delta_{\alpha \beta } -2 \sum_{ij } \left ( { \bf a}^\dagger_\beta { \bf a}_\alpha \right ) _ { ij } a^\dagger_i a_j ; \\ \left [ \hat{a}_\alpha , \hat{t } \right ] = 2\sum_\beta { \rm tr \ , } \left ( { \bf a}_\alpha { \bf t } { \bf a}^\dagger_\beta \right ) \hat{a}_\beta \\ \left [ \hat{t}_1 , \hat{t}_2 \right ] = \sum_{ij } \left ( { \bf t}_1{\bf t}_2 -{\bf t}_2{\bf t}_1 \right)_{ij } a^\dagger_i a_j \label{commutelast}\end{aligned}\ ] ] the method of belyaev and zelevinskii is to find boson images of the bifermion operators , @xmath16 where @xmath17 are boson creation and annihilation operators , @xmath18 = \delta_{\alpha \beta}$ ] , with the coefficients @xmath19 chosen so that the images @xmath20 have the same commutation relations as in ( [ commute1])-([commutelast ] ) . because the algebra is exactly matched , if one builds boson states in exact analogy to the fermion states then the full boson fock space is not spanned and one does not have nonphysical or spurious states . for the state - mapping methods described below , especially when the entire boson fock space is used , identifying and decoupling from spurious states is an important and problematic issue .
on the other hand , the belyaev - zelevinskii expansion is in general infinite . in the full boson fock space ,
that is , no truncation of the boson degrees of freedom , the image of one body operators is finite and given quite simply by @xmath21 .
since any fermion hamiltonian can be written in terms of one - body operators , the boson image of a finite fermion hamiltonian can be finite in the full space .
the states that one must use then are built from the pairs given in ( [ bzpair ] ) which will not just be products of bosons but will include exchange terms .
for example , for two bosons and using ( [ bzpair ] ) , @xmath22 these exchange terms are due to the antisymmetry .
we shall take care of such exchange effects by introducing a norm operator in the boson space .
for truncated spaces , however , the expansion of the bz hamiltonian is infinite .
the dyson mapping @xcite is a variant of belyaev - zelevenskii , in which one makes the mapping @xmath23 the operators are then clearly finite ; on the other hand they are just as clearly non - hermitian . from a computational viewpoint
non - hermiticity is only a minor barrier , but it is an obstacle to an understanding of the microscopic origin of hermitian ibm hamiltonians . furthermore the dyson operators mix spurious and physical spaces .
marshalek @xcite points out there exist mappings that are both finite and hermitian , but these in general require projection operators to eliminate spurious states .
we will regain this result later on in this paper .
the second major approach , pioneered by marumori @xcite , is to map fermion states and construct boson operators that preserve matrix elements . for fermion
many - body ( shell - model ) basis states one often uses slater determinants , antisymmetrized products of single - fermion wavefunctions which we can write using fock creation operators : @xmath24 for @xmath25 fermions . for an even number of fermions one
instead constructs states from products of fermion pairs , @xmath26 if the number of fermion is fixed at @xmath25 then @xmath27 runs from 1 to @xmath28 . the original work of marumori , however , focused on particle - hole excitations and so the number of pairs and consequently bosons is not fixed .
these states are not trivially orthonormal
. they must be orthogonalized .
exactly how this orthogonalization is accomplished will be a key theme in this paper .
marumori constructs the norm matrix @xmath29 and then the usui operator @xmath30 where bosons states are constructed in strict analogy to the fermion states , @xmath31 then the marumori expansion of any fermion operator is @xmath32 clearly marumori is best suited for particle - hole states with only a few excitations .
if one applied it to a system with numerous particle - particle pairs , as for the ibm , one obtains clumsy many - body terms .
kishimoto and tamura @xcite addressed this last issued by introducing a `` linked - cluster '' expansion which they then grafted into a bz - type scheme .
otsuka , arima , and iachello ( oai ) , along with talmi @xcite , investigated the microscopic origins of the interacting boson model through boson mappings .
although they also mapped states , they differed from marumori in some key details .
first of all , they built states built on a fixed number of particle - particle , not particle - hole , pairs , and restricted the pairs to one monopole ( @xmath33 ) and quadrupole ( @xmath34 ) pair .
these states were orthogonalized based on seniority .
that is , they construct , for @xmath35 fermions , low - seniority basis states of @xmath36 and @xmath37 fermion pairs , @xmath38 , and then orthonormalize the states such that the zero - seniority state is mapped to itself , and states of higher seniority @xmath39 are orthogonalized against states of lower seniority , @xmath40 then oai calculate the matrix elements @xmath41 for @xmath42 and obtain the coefficients for their one plus two - boson hamiltonian .
these coefficients have an explicit @xmath43-dependence ( and for large @xmath43 and arbitrary systems such matrix elements are not trivial to calculate , especially in analytic form ! ) and thus implicitly a many - body dependence . at first sight
this is not entirely unreasonable as it is well known the ibm parameters change substantially as a function of the number of bosons , even within a major ` shell ' .
nonetheless the oai mapping has three drawbacks .
the first is that it s not clear how to systematically calculate many - body contributions beyond that contained in the oai prescription , whereas the method we shall describe is fully and rigorously systematic .
the second is that the oai prescription can induce many - body effects where none are needed .
this point will be illustrated in section [ approximages ] .
thirdly , only the @xmath44 space is exactly mapped , but very deformed systems will involve large @xmath45 .
in fact , for an axial rotor limit , the average number of d - bosons in the ground state band is 2/3 the total number of bosons . as an alternative to oai , skouras , van isacker , and nagarajan @xcite proposed a `` democratic '' mapping where the orthogonalization is based on eigenvectors of the norm matrix rather than seniority . in what follows we attempt to rigorously unify all the state - mapping methods .
we have three strong results .
first , we give general expressions for fermion matrix elements via boson representations .
second , we show how in several cases one can have exact , finite , and hermitian boson images of fermion operators .
finally , we show how to extend both the oai and democratic mappings in a systematic and rigorous fashion , and illustrate how the choice of orthogonalization can affect the many - body dependence of the boson images .
the starting point of any state - mapping method is the calculation of matrix elements of fermion operators between states constructed from fermion pairs of the form ( [ statedefn ] ) , including the overlap : @xmath46 , @xmath47 , @xmath48 , and so on .
these matrix elements are much more difficult to compute than the corresponding matrix elements between slater determinants .
as we shall show , however , full and careful attention paid to the problem of calculation matrix elements can yield powerful results .
silvestre - brac and piepenbring @xcite , laboriously using commutation relations , derived a wick theorem for fermion pairs .
rowe , song and chen @xcite using ` vector coherent states ' ( we would say fermion - pair coherent states ) found matrix elements between pair - condensate wavefunctions , states of the form @xmath49 using a theorem by lang et al .
@xcite , we have generalized @xcite the method of rowe , song and chen and recovered ( actually discovered independently ) the expressions of silvestre - brac and piepenbring .
specifically , we construct generating functionals by taking the matrix element @xmath50^k \right ) .
\label{generating}\end{aligned}\ ] ] by taking derivatives of @xmath51 , etc .
, one computes the desired matrix elements in analytic form@xcite . for pair condensate wave functions one can calculate the matrix elements iteratively and propose a variational principle @xcite
. such a variational principle would be useful in determining the `` best '' microscopic structure for a truncated set of pairs .
alternately , otsuka and yoshinaga @xcite choose their @xmath36 and @xmath37 pairs from hartree - fock - bogoliubov states ; the two approaches can probably be related in some approximation .
this may be important is answering a basic question of ibm , the origin of algebraic limits : do they arise from changes in pair structure , or from effective many - body effects , or both ? we now want to translate the fermion matrix elements into boson space .
we take the simple mapping of fermion states into boson states @xmath52 where the @xmath53 are boson creation operators .
we construct boson operators that preserve matrix elements , introducing boson operators @xmath54 , @xmath55 , and most importantly the _ norm operator _ @xmath56 such that @xmath57 , @xmath58 and @xmath59 .
we term @xmath54,@xmath55 the boson _
representations _ of the fermion operators @xmath60 , @xmath61 .
one finds the ` linked - cluster ' ( a la kishimoto and tamura@xcite although with differences ) expansion of the representations to be of the form @xcite @xmath62 and similarly for @xmath63 . in the norm operator
the @xmath64-body terms embody the fact that the fermion - pair operators do not have exactly bosonic commutation relations , and act to enforce the pauli principle .
the coefficients @xmath65 etc . can be written in closed , albeit complicated , form @xcite
. the norm operator can be conveniently and compactly expressed @xcite in terms of the @xmath66th order casimir operators of the unitary group @xmath67 , @xmath68 , @xmath69 ( and so is both a matrix and a boson operator ; the trace is over the matrix indices and not the boson fock space ) @xmath70 where the colons ` : ' refer to normal - ordering of the boson operators .
this norm operator takes into account the exchange terms in the bz expansion of a fermion pair given in ( [ bzpair ] ) .
similarly and this is a new result we have not seen elsewhere in the literature the representations @xmath71 can also be written in compact form@xcite : @xmath72 b^\dagger_\sigma b_\tau \hat { \cal n}_b \colon \label{onebodyimage } \\
v}_b & = & \sum_{\mu , \nu } \left \langle \mu \left | v \right | \nu \right \rangle \sum_{\sigma , \tau } \colon \left \ { { \rm tr \ , } \left [ { \bf a}_\sigma { \bf a}^\dagger_\mu { \bf g } \right ] { \rm tr \ , } \left [ { \bf a}_\nu { \bf a}^\dagger_\tau { \bf g } \right ] \right .
\hspace*{1.00 in } & + & \left.4 \,{\rm tr \ , } \left [ { \bf a}_\sigma { \bf a}^\dagger_\mu { \bf p g a}_\mu { \bf a}^\dagger_\tau { \bf g } \right ] \right \ } b^\dagger_\sigma b_\tau \hat { \cal n}_b \colon , \label{twobodyimage}\end{aligned}\ ] ] where @xmath73 .
these compact forms are useful for formal manipulation .
furthermore they have the powerful property of exactly expressing the fermion matrix elements under _ any _ truncation , a fact not previously appreciated in the literature even for the norm operator @xcite . by this
we mean the following : suppose we truncate our fermion fock space to states constructed from a restricted set of pairs @xmath74 .
such a truncation need _ not _ correspond to any subalgebra .
then the representations in the corresponding truncated boson space , which still exactly reproduce the fermion matrix elements and which we denote by @xmath75_t$ ] etc . , are the same as those given above , retaining only the ` allowed ' bosons with unrenormalized coefficients .
for example @xmath76_t = \colon \exp \left(-\frac{1}{2 } \sum^{\infty}_{k=2 } \frac{(-1)^k}{k } \left[\hat{c}_k\right]_t\right):\ ] ] where @xmath77_t = 2 \colon { \rm tr \ , } ( \left [ { \bf p}\right ] _ t ) ^k \colon \ , , \ , \left [ { \bf p}\right ] _ t = \sum_{\bar{\sigma}\bar{\tau } } b^{\dagger}_{\bar{\sigma } } b_{\bar{\tau } } { \bf a}_{\bar{\sigma } } { \bf a}^{\dagger}_{\bar{\tau}}.\ ] ] this invariance of the coefficients under truncation will not hold true for the boson _
images _ introduced below . with the boson representations of fermion operators in hand
, one can express the fermion schrdinger equation ( [ schrodinger1 ] ) with @xmath78 as a generalized boson eigenvalue equation , @xmath79 here @xmath80 is the boson representation of the fermion hamiltonian .
every physical fermion eigenstate in ( [ schrodinger1 ] ) has a corresponding eigenstate , with the same eigenvalue , in ( [ schrodinger2 ] ) .
because the space of states constructed from pairs of fermions is overcomplete , there also exist spurious boson states that do not correspond to unique physical fermion states .
these spurious states will have zero eigenvalues and so can be identified .
the overcompleteness also means that ( [ schrodinger2 ] ) is harder to solve exactly than ( [ schrodinger1 ] ) .
so one truncates the model space .
in general the boson representations given in ( [ normexpand ] ) , ( [ onebodyimage ] ) and ( [ twobodyimage ] ) do not have good convergence properties , so that simple termination of the series such as ( [ normexpand ] ) in @xmath64-body terms is impossible and use of the generalized eigenvalue equation ( [ schrodinger2 ] ) , as written , is problematic .
instead we `` divide out '' the norm operator to obtain the _ boson image _
, i.e. schematically , @xmath81 that this is reasonable is suggested by the explicit forms of ( [ onebodyimage ] ) and ( [ twobodyimage ] ) .
the hope of course is that @xmath82 is finite or nearly so , so that a 1 + 2-body fermion hamiltonian is mapped to an image @xmath83 with the @xmath64-body terms , @xmath84 , zero or greatly suppressed .
we now discuss how this `` dividing out '' is to be carried out .
it turns out that for a number of cases the image of the hamiltonian is exactly finite . in particular , for the full boson fock space the representations factor in a simple way : @xmath85 and @xmath86 , where the factored operators @xmath87 , which we term the boson images of @xmath88 , have simple form @xcite : @xmath89 @xmath90\ ] ] this image hamiltonian @xmath91 is the one determined by bz if one decomposes the hamiltonian into multipole - multipole form and then maps these mulitpole operators . as discussed earlier ,
these bz multipole operators are finite in the full space .
this result , and its relation to other mappings , was noted by marshalek @xcite thus any boson representation of a hamiltonian factorizes : @xmath92 in the full space . since the norm operator is a function of the @xmath67 casimir operators it commutes with the boson images of fermion operators @xcite , and one can simultaneously diagonalize both @xmath80 and @xmath56 .
then eqn .
( [ schrodinger2 ] ) becomes @xmath93 where @xmath94 for the physical states , but @xmath95 for the spurious states is no longer necessarily zero .
the boson hamiltonian @xmath96 is by construction hermitian and , if one starts with at most only two - body interactions between fermions , has at most two - body boson interactions .
all physical eigenstates of the original fermion hamiltonian will have counterparts in ( [ schrodinger3 ] ) .
it should be clear that transition amplitudes between physical eigenstates will be preserved .
spurious states will also exist but , since the norm operator @xmath56 commutes with the boson image hamiltonian @xmath96 , the physical eigenstates and the spurious states will not admix . also the spurious states can be identified because , while they will no longer have zero energy eigenvalues , they will have eigenvalue zero with respect to the norm operator .
the boson schrdinger equation ( [ schrodinger3 ] ) , though finite , is not much use as the boson fock space is still much larger than the original fermion fock space , and we still must truncate the boson fock space . although the representations remain exact under truncation , the factorization into the image does not persist in general : @xmath97_t \neq \left [ \hat{\cal n}_b \right ] _
t \left [ \hat{h}_b \right ] _
this was recognized by marshalek @xcite .
( an alternate formulation @xcite does not require the complete fock space , but mixes physical and spurious states and so always requires a projection operator . ) if the truncation scheme represents a closed subalgebra ( specifically , if the truncated set of fermion pairs are closed under double commutations ) then a factorization @xcite @xmath98_t = \left [ \hat{\cal n}_b \right ] _
t \hat{h}_d = \hat{h}_d^\dagger \left [ \hat{\cal n}_b \right ] _
t \label{dysonfactor}\ ] ] _ does _ exist , with @xmath99 at most two - body , but not necessarily hermitian .
we term it a _ dyson _ image @xcite . under more restricted conditions on the structure of the pairs and
the hamiltonian one can guarantee @xmath99 is hermitian and commutes with @xmath100_t$ ] . in the full space , of course ,
all definitions of boson images coincide and yield the same result .
first , consider a partition of the single fermion states labeled by @xmath101 , where the dimension of each subspace is @xmath102 , @xmath103 so that @xmath104 .
we denote the amplitudes for the truncated space as @xmath105 and assume they can be factored , @xmath106 , with @xmath107 and @xmath108 , where @xmath109 ( symmetric ) or @xmath110 ( antisymmetric ) .
furthermore we assume the completeness relation ( [ completeness ] ) , which was crucial for proving that @xmath92 @xcite , is valid for the truncated space ; i.e. , @xmath111 . \label{newcomplete}\ ] ] the norm operator in the truncated space then becomes @xmath112_t = \colon { \rm exp } \sum_{k = 2 } \left(\frac{-1}{\omega_a}\right)^{k-1 } \frac{1}{k } { \rm tr } ( \bar{\bf p}^k ) \colon , \ ] ] where @xmath113 so that @xmath114_t = \left(\frac{1}{2 \omega_{a}}\right)\bar{\bf p}$ ] . in this case
the boson image of a one - body operator is the truncation of the boson image in the full space , @xmath115_t = \left [ \hat{\cal n}_b \right ]
_ t \left [ \hat { t}_b \right ] _ t\ ] ] @xmath116_t = 2\sum _ { \bar{\sigma } , \bar{\tau } } { \rm tr \ , } \left ( { \bf a}_{\bar{\sigma } } { \bf t } { \bf a}^\dagger_{\bar{\tau } } \right ) b^\dagger_{\bar{\sigma } } b_{\bar{\tau}}.\ ] ] the representation of a two - body interaction can be factored into a boson image times the truncated norm , @xmath117_t = \left [ \hat{\cal n}_b \right ] _
t \hat{v}_d ; \label{dysonimage2}\ ] ] however , @xmath118 , while finite ( 1 + 2-body ) , is not simply related to @xmath119_t$ ] as is the case for one - body operators .
if one writes @xmath120 then matrix elements of the two - boson interaction are @xmath121 \nonumber \\ - \omega_a ( 2\omega_a + ( -1)^p ) { \rm tr}_c ( { \bf a}_{\nu } { \bf k } \bar{\bf a}^\dagger_{\bar{\tau } } \bar{\bf a}_{\bar{\sigma}^{\prime } } \bar{\bf a}^\dagger_{\bar{\tau}^{\prime } } ) \delta_{\bar{\sigma } , \mu } \rbrace .\end{aligned}\ ] ] upon inspection one sees the image is not constrained to be hermitian .
consider the additional condition between the matrix elements of the interaction : @xmath122 where the factor @xmath123 is the number of pairs in the excluded subspace . while condition ( [ hermcondition ] ) looks complicated there are interactions that satisfy it ; for example , two - body interactions constructed from one - body operators @xmath124 where @xmath125 $ ] .
when ( [ hermcondition ] ) is satisfied then @xmath118 is hermitian and although @xmath126_t$ ] they are simply related : @xmath127 with @xmath128 renormalizing the two - boson part of @xmath129_t$ ] by a factor which ranges from unity ( full space ) to 2 for a very small subspace .
not all interactions satisfy ( [ hermcondition ] ) ; for example , the pairing interaction never does except in the full space . for the pairing interaction @xmath130 , and @xmath131 , and the image ( [ dysonimage2 ] )
@xmath132 becomes ( remembering @xmath104 ) @xmath133 + \sum_{\bar{\tau } \bar{\tau}^\prime \neq 0 , \bar{\sigma } } { \rm tr}\ ; ( \bar{{\bf a}}_{\bar{\sigma } } \bar{{\bf a}}^\dagger_{\bar{\tau } }
\bar{{\bf a}}_0 \bar{{\bf a}}_{\bar{\tau}^\prime}^\dagger ) b^\dagger_{\bar{\sigma } } b^\dagger_0 b_{\bar{\tau } } b_{\bar{\tau}^\prime } \right \rbrace , \label{pairingimage}\ ] ] where @xmath134 is the total number of bosons , @xmath135 , and @xmath136 .
the second term in the above is not hermitian but can be transformed away by a similarity transformation @xcite , leaving the first term as a finite hermitian image which gives the correct eigenvalues for all n. the so(8 ) and sp(6 ) models @xcite belong to a class of models which have a subspace for which ( [ newcomplete ] ) is valid and interactions which satisfy ( [ hermcondition ] ) . in these models
the shell model orbitals have a definite angular momentum @xmath137 and are partitioned into a pseudo orbital angular momentum @xmath138 and pseudospin @xmath139 , @xmath140 .
the amplitudes are then given as products of clebsch - gordon coefficients , @xmath141 where @xmath142 and @xmath143 are the total pseudo orbital angular momentum and pseudospin respectively of the pair of nucleons .
for the so(8 ) model @xmath144 and one considers the subspace of pairs with @xmath145 @xmath146 , @xmath147 ; in the sp(6 ) model @xmath148 and one considers the subspace with @xmath149 @xmath150 , @xmath151 .
the complicated conditions ( [ hermcondition ] ) hold true for important cases , for example , the quadrupole - quadrupole and other multipole - multipole interactions in the so(8 ) and sp(6 ) models ( that is , interactions of the generic form @xmath152 in the notation of @xcite ) have hermitian dyson images .
not all interactions in these models have hermitian dyson images . for example , pairing in any model ( see ( [ pairingimage ] ) ) and , in the so(8 ) model , the particular combination @xmath153 which is the so(7 ) limit .
it so happens that these particular cases nonetheless can be brought into finite , hermitian form as discussed in the next section .
the most general image hamiltonian one can define is @xmath154_t^{-1/2 } \left [ \hat{{\cal h \rm}}_b \right ] _
t \left [ \tilde{\cal n}_b \right ] _
t^{-1/2 } { \cal u}^\dagger , \label{himage}\ ] ] which is manifestly hermitian for any truncation scheme and any interaction , with @xmath155 a unitary operator .
( because the norm is a singular operator it can not be inverted .
instead @xmath156_t^{-1/2}$ ] is calculated from the norm only in the physical subspace , with the zero eigenvalues which annihilate the spurious states retained .
then @xmath157 does not mix physical and spurious states . ) if @xmath158 , this is the democratic mapping @xcite .
again , for the full space @xmath159 . this prescription is , we argue , useful for a practical derivation of boson image hamiltonians . ignoring for the moment the unitary transformation @xmath155 ,
consider the expansion ( [ htruncate ] ) of @xmath157 .
the operators @xmath160_t$ ] and @xmath156_t^{-1/2}$ ] have similar expansions , and by multiplying out ( [ himage ] ) one sees immediately that the coefficient @xmath161 depends only on up to @xmath64-body terms in @xmath162_t$ ] and @xmath156_t^{-1/2}$ ] , derived from @xmath163-fermion matrix elements which are tractable for @xmath64 small .
ideally @xmath157 would have at most two - body terms , and our success in finding finite images in the previous section gives us hope that the high - order many - body terms may be small ; at any rate the convergence can be calculated and checked term - by - term .
specifically , consider the convergence of the series ( [ htruncate ] ) as a function of @xmath64 .
a rough estimate is that , for an @xmath43-boson fock space , one can truncate to the @xmath64-body terms if for @xmath164 , @xmath165 is sufficiently small compared to @xmath166 ; the strictest condition is to require @xmath167 .
the hermitian image @xmath157 , defined in ( [ himage ] ) , is related to the dyson image @xmath99 , defined in ( [ dysonfactor ] ) , by a similarity transformation @xmath168_t^{1/2}$ ] , @xmath169 the similarity transformation @xmath170 orthogonalizes the fermion states @xmath171 inasmuch @xmath172 in the physical space ( and @xmath173 in the spurious space ) .
this is akin to gram - schmidt orthogonalization and the freedom to choose @xmath155 , and @xmath170 , corresponds to the freedom one has in ordering the vectors to be gram - schmidt orthogonalized .
the oai and democratic mappings are just two particular choices out of many ; the latter takes @xmath174 .
we can use the freedom in the choice of @xmath155 to our advantage . consider the so(8 ) model @xcite and its three algebraic limits : the pure pairing interaction , the quadrupole @xmath175 interaction , which can be written in terms of so(6 ) casimir operators , and the linear combination of pairing and quadrupole @xmath176 which can be written in terms of so(7 ) casimirs ( see @xcite for details and notation ) . as discussed in the last section ,
the dyson image of the quadrupole interaction is hermitian and finite , and @xmath177 with @xmath158 .
the dyson images of the pairing and so(7 ) interactions are finite but non - hermitian .
we have found @xmath155 s @xmath178 for both these cases ( but not the same @xmath155 ) such that their respective hermitian images @xmath157 are finite ; the one for pairing is exactly the oai prescription , while that for so(7 ) is exactly opposite , orthogonalizing states of low seniority against states of higher seniority .
these general hermitian images do not have a simple relation to the truncated full image , as do the hermitian dyson image ( [ hermdysonimage ] ) .
the one - body piece remains unchanged but there can be significant renormalization , and even change of sign , of the two - body piece .
for example , for the pairing interaction @xmath179_t = s^{\dagger } s + { 1\over 2 \omega^2 } s^{\dagger } s^{\dagger } s s+$ ] additional terms , including off - diagonal terms such as @xmath180 , whereas @xmath181 ( depending on the truncation ) terms such as @xmath182 but _ no _ off - diagonal terms .
hence we see here the renormalization is not just a simple overall factor @xmath183 as it was for the hermitian dyson for the @xmath184 term but @xmath185 for terms such as @xmath186 if one uses an `` inappropriate '' transform @xmath187 it can induce an unneeded and unwanted many - body dependence .
this principle we illustrate in the so(8 ) model with the so(7 ) interaction , whose spectrum is exactly known and for which we can derive a finite hermitian image with no many - body dependence ; this is the left - hand spectrum in figure 1 . for the right - hand side we took the @xmath155 appropriate for pairing , that is the oai prescription determined for @xmath188 , and calculated the spectrum for @xmath189 keeping only the strict two - body terms .
the distortion in the spectrum from the exact result , such as the overall energy shift and the large perturbation in the third band , indicates missing many - body terms .
that is , if one mapped the so(7 ) interaction using the canonical oai procedure one would find of necessity a many - body dependence in the interaction coefficients . by orthogonalizing the basis in the a different way , however , as expressed by a different choice of @xmath155 , the many - body dependent vanishes .
therefore it is possible that some of the @xmath43-dependence of oai is induced by their choice of orthogonalization and could be minimized with a different choice .
we are currently exploring how to exploit this freedom to best effect .
in order to investigate rigorous foundations for the phenomenological interacting boson model , we have presented a rigorous microscopic mapping of fermion pairs to bosons , paying special attention to exact mapping of matrix elements , hermiticity , truncation of the model space , and many - body terms .
first we presented new , general and compact forms for boson _ representations _ that preserve fermion matrix elements .
we then considered the boson _ image _ hamiltonian which results from `` dividing out '' the norm from the representation ; in the full boson fock space the image is always finite and hermitian ; in addition we discussed several analytic cases for truncated spaces where the image is also finite and hermitian .
finally , we give a prescription which is a generalization of both the oai and democratic mappings ; in the most general case for truncated spaces the hermitian image hamiltonian may not be finite but we have demonstrated there is some freedom in the mapping that one could possibly exploit to minimize the many - body terms .
this freedom , which manifests itself in a similarity transformation that orders the orthogonalization of the underlying fermion basis , depends on the hamiltonian .
999 s.t .
belyaev and v.g .
zelevinskii , _ nucl .
* 39 * ( 1962 ) 582 .
f. iachello and a. arima , _ the interacting boson model _ ( cambridge university press , 1987 ) .
t. otsuka , a. arima , f. iachello and i. talmi , _ phys .
lett._*76b * ( 1978 ) 139 .
t. otsuka , a. arima , and f. iachello , _ nucl .
phys . _ * a309 * ( 1978 ) 1 .
i. talmi , _ simple models of complex nuclei _ , ( harwood academic publishers , 1994 ) .
a. klein and e. r. marshalek , _ rev .
* 63 * ( 1991 ) 375 ; p. ring and p. schuck , _ the nuclear many - body problem _ ( springer - verlag , 1980 ) . t. marumori , m. yamamura , and a. tokunaga , _ prog . theor . phys . _
* 31 * ( 1964 ) 1009 ; t. marumori , m. yamamura , a. tokunaga , and a.takeda , _ prog .
theor . phys . _
* 32 * ( 1964 ) 726 .
dyson _ phys .
* 102 * ( 1956 ) 1217 .
e. r. marshalek , _ phys .
c _ * 38 * ( 1988 ) 2961 .
t. kishimoto and t. tamura , _ phys .
c _ * 27 * ( 1983 ) 341 ; h. sakamoto and t. kishimoto , _ nucl .
phys . _ * a486 * ( 1988 ) 1 . l. d. skouras , p. van isacker , and m. a. nagarajan , _ nucl .
phys . _ * a516 * ( 1990 ) 255 . b. silvestre - brac and r. piepenbring _ phys .
c _ * 26 * ( 1982 ) 2640 .
rowe , t. song and h. chen , _ phys .
c _ * 44*(1991 ) r598 .
lang , c.w .
johnson , s.e .
koonin , and w.e .
ormand , _ phys .
c _ * 48 * ( 1993 ) 1518 .
ginocchio and c.w .
johnson , to be published .
t. otsuka and n. yoshinaga , _ phys .
_ * 168b * ( 1986 ) 1 .
j. dobaczewski , h.b .
geyer , and f.j.w .
hahne , _ phys .
c _ * 44 * ( 1991 ) 1030 .
johnson and j.n .
ginocchio , in press , _ phys . rev .
c _ , august 1994 .
ginocchio and c.w .
johnson , to be published in `` frontiers of nuclear structure physics '' , t. otsuka , ed .
( world scientific , 1994 ) .
ginocchio and i. talmi , _ nucl .
phys . _ * a337 * ( 1980 ) 431 .
ginocchio , _ ann .
* 126 * ( 1980 ) 234 .
o. scholten in _ computational nuclear physics 1 _ , k. langanke , j.a .
maruhn and s.e .
koonin , eds . , ( springer - verlag 1991 ) chapter 5 . | we briefly review various mappings of fermion pairs to bosons , including those based on mapping operators , such as belyaev - zelevinskii , and those on mapping states , such as marumori ; in particular we consider the work of otsuka - arima - iachello , aimed at deriving the interacting boson model .
we then give a rigorous and unified description of state - mapping procedures which allows one to systematically go beyond otsuka - arima - iachello and related approaches , along with several exact results . |
is a small irregular galaxy in the local group ( see hodge 1994 for a brief history and references ) .
it is rich in newly formed massive blue stars and is peppered with over 140 h ii regions ( hodge & lee 1990 ) .
recently , massey , armandroff & conti ( 1992 ) and massey & armandroff ( 1995 ) have identified 15 wolf - rayet stars in .
since wolf - rayet stars are descended from only the most massive ( @xmath23040 m@xmath3 ) stars , this argues that has an unusually large unevolved massive ( ob ) star population as well , with a _ galaxy - wide _ surface density of massive stars that is higher by a factor of 2 than any other local group galaxy .
the galaxy - wide massive star density of is comparable to that observed in isolated regions of recent star formation in other local group galaxies , in accord with earlier suggestions that is currently undergoing a starburst .
the radio continuum emission from is strong ( e.g. klein & gr " ave 1986 ) .
of particular relevance to this work is the fact that yang & skillman ( 1993 ; hereafter ys93 ) have discovered a non - thermal bubble of radio emission in .
this bubble has a diameter of about 250 pc , and comparison with data at other wavelengths shows that it is associated with a region of star formation as well as the densest h i concentration in .
ys93 argue that the bubble is a collection of supernova remnants ( i.e. a ` superbubble ' ) .
the 408 mhz radio luminosity of the bubble is about 2.6 times that of n132d , one of the brightest supernova remnants ( snr ) in the large magellanic cloud .
also has two h@xmath4o maser complexes which include a flaring megamaser as well as a maser with intraday variability ( e.g. argon et al .
1994 ; baan & haschick 1994 ) . due to
the fact that is the closest starburst galaxy to our own galaxy and has interesting properties at a variety of wavelengths , we performed a deep pointing towards it with the hri detector ( david et al .
1995 ) on ( tr " umper 1983 ) .
our main goals were to look for x - ray binaries , strong snr , evidence for a hot interstellar medium and any x - ray emission near the maser complexes .
this is the first pointed x - ray observation of to our knowledge .
distance estimates to range from 0.73 mpc , although a consensus is developing that the distance is @xmath5 mpc ( see massey & armandroff 1995 ) .
we shall adopt this distance , for which the hri spatial resolution of @xmath6 arcsec corresponds to @xmath7 pc .
the galactic column density towards is high due to its low galactic latitude , and almost certainly has significant intrinsic absorption ( see below ) . from the data of stark et al .
( 1992 ) , we obtain a galactic column density of @xmath8 @xmath9 .
this column blocks almost all photons with energies @xmath10 kev .
ngc 147 and ngc 185 are a pair of dwarf elliptical galaxies that are about 100 kpc from m31 .
ngc 147 has a predominantly old stellar population ( @xmath11 gyr ) and 4 known globular clusters .
its h i content is low .
most of the stars in ngc 185 are also old , but in addition it has ob stars , dust clouds and a fairly massive amount of h i in its interstellar medium .
it probably contains a snr ( gallagher , hunter & mould 1984 ) , and it has 6 known globular clusters . unlike ngc 205 , ngc 185 is too far from m31 to have had its star formation triggered by a recent interaction with this galaxy .
hodge ( 1994 ) gives further details about these galaxies and references to the literature .
helfand ( 1984 ) suggested that each of these galaxies might have @xmath12 x - ray binaries ( see his table 1 ) , so we performed hri pointings towards them .
we also wanted to look for any evidence of a hot galactic wind , as discussed in section ivb of ford , jacoby & jenner ( 1977 ) .
for these galaxies we adopt a distance of 600 kpc and a galactic column of @xmath13 @xmath9 ( stark et al .
[ [ section ] ] was observed with the hri starting on 1996 january 18 ( rh600902 : total raw exposure of 32.7 ks spread over 10.8 days ) .
the observation was performed in the standard ` wobble ' mode , and reduction and analysis of the resulting data were performed with the starlink asterix x - ray data processing system . in figure 1
we show contours of part of the full band ( @xmath140.12.5 kev ) hri image overlaid on the image of from the palomar schmidt @xmath15 plate [ see minkowski & abell 1963 for more information on the optical image ] .
we also show the positions of interesting objects in identified at other wavelengths ( see the figure caption ) .
we have checked the astrometry of the hri image using a bright star that is also the brightest x - ray source in the field , and it appears to be nominal ( see figure 2 ) .
this star is identified as tv cas ( hd 1486 ; @xmath16 ; orbital period of 1.81 days ) , a well - studied , algol - type , eclipsing binary ( see khalesseh & hill 1992 and references therein ) .
in addition , we have also detected the w uma - type binary bh cas ( @xmath17 ; orbital period of @xmath18 days ; see metcalfe 1995 and references therein ) .
the x - ray and simbad database positions for bh cas agree to within 3 arcsec , further confirming the hri astrometry .
these are the first x - ray detections of tv cas and bh cas to our knowledge , and we give their x - ray details in appendix a. _ basic observed x - ray properties _ 0.1 in the hri image shows a clear detection of a pointlike source coincident with ( hereafter referred to as x-1 ) , and we have used the asterix point source searching program pss ( allan 1995 ) to quantify the source properties .
the centroid position of x-1 is @xmath19 00@xmath20 20@xmath21 29.0@xmath22 , @xmath23 @[email protected]@xmath26 , with an error radius ( statistical plus systematic ) of approximately 5 arcsec .
the source has @xmath27 counts ( after background subtraction ) , giving a mean hri count rate of @xmath28 count s@xmath1 .
the detection statistical significance is @xmath29 .
it is possible to use pimms ( mukai 1995 ) and a spectral model to convert the observed mean hri count rate into a mean source flux .
we adopt a power - law model with a photon index of @xmath30 .
the galactic column density is @xmath31 @xmath9 , and if we adopt this column density we obtain a power - law normalization of @xmath32 photons @xmath9 s@xmath1 kev@xmath1 at 1 kev and a 0.12.5 kev absorbed flux of @xmath33 erg @xmath9 s@xmath1 .
the 0.12.5 kev absorbed flux is not a strong function of the adopted column density .
for example , if we use a much lower column of @xmath34 @xmath9 , then we obtain a power law normalization of @xmath35 photons @xmath9 s@xmath1 kev@xmath1 at 1 kev and a 0.12.5 kev absorbed flux of @xmath36 erg @xmath9 s@xmath1 .
we estimate absorption - corrected fluxes and luminosities below , after we have considered how likely it is that x-1 is indeed associated with .
we have searched the literature for sources at other wavelengths that are coincident with the detected x - ray source , and we find that the x - ray source lies in the middle of a hive of activity in . ys93 quote an approximate central position for their non - thermal radio superbubble of @xmath19 00@xmath20 20@xmath21 28@xmath22 , @xmath23 @xmath2416@xmath2549@xmath26 , and x-1 is within 7.5 arcsec of this position ( see figure 3 ) .
it is within 4.1 arcsec of the strongest ` snr candidate ' listed in table 3 of ys93 .
the h ii region # 113 from hodge & lee ( 1990 ) has a central position of @xmath19 00@xmath20 20@xmath21 27.7@xmath22 , @xmath23 @xmath2416@xmath2550@xmath26 ( after making the correction described in ohta et al .
1992 ) , which is about 10 arcsec from the hri centroid position .
this h ii region is extended and the catalogued h ii regions in this area are underlain by widespread diffuse emission .
in addition , the peculiar star wr17 from massey & armandroff ( 1995 ) is located at @xmath19 00@xmath20 20@xmath21 29@xmath22 , @xmath23 @xmath2416@xmath2552@xmath26 , which is within 2 arcsec of the x - ray centroid .
massey & armandroff ( 1995 ) give an optical spectrum of this star and classify it as a wn type wolf - rayet star based on the detection of a broad feature which they identify as he ii 4686 .
they find a @xmath37 magnitude of 21.76 and @xmath38 .
using the 3.5-m telescope of the apache point observatory , we have performed independent optical spectroscopy which confirms that wr17 is an emission - line star of some type .
the h@xmath39 and [ n ii ] emission lines of wr17 are systematically shifted by @xmath40 km s@xmath1 to the red compared to those of the aforementioned diffuse emission .
this rules out the possibility that wr17 is a normal supergiant star with emission lines superimposed on it from the diffuse emission .
the measured velocity for wr17 is @xmath41 km s@xmath1 , and this agrees better with that of the surrounding h i than that of the h ii .
wr18 is also located within 6 arcsec of the hri centroid , although massey & armandroff ( 1995 ) state that this is probably not a wolf - rayet star .
our independent spectroscopy suggests that wr18 may be an ordinary supergiant star with emission lines superimposed on it from the h ii region .
0.1 in _ confusing foreground and background sources _ 0.1 in is at low galactic latitude so confusing foreground sources are a natural concern .
based on results from the galactic plane survey ( see section 9.1 of motch et al .
1997 ) , we expect about 0.9 sources per square degree with count rates as large as or larger than that of x-1 .
has an angular size of about @xmath42 arcmin@xmath43 , so we estimate that there is less than about a 1 per cent chance of having a confusing foreground source this bright in front of
. furthermore , the hri source is found at a position where one might expect x - ray activity in based on its emission in other wavebands .
for example , the x - ray source is almost exactly in the centre of the ys93 non - thermal superbubble ( see ys93 for good arguments , but not rigorous proof , that the superbubble itself is in ) and near a region of recent star formation .
the superbubble has an angular size of about @xmath44 arcmin@xmath43 , so there is less than about a 0.02 per cent chance of having a confusing foreground source as bright as x-1 in front of it .
we can further argue against some types of possible foreground sources based on examination of optical images .
for example , we can use the nomograph shown in figure 1 of maccacaro et al .
( 1988 ) which relates the expected 0.33.5 kev x - ray flux to the visual magnitude for various types of astronomical objects .
if we use the power - law model with a column of @xmath34 @xmath9 ( see above ) , we predict a 0.33.5 kev flux of @xmath45 erg @xmath9 s@xmath1 .
there are no optical stars in the x - ray error circle with @xmath37 magnitudes brighter than 20.8 ( based on analysis of a @xmath37 image taken with the 6-m telescope of the russian academy of sciences ) , and thus from figure 1 of maccacaro et al .
( 1988 ) we see that the combined x - ray flux and @xmath37 magnitude limit do not agree well with those expected from a normal foreground star ( this is true even for 0.33.5 kev fluxes as low as @xmath46 erg @xmath9 s@xmath1 ) .
motch et al .
( 1997 ) state that about 85 per cent of their sources from the galactic plane survey are stars with x - ray emitting coronae , so we have ruled out the most probable type of confusing foreground source . in summary , although we can not formally prove that a foreground contaminating source is impossible , we consider one to be quite unlikely .
confusing background sources ( e.g. seyfert galaxies , quasars and clusters of galaxies ) are also unlikely .
examination of figure 18 of motch et al .
( 1997 ) shows that the probability of having an extragalactic source as bright as x-1 behind the superbubble is @xmath47 .
this is a conservative probability estimate since any extragalactic source would also suffer from x - ray absorption by matter within the plane of .
some types of possible extragalactic confusing sources ( e.g. clusters of galaxies ) can also be ruled out based on the x - ray variability described below . 0.1 in _ luminosity , spatial extent and variability _ 0.1 in given that x-1 is probably located in the plane of , we expect that it will have some absorption over the galactic column of @xmath31 @xmath9 .
we expect an intrinsic column in of @xmath48 @xmath9 based on figure 5 of ys93 .
we thus adopt a total column of @xmath49 @xmath9 . using this column and a power - law model with a photon index of @xmath30
, we obtain a power - law normalization of @xmath50 photons @xmath9 s@xmath1 kev@xmath1 at 1 kev . the 0.12.5 kev absorbed flux for this model
is @xmath51 erg @xmath9 s@xmath1 , and the 0.12.5 kev absorption - corrected flux is @xmath52 erg @xmath9 s@xmath1 .
while we note that the flux correction factor used above is large ( about 4.9 ) , the model we have used seems physically appropriate given the probable nature of x-1 ( see below ) , and the presence of a large column between the plane of and earth is without question . at a distance of 1 mpc
the absorption - corrected flux corresponds to a mean isotropic luminosity of @xmath53 erg s@xmath1 .
if , for comparison , we only use the galactic column and neglect the probable absorption in then we obtain a 0.12.5 kev absorption - corrected flux of @xmath54 erg @xmath9 s@xmath1 and an isotropic luminosity of @xmath55 erg s@xmath1 ( the power - law normalization and absorbed flux for this model are given above ) .
inspection of the x - ray morphology of x-1 does not show any strong evidence for deviations from a circular shape .
as the ys93 superbubble has an extent of @xmath56 arcsec , we have examined the spatial extent of x-1 ( see figure 4 ) .
x-1 is much less extended than the radio emission from the superbubble . while the radial profile shown in figure 4 is not formally consistent with the nominal point spread functions described in david et al .
( 1995 ) , it is known that aspect solution errors can lead to a significant broadening of the profiles of point sources in some hri observations ( see the discussion in morse 1994 ) .
we do not have enough counts to perform aspect error correction using the morse code hriaspcor in ftools , and tv cas is too far off axis to be a useful comparison point source .
the current data do not give strong evidence for spatial extent over that from a point source which suffers from aspect solution errors .
furthermore , the variability discussed below suggests that most of the x - ray flux originates from a compact region with an extent of less than @xmath57 light day . count rates from should be averaged over an integer multiple of the 400-s wobble period when used to search for rapid variability of cosmic x - ray sources ( see brinkmann et al . 1994 ; although note that the effects of the wobble are less serious for the hri than the pspc ) .
we have used a source cell with a radius of 20 arcsec to extract a light curve of x-1 .
we use 400-s bins and neglect partially filled bins . in figure 5
we show part of this light curve where there is evidence for x - ray variability .
the count rate from x-1 appears to have dropped by over a factor of 2 within a day .
we have calculated the poisson probability that the variability is merely due to photon statistics , and it is less than @xmath58 . even taking into account issues such as those described in press & schechter ( 1974 ) , the variability appears to be of high statistical significance .
we have used a double panel format in figure 5 for clarity because our observations are spread over 10.8 days , and there are no data in between the two panels shown in figure 5 .
we have verified the variability shown in figure 5 using two independent analysis software systems .
the mean background count rate in the source cell , computed using a large , nearby , circular , background cell , is @xmath59 count s@xmath1 for the left panel and @xmath60 count s@xmath1 for the right panel .
we have verified that there is no positional error ( e.g. a shift or rotation ) that could be moving the source out of our source cell during the times shown in our right hand panel ( see figure 6 ) .
if we bin an image using only the data from the left hand panel , we can clearly see x-1 over the background level . however , if we bin an image using only the data from the right hand panel , we can not see x-1 despite the fact that the right hand panel has twice as much exposure and a slightly lower background .
there is evidence for variability similar to that shown in figure 5 throughout our observation .
the highest count rate observed is about twice the mean count rate , which corresponds to a 0.12.5 kev isotropic luminosity of @xmath61 erg s@xmath1 .
as discussed below , it appears likely that x-1 is an x - ray binary system in .
unfortunately , our data do not appear to provide strong constraints on the binary orbital period via periodic x - ray flux variations ( e.g. eclipses ) .
this is due partially to the low count rate , but also to the fact that our 32.7 ks of exposure is broken into numerous , small , separated blocks with durations between 12003000 s. we have used pss and the hri point spread function to search our hri field for any other x - ray point sources coincident with the extent of . we do not detect any point x - ray sources stronger than @xmath62 within the extent shown in figure 5 of ys93 .
we do not detect any significant point sources near the maser sites , and we do not detect the radio continuum source described in section 3.3 of argon et al .
the standard analysis software system ( sass ) reports a possible diffuse x - ray source at @xmath19 00@xmath20 20@xmath21 18.6@xmath22 , @xmath23 @xmath2418@xmath2550@xmath26 with a positional error of 20 arcsec .
careful examination of the x - ray image and the sass output shows that this source is not of high statistical significance in the current data ( e.g. there are background fluctuations at almost comparable levels ) .
however , the fact that the sass x - ray source is within 12.5 arcsec of a bright radio source discussed in section 3.4 of klein & gr " ave ( 1986 ) suggests that it may well be real ( note that some of the source labeling in section 3.4 of klein & gr " ave 1986 is incorrect ) .
the sass x - ray source is also quite close to two of the sources listed in table 1 of ys93 .
this is the location of an h i hole in which may have been created by supernova events .
the ` best guess ' hri count rate for this source is @xmath63 count s@xmath1 .
further x - ray observations are needed to study this emission .
we have used pss and the hri point spread function to place upper limits on point sources at several representative positions in that lie @xmath64 arcsec from x-1 .
the typical @xmath65 upper limit is about @xmath66 count s@xmath1 . using a @xmath30 power law with a column of @xmath67 @xmath9 ( motivated by the galactic column and figure 5 of ys93 )
, this corresponds to an absorbed 0.12.5 kev flux upper limit of about @xmath68 erg @xmath9 s@xmath1 and an absorption - corrected 0.12.5 kev flux upper limit of about @xmath69 erg @xmath9 s@xmath1 .
the isotropic luminosity upper limit is about @xmath70 erg s@xmath1 .
this upper limit is appropriate for x - ray sources that are extended on scales smaller than or comparable to that of the hri spatial resolution ( about 5 arcsec which corresponds to about 24 pc at the distance of ) .
thus , this limit can be applied to supernova remnants in the free - expansion or sedov - taylor phases ( with ages @xmath71 yr ) , although many supernova remnants have luminosities lower than our upper limit ( see table 3.2 of charles & seward 1995 ) .
of course , it is possible that there are even more highly obscured x - ray sources in ( i.e. sources that suffer even stronger intrinsic absorption in itself ) with larger isotropic x - ray luminosities than the limit given above .
a hard x - ray observation of could search for such sources . in order to search for diffuse emission from ( e.g. from a hot interstellar medium ) ,
we have carefully examined moderately smoothed and heavily smoothed hri images .
we have used both full - band hri images as well as images made using only hri channels 38 ( this can help to maximize the signal - to - noise level ; see section 2.4 of david et al .
there does not appear to be any highly significant diffuse x - ray emission coincident with the optical or radio extents of . to obtain a characteristic constraint on the amount of large - scale diffuse emission
, we consider a circular ` source ' aperture centred at @xmath19 00@xmath20 20@xmath21 16.2@xmath22 , @xmath23 @xmath2418@xmath2523@xmath26 with a radius of 100 arcsec .
this aperture includes many of the h ii regions of , and it does not include x-1 .
the mean count rate density within this aperture is @xmath72 count s@xmath1 arcmin@xmath73 . for comparison , a circular ` background ' aperture that is located outside the extent of ( at @xmath19 00@xmath20 20@xmath21 55.6@xmath22 , @xmath23 @xmath2423@xmath2557@xmath26 and with a radius of 112 arcsec )
has a mean count rate density of @xmath74 count s@xmath1 arcmin@xmath73 ( this mean count rate density may be compared with the hri background values listed in table 5 of david et al .
thus , after background subtraction , our source aperture has a residual mean count rate density of @xmath75 count s@xmath1 arcmin@xmath73 ( i.e. it is consistent with zero ) . adopting the same spectral model
as used above for point - source upper limits , we find a mean ( absorbed ) surface brightness of @xmath76 erg @xmath9 s@xmath1 arcmin@xmath73 .
of course , this value is an average constraint on large - scale diffuse emission , and it can not be used on a point - by - point basis . is listed ( without discussion ) in boller et al .
( 1992 ) as having been detected during the all - sky survey ( rass ) .
revision of the all - sky survey processing has led to an improved position for the rass source of @xmath19 00@xmath20 20@xmath21 21@xmath22 , @xmath23 @xmath2416@xmath2518@xmath26 ( th .
boller , private communication ) , and we show this position in figure 1 . the positional sensitive proportional counter ( pspc ; pfeffermann et al .
1987 ) registered @xmath77 count s@xmath1 from the claimed all - sky survey source ( th .
boller , private communication ) , which corresponds to an equivalent hri count rate of about @xmath78 count s@xmath1 .
the all - sky survey exposure was short ( 494 s ; th .
boller , private communication ) and places no useful spectral constraints on the pspc source .
only about 13 pspc photons in total ( including background ) were collected .
the revised rass position is 1.15 arcmin from our hri position .
the equivalent hri count rate for the rass source and our mean pointed hri count rate for x-1 are similar , but the positional offset is somewhat larger than would be expected from the typical @xmath6 arcsec hri error circle radius and the typical @xmath79 arcsec pspc error circle radius .
there are no hri sources formally consistent with the rass position , despite the fact that the hri image probes much deeper than the rass .
our position for x-1 is almost certainly not in error ( see figure 2 and the associated discussion ) , and the rass data do not place a strong limit on the brightness of x-1 during the all - sky survey ( x-1 lies between the 58 and 69 per cent contour levels for the rass source ) . ngc 147 was observed with the hri starting on 1995 january 19 ( rh400744 : total raw exposure of 14.8 ks spread over 63.6 ks ) , and ngc 185 was also observed starting on 1995 january 19 ( rh400743 : total raw exposure of 21.0 ks spread over 91.1 ks ) .
each observation had its pointing position coincident with the optical centre of the appropriate galaxy .
.x - ray sources near ngc 147 ( the first three sources ) and ngc 185 ( the last two sources ) in the hri fields . note that many of these sources are probably not related to ngc 147 or ngc 185 ( see the text ) .
the last column gives a brief description of the optical morphology based on inspection of palomar schmidt plates , palomar 200-inch plates and lick 3-m plates .
[ cols="^ , < , < , < , < " , ] we have used pss and the hri point spread function to search for x - ray point sources coincident with these galaxies . we do not detect any x - ray sources coincident with the cores of these galaxies , and we do not detect any x - ray sources coincident with their known globular clusters either ( we use the globular cluster positions from table 9 of ford et al . 1977 ) . as with , we have scrutinized smoothed x - ray images of each of these galaxies and do not see any evidence for diffuse emission .
we have placed upper limits on point sources at several representative positions in ngc 147 , and the typical 2@xmath80 upper limit is about @xmath81 count s@xmath1 . using a @xmath30 power law with the galactic column ( @xmath82 @xmath9 )
, this corresponds to an absorbed 0.12.5 kev flux upper limit of about @xmath83 erg @xmath9 s@xmath1 and an absorption - corrected 0.12.5 kev flux upper limit of about @xmath69 erg @xmath9 s@xmath1 .
the isotropic luminosity upper limit is about @xmath84 erg s@xmath1 .
we have performed the same procedure for ngc 185 , and the corresponding upper limits are @xmath85 count s@xmath1 ( count rate ) , @xmath86 erg @xmath9 s@xmath1 ( absorbed flux ) , @xmath87 erg @xmath9 s@xmath1 ( absorption - corrected flux ) and @xmath88 erg s@xmath1 ( isotropic luminosity ) . in table 1
we list all x - ray sources detected within 12 arcmin of the centres of ngc 147 and ngc 185 .
all of the sources in this table lie outside the central optical extents of ngc 147 and ngc 185 , and all but source 5 lie outside the regions previously searched for globular clusters ( see baade 1951 ; hodge 1974 ; hodge 1976 ; ford et al .
. however , these galaxies both are optically very large at faint surface brightness levels .
most of the x - ray sources in table 1 are probably unrelated foreground or background sources , but a couple of them may be associated with these galaxies or currently unknown globular clusters .
none of the sources in table 1 is currently listed in the nasa extragalactic database ( ned ) or the simbad database .
we give optical morphologies of possible counterparts to the x - ray sources in table 1 .
source 4 is near the center of a ring of stellar images and there appears to be a very faint object about 2 arcsec southwest of the x - ray position .
source 5 is almost certainly not a globular cluster , as it is not resolved into stars on red palomar plates .
it appears to be a foreground star .
x-1 appears to be located in near the centre of the ys93 non - thermal radio superbubble , and it has a powerful mean 0.12.5 kev isotropic luminosity of @xmath89 erg s@xmath1 .
single ( nondegenerate ) stars do not produce this much x - ray emission ( e.g. rosner , golub & vaiana 1985 ; pollock 1987 ; kashyap et al .
1992 ; and references therein ) , and the x - ray luminosities of cataclysmic variables also fall short by several orders of magnitude ( e.g. eracleous , halpern & patterson 1991 ) .
the observed variability argues against most of the emission coming from a cluster of stars or a supernova remnant .
thus it appears most likely that we have discovered an accreting neutron star or black hole in a binary system .
the highest 0.12.5 kev luminosity we observe from x-1 is @xmath61 erg s@xmath1 , and the total x - ray luminosity is likely to be @xmath90 erg s@xmath1 if the source is indeed an accreting neutron star or black hole ( see white , nagase & parmar 1995 for a discussion of the x - ray spectral energy distributions of x - ray binaries ) .
this is the eddington limit luminosity for a @xmath6 m@xmath91 object which is above the maximum mass for a neutron star .
while this may suggest the compact object is a black hole , we can not rule out a neutron star which emits anisotropically or at a super - eddington rate . this is especially true in light of the low heavy element abundances in ( e.g. lequeux et al . 1979 ) and the apparent x - ray luminosity / abundance relation for accreting compact objects ( clark et al .
1978 ; section 2.5 of van paradijs & mcclintock 1995 ) . as a peculiar emission - line star ( wr17 ) has been located in the tiny hri error circle by massey & armandroff ( 1995 ; see above ) , it is worth investigating whether this could be the binary companion star to the accreting compact object .
if , as claimed by massey & armandroff ( 1995 ) , this star is a wn type wolf - rayet star , then x-1 would be a more x - ray luminous version of the possible wolf - rayet x - ray binaries reported near the centre of 30 doradus ( wang 1995 ; also see van kerkwijk et al .
1996 and references therein regarding the possible wolf - rayet nature of cyg x-3 ) .
the system would have originally been a binary with two massive stars , and a supernova in this system would probably have contributed to the creation of the ys93 superbubble ( such a system might plausibly remain bound after a supernova ; see brandt & podsiadlowski 1995 ) .
of course , the identification of x-1 with wr17 has not yet been proven and x-1 could also be a low mass x - ray binary in which the optical companion is below the current threshold of detectability ( cf .
stocke , wurtz & khr 1991 ; but note that petre 1993 suggests that low mass x - ray binaries are rare in star forming regions ) .
finally , the presence of a powerful x - ray source in the centre of an unusually large bubble of radio emission invites a brief comparison with ss433 and w50 . ys93
compared the radio surface brightness ( @xmath92 ) and diameter ( d ) of their superbubble to the so - called ` @xmath92-d relationship ' for snr in the magellanic clouds . while the physical meaning , if any , of the @xmath92-d relationship is somewhat obscure ( especially when transported wholesale from the magellanic clouds to ) , the ys93 superbubble does appear to have a much larger diameter than expected given its radio surface brightness .
ys93 proposed a multiple snr model to explain this fact .
while the ys93 model appears entirely plausible , w50 also lies off the @xmath92-d relation ( margon 1984 ; its largest dimension is about 65 pc ) .
the growth of the ys93 superbubble could have been influenced by x-1 ( see section 5 of margon 1984 ) .
x-1 does not appear to have redshifted / blueshifted emission lines in the spectrum of massey & armandroff ( 1995 ) and there is no strong radio point source at its position , so if there were ss433-like activity in the past it appears to have ended . or
_ sax _ observations of would be very useful as they would ( 1 ) probably obtain a significantly higher count rate from x-1 due to the heavy column which strongly affects the hri band , ( 2 ) allow a better variability study including searches for eclipses from a high mass companion and x - ray pulsations , ( 3 ) allow a measurement of the spectral shape of x-1 and ( 4 ) search for additional variable and highly absorbed x - ray sources in .
they would also allow a much more precise determination of the luminosity of x-1 since the column correction factor would not be so large .
we are attempting to obtain such observations .
we are also trying to search for evidence of a binary star ( especially wr17 ) within the x - ray error circle and obtain higher quality spectra to further examine the wolf - rayet classification of wr17 by massey & armandroff ( 1995 ) .
our current spectra do not allow definitive discussions of either of these issues . if there are any x - ray binaries in the centres of ngc 147 or ngc 185 , they were quite faint at the times of these observations ( see section 2.2 for upper limits ) .
we do not detect the putative snr of gallagher et al .
( 1984 ) , although our data do not rule out the possibility that this object is a snr .
we gratefully acknowledge financial support from the smithsonian institution and the united states national science foundation ( wnb ) and the royal society ( acf ) . we thank h. ebeling for the use of his imcont software , and we thank d. zucker for reducing the optical spectra of stars in the region of x-1 .
we thank th .
boller , d. harris , l. ho , p. massey , j. mcclintock , p. obrien , r. petre , ph .
podsiadlowski and h. yang for helpful discussions .
the palomar schmidt plate image was obtained via the digitized sky survey which was produced at the space telescope science institute under united states government grant nag w-2166 .
the palomar observatory sky survey was funded by the national geographic society .
allan d.j . , 1995 ,
asterix user note 004 : source searching and parameterisation .
univ . of birmingham , birmingham argon a.l . , greenhill l.j .
, moran j.m . , reid m.j . ,
menten k.m .
, henkel c. , inoue m. , 1994 , apj , 422 , 586 baade w. , 1951 , u. mich .
obs . pub .
, 10 , 7 baan w.a .
, haschick a. , 1994 , apj , 424 , l33 boller th .
, meurs e.j.a . ,
brinkmann w. , fink h. , zimmermann u. , adorf h .- m . , 1992 , a&a , 261 , 57 brandt w.n .
, podsiadlowski ph . , 1995 , mnras , 274 , 461 brinkmann w. , et al . , 1994 , a&a , 288 , 433 charles p.a . ,
seward f.d . , 1995 , exploring the x - ray universe .
cambridge university press , cambridge clark g. , doxsey r. , li f. , jernigan g. , van paradijs j. , 1978 , apj , 221 , l37 david l.p .
, harnden f.r . , kearns k.e . ,
zombeck m.v . , 1995 , the rosat high resolution imager .
gsfc , greenbelt eracleous m. , halpern j. , patterson j. , 1991 , apj , 382 , 290 ford h. , jacoby g. , jenner d. , 1977 , apj , 213 , 18 gallagher j.s . , hunter d.a .
, mould j. , 1984 , apj , 281 , l63 helfand d.j .
, 1984 , pasp , 96 , 913 hodge p.w . , 1974 ,
pasp , 86 , 289 hodge p.w . , 1976 ,
aj , 81 , 25 hodge p.w . , lee m.g . , 1990 , pasp , 102 , 26 hodge p.w . , 1994 , in munoz - tunon c. , sanchez f. , eds , the formation and evolution of galaxies
cambridge univ .
press , cambridge , p. 1
kashyap v. , rosner r. , micela g. , sciortino s. , vaiana g.s .
, harnden f.r . , 1992 , apj , 391 , 667 khalesseh b. , hill g. , 1992 , a&a , 257 , 199 klein u. , gr " ave r. , 1986 , a&a , 161 , 155 lequeux j. , peimbert m. , rayo j.f . , serrano a. , torres - peimbert s. , 1979 , a&a , 80 , 155 maccacaro t. , gioia i.m . , wolter a. , zamorani g. , stocke j.t . , 1988 ,
apj , 326 , 680 margon b. , 1984 , ara&a , 22 , 507 massey p. , armandroff t.e . , conti p.s .
, 1992 , aj , 103 , 1159 massey p. , armandroff t.e . , 1995 , aj , 109 , 2470 mcgale p.a . , pye j.p . , hodgkin s.t . , 1996 , mnras , 280 , 627 metcalfe t.s . , 1995 ,
iau inform .
bull . var . stars .
4197 minkowski r. , abell g.o .
, 1963 , in strand k. , basic astronomical data .
chicago press , chicago , p. 481
morse j.a . , 1994 ,
pasp , 106 , 675 motch c. , guillout p. , haberl f. , pakull m. , pietsch w. , reinsch k. , 1997 , a&a , 318 , 111 mukai k. , 1995 , pimms users guide .
nasa / gsfc , greenbelt ohta k. , sasaki m. , yamada t. , saito m. , nakai n. , 1992 , pasj , 44 , 585 petre r. , 1993 , in beckman j. , colina l. , netzer h. , eds , the nearest active galaxies .
csic press , madrid , p. 117
pollock a.m.t . , 1987 , apj , 320 , 283 pfeffermann e. , et al . , 1987 , proc .
spie , 733 , 519 press w.h . , schechter p. , 1974 ,
apj , 193 , 437 rosner r. , golub l. , vaiana g.s . , 1985 , araa , 23 , 413 singh k.p . ,
drake s.a . ,
white n.e .
, 1996 , aj , 111 , 2415 stark a.a . , gammie c.f . , wilson r.w .
, bally j. , linke r. , heiles c. , hurwitz m. , 1992 , apjs , 79 , 77 stocke j.t .
, wurtz r. , khr h. , 1991 , aj , 102 , 1724 trmper j. , 1983 , adv . space res . , 4 , 241 van kerkwijk m.h . , geballe t.r . , king d.l . , van der klis m. , van paradijs j. , 1996 , a&a , 314 , 521 van paradijs j. , mcclintock j.e . , 1995 , in lewin w.h.g . , van paradijs j. , van den heuvel e.p.j . , eds , x - ray binaries .
cambridge univ . press , cambridge ,
wilson c.d . , 1995 ,
apj , 448 , l97 wang q.d .
, 1995 , apj , 453 , 783 white n.e . , nagase f. , parmar a.n .
, 1995 , in lewin w.h.g . ,
van paradijs j. , van den heuvel e.p.j . ,
eds , x - ray binaries .
cambridge univ . press , cambridge ,
yang h. , skillman e.d . , 1993 ,
aj , 106 , 1448 ( ys93 )
in this appendix we give the first x - ray data on the serendipitously detected binaries tv cas and bh cas ( see section 2.1.1 ) .
this will allow their inclusion in future systematic studies of the x - ray emission from algol and w uma - type binaries ( e.g. singh , drake & white 1996 ; mcgale , pye & hodgkin 1996 ; and references therein ) .
tv cas is offset from the centre of the hri field of view by about 12 arcmin , and we detect about 1080 counts from it after background subtraction ( the detection significance is @xmath93 ) .
the mean count rate after correction for vignetting is @xmath94 count s@xmath1 , and there is evidence for count rate variability by a factor of about 2 . the low count rate and large data gaps prevent a detailed x - ray variability analysis . based on the count rate conversion factor in singh et al .
( 1996 ) and the relative responses of the pspc and hri , we estimate that 1 hri count s@xmath1 corresponds to about @xmath95 erg @xmath9 s@xmath1 . using a distance of 275 pc ( from table 7 of khalesseh & hill 1992 ) , we calculate a 0.12.5 kev luminosity of about @xmath96 erg s@xmath1 , which suggests fairly typical coronal activity for an algol - type binary .
bh cas is offset from the centre of the hri field of view by about 11 arcmin , and we detect about 42 counts from it after background subtraction ( the detection significance is @xmath97 ) .
the count rate after correction for vignetting is @xmath98 count s@xmath1 , and we are not able to probe for variability due to the low count rate .
we adopt the same count rate to flux conversion factor as above and thus derive a 0.12.5 kev flux of @xmath99 erg @xmath9 s@xmath1 . the distance to bh cas
is not well known , but an upper limit on the distance is @xmath100 pc ( t. metcalfe , private communication ) .
thus the x - ray luminosity appears to be less than about @xmath101 erg s@xmath1 .
the observed x - ray flux does not allow a strong constraint on the distance via the implied luminosity ( compare the range of w uma x - ray luminosities seen in mcgale et al . | we report on pointed x - ray observations of , ngc 147 and ngc 185 made with the high resolution imager ( hri ) .
these are three local group galaxies that have never been previously studied in detail in the x - ray regime .
is the closest starburst galaxy to our own galaxy , and ngc 147 and ngc 185 are companions to m31 .
we have discovered a variable x - ray source coincident with .
the source is located near the centre of a large , non - thermal bubble of radio emission , and it is positionally coincident with an emission line star in which has been classified as a wn type wolf - rayet star .
we demonstrate that a confusing foreground or background source is improbable .
the x - ray source is probably an x - ray binary in , and it may be a wolf - rayet + black hole binary .
the source has mean and maximum 0.12.5 kev isotropic luminosities of about 2 and 4 times @xmath0 erg s@xmath1 .
we do not detect any sources in the central regions of ngc 147 or ngc 185 .
we place upper limits on their central x - ray emission , and we list all x - ray sources coincident with their outer extents .
we also present the first x - ray detections of the well - studied algol - type binary tv cas and the w uma - type binary bh cas , which were both serendipitously observed during our pointing .
galaxies : individual : galaxies : individual : ngc 147 galaxies : individual : ngc 185 stars : individual : tv cas stars : individual : bh cas x - rays : galaxies x - rays : stars . |
integration of a rigid body motions is well known . in this paper
we recall in sec.[sec2 ] the kinematic description of the coordinates in which the system is obviously integrable by reduction to quadratures ( deprit s coordinates , described in eq . ) .
we try to treat simultaneously the motions near all proper rotations , at the expense of having to append labels @xmath1 to most quantities thus making the notation a little heavier .
this is done to keep always clear that the stable and unstable motions are in some sense described by the same analytic functions .
our first aim is to obtain canonical coordinates @xmath2 , in the vicinity of the proper rotations , such that the hamiltonian , at fixed total angular momentum @xmath3 , becomes a function of the product @xmath4 ( near unstable proper rotations ) or of @xmath5 ( near the stable ones ) .
although existence of such coordinates is well known and their determination is in principle controlled by appropriate ( elliptic ) integrals , ( * ? ? ?
* sec.50 ) , it remains , to our knowledge , only implicit in the literature , ( * ? ? ?
* sec.69 ) . in sec.[sec3 ]
explicit expressions for the elliptic integrals are derived , eq . , which give the motions in terms of variables @xmath6 evolving exponentially in time on the hyperbolae @xmath7 or rotating uniformly on the circles @xmath8 and `` reducing '' the problem to determine the nontrivial scaling function @xmath9 ( depending on @xmath10 or @xmath11 ) which leads to the above mentioned canonical coordinates via @xmath12 , @xmath13 . in appendix
[ appa ] given details on the computation of appropriate elliptic integrals : this should also clarify the advantages of the cohomological analysis of the following sec.[sec4 ] and [ sec6 ] . in sec.[sec4 ]
the scaling function @xmath9 is related to a jacobian determinant between two differential forms , and it is computed : this is done , avoiding evaluation of other elliptic integrals , by computing the cohomology of the forms @xmath14 ( where @xmath15 are two deprit s coordinates , see sec.[sec2 ] ) and @xmath16 relative to the functions @xmath10 or @xmath11 .
the canonical coordinates are therefore completely determined , eq . , ( together with eq . ) .
the latter expressions are still quite implicit and in sec.[sec5 ] we bring them to a form suitable for the evaluation of the integrating coordinates via computable power series ; and as an example a few terms of the scaling functions @xmath9 and of the normal form of the hamiltonian are evaluated , eq . , .
doing so various families of polynomials , in a variable @xmath0 depending on the inertia moments , arise .
the polynomials appear to have zeros on very special locations suggesting possible conjectures .
to understand the properties of the polynomials which arise in the construction of the normal forms we show , sec.[sec6 ] , that the normal form can be generated by solving another problem of relative cohomology : this leads to an explicit determination of the normal forms in terms of a further family of polynomials .
in this case we prove that some of the new polynomials have all roots on the unit circle ( with the help of theorems on the location of zeros of symmetric polynomials ) : and we conjecture a relation with the lee - yang theorem for the zeros of `` ferromagnetic polynomials '' . in sec.[sec7 ]
we determine the action angle coordinates for the other two deprit variables ( which either do not appear in the hamiltonian or are constants of motion ) . in sec.[sec8 ]
we apply the relative cohomology method of sec.[sec4 ] and [ sec6 ] to the pendulum , to illustrate the cohomological method . and to the geodesic flow on revolution ellipsoids , with the purpose of showing that in the latter case also appear natural families of polynomials in a parameter @xmath0 ( depending on the ellipsoid equatorial and polar axes ) : but in this case the zeros are not located on the unit circle , _ although almost so _ : eq .. appendices contain supplements ( _ e.g. _ the proof of chen s theorem ) .
the theory of jacobian elliptic functions , for reference see @xcite , yields a complete calculation for the motion of a solid with a fixed point .
this is revisited here , to exhibit a few interesting properties of the relevant elliptic integrals .
the hamiltonian of a solid with inertia moments @xmath17 , in deprit s canonical coordinates @xmath18 , see problems in sec.4.11 of @xcite , is h(k_z , a , b,,,)=12 + 12(+ ) ( a^2-b^2).and it will be convenient to order the proper axes of the solid body so that @xmath19 . in this way the motions with @xmath20 and @xmath21 or @xmath22 are stable or unstable rotations , respectively
. for convenience the unstable rotations will be studied by shifting by @xmath23 the origin of the angle @xmath24 , so that for the unstable rotations we shall use the hamiltonian : h(k_z , a , b,,,)=12 + 12(+ ) ( a^2-b^2).motions can be described by referring them to the two node lines @xmath25 and @xmath26 : @xmath25 being the node between the angular momentum plane , orthogonal to the angular momentum @xmath27 , and the fixed reference plane @xmath28 , and @xmath26 being the node between the angular momentum plane and the inertial plane @xmath29 .
the angle @xmath30 locates the node @xmath25 on the fixed plane and the angle @xmath31 locates @xmath32 on the angular momentum plane with respect to @xmath25 .
[ fig1 ] with reference to the hamiltonian in the form eq .
, where @xmath21 correspond to _ unstable _ rotations around the @xmath33 proper axis , the pairs of canonical variables are @xmath34 , @xmath35 and @xmath36 ; @xmath37 while the angle @xmath38 is constant . at the `` fixed points '' @xmath39 or @xmath40 the angles @xmath31 and @xmath30 rotate at constant speeds : @xmath41 and @xmath42 or @xmath43 and @xmath42 respectively ; the angle between the angular momentum and the inertial axis @xmath44 is @xmath23 and @xmath45 . for the stable rotations
the role of the axes @xmath46 and @xmath47 are exchanged .
hence the coordinates are adapted to describe proper rotations around the axis @xmath48 ( @xmath21 , unstable ) of intermediate inertia or @xmath49 ( @xmath50 , stable ) of largest inertia .
the inertia moments @xmath19 will be used to define the quantities the lyapunov coefficients at the unstable fixed point are @xmath51 real with @xmath52 and at the stable are @xmath53 imaginary with @xmath54 .
we try to use a notation valid for _ both _ the stable and the unstable rotations : therefore a label @xmath1 , often abridged into @xmath55 , is appended to most quantities : with @xmath56 referring to the unstable case and @xmath57 referring to the stable one . calling @xmath58 the total energy ,
define for @xmath1 : & & b^2_a , g^2_a(2u - a^2/j_a)j_a^-1b_a^2(j_a^-1)^2 , + & & s^2_aa^2,r^2_a , k^2_a , + & & g^2_a = g^2_a(1+a s^2_a ) , b^2_a(1+a s^2_a)=then @xmath59 is a characteristic inverse time scale and , referring to eq . for @xmath60 and to eq . for @xmath61 , with @xmath62 constants of motion . if @xmath61 it is @xmath63 .
suppose first @xmath64 and small : then & & g^2_a , ag^2_a>0,b^2_a>0 , , + & & s_a^21 , , 1+a r_a^2== , + & & a s_a^2=,==g_a^2j_+j_- and motions near @xmath65 and with energy @xmath58 close to @xmath66 and can be expressed in terms of jacobian elliptic integrals ( see @xcite for notations ) .
rotating the positions of the inertia moments in @xmath67 the meaning of the variables changes but the following analysis remains essentially unchanged ( up to renaming variables ) : it appears in this way that the proper rotations around axis @xmath46 are ( linearly ) unstable while the rotations around the highest and lowest inertia axes ( @xmath68 and @xmath44 ) are ( linearly ) stable .
`` linearization '' of rigid body motions is , of course , well known to lead to elliptic integrals , see @xcite .
here we rederive the integration in a form suitable for our purposes of action - angle coordinates construction .
let @xmath69 so that @xmath70 ; for @xmath60 eq .
describe the separatrix branches emerging from the proper rotation around @xmath46 controlled by eq .
, while for @xmath61 they describe the motion near the stable rotation around @xmath68 controlled by eq .. let @xmath71 close to @xmath72 ; from the definitions of the jacobian elliptic integrals , see @xcite , and setting k_a=,u_a= g_a tg_a tit follows ( choosing the sign @xmath73 in the eq .
, for instance ) from ( @xcite for @xmath60 and @xcite for @xmath61 after correcting the obvious typo in @xcite and by @xmath74 ) , see appendix [ appa ] for details , & & -4 mm b(t)== b_a , + & & -4 mm ^2(t)= = + & & -4 mm ( t)==g_a where the last relation is deduced from the equations of motion and the expression for @xmath75 .
notice that @xmath76 and if @xmath77 it is @xmath78 , @xmath79 , @xmath80 and motions with this energy are `` like '' the motions close to the separatrix or , respectively , close to equilibrium of a pendulum . for @xmath81 , or for @xmath82
hence @xmath83 imaginary , the above relations have to be interpreted via suitable analytic continuations .
some of the following formulae become singular as @xmath84 : the singularity is only apparent and it will disappear from all relevant formulae derived or used in the following .
introduce variables useful in the following ( most of them appear in the theory of jacobi s elliptic functions ) : the @xmath85 depend on @xmath86 and can , and will , be imagined as functions of @xmath87 .
let @xmath88@xmath89 and the integral can be evaluated by series , @xcite and @xcite , defining @xmath90 and @xmath91 with @xmath92 .
this is well defined near the proper rotations , where @xmath93 , as @xmath83 tends to @xmath72 .
therefore , as derived in detail in appendix [ appa ] , _ a(ar_a^2,k^2_a)e^v_a= ( ) e^i v(11+a r_a^2,k^2_a ) and @xmath94 with @xmath95 real and @xmath96 real , with @xmath95 real .
introduce the new variables @xmath97 defined by p _
e^ g_0,at,= , x_ap_+p_- .
the above relations can be algebraically elaborated into the transformation of coordinates @xmath98 , see again details in appendix [ appa ] : and in the new coordinates the _ flow @xmath99 generates a solution _ of the equations of motion if @xmath100 is small and if the real parts of @xmath101 are @xmath102 .
however the equations of motion are analytic and so are the eq . : hence eq .
yield a solution under the only condition that @xmath100 are small , to be required for its convergence .
eq . give a complete parametrization of the motions _ but the coordinates @xmath103 are not canonical _ : the determination of the canonical coordinates will be discussed and done in sec.[sec4 ] .
\(1 ) via identities relating jacobian elliptic functions it would be possible to avoid considering functions that are defined by analytic continuations , for instance when the modulus @xmath104 or @xmath105 , and reduce instead to considering only cases with the `` simple '' arguments ( expressing functions by elliptic functions of other arguments , like @xmath106 , when necessary ) : we avoid this because it would hide the nice property that the stable and unstable cases can be seen as analytic continuations of each other .
\(2 ) the formula for @xmath24 reminds of one found by jacobi which he commented by saying that `` _ inter formulas elegantissimas censeri debet _ '' , @xcite ( _ i.e. _ _ it should be counted among the most elegant formulae _ , see also @xcite .
the motions @xmath107 solve the equations of motion if the real and imaginary parts of @xmath108 are positive .
but the equations of motion are analytic , hence the formulae of the previous sections give solutions of the equations independently of the sign of @xmath109 , provided the series converge .
convergence requires @xmath110 : which represents many data , in particular those in the vicinity of the separatrix or of the stable proper rotation .
the coordinates can be called `` hyperbolic '' or `` elliptic '' .
we also see that time evolution preserves both volume elements @xmath111 and @xmath112 ; which means that the jacobian determinant @xmath113 must be a function constant over the trajectories , hence a function @xmath114 of @xmath115 in the two cases .
note that @xmath114 has dimension of an action .
introduce the variables @xmath6 , related to @xmath101 in eq . by : , to deal with real coordinates when useful ; it is then possible to change coordinates setting @xmath116 , ( or @xmath117 , @xmath118 ) and choose @xmath119 so that the jacobian determinant @xmath114 for @xmath120 @xmath121 is @xmath122 .
a brief calculation shows that this is achieved by fixing c^2_a(x)=_0^x d_a(y)dy , which is possible for @xmath123 small because , from the equations of motion it is @xmath124 finite ( _ e.g _ @xmath125 ) .
therefore map @xmath120
@xmath126 is area preserving , hence _ canonical_. the hamiltonian eq . becomes a function @xmath127 of @xmath128 and the derivative of the energy with respect to @xmath129 has to be @xmath130 ( because the @xmath131 are canonically conjugated to @xmath15 ) .
note that @xmath129 has the dimension of an action , while @xmath2 are , dimensionally , square roots of action .
this allows us to find @xmath114 : by imposing that the equations of motion for the @xmath126 canonical variables have to be the hamilton s equations with hamiltonian @xmath132 it follows that @xmath133 , _
i.e. _ @xmath134 or @xmath135 by the above expression for @xmath136 .
the just obtained relation gives d_a(x_a)=g_0,a(x_a)^-1 u_a(x_a)which is an expression for the jacobian @xmath137 ( note that the jacobian between @xmath34 and @xmath126 is identically @xmath48 by construction ) .
to proceed it is necessary to determine the function @xmath138 .
the idea in @xcite is to make use of the general theory of relative cohomology classes to determine @xmath139 and , therefore , to find a complete expression of the action angle variables following a procedure employed in the theory of limit cycles of planar vector fields , @xcite .
a derivation can be based directly on the equalities on symplectic forms : to simplify the notation we drop for a while the label @xmath140 from @xmath123 s .
the idea is to compute only the _ cohomology class _ of the volume forms in the relative cohomology of the function @xmath141 , @xcite . for any @xmath49-form @xmath142 with @xmath143
the `` _ cohomology class _ '' relative to @xmath141 is the function defined by @xmath144 @xmath145 .
therefore an expression for the jacobian @xmath114 is obtained by finding an expression for the cohomology class of the form @xmath146 relative to @xmath141 .
remark that eq .
imply where @xmath147 is a form with @xmath72 cohomology ( _ i.e _ the first term is @xmath148 `` modulo @xmath149 '' because the cohomology of @xmath150 is , for all @xmath151 s , @xmath72 ) .
the terms with @xmath152 and @xmath153 do not contribute to the cohomology and , up to terms ( denoted @xmath154 ) not contributing to the cohomology , the form in eq . becomes ( as only terms with @xmath155 contribute ) if @xmath156 and @xmath157 , only the terms with @xmath158 contribute and we can take hence the cohomology with respect to @xmath159 is that of @xmath160 times -4 mm which contributes -(2i xg_0(x ) _
= 1i_=0^ ) dp_+dp_- because the cohomology of @xmath161 is @xmath162 .
therefore , by eq . and
by @xmath163 , see eq .
: summarizing we set , with @xmath101 defined in eq .
, , where @xmath164 if @xmath60 and @xmath165 if @xmath61 .
[ sec5 ] it it possible to evaluate @xmath166 and the normal form for the energy @xmath167 in the canonical coordinates @xmath121 up to large orders in @xmath129 .
this is done simply by deriving power series expansions , in @xmath87 first and then @xmath87 in terms of @xmath129 , for the functions in eq . .
consistently with eq .
@xmath166 can be expanded into a power series of @xmath168 or @xmath123 whose coefficients are functions of the parameter @xmath169 .
there is a simple relation between the cases @xmath55 and the coefficients of the expansions in @xmath168 can be computed exactly with a finite computational algorithm up to any prefixed order : this is described below .
the analysis of the jacobian @xmath170 can be based on the computable expressions of @xmath171 in eq .. normal forms express @xmath172 . from eq .
, we derive @xmath173 ( see eq . ) in terms of @xmath174 and , by eq .
, @xmath87 in terms of @xmath174 : which can be inverted in terms of the functions @xmath175 as k_a(x)^2= ( ) ^4w(ax ) , a=1as shown in @xcite .
therefore the above relations give explicitly the functions @xmath176 in terms of @xmath123 , hence they determine the functions @xmath177 .
to express @xmath176 in terms of @xmath168 we have to determine the jacobian function and invert the relation @xmath178 in the two cases .
this is done by remarking that @xmath179 and @xmath177 can be written as & & = ag_a(x)^2= 1j_12 ^ 2 , + & & h(x)=a^2 1as_a^2=a^2 + & & c^2(x,r_a^2)= 2a r_a ( w(ax)+a r_a^2)^-12 + & & 2_h=0^_= where the relations @xmath180 , @xmath181 , following from eq . or , have been used and the function @xmath182 is defined in eq .. this is remarkable because the summation over @xmath183 is analytic in @xmath184 in spite of the fact that each addend has a branch point in @xmath185 at @xmath186 .
therefore eq .
shows that the two cases @xmath55 are described by the same functions ( _ of course evaluated at completely different points _ ) and the power series for @xmath9 and @xmath58 can be computed setting @xmath60 and then the case @xmath61 is obtained by replacing @xmath168 by @xmath187 and @xmath0 by @xmath188 , @xmath189 by @xmath190 .
the simple relation between the unstable and stable cases implies that that we only need to study one of the two cases : we shall only make computations in the hyperbolic case @xmath60 . at this point the jacobian and
the normal forms can be computed as formal series in @xmath168 after solving the implicit function problem @xmath191 in the form @xmath192 ( in a power series in @xmath193 ) .
an algebraic numerical evaluation for @xmath60 ( unstable case with , for simplicity , @xmath194 ) yields results that for @xmath195 , if @xmath196 , are & & xc(x , r^2)^2=4 - 24 ( -1 + r^2 ) ^2 + 32 ( 3 - 14 r^2 + 3 r^4 ) ^3 + & & - 16 ( -1 + r^2 ) ( 19 - 242 r^2 + 19 r^4 ) ^4 + & & + 24 ( 35 - 1140 r^2 + 3026 r^4 - 1140 r^6 + 35 r^8 ) ^5 + & & - 192 ( -1 + r^2 ) ( 11 - 740 r^2 + 3426 r^4 - 740 r^6 + 11 r^8 ) ^6 + & & + 64 ( 77 - 10206 r^2 + 103635 r^4 - 211460 r^6 + 103635 r^8 + & & - 10206 r^10 + 77 r^12 ) ^7+ finally the polynomials in @xmath0 in eq . , coefficients of order @xmath197 for the power expansion in @xmath198 , _ can be conjectured to have all roots real _ : this has been numerically checked for @xmath199 .
the normal forms of the hamiltonians ( _ i.e. _ their expression in the variables @xmath2 ) are derived by expressing @xmath200 in eq . in terms of @xmath0 and @xmath201 and then expanding @xmath201 in powers of @xmath123 using @xmath202 in eq .
, via @xcite ( notice that in our hyperbolic case @xmath123 is the @xmath203 of the quoted tables with @xmath204 and @xmath205 exchanged and finally computing @xmath206 by inverting the jacobian relation @xmath207 .
this can be implemented algebraically on a computer . with the definitions in eq .
@xmath208 , in the unstable case ( @xmath56 ) or in the stable ( @xmath57 ) case , can be expressed as function of @xmath209 or , respectively , of @xmath210 obtained by solving @xmath191 as @xmath211 and replacing @xmath123 by @xmath212 in @xmath213 in eq.(notice that @xmath214 ) .
the resulting functions are normal forms for the hamiltonian since @xmath2 are conjugated variables . hence h(x_a)=+ ( ax_a , ar_a^2 ) for @xmath1 where @xmath215 , where @xmath216 is in eq .. one obtains & & ( x , r^2)=4 x - 2 ( -1 + r^2)x^2 - ( 1 + r^2)^2 x^3 + & & - 54 ( -1 + r^2 ) ( 1 + r^2)^2 x^4 - 316 ( 1 + r^2)^2 ( 11 - 10 r^2 + 11 r^4 ) x^5 + & & - 716 ( -1 + r^2 ) ( 1 + r^2)^2 ( 3 - 4 r + 3 r^2 ) ( 3 + 4 r + 3 r^2 ) x^6 + & & - 164 ( 1 + r^2)^2 ( 527 - 332 r^2 + 330 r^4 - 332 r^6 + 527 r^8 ) x^7 + & & - ( -1 + r^2 ) ( 1 + r^2)^2 ( 1043 + 548 r^2 + 1058 r^4 + & & + 548 r^6 + 1043 r^8 ) x^8+ it is remarkable that _ the polynomials in @xmath217 appearing as coefficients of order @xmath218 in the power series expansion in @xmath168 of the normal form have , for all @xmath219 s , just roots of unity _ as it appears by evaluating them numerically .
[ sec6 ] the function @xmath220 being the normal form of the hamiltonian in action - angle variables is unique , while the action - angle variables themselves are not unique .
therefore it makes sense to see if the normal form itself could be derived without actually determining the angles ( notice that the actions being a function of the energy are uniquely determined ) .
here we show that this can be achieved quite directly without having to deal with elliptic integrals : while in the elliptic case , for instance , the classical prescription , ( * ? ? ?
* sec.50 ) , would be to study the hamitonian as a function of the integral @xmath221 with @xmath222 defined in eq . ) .
as it is well known the full determination of the action - angle coordinates is much more difficult than the determination of the normal form of the hamiltonian , see for instance the `` conjecture '' in @xcite and its proof in @xcite .
therefore a direct determination of the normal form is worth investigating : and as it will appear here , the construction yields unexpected insights into our problem . consider first the normal form near the stable rotation .
in contrast with our first method , we do not start with the jacobi s coordinates in which the motion is linear .
the hamiltonian value @xmath58 will be written as ( see sec.[sec2 ] , @xmath61 ) : = b^2(1-r^2_a ^2 ) + r^2_a a^2 ^ 2 for simplicity of notation the label @xmath61 will be omitted . introducing @xmath223 ,
@xmath224 the hamiltonian becomes @xmath225 and the form @xmath226 is , if @xmath227 , & & dbd=1 + & & = _
n=0^_h+k = n -12h -12 k(-1)^k+h r^2h()^2(k+h ) + & & = _
n=0^p_n(r^2 ) ( ) ^n + * 0 * , p_n(r^2 ) + & & = ( -1)^n2nn _ h+k = n -12h -12k r^2h=2nn_h+k = n2hh2k kr^2h and @xmath228 is a form with @xmath72cohomology relative to @xmath229 .
therefore , see ( * ? ? ?
* proposition 1),@xcite , or theorems 1.1 , 1.2 in @xcite , there is a change of coordinates @xmath230 which , if @xmath231 and @xmath232 , keeps @xmath233 equal to @xmath234 and makes @xmath235 proportional to @xmath236 : let @xmath237 with @xmath238 chosen , see also sec.[sec4 ] , so that @xmath239 , _
i.e. _ as @xmath240 or , for instance , up to @xmath241 : and a normal form of the hamiltonian is u(p , q)= + ( x , r^2)where @xmath242 is @xmath243 times the inverse function to @xmath244 and @xmath245 ( with @xmath246 ) : this gives a normal form , because the variables @xmath247 are canonically conjugated to @xmath15 .
for instance ( and in agreement with eq . )
: it should be noted that the above analysis does not determine the canonical map into action - angle variables of the initial coordinates and , in particular , a further change of coordinates which transforms the plane @xmath248 into itself by rigidly rotating all circles centered at the origin by a radius dependent ( and even @xmath217 dependent ) angle will lead to a normal form .
for instance calling @xmath249 the change of variables @xmath250 is a canonical map , for any choice of @xmath251 , and it keeps the system in normal form .
since the hamiltonian is uniquely cast in normal form , the polynomials @xmath252 , coefficients of the expansion of @xmath253 in powers of @xmath168 , have an intrinsic significance .
it is therefore interesting that they seem to enjoy the property of having all zeros on the unit circle : we have checked this for @xmath254 : are really the _ zeros on the unit circle for all @xmath219?_. the location on the unit circle of the roots of our polynomials is certainly related to the property that the polynomials p_n(r^2)_h+k = n -12h -12k r^2h12^n_h+k = n 2h h 2k k r^2h appearing in the above expression for @xmath255 have all the roots on the unit circle .
the latter is a special case of the general result on polynomials , that we call `` polynomials with _
symmetric , positive and monotonic coefficients _ '' , of the form @xmath256 , with @xmath257 and @xmath258 for @xmath259 $ ] : the theorem can be found in @xcite , and it is described , for completeness in appendix [ appb ] .
this is analogous to the conjecture formulated after eq . , and it is an algebraic property which could , we think , also be seen as follows .
consider the polynomials q(z^2)=z^n__1,
,_n=1 e^_i , j j_ij(2)^2 z^_i_iwith @xmath260 .
then the coefficients @xmath261 of @xmath262 in @xmath263 have the form of a sum of @xmath264 exponentials of sums of some of the @xmath265 : the sums define quantities @xmath266 . by adjusting the constants @xmath265 the parameters @xmath267 for @xmath268 $ ] can independently take any value @xmath102 so that @xmath269 can be achieved by suitable choices of the @xmath260 .
therefore @xmath263 would have all zeros in the unit circle by lee - yang s theorem , @xcite , achieving a alternative proof of the above theorem .
the latter is a conjecture that can be checked rigorously for @xmath270 and we found some preliminary numerical evidence that it should also work at least for @xmath271 .
the proof for @xmath272 emerged from a discussion with a. giuliani ; he also has a proof solving the case @xmath273 and gave us the reference to chen s theorem .
the polynomials in @xmath0 appearing as coefficients of @xmath274 in the normal form eq . are sums of products of polynomials @xmath275 of degrees which add up to @xmath276 : an expression for the inverse function to @xmath277 can be found in terms of a `` tree expansion '' ( quickly sketched in appendix [ appc ] ) .
sums of products of symmetric polynomials with symmetrically decreasing positive coefficients is a polynomial with the same symmetry but it will not be , in general , decreasing .
hence the above theorem on the zeros of polynomials does not allow us any conclusion on the location of the zeros of the coefficients of the normal form : _ however we see empirically that the zeros are apparently still on the unit circle _ ( up to @xmath278 at least ) .
this is a special property of our polynomials @xmath279 because , in general , the inverse function to the map @xmath280 with @xmath281 a degree @xmath219 symmetric decreasing polynomial with positive coefficients ( hence with all zeros on the unit circle ) can be written @xmath282 but the @xmath283 _ do not have _ , in general , all zeros on the unit circle .
the analysis of the hyperbolic case is essentially identical : introducing the variable @xmath284 the form @xmath285 is and the normal form is related to the elliptic case and given by @xmath286 . where @xmath287 and @xmath288 .
\(1 ) the normal form in this section coincides with the one in sec.[sec5 ] because of the normal form _
uniqueness_. the agreement continues up to order @xmath289 : however the above analysis shows that this _ must hold _ to all orders : _
i.e. _ @xmath290 coincides with @xmath129 and our construction in sec.[sec2]-[sec5 ] determines explicitly one among the changes ( _ which are not unique _ ) of variables @xmath291 realizing the cohomological equivalence of the symplectic forms @xmath236 and @xmath292 .
( 2 ) it would be interesting to find a mechanical interpretation of the properties of the roots of the polynomials @xmath279 .
[ sec7 ] consider the elliptic motions , _
i.e. _ rotations close to the proper rotation about the axis @xmath293 . here
we compute the second pair of canonical variables corresponding to the second degree of freedom and described canonically by @xmath35 . setting @xmath294 , @xmath295 @xmath296 , + eq .
implies & & ^2=h^2 = 1k^2 ) + & & = 1k^2 + & & = 1k^2 = 1k^2 ( 1-s^2 ) = and this yields = + ( -1)^2therefore , @xcite , the expression of @xmath297 and of the uniformly rotating angle @xmath298 with angular velocity @xmath299 ( with @xmath300 ) are with @xmath301 . since the generating function of the canonical map integrating the motions has the form @xmath302 , ( *
* problem 4.11.5 ) , it is @xmath303 and the angle conjugated to @xmath3 must differ by a constant from @xmath304 and therefore it can be taken equal to @xmath304 , _ i.e _ : with @xmath305 , @xmath306 and @xmath307 . this completes the construction of the elliptic canonical coordinates conjugated to @xmath308 . the hyperbolic case , _ i.e. _ rotations close to the proper rotation around @xmath309 , is treated in the same way .
the case of the stable rotations close to the stable proper rotation around @xmath44 is also interesting and could be treated in a similar way leading to a complete parametrization of the motion along the unstable manifolds emerging from the rotations around the axis @xmath44 which correspond @xmath310 .
however the coordinates @xmath34 have a singularity on such rotations ( notice that the @xmath24 coordinates disappear from @xmath233 ) : hence new @xmath34 coordinates have to be introduced .
the simplest is to use the symmetry on the permutations of the inertia moments and make use of the equivalent hamiltonian @xmath311 in which the coordinates @xmath15 have a suitably different meaning but can be used to describe the motions close to proper rotations around the axes @xmath312 and @xmath46 ( respectively stable and unstable ) .
the analysis is then a repetition of the case @xmath60 with the hamiltonian eq ..
[ sec8 ] the method of sec.[sec6 ] can be applied to the case of the pendulum ( @xmath313 , @xcite ) or to the geodesic motion on a revolution ellipsoid ( @xmath314 , ( * ? ? ?
* ( 4.12.6 ) ) ) . the first case is considered in an attempt at proving the sign property of the normal form coefficients suggested in @xcite . in the second case it is of interest
because the normal form has the form of a series with coefficients polynomials in the ratio @xmath315 pf the equatorial axis @xmath140 to the polar axis @xmath316 and in this case the zeros of the polynomials do not seem to have special properties ( in particular they are not located on the unit circle ) . the normal form at the stable equilibrium of the pendulum is easily evaluated ( details are skipped ) to be @xmath317 if @xmath220 is the inverse function to @xmath318 with : c^2(z)_n=0^ _ n=0^^2 ( which is @xmath319 , see @xcite ) . and
to order @xmath320 the @xmath321 is : ( x)=x-12x^2 - 14x^3-x^4 -x^5-x^6 -x^7-x^8- ... this is consistent with the first few terms computed in ( * ? ? ?
* ( c.8 ) ) : due to different meaning of the symbols the relation between the function @xmath220 above and @xmath216 in the quoted reference is @xmath322 .
however constructing @xmath220 via the algorithm in appendix [ appc ] the sign property suggested in @xcite that @xmath220 might have all coefficients , but the first , negative does not follow ; see the final comments in appendix [ appc ] . in the case of the geodesic motion near the equatorial circle of the ellipsoid let @xmath323 be the equatorial and polar radii of an ellipsoid of revolution .
the hamiltonian for the geodesic motion is h=12 + 12 where @xmath3 is the polar component of the angular momentum and @xmath324 the polar angle , ( * ? ? ?
* problem 4.11.5 ) .
let @xmath325 , @xmath326 , @xmath327 , @xmath328 ; then @xmath329 and : therefore the normal form is @xmath330 with @xmath331 and @xmath220 the inverse function to @xmath332 with @xmath333 and p_n(r^2)1n+12nn _
n 12h-32k r^2hwhich is interesting as it shows that the location of the zeros on the unit circle is a peculiarity of the particular case of the rigid body .
_ nevertheless _ a numerical investigation for @xmath334 with an algebraic manipulator indicates , as a reasonable conjecture , that the zeros become asymptotically located on the unit circle : for @xmath219 large they are within an annulus of radii @xmath335 with : , c1 , c.88with @xmath336 , @xmath337 .
in this appendix we give details for the derivation of the formulae in sec .
[ sec3 ] .
let @xmath69 so that @xmath70 ; for @xmath60 eq .
describe the separatrix branches emerging from the proper rotation around the axis @xmath46 and for @xmath61 they describe the motion near the stable rotation around the axis @xmath68 . for @xmath71 close to @xmath72 and from the definitions of the jacobi elliptic integrals ,
see @xcite , it follows ( choosing the sign @xmath73 in the eq . , for instance ) : = sn(g_at , k_a)(t)=.and define @xmath338 in terms of @xmath339 so that @xmath340 as : h^2_a=1k_a^2,h^2_a= 1 - 1k_a^2 then ( @xcite for @xmath60 and @xcite for @xmath61 after correcting the obvious typo in @xcite and by @xmath74 ) : & & -4 mm b(t)== b_a= b_a cn(-iu_a h^-1_a , h_a ) , + & & -4 mm ^2(t)= = + & & -4 mm ( t)==g_a where the last relation is derived from the equations of motion and the previous expression for @xmath75 . also the second and third relations in eq . could be given expressions involving the moduli @xmath338 so that one could choose to use the formulae with @xmath86 for @xmath60 and with @xmath341 for @xmath61 .
remark also that , @xcite,@xcite , @xmath342 , and also that @xmath343 .
notice that @xmath76 and if @xmath77 it is @xmath78 , @xmath79 , @xmath80 and motions with this energy are `` like '' the motions close to the separatrix or , respectively , close to equilibrium of a pendulum . for @xmath81 , or for @xmath82
hence @xmath83 imaginary , all the above relations have to be interpreted via suitable analytic continuations .
some of the following formulae become singular as @xmath84 , but the singularity is only apparent and it will disappear from all relevant formulae derived or used in the following . denoting @xmath344 , @xmath345 , @xmath346 , @xmath347 @xmath348 , from the equation of motion ( and @xmath349 @xmath350 )
it follows identically for all @xmath351 : convenient choice of @xmath351 is @xmath352 with @xmath353 , _
i.e. _ @xmath354 and @xmath355 , then ( notice that @xmath356 ) : which follows by the addition formulae , @xcite , from the last of eq .. if @xmath88 , _ i.e. _ @xmath357 , it is @xmath358 and the integral can be evaluated by series , @xcite and @xcite , leading to eq . , . as remarked in sec.[sec3 ] @xmath94 with @xmath95 real and @xmath96 real , with @xmath95 real .
it is useful to check that if @xmath359 : from eq .
, via @xcite and @xcite and @xmath360 in eq .
, we get : & & _ a= _ k=0^_n=1^ _ , = 1 ( ) ( ( ax_a)^k e^ g_0,at _ a^)^2n-1 + & & = 1a _ k=1^_,=1(p_(ax_a)^k _ a^ ) + & & p _ e^ g_0,at , x_ap_+p_- therefore this implies : ( t)_a = a_k=0^_,=1 i a^k arctanh ( x^k_a p__a^ ) _ i.e. _ the second of eq .. likewise from the first of eq .
we derive : via @xcite , hence we obtain the first of eq .. via identities relating jacobian elliptic functions it would be possible to avoid considering functions that are defined by analytic continuations , for instance when the modulus @xmath104 or @xmath105 , and reduce instead to considering only cases with the `` simple '' arguments ( expressing functions by elliptic functions of the arguments @xmath361 , see for instance the first of eq . , when necessary ) : we avoid this as it would hide the nice property that the stable and unstable cases are in a sense analytic continuation of each other . summarizing the transformation of coordinates @xmath98 in eq .
is such that @xmath362 generates a solution of the equations of motion ( if @xmath363 ) .
the proof of the theorem is in the literature : first remark ( a classical result , see @xcite ) that if @xmath364 , and @xmath365 , then @xmath366 can only vanish if @xmath367 . because following word by word the proof in @xcite , write the polynomials @xmath368 & & a_0+a_1 z+
a_n z^n+a_n z^n+1+
+a_0 z^2n+1 , + & & a_0+a_1 z+
a_n-1 z^n-1 + 12 a_n z^n + 12 a_n z^n + a_n-1 z^n+1+
+a_0 z^2nas @xmath369 with @xmath370 or @xmath371 respectively and q_n(z)=_k=0^n _ k z^k , q_n^*(z)z^n z^n q_n(z^-1 ) , with @xmath372 ( with @xmath373 for @xmath374 and @xmath375 or @xmath376 ) .
hence @xmath377 with @xmath378 by the above remark , and @xmath379
. if @xmath380 is a zero for @xmath279 it is @xmath381 or = = 1,possible only if @xmath382 : because if @xmath383 then it is @xmath384 for @xmath385 and @xmath386 for @xmath387 ( and @xmath388 ) ; or if @xmath389 eq .
means @xmath390 .
to invert the relation @xmath391 let @xmath392 , see eq . , with @xmath393 a polynomial in @xmath0 of degree @xmath394 .
consider the equation z = x - f(z ) , with z = x+h(x),h(x)=- f(x+h(x ) ) and look for a solution @xmath395 .
let @xmath396 be a tree graph with a root and nodes @xmath351 into each of which merge @xmath397 branches oriented towards the root ; each node @xmath351 carries a label @xmath398 with the restriction @xmath399 unless @xmath351 is an end node in which case @xmath400 and @xmath401 .
such labeled tree graphs will form a family @xmath402 of trees identified by the rule that two trees are regarded as identical if they can be overlapped by pivoting the branches avoiding that any two branches overlap in the course of the process .
define the `` value '' , @xmath403 , of @xmath396 as @xmath404 with @xmath405 and @xmath406 .
notice that if he coefficients defining @xmath407 , @xmath408 , are all _ negative _ then @xmath409 for all @xmath396 .
then h(x)=_k=1^_(k )
val ( ) x^kas it is checked by induction ( for instance ) . in our case @xmath410 and the contributions of the value of a tree @xmath396 to the coefficient of @xmath274 is a product of symmetric polynomials @xmath411 with symmetrically decreasing coefficients and of total degree @xmath412 in @xmath0
; consequently the polynomials in @xmath0 being sums of products of polynomials enjoying the @xmath413 symmetry property do enjoy it as well .
however positivity and monotony is in general lost .
notice that eq .
, suggest that the coefficients of @xmath414 with @xmath197 are negative polynomials @xmath415 : in fact @xmath416 are either squares of @xmath417 times polynomials with positive coefficients for @xmath219 odd or , for @xmath219 even , @xmath418 times @xmath419 which after performing the multiplication become a polynomial with positive coefficients .
this property , however , does not follow from the above tree expansion and we wonder whether it continues to be true at the higher orders . | revisiting canonical integration of the classical solid near a uniform rotation , canonical action angle coordinates , hyperbolic and elliptic , are constructed in terms of various power series with coefficients which are polynomials in a variable @xmath0 depending on the inertia moments .
normal forms are derived via the analysis of a relative cohomology problem and shown to be obtainable without the use of ellitptic integrals ( unlike the derivation of the action - angles ) .
results and conjectures also emerge about the properties of the above polynomials and the location of their roots .
in particular a class of polynomials with all roots on the unit circle arises . |
this story starts with cousins ( see the review by kilkenny these proceedings ) & warren s ( 1963 ) suggestion of having found photometric variability in the f0v star @xmath0dor .
cousins ( 1966 ) himself confirmed unambiguously his discovery of the variability in @xmath0dor , the star that later would become the prototype of a new class of pulsating stars .
much later , cousins ( 1992 ) was able to detect two close periods in @xmath0dor : 0.7570d & 0.7334d .
balona et al . (
1996 ) undertook a very detailed photometric and spectroscopic follow - up study of @xmath0doradus and found one additional frequency .
moreover , they refined the frequency values to 1.32098d@xmath2 , 1.36354d@xmath2 , 1.47447d@xmath2 . for illustrative purpose ,
we show in fig .
1 the phase diagrams for the three detected frequencies of both the photometric and spectroscopic data of balona et al .
earlier on , balona et al . ( 1994 ) already proposed @xmath0dor to be a member of `` a new class of non - radial gravity - mode pulsators '' . with their 1996 paper , balona et al . left no longer doubt about the cause of @xmath0dor s variability : the radial - velocity variations excluded binarity , while spot models were also excluded since : 1 ) they required a too large differential rotation , 2 ) the periods were too stable , and 3 ) there was no phase difference between the radial - velocity and light maxima .
it therefore seemed unavoidable to conclude that @xmath0dor pulsates multiperiodically , with pulsation periods an order of magnitude longer than those of pressure modes .
meanwhile , several comparison stars for other variables turned out to have similar variability to @xmath0dor : * hd164615 ( burke et al . , 1977 ) ;
* hd55892 ( hensberge et al . , 1981 ) ; * four variables in ngc2516 ( antonello & mantegazza , 1986 ) ; * 9aur ( krisciunas & guinan , 1990 ) ; * hd111828 ( mantegazza et al . , 1991 ) ; * hd224638 & hd224945 ( mantegazza & poretti , 1991);. this led krisciunas & handler ( 1995 ) to publish a list of 17 candidate @xmath0dor stars . since the publication of the list by krisciunas & handler , several additional candidates were found from ground - based photometry ( henry , 1995 , 1999 ; breger et al . , 1997 ; kaye et al . , 1999 ; among others ) .
presently , an updated list is maintained by handler at http://www.saao.ac.za/@xmath3gerald/gdorlist.html .
it should be noted that the periodic variability of only very few stars is illustrated by phase plots of the quality shown in fig .
[ fig : gdor ] . for most stars
the scatter is considerably larger .
the cause of the non - radial gravity modes is still a matter of debate .
guzik et al .
( 2000 ) have proposed an excitation mechanism , but this certainly is not generally accepted ( see , e.g. , wu 2002 ; lffler 2002 ) .
it is therefore of importance to find as many members as possible and to derive their basic physical stellar parameters , besides their frequencies and mode identification .
one of the early ideas about the general properties of @xmath0dor stars is that such pulsators would be young .
this suggestion was based on the fact that circumstellar matter was discovered around some member stars , among which is @xmath0dor itself .
inspired by this proposal , searches for @xmath0dor stars in clusters of different ages were undertaken .
krisciunas & patten ( 1998 ) discovered 2 @xmath0dor candidates ( out of 15 stars ) in m34 ( age@xmath4195myr ) , while zerbi et al .
( 1998 ) came up with 8 candidates ( out of 44 stars ) in ngc2516 ( age@xmath4107myr ) .
an enormous effort in this respect was done by martin ( 2000 ) , who checked 149 stars in 9 clusters .
she found only 3 @xmath0dor stars : 2 in the pleiades ( age@xmath483myr ) and 1 in coma ber ( age@xmath4490myr ) .
the conclusion therefore must be that current studies do not reveal a clear relationship between age and the @xmath0dor phenomenon .
the flansa @xmath0dor campaign found its origin at the iau general assembly in kyoto in 1997 , when ca convinced le to come to leuven for a scientific visit to search in detail for new @xmath0dor stars in the hipparcos data base .
le s visit resulted in a chain of initiatives with respect to the study of @xmath0dor stars , which we highlight in this paper .
first we report on the status of the flansa @xmath0dor campaign in section3 , after having highlighted the contribution into this field by the hipparcos mission ( section2 ) .
the analyses of the time series are described in section4 , while an attempt to identify the modes in two stars is explained in section5 .
we end this paper with future prospects and additional current studies of @xmath0dor stars at the institute in leuven and worldwide .
we recall that the hipparcos photometry provides a homogeneous all - sky survey down to @xmath5 magnitude 7.3 - 9.0 ( depending on the galactic latitude and the spectral type of the observed star ) .
the hipparcos satellite took 13 million photometric measurements of about 118000 stars over a period of 3.3 yr ( 52000 stars are from the survey and 66000 stars were selected objects ) .
the hipparcos photometry proved to be very useful to reveal large numbers of new variables of many different kinds ( see esa 1997 : fields h6 , h49 - 54 of the main catalogue and volumes 11 and 12 ) . in specific cases
the number of known variable stars increased dramatically , e.g. the number of slowly pulsating b stars increased nearly by a factor 10 ( waelkens et al .
1998 ) .
hipparcos has rather good detection capabilities for determining correctly the characteristic periods of @xmath0dor stars ( 0.5 - 3 d ) .
the sampling does not suffer from the alias problems of earth observations for such stars .
indeed , the spectral windows of hipparcos and ground - based data show very different patterns .
however , the precision of photometric individual measurements degrades quite rapidly for fainter stars ( although more time was allocated to observe them ) , impeding the unambiguous variability detection of fainter objects .
another difficulty with the hipparcos data occurred specifically for @xmath0dor stars , since their variability behaviour is not always strictly periodic .
this , in combination with the rather low density of data points of the hipparcos sampling , prohibited an easy search for new candidate members of the class .
for instance , when we compare the bona - fide member list ( which contained 13 stars in the version of 2000 ) and the hipparcos catalogue , we find that all stars listed were measured by hipparcos and were classified as follows : 4 stars are in the variability annex as periodic variables , one star as unsolved variable , three stars are flagged as microvariable , four are without classification and one flagged as constant .
a remark worth making is that the four stars in the periodic variables of the variability annex have periods at the short end of the interval 0.5 - 3 d. eyer ( 1998 ) and eyer & grenon ( 1998 ) did a selection of @xmath0dor candidates , which was made from the published periodic variable star annex from the hipparcos catalogues ( volume 11 and 12 , esa 1997 ) .
this preliminary study was based on stars which had precise parallax and colour @xmath6 . in the hr diagram , there was clearly at the cool edge of the region occupied by the @xmath7scuti stars a very clumped group of stars with periods and amplitudes compatible with the @xmath0dor stars and thus these were declared good candidate @xmath0dor stars .
aerts , eyer & kestens ( 1998 ) , in an effort to be more general and to settle better the case of membership , performed a multivariate discriminant analysis by using geneva photometric indices in order to determine the physical properties of the stars . as a supplementary test , all stars were checked against multiperiodicity .
the stars were drawn from the periodic variable star annex .
the two first studies listed the name of 29 stars identified with hipparcos photometry ( three being redundant with the list of handler & krisciunas 1997 ) .
because of the sometimes irregular nature of @xmath0dor light curves , candidates could have been rejected from the periodic variable annex by the hipparcos groups analyzing and producing the catalogues . handler ( 1999a ) decided to go through the hipparcos data once more , but looked at the periodic variability annex as well as at the unsolved catalogues .
he also widened the criteria of acceptable spectral types .
his analysis permitted the addition to the list of 31 non - redundant prime candidates .
we recall here that only 6 years ago , the number of total objects was 17 ( 6 bona fide members and 11 candidates krisciunas & handler 1995 ) . as we can see from the successive lists provided in publications and/or on the internet
, this field has undergone intense studies . at the time of this oral presentation
there were 13 bona - fide members listed by handler , and the number of candidates amounted 122 .
we assisted in the explosion of the number of candidates and therefore undertook the initiative of settling the origin of the variability for each candidate @xmath0dor star .
this somewhat unrewarding task of removing objects from a list of potentially very exciting stars requires extensive new photometric campaigns .
the institute of astronomy in leuven decided to undertake this effort .
c. aerts took the initiative of applying for several weeks of observing time in south africa in the framework of the flansa project .
l. eyer became the driving force of the planning and exploration of the observing time allotted in sutherland .
at the same time the 1.2-m swiss at la silla was used to monitor each of the candidates spectroscopically .
furthermore the leuven 1.2-m telescope mercator ( a twin of euler ) at la palma has become operational ( since a few months prior to this writing ) and is currently used to monitor the northern @xmath0dor candidates photometrically in the seven - colour geneva system .
we here report on the results derived from the data gathered in south africa .
the time allocated for this project was 15 weeks spread over one year ( from 1999 october to 2000 september ) in order to observe 35 stars on the 0.5-m telescope located in sutherland .
the observations were done in the @xmath8 , @xmath5 , @xmath9 bands in the sequence @xmath10 ( 30-s exposures in each filter ) and then sky measurements in @xmath11 ( 10 s in each filter ) .
the whole sequence for observing an individual star took about 3 min and 50 s. the photometry obtained is differential .
two stars were chosen as comparison stars for each programme star .
their selection was done using hipparcos epoch photometry .
particular attention was paid to remove stars with peculiar chemical composition . for those stars
the variation of the luminosity is a function of the wavelength .
more precisely , for the cooler ap stars , the variations can be in anti - phase in blue and red light and the signal is washed out in a broad band filter like the @xmath12 filter . therefore the star appears to be a constant .
the physical reason of this is that the blocking effect is more severe in the blue , while the re - emission is in the red for the cool ap stars ( in the case of hot ap stars the blocking effect is in the ultraviolet and the backwarming in the visible ) .
the comparison stars were generally chosen to be at a distance less than 7@xmath13 and of similar magnitude .
furthermore , one comparison star was chosen to be slightly bluer than the programme star and the other slightly redder .
the measurements consisted of three block sequences of @xmath14 where @xmath15 is the programme star and @xmath16 and @xmath8 are the comparison stars .
typically a star was measured 2 - 4 times per night .
however the sampling was sometimes denser with short sequences @xmath17
(where @xmath15 is the programme star and @xmath18 the two comparison stars ) in order to be able to distinguish periods of @xmath7sct stars from periods of @xmath0dor stars .
some tests were also done to take photometry only in two bands at the beginning or end of the night ( before and after the astronomical night ) focusing on only one star . the campaigns in sutherland suffered from relatively bad weather .
normally the fraction of photometric nights is about 55% , while we had about 4045% .
there were a few exceptional and rare events : * the tail of the bad weather causing the tragic flood in zimbabwe surprisingly reached sutherland , * a beautiful aurora occurred , which had the sad effect of rendering void a large part of a night .
the time scale of the light variability caused by this aurora is short as seen in fig .
[ fig : aurora ] .
we recall here that sutherland is not very far south ( latitude is @xmath19 ) .
we give below by affiliation the list of observers , with ( in parentheses ) the number of weeks observed . from the katholieke
universiteit leuven ( kul ) : l. eyer ( 6 ) , p. de cat ( 3 ) , from the south african astronomical observatory ( saao ) : f. van wyk ( 2 ) , g. handler ( 1 ) , from the university cape town ( uct ) : m. mller ( 2 ) , m. hempel ( 1 ) .
the people having no experience with the 0.5-m received an introduction by either d. kurtz or g. handler .
the initial target list contained thirty - seven stars .
however , at the beginning of the observing run two stars were removed from the list thanks to the cross - check of the list by g. handler ( 1999b ) .
their @xmath5 magnitudes range from 5.3 to 9.5 .
five stars needed to be observed with neutral density filters .
the spectral types range from a1 to f4 , but most stars are classified as f0 or f2 . altogether 105 stars ( programme stars and comparison stars ) were observed during the observing run .
finally we note that twenty - six of the stars are being measured with the echelle spectrograph coralie attached to the swiss euler telescope located at la silla .
the spectral analysis is still in progress , so we will not refer to it here .
thirty stars were analyzed in total in the following section .
the number of observations per star varies between 14 and 97 in the @xmath5 band .
seventeen stars were from the lists of eyer ( 1998 ) and aerts et al .
( 1998 ) , eleven from handler ( 1999a ) , one from the bona fide member list , and one from paunzen et al .
the analysis of the time series was performed by f. bouckaert in the framework of his master thesis in leuven and was supervised by j. cuypers and c. aerts . for those stars which are also present in the geneva data base ,
a general comparison was done with geneva photometry . for all stars ,
the variability in the south african data was cross - checked also with the hipparcos data and with results from the literature , whenever possible .
there is a point worth mentioning here : in simbad the origin of the discovery of the variability is not reported and sometimes it is difficult to trace back the source of the claimed variability and subsequently check more carefully the data .
the period searches were done with different programs and methods , both based on the pdm technique and on fourier analysis .
moreover , we used the programme developed by j. cuypers which fits simultaneously several frequencies .
special emphasis was put on the search for multiperiodicity for every star .
the stars were classified in several categories : multiperiodic @xmath0dor stars , singly periodic @xmath0dor stars , non-@xmath0dor stars ( @xmath7sct stars , eclipsing binaries ) and dubious cases .
we now sum up the results for each of these categories .
our analysis resulted in the classification of 9 targets as multiperiodic @xmath0dor stars : hd12901 ( hip9807 ) , hd34025 ( hip24215 ) , hd48501 ( hip32144 ) , + hd65526 ( hip39017 ) , hd110606 ( hip62105 ) , hd112685 ( hip63372 ) , + hd135828 ( hip74825 ) , hd206481 ( hip107443 ) , and hd209295 ( hip108976 ) .
as an illustration , we present in fig .
[ fig : multiper ] the data for the star hd112685 .
it is clearly multiperiodic , with the periods 0.623d ( main period ) , 0.349d and 0.737d .
although we have far fewer data points , the quality of the phase diagrams is similar to the one of the prototype shown in fig .
[ fig : gdor ] .
the hipparcos periodic variable annex quotes a period of 0.600d for this object .
the phase diagrams of the eight other stars are similar to the ones shown here for hd112685 .
there are 8 stars classified as monoperiodic @xmath0dor stars : hd10167 ( hip7649 ) , hd14940 ( hip11192 ) , hd26298 ( hip19383 ) , hd40745 ( hip28434 ) , hd111709 ( hip62774 ) , hd112934 ( hip6349 ) , hd187028 ( hip97590 ) , and hd197451 ( hip102329 ) .
one might be tempted to conclude that about half of the @xmath0dor stars turn out to be monoperiodic .
however , we note that the stars of small amplitude will have a tendency to be classified as monoperiodic @xmath0dor stars for statistical reasons , particularly when a relatively small amount of data is available .
the monoperiodicity of our targets does not necessarily imply that the fraction of multiperiodic @xmath0dor stars is effectively higher at higher amplitude .
hd41448 ( hip28778 ) , hd81421 ( hip46223 ) , and hd85964 ( hip48580 ) are thought to be eclipsing binaries .
this conclusion was drawn because of the difference in depth of successive minima in the light curve , which is typical for eb type eclipsing binaries .
the phase plots of the three stars are shown in the fig .
[ fig : eclbin ] .
handler ( 2001 ) would classify the star hd41448 as a @xmath0dor star , though .
its light curve does indeed also support that interpretation since it shows a clear sign of multiperiodicity .
it remains to be confirmed if the unequal minima are true , or merely a consequence of the multiperiodicity .
it may very well be that this star is a binary with a @xmath0dor star as a component .
the spectra currently taken by coralie will probably resolve the ambiguity .
there are two objects classified as @xmath7sct stars : hd125081 ( hip69848 ) and hd181998 ( hip95358 ) .
the spectrum for the former star is shown in fig .
[ fig : deltascuti ] .
the main frequency is 6.4941d@xmath2 . for the latter object
our analysis is in contradiction with the conclusion of handler ( 1999b ) .
we find evidence for a frequency of 6.742d@xmath2 , while handler lists a typical @xmath0dor frequency of 0.7496d@xmath2 .
very probably both frequencies are aliases of each other . for three additional stars
our study did not lead to firm results , mostly because of a lack of data .
this concerns hd171813
( hip91412 ) , hd197187 ( hip102217 ) , and hd216910 ( hip113402 ) .
for the latter two stars our analysis does not confirm the previously published periods found in hipparcos ( those two stars also have few measurements ) .
there is one star which shows longer term behaviour in our data : hd27377 ( hip20036 ) for which we find a frequency of 0.100d@xmath2 .
this result is in conflict with handler ( 1999a ) , who reports the two frequencies 0.3511d@xmath2 and 0.548d@xmath2 from the hipparcos data .
additional data are needed to understand the variability pattern for this object .
for 4 additional studied stars no conclusions were reached . all the results presented here will be reviewed and published with the addition of coralie spectra which are currently being gathered and analysed .
special care will be taken for the claim of eclipsing binaries and for the distinction between ew eclipsing binaries and monoperiodic @xmath0dor stars .
for two of the target stars , we also have extensive geneva data at our disposal .
as this leads to amplitudes at seven different wavelengths , we decided to try to identify the pulsation modes in these two @xmath0dor stars .
this work was performed by m. van loon in the framework of her master s thesis and was supervised by c. aerts .
the goal of the work is to identify the non - radial pulsation modes thanks to the different behaviour of the pulsation as observed through different filters .
this photometric method is able to discriminate between different ( low ) degrees @xmath20 but can not derive the azimuthal number @xmath21 . for a review of the method of photometric mode identification we refer to garrido ( 2000 ) .
although the method has severe limitations because of the inherent different results introduced by using different atmosphere models ( see the discussion by garrido ) , we applied it to the two stars to get a crude idea about the pulsational degree .
as there exists currently no real overall accepted excitation mechanism , any estimate of the observed degree is welcome for as many stars as possible .
only two stars were considered , hd12901 and hd48501 , since these are the only ones for which we have data in many filters .
they were measured in the seven filters of geneva photometry ( @xmath22 , @xmath8 , @xmath23 , @xmath24 , @xmath5 , @xmath25 , @xmath26 ) .
these two stars were already discussed in the article by eyer & aerts ( 2000 ) , who classified them as new @xmath0dor stars .
bouckaert found three frequencies for each of the two stars : 1.21564d@xmath2 , 1.39595d@xmath2 , and 2.18374d@xmath2 for hd12901 and 1.29054d@xmath2 , 1.09408d@xmath2 , and 1.19924d@xmath2 for hd48501 .
the physical parameters of both stars were determined from the calibration by knzli et al .
( 1997 ) , making use of the stellar models published by schaller et al .
( 1992 ) and schaerer et al .
they are displayed in table [ eyertab1 ] .
we remark that the stars are nearly twins , except for their metallicity .
indeed their temperature , gravity , mass , and radius are equal within the uncertainty induced by the method of determination .
the time series analysis reveals also that the detected frequencies are rather similar , as well as their amplitude ratios .
one expects therefore , their pulsation modes to be of similar nature as well .
.[eyertab1 ] physical properties of two @xmath0dor stars [ cols="^,^,^,^,^,^,^",options="header " , ] the method of heynderickx et al .
( 1994 ) was used for the mode identification . with this method ,
the amplitude ratios with respect to the @xmath22 filter are concerned .
the choice of the @xmath22 filter is arbitrary , but was chosen in this method because it was designed originally for applications to @xmath27cep stars , which have by far the largest amplitude in the bluest filter .
the main hypotheses and ingredients of the method are : * the unknown non - adiabatic effects are taken into account by a free parameter , which is determined such that the difference between observed and theoretical amplitudes is minimal ( same as in the method by garrido ) , * rotational effects are completely neglected , * the phase difference between the light curves of different bands is zero ( unlike the method by garrido ) .
we prefer to make use of this method , as it uses the variability in seven filters ( while garrido s method only makes use of one colour and one phase difference ) . as it turned out , for both stars
, we observe no phase difference at all in the seven filters .
this is quite a surprise , since one expects to have severe phase differences due to non - adiabaticity .
however , the data are of very high quality and leave no doubt that the phase differences are zero . in that respect , it is better to use as many amplitude ratios as possible to derive the degree of the mode .
the work of m. van loon consisted of several steps : 1 . she translated the code of heynderickx , which was written in pascal , to c@xmath28 , 2 .
she included improved numerical interpolation tools to calculate the derivatives of the functions ( such as the flux and the limb darkening function ) with respect to the effective temperature and the gravity , 3 .
she cross - checked if her results were compatible with those by heynderickx for a large number of @xmath27cep stars , 4 .
she added to the code the grid of stellar atmosphere models for the temperature range of the @xmath0dor stars .
for all three frequencies in both stars , the study favours very strongly the solution @xmath1 .
all other @xmath20-values lead to amplitude ratios very different from the observed ones .
these results remain to be confirmed , of course , preferably by spectroscopic studies , but they do help to discriminate between future excitation models for the @xmath0dor stars .
we remark that the models by guzik et al .
( 2000 ) imply the excitation of mainly @xmath29 modes , which is thus compatible with our observational results .
there are a number of future projects in the pipeline , both observational and theoretical .
we mention here that a large observational campaign is being organized and performed by p. mathias & collaborators .
it consists of photometric and spectroscopic observations in the northern hemisphere .
this , together with our spectroscopic study with euler from la silla for the southern stars and our photometric study with mercator from la palma should allow us to constrain the observational behaviour of the @xmath0dor stars with high precision during the forthcoming years .
moreover , the observational results will allow a critical confrontation with the theoretical works that are currently being worked out ( lffler 2002 ; wu , 2002 and maybe others ) .
we are thankful for the warm and generous hospitality of the researchers and staff in cape town and sutherland .
aerts , c. , eyer , l. , kestens , e. 1998 , , 337 , 790 antonello , e. , mantegazza , l. 1986 , , 164 , 40 balona , l.a . , bhm , t. , foing , b.h . , et al .
1996 , mnras , 281 , 1315 balona , l.a . , krisciunas , k. , cousins , a.w .
1994 , mnras , 270 , 905 breger , m. , handler , g. , garrido , r. , et al .
1997 , a&a , 324 , 566 burke , e.w .
jr , burke , e.w .
iii , lady , s. 1977 , ibvs , 1279 cousins , a.w .
1966 , observatory , 86 , 69 cousins , a.w . 1992 , observatory , 112 , 53 cousins , a.w . ,
warren , p.r .
1963 , mnassa , 22 , 65 esa 1997 , the hipparcos and tycho catalogues , esa sp-1200 eyer , l. 1998 , ph.d .
thesis , geneva university eyer , l. , grenon , m. 1998 , in : f .-
l.deubner , j. christensen - dalsgaard , d.w .
kurtz ( eds . ) , new eyes to see inside the sun and stars , iau symp .
185 , dordrecht : kluwer academic publishers , 291 eyer , l. , aerts , c. 2000 , , 361 , 201 garrido , r. 2000 , in : m.breger & m.h.montgomery ( eds . ) , delta scuti and related stars , @xmath30 vienna workshop in astrophysics , aspcs , 210 , 67 guzik , j. , kaye , a. b. , bradley , p.a . ,
cox , a.n . , neuforge , c. 2000 , apj , 542 , l57 handler , g. , krisciunas , k. 1997 , delta scuti star newsletter 11 , 3 handler , g. 1999a , mnras , 309 , l19 handler , g. 1999b , private communication handler , g. 2001 , private communication henry , g. 1995 , robotic telescopes , aspcs , 79 , 44 henry , g. 1999 , pasp , 111 , 845 hensberge , h. , deridder , g. , doom , c. , et al .
1981 , a&as , 46 , 151 heynderickx , d. , waelkens , c. , smeyers , p. 1994 , a&as , 105 , 447 kaye , a.b . ,
henry , g.w . ,
fekel , f.c . , hall , d.s .
1999 , mnras , 308 , 1081 krisciunas , k. , guinan , e.f .
1990 , ibvs , 3511 krisciunas , k. , handler , g. 1995 , ibvs , 4195 krisciunas , k. , patten , b.m .
1998 , dssn , 12 , 25 knzli , m. , north , p. , kurucz , r.l . , nicolet , b. 1997 , a&as , 122 , 51 lffler , w. 2002 , in : c.aerts , t.bedding & j.christensen-dalsgaard ( eds . ) , radial and nonradial pulsations as probes of stellar physics , aspcs , in press mantegazza , l. , poretti , e. 1991 , ibvs , 3690 mantegazza , l. , poretti , e. , antonello , e. 1991 , ibvs , 3612 martin , s. 2000 , phd thesis , university of granada paunzen , e. , maitzen , h.m .
1998 , a&as , 133 , 1 schaller , g. , schaerer , d. , meynet , g. , maeder , a. 1992 , a&as , 96 , 269 schaerer , d. , meynet , g. , maeder , a. , schaller , g. 1993 , a&as , 98 , 523 waelkens , c. , aerts , c. , kestens , e. , grenon , m. , eyer , l. 1998 , , 330 , 215 wu , y. 2002 , in : c.aerts , t.bedding & j.christensen-dalsgaard ( eds . ) , radial and nonradial pulsations as probes of stellar physics , aspcs , in press zerbi , f. , mantegazza , l. , campana , s. , antonello , e. 1998 , pasp , 110 , 804 eyer in the light of hipparcos data , pierre north checked the @xmath31 values given by the photometric calibration by knzli et al .
( 1997 ) for a sample of @xmath7sct stars . indeed
, the estimates of @xmath34 seem affected by a systematic trend and to be too large by about @xmath35 . the source of this bias
is neither understood nor identified but is under investigation .
aerts i also point out that , in general , the determination of the effective temperature is much more accurate than the one of the gravity .
this is true for every photometric system . for b stars ,
the typical uncertainty of @xmath31 is 0.2 0.5 , even for bright stars .
balona 1 ) how can you tell if a singly - periodic star may not be a close binary rather than a @xmath0dor star ? + 2 ) you mentioned that two @xmath0dor stars do not show a variation of phase with waveband .
this is rather strange since one would expect such a variation owing to non - adiabatic effects .
can you comment ?
we also looked at the colour variations , but for unambiguous answers we will have to wait for the spectroscopy results .
+ 2 ) yes , i agree , that is what we expected .
but that is not what the stars are telling us .
we clearly find that there is no phase difference for the two stars , of which we have numerous high - quality geneva photometry .
handler with the increasing number of @xmath0dor candidates ( 17 in 1995 , @xmath36 140 in 1999 ) and the increasing interest in these stars it has become very difficult to keep the @xmath0dor database up to date .
i d therefore like to encourage everybody to communicate their new discoveries to me for immediate inclusion in the @xmath0dor master list . simply send an e - mail to [email protected]
. thank you for your help to keep the database up to date . | we report on the results from a large photometric campaign on thirty five @xmath0dor star candidates undertaken at the institute of astronomy of the university of leuven in the framework of the flanders south - africa grant .
an overview of the data , as well as the results of the analysis of the obtained time series are presented , the main conclusion being that nine stars are thought to be multiperiodic @xmath0dor stars and eight monoperiodic .
we also performed a photometric mode identification for two stars of the sample by comparing the amplitude ratios in the different passbands of the geneva photometric system .
both stars seem to pulsate in non - radial modes of degree @xmath1 . |
we present studies of baryon acoustic oscillations ( bao ) with sloan digital sky survey ( sdss ) data release dr12 @xcite .
part i of this ongoing study is ref .
we refer the reader to part i for the notation and methods .
this is an outline of the study : \(i ) we define and measure bao observables @xmath0 , @xmath1 , and @xmath2 that do not depend on any cosmological parameter , see part i @xcite . from each of these observables
we obtain the bao correlation distance @xmath3 with its respective dependence on the cosmological parameters .
we present many redundant measurements with different galaxy selections , i.e. different background fluctuations , to gain confidence in the results .
\(ii ) we use the measured bao distances , and the correlation angle @xmath5 of the cosmic microwave background ( cmb ) , as an uncalibrated standard ruler to constrain the correlated parameters @xmath6 and @xmath7 .
the cosmological parameters @xmath9 , @xmath10 and @xmath11 drop out of this analysis .
\(iii ) finally we use the measured bao distances and @xmath5 as a calibrated standard ruler to further constrain @xmath6 with external measurements of @xmath9 and @xmath10 . the results of this analysis are compared with the final consensus results corresponding to the dr12 data @xcite . [
cols="^,^,^,^,^,^,^,^,^",options="header " , ]
the results of these studies are : \(i ) we define and measure bao observables @xmath0 , @xmath1 , and @xmath2 that do not depend on any cosmological parameter . from each of these observables
we obtain the bao correlation distance @xmath3 in units of @xmath12 with its respective dependence on the cosmological parameters .
it is difficult to distinguish the bao signal from the background fluctuations due to the clustering of galaxies . to gain confidence in the results we repeat the measurements many times with different galaxy selections to obtain different background fluctuations .
the measured bao observables in tables iii and iv of part i @xcite and table [ cs ] are the main result of these studies . these measurements in combination with independent observations
constrain cosmological parameters .
\(ii ) from the bao measurements alone we obtain the constraint @xmath13 for constant @xmath7 , or @xmath14 when @xmath7 is allowed to depend on @xmath8 as in scenario 4 .
see table [ bao_fit ] for fits in several scenarios .
\(iii ) from the bao measurements plus @xmath15 from the cmb we obtain the constraints @xmath16 and @xmath17 when @xmath7 is allowed to vary as in scenario 4 . see tables [ bao_thetamc_fit_18 ] and [ bao_thetamc_fit_17 ] for fits in several scenarios .
the cosmological parameters @xmath9 , @xmath10 and @xmath11 drop out of this analysis .
\(iv ) from the bao measurements plus @xmath15 from the cmb plus external measurements of @xmath18 defined in eq . (
[ a ] ) we obtain @xmath19 for @xmath20 , and @xmath21 for @xmath22 , when @xmath7 is allowed to vary as in scenario 4 . for more details
see table [ calibrated_bao ] .
note that the external constraint from @xmath18 reduces the uncertainty of @xmath6 .
\(v ) the data is consistent with the constraint @xmath23 or the constraint @xmath7 constant , but there is some tension when both constraints are required .
these tensions are presented in table [ tension ] and in fig .
[ o_de_z_all ] . the measured excess of @xmath7 for @xmath24 is not understood .
\(vi ) we note that the final consensus results of dr12 data @xcite also show this tension , see last column in table [ dr12 ] which is in agreement with fig .
[ o_de_z_all ] .
finally we note that the bao measurements in this study are in agreement with @xcite : compare table [ dr12 ] with tables iii or iv of part i @xcite or table [ cs ] .
the two studies are complementary .
funding for sdss - iii has been provided by the alfred p. sloan foundation , the participating institutions , the national science foundation , and the u.s .
department of energy office of science .
the sdss - iii web site is http://www.sdss3.org/. sdss - iii is managed by the astrophysical research consortium for the participating institutions of the sdss - iii collaboration including the university of arizona , the brazilian participation group , brookhaven national laboratory , carnegie mellon university , university of florida , the french participation group , the german participation group , harvard university , the instituto de astrofisica de canarias , the michigan state / notre dame / jina participation group , johns hopkins university , lawrence berkeley national laboratory , max planck institute for astrophysics , max planck institute for extraterrestrial physics , new mexico state university , new york university , ohio state university , pennsylvania state university , university of portsmouth , princeton university , the spanish participation group , university of tokyo , university of utah , vanderbilt university , university of virginia , university of washington , and yale university . | we define baryon acoustic oscillation ( bao ) observables @xmath0 , @xmath1 , and @xmath2 that do not depend on any cosmological parameter . from each of these observables
we recover the bao correlation length @xmath3 with its respective dependence on cosmological parameters .
these bao observables are measured as a function of redshift @xmath4 with the sloan digital sky survey ( sdss ) data release dr12 . from the bao measurements alone , or together with the correlation angle @xmath5 of the cosmic microwave background ( cmb ) , we constrain the curvature parameter @xmath6 and the dark energy density @xmath7 as a function of the expansion parameter @xmath8 in several scenarios . these observables are further constrained with external measurements of @xmath9 and @xmath10 .
we find some tension between the data and a cosmology with flat space and constant dark energy density @xmath7 . |
we will call _ bipartite _ a knot that can be represented by a _ matched _ diagram , that is , a diagram whose crossings are split in pairs of the types depicted in fig . [
fig : p1 ] .
the pairs in the upper line are said to be _ positive _ , those in the lower line , _
negative_. note the signs of the pairs do not change when the orientation on the knot is reversed .
note , moreover , that if the crossings of an unoriented knot are split into matched unoriented pairs , then , introducing any orientation , we always get counter - directed pairs shown in fig .
[ fig : p1 ] .
* examples .
any rational knot has a matched diagram , because any rational number can be represented by a continued fraction with even ( positive or negative ) denominators ( see @xcite ) .
the standard diagram of the knot @xmath0 ( which is not rational ) can be easily transformed to a matched form : \3 .
we managed to find matched diagrams for all table knots with up to 8 crossings , save for the knot @xmath1 .
the problem raised by jozef przytycki in 1987 is to investigate which knots possess a matched diagram .
this question appears in the well - known collection `` open problems in topology '' maintained by rob kirby @xcite , as part of problem 1.60 .
more exactly , conjecture 1(a ) therein ( belonging to przytycki @xcite ) reads : `` there are oriented knots without a matched diagram '' .
as the reader understands , the word `` oriented '' can be here omitted without any loss of meaning .
this conjecture stayed open for 24 years , notwithstanding the effort of several excellent mathematicians , including its author and j. conway @xcite .
we give a positive solution to the conjecture , that is , demonstrate that some knots , e.g. pretzel knot with parameters @xmath2 , are not bipartite . in the next section ,
we introduce our main construction , which also explains the meaning of the word `` bipartite '' in this context .
consider a matched diagram of a knot @xmath3 . replace every matched pair of crossings by two parallel segments , directed as the knot and joined by a common perpendicular ,
see fig .
[ fig : p2 ] . the parallel segments
are then joined by the remaining fragments of the knot diagram into a simple closed line on the plane , straightenable into a circle , whereas the common perpendiculars become _ chords_. an example for the matched diagram of the knot @xmath0 mentioned above , is given in fig .
[ fig : md - cd ] , where the inner and outer chords are changed places : this leads to turning the knot diagram inside outside with respect to some point and does not alter the isotopy type of the knot .
chord diagram , height=151 ] the chord diagrams obtained in this way , are rather special : the set of all chords is split into two parts ( inner chords and outer chords ) , so that the chords in each part do not intersect between themselves , and the intersection graph of the whole diagram is bipartite .
this procedure is reversible : from a bipartite signed chord diagram one can reconstruct the knot diagram in a unique way ( see fig .
[ fig : cd - md ] ) . matched diagram , height=151 ]
a seifert surface @xmath4 of a knot is a compact oriented surface embedded in @xmath5 so that its boundary is the given knot . choosing a basis in @xmath6
, one can construct a matrix of the bilinear form @xmath7 , where @xmath8 is the linking number , and @xmath9 is a small shift in the positive direction along the normal of @xmath4 .
this matrix is called a seifert matrix of the given knot .
there is a standard procedure to construct a seifert from any diagram , using checkerboard coloring and seifert circles . for matched diagrams there
exists a different construction , which is crucial for our needs : it yields a seifert matrix of a special type , which , in turn , produces an alexander matrix with extraordinary properties .
[ lemma1 ] any bipartite knot has a seifert surface such that its seifert matrix has the form @xmath10 , where @xmath11 , @xmath12 , @xmath13 , @xmath14 are square matrices of the same size , @xmath11 is a unit matrix , @xmath12 is a zero matrix , and @xmath13 and @xmath14 are both symmetric integer matrices . consider a bipartite knot @xmath3 and its plane diagram , construct the chord diagram as indicated above .
start constructing the seifert surface from the inner circle of the chord diagram , out of which we cut out every chord together with a small open neighborhood and glue instead a double twisted band , so that the direction of the twists corresponds to the sign of the chord ( see fig .
[ fig : p3 ] ) .
so far the surface remains orientable , and its boundary follows the knot as much as it can .
now we must add the bands along the outer chords . here
one must be cautious , because simply connecting the ends of the two half - chords by two half - twisted bands results in an unorientable surface .
we will do as follows : first attach a band along each outer chord , then , around the middle of that band , we attach a perpendicular small band , which is twice twisted opposite to the sign of the chord .
we show this procedure on fig .
[ fig : p4 ] for one chord on big scale and on fig .
[ seifert ] for a whole seifert surface of a certain knot . in the latter picture
, the narrow twice twisted bands on the left and on the right should be though of as lying above the surface of the corresponding perpendicular wide bands .
the thick solid lines indicate the boundary of the seifert surface ; the four small dashed segments show parts of the visible contour of the surface which do not belong to its boundary .
the two sides of the seifert surface ( which is two - sided by definition ) are indicated by different shades of gray .
, width=226 ] let @xmath15 be the number of outer chords ( if this number is greater than the number of inner chords , we can turn the chord diagram inside outside to simplify computations ) . then
, as a basis of @xmath6 , one can take the set @xmath16 corresponding to the outer chords and shown in fig .
[ fig : p5 ] .
( left ) and @xmath17 ( right),scaledwidth=70.0% ] it follows that the seifert matrix for this seifert surface ( see @xcite ) has the form @xmath18 where @xmath12 , @xmath11 , @xmath13 , @xmath14 are matrices of size @xmath19 , @xmath11 is the unit matrix , @xmath12 is the zero matrix , and @xmath20 , @xmath21 .
it is clear that @xmath13 is a diagonal matrix with numbers @xmath22 on the diagonal ( the sign is inverse to the sign of the outer chord number @xmath23 ) .
the cycles @xmath24 and @xmath25 can be chosen so that not to have common points , if @xmath26 , therefore , @xmath27 and thus the matrix @xmath14 is symmetric .
this construction shows that , on a practical side , it is advisable to turn the diagram inside outside , if the number of outer chords is bigger than that of inner chords as we did before for the example knot @xmath0 .
the determinant of the alexander matrix @xmath28 is equal to the alexander polynomial of the knot @xmath3 ; it is an element of the ring of laurent polynomials @xmath29 $ ] , determined up to multiplication by invertible elements of the ring , that is , monomials @xmath30 .
the alexander matrix is not determined by the knot uniquely ; in fact , to any knot there corresponds a big family of alexander matrices related between themselves by a set of equivalence transformations which is well known ( see @xcite . in particular , even the size of the matrix @xmath31 is not invariant .
what is invariant , however , is the sequence of alexander ideals , the @xmath32-th ideal being defined as the ideal in @xmath29 $ ] generated by all minors of an arbitrary alexander matrix of size @xmath33 , where @xmath15 is the smallest among the number of columns and rows in @xmath31 ( see @xcite ) .
it is well known that the alexander polynomial can be rewritten in terms of the conway variable @xmath34 ; in general , this is no longer true about the generators of all alexander ideals . in the case of bipartite knots , we can prove a stronger assertion .
[ lemma2 ] if the knot @xmath3 is bipartite , then there exists a square integer matrix @xmath35 such that the matrix @xmath36 is an alexander matrix for @xmath3 ( here @xmath11 is the unit matrix ) .
consider the seifert matrix @xmath37 from lemma [ lemma1 ] .
put @xmath28 , multiply the left block column by @xmath38 , the second by @xmath39 , then interchange both columns . using the symmetry of @xmath13 and @xmath14 ,
we get : @xmath40 by a sequence of elementary transformations , we can make zero the upper right block of this matrix , using its right lower block : in doing so , we will be always adding polynomials @xmath41 to the elements of left upper block . in the end ,
the matrix will become @xmath42 to achieve our goal , it only suffices to prove one technical proposition .
[ lemma3 ] let @xmath43 $ ] be a set of ordinary polynomials and @xmath44 be the corresponding ideal in @xmath29 $ ] .
suppose that @xmath11 contains the binomial @xmath45 .
the the ideal @xmath11 is trivial : @xmath46 $ ] .
it is clear that @xmath47 .
then division gives @xmath48 , where @xmath49 are some integers .
then our ideal coincides with @xmath50 , where @xmath51 is the greatest common divisor of all @xmath52 s . expand the element @xmath53 in the new generators : @xmath54 .
then @xmath55 is divisible by @xmath53 .
therefore , @xmath56 and the ideal is trivial .
the last lemmas show that the alexander ideals of bipartite knots can not be arbitrary .
in particular , they are always generated by polynomials in @xmath57 .
let @xmath3 by a bipartite knot .
if the alexander ideal @xmath58 is non - trivial , then it can not contain the polynomial @xmath53 .
this is a direct consequence of lemmas [ lemma2 ] and [ lemma3 ]
. this condition immediately give a series of knots which are not bipartite .
rolfsen table knots @xmath59 are not bipartite . for the knot @xmath60 ,
also known as the pretzel knot with parameters @xmath2 , a detailed calculation of the second alexander ideal is available from @xcite . for other knots from the given list
, we borrowed the result from computer generated tables of knot atlas @xcite .
contrary to the universal tradition , we allow ourselves to raise the carpet and explain how we actually arrived at this solution .
it was clear to us from the beginning that the rational knots are all bipartite .
then we designed a procedure to very quickly compute the conway polynomial of a bipartite graph , starting from the corresponding signed intersection graph . looking through the table of all knots through 8 crossings , we managed to find the bipartite graphs that would give the same conway polynomials , and after one or two tries , using knotscape @xcite and knotinfo @xcite , we obtained a bipartite representation for the corresponding knots .
this worked for all knots , save for @xmath1 .
now , this is the only knot until 8 crossings with nontrivial second alexander ideal .
we looked at other knots with nontrivial second alexander ideal and found some that can not be expressed through the conway variable @xmath57 . on the other hand
, we devised a procedure to represent the alexander matrix of a bipartite graph in terms of @xmath57 .
after this work was finished , the second author ( m. sh . ) invented another argument showing that a knot with second alexander ideal @xmath61 , e.g. the table knot @xmath62 , can not be bipartite .
a separate publication is being prepared to this end . to summarize ,
we have presented a sufficient condition for a knot not to have any matched diagram .
we do not know , however , of any reasonable necessary condition in terms of alexander ideals .
the simplest knot which still stands our efforts , is @xmath1 . | we give a solution to a part of problem 1.60 in kirby s list of open problems in topology @xcite thus answering in the positive the question raised in 1987 by j.przytycki @xcite . |
the detection of neutral interstellar clouds at large @xmath0-distances ( mnch 1957 ) led to the hypothesis of a hot gaseous galactic halo ( spitzer 1956 ) . this high temperature ( @xmath1 k ) low volume density gas was proposed to confine the neutral clouds high above the galactic disk .
alternatively , a low temperature , almost neutral halo ( @xmath2 k ) was postulated by pikelner & shklovsky ( 1958 ) .
their model predicted emission lines of neutral species with velocity dispersions of 70due to turbulent motions . in the early fifties optical polarization studies revealed that the galaxy hosts an interstellar magnetic field , with a field strength of a few @xmath3 g .
this is , on the average , oriented parallel to the galactic disk .
several years later , radio continuum surveys ( beuermann et al . 1985 ) gave clear evidence for radio synchrotron emission originating at distances of a few kpc above the galactic plane .
this indicates that magnetic fields are also present within the galactic halo . already in 1966
parker pointed out that magnetic fields must be associated with gaseous counterparts .
uv absorption line measurement with the _ copernicus _ satellite showed the presence of highly - ionized species within the galactic halo .
however , the detection of an ubiquitous gaseous component was not easy to establish .
it took nearly 40 years , since spitzer s prediction .
rosat _ all - sky survey revealed the presence of an ubiquitous hot x - ray emitting plasma within the galactic halo .
pietz et al .
( 1998 ) and kerp et al .
( 1999 ) showed that the x - ray emitting halo gas is smoothly distributed across the whole sky .
they performed a quantitative correlation the new leiden / dwingeloo hi 21-cm line survey and the _ rosat _ all - sky survey .
these analyses gave evidence that the galactic halo is the brightest diffuse soft x - ray source .
also recently , evidence for an extended neutral galactic halo was presented by albert et al .
lockman & gehman ( 1991 ) found gas with a velocity dispersion of up to 35towards the galactic poles , which was interpreted as emission originating from a galactic halo .
kalberla et al .
( 1998a ) disclosed the existence of a pervasive component with a velocity dispersion of 60 from the leiden / dwingeloo survey ( lds , hartmann & burton 1997 ) . the x - ray as well as the neutral gas within the galactic halo are consistent with a hydrostatic equilibrium model of the milky way on large scales ( kalberla & kerp 1998 ) .
the x - ray data constrain the extent of the gaseous halo , while the high - velocity dispersion component determines the pressure balance of the equilibrium model .
according to this equilibrium model of kalberla & kerp ( 1998 ) , the galactic halo has a vertical scale height of 4.4kpc and a radial scale length of 15kpc .
these values are in agreement with those derived from the distribution of the highly - ionized atoms within the galactic halo published by savage et al .
( 1997 ) : @xmath4kpc , @xmath5kpc and @xmath6kpc .
kalberla et al .
( 1998a ) as well as savage et al .
( 1997 ) find that the observational data are fitted best by assuming a co - rotation of the galactic disk and halo in addition to a turbulent motion of the gas with an averaged velocity dispersion of 60 .
taking all in one , it is possible to construct a consistent model of the galactic halo , consisting of gas , magnetic fields and cosmic rays . in this paper
we discuss the evidence for a gaseous halo in some detail . in sect .
2 we describe observational evidence for the existence of soft x - ray plasma within the galactic halo . in sect .
3 we focus on observations and model calculations of the galactic disk / halo . in sect .
4 we discuss the stability of a multi - phase disk - halo system .
since the first detection of the diffuse soft x - ray background in 1968 ( bowyer , field and mack ) it was a matter of debate whether the radiation originates within the galactic halo or in the local vicinity of the sun .
because of the strong energy dependence of the photo - electric cross section on energy ( @xmath7 ) the soft x - ray emission is strongly attenuated by the ism on the line of sight . in the so called 1/4kev band the cross section is about @xmath8 ; thus , towards most of the high latitude sky we have at least an opacity of unity . with the _ rosat _ x - ray telescope it was possible to study the diffuse soft x - ray background emission on a high signal - to - noise ratio .
the angular resolution of 30 " was sufficient to reveal individual shadows of clouds in front of an x - ray emitting source . here , we like to compile the major findings of two recent papers on the galactic x - ray halo . both are based on the _ rosat _ all - sky survey data , both assume that the soft x - ray background radiation is a superposition of the local hot bubble , the galactic halo and extragalactic background emission . despite these similarities , they derive a totally different galactic halo x - ray intensity distribution .
what may cause these differences and what are the implications ? _
the data - set : _ snowden et al .
( 1998 ) analyzed _ rosat _ 1/4kev band data , while pietz et al .
( 1998 ) studied the x - ray intensity distribution of the _ rosat _ 3/4kev and 1/4kev energy band . because of the strong energy dependence of the absorption cross section , the 3/4kev emission is only attenuated weakly , while the 1/4kev photons are strongly absorbed by the same amount of interstellar matter .
for example a neutral hydrogen column density of @xmath9 attenuates the 3/4kev radiation by only 15% , while the same column density absorbs 75% of the 1/4kev photons .
this major difference indicates , that the 3/4kev is more appropriate to study the intensity distribution of a hot galactic halo gas . _
the plasma temperature : _ because of the moderate energy resolution of the _ rosat _ pspc detector , the practical way to determine the plasma temperature across the galactic sky is to evaluate energy - band ratios .
especially the 1/4kev to 3/4kev band ratio is very sensitive to variations of the plasma temperature .
a @xmath10})\,=\,6.0 $ ] plasma will emit most of the photons in the 1/4kev range , but a minor increase in temperature , for instance to @xmath10})\,=\,6.3 $ ] produces about an order of magnitude more 3/4kev photons than the @xmath11})\,=\,6.0 $ ] plasma .
snowden et al . ( 1998 ) derived a lhb temperature of @xmath10})\,=\,6.07 $ ] while the galactic halo temperature is lower and about @xmath10})\,=\,6.02 $ ] .
both plasma do _ not _ contribute significantly to the 3/4kev energy range .
in contrast to this finding , pietz et al .
( 1998 ) claimed that the temperature of the lhb is about @xmath10})\,=\,5.85 $ ] and that of the halo @xmath10})\,=\,6.2 $ ] .
the lhb plasma does not emit significant amounts of 3/4kev photons , while the halo plasma accounts for about half of the observed 3/4kev emission .
accordingly , the gap between the observed 3/4kev intensity and the known sources of diffuse 3/4kev radiation , already identified within the wisconsin survey ( mccammon & sanders 1990 ) , is overcome by the approach of pietz et al . _
the method : _ snowden et al .
fitted scatter - diagrams ( column density @xmath12 versus x - ray intensity @xmath13 ) . for this approach
it is necessary to assume an _ a priori _ plasma temperature to determine the photo - electric cross section of the ism .
snowden et al .
analyzed the _
1/4kev radiation splitted into the so - called r1 and r2 energy bands . in the second step , they derived the temperature of the plasma components , by evaluating the r1/r2 energy - band ratio .
the r1/r2 band ratio is only a weak function of temperature . using the same plasma temperatures as in the example above we find for @xmath10})\,=\,6.0 $ ] a ratio r1/r2 = 0.9 , while for @xmath10})\,=\,6.3 $ ] , r1/r2 = 0.6 .
the uncertainties within of the _ rosat _ r1 and r2 band are about 10% , accordingly the r1/r2-band ratio has a statistical uncertainty of about 20% .
thus , it is a difficult task to constrain the plasma temperature by analyzing the r1/r2-band ratio . in the final step , they derived the galactic halo intensity by inverting the radiation transfer equation . within the 1/4kev band
the product of @xmath12 and @xmath14 is close to or larger as unity , the quotient @xmath15 becomes uncertain , because the divisor is a small value .
the approach of pietz et al .
was totally different .
they found evidence for a large scale 3/4kev intensity gradient between the galactic center and the galactic anti - center direction . because the 3/4kev radiation is only to a minor fraction attenuated by the galactic interstellar matter towards the high latitude sky
, they attributed this gradient to a variation of the emissivity of the galactic halo plasma .
they modeled this 3/4kev gradient and optimized the shape of the galactic halo to fit the 3/4kev diffuse x - ray background intensity distribution .
the attenuation of the 3/4kev radiation is much weaker than in the 1/4kev energy range , but present .
pietz et al .
estimated the galactic halo plasma temperature by the spectral fit of the diffuse x - ray background data of a deep pointed pspc observation .
this is the main source of uncertainty , because it is unknown whether this pointed observation is towards a representative direction or not . to overcome this source of uncertainty
, they investigated the 1/4kev to 3/4kev energy ratio , which is , as shown above , a sensitive measure of the plasma temperature .
they cross - correlated the 1/4kev galactic halo intensities of kerp et al .
( 1999 ) with that of the modeled 3/4kev sky and deduced a plasma temperature of @xmath11})\,=\,6.2\,\pm\,0.02 $ ] ( compare their fig.10 ) . in following step , they scaled the emissivity of 3/4kev model to the 1/4kev energy range .
_ the rle of the extragalactic background radiation : _ the radiation transfer equation of snowden et al . and pietz et al .
is the same .
an important , but up to now not discussed diffuse soft x - ray component is the extragalactic background .
the 1/4kev to 3/4kev energy - band ratio is not only a sensitive measure of the plasma temperature , but also a measure for the brightness of the extragalactic background . as pointed out by kerp & pietz ( 1998 ) , if the extragalactic background spectrum is steep @xmath16 ( hasinger et al .
1993 ) and the intensity within the 1/4kev band high @xmath17 ( cui et al . 1996 ) , the quantitative analysis of the _ rosat _ energy - band ratios allows only a low galactic halo plasma temperature ( @xmath10})\,<\,6.1 $ ] ) . if one assumes a flatter extragalactic background spectrum ( @xmath18 , gendreau et al . 1995 ) and a fainter 1/4kev intensity ( @xmath19 , barber et al . 1996 ) , a higher galactic halo plasma temperature is consistent with the energy band ratios .
snowden et al .
adopt the steep spectrum high 1/4kev intensity set of parameters , and found accordingly a low temperature halo .
moreover , this choice of parameters for the extragalactic background determines that 66% of the 3/4kev diffuse x - ray emission is of extragalactic origin , while low plasma temperatures derived for the galactic halo and the lhb _ can not _
account for the rest of the 3/4kev emission .
thus , a significant fraction of the 3/4kev diffuse x - ray radiation is not investigated .
pietz et al . used the flatter background spectrum and low intensities for the extragalactic background , the second set of parameters , which allows a higher temperature for the galactic halo plasma .
the difference between the 3/4kev extragalactic background level ( also 66% of the total observed 3/4kev intensity ) and the observed 3/4kev intensity is attributed to the galactic halo plasma .
apparently , the 3/4kev gap is filled with the galactic halo plasma emission . _
the model predictions : _ snowden et al .
( 1998 ) confirm the lhb model of ( snowden et al .
1990 ) only slightly modified because of the galactic halo emission . in the snowden et al .
( 1998 ) model the galactic halo contains some patches of soft x - ray emitting halo gas , which are distributed across the high latitude sky .
the 3/4kev emission is not entirely explained by the model , because the gap between the extragalactic x - ray background intensity level and the observed intensity is still present . the pietz et al .
model explains entirely the 3/4kev and 1/4kev x - ray background emission towards high latitudes .
the derived galactic halo plasma temperature of @xmath10})\,=\,6.2 $ ] emits most of the soft x - ray photons in the 1/4kev energy band .
accordingly , they scaled the emissivity of the 3/4kev model to that of the 1/4kev x - ray emission .
they attribute the difference between the large - scale averaged observed 1/4kev and modeled 1/4kev galactic halo emission to the lhb .
the derived lhb emission is smoothly distributed across the sky . _
conclusions : _ snowden et al . confirmed the lhb model , in which a displacement of x - ray plasma and @xmath20 produces the apparent galactic plane to galactic pole x - ray intensity gradient .
they found evidence for some low temperature soft x - ray emitting plasma patches within the galactic halo gas .
they conclude that temperature and intensity of the halo emission must be highly variable which excludes any pervasive galactic halo plasma .
pietz et al . constructed a model of the x - ray halo which is consistent solution for the detected diffuse soft x - ray emission across the entire _ rosat _ x - ray energy range .
they proposed , that the 3/4kev emission is a superposition of the galactic halo and extragalactic background emission .
the galactic halo plasma appears to be isothermal on large angular scales , accordingly it is possible to scale the 3/4kev model to the 1/4kev energy regime .
they derived a smooth intensity distribution of the lhb , which is significantly cooler than the galactic halo plasma .
the smooth lhb intensity distribution is a function of the galactic latitude . towards the northern galactic pole ,
the lhb is about 1.5 times brighter than towards the southern galactic pole .
the residuals between the observed and modeled x - ray intensity distributions show some individual excess x - ray emitting structures which are identified as star forming regions , supernova remnants or attributed to the high - velocity cloud phenomenon ( kerp et al .
the pietz et al . galactic x - ray halo model predicts the existence of a hot gaseous phase within the galactic halo .
this hot gas phase is the environment of cosmic - rays , magnetic fields and neutral clouds , accordingly , we can prove the x - ray model parameters by the quantitative comparison with recent all - sky surveys of , radio continuum and @xmath21-ray emission ( kalberla & kerp , 1998 ) .
on large scales , the vertical distribution of gas in the galaxy can be characterized by a layered structure ( dickey & lockman 1990 ) .
the layers which are associated with the disk have an exponential scale height of up to 400 pc . until recently there was also agreement that gas associated with the halo must have a scale height of @xmath22 kpc ( e.g. kulkarni & fich 1985 or lockman & gehman 1991 ) .
such a scale height corresponds to a component with a velocity dispersion of 35 .
this was questioned by westphalen ( 1997 ) and kalberla et al .
( 1998a ) who found a velocity dispersion of 60 , corresponding to a scale height of 4.4 kpc .
here we discuss the major differences between both approaches .
_ the data - set : _ a component with a velocity dispersion of @xmath23 35 was derived by different authors from profiles extracted from the bell laboratories survey ( bls , stark et al .
the main reason for using this horn - antenna for such a kind of analysis was the high main beam efficiency of 92% , which means that the observations are only little affected by stray radiation from the antenna side - lobes . a high velocity dispersion component of @xmath23 60 was found by westphalen ( 1997 ) using the lds .
this survey was observed with a parabolic dish .
usually , such a telescope suffers from spurious emission which disables any analysis of components with high - velocity dispersion . in the case of the lds ,
however , the stray radiation was removed in a numerical way by hartmann et al .
these authors claim that they reduce spurious emission on average by two orders of magnitude .
they demonstrated , that even reflections from the ground can affect observations .
subsequently , such effects have been corrected by kalberla et al .
( 1998a ) who suggested that systematic uncertainties due to residual stray radiation in wings of the spectra are below 15 mk .
_ data reduction : _ emission lines with high - velocity dispersion may be seriously affected by the instrumental baseline and the correction algorithm which was used .
individual bls profiles were corrected by fitting a second order polynomial to emission free profiles regions which have been determined in an iterative way . in case of the lds ,
a third order polynomial fit was applied after determining the emission free regions with polynomials up to @xmath24 order .
the resulting bls baselines are believed to be accurate to 50 mk .
averaging observations in time should improve the baseline accuracy .
however , the averaging of spectra of the bls does not show this expected behavior ( stark et al . 1992 ) .
in opposite to this finding , the individual spectra of the lds are noisier and have a typical rms noise level of 70 mk .
averaging the lds spectra however leads to the expected improvement of the rms noise up to a limit of a few mk . in comparing bls and lds spectra after averaging across regions of @xmath25 kalberla et al .
( 1998b ) found significant deviations , indicating incompatible baselines between both surveys .
the majority of the averaged bls spectra reveal negative brightness temperatures in the baseline regions .
kalberla et al .
( 1998b ) concluded , that probably an improper baseline correction algorithm was used by the reduction of the bls data . in turn
this led to a subtraction of the faint and broad line emission of the high - velocity dispersion component .
polynomial baseline corrections are designed to _ remove _ features from the spectra which are assumed to be artificial .
accidentally , broad emission line components may be affected .
the question , whether the baseline correction procedure applied to the lds may have _ added _ an artificial broad emission line component was studied by kalberla et al .
( 1998a ) in detail .
the complete lds data analysis was repeated .
no evidence was found that the high - velocity dispersion component is due to instrumental or computational artifacts .
_ data analysis : _ both surveys have been analyzed regarding low surface brightness components with high - velocity dispersion . in case of the bls kulkarni & fich ( 1985 ) as well as lockman & gehman ( 1991 ) averaged large regions in direction to the galactic poles ( @xmath26 and @xmath27 ) .
such averages were decomposed into gaussian components .
westphalen ( 1997 ) considered individual @xmath28 fields in all directions with latitudes @xmath29 .
after a gaussian decomposition of the individual averaged profiles , the average property of a component with high - velocity dispersion was determined . in each case
the rms deviations between the individual observations have been determined on a channel by channel basis ; rms - peaks indicating interference , hvcs or instrumental problems have been used to flag velocity intervals , which need not to be fitted by a gaussian component . to verify , that the determined components are not affected by residual stray radiation , the stray radiation itself was decomposed into gaussian components . since a gaussian analysis may be biased by initial constraints , each fit was repeated several times using different criteria for the derivation of components with high - velocity dispersion . to compare the bls and lds datasets and the models based on a high - velocity dispersion component , we plot data and corresponding models . fig .
1 is based on lds data .
we include the model derived by kalberla et al .
( 1998a ) .
2 represents the bls data and the model 2 published by lockman & gehman ( 1991 ) for comparison .
the basic hydrostatic halo model is due to parker ( 1966 ) . in this classical publication
it was shown that a single phase halo must be unstable . since then
, attempts have been made by various authors to derive conditions for a stable halo .
we consider only the most recent papers .
bloemen ( 1987 ) concluded that a gaseous halo is needed to provide the pressure necessary to stabilize the halo .
he proposed an extended high temperature halo . at a vertical distance of @xmath30kpc
the halo gas temperature should be @xmath31k while within the disk and above @xmath32kpc the temperature should be @xmath33k .
this halo plasma should emit significant amounts of x - ray photons within the 1/4kev and 3/4kev energy bands .
boulares & cox ( 1990 ) studied the galactic hydrostatic equilibrium including magnetic tension and cosmic - ray diffusion .
they found , that the gas pressure , based on recent observations at that time , was not sufficient to stabilize the halo against parker instabilities .
for a stable galactic halo , they proposed a halo phase with a velocity dispersion of 20 in the disk ( @xmath34 ) and of 60 at @xmath35 kpc .
this is close to the high - velocity dispersion component established by kalberla et al .
( 1998a ) .
the lds observations indicate however , that the emission in the halo seems to be isothermal with a velocity dispersion which is independent from @xmath0-height .
we calculated the distribution as proposed by boulares & cox ( 1990 ) and found that this distribution is similar to the distribution derived from the lockman & gehman ( 1991 ) model which is plotted in fig .
2 . to describe the large scale properties of the milky way in the sense of a steady - state solution , kalberla & kerp ( 1998 ) assumed a hydrostatic halo model , as proposed by parker ( 1966 ) . kalberla & kerp ( 1998 ) re - analyzed the hydrostatic equilibrium model of the milky way , including the most recent all - sky surveys , ranging from the @xmath21-ray to the radio synchrotron emission regime .
they distinguished three main components of the galaxy which can be characterized by their vertical scale heights @xmath36 : the cold and warm neutral interstellar medium ( ism ) with @xmath36 = 150 pc and @xmath36 = 400 pc ( dickey & lockman , 1990 ) , the diffuse ionised gas ( dig ) with @xmath36 = 1 kpc ( reynolds , 1997 ) , and halo gas with @xmath36 = 4.4 kpc . the major difference to previous investigations ( e.g. bloemen 1987 , boulares & cox 1990 ) is , that kalberla & kerp ( 1998 ) included the neutral ( kalberla et al . 1998a ) and x - ray ( pietz et al .
1998 ) gas phase , located within the galactic halo , into their calculations .
such a layered disk - halo model was found to fit the large scale distribution of the neutral and ionized hydrogen gas throughout the whole galaxy .
gas , magnetic fields and cosmic rays were found to be in pressure equilibrium .
as discussed by kalberla & kerp ( 1998 ) in detail , such an equilibrium situation does not only explain the large scale distributions of gas , magnetic fields and cosmic rays , but results also in a stable configuration .
it is essential , that the multi - phase halo is supported by the pressure of the gaseous layers with low @xmath0 scale - heights , belonging to the galactic disk and to the disk - halo interface . without such layers ,
the halo would be instable as described by parker ( 1966 ) .
kalberla & kerp ( 1998 ) found that the halo is stable on _ average _ only .
pressure fluctuations may cause instabilities above @xmath0 distances of @xmath37 4 kpc , but not below .
considering activities in the galactic disk ( e.g. norman & ikeuchi 1989 ) one expects that such pressure fluctuations will affect the halo gas .
necessarily , instabilities will trigger the condensation of hvcs in the halo .
kalberla et al .
( these proceedings ) studied the large scale distribution of hvcs and found that turbulent motions within the halo have considerable effects on the observed column density distribution . except from turbulence ,
the scenario , derived from hydrostatic considerations , is in close agreement with predictions of hydro - dynamical models .
fountain parameters which have been favored by bregman ( 1980 ) are in close agreement with the parameters derived by pietz et al .
( 1998 ) from observations .
a gaseous multi - layer structure was found also by avillez ( these proceedings ) from 3d hydro - dynamical simulations . in his calculations a halo , and at the same time a disk - halo interface is built up after a short period of time , reaching a steady state situation soon .
scale heights , densities and temperatures of these layers are found to be comparable to the values listed by kalberla & kerp ( 1998 ) .
albert , c.e . , welsh , b.y .
, danly , l. 1994 , apj 437 , 204 barber c.r .
, roberts t.p . , warwick r.s . , 1996 , mnras 282 , 157 beuermann k. , kanbach g. , berkhuijsen e.m . , 1985 ,
a&a 153 , 17 bloemen j.b.g.m . , 1987 ,
apj 322 , 694 boulares a. , cox d.p . , 1990 , apj 365 , 544 ( b&c ) bowyer c.s , field g.b . , mack j.e . , 1968 ,
nat 217 , 32 bregman j.n .
, 1980 , apj 236 , 577 cui w. , sanders w.t . , mccammon d. , et al .
1996 , apj 468 , 117 dickey j.m . ,
lockman f.j . , 1990 ,
ara&a , 28 , 215 gendreau k.c . , mushotzky r. , fabian a.c . , et al .
1995 , pasj 47 , l5 hartmann dap , burton w.b .
1997 , `` atlas of galactic neutral hydrogen '' , cambridge university press hartmann dap , kalberla p.m.w .
, burton w.b .
et al . , 1996 a&as 119 , 115 hasinger g. , burg r. , giacconi r. et al .
1993 , a&a 275 , 1 kalberla p.m.w .
, kerp j. , 1998 , a&a 339 , 745 kalberla p.m.w . , westphalen g. , mebold u. , hartmann dap .
, burton w.b . , 1998 , a&a 332 , l61 kalberla p.m.w . , westphalen g. , mebold u. , hartmann dap , burton w.b .
, 1998b , in proceedings of the iau colloquium no .
166 , eds .
d. breitschwerdt et al . ,
lecture notes in physics 506 , 475 kerp j. , burton , w.b . , egger r. , et al . , 1999 ,
a&a in press ( astro - ph/9810307 ) kerp j.,pietz j. , 1998 , in proceedings of the iau colloquium no .
166 , eds .
d. breitschwerdt et al . ,
lecture notes in physics 506 , 337 kulkarni , s.r .
, fich , m. 1985 , apj 289 , 792 .
lockman , f.j .
, gehman , c.s . 1991 , apj 382 , 182 mnch , g. 1957 , apj 125 , 42 mccammon d. , sanders w.t . , 1990 , ara&a 28 , 657 norman c.a . , ikeuchi s. , 1989 , apj 345 , 372 parker e.n . , 1966 ,
apj 145 , 811 pietz j. , kerp j. , kalberla p.m.w . , et al . , 1998 , a&a 332 , 55 pikelner s.b . , shklovsky i.s . , 1958 , iau symposium 8 , reviews of modern physics , 30 , 935 reynolds r.j . 1997 , in proc
we - heraeus - seminar on `` the physics of galactic halos '' , eds .
lesch et al . , akademie verlag , berlin , 57 savage , b.d . ,
sembach , k.r . ,
lu , l. , 1997 , aj 113 , 2158 snowden s.l .
, cox d. , mccammon d. , sanders w.t .
, 1990 , apj 354 , 211 snowden s.l . , egger r. , finkbeiner d.p . , et al .
, 1998 , apj 493 , 715 spitzer , l. 1956 , apj 124 , 20 stark , a.s . ,
gammie , c.f .
, wilson et al . , 1992 , apjs 79 , 77 westphalen , g. , 1997 , phd thesis , university of bonn | in 1998 several papers claim the detection of an ubiquitous gaseous phase within the galactic halo . here
we like to focus on the detections of x - ray emitting gas within the galactic halo as well as the discovery of a pervasive neutral galactic halo gas . we discuss critically the major differences between the recent publications as well as the limitations of the analyses . |
two decades after pioneering contributions of the likes of quantum cryptography @xcite and teleportation @xcite , quantum information science has nowadays acquired a status of maturity @xcite , having witnessed theoretical and experimental advances which demonstrated in several ways the power of quantum technology @xcite .
nonetheless , fundamental questions of inherent practical relevance remain open , broadly concerning the proper identification of the ultimate _ resources _ behind such a power .
paradigmatic instances of systems suitable for quantum protocols have been , for instance , registers of pure qubits for discrete - variable ( `` digital '' ) quantum computation and cryptography @xcite , and multimode gaussian states of radiation fields for continuous - variable ( `` analog '' ) quantum communication and information processing @xcite .
entanglement @xcite has been the crucial resource on which most studies have focused so far as its presence is crucial for better - than - classical communication performances .
recently , however , there has been more of a push for research in quantum information to go beyond the boundaries of its first generation . in particular , the need to consider more realistic setups where mixedness affects quantum computations in the discrete variable framework @xcite , and the need to go beyond the nutshell of gaussian states and operations to achieve universality in continuous variable computation @xcite have both been strong motivations .
nonclassicality ( i.e. the quantumness of correlations beyond and without entanglement ) @xcite and non - gaussianity @xcite are currently under the limelight due to being recognized as `` power - ups '' for quantum technology , in particular in applications such as quantum computation @xcite , quantum communication @xcite and metrology @xcite . in this paper
we investigate the nonclassicality of correlations in continuous - variable ( cv ) systems beyond gaussian states .
nonclassical correlations , including and beyond entanglement , can be quantified for instance by the _ quantum discord _
@xcite , a measure that aims at capturing more general signatures of quantumness in composite systems .
they are associated e.g. with the non - commutativity of quantum observables and with the fact that local measurements generally induce some disturbance on quantum states , apart from very special cases in which those states admit a fully classical description .
the interest in quantum discord has risen significantly since it was recognized as the potential resource behind the quantum speed - up in certain mixed - state models of quantum computing such as the dqc1 @xcite , in which entanglement is negligible or strictly vanishing @xcite .
quantum discord @xcite , along with other similar nonclassicality measures @xcite such as the ( ameliorated ) measurement - induced disturbance @xcite and distance - based quantifiers such as the geometric discord @xcite and the relative entropy of quantumness @xcite , typically feature nontrivial optimizations in their definitions .
in particular , it is necessary to identify the least disturbing measurement to be applied on one or more subsystems to extract those nonclassical correlations , rendering their exact computation a formidable task .
closed formulae are available for the quantum discord of a special subclass of two - qubit states @xcite and of general two - mode gaussian states under the restriction of gaussian measurements @xcite .
we consider here a family of two - mode states that are the cv counterparts of two - qubit werner states @xcite .
they are non - gaussian states obtained as mixtures of two gaussian states @xcite , an entangled two - mode squeezed state and a two - mode thermal product state , and find applications in cv quantum cryptography @xcite .
studying their nonclassicality beyond entanglement is particularly interesting from a fundamental point of view , as they offer a unique testbed to compare the role of gaussian versus non - gaussian measurements to extract correlations with minimum disturbance from general two - mode cv states .
gaussian states and operations are known to satisfy sharp extremality properties @xcite in the space of all cv states .
however , such results do not apply to quantum discord , rendering the understanding of the structure of ( non)classical correlations even subtler in cv systems compared to finite - dimensional systems , for which general results are known instead @xcite .
we show that the states analyzed here represent instances of quantum states carrying _ genuine _ cv nonclassical non - gaussian correlations , for which the optimization in the discord is achieved _ only _ by an infinite - dimensional component , hence genuinely cv , non - gaussian measurement . in the following
we prove that for a particular type of the non - gaussian cv werner state @xcite ( obtained by mixing a two - mode squeezed state with the vacuum ) quantum discord can be computed _
analytically_. the global optimization in its definition is achieved exactly by photon counting , and such a measurement is strictly less disturbing than any gaussian measurement .
the gap between optimal gaussian measurements ( homodyne detection ) and optimal non - gaussian measurements ( photon counting ) is quantified by the difference of the associated conditional entropies .
the gap is then compared with an entropic measure for the non - gaussianity of the states @xcite , and found to be proportional to it in the low squeezing regime .
the exactly computed discord turns out to be equal to other measures for the quantumness of correlations as well in this special case @xcite .
next , we analyze discord in the most general cv werner state , ideally complementing the analysis of the two - qubit werner state performed originally in @xcite ( see also @xcite ) .
we show that photon counting provides in general a ( not necessarily optimal ) upper bound on quantum discord , that coincides with the measurement - induced disturbance @xcite , and we derive a lower bound on discord as well . finally , we give an example of a cv state which is positive under partial transposition ( ppt @xcite ) , and show that it carries `` weak '' nonclassical correlations , signaled by analytically computable upper and lower bounds on discord which are close to each other and stay small and finite even for infinite squeezing .
our results provide insights into the relationship between the general quantumness of correlations , entanglement distillability and separability in cv systems outside the gaussian scenario . from a practical perspective
, our results identify the key role of non - gaussian measurements such as photon counting to access and extract all nonclassical correlations in general cv states , even in the particular case of non - gaussian states with a positive - everywhere wigner function such as the mixtures of gaussian states studied here .
the paper is organized as follows . in sec .
[ secbasic ] we set up notation and recall the definitions of quantum discord @xcite and of the cv werner states @xcite . in sec .
[ secexact ] we study a special type of cv werner state and calculate its discord ( and other nonclassicality indicators ) exactly , proving that photon counting constitutes a globally optimal measurement strategy , and issuing comparisons between the degree of non - gaussianity @xcite and the amount by which gaussian measurements fail to be optimal on the considered class of states . in sec .
[ secgeneral ] we address the general case and provide exact lower and upper bounds for the quantum discord of arbitrary cv werner states . in sec .
[ secppt ] we study a special type of cv state obtained by partial transposition from a cv werner state , and investigate its degree of nonclassical correlations in different entanglement regimes .
we draw our conclusions in sec .
[ secfinal ] .
quantum discord is a measure of the nonclassicality of correlations in a bipartite quantum state @xmath0 defined as a difference @xcite @xmath1 between the quantum generalizations of two classically equivalent expressions for the mutual information , namely the quantum mutual information @xmath2 where @xmath3 is the von neumann entropy , and the so called one - way classical correlation @xcite @xmath4 with @xmath5 being the conditional entropy of @xmath6 given a positive operator valued measurement ( povm ) @xmath7 has been performed on @xmath8 . here
@xmath9 $ ] are reduced states of subsystems @xmath6 and @xmath8 , respectively , @xmath10/p_{b}(i)$ ] is the conditional state obtained upon detecting the povm element @xmath11 on @xmath8 and @xmath12 $ ] is the corresponding probability . without loss of generality one can restrict to rank 1 povms in what follows @xcite .
subtracting eq .
( [ j ] ) from eq . ( [ i ] ) one gets for quantum discord the following formula : @xmath13 note that quantum discord is generally asymmetric under the change @xmath14 , but in this paper we will consider only symmetric states for which the discord is naturally invariant under the choice of the measured subsystem .
the hard step in the evaluation of the quantum discord ( [ d2 ] ) is the optimization of the conditional entropy @xmath15 over all povms .
although this can not be done in full generality analytically even for two qubits @xcite the problem is tractable using analytical tools for certain subsets of states and povms , e.g. , for gaussian ones @xcite .
however , the question as to whether gaussian measurements are globally optimal for the extraction of nonclassical correlations from general cv quantum states is essentially open and has not been settled even for gaussian states themselves .
we consider the cv werner state @xcite defined as @xmath16 where @xmath17 , @xmath18 is the two - mode squeezed vacuum state with @xmath19 ( @xmath20 is the squeezing parameter ) and @xmath21 is the thermal state with @xmath22 , where @xmath23 is the mean number of thermal photons in mode @xmath24 .
we observe that the infinite - dimensional state ( [ rho ] ) also possesses the same structure as @xmath25-dimensional states invariant under the maximal commutative subgroup of @xmath26 introduced in @xcite .
first , let us consider the simplest special case of a werner state with @xmath28 which gives using eq .
( [ rho ] ) @xmath29 representing just a mixture of a two - mode squeezed vacuum state with the vacuum . for @xmath30 the partially transposed matrix @xmath31 ( obtained by transposing @xmath32 with respect to the degrees of freedom of subsystem
@xmath6 only ) has negative eigenvalues @xcite and therefore , according to the ppt criterion @xcite , the state ( [ rho0 ] ) is entangled .
note that the state ( [ rho0 ] ) has been further studied in @xcite from the point of view of its entanglement properties , as measured by the negativity @xcite , highlighting its applications for quantum key distribution . in order to calculate the entropies arising in the expression of quantum discord ( [ d2 ] ) we need to determine the eigenvalues of the reduced state @xmath33 , the global state ( [ rho0 ] ) and the conditional state @xmath34/p_{b}(i)$ ] .
the latter two states attain the form @xmath35 where @xmath36 and @xmath37 are generally nonorthogonal normalized pure state vectors .
the state ( [ sigma ] ) has at most two - dimensional support spanned by vectors @xmath37 corresponding to eigenvalues @xmath38 that read as @xmath39 on inserting the eigenvalues ( [ nu ] ) into the formula for the von neumann entropy @xmath40 we get analytically the entropy of the state ( [ sigma ] ) .
returning back to the state ( [ rho0 ] ) we get , in particular , @xmath41 and @xmath42 ; the eigenvalues thus amount to @xmath43 hence , we can immediately calculate the entropy @xmath44 using formula ( [ s ] ) , @xmath45 tracing , further , the state ( [ rho0 ] ) over mode @xmath6 yields a diagonal reduced state @xmath46 with @xmath47 given in eq .
( [ rhoth ] ) , possessing the eigenvalues @xmath48 for @xmath49 . making use of the definition ( [ s ] )
we get also the entropy of the reduced state in the form : @xmath50\right.\nonumber\\ & & \left.+\frac{2p\lambda^2\ln\lambda}{1-\lambda^2}\right\}.\end{aligned}\ ] ] we are left with the minimization of the conditional entropy @xmath51 over all povms .
the state ( [ rho0 ] ) is constructed in such a way that we can `` guess '' the optimal measurement before attempting to solve this anyway hopeless task with brute - force .
a close look at the state reveals that the optimal measurement on mode @xmath8 is just photon counting ( i.e. , in this case , projection onto the eigenstates of @xmath33 ) characterized by the set of projectors @xmath52 , where @xmath53 is the @xmath54-th fock state . clearly , if we detect @xmath54 photons in mode @xmath8 in the state ( [ rho0 ] ) , the conditional state @xmath55 is simply a _
pure _ fock state @xmath56 which implies , immediately , that the conditional entropy attains its minimal possible value @xmath57 .
this proves rigorously that photon counting is the globally optimal measurement strategy for the state @xmath32 , thus giving the discord exactly equal to @xmath58 where the involved entropies are defined in equations and .
the discord is an increasing function of both @xmath59 and @xmath60 and is plotted in fig .
[ fig13d ] .
note , that the considered werner state ( 8) belongs to the class of maximally correlated states for which eq . can be proved alternatively @xcite using the duality relation between classical correlations and entanglement of formation @xcite .
( [ drho0 ] ) ] versus the probability @xmath60 and the squeezing factor @xmath59 for the cv werner state @xmath32 [ eq . ( [ rho0 ] ) ] .
all the quantities plotted are dimensionless.,width=321 ] a relevant question is whether photon counting is the only measurement realizing the global minimum in the evaluation of quantum discord , or , more specifically , whether there can exist gaussian povms that attain instead an equally optimal performance on the state .
the answer , as we argue in the following , is that nonclassical correlations in the state ( [ rho0 ] ) captured by discord ( [ drho0 ] ) can not be extracted equally well by any gaussian measurement . in other words ,
gaussian discord ( @xmath61 ) , defined by eq .
( [ d2 ] ) by imposing the minimization be restricted only to gaussian measurements @xcite , is just a strictly larger upper bound on the true discord ( [ drho0 ] ) for the mixed non - gaussian cv werner state . to show this ,
consider the following gaussian povm @xcite consisting of elements @xmath62 where @xmath63 with @xmath64 , is a pure normalized momentum - squeezed vacuum state with squeezing parameter @xmath65 , that is rotated counterclockwise by a phase @xmath66 and that is subsequently displaced by @xmath67 .
here , @xmath68 is the displacement operator and @xmath69/2\right\}$ ] is the squeezing operator .
note that @xmath70 corresponds to heterodyne detection , whereas homodyne detection is obtained in the limit @xmath71 .
if the povm element @xmath72 is detected on mode @xmath8 of the two - mode squeezed vacuum state ( [ psinopa ] ) , then mode @xmath6 collapses into the state @xmath73 , which is a pure momentum - squeezed state with squeezing parameter @xmath74\ ] ] that is rotated clockwise by phase @xmath75 and that is displaced by @xmath76,\ ] ] where @xmath77^{-1}$ ] .
thus , the obtained conditional state @xmath78/q\left(\alpha\right)$ ] is again a convex mixture of the form ( [ sigma ] ) where @xmath79 , @xmath80 and @xmath81 for the cv werner state ( [ rho0 ] ) with ( from bottom to top )
all the quantities plotted are dimensionless.,width=283 ] here @xmath83/\pi$ ] is the probability density of obtaining the measurement outcome @xmath84 , @xmath85 , where @xmath86 is given in eq .
( [ rhoth ] ) , and @xmath87 .
the latter function @xmath88 can be computed straightforwardly using the formula @xcite @xmath89 the overlap @xmath90 appearing in eigenvalues ( [ nu ] ) can be calculated exactly along the same lines .
expressing finally the thermal state @xmath91 as @xmath92 we get the function @xmath93 in the form @xmath94d(\alpha-\varsigma)$ ] .
using once again the formula ( [ overlap ] ) to express the overlap @xmath95 and performing the integration over @xmath96 we arrive at the formula @xmath97 } { 1-\lambda^{4}\tanh^{2}(t)}|\alpha|^{2}\nonumber \\ & & + \frac{\left(1-\lambda^2\right)^{2}\tanh(t)}{2\left[1-\lambda^{4}\tanh^{2}(t)\right ] } \left({\rm e}^{-i2\varphi}\alpha^2+{\rm e}^{i2\varphi}\alpha^{\ast 2}\right)\bigg\}. \nonumber\end{aligned}\ ] ] substituting the obtained explicit expressions for functions @xmath98 and @xmath99 into eqs .
( [ ab2 ] ) and using the explicit expression for the overlap @xmath90 we get from eq .
( [ nu ] ) the eigenvalues and hence the entropy @xmath100 of the conditional state @xmath101 .
subsequent averaging of the entropy over the density @xmath99 finally yields the gaussian conditional entropy @xmath102 as a function of the squeezing parameter @xmath103 and phase @xmath75 of the gaussian measurement ( [ gpovm ] ) . due to the complicated dependence of the conditional entropy @xmath100 on @xmath84 , the remaining integration over the complex plane @xmath104 , where @xmath105 , has to be performed numerically .
likewise , minimization of the entropy ( [ halpha ] ) with respect to variables @xmath103 and @xmath75 also requires numerics .
this analysis reveals that , within the gaussian povm set , the entropy is minimized by homodyne detection on mode @xmath8 .
the resulting plots of gaussian discord and the true quantum discord ( [ drho0 ] ) ( the latter obtained by photon counting on @xmath8 ) are shown in fig .
the figure clearly shows that apart from trivial cases @xmath106 the gaussian discord is always strictly larger than the discord ( [ drho0 ] ) , meaning that general gaussian measurements are strictly suboptimal ( or , in other words , non - minimally disturbing ) for the extraction of nonclassical correlations in the non - gaussian state . to shine light on the reasons behind this , we compare the gap @xmath107 quantifying how much gaussian measurements overestimate the actual amount of nonclassical correlations , with a measure of the non - gaussianity @xmath108 of the state @xcite .
the latter is defined as the quantum relative entropy between @xmath32 and a reference gaussian state @xmath109 with the same first and second moments , @xmath110 where @xmath111 is given in . to find @xmath112 , we note that @xmath32 has zero first moments but possesses a covariance matrix of the form @xmath113 with @xmath114 , @xmath115 .
the von neumann entropy of the gaussian state @xmath109 can be then written as @xcite @xmath116 where @xmath117/({1-\lambda ^2})}$ ] is the doubly - degenerate symplectic eigenvalue @xcite of the covariance matrix @xmath118 .
the non - gaussianity @xmath108 is a concave function of @xmath60 and increases with @xmath59 , diverging in the limit of infinite squeezing . in the regime of low squeezing
, @xmath119 , a series expansion ( up to the quadratic term in @xmath59 ) returns an approximate expression for the non - gaussianity , @xmath120.\ ] ] on the other hand , the gap in discord just coincides with the conditional entropy corresponding to homodyne detection , @xmath121 . in the low squeezing regime , we can also expand in series ( up to the quadratic term in @xmath59 ) the integrand in , so as to obtain an approximate analytic expression for the gap , @xmath122,\ ] ] where @xmath123 is euler s constant .
defining the ratio @xmath124 } { \ln \left(\frac{8}{\lambda ^2}\right)+\gamma -1}\ ] ] between the approximate expressions for @xmath108 and @xmath125 ( at @xmath126 ) , we see that the linearly dependent relationship @xmath127 holds with good approximation for small @xmath59 .
in other words , for low squeezing ( say @xmath128 ) , the ( normalized ) gap @xmath129 between optimal gaussian ( homodyne ) and globally optimal non - gaussian ( photon counting ) measurements for the extraction of nonclassical correlations , correctly characterizes and quantitatively reproduces the non - gaussianity @xmath108 of the considered state .
interestingly , @xmath130 this intriguing connection between the nonclassicality gap and non - gaussianity fails to hold for larger values of @xmath59 ; the discrepancy between the two parameters becomes extreme in the limit @xmath131 , when the non - gaussianity diverges while the gap @xmath125 closes to zero . a comprehensive comparison between the normalized gap in discord and the non - gaussianity measure is shown in fig . [ figap ] .
one may argue , that even a simpler non - gaussian state than that of given in eq .
( [ rho0 ] ) can be found possessing a strictly lower discord for a non - gaussian measurement than for the best gaussian measurement .
for instance , the optimal measurement minimizing the discord in the qubit werner state @xcite , studied in the seminal paper on quantum discord @xcite , is a simple non - gaussian projection onto the first two fock states @xmath132 and @xmath133 .
one can easily check that the optimization over all gaussian measurements gives a strictly higher discord .
let us stress that the nonclassical correlations captured by discord are fundamentally different for the cv werner state ( [ rho0 ] ) considered here and for the qubit werner state .
namely , although our cv werner @xcite is formally a qubit - like state , the globally optimal povm has an infinite number of elements given by projectors onto all fock states .
moreover , it is not difficult to show that no povm measurement on mode @xmath8 possessing a finite number @xmath134 of elements @xmath135 , @xmath136 can nullify the conditional entropy @xmath137 and hence also be globally optimal .
namely , the conditional state corresponding to detection of the element @xmath135 has to be a pure state , that is , @xmath138=|\chi_{i}\rangle\langle\chi_{i}|$ ] in order for the entropy of the conditional state to vanish .
now , consider the element @xmath139 .
the corresponding conditional state @xmath140=\rho_{0,a}-\sum_{i=1}^{n}|\chi_{i}\rangle\langle\chi_{i}|$ ] , where @xmath141 is obtained from eq .
( [ rho0b ] ) by replacing @xmath8 with @xmath6 , can not have neither zero , nor one , nor even any finite number of strictly positive eigenvalues , as this would imply that the state @xmath141 also has a finite number of strictly positive eigenvalues which is not the case [ see equation ] .
therefore , for any finite @xmath134 the conditional state @xmath142 is definitely a mixed state possessing strictly positive entropy and hence resulting in a strictly positive and therefore suboptimal conditional entropy ( i.e. , non - optimal discord ) .
consequently , the globally optimal povm of the qubit werner state is only two - component and thus the qubit werner state is only a trivial embedding of a two - qubit state into an infinitely - dimensional two - mode state space carrying _ only _ qubit - type non - gaussian nonclassical correlations .
in contrast , the cv werner state ( [ rho0 ] ) carries genuinely cv non - gaussian nonclassical correlations that can be optimally extracted only by a _
non - gaussian _ povm measurement with an _ infinite _ number of elements : in this particular case , photon counting . before going
further let us note that , for a state of the form ( [ sigma ] ) , the global optimality of photon counting ( for the calculation of quantum discord ) follows immediately from the fact that the projection of one of its modes onto a fock state projects the other mode onto the same fock state .
as a consequence the conditional entropy achieves the minimum possible value @xmath144 and the discord is of the form ( [ drho0 ] ) .
it is not difficult to find more general mixed states with the same property , for instance , states with the structure @xmath145 with @xmath146 , @xmath147 and the matrix @xmath148 with elements @xmath149 being positive - semidefinite .
a particular example of such states is a convex mixture of an arbitrary number of two - mode squeezed vacua ( [ psinopa ] ) with different squeezing parameters @xmath150 obtained for @xmath151 , where @xmath152 are probabilities and @xmath153 for @xmath154 .
we remark that for _ all _ non - gaussian states of the form the quantum discord can be computed exactly .
the quantum state ( [ rho0 ] ) admits furthermore an analytical calculation of other optimized entropic quantifiers of nonclassical correlations @xcite , encompassing the ` ameliorated ' measurement - induced disturbance @xcite and the relative entropy of quantumness @xcite .
interestingly , as we will now show , both of them coincide with the discord ( [ drho0 ] ) .
this is in complete analogy with the corresponding werner state for two qubits @xcite and for two qu@xmath25its with arbitrary @xmath25 @xcite .
the optimized nonclassicality indicator called ` ameliorated ' measurement - induced disturbance ( amid ) is a two - way measure of quantum correlations defined as @xcite @xmath155 where @xmath156 is the quantum mutual information and @xmath157 is the classical mutual information @xcite of the quantum state @xmath158 . here
@xmath159 is the classical mutual information of the joint probability distribution @xmath160 $ ] of outcomes of local measurements @xmath161 and @xmath162 on @xmath158 , where @xmath163 and @xmath164 @xmath165 are the shannon entropies of the joint distribution and marginal distribution @xmath166 ( @xmath167 ) , respectively .
the classical mutual information is upper bounded as @xcite @xmath168 since @xmath169 for the state , we need to compare @xmath170 with @xmath171 which can be done using majorization theory for infinite - dimensional density matrices @xcite .
consider two such density matrices @xmath172 and
@xmath173 with @xmath174 and @xmath175 being their non - zero eigenvalues arranged in a decreasing order and repeated according to their multiplicity . if some of the matrices , for example , @xmath172 has only a finite number @xmath176 of non - zero eigenvalues , we set @xmath177 .
we then say that @xmath172 is more mixed than @xmath173 , and write @xmath178 , if @xmath179 it holds further @xcite , that if @xmath180 then their von neumann entropies satisfy @xmath181 .
taking now the eigenvalues @xmath182 , @xmath183 given in eq .
( [ nurho0b ] ) instead of eigenvalues @xmath184 , and @xmath38 given in eq .
( [ nurho0 ] ) instead of eigenvalues @xmath185 , one easily finds that they satisfy eq .
( [ majorization ] ) .
this implies that the density matrices @xmath158 and @xmath33 of eqs .
( [ rho0 ] ) and ( [ rho0b ] ) satisfy @xmath186 and therefore @xmath187 .
hence , one gets @xmath188 which leads , using the inequality ( [ iabbound ] ) , to the upper bound on the classical mutual information in the form @xmath189 .
considering now photon counting on both modes in the state ( [ rho0 ] ) , one gets @xmath190 which gives the classical mutual information in the form @xmath191 .
hence , the latter inequality is saturated by photon counting which finally yields @xmath192 , that is , amid coincides with the quantum discord , eq .
( [ drho0 ] ) .
the measure quantifies the minimum distance , in terms of relative entropy , between a quantum state @xmath32 and the set of completely classically correlated states @xcite .
it is operationally associated to the amount of distillable entanglement that can be generated between a quantum state @xmath158 ( carrying nonclassical correlations ) and a set of ancillary systems in the worst case scenario of an activation protocol @xcite .
the relative entropy of quantumness is defined as @xcite @xmath193,\ ] ] where the minimization is performed over local measurements @xmath161 and @xmath162 consisting of collections of one - dimensional projectors @xmath194 and @xmath195 . the minimum in eq .
( [ q ] ) can be found using the same trick as in the previous cases .
first , we take a suitable tight bound on the quantity that is to be optimized and then we guess a measurement saturating the bound . in the present case we look for the minimum of the joint shannon entropy @xmath196 satisfying the inequalities @xmath197 . using the previous result that for photon counting we get @xmath198 , we see that the measurement saturates the lower bound and therefore @xmath199 . thus all three considered nonclassicality indicators coincide for the state ( [ rho0 ] ) , i.e. , @xmath200
let us now move to the analysis of nonclassical correlations in a generic cv werner state ( [ rho ] ) with @xmath201 , complementing the seminal analysis of nonclassical correlations in a two - qubit werner state performed in @xcite .
this can be interesting in particular because , in contrast to the qubit case , there can potentially exist ppt entangled cv werner states @xcite .
in the present general case we do not have any tight bounds , similar to those of the previous section , allowing us to perform exact optimizations in eqs .
( [ d2 ] ) , ( [ ic ] ) and ( [ q ] ) .
for this reason , we can not prove the global optimality of photon counting or any other measurement strategy analytically .
we then resort to computing upper bounds on discord , amid and relative entropy of quantumness , obtained for ( possibly nonoptimized ) measurements in the local eigenbasis of the reduced state(s ) of the two - mode cv werner state .
interestingly , all the upper bounds on the different quantities again coincide as we show later in this section : this hints at the conjecture that they might be indeed tight for the considered states , although we can not provide conclusive evidence of this claim .
we also derive nontrivial lower bounds for the nonclassical correlations .
we consider a nonoptimized upper bound on quantum discord defined for a density matrix @xmath202 as @xmath203 where @xmath204 is the conditional entropy for the measurement of mode @xmath8 in the local eigenbasis of the reduced state @xmath205 ( see also @xcite ) .
for the general cv werner state of , tracing @xmath202 over mode @xmath6 gives the reduced state @xmath206 where @xmath207 is defined in eq .
( [ rhoth ] ) .
this state is diagonal in the fock basis with eigenvalues @xmath208 that give , after substitution into eq .
( [ s ] ) , the marginal entropy @xmath209 appearing in eq .
( [ u ] ) .
the local eigenbasis is a fock basis and so the projection on it corresponds again , even in the present general case , to photon counting .
the conditional state @xmath210/p_{b}(m)$ ] , where @xmath211 , obtained by projecting mode @xmath8 onto fock state @xmath53 reads explicitly @xmath212 with @xmath213 .
it has the eigenvalues @xmath214 where @xmath215 is the kronecker symbol , that give the following entropy of the conditional state @xmath216\right.\nonumber\\ & & \left.\times\left(\frac{1}{1-\mu^{2}}-\mu^{2m}\right)+\ln\left(\mu^2\right)\right.\nonumber\\ & & \left.\times\left[\frac{m}{1-\mu^2}+\frac{\mu^2}{\left(1-\mu^2\right)^{2}}-2m\mu^{2m}\right]\right\}\nonumber\\ & & -\eta_{m}^{(m)}\ln\eta_{m}^{(m)}.\end{aligned}\ ] ] hence , one gets the conditional entropy @xmath217 .
it remains to calculate the global entropy of the state ( [ rho ] ) .
for this purpose it is convenient to express the state as @xmath218 where @xmath219 the state ( [ rho ] ) thus possesses the eigenvalues @xmath220 corresponding to the eigenvectors @xmath221 , @xmath222 and the remaining eigenvalues @xmath223 are the eigenvalues of the infinite - dimensional matrix @xmath224 with elements ( [ mmn ] ) .
this gives the global entropy @xmath225\right .
\nonumber\\ & & \left.+\frac{2\ln\left(\mu\right)\left(1+\mu^2 + 2\mu^4\right)}{1-\mu^4}\right\ } -\sum^\infty_{l=0}f_l\ln f_{l}.\nonumber\\\end{aligned}\ ] ] the eigenvalues @xmath226 of matrix @xmath224 appearing in the last expression of the previous equation can not be calculated analytically and one has to resort to numerical diagonalization of a sufficiently large truncated matrix .
hence , one gets using eq .
( [ globals ] ) , and expressions for @xmath204 and @xmath209 , the sought upper bound ( [ u ] ) on the true quantum discord .
the true discord can be also bounded from below in the following way .
let us observe first , that apart from the trivial case @xmath227 ( corresponding to a product state ) all other cv werner states have nonclassical correlations as they possess a strictly positive quantum discord @xmath228 .
this can be proven using the sufficient condition on strict positivity of quantum discord @xcite according to which @xmath228 for a state @xmath202 if at least one off - diagonal block @xmath229 , @xmath154 is not normal , i.e. , it does not commute with its adjoint . in the present case of the werner state ( [ rho ] )
we have explicitly @xmath230 .
assuming @xmath30 and @xmath231 this gives immediately a nonzero commutator @xmath232&=&p^2\left(1-\lambda^2\right)^2\lambda^{2(i+j ) } \left[|i\rangle_{b}\langle i|-|j\rangle_{b}\langle j|\right]\nonumber\\ & \ne & 0 \quad \mbox{for}\,\,\ , i\ne j , \ , \ , p\neq0,\end{aligned}\ ] ] as required .
an explicit , non - tight lower bound can be derived that is nonnegative ( and thus nontrivial ) at least on some subinterval of probabilities @xmath60 .
namely , assume a povm on mode @xmath8 given by a collection of rank-1 operators @xmath233 .
if the component @xmath234 is detected on mode @xmath8 in the state ( [ rho ] ) , then mode @xmath6 collapses into the normalized conditional state @xmath235 where @xmath236 is a pure unnormalized state that is not specified here and @xmath237 with @xmath205 given in eq .
( [ rhob ] ) is the probability of measuring the outcome @xmath238 .
the state is a convex mixture of a pure state and a thermal state . making use of the concavity of the von neumann entropy @xmath239 and the fact that the entropy vanishes on pure states we arrive at the following inequality : @xmath240.\ ] ] by multiplying both sides of the inequality by @xmath152 and summing over
@xmath238 one finds the classical conditional entropy to be lower bounded as @xmath241 $ ] which yields finally using eq .
( [ d2 ] ) the lower bound @xmath242.\ ] ] in what follows we evaluate the upper and lower bounds in eqs .
( [ u ] ) and ( [ l ] ) , respectively , for two particularly important two - parametric subfamilies of the set of cv werner states .
first , we consider the case @xmath243 .
then the reduced state ( [ rhob ] ) is just a thermal state @xmath244 with a well - known entropy @xmath245 in this case the cv werner state is a mixture of a pure two - mode squeezed vacuum state ( [ psinopa ] ) with a product of its marginals ; in the strong squeezing limit @xmath131 , this state approaches a mixture of a maximally entangled epr state and a maximally mixed ( infinitely thermal ) state , which is a direct cv counterpart to the qubit werner state @xcite .
further , it can be shown that the cv werner state ( [ rho ] ) with @xmath246 is entangled for any @xmath30 @xcite . the upper bound ( [ u ] ) and lower bound ( [ l ] ) on quantum discord are depicted in fig .
note that in this and in the following plots , only nonzero values of the lower bound ( [ l ] ) will be shown . , eq .
( [ u ] ) ( solid curve ) , and lower bound @xmath247 , eq .
( [ l ] ) ( dashed curve ) , on quantum discord ( [ d2 ] ) versus probability @xmath60 for the cv werner state ( [ rho ] ) with @xmath248 .
all the quantities plotted are dimensionless.,width=283 ] the case for which @xmath249 is interesting because the cv werner state ( [ rho ] ) runs through three different separability regions as the parameter @xmath60 increases @xcite : 1 . if @xmath250 , then the state @xmath202 is separable ( dark gray strip in fig . [ fig3 ] ) .
2 . if @xmath251 , the state @xmath202 is ppt ( i.e. , it has positive partial transposition @xcite ) and it is unknown whether it is undistillable - entangled or separable ( light gray strip in fig .
[ fig3 ] ) .
3 . if @xmath252 , then the state @xmath202 is non - ppt and therefore entangled ( white region in fig .
[ fig3 ] ) .
next we focus on the determination of the nonoptimized ( upper bound ) version of amid ( [ amid ] ) called measurement - induced disturbance ( mid ) defined as @xcite @xmath253 where @xmath254 is the classical shannon mutual information of a probability distribution of results of measurements in eigenbases of the reduced states @xmath255 , which in the present case coincides with the joint photon - number distribution @xmath256 .
its marginal distributions @xmath257 coincide with the eigenvalues of the reduced states ( [ gm ] ) , i.e. , @xmath258 whence we get the equality between local shannon and von neumann entropies @xmath259 hence mid simplifies to @xmath260 the global shannon entropy can be derived easily by noting that the eigenvalues ( [ etamn ] ) of the conditional state ( [ rhoam ] ) satisfy @xmath261 thus representing a conditional probability @xmath262 of detecting @xmath143 photons in mode @xmath6 given @xmath54 photons have been detected in mode @xmath8 .
this implies immediately that @xmath263 , where we have used eq .
( [ shequality ] ) and the equality @xmath264 . substituting from here for @xmath196 into eq .
( [ midsimple ] ) finally leads to the equality of the upper bound on discord ( [ u ] ) and mid ( [ mid ] ) @xmath265 note that the two coincident quantities also provide an upper bound for the relative entropy of quantumness @xcite of the states .
note further that the lower bound on discord is also a lower bound for the other measures of nonclassical correlations such as amid , mid , and relative entropy of quantumness , since quantum discord is in general smaller than those mentioned quantities for arbitrary bipartite quantum states @xcite . , eq .
( [ u ] ) ( solid curve ) , and lower bound @xmath247 , eq .
( [ l ] ) ( dashed curve ) , on quantum discord ( [ d2 ] ) versus probability @xmath60 for the cv werner state ( [ rho ] ) with @xmath249 and @xmath266 .
the dark gray shaded region corresponds to separable states , the light gray shaded region corresponds to ppt states with unknown separability properties , and the white region corresponds to entangled non - ppt states .
the boundary probabilities @xmath267 and @xmath268 are depicted by vertical solid line and dashed line , respectively .
all the quantities plotted are dimensionless.,width=283 ]
one of the main technical disadvantages of the cv werner state with @xmath201 is that its eigenvalues , and consequently its global von neumann entropy , can not be calculated analytically .
interestingly , this ceases to be the case if the state is partially transposed @xcite .
then , one can find regions of parameters @xmath269 and @xmath270 for which the partial transposes are positive - semidefinite , and so represent a legitimate quantum state in their own right .
thus , one obtains another , sometimes simpler to treat family of non - gaussian quantum states for which we can get further with analytical tools than in the case of the original state .
let us illustrate this on a simple example of the cv werner state ( [ rho ] ) with @xmath271 . as was shown in ref .
@xcite for @xmath272 the partial transposition @xmath273 of the werner state @xmath202 with respect to mode @xmath6 , @xmath274 possesses the following nonnegative nonzero eigenvalues @xmath275 where @xmath276 , and is therefore a valid two - mode density matrix corresponding to a different non - gaussian state . direct substitution of the eigenvalues into eq .
( [ s ] ) gives the analytical expression for the global entropy of @xmath277 of the form @xmath278.\end{aligned}\ ] ] tracing the state ( [ tilderho ] ) over mode @xmath6 one gets the reduced state of mode @xmath8 @xmath279 with entropy @xmath280.\end{aligned}\ ] ] similarly one can find a reduced state @xmath281 of mode @xmath6 which coincides with the reduced state ( [ tilderhob ] ) and yields the local entropy @xmath282 . upon detecting @xmath54 photons in mode @xmath8 in the state ( [ tilderho ] ) ,
mode @xmath6 collapses into the normalized conditional state @xmath283 where @xmath284 is the probability of detecting @xmath54 photons on mode @xmath8 in the state ( [ tilderho ] ) .
after some algebra , the corresponding nonoptimized conditional entropy @xmath285 attains then the form @xmath286 where we have used eqs .
( [ stilderho ] ) and ( [ stilderhob ] ) .
substituting finally the latter formula into eq .
( [ u ] ) we arrive at a very simple analytical expression for the upper bound on the quantum discord of the state , @xmath287 even for the partially transposed cv werner state @xmath277 , it is possible to derive a nontrivial lower bound on quantum discord ( [ d2 ] ) . repeating the algorithm leading to eq .
( [ l ] ) for the state ( [ tilderho ] ) , one gets the lower bound in the form : @xmath288.\ ] ] the upper bound ( [ utilderho ] ) and lower bound ( [ ltilderho ] ) on discord are depicted in fig .
[ fig4 ] as a function of the parameter @xmath59 .
note that in this case they are quite close to each other , with the lower bound being always faithful in the whole considered parameter space . also , nonclassical correlations in this ppt state ( which may be separable or at most contain undistillable entanglement ) are quite weak ( yet always nonzero ) , increasing slowly with the squeezing @xmath20 ( recall that @xmath19 ) and converging to the small , finite value @xmath289 in the limit @xmath290 . , eq . ( [ utilderho ] ) ( solid curve ) , and lower bound @xmath247 , eq . ( [ ltilderho ] ) ( dashed curve ) , on quantum discord ( [ d2 ] ) versus squeezing parameter @xmath19 for the partially transposed cv werner state ( [ tilderho ] ) .
all the quantities plotted are dimensionless.,width=283 ] as for other nonclassicality indicators , we get equivalent results . moving for instance to the evaluation of the mid ( [ mid ] ) , one gets the joint photon - number distribution for the state ( [ tilderho ] ) to be @xmath291 the shannon entropy of the distribution reads @xmath292 where the global entropy @xmath293 is given in eq .
( [ stilderho ] ) .
the marginal distributions on each mode coincide and they are given by eq .
( [ tildepm ] ) .
hence for the state @xmath277 also , the local shannon and von neumann entropies satisfy eq .
( [ shequality ] ) . making use of the latter equality in the definition of mid ( [ mid ] ) we can express it as @xmath294 where in the derivation of the second equality we used eq .
( [ htildepab ] ) .
thus , as for the generic cv werner state of the previous section , for the considered partially transposed cv werner state the mid coincides with the upper bound on discord associated with local photon counting , i.e. , @xmath295 .
in this paper we studied the nonclassicality of correlations in a class of non - gaussian states of a two - mode cv system .
we adopted primarily quantum discord @xcite , as well as other quantifiers such as measurement - induced disturbance and relative entropy of quantumness , and provided exact results ( when possible ) and in general upper and lower bounds on their quantification in the considered states .
our analysis enabled us to venture beyond gaussian states and operations to access general features of quantum correlations in infinite - dimensional states .
we focused in particular on two - mode cv werner states @xcite , constructed as mixtures of gaussian states and thus possessing a positive , yet non - gaussian - shaped wigner function on the quantum phase space .
such states provide important testbeds for the understanding of nonclassicality beyond entanglement and its interplay with entanglement itself . for the special case of a two - mode squeezed state mixed with a vacuum @xcite
, we were able to calculate quantum discord exactly , proving that local photon counting constitutes the globally optimal strategy to extract nonclassical correlations , and any gaussian measurement strategy turns out to be suboptimal for the task .
the considered states constitute probably the simplest example of bipartite states possessing genuinely non - gaussian cv nonclassical correlations beside entanglement . the relationship between nonclassicality and non - gaussianity
was further highlighted by observing that the gap between the quantum discord and its ( non - optimal ) counterpart , restricted to gaussian measurements only , scales linearly with an entropic measure of non - gaussianity @xcite for the considered class of states in the regime of low squeezing .
for general cv werner states , photon counting measurements provide in general upper bounds on quantum discord and related nonclassicality measures , but we could not prove the tightness of such bounds analytically .
we are tempted to conjecture that our upper bound does yield the true quantum discord for all cv werner states , and future progress on validating or disproving this claim would be valuable . on the other hand , we derived nontrivial lower bounds on quantum discord as well for the considered states , which we do not expect to be tight .
we finally constructed a particular instance of a non - gaussian state which is positive under partial transposition , and whose upper and lower bounds on discord are analytically computable and allow us to pin down its nonclassical correlations as being quite weak , staying finite even in the limit of infinite squeezing .
our study evidences a trend for quantum correlations to be generally limited in the absence of ( distillable ) entanglement , as originally noted for gaussian states @xcite .
it would be interesting to provide tight upper bounds on the attainable amount of quantum discord for all separable cv states .
such a bound is known for two - mode gaussian states and corresponds to one unit of discord , on a scale ranging to infinity @xcite . on finite - dimensional two - qu@xmath25it systems , on the other hand
, it is known that separable mixed states can have as much nonclassical correlations as maximally entangled pure states , approaching the bound @xmath296 when measured by the relative entropy of quantumness @xcite .
nonclassicality and non - gaussianity are two of the most important resources for the optimization and realistic implementation of present - day and next - generation quantum technology @xcite .
this paper realizes a first step to explore the interplay between the two in physically relevant cv states .
an important next target for future work would be to study the structure of nonclassical correlations in other practically useful states deviating from gaussianity , such as photon - subtracted states @xcite . comparing the performance of non - gaussian measurements , such as photon counting , with that of gaussian strategies , such as homodyne and heterodyne detection , for accessing nonclassical correlations , and studying how the gap between the two scales with the non - gaussianity of the states @xcite , and with other nonclassical parameters widely used in quantum optics @xcite , could provide novel insight into the nature of quantumness ( in its broadest sense ) and its potential exploitation for cv quantum information .
99 a. k. ekert , phys . rev .
lett . * 67 * , 661 ( 1991 ) . c. h. bennett , g. brassard , c. crpeau , r. jozsa , a. peres , and w. k. wootters , phys .
lett . * 70 * , 1895 ( 1993 ) .
m. a. nielsen and i. l. chuang , _ quantum computation and quantum information _ ( cambridge university press , cambridge , 2000 ) .
j. kimble , nature * 453 * , 1023 ( 2008 ) ; see also _ the age of the qubit _ , iop publications ( 2011 ) , ` www.iop.org/publications/iop/2011/page_52065.html ` .
c. h. bennett and d. p. divincenzo , nature * 404 * , 247 ( 2000 ) . s. l. braunstein and p. van loock , rev .
phys . * 77 * , 513 ( 2005 ) ; g. adesso and f. illuminati , j. phys . a : math .
40 * 7821 , ( 2007 ) .
d. cavalcanti , l. aolita , s. boixo , k. modi , m. piani , and a. winter , phys .
a * 83 * , 032324 ( 2011 ) ; v. madhok and a. datta , phys .
a * 83 * , 032323 ( 2011 ) ; v. madhok and a. datta , arxiv:1107.0994 ( 2011 ) .
t. opatrny , g. kurizki , and d .-
welsch , phys .
a * 61 * , 032302 ( 2000 ) ; p. t. cochrane , t. c. ralph , and g. j. milburn , _ ibid . _ * 65 * , 062306 ( 2002 ) ; s. olivares , m. g. a. paris , and r. bonifacio , _ ibid . _ * 67 * , 032314 ( 2003 ) ; a. kitagawa , m. takeoka , m. sasaki , and a. chefles , phys . rev .
a * 73 * , 042310 ( 2006 ) ; f. dellanno , s. de siena , l. albano , and f. illuminati , phys .
a * 76 * , 022301 ( 2007 ) ; f. dellanno , s. de siena , g. adesso , and f. illuminati , phys .
a * 82 * , 062329 ( 2010 ) .
a. datta , a. shaji , and c. m. caves , phys .
lett * 100 * , 050502 ( 2008 ) ; b. p. lanyon , m. barbieri , m. p. almeida , and a. g. white , phys . rev
* 101 * , 200501 ( 2008 ) ; a. brodutch and d. r. terno , phys .
a * 83 * , 010301 ( 2011 ) ; b. eastin , arxiv:1006.4402 ( 2010 ) ; r. chaves and f. de melo , phys .
rev . a 84 , 022324 ( 2011 ) ; g. passante , o. moussa , d. a. trottier , and r. laflamme , arxiv:1105.2262 ( 2011 ) ; r. auccaise , j. maziero , l. celeri , d. soares - pinto , e. deazevedo , t. bonagamba , r. sarthour , i. oliveira , and r. serra , phys .
lett . * 107 * , 070501 ( 2011 ) .
s. luo , phys .
a * 77 * , 042303 ( 2008 ) ; m. ali , a. r. p. rau , and g. alber , phys . rev .
a * 81 * , 042105 ( 2010 ) ; see also erratum , phys . rev .
a * 82 * , 069902(e ) ( 2010 ) ; x. liu , j. ma , z. xi , and x. wang , phys . rev .
a * 83 * , 012327 ( 2011 ) ; d. girolami and g. adesso , phys .
a * 83 * , 052108 ( 2011 ) .
b. m. terhal , m. horodecki , d. w. leung , and d. p.divincenzo , j. math .
* 43 * , 4286 ( 2002 ) ; d. p. divincenzo , m. horodecki , d. leung , j. smolin , and b. m. terhal , phys . rev . lett .
* 92 * , 067902 ( 2004 ) . | we study nonclassical correlations beyond entanglement in a family of two - mode non - gaussian states which represent the continuous - variable counterpart of two - qubit werner states .
we evaluate quantum discord and other quantumness measures obtaining exact analytical results in special instances , and upper and lower bounds in the general case .
non - gaussian measurements such as photon counting are in general necessary to solve the optimization in the definition of quantum discord , whereas gaussian measurements are strictly suboptimal for the considered states .
the gap between gaussian and optimal non - gaussian conditional entropy is found to be proportional to a measure of non - gaussianity in the regime of low squeezing , for a subclass of continuous - variable werner states .
we further study an example of a non - gaussian state which is positive under partial transposition , and whose nonclassical correlations stay finite and small even for infinite squeezing .
our results pave the way to a systematic exploration of the interplay between nonclassicality and non - gaussianity in continuous - variable systems , in order to gain a deeper understanding of and to draw a bigger advantage from these two important resources for quantum technology . |
cosmic rays have the spectrum of @xmath7 up to the so - called knee - energy of @xmath8 .
cosmic ray propagation theories suggest @xmath9 energy spectra at the cosmic ray acceleration sites ( e.g. * ? ? ?
the current description of cosmic ray acceleration up to knee energy @xmath10 is the well known first - order fermi acceleration @xcite . in the first - order fermi acceleration model
, magnetic turbulence is an important agent for particle acceleration .
turbulence makes the particle momenta isotropic , thus portion of the particles cross the shock front many times .
expectation value of the kinetic energy after @xmath11 times of shock crossing is @xmath12 . on the other hand ,
the probability for a particle to survive @xmath11 shock crossing can be roughly estimated as @xmath13 .
this gives us the power - law spectrum of @xmath14 .
@xcite , @xcite , and @xcite have done simulations to describe the self - consistent generation of turbulence , with approximations such as gyro - center approximation , random walk approximation , or lowering the dimension . on the other hand ,
the recent development of the particle - in cell simulation has made it possible to describe the particle acceleration in electron - positron plasma self - consistently ( e.g. * ? ? ? * ) . in this letter
, we propose an alternative approach to the simulation of cosmic ray acceleration .
we have calculated the motion of particles deterministically , solving the particles cyclotron motion from larmor radii of thermal particles @xmath15 to that of knee energy particles @xmath16 . according to the theories , we assume turbulence spectrum in @xmath17 space .
this allows us to cover a large dynamic - range of space and energy , which enables us a direct comparison of the accelerated cosmic ray spectra with the observations .
[ [ upstream - and - downstream - regions ] ] upstream and downstream regions + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + in our method , the electromagnetic field and velocity field of a continuous region are given by @xmath18 where the amplitude and the wavenumber of the @xmath19-th mode are @xmath20 and the initial phase of the @xmath19-th mode is @xmath21 .
here @xmath22 is the spectral index that reflects the nature of the turbulence and @xmath23 is the total number of the modes ( @xmath24 ) , @xmath25 is the amplitude for each mode , @xmath26 is its wavenumber , and @xmath27 are two mutually perpendicular unit vectors that are perpendicular to @xmath28 .
we choose @xmath26 to be either parallel or antiparallel to @xmath6.equation ( [ eq_log_k ] ) means that @xmath29 are logarithmically discrete .
we use @xmath30 as the measure of the strength of the fluctuation , independent of @xmath23 .
because increasing @xmath23 while keeping @xmath31 ( 1 ) keeps the magnetic energy in fluctuation mode , and ( 2 ) keeps the expectation value of the fluctuation field @xmath32 the same , if @xmath33 are independent .
we will confirm these properties in section [ result ] .
the argument to derive @xmath34 in @xmath17 space is summarized below : variables in @xmath17 space are marked by tilde .
the power law energy spectrum is @xmath35 in kolmogorov turbulence case .
this energy spectrum is in linear bin . in log energy bin
the spectral power is @xmath36 ; and since @xmath37 , @xmath38 .
thus , we get @xmath34 for our discretization of the turbulent magnetic field .
[ [ junction - conditions ] ] junction conditions + + + + + + + + + + + + + + + + + + + we assumed strong shock junction condition with low plasma @xmath39 limit at the shock front : @xmath40 , @xmath41 , and @xmath42 , where @xmath43 is the shock compression ratio , @xmath44 and @xmath45 are components of the @xmath46 normal and tangential to the shock , respectively . and @xmath47 , @xmath48 at @xmath49 , @xmath50 , and @xmath51 years of time evolution .
time integrated energy spectrum is shown with the bold curve . ] and @xmath47 , @xmath48 at @xmath49 , @xmath50 , and @xmath51 years of time evolution .
time integrated energy spectrum is shown with the bold curve . ] [ [ initial - condition ] ] initial condition + + + + + + + + + + + + + + + + + for each set of initial condition we introduce electromagnetic fields described in section [ representation ] .
we choose a set of initial turbulence phase @xmath52 , sign of @xmath53 and @xmath54 from uniform distribution .
then we put @xmath55 protons in boltzmann distribution of temperature @xmath56 at the upstream side of the shock .
we use the values in table [ parameter_table ] , based on @xcite s observation of sn 1006 .
cc @xmath57 & fluid speed in shock frame + @xmath58 & alfven speed in fluid frame + @xmath59 & unperturbed magnetic field strength + @xmath60 & torsional alfven mode energy + @xmath61 & angle of @xmath6 to shock normal + @xmath62 & maximum wavelength of turbulence + @xmath63 & minimum wavelength of turbulence + @xmath64 & temperature of the particles + @xmath65 & particle mass + @xmath66 & particle charge + [ [ evolution ] ] evolution + + + + + + + + + we make each turbulence mode propagate at alfven velocity of the uniform field @xmath67 as in equation [ equationb ] - [ equatione ] , and updated the particle with lorenz force , with 4-th order runge - kutta method .
we choose time discretization @xmath68 for each particle at every timestep , so that @xmath69 and @xmath70 always hold .
typical time step is @xmath71 whereas the larmor period of thermal particle for @xmath72 is @xmath73 .
we have made following examinations for the results of our method .
first , we have traced the particles energy @xmath75 as the function of shock crossing number @xmath11 ( figure [ jump - energy ] ) .
the inclination of the curves match the inclination of the theoretical prediction , @xmath76 .
secondly , we have traced the spatial location where the particles gained their kinetic energy .
we have found that 94% of the final kinetic energy have been earned within @xmath49 final larmor radius away from the shock .
this is consistent with the first - order fermi acceleration picture .
thirdly , we have studied the validity of our fourier representation in @xmath17 space .
we have kept the physical parameters and increased the number of modes per decade @xmath77 ; we see that the spectra converge , and converge to the theoretical power - law spectrum ( figure [ energy - spectrum - convergence ] ) .
this justifies our use of @xmath78 space discretization .
we have done a large number of simulations while varying the background conditions , @xmath79 from @xmath80 to @xmath81 , @xmath4 , the ratio of magnetic energy in fluctuation mode to that in background field from @xmath82 to @xmath83 , @xmath84 , the angle between the background field and the shock normal from @xmath85 to @xmath86 .
figure [ energy - spectrum - age ] shows the time evolution of the energy spectrum for @xmath51 years .
the high - energy end of the spectrum gradually grows , and reaches @xmath87 by @xmath51 years
. in our simulation all the particles start its motion in the given time .
since we do nt include the back - reaction from the particles to the electromagnetic field in our simulations , time - integral of energy spectra at each time - slice gives the steady state energy spectra .
this steady state spectrum is also shown in figure [ energy - spectrum - age ] with thick curve .
the nonthermal component has @xmath88 power - law spectrum that meets the observational requirement mentioned in section [ introduction ] .
we can also estimate the `` injection rate '' to be the proportion of particles that have more than @xmath89 of energy after 1 years . for @xmath90 , @xmath91 , @xmath92 , and @xmath93 ,
the injection rate was @xmath94 , @xmath95 , @xmath96 , and @xmath97 , respectively.the other parameters are @xmath98 , @xmath48 and @xmath99 . in figure [ dsa - map ]
we show for all the parameter range the ratio of the particle numbers that were accelerated to have energy greater than @xmath100 .
we see that the acceleration is most efficient at polar region ( @xmath101 ) when @xmath102
. we can understand this dependence of the spectra with background fluid parameters as follows ; particles are trapped in larmor motion and tend to move in direction of @xmath6 . thus particles more easily cross the shockfront when @xmath6 is parallel to shock normal .
if the turbulence amplitude is much weaker , fewer particles get reflected by pitch angle scattering , and fermi acceleration is suppressed .
the injection is more efficient for smaller @xmath79 , because more energy is distributed to modes with smallest wavelengths which are resonant with the thermal particles .
if a spherical shock emerges in a uniform mean magnetic field , there are two polar region where the mean magnetic field is parallel to the shock normal , and the equatorial region has the mean magnetic field perpendicular to the shock normal .
thus , we expect the fermi acceleration process to be only active in the pair of polar region .
this might explain the bipolar structure we see at sn 1006 .
we have also checked the acceleration rate in three - dimensional(isotropic ) , rather than one - dimensional(anisotropic ) distribution of @xmath103 .
we have found that less significant dependence of injection rate on @xmath84 with larger @xmath4 .
we can interpret this as follows ; if the turbulence spectrum is isotropic and the maximum turbulence wavelength is large , turbulence modes with largest wavelength and strongest amplitude play the role of local @xmath6 .
thus we observe almost isotropic fermi acceleration .
this might explain the many snrs with no typical orientation .
is the longest wavelength of the turbulence modes , @xmath104 is the ratio of turbulent magnetic field to unperturbed magnetic field , @xmath84 is the angle between shock normal and @xmath6 . ]
some might question the validity of @xmath4 value much greater than unity .
however , @xcite observed extremely fast varying x - ray images at snr rx j1713.7 - 3946 .
their observation may indicate that magnetic field is locally enhanced up to @xmath105 in @xmath106 year in snr , which corresponds to @xmath107 case in our model .
our simulations suggest that such fast - variating spots in snrs might be the sites of galactic @xmath108 cosmic ray acceleration .
although we have ignored many of the fourier modes by adopting @xmath17 space discretization of the turbulence spectrum , the validity of the approximation can be argued in many ways .
most importantly we have confirmed that our measure in @xmath17 space lead to convergence .
figure [ energy - spectrum - convergence ] shows energy spectra for @xmath31 , with increasing @xmath109 .
turbulent cascade , from which the very turbulence arises , is by nature a logarithmic process : a mode of a certain wavelength couples with the mode of half the wavelength by nonlinear term of euler equation .
first order fermi acceleration also is a logarithmic process : particles gain energy as an exponential function of shock crossing number @xmath12 . all these reason combined , waves in logarithmically discretized wavenumber space act as a sufficient ladder to carry up cosmic ray particles .
the authors thank s. inoue and k. murase for useful comments .
they also thank center for computational astrophysics ( cfca ) of national astronomical observatory of japan and yukawa institute for theoretical physics ( yitp ) in kyoto university for their computing facilities .
the numerical calculations were carried out on cray xt4 at cfca and altix3700bx2 at yitp .
s. i. is supported by grants - in - aid ( 15740118 , 16077202 , and 18540238 ) from ministry of education , culture , sports , science and technology ( mext ) of japan .
this work was supported by the grant - in - aid for the global coe program `` the next generation of physics , spun from universality and emergence '' from the mext of japan . | we simulate the acceleration processes of collisionless particles in a shock structure with magnetohydrodynamical ( mhd ) fluctuations .
the electromagnetic field is represented as a sum of mhd shock solution ( @xmath0 ) and torsional alfven modes spectra ( @xmath1 ) .
we represent fluctuation modes in logarithmic wavenumber space .
since the electromagnetic fields are represented analytically , our simulations can easily cover as large as eight orders of magnitude in resonant frequency , and do not suffer from spatial limitations of box size or grid spacing .
we deterministically calculate the particle trajectories under the lorenz force for time interval of up to ten years , with a time step of @xmath2 .
this is sufficient to resolve larmor frequencies without a stochastic treatment .
simulations show that the efficiency of the first order fermi acceleration can be parametrized by the fluctuation amplitude @xmath3 .
convergence of the numerical results is shown by increasing the number of wave modes in fourier space while fixing @xmath4 .
efficiency of the first order fermi acceleration has a maximum at @xmath5 .
the acceleration rate depends on the angle between the shock normal and @xmath6 , and is highest when the angle is zero .
our method will provide a convenient tool for comparing collisionless turbulence theories with , for example , observations of bipolar structure of super nova remnants ( snrs ) and shell - like synchrotron - radiating structure . |
the edge - on spiral galaxy ngc 253 is one of the two prototypical , nearby starburst galaxies ( along with m82 ) .
its distance of 2.5 mpc @xcite enables its starburst disk to be studied at very high linear resolution , since 1 corresponds to only 12 pc .
the galaxy contains an inner starburst disk roughly 1520 ( 180240 pc ) in extent that has been studied in detail , especially by virtue of its emission in the infrared @xcite and in molecular lines @xcite .
subarcsecond - resolution imaging at centimeter wavelengths reveals at least 64 compact radio sources in the inner disk @xcite ; these are thought to be roughly equally divided between 2 regions dominated by thermal emission , and supernova remnants dominated by nonthermal emission @xcite . at the centimeter wavelengths where radio telescopes are most sensitive , the poorer resolution and the increasing strength of the more diffuse galaxy emission lead to confusion that prevents complete source identification in the inner 200-pc starburst .
the 200-pc region clearly dominates the current star formation in ngc 253 , as shown by the infrared , millimeter , and centimeter observations .
centimeter imaging of this inner starburst indicates a supernova rate of @xmath60.3 yr@xmath4 @xcite , which is consistent with results of 0.10.3 yr@xmath4 inferred from models of the infrared emission of the entire galaxy @xcite .
near- and mid - infrared imaging at arcsecond resolution shows a number of emission peaks that do not generally line up with the radio sources @xcite , while hubble space telescope images show at least four compact star clusters , which also do not coincide with individual compact radio sources @xcite .
thus , it is apparent that extinction and confusion play major roles in the inner disk at different wavebands , and it may be that only the centimeter and millimeter images reveal the regions of most recent star formation . outside the 200-pc starburst disk , ngc 253 may have a significant amount of star formation , as shown by the existence of larger - scale radio emission @xcite .
since this part of the galaxy is much less affected by confusion and extinction , there is greater potential for identifying the strongest regions of star formation and studying them in multiple bands .
however , recent high - resolution studies of the galaxy outside the inner few hundred parsecs are uncommon , since most of the imaging has been of small fields centered on the main starburst disk .
one recent study on the larger scale is that of @xcite , who used rosat to identify 73 x - ray sources in the bulge , disk , and halo of ngc 253 , attributing most of them to x - ray binary stars . to study the population of compact radio sources on this large scale
, we have reprocessed the high - resolution 6-cm and 20-cm very large array ( vla ) data obtained in 1987 @xcite in order to image the entire primary beam of the individual vla antennas at high resolution ; limited computing resources prevented this from being done at the time of the observations . in this paper , we report the locations of the compact radio sources outside the inner 200-pc starburst , compare these locations to images made in other wavebands , and estimate the supernova rate outside the inner disk .
the observations of ngc 253 that we use here were made in 1987 , using the vla @xcite in its a configuration .
results from imaging the inner starburst at 6 cm ( 4860 mhz ) were reported by @xcite , while the 20-cm ( 1490 mhz ) image of the same region was discussed briefly by @xcite .
details of the observations were reported in those papers .
other 6-cm observations had much poorer ( _ u , v _ ) coverage than those made in 1987 , so only the 1987 data are considered here . for self - consistency ,
the data were re - calibrated and self - calibrated using the same procedures reported previously .
we then made large - scale ( @xmath7 pixels ) images of the radio emission of ngc 253 to reveal the compact radio sources seen throughout the primary beam of the individual 25-m telescopes of the vla .
these images covered @xmath8 at 20 cm and @xmath9 at 6 cm . outside the starburst disk , confusion was not a significant issue , but sensitivity was critical .
we therefore used natural weighting of the data in the ( _ u , v _ ) plane @xcite ; this scheme produces the maximum sensitivity , at the price of a modest loss in resolution .
a zero - spacing ( total ) flux density was specified in the imaging , in order to partially compensate for the lack of short interferometer spacings .
this reduced , but did not completely remove , systematic effects caused by diffuse emission that was not sampled well in our observations .
table [ tab : obssum ] gives the resolution and the r.m.s .
noise achieved for the final images .
several effects contribute to reductions in the sensitivity far from the field center ; each is discussed briefly here .
first , the attenuation of the primary beam of the individual vla antennas reduces the sensitivity by more than 10% at 6 cm and 20 cm for respective distances from the pointing direction of greater than @xmath10 and @xmath11 .
corrections for this attenuation were made using the aips @xcite software developed by nrao , and were less than 10% except for several sources ( see section [ background ] ) far from the main area of radio emission . a second effect , chromatic aberration ( `` bandwidth smearing '' ) due to the non - zero observing bandwidth @xcite , is more deleterious .
it reduces the peak flux density by more than a factor of two at distances greater than 2 from the phase center , and is progressively worse farther out in the field .
the actual phase center of the observations was @xmath12 north of the apparent galaxy nucleus , at a b1950 position of @xmath13 ; since most of the large - scale sources are north of the nucleus , this actually was slightly beneficial in the final data analysis . compared to the chromatic aberration , the reduction in flux density due to delay smearing @xcite over the observations 10-second averaging period is negligible .
finally , the lack of short interferometer spacings causes correlated positive or negative emission as high as 0.080.1 mjy over areas of @xmath14 pixels away from the center of the field .
the final source detection threshold that was used was four times the quadrature sum of this systematic value and the r.m.s .
noise , or approximately 0.4 mjy at each wavelength .
however , this threshold rises substantially at distances more than a few arcminutes from the phase center .
figure [ lbig ] is a portion of the 20-cm image showing the circumnuclear region of ngc 253 at high resolution .
the locations of compact x - ray sources detected by @xcite , using the rosat high resolution imager , are shown as crosses whose sizes indicate the errors in the x - ray positions .
we have identified all radio sources that have apparent peak flux densities of at least 0.4 mjy beam@xmath4 at either 6 cm or 20 cm , outside the central 200-pc starburst .
figures [ close]a and [ close]b are respective enlargements of the 20-cm and 6-cm images showing the radio emission from the 200-pc starburst disk and its surroundings , including a number of the circumnuclear compact sources .
note that the 6-cm image shows little detail , due to the larger number of beam areas needed to cover the given region , but it does serve to illustrate the relative locations of the detected sources .
table [ tab : sor ] lists the source identifications in the circumnuclear region at the two wavelengths .
( each source listed in this table can be seen in the region shown in figure [ lbig ] . )
included are the source positions , flux densities at 6 cm and 20 cm , measured sizes , and a few comments .
when compact sources are identified at both wavelengths , the higher resolution 6-cm position is quoted ; typical position accuracies are @xmath15@xmath16 . at 6 cm ,
flux densities as low as 0.2 mjy are given for sources that meet the 0.4-mjy threshold at 20 cm , but these flux densities have very large fractional errors .
source flux densities and sizes have been derived by making gaussian fits to the individual sources in the image plane , and confirmed by integrating over the region of significant emission .
most sources are unresolved in the full - resolution images at both 6 cm and 20 cm , as indicated by a `` u '' in the table
. the sizes of resolved sources are derived from the 6-cm full - resolution image .
the total 6-cm flux densities and approximate sizes found at full resolution generally are consistent with values found from a 6-cm image tapered to the 20-cm resolution .
we estimated flux density errors due to several causes .
first , the areas of apparent systematic negative or positive flux near the main starburst disk , caused by undersampling in the aperture plane , average as much as 0.08 mjy beam@xmath4 at 6 cm and 0.10 mjy beam@xmath4 at 20 cm ; this offset is denoted by @xmath17 .
second , there is a @xmath18% error in the absolute flux density scales of each of the two maps , so @xmath19 , where @xmath20 is the source flux density .
third , there is a fitting error , @xmath21 , caused by confusion with underlying diffuse emission and uncertainties in the source fits .
since the newly identified sources reported here are outside the main starburst disk and fairly well isolated , we take the confusion error to be negligible , so @xmath21 is simply the uncertainty reported by the least - squares fitting program .
this value includes the r.m.s .
noise given in table [ tab : obssum ] , so the noise is not included separately .
we combine these errors in quadrature to get the total flux - density error , @xmath22 , using @xmath23 thus , the total flux - density error for a 0.4-mjy source is typically 0.09 mjy at 6 cm and 0.11 mjy at 20 cm .
there are 22 sources above the 0.4-mjy limit at 20 cm , inside a box 4 ( 2.8 kpc ) on a side and centered on the nucleus of ngc 253 .
the density of background sources above this flux - density threshold is expected to be @xmath24 sr@xmath4 @xcite .
therefore , we expect about one background source stronger than 0.4 mjy within a field of 16 square arcminutes , implying that almost all of the detected sources are associated with ngc 253 . the apparent lack of weak sources at greater distances from the nucleus may be caused primarily by chromatic aberration ( see section [ obs ] ) ; sources far from the phase center of the observation have their peak flux densities reduced substantially , and therefore may fall below the detection threshold .
observations using narrow spectral - line channels rather than a broadband continuum would be needed to determine whether other weak compact sources exist farther from the phase center .
five additional sources were detected at 20 cm within an area 12 on a side , centered on ngc 253 , and are listed in table [ tab : back ] .
these are within the size of the optical galaxy , which has a measured diameter of 25 to 25th magnitude @xcite
. images of these sources are not shown , since they are substantially degraded by chromatic aberration , which prevents measurement of the source sizes .
therefore , only the total flux densities are given in table [ tab : back ] .
the source locations are best seen in previous lower - resolution radio images of ngc 253 , particularly that published by @xcite , though none of those images resolves source b2 into the three components distinguished here .
only sources much stronger than 0.4 mjy can be detected in the large region due to the chromatic aberration .
the expected density of sources stronger than 15 mjy at 20 cm is @xmath25 sr@xmath4 @xcite , so we expect to detect less than one background source above this threshold .
source b3 appears to be associated with a `` spur '' noted by @xcite , and the three - component source b2 lies along the large - scale galaxy disk , making it probable that at least b2 and b3 are associated with ngc 253 .
the detected circumnuclear sources ( table [ tab : sor ] ) outside the central starburst have typical flux densities of 0.43 mjy at 6 cm and 20 cm , corresponding to radio powers of @xmath26 to @xmath27 w hz@xmath4 .
most of them seem to have relatively steep spectra ; for a source with a flux density of 0.4 mjy at 6 cm and a spectral index of @xmath28 ( with @xmath29 ) , the total luminosity between 10 mhz ( or 100 mhz ) and 100 ghz is @xmath30 erg s@xmath4 .
this luminosity is a factor of 1001000 lower than the luminosities of the point x - ray sources detected by @xcite , which they generally attribute to x - ray binaries .
since none of the radio sources listed in table [ tab : sor ] coincides with a compact x - ray source , the radio emission is probably not associated with evolved binaries . in the main 200-pc starburst disk of ngc 253
, @xcite deduced that @xmath31 compact radio sources exist . a supernova rate of @xmath32 yr@xmath4 was estimated from the radio emission @xcite , while supernova rates of 0.10.3 yr@xmath4 have been derived on other grounds @xcite . here , we report an additional 22 circumnuclear compact radio sources that lie outside the central 200-pc starburst , but within 2 kpc ( 170 ) of the galaxy center .
since they do not coincide with x - ray binaries , and most are less than 510 pc ( @xmath33@xmath34 ) in diameter , they most likely are due to supernova remnants and 2 regions , as are the sources in the nuclear region .
the sources can be compared to the galactic supernova remnant cas a. for an assumed distance of 2.8 kpc @xcite and a flux density of @xmath35 jy @xcite , cas a has a 6-cm radio power of @xmath36 w
hz@xmath4 , comparable to the powers of the weaker circumnuclear sources in ngc 253 . there are 1015 circumnuclear sources with steep radio spectra , likely to be supernova remnants .
the number of steep - spectrum sources is uncertain because some are too weak for good spectral - index determination .
in addition , since the ( _ u , v _ ) coverage at 6 cm is not matched to that at 20 cm , more flux density may be `` resolved out '' at 6 cm , which could lead to estimates of spectra that are overly steep .
( this resolution effect is the primary reason that spectral indices are not quoted in table [ tab : sor ] . ) by comparison , there are at least 32 steep - spectrum sources in the inner 200-pc starburst @xcite ; correcting for confusion would probably increase this number to 50 or more . if the circumnuclear radio sources have the same general character as those in the main starburst disk , a simple comparison with the analysis of @xcite indicates that the estimated circumnuclear supernova rate outside the central starburst is @xmath37 yr@xmath4 .
( of course , some sources outside the strong starburst could have been missed , due primarily to chromatic aberration . ) in any case , based solely on the ratio of steep - spectrum sources inside and outside the 200-pc starburst , we infer that at least 20%30% of the global star - formation and supernova remnants are outside that central starburst .
no supernovae have been detected in the circumnuclear region during limited searches made at optical wavelengths in 19881991 @xcite , and at near - infrared wavelengths in late 1993 @xcite .
however , such searches did not cover long enough time periods to expect supernova detections in ngc 253 , and also could have been affected by obscuration and confusion .
figure [ blowup ] shows a complex of sources to the west of the main starburst disk , roughly 800 pc from the galaxy nucleus .
figure [ blowup]a is the 20-cm image , showing four individual radio sources . these sources are numbered 4 , 6 , 7 , and 8 in table [ tab : sor ] and similarly labeled in the figure .
two 6-cm images also are shown .
figure [ blowup]b is an image for which the visibility data were tapered and then restored with a point - spread function identical to that at 20 cm , while figure [ blowup]c is a full - resolution image . inspection of table [ tab : sor ] reveals that the two strongest 6-cm sources shown in figure [ blowup]b have flat or inverted spectra . summing the four individual compact sources
gives a total 6-cm flux density of @xmath38 mjy and a total 20-cm flux density of @xmath39 mjy , while integration of the images over the region of figure [ blowup ] yields a total 6-cm flux density of 7.7 mjy and a total 20-cm flux density of 9.0 mjy , slightly higher than the fits to the compact sources .
we take the somewhat simplistic step of making a two - component decomposition of the flux density in the radio sources shown in figure [ blowup ] . to do so
, we assume the presence of a thermal , flat - spectrum component having @xmath40 , and a nonthermal , steep - spectrum component due to supernova remnants .
the steep - spectrum sources in the inner starburst of ngc 253 have typical spectral indices of @xmath41 @xcite , while galactic supernova remnants have a median spectral index of @xmath42 @xcite , so there is some range in the estimates of the flat - spectrum component .
the simple spectral decomposition gives thermal radio emission containing 5.3 to 5.9 mjy out of a total of 6.6 mjy at 6 cm , for a steep - spectrum component having a spectral index ranging from @xmath43 to @xmath28 .
thus , 80% to 90% of the 6-cm flux density in the region shown in figure [ blowup ] appears to be thermal in origin .
the natural conclusion is that the flat - spectrum radio component is thermal radiation from 2 regions energized by young stars .
applying the analysis of the strongest flat - spectrum source in the starburst disk , which was described by @xcite , we find that the equivalent of @xmath44 o5 stars is necessary to energize the radio emission within a region of about @xmath45 ( @xmath46 pc ) .
this is comparable to the number of young stars needed to account for the strongest thermal source in the inner starburst , but those stars are contained in a volume @xmath47 times higher .
therefore , the average thermal gas density in the 130-pc region shown in figure [ blowup ] is @xmath48 @xmath5 , much lower than the values of @xmath25 @xmath5 found in the dense part of the inner starburst of ngc 253 @xcite and in ngc 5253 @xcite , or of @xmath49 @xmath5 for the typical thermal radio sources in ngc 4038/4039 @xcite .
few high - resolution observations at other wavelengths are available for the region shown in figure [ blowup ] .
however , a moderate - resolution ( @xmath50 ) vla image @xcite shows a source at the same location , with an apparent 3.6-cm flux density between 5 and 6 mjy .
this is entirely consistent with the value expected for for the flat - spectrum component deduced above .
near - infrared images shown in figures 1 and 10 of @xcite indicate a slight enhancement in k ( 2.2 @xmath51 m ) at the location of the complex of compact radio sources , apparently along an inner spiral arm .
this infrared source may represent a highly reddened set of 2 regions containing the numerous young stars that energize the local radio complex .
a similar argument can be made for source 15 , which lies nearly 2 kpc from the galaxy nucleus , and also appears to have a thermal spectrum .
ionization of this radio source requires the equivalent of about 35 o5 stars in a region about 1530 pc in diameter , and an average ionized density near 200 @xmath5 .
chromatic aberration makes these estimates somewhat uncertain , but they should be correct to 50% or better .
the relatively low average densities derived for the two strongest thermal - emitting complexes in the circumnuclear region imply that they contain more `` normal '' star formation , rather than the intense starbursts characteristic of the central 200 parsecs .
the number and the density of massive young stars are significantly higher than in orion and other nearby galactic o - b associations @xcite , and may be comparable to the richest star - forming regions in our galaxy , such as w49 @xcite
. however , the intensity of massive star formation is considerably lower than in 30 doradus @xcite or in the super star clusters seen in a number of starburst galaxies @xcite .
we have used archival vla data to image the circumnuclear region of ngc 253 at arcsecond resolution .
twenty - two compact radio sources have been found in the inner 2 kpc of the galaxy , but outside the well - known 200-pc disk , and most of these are probably associated with regions of recent star formation .
the supernova rate inferred outside the central starburst is @xmath30.1 yr@xmath4 ; this may be a slight underestimate due to the decreasing sensitivity of the radio observations at distances more than @xmath1 from the nucleus .
therefore , the region outside the well - studied inner starburst seems to account for a significant fraction of the recent star formation in ngc 253 .
a collection of sources located 800 pc to the west of the nucleus appears to be a complex of 2 regions energized by the equivalent of 70 o5 stars , but with an average ionized gas density @xmath49 times lower than that found in the inner starburst of the galaxy .
i thank k. anantharamaiah , r. antonucci , n. mohan , and w. pietsch for useful discussions and for providing data in advance of publication .
i especially thank the referee , jean turner , for some perceptive comments and for pointing out errors in table [ tab : sor ] .
this research has made use of the nasa / ipac extragalactic database ( ned ) which is operated by the jet propulsion laboratory , california institute of technology , under contract with the national aeronautics and space administration .
in addition , this research has made use of nasa s astrophysics data system .
lllccc 1&@xmath52 & @xmath53 & @xmath54 & @xmath55 & u + 2&00 44 59.498 & @xmath5625 33 54.91 & @xmath57 & @xmath58 & u + 3&00 45 00.117 & @xmath5625 34 20.89 & @xmath59 & @xmath55 & u + 4&00 45 00.580 & @xmath5625 33 37.30 & @xmath60 & @xmath61 & 0.9 @xmath62 0.7 , 89 + 5&00 45 00.619 & @xmath5625 34 15.25 & @xmath63 & @xmath54 & u + 6&00 45 00.785 & @xmath5625 33 35.66 & @xmath64 & @xmath65 & 1.1 @xmath62 0.9 , 7 + 7&00 45 01.007 & @xmath5625 33 31.19 & @xmath66 & @xmath67 & 0.6 @xmath62 0.6 + 8&00 45 01.187 & @xmath5625 33 35.61 & @xmath68 & @xmath69 & 1.9 @xmath62 1.5 , 145 + 9&00 45 02.495 & @xmath5625 34 00.40 & @xmath66 & @xmath70 & u + 10&00 45 02.659 & @xmath5625 33 04.87 & @xmath63 & @xmath54 & u + 11&00 45 03.346 & @xmath5625 32 59.96 & @xmath54 & @xmath71 & u + 12&00 45 03.497 & @xmath5625 33 00.15 & @xmath63 & @xmath55 & u + 13&00 45 04.236 & @xmath5625 32 46.97 & @xmath57 & @xmath72 & u + 14&00 45 05.998 & @xmath5625 32 29.16 & @xmath59 & @xmath54 & u + 15&00 45 06.097 & @xmath5625 31 55.60 & @xmath73 & @xmath74 & + 16&00 45 06.153 & @xmath5625 32 26.66 & @xmath72 & @xmath55 & u + 17&00 45 09.602 & @xmath5625 33 55.91 & @xmath63 & @xmath72 & u + 18&00 45 09.804 & @xmath5625 33 06.73 & @xmath75 & @xmath76 & 1.2 x 0.9 , 2 + 19&00 45 12.084 & @xmath5625 32 56.00 & @xmath77 & @xmath74 & + 20&00 45 13.136 & @xmath5625 33 00.97 & @xmath78 & @xmath79 & + 21&00 45 13.308 & @xmath5625 33 08.83 & @xmath63 & @xmath55 & u + 22&00 45 15.846 & @xmath5625 32 01.53 & @xmath80 & @xmath81 & + [ tab : sor ] lllccl b1&@xmath82 & @xmath83 & & @xmath84 & outside 6 cm window + b2a&00 45 17.08 & @xmath5625 29 58.6 & @xmath85 & @xmath86 & undetected at 6 cm + b2b&00 45 17.45 & @xmath5625 29 59.7 & @xmath87 & & blends with b2c at 20 cm + b2c&00 45 17.54 & @xmath5625 30 00.6 & @xmath88 & @xmath89 & blends with b2b at 20 cm + b3&00 45 31.33 & @xmath5625 34 52.6 & & @xmath90 & outside 6 cm window + [ tab : back ] | archival vla data has been used to produce arcsecond - resolution 6- and 20-cm images of the region surrounding the nuclear 200-pc ( @xmath0 ) starburst in ngc 253 .
twenty - two discrete sources stronger than 0.4 mjy have been detected within @xmath1 kpc ( @xmath23 ) of the galaxy nucleus ; almost all these sources must be associated with the galaxy .
none of the radio sources coincides with a detected x - ray binary , so they appear to be due to supernova remnants and 2 regions .
the region outside the central starburst has a derived radio supernova rate of @xmath30.1 yr@xmath4 , and may account for at least 20% of the recent star formation in ngc 253 .
most of the newly identified sources have steep , nonthermal radio spectra , but several relatively strong thermal sources also exist , containing the equivalent of tens of o5 stars .
these stars are spread over tens of parsecs , and are embedded in regions having average ionized gas densities of @xmath220200 @xmath5 , much lower than in the most active nuclear star - forming regions in ngc 253 or in the super star clusters seen in other galaxies .
the strongest region of thermal emission coincides with a highly reddened area seen at near - infrared wavelengths , possibly containing optically obscured 2 regions . |
bepcii / besiii is an upgrade facility from the previous bepc / bes @xcite .
the collider experiment besiii is monitoring the double - ring electron - positron head - on collisions produced by the bepcii machine .
the luminosity is optimized at a center - of - mass energy of 3.78and the peak record has reached to @xmath3@xmath4s@xmath5 , which luminosity is one order of higher than that at cesr - c @xcite .
the beam energy ranges from 1.0@xmath6 to 2.3@xmath6 .
therefore , the physics in besiii cover the @xmath7-charm physics . the besiii spectrometer @xcite as shown in fig .
[ fig : bes ] , consists of the following main components with order of the distance to the interaction position : a main draft chamber with momentum resolution 0.5% at 1 ; an electromagnetic calorimeter with energy resolution 2.5% at 1.0 ; a time - of - flight counters ; a superconducting magnet with a field of 1 t ; a muon chamber system made of resistive plate chambers . in 2009 ,
the besiii experiment has collected about 225 m @xmath0 and 106 m @xmath1 samples , both of which are the world largest on - peak charmonium production .
based on these dataset , besiii has made great effort on the study of the charmonium decays , some important of which have been reviewed in this proceeding .
in addition , a search for new physics through the @xmath2 violation process is reported .
the radiative decay of the @xmath1 to a pseudo - scalar meson provide important tests for various mechanisms in understanding the low-@xmath12 phenomena , such as the vector meson dominance model ( vdm ) , the two - gluon coupling to @xmath13 states , the @xmath14 mixing angle , the final - state radiation by light quarks .
recently , the cleo - c collaboration reported measurements for the decays of @xmath0 , @xmath1 , and @xmath15 to @xmath16 @xcite , and no evidence for @xmath17 or @xmath18 was found .
besiii studied the processes @xmath19 with @xmath20 , and @xmath17 with @xmath21 $ ] , and @xmath22 with @xmath23 .
the analyses use the 106 m @xmath1 data sample .
the results show that @xmath19 and @xmath17 are observed for the first time with significance of 4.6@xmath24 and 4.3@xmath24 , respectively , and with branching fractions of @xmath25 and @xmath26 . branching fraction of the process @xmath22
is measured with improved accuracy to be @xmath27 .
the mass distributions of the pseudoscalar meson candidates are shown in fig .
[ fig : psip - gp ] . for the first time
, besiii determined the ratio of the @xmath10 and @xmath11 production rates from @xmath1 decays , @xmath28% .
this ratio is in agreement with 90% c.l .
upper limit determined by cleo - c and in contradiction to predictions of leading - order perturbative qcd .
this is also smaller than the corresponding ratio for the @xmath29 decays by an order of magnitude , which was measured to be @xmath30% .
more details on this study can be found in ref .
@xcite .
: a ) @xmath18 ; b ) @xmath31 ; c ) @xmath32 ; d ) @xmath33 ; e ) @xmath34 .
refer to ref .
@xcite for details.,scaledwidth=70.0% ] ( -160 , 105 ) * ( a ) * : a ) @xmath18 ; b ) @xmath31 ; c ) @xmath32 ; d ) @xmath33 ; e ) @xmath34 .
refer to ref .
@xcite for details.,scaledwidth=140.0% ] : a ) @xmath18 ; b ) @xmath31 ; c ) @xmath32 ; d ) @xmath33 ; e ) @xmath34 .
refer to ref .
@xcite for details.,title="fig:",scaledwidth=140.0% ] ( -350,250 ) * ( b ) * ( -350,150 ) * ( c ) * ( -155,250 ) * ( d ) * ( -155,150 ) * ( e ) * ( -235,39 ) * mass ( ) * ( -380,110 )
besiii has world largest @xmath1 on - peak production , which endow besiii with the largest radiative decay @xmath35 data .
@xmath35 hadronic decay helps in understanding the @xmath36-wave charmonium decay dynamics .
most of hadronic decay channels of @xmath35 are not still known and the pursued measurement provide a good laboratory to test the color singlet and octet mechanism in interpreting @xmath35 hadronic decays . at besiii , @xmath35 decays to @xmath37 @xcite , @xmath38 @xcite , @xmath39 @xcite and @xmath40 @xcite have been studied through @xmath41 decays .
the decay @xmath35 decays to @xmath38 were studied with @xmath42 reconstructed form @xmath43 or @xmath44 , @xmath45 from @xmath44 , and @xmath46 from @xmath47 .
the results are shown in table [ tab : chicjvvres ] .
the decays of @xmath48 $ ] , which were suppressed in helicity sum rule ( hsr ) @xcite , and the doubly ozi suppressed decay @xmath49 are observed for the first time .
evidence for the @xmath50 decay is found with a signal significance of @xmath51 .
the branching fractions for @xmath52}\to\phi\phi[\omega\omega]$ ] decays are remeasured with improved precision .
these precise measurements will be helpful for understanding @xmath35 decay mechanisms .
in particular , the measured branching fractions for @xmath53 indicate that hsr is significantly violated and that long distance effects play an important role in this energy region .
details about this study can be found in ref .
@xcite .
.summary of the branching fractions ( @xmath54 ) for @xmath55 measured at besiii @xcite . also listed are the world average in pdg @xcite .
the upper limit is estimated at the 90% c.l.[tab : chicjvvres ] [ cols="<,<,<",options="header " , ]
we report the recent progress on the charmonium decays studied at besiii , including @xmath1 radiative decays to a pseudoscalar meson , @xmath35 hadronic decays , @xmath35 radiative decays to a vector meson , and @xmath56 decays to vector meson pairs .
in addition , the search effort for new physics through @xmath2 violation process , pseudoscalar meson decay to @xmath57 , is presented .
this work is supported in part by the national natural science foundation of china ( 10905091 ) , srf for rocs of sem , and china postdoctoral science foundation .
m. ablikim _ et al . _
[ besiii collaboration ] , nucl .
instrum .
a * 614 * , 345 ( 2010 ) [ arxiv:0911.4960 [ physics.ins-det ] ] .
y. kubota _ et al . _
[ cleo collaboration ] , nucl .
instrum .
* a320 * , 66 - 113 ( 1992 ) . s. j. brodsky and g. p. lepage , phys .
d * 24 * , 2848 ( 1981 ) .
k. nakamura _ et al .
_ [ particle data group ] , j. phys .
g * 37 * , 075021 ( 2010 ) . m. ablikim _ et al . _
[ besiii collaboration ] , phys .
rev . * d83 * , 112005 ( 2011 ) .
[ arxiv:1103.5564 [ hep - ex ] ] . | in 2009 , the besiii experiment has collected about 225 m @xmath0 and 106 m @xmath1 samples , both of which are the world largest on - peak charmonium production .
based on these dataset , besiii has made great effort on the study of the charmonium decays , some important of which have been reviewed in this proceeding .
in addition , a searching for new physics through the @xmath2 violation process is reported . |
type ia supernovae ( sne ia ) are one of the most powerful distance indicators for distant galaxies in the cosmic expansion dominated field ( @xmath16 mpc ) .
three advantages of sne ia - a bright peak luminosity , a small luminosity dispersion , and a homogeneity in the universe - have enabled us to investigate the cosmic expansion rate ( @xmath17 ) @xcite and the cosmic acceleration rate ( @xmath18 ) @xcite . although great progress has been made over the last two decades , more accurate calibration of peak luminosity of sne ia is needed to derive more accurate values of cosmological parameters .
recently , two large projects , supernovae and @xmath17 for the equation of state ( sh0es , @xcite ) and carnegie hubble program ( chp , @xcite ) , have been working on the calibration of sne ia to improve the measurement of cosmological parameters .
they performed the luminosity calibration of sne ia based on the cepheid distance estimates of 8 sne ia host galaxies and derived similar values of the hubble constant : @xmath19 km s@xmath13 mpc@xmath13@xcite and @xmath20 km s@xmath13 mpc@xmath13 @xcite .
however , these values of the hubble constant are somewhat larger than those based on recent analysis of the cosmic microwave background radiation ( cmbr ) .
@xcite derived the cosmological parameters from the analysis of the angular power spectrum of cmbr in wmap9 assuming a flat @xmath18cdm cosmology , and presented @xmath21 km s@xmath13 mpc@xmath13 .
later , the planck collaboration lead to an even smaller value of the hubble constant , @xmath22 km s@xmath13 mpc@xmath13 , using new cmbr data from the planck satellite @xcite .
thus the traditional distance ladder method and the cmbr analysis based on @xmath18cdm cosmology show a 23@xmath23 level of difference in the estimates of the hubble constant @xcite .
this is called `` the hubble tension '' , one of the most critical issues in modern cosmology .
in addition , recent studies of the baryon acoustic oscillation ( bao ) in the range of redshift combined with sne ia data presented similar values to those from the cmbr analysis @xcite .
we have been working on the luminosity calibration of sn ia by measuring the precise distances to sn ia host galaxies based on the tip of the red giant branch ( trgb ) @xcite .
the trgb method uses old red giant branch ( rgb ) stars , classified as population ii .
the metallicity dependence of @xmath1-band luminosities of trgb stars is known to be small @xcite . moreover
, the trgb method often uses rgb stars located in the halo , where the internal extinction is expected to be negligible and the crowing of stars is much less than the disk or the bulge .
@xcite ( paper i ) measured the trgb distance to m101 hosting sn 2011fe , which is the nearest sn ia with modern ccd photometry .
@xcite ( paper ii ) determined the trgb distances to m66 and m96 in the leo i group , hosting sn 1989b and sn 1998bu . by combining the trgb distance estimates of three sne ia from papers
i and ii , and optical light curves for three sne ia , we derived the hubble constant , @xmath24(random)@xmath25(systematic ) km s@xmath13 mpc@xmath13 @xcite , which is slightly smaller than values from the cepheid calibrated sne ia @xcite . in this paper , we present the trgb distances to two additional sn ia host galaxies , ngc 4038/39 hosting sn 2007sr and ngc 5584 hosting sn 2007af .
ngc 4038/39 are a well known pair of galaxies showing early stage merger of disk galaxies . in 2007
, a supernova was discovered near the southern tidal tail of this galaxy , sn 2007sr @xcite .
spectroscopic observations confirmed that it is a sn ia @xcite .
photometric follow - up observations were made from optical to near - infrared @xcite .
light curve fits on this sn suggested that its internal reddening is expected to be small , @xmath26 @xcite .
there are three distance estimates to ngc 4038/39 based on the trgb in the literature .
@xcite analysed hubble space telescope @xmath27/wide field planetary camera 2 @xmath28 images of a field on the southern tidal tail and derived a trgb distance , @xmath29 ( @xmath30 mpc ) .
this value is significantly smaller than previous estimates based on radial velocities of ngc 4038/39 ( @xmath31 mpc ( @xmath32 @xmath33 @xmath34 , assuming @xmath35 km s@xmath13 mpc@xmath13 ) ) .
later , @xcite presented a similar distance estimate , @xmath36 ( @xmath37 mpc ) , using the new @xmath38/advanced camera for surveys ( acs ) image data @xcite .
however , @xcite presented a significantly larger value , @xmath39 ( @xmath40 mpc ) , based on the same acs image data used in @xcite . on the other hand
, @xcite detected cepheid variables in the main bodies of ngc 4038/39 and derived a distance modulus , @xmath41 ( @xmath42 mpc ) .
this is similar to the value in @xcite , but much larger than the value given by @xcite .
ngc 5584 is a moderately inclined ( @xmath43 ) spiral galaxy hosting sn 2007af , which is one of the most well monitored sne ia . before its pre - maximum , spectroscopic and photometric observations from optical to near infrared
have been made @xcite .
moreover the internal reddening for sn 2007af is known to be small , @xmath44 @xcite .
thus , sn 2007af is an ideal target for the luminosity calibration of sne ia .
previous distance estimates based on the tully - fisher relation show a wide range of values with large uncertainty : @xmath45 with uncertainties of @xmath46 @xcite .
@xcite detected 94 cepheid variables based on the optical / near - infrared image data taken with @xmath38/wfc3 , and derived a distance , @xmath47 ( @xmath48 mpc ) .
this value is much larger than those based on the tully - fisher relations . to date
, there is no trgb distance estimate for ngc 5584 .
this paper is composed as follows .
section 2 describes data reduction .
section 3 presents color - magnitude diagrams and distance estimates for each galaxy . implications of our results are discussed in section 4 .
lcc r.a.(2000 ) & @xmath49 & @xmath50 + dec(2000 ) & @xmath51 & @xmath52 + prop i d . &
10580 & 11570 + detector & acs / wfc & wfc3/uvis + exposure , v & 10,870s , @xmath53 & 45,540s , @xmath54 + exposure , i & 8,136s , @xmath55 & 14,400s , @xmath55 + [ observation ] lcccc r.a.(2000 ) & @xmath56 & @xmath57 & @xmath58 & @xmath59 + dec(2000 ) & @xmath60 & @xmath61 & @xmath62 & @xmath63 + prop i d .
& 9666 & 11903 & 10048 , 11677 & 11903 , 11452 + exposure , @xmath54 & 720s & 1,160s & 2,139s & 1,160s + exposure , f606w & 676s & 800s & 163,676s & 18,376s + exposure , @xmath55 & 676s & 1,300s & 184,242s & 3,400s + [ observation_gc ] the image data for ngc 4038/39 and ngc 5584 were acquired from the @xmath38 archive .
table [ observation ] lists the information for the @xmath38 image data used in this study .
the image data for ngc 4038/39 covering the southern tidal tail , as shown in figure [ fig_finder ] ( a ) , were taken with the acs in wide field channel ( wfc ) mode ( proposal i d : 10580 ) with exposure times of @xmath64 @xmath65s for @xmath53 and @xmath66 @xmath65s for @xmath55 .
corresponding total exposure times were 10870s and 8136s for @xmath53 and @xmath55 , respectively . charge transfer efficiency ( cte ) corrected and flat fielded images ( indicated by suffix _
flc.fits ) were combined to make a single drizzled image using the drizzlepac . because individual _
flc images were severely contaminated by cosmic rays , the automatic image aligning tool , tweakreg task , failed to provide fine solutions .
thus , we used an alternative strategy , as described in the following .
first , we performed an aperture photometry on the _ flc images with aperture radii of @xmath67 and @xmath68 using daophot in the iraf package .
the magnitude difference between these two aperture radii is defined as the concentration index , @xmath69 .
next , sources with @xmath70 were selected . through this process
, sources with abnormally narrow ( e.g. , cosmic rays ) or abnormally broad ( e.g. , background galaxies ) radial profiles were rejected .
second , we converted image coordinates for the selected sources to the world coordinate system ( wcs ) based on the information listed in the header .
the wcs coordinates for a pair of _ flc images were matched with a matching criteria of 005 ( 1 pixel ) .
this process rejects most non - stellar objects such as cosmic rays and hot pixels .
finally , the files to be used with the tweakreg task were created .
these files contain source coordinates and flux counts . using these files , the tweakreg task output better solutions of @xmath71 and @xmath72 with a rms value of @xmath73 pixel ( @xmath74 ) .
+ the aligned images were combined into a single drizzled images using the astrodrizzle task .
the final_pixfrac and final_scale values used are 1.0 and @xmath75pixel , respectively .
output images show fwhm of 2.64 pixels ( @xmath76 ) for @xmath53 and 2.38 pixels ( @xmath77 ) for @xmath55 .
observations for ngc 5584 were done with the wide field camera 3 ( wfc3 ) in ultraviolet - visible ( uvis ) mode with the primary aim of detecting cepheid variables ( proposal i d : 11570 ) .
centered on ngc 5584 , the @xmath78 field covering the main body and inner halo of the galaxy , as shown in figure [ fig_finder ] ( b ) , was observed with exposure times of @xmath79 @xmath80s for @xmath54 and @xmath81 @xmath80s for @xmath55 .
total exposure times are 45,540s and 14,400s for @xmath54 and @xmath55 , respectively .
flc images were made from the flat - fielded images ( indicated by suffix _
flt.fits ) and the raw images ( indicated by suffix _
raw.fits ) by applying pixel - based cte correction software . derived _
flc images were aligned in the same way as was done for ngc 4038/39 images .
mean rms values for @xmath71 and @xmath72 from the tweakreg task are @xmath82 pixels ( @xmath83 ) for both filters .
aligned _ flc images were then combined using the astrodrizzle task with a final_pixfrac value of 0.7 and a final_scale value of @xmath84pixel .
rms values of weight images for each filter are smaller than 4% of the median value .
fwhms for point sources in the final drizzled images are 3.00 pixels ( @xmath85 ) for @xmath54 and 3.22 pixels ( @xmath86 ) for @xmath55 .
figure [ fig_resolve ] shows @xmath87 sections of the output @xmath55 images of ngc 4038/39 ( a ) and of ngc 5584 ( b ) .
point sources and extended sources ( mostly background galaxy ) can be clearly seen in the images .
psf photometry was done on the drizzled images using the standard routine : daofind , phot , and allstar tasks in iraf / daophot @xcite .
initial aperture photometry with aperture radii of @xmath67 and @xmath68 were done to derive the concentration index for each stellar object . with this combination of aperture radii ,
point sources show mean concentration index values of @xmath88 for ngc 4038/39 and @xmath89 for ngc 5584 .
psf images were constructed for each galaxy based on @xmath90 bright , isolated , and @xmath91 or @xmath92 stellar objects .
source coordinates to be used in the photometry were determined from @xmath55-band images for each galaxy with a detection threshold of 2 .
we used revised photometric zeropoints for acs / wfc ( 26.416 for @xmath53 and 25.529 for @xmath55 ) and wfc3/uvis ( 25.81 for @xmath54 and 24.67 for @xmath55 ) , provided by stsci .. aperture correction values to the aperture radii of @xmath93 ( for acs / wfc ) or @xmath94 ( for wfc3/uvis ) were calculated by checking the curve of growth for bright isolated stars .
derived values are @xmath95 , @xmath96 for @xmath53 , @xmath55 on acs / wfc and @xmath97 , @xmath98 for @xmath54 , @xmath55 on wfc3/uvis with an uncertainty of 0.02 .
additional aperture corrections up to the infinite aperture radius were also applied .
corresponding values derived from the enclosed energy curves provided by stsci are : @xmath99 for @xmath53 , @xmath55 on acs / wfc and @xmath100 for @xmath54 , @xmath55 on wfc3/uvis , respectively .
finally , acs magnitudes were converted to standard johnson - cousins magnitudes using the photometric transformation equations listed in @xcite .
the photometric systems for acs / wfc and wfc3/uvis are not the same .
there is no trgb calibration for the wfc3/uvis system yet . to apply the trgb calibration , and to facilitate comparisons of the two galaxies in this study and those in our previous studies ( @xcite )
, we derived photometric transformations from the wfc3/uvis system to the acs / wfc system as follows .
we used archival image data for two milky way globular clusters , ngc 2419 and 47 tuc , which were observed using both acs / wfc and wfc3/uvis , as listed in table [ observation_gc ] and shown in figure [ fig_finder_gc ] .
these two globular clusters have been used for the photometric transformation from the acs / wfc system to the johnson - cousin system in @xcite , @xcite , and @xcite .
we calibrated raw image data to construct cte corrected drizzled images and carried out psf photometry as described in section 2 .
we then matched the photometric catalogues from acs / wfc and wfc3/uvis image data .
constructed color - magnitude diagrams ( cmds ) for resolved stars in ngc 2419 and 47 tuc are shown in figure [ fig_cmd_gc ] . to reduce the uncertainty due to the stars in the crowded field
, we used only the stars located in relatively low density environments at @xmath101 and @xmath102 , indicated by the hatched lines in figure [ fig_finder_gc ] .
cmds from acs / wfc data ( open squares ) and wfc3/uvis data ( filled squares ) look very similar .
stars with @xmath103 , wfc3/uvis data show in general redder colors than acs / wfc data . for blue stars with @xmath104
, however , wfc3/uvis data shows slightly bluer color than acs / wfc data .
this trend is seen in both globular clusters .
in order to compare the difference quantitatively , we selected bright and non - saturated stars in the shaded regions ( @xmath105 for ngc 2419 , and @xmath106 or @xmath107 for 47 tuc ) of figure [ fig_cmd_gc ] and compared them in figure [ fig_transformation ] . figure [ fig_transformation ] shows comparisons of magnitudes between the two photometric systems in terms of their colors .
@xmath108 colors range from 0.0 to 2.2 , and their mean offsets are @xmath109 for @xmath54 and @xmath110 for @xmath55 .
we plot isochrones , as a reference , for the 12 gyr age and @xmath111=-1.4 $ ] , which is a mean metallicity of ngc 2419 and 47 tuc , provided by dartmouth group @xcite .
data points and stellar isochrones are in good agreement , mean offsets are estimated to be @xmath112 0.01 in both filters .
we derive a photometric transformation as follows : + @xmath113 = @xmath114 , + and + @xmath115 = @xmath116 + where @xmath117 .
+ note that the amount of correction is very small : 0.05 mag in @xmath54 and smaller than 0.01 mag in @xmath55 for @xmath118 .
we applied this photometric transformation to the photometry of ngc 5584 .
transformation uncertainties are estimated to be @xmath110 mag considering uncertainties of aperture corrections , although the fitting uncertainties are much smaller than 0.01 mag .
we derived the same photometric transformation for @xmath53-band for the future studies .
+ @xmath119 = @xmath120 , + and + @xmath115 = @xmath121 + where @xmath122 .
we checked the photometric completeness and systematic errors of our photometry through the artificial star experiment . to do this , we selected test fields with a size of @xmath123 from the hatched regions in figure [ fig_finder ] . next , we added about @xmath124 and @xmath125 artificial stars into the test fields of ngc 4038/39 and ngc 5584 , which correspond to @xmath126% of the total amount of stars detected in each field .
these added stars were uniformly distributed throughout the fields and have a magnitude range of @xmath127 .
we set @xmath128 for input artificial stars in order to make them similar to the estimated median @xmath129 color at the trgb magnitude ( see section 3.2 ) .
then , we preformed psf photometry on the fields in the same way as done on the original images .
we then checked recovery rates and photometric systematics .
we iterated this process 600 and 300 times for ngc 4038/39 and ngc 5584 , respectively , to keep a larger number of artificial stars and to reduce statistical uncertainties . in total , about @xmath130 artificial stars were used for each field in the test .
figures [ fig_comp ] ( a ) and ( d ) show the recovery rates for artificial stars as a function of input @xmath1-band magnitudes for these two galaxies .
about 90% and 70% of stars are expected to be recovered at the trgb magnitude of ngc 4038/39 ( @xmath131 ) and ngc 5584 ( @xmath132 ) .
the 50% recovery rates are estimated to be @xmath133 for ngc 4038/39 and @xmath134 for ngc 5584 .
it is noted that , although exposure times of ngc 4038/39 are much shorter than those of ngc 5584 , the recovery rates are higher in the same magnitude range . this can be explained as follows .
first , the field used in the analysis of ngc 4038/39 is much farther from the center of the host galaxy than that of ngc 5584 , so that the expected stellar density is much lower .
second , ngc 4038/39 was observed with acs / wfc , which has higher sensitivity than the wfc3/uvis at longer wavelengths of @xmath112 400 nm .
third , ngc 4038/39 was observed with @xmath53-band , which has a much wider spectral window than the @xmath54-band used for ngc 5584 .
figures [ fig_comp ] ( b ) , ( c ) , ( e ) and ( f ) show random uncertainties and systematic offsets of @xmath1-band magnitudes ( @xmath135 ) and @xmath136 colors ( @xmath137 ) for the two galaxies .
random uncertainties increase gradually , being proportional to the input @xmath1-band magnitudes .
systematic offsets are noted . for @xmath138 in ( b ) and ( e ) , @xmath139 values are positive .
this means that those faint stars are measured slightly brighter than their intrinsic brightness . at the trgb magnitude of ngc 4038/39 and ngc 5584 , such magnitude and color offsets are smaller than 0.02 mag ,
so we ignored these offsets .
the @xmath38 fields used in this study cover the tidal tail in ngc 4038/39 , and the disk and bulge in ngc 5584 . therefore the color - magnitude diagrams of these fields show various populations of stars , such as young main sequence , blue and red super giant , asymptotic giant branch ( agb ) and rgb stars .
these mixed populations increase uncertainties in determining the trgb . in order to reduce the contamination due to multiple stellar populations and to sample the rgb population as much as possible , we selected regions with the lowest sky background level , avoiding spiral arms and tidal tails .
chosen regions are marked by the hatched regions in figure [ fig_finder ] .
figure [ fig_cmd ] show the cmds of resolved stars in the selected regions of ngc 4038/39 ( a ) , and ngc 5584 ( b ) .
we calculated the expected number of galactic foreground stars in each cmd using the trilegal program provided by the padova group @xcite .
derived numbers of stars are smaller than ten for both figures , indicating that a contamination by the foreground stars is negligible .
we also checked the background galaxy contamination by comparing our cmd with the cmd of the hubble extreme deep field ( hxdf ) , where background galaxies are dominated .
we used @xmath53 and @xmath55 combined images provided by @xcite .
total exposure times are 174,400s for @xmath53 , and 50,800s for @xmath55 .
we then carried out the psf photometry in the same method as for ngc 4038/39 and ngc 5584 .
constructed cmd for the point sources in the entire field of hxdf is shown in figure [ fig_cmd ] ( c ) .
most of sources are expected to be unresolved background galaxies .
we then count the number of sources inside the slanted boxes of each panel , which will be used in the trgb determination ( see next section ) : 5247 sources for ngc 4038/39 , 4491 sources for ngc 5584 , and 318 sources for hxdf .
after the correction for the field area differences ( 4.9 arcmin@xmath140 for ngc 4038/39 field , 1.0 arcmin@xmath140 for ngc 5584 field , and 10.8 arcmin@xmath140 for hxdf ) , we conclude that the background galaxy contaminations in the cmds are very small : @xmath141% for ngc 4038/39 and @xmath142% for ngc 5584 .
we ignored this contamination in the following analysis .
the cmds for ngc 4038/39 and ngc 5584 show clear rgb populations as well as relatively weak main sequence and agb populations .
both cmds show a clear number density enrichment of rgb stars at @xmath143 for ngc 4038/39 , and @xmath144 for ngc 5584 , which correspond to the trgb .
our photometry of ngc 4038/39 and ngc 5584 are deep enough to reach more than one magnitude below the trgb .
this enables us to measure distances with the trgb , a known precise distance indicator @xcite .
we derived @xmath1-band luminosity functions for resolved stars inside the slanted boxes of figure [ fig_cmd ] , designed to preferentially select the blue rgb stars .
+ the use of the blue rgb stars in the analysis has several advantages , as noted in @xcite .
first , the trgb for the bluer rgb stars is less curved than that of the redder rgb stars in the @xmath145 cmd .
second , the bluer rgb stars have a relatively higher photometric recovery rate than the redder rgb stars .
third , the bluer rgb stars are relatively less affected by the host galaxy reddening .
figure [ fig_trgb ] shows derived @xmath1-band luminosity functions and corresponding edge - detection responses .
we used the logarithmic form of edge - detection algorithm @xcite , which is described by @xmath146-log[\phi(i-\sigma_i ) ] ] \ ] ] , where @xmath147 is gaussian - smoothed luminosity function and @xmath148 is the mean photometric error .
edge - detection response functions show strong peaks at @xmath143 for ngc 4038/39 and @xmath144 for ngc 5584 , as well as relatively weak peaks at between @xmath149 and @xmath150 . in a galaxy that contain various stellar populations , the number of agb stars is much smaller than that of the rgb stars .
therefore , it is expected that the luminosity function of the red giant stars shows a weak jump at the agb tip , and a much stronger jump at the rgb tip .
we determined the prominent peaks to be the trgb of each galaxy .
the relatively weaker peaks in the brighter part of the trgb are thought to be the tips from intermediate agb populations or noise .
we derive our best estimates of the trgb magnitudes and their measurement errors from ten thousand simulations of bootstrap resampling .
derived values are @xmath151 for ngc 4038/39 and @xmath152 for ngc 5584 .
+ lcc trgb magnitude , @xmath153 & @xmath154 & @xmath155 + median trgb color , @xmath156 & @xmath157 & @xmath158 + galactic extinction , @xmath159 , @xmath160 & 0.06 , 0.05 & 0.06 , 0.05 + intrinsic trgb magnitude , @xmath161 & @xmath162 & @xmath163 + intrinsic trgb color , @xmath164 & @xmath165 & @xmath166 + absolute trgb magnitude , @xmath167 & @xmath168 & @xmath169 + distance modulus , @xmath170 & @xmath171 & @xmath172 + distance , d & @xmath173 mpc & @xmath174 mpc + [ distance ] we adopted the trgb calibration suggested by @xcite , @xmath175 $ ] , which was also used in our previous studies @xcite .
median trgb colors estimated from @xmath176 rgb stars in the @xmath177 magnitude range of the trgb are @xmath157 for ngc 4038/39 and @xmath158 for ngc 5584 .
we adopted foreground extinction values , @xmath178 for ngc 4038/39 and @xmath179 for ngc 5584 from @xcite .
host galaxy extinctions for the blue rgb stars are expected to be very small , so they are ignored . by combining the trgb magnitudes , extinction values , and the trgb calibration , we obtained values of the distance moduli , @xmath4 ( random ) @xmath5 ( systematic ) for ngc 4038/39 , and @xmath180 ( random ) @xmath5 ( systematic ) for ngc 5584 , as summarized in table [ distance ] .
we checked the effect of trgb color selection to the trgb distance estimates .
we divided the rgbs into four ( for ngc 4038/39 ) and three ( for ngc 5584 ) subregions with different @xmath136 ranges .
the trgb colors of these subregions cover @xmath181 for ngc 4038/39 and @xmath182 for ngc 5584 , in steps of 0.3 mag .
we then derived the trgb distances for each subregion by applying the same method as done for the blue rgb stars .
the trgb distances for four subregions of ngc 4038/39 show similar values : @xmath183 to 31.77 with uncertainties of 0.04 to 0.09 .
a weighted mean of these four estimates is @xmath184 , which is in good agreement with the estimate from the blue rgb stars , @xmath4 .
the three subregions of ngc 5584 also show similar values : @xmath185 to 31.78 with uncertainties of 0.04 to 0.06 .
a weighted mean of these estimates is @xmath186 , showing an excellent agreement with the estimate from the blue rgb stars , @xmath187 .
therefore , we conclude that the trgb distance estimation depends little on the color selection of rgb stars .
tables [ distance_4038 ] and [ distance_5584 ] summarize the distance estimates for each galaxy in this study and in previous studies
. distance estimates for ngc 4038/39 show a large range : @xmath188 to @xmath189 .
@xcite presented a distance estimate of @xmath190 from the photometry of 35 cepheid variables .
they derived periods of the cepheid variables from the image data taken with the @xmath38/wfpc2 and apparent magnitudes from the @xmath191 image data taken with the @xmath38/wfc3 .
later , @xcite applied their theoretical model to the observational data in @xcite , and derived a similar value , @xmath192 .
@xcite reported relatively short distance estimates based on the trgb method : @xmath193 from the wfpc2 @xcite , and @xmath194 from the acs image data @xcite .
however , @xcite reported a much larger value of the trgb distance , @xmath195 from the same acs image data as used in @xcite .
this @xmath196 mag difference comes from the different trgb magnitudes : @xmath197 in @xcite and @xmath198 in @xcite . @xcite
derived even larger values for the distance moduli : @xmath199 from the peak luminosity of sn ia , sn 2007sr , and @xmath200 from the systematic recession velocity relative to the local group , applying the large scale flow model by @xcite . by considering the distance moduli from three independent methods , @xcite suggested to use a conservative value , @xmath201 mpc ( @xmath199 ) as the distance to ngc 4038/39 .
llccl 1 & @xcite & sn ia & @xmath202 & salt2 , @xmath203 km s@xmath13 mpc@xmath13 + 2 & @xcite & sn ia & @xmath204 & @xmath205 km s@xmath13 mpc@xmath13 + 3 & @xcite & recession velocity & @xmath206 & @xmath205 km s@xmath13 mpc@xmath13 + 4 & @xcite & cepheid & @xmath207 & @xmath208 cepheids = 35 + 5 & @xcite & cepheid & @xmath209 & @xmath208 cepheids = 32 + 6 & @xcite & trgb & @xmath210 & @xmath211 + 7 & @xcite & trgb & @xmath212 & @xmath197 + 8 & @xcite & trgb & @xmath213 & @xmath198 + 9 & this study & trgb & @xmath214 & @xmath2 + [ distance_4038 ] llccl 1 & @xcite & tf & @xmath215 & mean of @xmath216-band + 2 & @xcite & tf & @xmath217 & + 3 & & tf & @xmath218 & + 4 & @xcite & tf & @xmath219 & @xmath1-band + 5 & @xcite & tf & @xmath220 & @xmath221 km s@xmath13 mpc@xmath13 + 6 & & tf & @xmath222 & @xmath223 km s@xmath13 mpc@xmath13 + 7 & @xcite & tf & @xmath224 & @xmath225 m , @xmath226 + 8 & @xcite & tf & @xmath227 & @xmath225 m , @xmath226 + 9 & @xcite & sn ia & @xmath228 & mlcs2k2 , @xmath229 , @xmath223 km s@xmath13 mpc@xmath13 + 10 & & sn ia & @xmath230 & mlcs2k2 , @xmath231 , @xmath223 km s@xmath13 mpc@xmath13 + 11 & @xcite & sn ia & @xmath232 & snoopy , @xmath223 km s@xmath13 mpc@xmath13 + 12 & @xcite & sn ia & @xmath233 & salt2 , @xmath203 km s@xmath13 mpc@xmath13 + 13 & @xcite & cepheid & @xmath234 & @xmath208 cepheids = 94 + 14 & @xcite & cepheid & @xmath235 & @xmath208 cepheids = 78 + 15 & this study & trgb & @xmath236 & @xmath3 + [ distance_5584 ] our best estimate of the trgb distance to ngc 4038/39 is @xmath4 .
it is roughly one mag larger than that found in @xcite .
our estimate is not much different , within errors , from the trgb distance given by @xcite ( @xmath195 ) .
we investigate any cause for the small difference of 0.16 mag between our best estimate and that of @xcite .
there are three differences in the methods between the two studies .
first , @xcite used relatively redder rgb stars ( @xmath237 ) than those used in this study .
second they used the entire acs / wfc field , including tidal tail regions , while we used only the region with low sky values , avoiding tidal tail regions .
third , @xcite applied t magnitude ( = @xmath238 $ ] ) , while we used i magnitude .
using the same t magnitude as used in @xcite , we derived the trgb magnitude of @xmath239 , which is almost the same as the value given by @xcite ( @xmath198 ) .
when we used the stars in the lowest sky background regions , avoiding tidal tail regions , as shown in figure [ fig_finder ] ( a ) , we obtained a 0.05 mag fainter value , @xmath240 .
when we used the blue rgb stars ( @xmath241 , inside the slanted box of figure [ fig_cmd ] ) in the lowest sky background regions , we derive another 0.05 mag fainter value , @xmath242 .
the corresponding distance modulus is @xmath243 , which is in agreement with our best estimate , @xmath4 , within an uncertainty .
thus , within the 0.16 magnitude difference between our best estimate and that of @xcite , 0.06 mag comes from field selection , 0.05 mag comes from the trgb color selection , and 0.05 mag comes from the measurement error . our estimate is consistent with the cepheid estimate by @xcite ( @xmath190 ) .
distance estimates for ngc 5584 in the previous studies also show a large range from @xmath244 to @xmath245 .
distance estimates using the tully fisher relation give a large range ( @xmath246 ) with large uncertainties @xcite .
sn ia distance estimates show a much larger range from @xmath247 to @xmath245 , depending on the selection of the light curve fitting tool , the total to selective extinction ratio ( @xmath248 ) , and the value of the hubble constant .
@xcite found 94 cepheid variables from the @xmath38/wfc3 image data and derived a distance modulus of @xmath249 .
distance modulus derived in this study , the first distance estimate using the trgb method , is @xmath180 .
this is in good agreement with the cepheid estimate by @xcite , considering measurement errors . to date , ngc 5584 is the most distant galaxy among the galaxies to which distances were measured using the trgb method .
llcccccc ngc 4038/39 & sn 2007sr & ubvri & @xmath250 & @xmath251 & @xmath252 & @xmath214 & @xmath253 + ngc 5584 & sn 2007af & ubvri & @xmath254 & @xmath255 & @xmath256 & @xmath236 & @xmath257 + m101 ( ngc 5457 ) & sn 2011fe & ubvri & @xmath258 & @xmath259 & @xmath260 & @xmath261 & @xmath262 + m66 ( ngc 3627 ) & sn 1989b & ubvri & @xmath263 & @xmath264 & @xmath265 & @xmath266 & @xmath267 + m96 ( ngc 3368 ) & sn 1998bu & ubvri & @xmath268 & @xmath269 & @xmath270 & @xmath271 & @xmath272 + & @xmath273 + & @xmath274 + [ luminosity ] the trgb distance estimates of sne ia host galaxies in this study and our previous studies are very useful in determining the peak luminosity of sne ia and the hubble constant .
m101 ( ngc 5457 ) is a nearby face - on spiral galaxy hosting one sn ia , sn 2011fe , the nearest low - reddened sne ia with modern photometry .
it was discovered just 4 hours after its explosion , so complete monitoring observations have been made .
thus , it is an ideal target for the calibration of sne ia .
@xcite investigated nine @xmath38/acs fields around the inner region of m101 and derived a trgb distance of @xmath275 ( random)@xmath5 ( systematic ) .
m66 ( ngc 3627 ) and m96 ( ngc 3368 ) are nearby spiral galaxies in the leo i group hosting sne ia , sn 1989b and sn 1998bu , respectively . the host galaxy extinction values of these two sne ia are known to be relatively high ( @xmath276 ) .
however , since they are nearby and their optical light curves from before their maximum magnitudes are available , they are still useful for the calibration of sne ia .
@xcite presented new trgb distances to m66 and m96 based on the deep @xmath38/acs images : @xmath277(random)@xmath5(systematic ) for m66 and @xmath278 ( random)@xmath5 ( systematic ) for m96 .
we included the trgb distance estimates for these three galaxies in the luminosity calibration of sne ia as well as the two galaxies in this study , ngc 4038/39 and ngc 5584 .
we compiled optical light curves of five sne ia in the following literature : ubvri photometric data for sn 2007sr and sn 2007af from the carnegie supernova project ( @xmath279-band ) and @xcite ( bvri bands ) , ubvri photometric data for sn 2011fe , sn 1989b and sn 1998bu from @xcite , @xcite , and @xcite , respectively .
we added three @xmath280-band data points before the pre - maximum of sn 2007sr , given by the all - sky automated survey ( asas ) and @xcite .
all the photometric data for these five sne ia were done with ccd detectors , not photoelectric detectors .
then we ran mlcs2k2 ( version 0.07 ) @xcite to derive their light curve parameters .
we set the total to selective extinction ratio , @xmath231 for the milky way and @xmath281 for sn ia host galaxies , to be consistent with @xcite .
we added 0.08 mag as a light curve fit error , and 10% of the milky way and host galaxy extinction values as extinction errors in the error propagation .
we also included the intrinsic luminosity dispersion of sne ia of 0.14 mag @xcite in the error calculation .
-band host galaxy extinction values of sne ia and their trgb distances . note that the extinction values for sn 1989b and sn 1998bu are much higher than other three sne ia .
] the values of the parameters derived from the light curve fits are summarized in table [ luminosity ] .
our light curve fit results are consistent with those in @xcite and @xcite , within fitting uncertainty .
figure [ fig_5sne ] shows a comparison of trgb distance estimates , @xmath280-band absolute peak magnitudes , and host galaxy extinction values of the five sne ia .
two features are noted .
first , the host galaxy extinction values for sn 1989b and sn 1998bu are much larger than those of the other three sne ia .
thus the derived light curve parameters from these two sne may be less reliable , although we corrected for their extinctions .
second , sn 2007af is as bright as those in our previous studies , while sn 2007af is somewhat fainter . the standard deviation for five sne ia is 0.21 mag , slightly larger than the luminosity dispersion of the distant sne ia , 0.14 mag @xcite .
this may be due to small number statistics .
further studies adding more sne ia in the luminosity calibration is needed .
a weighted mean value of the @xmath280-band absolute peak magnitudes of five sne is estimated to be @xmath9 .
it is @xmath282 mag brighter than that of the sh0es project ( @xmath283 , @xcite ) and 0.2 mag fainter than those of the hubble key project ( @xmath284 , @xcite ) and the @xmath38 key project ( @xmath284 , @xcite ) . if we adopt 3 low - reddened sne ( sn 2007sr , sn 2007af , and sn 2011fe ) , then the absolute peak magnitude would be slightly fainter , @xmath10 , getting closer to the estimate given by @xcite .
@xcite presented the relation between the absolute magnitude of sne ia and the hubble constant based on the mlcs2k2 fits of 140 sne ia at @xmath285 : @xmath286 using the absolute magnitudes derived in this study to the above equation , we derived a hubble constant : @xmath287 ( random)@xmath288 ( systematic ) km s@xmath13 mpc@xmath13 from the five sne , and @xmath289 ( random)@xmath290 ( systematic ) km s@xmath13 mpc@xmath13 from the three low - reddened sne . our estimate based on the five sne is similar to the values from the cmbr analysis , @xmath291 km s@xmath13 mpc@xmath13 @xcite and not much different , within the errors , from those of recent cepheid calibrations of sne ia , @xmath292 km s@xmath13 mpc@xmath13 @xcite .
our estimate based on the three low - reddened sne is between the values from the two independent methods .
we present the trgb distance estimates for sn ia host galaxies ngc 4038/39 and ngc 5584 from the psf photometry of archival @xmath38 images .
a summary of the main results in this study is as follows .
* we detected rgb stars in the outer region of ngc 4038/39 and ngc 5584 based on the combined archival acs / wfc and wfc3/uvis image data .
* we derived photometric transformation between @xmath54 and @xmath55 bands of wfc3/uvis and that of acs / wfc , by comparing the photometry of the resolved stars in two milky way globular clusters , ngc 2419 and 47 tuc .
* we found @xmath1-band trgb magnitudes to be at @xmath131 for ngc 4038/39 and @xmath293 for ngc 5584 . from these
, we derived the trgb distance estimates of @xmath294 for ngc 4038/39 and @xmath295 for ngc 5584 . * based on distance estimates for the two sne derived in this study and that of the three sne in our previous studies , we calibrate the @xmath280-band peak absolute magnitude of sne ia . * from the mean absolute magnitude of five sne ia , the hubble constant is determined to be @xmath296 km s@xmath13 mpc@xmath13 , which is similar to the values from the cmbr analysis , @xmath291 km s@xmath13 mpc@xmath13 @xcite .
our estimate is slightly smaller , but agrees within errors , with the values from the recent calibrations of sne ia with the cepheid variables , @xmath297 km s@xmath13 mpc@xmath13 @xcite .
if we adopt 3 low - reddened sne , corresponding the hubble constant is slightly increased , @xmath289 ( random)@xmath290 ( systematic ) km s@xmath13 mpc@xmath13 .
this work was supported by the national research foundation of korea ( nrf ) grant funded by the korea government ( msip ) ( no .
2012r1a4a1028713 ) .
this paper is based on image data obtained from the multimission archive at the space telescope science institute ( mast ) .
thanks to brian s. cho for his comments and suggestions for the manuscript .
amanullah , r. , lidman , c. , rubin , d. , et al .
2010 , , 716 , 712 aubourg , . ,
bailey , s. , bautista , j. e. , et al .
2014 , arxiv:1411.1074 bellazzini , m. , ferraro , f. r. , & pancino , e. 2001 , , 556 , 635 bennett , c. l. , larson , d. , weiland , j. l. , et al .
2013 , , 208 , 20 bennett , c. l. , larson , d. , weiland , j. l. , & hinshaw , g. 2014 , , 794 , 135 fiorentino , g. , musella , i. , & marconi , m. 2013 , , 434 , 2866 freedman , w. l. , madore , b. f. , gibson , b. k. , et al .
2001 , , 553 , 47 freedman , w. l. , & madore , b. f. 2010 , , 48 , 673 freedman , w. l. , madore , b. f. , scowcroft , v. , et al .
2012 , , 758 , 24 girardi , l. , barbieri , m. , groenewegen , m. a. t. , et al . 2012 , red giants as probes of the structure and evolution of the milky way , 165 pereira , r. , thomas , r. c. , aldering , g. , et al .
2013 , , 554 , aa27 perlmutter , s. , aldering , g. , goldhaber , g. , et al . 1999 , , 517 , 565 phillips , m. m. , simon , j. d. , morrell , n. , et al .
2013 , , 779 , 38 planck collaboration , ade , p. a. r. , aghanim , n. , et al .
2014 , , 571 , aa16 planck collaboration , ade , p. a. r. , aghanim , n. , et al .
2015 , arxiv:1502.01589 salaris , m. , & cassisi , s. 1997 , , 289 , 406 sandage , a. , tammann , g. a. , saha , a. , et al .
2006 , , 653 , 843 saviane , i. , hibbard , j. e. , & rich , r. m. 2004 , , 127 , 660 saviane , i. , momany , y. , da costa , g. s. , rich , r. m. , & hibbard , j. e. 2008 , , 678 , 179 schlafly , e. f. , & finkbeiner , d. p. 2011
, , 737 , 103 schweizer , f. , burns , c. r. , madore , b. f. , et al .
2008 , , 136 , 1482 simon , j. d. , gal - yam , a. , penprase , b. e. , et al .
2007 , , 671 , l25 sirianni , m. , jee , m. j. , bentez , n. , et al .
2005 , , 117 , 1049 springob , c. m. , masters , k. l. , haynes , m. p. , giovanelli , r. , & marinoni , c. 2009 , , 182 , 474 sorce , j. g. , tully , r. b. , & courtois , h. m. 2012 , , 758 , l12 sorce , j. g. , tully , r. b. , courtois , h. m. , et al .
2014 , , 444 , 527 | we present the tip of the red giant branch ( trgb ) distances to type ia supernova ( sne ia ) host galaxies ngc 4038/39 and ngc 5584 . based on the deep images constructed using archival hubble space telescope data , we detect red giant branch stars in each galaxy .
@xmath0 photometry of the resolved stars and corresponding @xmath1-band luminosity functions show the trgb to be at @xmath2 for ngc 4038/39 and @xmath3 for ngc 5584 . from these estimates , we determine the distance modulus to ngc 4038/39 to be @xmath4 ( random)@xmath5 ( systematic ) ( corresponding to a linear distance of @xmath6 mpc ) and the distance modulus to ngc 5584 to be @xmath7(random)@xmath5 ( systematic ) ( corresponding to a linear distance of @xmath8 mpc ) . we derive a mean absolute maximum magnitude of sne ia of @xmath9 from the distance estimates of five sne ia ( including two sne in this study and three sne ia from our previous studies ) , and we derive a value of @xmath10 using three low - reddened sne ia among the five sne ia . with these estimates ,
we derive a value of the hubble constant , @xmath11(random)@xmath12 ( systematic ) km s@xmath13 mpc@xmath13 and @xmath14(random)@xmath15 ( systematic ) km s@xmath13 mpc@xmath13 , respectively .
the value from the five sne is similar to those from the cosmic microwave background analysis , and not much different within errors , from those of recent cepheid calibrations of sne ia .
the value from the three sne is between the values from the two methods . |
it is currently commonly accepted that structure in the universe is formed through hierarchical processes , in which smaller structures collapse to form galaxies , which then group together or coalesce to form larger galaxies and galaxy groups and clusters . within this general background , anti - hierarchical "
behaviour has been observed , in the sense of the most dense large - galaxy - size structures collapsing earlier and evolving faster than smaller galaxies ( e.g. * ? ? ?
* and references therein ) .
these processes are thought to be accompanied by channeling of material to the central regions , where it form stars and feeds a growing black hole ( bh ) .
the feedback of the latter regulates galaxy formation in an
evolutionary sequence " ( e.g. * ? ? ? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
according to these models galaxies grow through star formation and at first host small " bh possibly in very dense and obscured nuclear environments . as time progresses the combined radiative and mechanical output from star formation and the bh accretion grows too , until the agn reaches qso luminosities and literally blows away the circumnuclear material .
this effectively terminates star formation and the smbh shines briefly ( in cosmic terms ) as an unobscured qso .
once it has accreted the remaining material , the qso switches off , leaving a passively evolving galaxy with a dormant " smbh , such as those observed in present - day massive galaxies @xcite .
the far - infrared ( fir)-submillimetric ( submm ) spectral region is ideally suited for searching for star formation , since the black - body - like ( greybody ) emission by cold dust associated with star formation peaks at those wavelengths . at the same time it suffers very modest
obscuration from the circumnuclear material .
this is proved by the large numbers of strongly star forming submm galaxies ( smgs ) detected by scuba , aztec , sma and , lately , herschel , among other facilities .
there are many multi - wavelength studies of smgs , both from surveys in `` blank fields '' ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) and targeting overdensities of such sources around high redshift objects ( e.g. , @xcite ) .
these studies probe the properties of the star formation and of the agn , and the association between both phenomena , although there are not many cases in which this association has been proved .
smgs have substantial stellar masses in place by @xmath10 . a relatively high fraction of smgs host agn ( e.g. @xcite , but see @xcite ) , which seem to have bh - to - stellar mass ratios smaller than local galaxies @xcite .
this implies that bh growth lags galaxy growth , since the host galaxies are already mature , while the bhs still require substantial growth ( e.g. , by about a factor of 6 ) , which however can be accommodated within the limits imposed by the lifetime of the submm - bright phase @xcite , if accretion occurs close to the eddington limit @xcite .
the observed population of @xmath10 smgs would be sufficient to account for the formation of the population of bright ellipticals seen in the present universe @xcite .
we have found @xcite a sample of x - ray - obscured qsos at @xmath11 ( when most of the star formation and bh growth is occuring in the universe ) with strong submm emission , much higher than x - ray - unobscured qsos at similar redshifs and luminosities , which however represent 85 - 90% of the x - ray qsos at that epoch .
this submm emission is too high to come from the agn @xcite .
the fields around these objects show strong overdensities of smgs @xcite .
the uv and x - ray spectra of the central qsos @xcite show evidence for strong ionized winds which produce the x - ray obscuration . piecing all these clues together
we infer that the host galaxies of these qsos are undergoing strong sfr , while the central smbh are also growing through accretion .
the ionized winds are strong enough to quench the star formation , so the qsos must be just emerging from a strongly obscured accretion state to become normal " unobscured qsos with passively evolving galaxies , in an evolutionary stage which might last about 10 - 15% of the qso lifetime .
they appear to be in the centres of high density peaks of the universe .
in this paper we endeavour to prove the physical association of one of those overdensities to its central qso ( rx j094144.51 + 385434.8 , henceforth rx j0941 ) .
we also investigate the nature and evolutionary stage of those smgs through their rest frame uv - fir seds .
the outline of this paper is as follows : in section [ data ] we summarize the data used in this paper and the reduction process , in section [ zphot ] we describe how the source catalogues in each band have been merged into a single catalogue , and how that catalogue has been used to obtain photometric redshifts .
these are then used in section [ results ] to show which objects are associated with the structure around the qso and to the submm sources , and how we have obtained the different physical parameters for each object .
the evolutionary stage of our sources is discussed in section [ discussion ] .
finally , in section [ conclusions ] we summarize our results .
we have assumed throughout this paper a hubble constant @xmath12 km / s / mpc , and density parameters @xmath13 and @xmath14 . the spectral index @xmath15 is defined as @xmath16 . | we present here photometric redshift confirmation of the presence of large scale structure around the @xmath0 qso rx j0941 , which shows an overdensity of submm sources .
radio imaging confirms the presence of the submm sources and pinpoints their likely optical - nir counterparts .
four of the five submm sources present in this field ( including the qso ) have counterparts with redshifts compatible with @xmath0 .
we show that our photometric redshifts are robust against the use of different spectral templates .
we have measured the galaxy stellar mass of the submm galaxies from their rest - frame @xmath1-band luminosity obtaining @xmath2 , slightly larger than the schechter mass of present day galaxies , and hence indicating that most of the stellar mass is already formed .
we present optical - to - radio spectral energy distributions ( seds ) of the five scuba sources .
the emission of rx j0941 is dominated by reprocessed agn emission in the observed mir range , while the starburst contribution completely dominates in the submm range .
the seds of the other three counterparts are compatible with a dominant starburst contribution above @xmath3 , with star formation rates sfr@xmath4/yr , central dust masses @xmath5 and hence central gas masses @xmath6 .
there is very little room for an agn contribution . from x - ray upper limits and the observed 24 flux , we derive a maximum 2 - 10kev x - ray luminosity of @xmath7 erg / s for any putative agn , even if they are heavily obscured
this in turn points to relatively small black holes with @xmath8 and hence stellar - to - black hole mass ratios about one order of magnitude higher than those observed in the present universe : most of their central black hole masses are still to be accreted .
local stellar - to - black hole masses ratios can be reached if @xmath9% of the available nuclear gas mass is accreted .
= 1 [ firstpage ] galaxies : formation - galaxies : starburst - galaxies : evolution - galaxies : high redshift - submilimetre |
the question of the influence of the nuclear medium on the properties of the nucleon and its resonances has attracted a lot of attention recently both in experiment and theory .
it has also been one of the challenges to understanding strong qcd @xcite where the connection between the nucleon internal degrees of freedom and nuclear medium effects has been a key issue .
historically , nucleon resonances have been among the richest sources of information on low energy qcd phenomena . in particular ,
nucleon resonance excitations by electromagnetic probes and their decays via the strong interaction have been an ideal method for such a study . in experiment ,
the advent of high intensity photon beams at jlab , esrf , elsa , and mami has greatly improved the data for pion production on single nucleons at the resonance regions over the last decade .
these precise measurements significantly improved our knowledge about nucleon resonances and further constrained theoretical phenomenologies @xcite .
however , our knowledge of resonance properties in nuclear reactions @xcite is still very limited , although great strides have been made by several groups to tackle this issue with various prescriptions @xcite .
because of the complexity of the strong interaction in nuclear reactions , many phenomena may contribute , making the theoretical analyses difficult . for instance , recent experimental work @xcite on @xmath0 and @xmath1 suggested a controversial role for the @xmath2 by an empirical property change of the resonance @xcite .
it is essential for any theoretical model to introduce as few parameters as possible .
this is our motivation for providing a quark - level prescription as a first step in this direction .
we propose a quark model approach for meson photoproduction on light nuclei , based on the recent success of a chiral effective theory for the pseudoscalar meson ( @xmath3 , @xmath4 , @xmath5 ) photo- and electroproduction @xcite .
starting with an effective chiral lagrangian for quark - pseudoscalar - meson coupling , this framework would in principle permit a self - consistent inclusion of all the _ s_- and _ u_-channel resonances .
this is a sensible approach to the investigation of the effects of the nuclear medium on the resonances since we believe that such effects should have an overall influence on all the active baryons , i.e. both the nucleon and its resonances . in this work , we will re - visit @xmath6 .
the deuteron , as the simplest composite nucleon system , provides the simplest example for which the effective theory can be directly applied .
we will then explore the response of baryons ( described by the quark model ) to the nuclear medium , and establish the relation between nuclear medium effects and baryon intrinsic dynamics .
of course while we acknowledge that medium effects are hardly likely to have a significant effect for the case of photoproduction on the deuteron , the simplicity of the system in this preliminary work allows for a systematic study of the way such medium effects will be manifest in heavier systems , the subject of future work
. it will be helpful to briefly review the theoretical approaches to @xmath6 available in the literature .
the widely applied isobaric model has been successful in giving a reasonable description of the pion cross sections ( see e.g. ref .
phenomenological methods have also been carried out based on analyses of single nucleon interactions @xcite .
a relativistic quantum hadronic dynamical approach has also produced similar results @xcite .
more recently , a coupled channel approach adopting the blomqvist - laget operators @xcite was developed @xcite . in general , these approaches have reproduced the experimental data reasonably well in the delta resonance regions due to the dominance of the delta magnetic dipole .
a key question in nuclear reactions is the interplay of ambiguities from the reaction mechanism and short - distance components of the nuclear wavefunctions .
promisingly , for the two nucleon system , such a problem can be clarified due to the sensitivities of transition mechanisms to certain kinematics .
this feature simplifies the problem and highlights the physics of interest . here
, we summarize several points which will support the simplifications , and enable us to derive novel information about the nucleon and its resonances in @xmath6 .
firstly , concerning the reaction mechanism , we neglect the rescattering contributions at the delta resonance region due to the dominance of the direct scattering process above the production threshold .
this is consistent with the findings of other models @xcite .
secondly , concerning the nuclear wavefunction , we are interested in the role played by the short - distance components and their effects on the cross section
. we will show that , in the @xmath7 resonance region , the deuteron @xmath8-state generally gives corrections to the backward angle cross sections , while the @xmath9-state dominates at forward angles .
we will then investigate the impact of the nuclear medium effects on both of the kinematical regions .
let us start with an effective lagrangian for the quark - goldstone - boson interaction @xcite @xmath10\psi + \cdot\cdot\cdot,\ ] ] where @xmath11 and @xmath12 denote the vector and axial currents which have the following expressions : @xmath13 and the chiral transformation is , @xmath14 where @xmath15 is the decay constant of the meson @xmath16 , which is the goldstone boson field . in the su(3 )
symmetry limit , the quark field @xmath17 is @xmath18 and the meson field @xmath16 is a 3@xmath193 matrix : @xmath20 thus , the lagrangian in eq .
( [ lagrangian ] ) is invariant under the chiral transformation . expanding the field @xmath21 in eq .
( [ chiral ] ) in terms of @xmath16 , i.e. @xmath22 , we obtain the standard quark - meson pseudovector coupling at tree level @xcite : @xmath23 where the goldstone boson field @xmath16 is now presented by the su(3 ) octet mesons ( @xmath3 , @xmath4 and @xmath5 ) .
the quark - photon electromagnetic coupling is @xmath24 where the photon has three momentum @xmath25 , and the constituent quark carries a charge @xmath26 .
this approach naturally places a constraint on the relative phases of those octet mesons in the su(3 ) flavor symmetry limit , and relates the quark operators to the hadronic ones via explicit quark model wavefunctions for the baryons states . in this sense ,
the quark model transition amplitudes can be compared with hadronic ones , from which information about the baryon structure can be learned .
a systematic investigation of the four channels of pion photoproduction on a bare nucleon was carried out in @xcite , where it was shown that , up to 500 mev photon energies , all the available observables could be reasonably accounted for with only one parameter for the delta resonance coupling .
great effort was devoted in that work to restrict the number of parameters by taking advantage of the quark model symmetry in order that a complete set of resonances could be systematically taken into account .
overall quantitative agreement of the model with experimental data can be seen in @xcite . in the present study , however , due to the fast fall - off of the deuteron momentum space wavefunction at large momenta and the photon energies considered ( @xmath27 mev ) , contributions from resonances in the second resonance regions , ( @xmath28 , @xmath29 , @xmath2 , @xmath30 , @xmath31 , @xmath32 , @xmath33 , @xmath34 @xcite , etc ) are generally negligible . in the first resonance region , where the delta and born term are dominant , it is a good approximation to include only these two contributions in the direct scattering process .
in the @xmath35-@xmath36 c.m . system the unpolarized cross section for @xmath6 can be written as @xmath37 where @xmath25 is the photon momentum defining the @xmath38-axis , and @xmath39 is the squared total c.m .
energy ; @xmath40 is the deuteron total spin , while @xmath41 , @xmath42 , and @xmath43 denote the spin projections for the initial , final state deuterons , and the incoming photon , respectively .
the amplitudes with @xmath44 and @xmath45 are not independent due to parity conservation . the invariant amplitude @xmath46 in the direct scattering ( fig .
[ fig:(1 ) ] ) can be expressed as @xmath47 where @xmath48 denotes the deuteron ground state wavefunction with spin projection @xmath49 ; @xmath50 is the transition operator for the single nucleon interaction derived within the quark model , and is the nucleon photoproduction operator averaged over the isospin wavefunction of the deuteron .
the kinematical variables are defined as , @xmath51 , and @xmath52 .
the deuteron wavefunction is defined as @xmath53 where @xmath54 and 2 denote the @xmath9 and @xmath8-state components and the spin wavefunction is @xmath55 where @xmath56 is the nucleon spin wavefunction , and @xmath57 and @xmath58 are spin projections for the spectator and active nucleon .
the well - established paris model @xcite wavefunction for the deuteron is adopted , which means ambiguities from the nuclear wavefunctions are minimised in this examination .
the paris potential is obtained from a phenomenological momentum - dependent meson exchange model for the nucleon - nucleon ( @xmath59-@xmath59 ) interaction with parameters chosen to fit @xmath60-@xmath61 scattering data for energies from the deuteron binding energy up to several hundred mev , thus giving a reliable description of the deuteron wavefunction down to short @xmath60-@xmath61 separations . to check the sensitivity of the reaction mechanism to the nuclear structure
, we also carry out the calculation with a hulthn wavefunction , for which the detailed description and parameters are given in ref .
this simpler wavefunction is generated by the yamaguchi - yamaguchi separable @xmath60-@xmath61 interaction @xcite which has the useful feature that the @xmath9 and @xmath8 components of the wavefunction have simple analytical forms in momentum space .
the @xmath60-@xmath61 potential in this model has parameters adjusted to give the experimental deuteron binding energy , quadrupole moment , and @xmath60-@xmath61 triplet state effective range , and scattering length , all to high accuracy .
however , it is not expected to be as accurate as the paris wavefunction at larger momenta . thus , since the long - distance asymptotic behavior of the deuteron wavefunction is well - known , the contrast will further pin down the nuclear structure effects arising from short - distance components .
we neglect the @xmath8-state to @xmath8-state transition term in the amplitude in eq.(4 ) but retain the @xmath9 to @xmath8 transition .
the @xmath8-@xmath8 transition may become important at high momentum transfers and show up in the backward - angle cross sections . however , this is not the focus in this work .
firstly , we examine the single nucleon reaction operators in @xmath62 by applying the transition operators derived in the free nucleon reaction to the direct scattering process ( fig .
[ fig:(1 ) ] ) . note that the quark model parameters are determined in the free nucleon reactions while the deuteron wavefunction used is from independent studies .
our calculations are therefore parameter - free predictions of the effective theory in the quark model framework .
it should also be pointed out that since the final state pions are treated as point - like particles , the size effects and possible medium effects on the mesons are absorbed into the quark model parameters .
this is different from isobaric prescriptions , where the meson and baryon degrees of freedom could be influenced by the nuclear medium separately ( see e.g. refs .
@xcite ) . in this study
, we will present a new dynamic view of this phenomenon by relating the nuclear medium s influence on the baryon properties to the intrinsic dynamic aspects of baryons via the quark model parameters .
we note that similar ideas are also proposed in ref .
@xcite . in fig .
[ fig:(2 ) ] , the results for the differential cross sections at @xmath63 mev with both wavefunctions are presented . as shown by the solid ( paris wavefunction ) and dashed curve ( hulthn wavefunction ) in fig .
[ fig:(2 ) ] , these two wavefunctions produce the same cross sections at forward angles due to the same asymptotic behavior , while discrepancies arise at large angles due to differences between these two models at short distances .
although the experimental data @xcite are still sparse , the cross sections using both wavefunctions are in good agreement with the data over a wide angle region .
we also present the results without the @xmath8-state contributions ( the dotted and dot - dashed curve stand for the @xmath9-state paris and hulthn wavefunctions , respectively ) , where the large - angle cross section differences highlight the @xmath8-state contributions .
the above result is a direct examination of the quark model treatment for the elementary process , which deserves more careful consideration . since nucleons in the deuteron
are weakly bound , one would expect a weak effect from the nuclear medium .
the above result indeed suggests that the baryons behave as they do in the single nucleon reactions .
therefore , in principle no parameter change to the baryon properties is needed here .
this conclusion is consistent with other previous studies @xcite .
however , where our approach has an advantage is that it also allows us to examine the impact of the nuclear medium more systematically .
we consider a mechanism that changes the properties of the delta resonance due to the change of quark model parameters , e.g. quark potential strength within the baryons .
the quark model potential strength @xmath64 and constituent quark mass @xmath65 are explicitly related to the mass of the baryons in the nonrelativistic constituent quark model ( nrcqm ) , where the nucleon and delta resonance belong to the same representation @xcite .
a simple mass relation from the quark hyperfine interaction will allow us to make an empirical connection : @xmath66 where @xmath67 is the degenerate mass between the nucleon and delta before the hyperfine splitting , and @xmath68 and @xmath69 are the physical masses of the nucleon and delta .
the quantity @xmath70 is determined by the contact coupling : @xmath71 where @xmath72 is the electromagnetic coupling .
it is reasonable to assume , for the case of the deuteron , that the nucleon mass @xmath68 is fixed for any set of parameters .
we can then express the mass formula as : @xmath73+m_n \ , \ ] ] where the barred " quantities refer to the values given by the medium - modified parameters .
the delta s strong decay width for @xmath74 can be also related to the quark model parameters : @xmath75 where @xmath76 and @xmath77 are the nucleon and pion energies in the delta c.m .
these variables are determined by the delta mass , which hence are related to eq .
( [ delta - mass ] ) . also , the change of the potential strength will influence the form factors for baryons , i.e. both nucleon and delta .
therefore , even though we fix the mass of the nucleon , its size may change due to the change of quark potential strength in the nuclear medium . for the delta
, we can simply express @xmath78 .
the @xmath79 coupling @xmath80 is determined in the single free nucleon reaction @xcite , where @xmath81 is the strength given by the experimental data .
the simple relations from the static properties of baryons introduces new aspects in the reaction mechanism . if the medium influence leads to a significant change to @xmath64 and @xmath65 , from eq .
( [ delta - mass ] ) and ( [ delta - total - width ] ) , the mass shift and the width change of the delta will be correlated . as a sensitivity test
, we assume the quark mass is fixed , and find that a change of the quark potential strength from @xmath82 mev ( determined in the free nucleon reaction @xcite ) to @xmath83 mev results in a mass increase of about 36 mev as suggested by other calculations and total width change from @xmath84 mev to @xmath85 mev . in fig .
[ fig:(3 ) ] , we present the calculated results ( dashed curve ) for the differential cross sections with such a property change to the baryons . by comparison with the unchanged - parameter calculation ( solid curve ) , we see that the cross section is very sensitive to the quark model parameters at forward angles . this can be understood as follows .
the overall shape of the cross section is governed by the behaviour of the deuteron wavefunctions , which ensure that the integrand in eq .
( [ impulse ] ) is only non - negligible when the arguments of both wavefunctions are small .
this is when @xmath86 is of the same order as both @xmath87 and @xmath88 , and when @xmath25 and @xmath89 are parallel ( forward scattering ) . in the c.m .
frame we also have @xmath90 and hence at forward angles where @xmath86 is large , @xmath91 , the momentum of the active nucleon , is at its smallest ; at larger scattering angles it carries away more momentum from its interaction with the photon . since the photon energy is chosen to be close to the delta threshold in the photon - active nucleon lab system , the forward angles correspond to the kinematics where the delta will contribute significantly to the cross section , and medium effects on the delta will also show up most here . at backward angles ,
most of the momenta carried by the deuteron will be on the active nucleon , resulting in a shift of the photon - active - nucleon c.m .
energy far away from the delta mass position leading to the small effects from the parameter change .
the fact that the overall shape of the cross section is governed by the behaviour of the deuteron wavefunctions , also accounts for the negligible effects from the energy conservation breaking at the initial photon - nucleon and final meson - nucleon interaction vertices in this reaction . for denser nuclear medium
, such effects may become important due to larger contributions from short - range components of nuclear wavefunctions .
this could be a possible mechanism by which the nuclear medium can change the quark model parameters . in fig .
[ fig:(4 ) ] , the medium influence on the delta magnetic dipole is compared with the free nucleon case in the photon - nucleon c.m .
the largest effects are from the region where the delta is excited close to its mass shell . as discussed above , for a photon energy @xmath63 mev , the close - to - threshold region corresponds to the forward scatterings , with the off - shell region to the backward angles .
when the delta goes far off - shell , the parameter change effects become negligible .
thus , one would expect that the forward angle scattering is the ideal place for the investigation of the delta property changes due to the medium .
it is noted that the major contributions to the small angle cross section come from deuteron configurations in which the neutron - proton ( @xmath60-@xmath61 ) separation is larger than the range of strong interactions . for these configurations
any medium modifications to @xmath64 will be negligible .
figure [ fig:(3 ) ] should therefore only be used as an indication of the relative sensitivity of the pion production process at different momentum transfers to a ` fixed ' density - independent change to @xmath64 . any density dependent mechanism which produced a change at small @xmath60-@xmath61 separations ( and large momentum transfer ) of the magnitude used in our sensitivity test
would be expected to be associated with much smaller changes at large @xmath60-@xmath61 separations and small momentum transfers .
we have studied the photoproduction of @xmath92 on the deuteron at the energy of the delta excitation in a quark model framework with an effective chiral lagrangian .
the advantage of this approach has been that the nucleon and its resonances were introduced into the nuclear reaction with only a limited number of parameters @xcite . in this approach
we propose that the nuclear medium effects on the baryon properties can be recognized through the change of quark model parameters for baryons .
in particular , the nuclear medium may influence the quark potential strengths of a baryon state , and further lead to changes to its mass and width .
it thus naturally relates the resonance property changes to their intrinsic dynamical aspects .
mechanisms by which the nuclear medium changes the quark model parameters need to be addressed by investigating reactions on heavier and denser nuclei .
we expect that a systematic investigation of reactions on heavier nuclei will establish the medium - density - dependence of the quark model parameters .
it is also important to extend this approach to the second resonance region , where higher resonances such as @xmath28 , @xmath2 , @xmath29 , etc .
, can be investigated .
consequent work will be reported elsewhere .
authors thank j.a .
tostevin for providing the paris wavefunction and useful discussions .
useful comments from z.p .
li , c. bennhold and r.l .
workman on an early version of this paper are greatly appreciated .
useful discussions with e. oset and t.t.s .
kuo are gratefully acknowledged .
financial supports of the u.k . engineering and physical sciences research council ( grant no .
gr / r78633/01 and gr / m82141 ) are gratefully acknowledged .
99 white paper on key issues in hadronic physics , jefferson lab , dec .
, 2000 , hep - ph/0012238 .
b. krusche and s. schadmand , prog .
phys . * 51 * , 399 ( 2003 ) [ nucl - ex/0306023 ] .
walker , phys . rev . * 182 * , 1729 ( 1969 ) .
d. drechsel , o. hanstein , s.s .
kamalov , and l. tiator , nucl .
phys . * a 645 * , 145 ( 1999 ) .
o. hanstein , d. drechsel , and l. tiator , nucl . phys . * a 632 * , 561 ( 1998 ) .
i.g . aznauryan and s.g .
stepanyan , phys .
d * 59 * , 054009 ( 1999 ) .
crawford , proceedings of the workshop on the physics of excited nucleons , mainz , germany , 2001 , ( nstar 2001 ) , p. 163 . r.m .
davidson , n.c .
mukhopadhyay , and r.s .
wittman , phys .
d * 43 * , 71 ( 1991 ) .
s. nozawa , b. blankleider , and t .- s.h .
lee , nucl . phys . * a 513 * , 459 ( 1990 ) . s. nozawa and t .- s.h .
lee , nucl . phys . * a 513 * , 511 ( 1990 ) . t. sato and t .- s.h .
lee , phys .
c * 54 * , 2660 ( 1996 ) ; * 63 * , 055201 ( 2001 ) .
h. garcilazo and e. moya de guerra , nucl
. phys . * a 562 * , 521 ( 1993 ) .
z. li , r.a .
arndt , l.d .
roper , and r.l .
workman , phys . rev . *
c 47 * , 2759 ( 1993 ) .
arndt , i.i .
strakovsky , and r.l .
workman , phys .
c * 53 * , 430 ( 1996 ) . z .-
li , h .- x . ye , and m .- h .
lu , phys .
c * 56 * , 1099 ( 1997 ) .
q. zhao , j.s .
al - khalili , z .- p .
li , and r.l .
workman , phys .
c * 65 * , 065204 ( 2002 ) .
li and b. saghai , nucl .
a * 644 * , 345 ( 1998 ) .
b. saghai and z .- p .
li , eur .
j. a * 11 * , 217 ( 2001 ) .
q. zhao , b. saghai and z .- p .
li , j. phys .
g * 28 * , 1293 ( 2002 ) .
g. van der steenhoven , _ et al .
57 * , 182 ( 1986 ) . v. metag , proceedings of the workshop on the physics of excited nucleons , mainz , germany , 2001 , ( nstar 2001 ) , p. 447
. f. rambo _
et al . _ ,
* a660 * , 69 ( 1999 ) .
carrasco , j. nieves , and e. oset , nucl . phys . *
a565 * , 797 ( 1993 ) ; r.c .
carrasco , e. oset , nucl .
a536 * , 445 ( 1992 ) ; j. nieves , e. oset , and c. garcia - recio , nucl . phys . *
a554 * , 554 ( 1993 ) .
lutz and c.l .
korpa , nucl .
a700 * , 309 ( 2002 ) ; c.l .
korpa and m.f.m .
lutz , nucl - th/0306063 .
j. lehr and u. mosel , phys .
c * 64 * , 042202 ( 2001 ) ; phys .
c * 68 * , 044601 ( 2003 ) .
i. blomqvist and j.m .
laget , nucl . phys . *
a280 * , 405 ( 1977 ) .
p. bosted and j.m .
laget , nucl . phys . *
a296 * , 413 ( 1978 ) .
c. lazard , r.j .
lombard , and z. maric , nucl . phys . *
a271 * , 317 ( 1976 ) .
koch , and r.m .
woloshyn , phys .
c * 16 * , 1968 ( 1977 ) .
h. garcilazo and e. moya de guerra , phys .
c * 52 * , 49 ( 1995 ) .
kamalov , l. tiator , and c. bennhold , phys .
rev . lett . *
75 * , 1288 ( 1995 ) ; phys .
c * 55 * , 98 ( 1997 ) .
a. manohar and h. georgi , nucl .
b234 * , 189 ( 1984 ) .
d. e. groom _ et al .
_ ( particle data group ) , eur .
j. c * 15 * , 1 ( 2000 ) .
m. lacombe , b. loiseau , r. vinh mau , j. cote , p. pires , and r. de tourreil , phys .
* b 101 * , 139 ( 1981 ) .
a.a . ioannides and r.c .
johnson , phys .
c * 17 * , 1331 ( 1978 ) .
y. yamaguchi and y. yamaguchi , phys .
95 * , 1635 ( 1954 ) .
d.o . riska and g.e .
brown , nucl . phys . * a 679 * , 577 ( 2001 ) . g. von holtey _
et al_. , z. phys .
* 259 * , 51 ( 1973 ) .
close , an introduction to quarks and partons " , academic press , london ( 1979 ) . said online database , http://gwdac.phys.gwu.edu | a quark - model framework for studying nuclear medium effects on nucleon resonances is described and applied here to pion photoproduction on the deuteron , which is the simplest composite nucleon system and serves as a first test case .
pion photoproduction on nuclei is discussed within a chiral constituent quark model in which the quark degrees of freedom are explicitly introduced through an effective chiral lagrangian for the quark - pseudoscalar - meson coupling .
the advantage of this model is that a complete set of nucleon resonances can be systematically included with a limited number of parameters .
also , the systematic description of the nucleon and its resonances at quark level allows us to self - consistently relate the nuclear medium s influence on the baryon properties to the intrinsic dynamic aspects of the baryons .
as the simplest composite nucleus , the deuteron represents the first application of this effective theory for meson photoproduction on light nuclei .
the influence of the medium on the transition operators for a free nucleon is investigated in the delta resonance region .
no evidence is found for a change of the delta properties in the pion photoproduction reaction on the deuteron since the nuclear medium here involves just one other nucleon and the low binding energy implies low nuclear density .
however , we show that the reaction mechanism is in principle sensitive to changes of delta properties that would be produced by the denser nuclear medium of heavier nuclei through the modification of the quark model parameters .
pacs numbers : 13.60.le , 25.20.-x , 25.10.+x |
typical erds type problems in extremal discrete mathematics ask for the minimum or maximum number of certain configurations over all inputs of a given size .
they are easy to formulate but often extremely hard to answer .
their impact on mathematics and computer science has been enormous , not only because of specific algorithms based on combinatorial bounds but also because they have triggered the development of theoretical and practical methods that turned out to be applicable elsewhere .
some of the most simply formulated yet notoriously hard erds type problems occur in combinatorial geometry . in 1946 , erds @xcite asked two questions on distances : ( 1 ) at most how many times can a given distance occur among @xmath0 points in the plane ; ( 2 ) what is the minimum number of distinct distances determined by @xmath0 points in the plane ?
the difficulty of the , so called , _ unit distance _ and _ distinct distance _ problems is still to be measured .
erds and purdy @xcite generalized the unit- and distinct distance problems to congruent ( or repeated ) simplices : what is the maximum number of congruent @xmath8-dimensional simplices among @xmath0 points in @xmath9 , for @xmath7 ?
no asymptotically tight bound is known for this problem for @xmath15 , ( and it is trivial for @xmath16 by lenz - type constructions ) .
pach and sharir @xcite derived bounds on the maximum number of occurrences of the same angle determined by a planar @xmath0-element point set ; recently apfelbaum and sharir @xcite studied the analogous problem in three - space . a recent book by bra , moser , and pach @xcite and a survey of pach and sharir @xcite provide substantial details on these and other similar problems .
in the sequel , we focus on problems about the extremal number of simplices with certain volume properties in a point set . in 1967 , a. oppenheim ( see @xcite ) asked what is the maximum number of unit area triangles determined by @xmath0 points in the plane .
erds and purdy gave an @xmath17 upper bound , and also showed that a suitable section of the integer lattice yields @xmath18 such triangles @xcite .
the currently best upper bound , @xmath19 , due to dumitrescu , sharir and cs .
tth @xcite , recently improved an older @xmath20 bound of pach and sharir @xcite .
answering further questions of erds and purdy @xcite , bra , rote , and swanepoel @xcite showed the following two results : ( 1 ) the maximum number of triangles of maximum area ( or of maximum perimeter ) determined by @xmath0 points in the plane is exactly @xmath0 .
( 2 ) the maximum number of triangles of minimum ( nonzero ) area determined by @xmath0 points in the plane is @xmath21 . in 1982 ,
erds , purdy , and straus @xcite considered the generalization of the problem of distinct triangle areas to higher dimensions and posed the following problem : let @xmath22 be a set of @xmath0 points in @xmath9 not all in a hyperplane .
what is the minimal number @xmath23 of distinct volumes of full - dimensional simplices with vertices in @xmath22 ?
it is easy to see that @xmath24 by taking @xmath1 sets of about @xmath25 equally spaced points on parallel lines through the vertices of an @xmath26-simplex .
erds , purdy , and straus conjectured that equality holds at least for sufficiently large @xmath0 ( see also @xcite ) .
very recently , pinchasi @xcite confirmed the conjecture in the plane after earlier work by burton and purdy @xcite and dumitrescu and cs .
tth @xcite : the minimum number of distinct ( nonzero ) triangle areas determined by @xmath0 noncollinear points is @xmath27 . in this paper we give a first partial result for @xmath28 by proving @xmath29 for every @xmath13 , which is optimal up to a multiplicative constant . as expected , the three dimensional analogues of combinatorial problems in the plane are often much harder . only the recent few years saw intensifying work on these problems in three - space @xcite .
algorithmic problems on sets of points in euclidean space have been often attacked using the duality transform and the machinery of constructing hyperplane arrangements .
the first applications of this technique can be found in early works of chazelle , edelsbrunner , guibas , lee , orourke and seidel @xcite : among others , the problem of computing a minimum area triangle in a given set of @xmath0 points .
a @xmath30 time algorithm for such a task was given independently in @xcite and @xcite .
if there are degeneracies in the set , the algorithm returns zero area and a triplet of collinear points .
an extension of the algorithm for finding a minimum volume simplex among @xmath0 points in @xmath9 was given in @xcite : it runs in @xmath31 time , but again reports zero volume if there are @xmath32 points on a hyperplane .
here we further extend the algorithms for two and three dimensions to report all simplices of minimum _ nonzero _ volume , within the same @xmath30 and @xmath5 running times , respectively .
[ [ our - contribution . ] ] our contribution .
+ + + + + + + + + + + + + + + + + in this paper , we address classical problems on minimum , unit , and distinct volume simplices .
we show that every set of @xmath0 points in three - space determines @xmath5 minimum ( positive ) volume tetrahedra , and there are points sets in @xmath2 that span @xmath33 minimum volume tetrahedra .
our techniques generalize to arbitrary dimensions : for every @xmath34 , @xmath35 , any set of @xmath0 points in @xmath9 determines @xmath36 minimum ( positive ) volume @xmath8-dimensional simplices ( each having @xmath37 vertices ) , and there are points sets in @xmath9 that span @xmath38 minimum volume @xmath8-dimensional simplices . in three - space , we also give a @xmath5 time algorithm for reporting all tetrahedra of minimum nonzero volume , and thereby extend an early algorithm of edelsbrunner , orourke , and seidel .
we prove that every set of @xmath0 points in @xmath2 determines at most @xmath11 unit - volume tetrahedra , and there are point sets that span @xmath39 unit - volume tetrahedra .
finally , we show that for every @xmath13 , any set of @xmath0 points in @xmath9 , not all in a hyperplane , determines at least @xmath40 full - dimensional simplices of distinct volumes , and there are point sets for which this number is @xmath41 ; this gives a first answer to the question of erds , purdy , and straus .
[ [ organization . ] ] organization .
+ + + + + + + + + + + + + section [ s - toolbox ] presents a collection of tools we use from euclidean geometry , previous results and extensions of previous results adapted to our goals .
we prove our main theorems on the number of minimum- , unit- , and distinct volume tetrahedra in sections [ s - minimum ] , [ s - unit ] , and [ s - distinct ] , respectively . to simplify notation , we assume in our proofs that @xmath42 is an integer .
we will frequently apply a classic result of szemerdi and trotter @xcite on the number of point - lines incidences in the plane .
( since collinear points remain collinear under any affine transformation , the result holds in arbitrary dimensions . ) the constant factor hidden in the asymptotic notation has been significantly improved by the cutting method and by the theory of crossing numbers @xcite ; the current best constant is due to pach et al .
the szemerdi - trotter bound comes in two equivalent formulations ; we also state two immediate corollaries that we use in our proofs .
given a point set @xmath22 in @xmath9 , for any integer @xmath43 , a line is called _
@xmath8-rich _ if it is incident to at least @xmath8 points of @xmath22 .
we denote by @xmath44 the set of @xmath8-rich lines .
( szemerdi - trotter @xcite ) .
[ thm : st-83b ] given @xmath0 points in @xmath9 , @xmath45 , the number of @xmath8-rich lines , @xmath43 , is @xmath46 ( szemerdi - trotter @xcite ) .
[ thm : st-83a ] the number of point - line incidences among @xmath0 points and @xmath47 lines in @xmath9 , @xmath45 , is @xmath48 [ cor : st-83b ] given @xmath0 points in @xmath9 , @xmath45 , the number of point - line incidences among the @xmath0 points and the @xmath8-rich lines , @xmath43 , is @xmath49 .
note that in theorem [ thm : st-83a ] , the term @xmath47 is responsible for the incidences on lines containing a single point .
[ cor : st-83 ] given @xmath0 points and @xmath47 lines in @xmath9 , @xmath45 , each line containing at least two points , the number of point - line incidences among them is @xmath50 .
incidence bounds have been key components in many results in combinatorial geometry .
some szemerdi - trotter - type bounds were found for point - plane incidences in the space but they are either not known to be tight or they hold in severely restricted settings only ( e.g. , @xcite ) . instead of these multidimensional bounds , we apply _ planar _ incidence bounds on the _ projections _ of a finite point set @xmath51 onto planes in certain directions determined by @xmath22 .
the szemerdi - trotter bound on point - line incidences is tight in the worst case , but does not hold for points with _
multiplicities_. since the planar projection of a three - dimensional point set may have an irregular distribution of multiplicities , we partition the point sets ( and the planar multiset ) into subsets of roughly the same multiplicities .
we demonstrate this technique in section [ s - unit ] , where we aggregate szemerdi - trotter bounds for various substructures and give an upper bound on the number of incidences of lines and points _ with multiplicities _ in the plane , and ultimately a bound on incidences in three - space .
we extensively apply the concepts of _ degenerate _ and _ nondegenerate _ planes in three - space .
they were introduced recently by elekes and tth @xcite here , we use the same concepts with a different terminology . consider a point set @xmath51 . for a constant @xmath52 , @xmath53
, we say that a plane @xmath54 is _ @xmath52-degenerate _ if at most @xmath55 points of @xmath56 are collinear . in this paper , we fix @xmath57 and say that a plane @xmath54 is _ nondegenerate _ if it is @xmath58-degenerate ; otherwise @xmath54 is _ degenerate_. a nondegenerate plane is always spanned by @xmath22 , while a degenerate plane may contain only one point , or collinear points . if @xmath22 lies in a plane @xmath54 , we refer to @xmath22 as degenerate or nondegenerate according to the above definition .
we apply several consequences of a well known combinatorial geometric result due to beck @xcite .
[ lem : beck2 ] ( beck @xcite ) .
consider a set @xmath22 of @xmath0 points in the plane .
if at most @xmath47 points of @xmath22 are collinear , then @xmath22 determines at least @xmath59 distinct lines .
[ cor : beck3 ] consider a set @xmath22 of @xmath0 noncollinear points in the plane .
if at most @xmath47 points of @xmath22 are collinear , then @xmath22 determines at least @xmath60 nondegenerate triangles .
if every line contains fewer than @xmath61 points , then by lemma [ lem : beck2 ] , @xmath22 determines @xmath62 line segments .
if a line contains @xmath61 points , then there are @xmath62 segments along this line . in either case ,
each of these line segments is the base for at least @xmath63 distinct nondegenerate triangles .
since we count every triangle at most three times , we obtain @xmath60 distinct triangles . for a set @xmath22 of @xmath0 points in the plane , let @xmath64 denote the number of unit area triangles determined by @xmath22 .
let @xmath65 be a constant such that @xmath66 for every set @xmath22 of @xmath0 points in the plane .
it is conjectured that @xmath65 can be arbitrarily small ; we can assume @xmath67 due to the bound @xmath68 of pach and sharir @xcite ( the current best bound @xmath69 gives @xmath70 , cf .
we use the general parameter @xmath71 when we further chisel the bound on unit area triangles in the special case of degenerate planes .
we apply these results in section [ s - unit ] in a charging scheme , where we distribute unit area triangles spanned by a planar point set among line segments in that plane .
[ lem : unit ] let @xmath22 be a set of @xmath0 points in the plane , and @xmath72 be a line incident to exactly @xmath47 points of @xmath22
. then @xmath22 determines at most @xmath73 unit area triangles .
put @xmath74 , and let us denote the lines parallel to @xmath72 containing at least one point of @xmath75 by @xmath76 , where @xmath77 is the number of such lines
. for every @xmath78 , let @xmath79 .
we partition the unit area triangles into three subsets : * @xmath80 counts triangles with all three vertices in @xmath75 ; * @xmath81 counts triangles with two vertices on @xmath72 and one in @xmath75 ; * @xmath82 counts triangles with one vertex on @xmath72 and two in @xmath83 ; \(i ) by the assumption , @xmath84 .
\(ii ) for every line @xmath85 , @xmath86 , there are at most @xmath87 unit area triangles @xmath88 , with @xmath89 and @xmath90 .
therefore , @xmath91 .
\(iii ) @xmath75 determines less than @xmath92 line segments @xmath93 . for segments @xmath93 non - parallel to @xmath72 , there are at most two vertices @xmath94 such that @xmath88 has unit area ( because @xmath95 has to lie on one of two lines parallel to @xmath93 ) , so we have less than @xmath96 unit area triangles of this kind . for segments @xmath93 on a line @xmath85 parallel to @xmath72 ,
there are at most @xmath97 unit area triangles @xmath88 , with @xmath98 and @xmath94 .
summing over all lines @xmath85 , there are at most @xmath99 triangles of unit area ( since @xmath100 ) of this second kind . consequently @xmath101 .
altogether we have at most @xmath102 unit area triangles in @xmath22 .
[ cor : dir ] there is a constant @xmath103 with the following property .
if a set @xmath22 of @xmath0 points in the plane determines @xmath104 nondegenerate triangles , then there is a subset @xmath105 of @xmath106 points and for each @xmath107 there is a set @xmath108 of @xmath109 pairwise non - overlapping line segments spanned by @xmath22 such that ( i ) all segments in @xmath108 have a common endpoint @xmath110 and ( ii ) the sets of segments @xmath108 , @xmath107 , are pairwise disjoint .
first assume that @xmath22 is nondegenerate .
by corollary [ cor : beck3 ] , we have @xmath111 and @xmath22 spans @xmath62 distinct lines . for any @xmath8 , @xmath112 ,
at most @xmath113 lines are incident to more than @xmath8 points by theorem [ thm : st-83b ] . there is a constant @xmath114 such that the set @xmath115 of lines incident to at least 2 and at most @xmath116 points still contains @xmath62 lines .
for each line @xmath117 , choose an arbitrary point @xmath110 among the at most @xmath116 points in @xmath118 and place @xmath72 into the set @xmath108 .
the resulting sets @xmath108 are disjoint and each contain at most @xmath119 lines . for at least @xmath40 points @xmath120 ,
the set @xmath108 contains at least @xmath40 lines .
next assume that @xmath22 is degenerate , that is , there is a line @xmath72 incident to @xmath121 points . in this case , every point off @xmath72 forms @xmath62 triangles with point pairs from @xmath72 , and so @xmath122 .
each of the @xmath63 points off the line @xmath72 is incident to @xmath40 lines spanned by @xmath22 . for any constant @xmath8 , @xmath123 ,
at most @xmath124 lines are incident to more than @xmath8 points in @xmath75 by theorem [ thm : st-83b ] . there is a constant @xmath125 such that the set @xmath126 of lines incident to at least 2 and at most @xmath127 points contains @xmath59 elements . for each line in @xmath126 , choose an arbitrary incident point @xmath110 and assign the line to @xmath108 . for at least @xmath128 points @xmath129 , the resulting set @xmath108 contains at least @xmath40 lines .
[ cor : unit+ ]
if a set of @xmath0 points in the plane determines @xmath104 nondegenerate triangles , then at most @xmath130 of them have unit area .
first , suppose that @xmath22 is nondegenerate .
by lemma [ lem : beck2 ] , @xmath111 , hence @xmath131 .
next , suppose that @xmath22 is degenerate .
a line @xmath72 passes through @xmath132 points .
put @xmath133 .
as in the proof of corollary [ cor : dir ] , we have @xmath134 .
it follows , using lemma [ lem : unit ] , that @xmath135 .
beck @xcite also showed that lemma [ lem : beck2 ] can be extended to higher dimensions : ( beck @xcite ) .
[ lem : b ] for any @xmath45 there exist constants @xmath136 $ ] such that , for any set @xmath22 of @xmath0 points in @xmath9 , at least one of the following holds . 1 .
a hyperplane contains more than @xmath137 points of @xmath22 ; or 2 .
the @xmath1-tuples of @xmath22 span at least @xmath138 distinct hyperplanes .
we show that @xmath0 points determine at most @xmath139 minimum volume tetrahedra in three space .
the upper bound is based on a new charging scheme which assigns each tetrahedron of minimum volume to one of its four faces .
we then extend our charging scheme and show that for any fixed @xmath34 , @xmath35 , the number of @xmath8-dimensional simplices of minimum ( nonzero ) volume in @xmath9 is @xmath36 , where the constant of proportionality depends only on @xmath8 and @xmath1 .
this bound is best possible apart from the constant factor .
[ t - min ] the number of tetrahedra of minimum ( nonzero ) volume spanned by @xmath0 points in @xmath2 is at most @xmath3 , and there are point sets for which this number is @xmath4 . given @xmath0 points in @xmath2 , all tetrahedra of minimum nonzero volume can be reported in @xmath5 time and @xmath30 working space .
[ [ lower - bound . ] ] lower bound .
+ + + + + + + + + + + + the lower bound construction is simple .
form a rhombus with unit sides in the @xmath140-plane from two equilateral triangles @xmath88 and @xmath141 with a common side @xmath93 .
extend it to a prism in 3-space , and make a unit volume tetrahedron with vertices @xmath142 , @xmath143 , @xmath95 , and @xmath1 , on the four vertical lines .
replace each of @xmath142 , @xmath143 , @xmath95 , and @xmath1 , with @xmath144 equally spaced points with inter - point distances @xmath145 along these lines , for a sufficiently small @xmath146 ( assume @xmath0 is divisible by 4 ) .
observe that each tetrahedron with one vertex on each of the four lines has volume very close to 1 , but the minimum volume is @xmath147 , given by tetrahedra with two consecutive vertices on a line , and two vertices on any two of the other three lines .
the number of such tetrahedra is @xmath148 .
all other tetrahedra have zero volume . [
[ upper - bound . ] ] upper bound .
+ + + + + + + + + + + + let @xmath22 be a set of @xmath0 points in @xmath2 .
denote by @xmath149 the tetrahedron determined by four non - coplanar points @xmath150 and by @xmath151 its volume .
similarly , let @xmath152 denote the tetrahedron determined by the endpoints of two line segments @xmath153 and @xmath154 .
the key to our upper bound is the following charging scheme : assign every ( nondegenerate ) tetrahedron determined by @xmath22 to one of its four faces as follows .
assign @xmath155 to a triangle face of maximum area among the faces adjacent to a diameter of @xmath155 .
we show that at most a constant number of minimum volume tetrahedra are charged to every triangle , and this yields the desired @xmath5 upper bound .
consider a ( nondegenerate ) triangle @xmath156 with @xmath157 .
let @xmath93 be a diameter of @xmath156 .
choose a coordinate system such that @xmath142 is the origin , @xmath93 lies on the @xmath158-axis , and @xmath159 lies in the @xmath140-plane .
refer to fig .
[ fig : min ] .
= 3.9 in let @xmath160 , and denote by @xmath161 the height from @xmath95 in @xmath159 .
assume that the minimum volume of a tetrahedra spanned by @xmath22 is @xmath162 .
all points @xmath1 with @xmath163 must lie in two horizontal planes at distance @xmath164 from the @xmath140-plane .
let us consider the plane @xmath165 for now . by our assignment , all points @xmath1 must lie in the interior of an axis - aligned rectangle @xmath166 $ ] in the plane @xmath167 ( otherwise @xmath93 would not be the diameter or @xmath168 would have larger area than @xmath156 ; note that point @xmath1 can not lie on the boundary of @xmath169 ) .
partition the open rectangle @xmath169 into two rectangles @xmath170\times [ z_0]$ ] and @xmath171.$ ] for @xmath172 , there is at most one point @xmath173 such that @xmath174 is minimum .
assume there are two points @xmath175 for some @xmath176 .
we will pick a side @xmath177 so that @xmath178 is not parallel with @xmath179
. then @xmath180 is a nondegenerate tetrahedron .
we show that @xmath181 .
assume first that @xmath179 is not parallel to @xmath93 , and pick @xmath182 .
the segment @xmath179 lies in rectangle @xmath183 , @xmath184 ; observe that at most one element of @xmath185 may lie on the boundary of @xmath183 .
denote by @xmath186 and @xmath187 the two lines containing the two sides of @xmath183 parallel to the @xmath158-axis .
the volume of @xmath180 strictly increases if we move @xmath1 to position @xmath188 and @xmath189 to position @xmath190 along the line @xmath179 such that @xmath191 and @xmath192 .
the volume of @xmath193 is @xmath194 , hence @xmath195 ( e.g. , the base triangle of @xmath193 lies in the @xmath196-plane , and its height is @xmath161 ) .
a contradiction .
next assume that @xmath179 is parallel to @xmath93 , and pick @xmath197 . in this case ,
@xmath1 and @xmath189 are strictly in the interior of @xmath183 . if we replace @xmath1 and @xmath189 by points @xmath188 and @xmath190 on the line @xmath179 such that @xmath198 has the same @xmath158-extent as the rectangle @xmath183 , then @xmath199 .
draw two lines parallel to @xmath197 through @xmath188 and @xmath190 .
let these two lines intersect the @xmath196-plane at points @xmath200 and @xmath201 .
( here we use the fact that if the vertices of a tetrahedron are on three parallel lines , one can shift the single points along the corresponding lines , and the volume remains the same . )
since we moved @xmath188 and @xmath190 along lines parallel to @xmath178 , the volume of @xmath202 is the same as that of @xmath203 , and this volume is @xmath194 ( e.g. , the base triangle of @xmath204 lies in the @xmath196-plane , and its height is @xmath161 ) .
a contradiction .
symmetrically , the plane @xmath205 also contains at most 2 points @xmath1 with @xmath163 , hence the number of minimum volume tetrahedra is at most @xmath206 .
[ [ reporting - minimum - nonzero - volume - tetrahedra . ] ] reporting minimum nonzero volume tetrahedra .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + next , we devise an algorithm for counting and reporting all minimum ( nonzero ) volume tetrahedra spanned by a given set @xmath22 of @xmath0 points in three - space .
a plane @xmath207 and a point @xmath208 form a _ slab _ , which is defined as the open region bounded by @xmath207 and the plane @xmath209 , which is parallel to @xmath207 and incident to @xmath110 .
we say that the slab is _ empty _ , if it is disjoint from @xmath22 .
bra , rote , and swanepoel @xcite observed that if @xmath156 is a minimum area triangle in the plane , then the relative interior of the line segment @xmath93 and the slab formed by the line through @xmath93 and the point @xmath95 is empty of @xmath22 .
this observation readily generalizes to three - space : if @xmath210 is a minimum volume tetrahedron , then the triangle @xmath156 is a minimum area triangle in the plane @xmath211 spanned by @xmath156 , and the open slab formed by @xmath211 and point @xmath1 is empty .
let @xmath212 denote the set of planes spanned by @xmath22 .
we compute for every empty slab formed by a plane @xmath213 and point @xmath110 , the number of tetrahedra whose base is a minimum area triangle in @xmath207 and whose 4th vertex lies in a plane @xmath209 parallel to @xmath207 and incident to @xmath110 ; we also record the volume of these tetrahedra ( which might not be the global minimum ) .
we then sum up these quantities for those empty slabs where this volume is _ the minimum_. we can compute the number of minimum volume tetrahedra , since each of them is counted exactly four times , once for each face . if we record all minimum area triangles in every plane @xmath207 and every point in plane @xmath209 , then we can also _ report _ all minimum volume tetrahedra ( we have already shown that there are only @xmath5 of them ) .
our algorithm works on the dual arrangement @xmath214 .
consider the ( well - known ) duality transform which maps a point @xmath215 to the ( nonvertical ) plane @xmath216 with equation @xmath217 .
conversely , a nonvertical plane @xmath77 with equation @xmath218 is mapped to the point @xmath219 .
duality @xmath220 satisfies @xmath221 , for any point @xmath110 , and @xmath222 , for any nonvertical plane @xmath77 .
the duality preserves point - plane incidences and it reverses the above - below relationship , which is understood with respect the the @xmath223-axis .
furthermore , it preserves vertical distances for point - plane pairs .
more precisely , point @xmath110 is above plane @xmath77 , if and only if point @xmath224 is above plane @xmath216 , and the vertical distance between the point and the plane in each pair is the same .
every plane @xmath207 spanned by the point set @xmath22 corresponds to a vertex @xmath225 of the arrangement .
an empty slab determined by @xmath207 and a point @xmath110 corresponds to a vertical line segment between @xmath225 and plane @xmath216 that does not pierce any plane of the arrangement @xmath214 .
first we solve a planar problem : count the minimum ( nonzero ) area triangles in a plane efficiently .
for a plane @xmath213 , let @xmath226 , let @xmath227 be the number of lines spanned by @xmath228 , and let @xmath229 denote the number of minimum ( nonzero ) area triangles determined by @xmath230 .
let @xmath231 denote the set of lines spanned by @xmath22 .
[ pro : time ] consider a plane @xmath213 .
given the list of @xmath226 points in @xmath207 , and for every line @xmath232 the length and the number of shortest ( that is , _ minimum length _ ) segments in @xmath118 , we can compute @xmath229 in @xmath233 time . for every line @xmath232 , @xmath234 , compute the number of minimum area triangles in the empty slabs formed by @xmath72 ( that is , triangles such that one side is a shortest segment @xmath118 and a third vertex is a closest point to @xmath72 ) .
since the number of shortest segments along @xmath72 is given , it suffices to check the distance of all points of @xmath228 to the line @xmath72 .
we can enumerate these triangles in @xmath235 time . summing this over
all @xmath227 lines , @xmath229 can be computed in @xmath233 time .
we use the szemerdi - trotter theorem ( theorem [ thm : st-83b ] ) to bound the time to compute @xmath229 for all planes @xmath213 .
[ cor : n3 ] with the notation defined above , we have @xmath236 .
denote by @xmath237 the set of lines incident to at least @xmath238 but fewer than @xmath239 points of @xmath22 , for @xmath240 . by corollary [ cor : st-83 ]
, we have @xmath241 we can now count the minimum ( nonzero ) volume tetrahedra spanned by @xmath22 .
assume that the entire dual arrangement of @xmath214 and its incidence graph is available ; it can be computed in @xmath5 time , and stored in @xmath5 space @xcite .
the arrangement is a cell complex with faces of dimension 0 , 1 , 2 , and 3 . for every vertex , we store the incident planes , and the incident ridges ( lines formed by the intersection of two planes ) ; for every ridge , we store the incident vertices ( sorted along the ridge ) and planes of the arrangement ( sorted around the ridge ) .
we perform the following three tasks : ( 1 ) .
preprocess the lines spanned by @xmath22 . for every line @xmath232 ,
compute the number of shortest segments in @xmath118 and record their ( common ) length .
the sorted list of points along each line can be extracted from the arrangement .
since there are at most @xmath242 lines , each containing no more than @xmath0 points , this can be done in @xmath5 time .
( 2 ) for every vertex @xmath225 , compute the number @xmath229 of minimum ( nonzero ) area triangles lying in the plane @xmath207 . by proposition
[ pro : time ] , this can be done in @xmath243 time for a plane @xmath207 . by proposition [ cor :
n3 ] , the total time over all planes in @xmath212 is @xmath244 .
the list of points in each plane @xmath207 is also available from the arrangement .
( 3 ) sweep the dual arrangement @xmath214 with a horizontal plane .
for every vertex @xmath225 , drop a vertical line through @xmath225 and find the face @xmath245 of the cell hit by this line , and the intersection point @xmath158 of the vertical line with @xmath245 : the face @xmath245 can be of dimension 0 , 1 , or 2 .
again , the number @xmath246 of planes in @xmath214 incident to @xmath245 can be extracted from the arrangement . the empty slab formed by @xmath207 and
@xmath247 determines @xmath248 tetrahedra with a minimum area base triangle in @xmath207 and a 4th point in plane @xmath247 . for every vertex @xmath225 of the arrangement , we record @xmath248 and the volume of the corresponding tetrahedra .
if this volume turns out to be the minimum volume , all these @xmath248 tetrahedra can also be reported in a second pass over the data .
since the total number of minimum volume tetrahedra is @xmath5 , all reporting takes @xmath5 time .
if the entire dual arrangement @xmath214 and its incidence graph is available , then our algorithm requires @xmath249 time and @xmath5 space in the real ram model of computation ( where algebraic operations over reals have unit cost ) .
the space requirement can be reduced to @xmath30 if we use a plane sweep algorithm to perform tasks ( 2 ) and ( 3 ) .
it sweeps the arrangement with a horizontal plane and keeps in memory the cells intersecting the sweep plane , whose total complexity is @xmath30 by the zone theorem @xcite .
when the sweep plane arrives at a vertex @xmath225 , it can provide the set of planes and ridges incident to @xmath225 and so we can perform task ( 2 ) .
the plane sweep algorithm takes @xmath250 time ( since it maintains an event queue , which is updated in @xmath251 time for each vertex of the arrangement ) .
the time can be reduced to @xmath5 while maintaining an @xmath30 working space by the _ topological sweep _ method of edelsbrunner and guibas @xcite ( originally designed for line arrangements in the plane , and adapted to three - dimensions by anagnostou _ et al . _
@xcite ) .
edelsbrunner and guibas @xcite and chazelle @xcite gave quadratic time sweep - line algorithms for the problem of finding a triangle of minimum ( possibly zero ) area .
it is easy to rewrite our 3d algorithm for the planar problem and report all triangles of minimum nonzero area determined by a set of @xmath0 points in @xmath30 time and space .
when the sweep line arrives at vertex @xmath252 of the dual arrangement , pairs of lines incident to @xmath252 with consecutive slopes correspond to minimum length segments determined by the points of @xmath22 whose duals intersect at @xmath252 . by drawing vertical segments at each vertex of the arrangement , the algorithm examines all triangles whose base is a minimum length segment on some line of @xmath231 , and the third vertex is closest to the line through the base .
[ [ generalization - to - k - simplices - in - d - space ] ] generalization to @xmath8-simplices in @xmath1-space + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + theorem [ t - min ] can be generalized to arbitrary dimensions .
[ d - min ] for every ( fixed ) @xmath34 with @xmath35 , the number of @xmath8-dimensional simplices of minimum ( nonzero ) volume spanned by @xmath0 points in @xmath9 is @xmath36 , and there are point sets for which this number is @xmath38 .
* lower bound . *
consider @xmath8 parallel lines such that any two are at unit distance apart from each other ; place @xmath253 equally spaced points on each line ( assume @xmath0 is a multiple of @xmath8 ) .
every @xmath8-simplex with positive volume has two vertices on one line and one vertex on each of the other lines .
if the @xmath8-simplex has minimum volume , the vertices along the same line must be consecutive .
conversely , every @xmath8-simplex with two consecutive vertices on one line and one arbitrary point on each of the other two lines has the same ( minimum ) volume .
the number of such @xmath8-simplices is @xmath254 .
[ [ upper - bound.-1 ] ] upper bound .
+ + + + + + + + + + + + let @xmath22 be a set of @xmath0 points in @xmath9 .
assign every ( nondegenerate ) @xmath8-simplex @xmath255 determined by @xmath22 to one of its @xmath37 faces as follows .
we choose the @xmath8 vertices of a face of @xmath255 in @xmath8 steps .
first choose an arbitrary vertex @xmath256 of @xmath255 .
in step @xmath257 , choose a vertex @xmath258 of @xmath255 which is at maximum ( euclidean ) distance from the @xmath259-dimensional affine subspace @xmath260 spanned by @xmath261 . we show that at most a constant number of minimum volume @xmath8-simplices are assigned to every @xmath262-simplex spanned by @xmath22 .
this yields an @xmath36 upper bound .
assume that the minimum volume of a @xmath8-simplex is 1 .
consider a ( nondegenerate ) @xmath262-simplex @xmath263 such that @xmath258 , @xmath264 , is a furthest point in @xmath265 from the affine subspace @xmath260 .
choose an orthogonal coordinate system such that @xmath256 is the origin and @xmath266 , @xmath257 , is the @xmath267-dimensional subspace spanned by the first @xmath268 coordinate axes .
denote the @xmath269 extents of the axis - aligned bounding box of @xmath270 by @xmath271 ; and choose @xmath272 so that @xmath273 .
so @xmath274 is the volume of the bounding box of any unit volume @xmath8-simplex with a @xmath262-dimensional face @xmath270 in the coordinate system described . by the choice of the points @xmath275
, we have @xmath276 . if our charging scheme assigns a minimum volume @xmath8-simplex @xmath277 to @xmath270 , then @xmath278 , since we chose @xmath279 such that it is at least as far from @xmath280 as @xmath143 .
every point @xmath143 such that @xmath277 is a @xmath8-simplex of a given volume lies in a hypersurface @xmath281 , which is the minkowski sum @xmath282 of the @xmath262-dimensional euclidean space @xmath283 spanned by @xmath270 ( in the first @xmath269 coordinates ) and a sphere @xmath284 of radius @xmath272 ( in the remaining @xmath285 coordinates ) .
every point @xmath286 has a unique decomposition into the vector sum @xmath287 with @xmath288 and @xmath289 .
if the @xmath8-simplex @xmath277 , @xmath290 , is assigned to @xmath270 , then @xmath143 must satisfy @xmath269 constraints : @xmath143 is at most as far from the subspace @xmath260 as @xmath258 for @xmath291 .
these constraints are satisfied if the @xmath292-th coordinate of @xmath293 is at most @xmath294 , @xmath295 .
that is , @xmath293 lies in an axis - aligned rectangle @xmath296 of extents @xmath297 , for @xmath298 . partition @xmath169 into @xmath299 congruent axis - aligned rectangles of extents @xmath300 , each for @xmath301 .
partition @xmath302 into @xmath303 convex regions of diameter @xmath304 each .
these two partitions give a partition of the minkowski sum @xmath305 into @xmath306 regions .
next we show that each region contains at most one point of @xmath22 .
this confirms that at most @xmath307 minimum @xmath8-simplices are assigned to @xmath270 , completing our proof .
suppose , by contradiction , that @xmath308 are assigned to @xmath270 and both lie in the same region of @xmath305 .
since @xmath309 , both @xmath310 and @xmath311 are @xmath8-simplices of minimum volume .
we show that @xmath312 spans a @xmath8-simplex of a strictly smaller positive volume , which is a contradiction .
consider the maximum index @xmath313 for which the set @xmath314 is a nondegenerate @xmath315-simplex .
( such an index @xmath313 exits , since @xmath316 is a noncollinear triple . )
we show that the @xmath8-simplex @xmath317 has smaller volume than @xmath277 .
note that it is enough to compare the volumes of the @xmath315-simplices @xmath318 and @xmath319 . to compare the volumes of @xmath320 and @xmath321 ,
it is enough to compare the distances of @xmath322 and @xmath95 to the affine subspace @xmath323 spanned by @xmath324 , that is , @xmath325 and @xmath326 .
if we shift the line segment @xmath327 in an affine subspace parallel to @xmath328 , then the volumes of @xmath320 and @xmath321 do not change , so we may assume that segment @xmath329 is orthogonal to @xmath328 .
then , we have @xmath330 .
the first @xmath331 coordinates of the vector @xmath332 are parallel to @xmath333 , and so they can be ignored when computing @xmath326 .
each of the coordinates @xmath334 of @xmath332 is at most a @xmath335 fraction of the corresponding coordinate of @xmath322 ; the @xmath8-th coordinate of @xmath336 is @xmath337 , and each of the last @xmath338 coordinates of @xmath332 is at most @xmath304 .
hence @xmath339 this confirms that @xmath95 is strictly closer to @xmath323 than @xmath322 , hence @xmath340 .
[ [ lower - bound.-1 ] ] lower bound .
+ + + + + + + + + + + + as mentioned in the introduction , erds and purdy showed that a suitable section of the integer lattice has @xmath18 unit - area triangles @xcite . this immediately gives a lower bound of @xmath341 unit volume tetrahedra ( e.g. , by placing two such lattice sections , with @xmath61 points each , in two parallel planes ) .
[ [ upper - bound.-2 ] ] upper bound .
+ + + + + + + + + + + + pach and sharir @xcite showed that every point @xmath142 of a set @xmath22 of @xmath0 points in the plane is incident to at most @xmath342 unit area triangles spanned by @xmath22 .
this bound follows easily from the szemerdi - trotter bound on the number point - line incidences : for every point @xmath343 , there is a unique line @xmath344 parallel to @xmath93 such that every point @xmath345 above @xmath344 gives a unit area triangle @xmath156 ; there are @xmath342 point line incidences between @xmath22 and the lines @xmath344 .
this bound is tight , since the szemerdi - trotter bound is tight : there are point sets with @xmath346 unit triangles having a common vertex : place a set @xmath347 of @xmath61 points and a set @xmath126 of @xmath61 lines with @xmath346 incidences in the plane ; @xmath126 determines a point set @xmath348 with @xmath349 , and so @xmath350 spans @xmath346 unit area triangles incident to @xmath142 .
we follow a similar strategy in three dimensions . instead of bounding directly the number of unit - volume tetrahedra with a common vertex or a common edge , we design a charging scheme .
we assign every unit volume tetrahedron to a line segment lying in the plane containing one of its four faces .
ideally , a tetrahedron is assigned to one of its edges , but this does not always hold in our charging scheme . we show that at most @xmath351 tetrahedra are assigned to every segment ; which immediately gives an @xmath11 bound on the number of unit volume tetrahedra .
[ thm : unit+ ] the number of unit - volume tetrahedra determined by @xmath0 points in @xmath2 is @xmath11 . let @xmath22 be a set of @xmath0 points in @xmath2 , and @xmath212 denote the set of all planes spanned by @xmath22 .
we follow the convention that in a tetrahedron @xmath352 , vertex @xmath1 lies above the plane @xmath211 spanned by @xmath156 . for every triangle @xmath156 , every vertex @xmath1 for which @xmath210 has
unit volume lies in a plane @xmath353 parallel to plane @xmath211 .
we decompose the set @xmath354 of unit volume tetrahedra into two subsets : * @xmath355 contains every @xmath356 where @xmath353 is nondegenerate ; * @xmath357 contains every @xmath356 where @xmath353 is degenerate . *
consider @xmath355 .
* for two distinct points @xmath358 , let @xmath359 denote the set of planes spanned by @xmath22 containing the segment @xmath360 .
[ [ assigning - triangles - to - segments . ] ] assigning triangles to segments .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + we design a charging scheme where we assign every triangle @xmath159 to a segment in the plane @xmath211 .
this induces a charging scheme for unit volume tetrahedra : if @xmath156 is assigned to segment @xmath360 , then we assign every unit volume tetrahedron @xmath210 to segment @xmath360 , too
. the assignment is done in each plane @xmath361 independently .
consider a plane @xmath361 .
we proceed in two stages : in the first stage we compute , for every triangle @xmath156 , a collection @xmath362 of segments in @xmath211 ; in the second stage , we assign @xmath156 to one of the segments in @xmath362 .
the assignment is fairly elaborate ; intuitively , we pursue two goals : ( 1 ) every segment @xmath360 should be assigned to `` few '' triangles of the same area , and ( 2 ) a segment @xmath363 should _ not _ be assigned to a triangle @xmath156 if the plane @xmath353 contains a `` rich '' line parallel to @xmath360 .
we continue with the details .
denote by @xmath364 the number of points in @xmath54 , and by @xmath365 the number of triangles spanned by @xmath56 .
corollary [ cor : dir ] tells us that there is a set @xmath366 of @xmath367 points , for some absolute constant @xmath103 , such that each @xmath368 is incident to a set @xmath108 of @xmath369 line segments determined by @xmath56 , and the line segments in @xmath108 , @xmath368 , are all distinct . denote the collection of all these segments by @xmath370 .
we have @xmath371 . for every triangle @xmath372 ,
let @xmath373 denote the set of segments in @xmath54 parallel to any one of the @xmath374 richest lines in the plane @xmath353 .
let @xmath375 be the set of segments to which @xmath156 may be assigned .
since @xmath376 for every @xmath368 , we have @xmath377 . for two points @xmath378 ,
let @xmath379 denote the number of points in @xmath56 that do not lie on the line through @xmath360 .
if @xmath380 , then one endpoint of @xmath360 is in @xmath381 , and by corollary [ cor : dir ] , we have @xmath382 .
hence , we have @xmath383 . by corollary [ cor : unit+ ]
, @xmath56 determines at most @xmath384 triangles of any given area , with ( say ) @xmath385 .
we are now ready to assign triangles to segments : assign every triangle @xmath156 determined by @xmath386 to a segment @xmath387 such that ( i ) every segment @xmath380 is assigned to @xmath388 triangles and ( ii ) every segment @xmath380 is assigned to @xmath389 triangles of any given area . [
[ multiplicities . ] ] multiplicities .
+ + + + + + + + + + + + + + + let @xmath390 denote the set of triangles @xmath372 for which @xmath353 is nondegenerate and @xmath156 is assigned to segment @xmath360 .
we say that the _ multiplicity _ @xmath391 of a triangle @xmath392 is the number of triangles in @xmath390 with the same area as that of @xmath156 .
for every @xmath393 , let @xmath394 denote the set of triangles @xmath395 for which @xmath396 .
since @xmath360 is assigned to at most @xmath397 triangles of any given area , the multiplicity of no triangle can exceed @xmath398 , for a sufficiently large constant @xmath399 ; that is , @xmath400 for @xmath401 .
aggregate the triangles assigned to the same segment @xmath360 in various planes @xmath402 by letting @xmath403 [ [ projection - to - a - plane . ] ] projection to a plane .
+ + + + + + + + + + + + + + + + + + + + + + fix a line segment @xmath360 .
project @xmath22 along lines parallel to @xmath360 onto a plane @xmath404 orthogonal to @xmath360 .
the image of @xmath22 under the projection is a multiset @xmath405 , where the multiplicity of each point @xmath406 is the number of points of @xmath22 on a line through @xmath407 parallel to @xmath360 .
similarly , the projection of a subset @xmath105 is a multiset @xmath408 .
the projection of the line through @xmath360 is a single point @xmath409 , and every plane @xmath54 parallel to @xmath360 is projected into a line @xmath410 in @xmath404 .
an illustration is provided in fig .
[ fig : projection ] .
sort the points of @xmath22 into @xmath411 buckets according to the multiplicities of their projections : for @xmath412 , let @xmath413 denote the set of points @xmath120 whose projection has multiplicity at least @xmath414 . for every plane @xmath54 parallel to @xmath360
, we have @xmath415 [ [ partition - of - planes - babc - according - to - four - parameters . ] ] partition of planes @xmath353 according to four parameters .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + we partition the nondegenerate planes parallel to @xmath360 into @xmath416 equivalence classes .
we will bound the number of unit volume tetrahedra assigned to @xmath360 corresponding to each equivalence class separately . for four parameters , @xmath417 , @xmath418 , @xmath419 , and @xmath420 , we define the set @xmath421 of nondegenerate planes @xmath422 parallel to @xmath360 such that there is a unit volume tetrahedra @xmath210 assigned to @xmath360 with @xmath423 and we have @xmath424 , @xmath425 , and @xmath426 .
we first give an upper bound on the number of planes in @xmath421 .
there are at most @xmath427 distinct planes @xmath402 with @xmath428 .
we have assigned at most @xmath429 triangles lying in plane @xmath54 to @xmath360 by property ( i ) of the assignment .
so there are at most @xmath430 different triangle areas which are each represented by at most @xmath431 triangles in @xmath54 assigned to @xmath360 .
triangles @xmath372 of distinct areas correspond to distinct planes @xmath353 parallel to @xmath432 .
hence , each such plane @xmath54 determines at most @xmath430 parallel planes @xmath353 in @xmath421 , and so @xmath433 .
we deduce several relation among the parameters @xmath292 , @xmath315 , @xmath77 , and @xmath434 , assuming @xmath435 . recall that we assigned @xmath436 triangles of the same area to @xmath360 in every plane @xmath361 with @xmath437 and @xmath438 , by property ( ii ) of the assignment , and so @xmath439 next , we deduce an upper bound on @xmath292 in terms of @xmath77 and @xmath434 . consider a plane @xmath440 corresponding to a triangle @xmath441 in a plane @xmath402 . by the choice of @xmath77 and @xmath434 , at least @xmath442 points of @xmath443 lie on @xmath414-rich lines parallel to @xmath360 .
since we assigned @xmath156 to segment @xmath360 only if @xmath360 is not parallel to the @xmath444 richest lines in @xmath422 , and @xmath422 contains several @xmath414-rich lines parallel to @xmath360 , we know that an @xmath414-rich line is not among the @xmath445 richest lines of @xmath422 . by theorem [ thm : st-83b ] , there are at most @xmath446 distinct @xmath414-rich lines in the plane @xmath422 that contains at most @xmath447 points .
hence , the number of @xmath414-rich lines in any @xmath440 is more than @xmath448 .
we have @xmath449 , that is , @xmath450 [ [ counting - incidences - in - the - plane . ] ] counting incidences in the plane .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + next , we estimate the number of incidences of the projection points @xmath451 and the lines of @xmath452 .
since @xmath453 and each line spanned by @xmath454 contains at least two points of @xmath451 , corollary [ cor : st-83 ] gives @xmath455 this bound is too weak if @xmath456 , and so we give another bound for @xmath457 . since every @xmath458 is an @xmath459-rich line in the planar point set @xmath451 , we can apply corollary [ cor : st-83b ] and we obtain @xmath460 .
the combination of these bounds gives that @xmath461 in every @xmath462 , every point of @xmath443 corresponds to @xmath463 unit - volume tetrahedra assigned to @xmath360 .
the number of points in @xmath422 is latexmath:[$|s\cap b| = o(\sum_{r=1}^{\log n } observed among the parameters @xmath292 , @xmath315 , @xmath77 , and @xmath434 , we can sum these quantities for all @xmath292 , @xmath315 , @xmath77 , and @xmath434 , and get @xmath465 summing over all segments @xmath360 , we obtain @xmath466 [ [ consider - u_2 . ] ] consider @xmath357 .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + for every line @xmath72 spanned by @xmath22 , consider all planes @xmath211 spanned by some triangle @xmath88 for which the plane @xmath353 is degenerate and its principal line is @xmath72 .
project @xmath22 onto a plane @xmath467 orthogonal to @xmath72 .
decompose the point set @xmath22 into @xmath411 buckets as described above : for @xmath468 , the bucket @xmath451 contains all points in @xmath467 whose multiplicity is at least @xmath414 . for a given plane @xmath54 , observe that the number of @xmath414-rich lines in @xmath54 that are parallel to @xmath72 is @xmath469 .
we know that @xmath470 .
it follows by jensen s inequality that @xmath471 for two indices @xmath472 and @xmath473 , let @xmath474 denote the set of planes @xmath361 such that @xmath54 is parallel to @xmath72 , it is incident to at least two points of @xmath475 , and @xmath476 .
the projection @xmath410 of a plane @xmath477 is a line incident to at least @xmath478 points in bucket @xmath451 .
we clearly have @xmath479 , and so the number of such lines can be bounded by theorem [ thm : st-83b ] : @xmath480 for a fixed line @xmath72 , denote by @xmath481 the set of planes spanned by @xmath22 and containing @xmath72 . by the fact that the number of unit area triangles in a plane @xmath482 is @xmath483 , the number of triangles @xmath156 whose degenerate plane @xmath353 contains @xmath72 is @xmath484 by theorem [ thm : st-83b ] , for every @xmath485 there are at most @xmath486 distinct @xmath281-rich lines .
moreover , the total number of incidences on these lines is at most @xmath487 .
therefore , the number of unit volume tetrahedra @xmath488 whose plane @xmath353 is degenerate and its principal line @xmath72 is @xmath281-rich is bounded by @xmath489 . on the other hand , the number of unit tetrahedra @xmath490 whose plane @xmath353 is degenerate and its principal line @xmath72 contains at most @xmath281 points , and so @xmath491 , is at most @xmath492 .
we balance the two upper bounds by choosing @xmath493 , and then we have @xmath494 . recalling that @xmath67 ,
we have proved that @xmath495
in 1982 , erds , purdy , and straus @xcite considered the analogue of the problem of distinct triangle areas in the plane to higher dimensions and posed the following problem : let @xmath22 be a set of @xmath0 points in @xmath9 not all in one hyperplane . what is the minimal number @xmath23 of distinct volumes of full - dimensional simplices with vertices in @xmath22 ? by taking @xmath1 sets of about @xmath25 equally spaced points on parallel lines through the vertices of a @xmath26-simplex , one gets @xmath496 .
erds , purdy , and straus conjectured that equality holds at least for sufficiently large @xmath0 ( see also @xcite ) . in this section ,
we give a first linear lower bound on the number of full - dimensional simplices in @xmath9 ( including tetrahedra in 3-space ) determined by @xmath0 points that span @xmath9 .
some remarks are in order : we have mentioned earlier erds s distinct distance problem that asks for the minimum number @xmath497 of distinct inter - point distances among @xmath0 points in the plane .
the currently known best lower bound for @xmath497 , due to katz and tardos @xcite , in fact gives a lower bound on the maximum number of interpoint distances measured _ from a single point_. for triangle areas in the plane , we have a similar phenomenon .
pinchasi proved that every set @xmath22 of @xmath0 noncollinear points in the plane contains a point pair @xmath498 such that the points of @xmath22 determine @xmath27 distinct ( nonzero ) distances measured from the line @xmath93 , and so there are at least @xmath499 triangles with distinct ( positive ) areas @xcite .
our proof follows a similar path .
we show that @xmath0 points in @xmath1-space , not all on a hyperplane , determine @xmath40 full - dimensional simplices with distinct volumes that share a common face .
we start with a useful lemma about triangle areas in the plane : [ lem : origin ] let @xmath22 be a set of @xmath0 points in the plane and let @xmath500 be a point such that each line @xmath501 , @xmath502 , is incident to no other points in @xmath22 .
then there is a point @xmath503 such that the triangles @xmath504 with @xmath505 determine @xmath40 distinct areas .
let @xmath506 be the smallest integer such that for every @xmath503 , the triangles @xmath22 with a common side @xmath507 have at most @xmath104 distinct areas . for every fixed @xmath508 , al points @xmath509 where the area of the triangle @xmath510 is one of the at most @xmath104 possible areas must lie on one of at most @xmath511 lines parallel to @xmath507 .
hence for every @xmath512 , all the remaining @xmath513 points of @xmath22 must lie on one of at most @xmath511 lines .
since the directions of the lines @xmath507 , @xmath514 , are all distinct , we obtain a total of @xmath515 distinct lines ( the lines @xmath507 are also counted here , accounting for collinear triples @xmath516 ) .
consequently there are @xmath517 point - line incidences between @xmath22 and these @xmath47 distinct lines .
[ thm : distinct ] the tetrahedra determined by a set @xmath22 of @xmath0 points in @xmath2 , not all on a plane , have at least @xmath40 distinct volumes .
furthermore , @xmath22 spans a triangle @xmath270 such that @xmath22 determines @xmath40 tetrahedra of distinct volumes , having a common face @xmath270 .
let @xmath22 be a set of @xmath0 points in @xmath2 , not all on a plane .
consider the two cases from lemma [ lem : b ] .
first assume that there are @xmath521 points in a plane @xmath54 . since @xmath22 is not contained in a plane
, we may assume further that @xmath22 spans the plane @xmath54 ( not all points of @xmath56 are collinear ) . by the planar result mentioned above , the points @xmath56 span at least @xmath522 triangles with distinct areas and with a common side @xmath523 .
let @xmath524 .
the triangles in @xmath54 together with @xmath434 determine @xmath40 tetrahedra of distinct volumes with a common face @xmath525 . next
, assume that the triples of @xmath22 span at least @xmath526 distinct planes .
there is a segment @xmath527 that participates in at least @xmath528 such triples .
let @xmath72 be the line through @xmath527 ; and let @xmath212 be the set of planes determined by @xmath22 and containing @xmath527 .
let @xmath529 , @xmath530 , be a set of points that contains exactly one point from each @xmath531 , @xmath361 .
project the point set @xmath529 onto a plane @xmath404 orthogonal to @xmath72 .
refer to fig .
[ fig : distinct ] .
let @xmath532 denote the projection points ( and the projection of @xmath72 onto @xmath404 is a point @xmath533 ) . by construction
, @xmath534 contains @xmath40 points such that each line @xmath535 , @xmath536 , is incident to a unique point of @xmath537 .
by lemma [ lem : origin ] , there is a point @xmath538 such that the triangles @xmath539 with @xmath540 determine @xmath40 distinct areas in the plane @xmath404 . for every triangle @xmath541
, the volume of the tetrahedron @xmath542 can be expressed as ( here we use again the fact that if the vertices of a tetrahedron are on three parallel lines , one can shift the single points along the corresponding lines while the volume remains the same ) : @xmath543 hence @xmath22 contains @xmath40 tetrahedra of distinct volumes sharing a common face @xmath544 .
[ thm : ddistinct ] for every @xmath13 , the full - dimensional simplices determined by a set @xmath22 of @xmath0 points in @xmath9 , not all on a hyperplane , have at least @xmath40 distinct volumes .
furthermore , @xmath22 spans a @xmath26-dimensional simplex @xmath270 such that @xmath22 determines @xmath40 full - dimensional simplices of distinct volumes , having a common face @xmath270 .
we proceed by induction on @xmath13 .
the theorem is obvious for @xmath545 and it was confirmed by burton and purdy for @xmath546 @xcite .
let @xmath28 , and assume that for every @xmath547 , there is a constant @xmath548 such that any @xmath549-element point set in @xmath550 , not all in a @xmath551-dimensional affine subspace , determines at least @xmath552 @xmath188-dimensional simplices of distinct volumes having a common @xmath551-dimensional face .
let @xmath22 be a set of @xmath0 points in @xmath9 , not all on a hyperplane .
consider the two cases from lemma [ lem : b ] : first assume that there are @xmath137 points in a hyperplane .
these points span a @xmath188-dimensional subspace @xmath54 , for some @xmath553 . by the induction hypothesis ,
the points @xmath56 determine @xmath554 @xmath188-dimensional simplices of distinct volumes having a common @xmath551-dimensional face .
since @xmath22 is not contained in a hyperplane , there is a set @xmath555 of @xmath556 points such that @xmath557 and @xmath54 together span the entire space @xmath9 . adding the vertices @xmath557 to each of the @xmath40 @xmath188-dimensional simplices of distinct volumes in @xmath54
, we obtain @xmath40 full - dimensional simplices of distinct volumes having a common face in @xmath9 . next ,
assume that the @xmath1-tuples of @xmath22 span at least @xmath138 distinct hyperplanes .
there is a @xmath26-tuple @xmath558 that participates in at least @xmath559 such @xmath1-tuples spanning distinct hyperplanes .
let @xmath54 be the @xmath560-dimensional affine subspace spanned by @xmath270 ; and let @xmath231 be the set of @xmath26-dimensional affine subspaces determined by @xmath22 and containing @xmath270 .
let @xmath529 , @xmath561 , be a set that contains exactly one point from @xmath562 , for each @xmath563 .
project the point set @xmath529 onto a plane @xmath404 orthogonal to @xmath54 .
refer to fig .
[ fig : distinct ] .
let @xmath532 denote the projection points ( and the projection of @xmath54 onto @xmath404 is a point @xmath533 ) . by construction
, @xmath534 contains @xmath40 points such that each line @xmath535 , @xmath536 , is incident to a unique point of @xmath537 . by lemma [ lem : origin ]
, there is a point @xmath538 such that the triangles @xmath539 with @xmath540 determine @xmath40 distinct areas in the plane @xmath404 .
for every triangle @xmath541 , the volume of the full - dimensional simplex @xmath564 spanned by @xmath565 can be expressed as : @xmath566 hence @xmath22 contains @xmath40 full - dimensional simplices of distinct volumes sharing a common face @xmath567 . our proof for theorem [ thm :
ddistinct ] crucially relies on the fact that we build _ full - dimensional _ simplices of distinct volumes ( i.e. , @xmath568 ) .
no tight bound is known for the order of magnitude of the minimum number of @xmath8-dimensional simplices of distinct volumes determined by @xmath0 points in @xmath9 , not all in a hyperplane , for any @xmath569 .
in particular , for @xmath570 this is erds s celebrated distinct distance problem in @xmath9 we mentioned earlier .
h. edelsbrunner and l. j. guibas , topologically sweeping an arrangement ,
_ j. comput .
* 38 * ( 1989 ) , 165194 .
corrigendum : topologically sweeping an arrangement , _ j. comput
_ * 42 * ( 1991 ) , 249251 . | we formulate and give partial answers to several combinatorial problems on volumes of simplices determined by @xmath0 points in 3-space , and in general in @xmath1 dimensions .
( i ) the number of tetrahedra of minimum ( nonzero ) volume spanned by @xmath0 points in @xmath2 is at most @xmath3 , and there are point sets for which this number is @xmath4
. we also present an @xmath5 time algorithm for reporting all tetrahedra of minimum nonzero volume , and thereby extend an algorithm of edelsbrunner , orourke , and seidel .
in general , for every @xmath6 , @xmath7 , the maximum number of @xmath8-dimensional simplices of minimum ( nonzero ) volume spanned by @xmath0 points in @xmath9 is @xmath10 .
( ii ) the number of unit - volume tetrahedra determined by @xmath0 points in @xmath2 is @xmath11 , and there are point sets for which this number is @xmath12 .
( iii ) for every @xmath13 , the minimum number of distinct volumes of all full - dimensional simplices determined by @xmath0 points in @xmath9 , not all on a hyperplane , is @xmath14 . |
accurate inferences of solar meridional flow are important for modelling the solar dynamo ( e.g. , @xcite , @xcite ) . at or near the surface
, the magnitude of the meridional flow is known to a good extent ( e.g. , @xcite , @xcite ) from a variety of methods such as doppler - shifts and feature tracking ( e.g. , @xcite ) , time - distance helioseismology ( e.g. , @xcite , ) ring - diagram analysis ( e.g. , @xcite ) , and fourier - hankel analysis @xcite .
the nature of meridional flows in deeper layers below about 0.9 solar radii still remains under debate .
such measurements were obtained using tracking of supergranules @xcite and helioseismologic techniques , such as time - distance helioseismology ( @xcite ; @xcite ) , fourier - hankel analysis @xcite , and global helioseismology @xcite .
the conclusions on the global pattern of the meridional flow are very different , ranging from multiple cells in depth @xcite and multiple cells in depth and latitude @xcite to a single - cell picture with a return flow starting in rather shallow layers ( @xcite ; @xcite , at about 0.9 solar radii ) or in deeper regions ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* below 0.85 solar radii ) .
however , there are several details to keep in mind concerning these inversion results @xcite , e.g. , that uncertainties in the results due to systematic effects are likely to be larger than the random errors in the inversion results , that the results were obtained using different instruments , and that they cover different periods in time .
furthermore , the impact of the radial flow component on the travel times was not taken into account in the results obtained by @xcite and @xcite . in time - distance helioseismology
@xcite , flows can be inferred using sensitivity functions ( kernels ) , which are a model for the impact of the flows on travel - time measurements of acoustic waves .
measurements of deep meridional flow have so far been done using the rather classical ray approximation ( @xcite ; @xcite ) . in this model , travel times are assumed to be sensitive to flows only along an infinitely thin ray path which connects the two observation points .
recently , @xcite extended a born approximation model for the travel - time measurements from cartesian to spherical geometry .
an alternative approach for computing born kernels was proposed very recently by @xcite .
these developments permit the use of born approximation kernels for inferring the deep meridional flow . in the born approximation ( e.g. , @xcite , @xcite )
, the full wave field in the solar interior is modelled using a damped wave equation which is stochastically excited by convection .
this wave equation is solved in zero - order and in its first - order perturbation , which includes advection in the presence of a flow field . when modelling travel times using the born approximation , the advection and first - order scattering of the wave field at any location inside the sun is thereby taken into account .
the ray approximation is expected to be accurate if the underlying flow field does not vary at length scales which are smaller than a wavelength ( e.g. , * ? ? ?
* ) . in the case of flows at the bottom of the convection zone
, this scale was estimated to be of the order of @xmath1 ( @xcite and references therein ) .
if the flow varies on smaller length scales , the born approximation is thought to be more accurate ( e.g. , @xcite , @xcite , @xcite , and @xcite ) .
in addition to modelling the perturbation to the full wave field in the solar interior , an advantage of the born approximation is that it also provides a model for additional observational quantities , such as disc - averaged cross - covariances and mean power spectra , which is not the case for the ray approximation .
the accuracy of the model can therefore easily be validated .
several methods for inferring the deep solar meridional flow @xcite have been validated using a linear numerical simulation of the solar interior wave field by @xcite , which includes a standard single - cell meridional flow profile .
the same simulation is to be used in this paper in order to validate the use of spherical born kernels for inferring solar meridional flows . in section [ secborn ]
, we will give a short introduction to the born approximation as used in time - distance helioseismology .
section [ secfast ] includes the details of a numerical optimization , which was necessary in order to obtain the required number of kernels in an acceptable amount of time .
the main comparison of the born approximation model with artificial data obtained from the simulation is presented in section [ secfwdcomp ] . in section [ secradial ]
, we discuss the relevance of the radial flow component on the measured travel times .
furthermore , we show inversion results of this data obtained with the born kernels in section [ secinversions ] .
conclusions are presented in section [ secconclusions ] .
we first briefly summarize the born approximation model introduced by @xcite , which is to be validated in the following sections . following @xcite ,
the measurement process in time - distance helioseismology is modelled as closely to observations as possible .
the objective of the model is to give a linear relationship between a small - magnitude flow field @xmath2 in the solar interior , which perturbs a mean observed or background model travel time , @xmath3 , according to @xmath4 = \int_\sun { { \bf k}}({{\mathbf r } } ) { { \ , \bm \cdot \,}}{{\bf v}}({{\mathbf r } } ) \ , { { \mathrm d}}^3 { { \mathbf r } } , \label{eqkernelgoal}\ ] ] where @xmath5 denotes the expectation value of a stochastic quantity .
vector quantities are printed in bold throughout this work . once an expression for the sensitivity function @xmath6 is found and travel times are measured , equation can be used to infer solar interior flows . in order to achieve this goal , a model is first developed for an unperturbed spherically symmetric non - rotating sun ( zero - order problem ) .
the stochastically driven and damped wave equation for solar oscillations is solved by expansion into solar eigenmodes .
the resulting wave field @xmath7 at location @xmath8 and time @xmath9 can be used to model the observed doppler signal @xmath10 via @xmath11 , where the line - of - sight is assumed to be radial .
after a spherical harmonic transform @xmath12 and a fourier transform ( indicating the fourier transform by the use of the variable @xmath13 instead of @xmath9 ) , a power spectrum can be computed both from observations and from the model , @xmath14 where the filtered @xmath15 were obtained by multiplying @xmath16 by a filter function @xmath17 . after reconstructing the filtered dopplergram time series , @xmath18 , @xmath19 cross - covariances
are computed by @xmath20 which are used for fitting travel - times from observations , see
@xcite , ( * ? ? ?
* hereafter gb04 ) , and @xcite .
as our goal is to measure flows , we use travel - time differences @xmath21 , and the terms travel - time and travel - time difference are used interchangeably in this work . in order to model the effect of a flow field on the travel times
, the zero - order model for the mean sun is perturbed .
the flow is assumed to be small and thus only linear effects are taken into account .
the perturbation of the flow is introduced in the wave equation by adding an advection term .
this first - order wave equation is solved for the perturbation to the wave field , @xmath22 . taking only first - order terms ( i.e. linear terms ) into account
, this corresponds to modelling the first - order scattering of the modes in the solar interior due to the flow field , see @xcite .
the perturbation to the wave field is then used to obtain the perturbation to the cross - covariance and the travel - time shift according to equation .
the result is @xmath23 where the last term is identical to the previous term , apart from complex conjugation and exchange of indices 1 and 2 . in equation ,
the sum is taken over all pairs of eigenmodes @xmath24 and the quantities @xmath25 and @xmath26 describe the scattering of mode @xmath27 into mode @xmath28 due to a flow at location @xmath29 , which is propagated to the observation points at the surface .
each mode is identified by its harmonic degree @xmath30 and radial order @xmath31 .
see appendix [ appendixoptimization ] for a recap of the definitions of @xmath25 and @xmath26 .
as the numerical evaluation of equation is quite costly for obtaining an accurate kernel ( about 2 days on 32 cpu cores for a three - dimensional kernel covering only about 5 % of the solar volume with a coarse spatial grid ) , it is necessary to optimize its computation .
this requires a little more mathematical detail , which the reader may well skip in order to understand the main results of this paper .
we here propose an approach in which we make use of the separability of the eigenfunctions into a horizontal and a radial dependence .
this is a consequence of the separation of variables performed when solving the oscillation equations ( e.g. , @xcite ) .
the horizontal dependence of the eigenfunction is then also separable from the radial order of the mode .
this allows us to rewrite the kernel formula ( see appendix [ appendixoptimization ] for details ) @xmath32 here , the quantity @xmath33 includes the horizontal dependence of the kernel from @xmath26 .
the quantity @xmath34 includes the sum over @xmath31 and @xmath35 , the factor @xmath25 as well as the @xmath36-dependent factors and the radial dependence from @xmath26 in equation .
if @xmath34 is evaluated before the main loop over @xmath30 , @xmath37 , and the spatial grid in equation , the computational burden of the sum over @xmath31 and @xmath35 does not play a considerable role in the total computation time anymore . in practice , for a given @xmath30 , the number of radial orders to be summed over is about 10 for probing the deep solar interior .
the approach presented here thus decreases the computation time by about two orders of magnitude . computing a full 3d sensitivity kernel using harmonic degrees @xmath38 and a spatial grid with @xmath39 grid points ( @xmath40 , covering depths until @xmath41 and the complete horizontal domain ) , therefore , takes about 1 day on 8 cpus .
a complete set consists of one kernel per travel distance , which can be reprojected to different latitudes and used to obtain , e.g. , kernels for point - to - arc travel times , see examples presented in figure [ figkernels ] .
computing such a set with 126 kernels as used in sections [ secfwdcomp ] - [ secinversions ] takes about 3 weeks on 96 cpus .
very recently , @xcite developed an alternative approach for computing born kernels .
this approach is based on numerically solving for the green s functions for individual modes and individual frequencies .
it allows for a rather flexible computation of kernels for different quantities such as sound - speed , density , flows , and damping properties , as well as a possible inclusion of axisymmetric perturbations to the model . while a detailed comparison of both methods is left to further studies , we consider as an example the computation of the radial solar model case from table 2 in @xcite using the frequency resolution from this study ( about 92,000 grid points spread over a range of @xmath42 , see also @xcite ) and 149 modes
this would thus take about eight times longer than using our method .
in general , the computation time for our method scales with the number of modes squared , and the method of @xcite scales with the number of modes times the number of frequency bins used . as a very fine frequency resolution is necessary to adequately sample the frequency domain in the case of deeply - penetrating modes @xcite , our method is thus advantageous for computing kernels for filtered observations of deep flows , and theirs for computing kernels using many modes or a coarser frequency resolution .
furthermore , it is possible to reach an even greater optimization when computing individual kernels which were integrated over the azimuthal coordinate @xmath43 . for inferring flow fields which are rotationally symmetric such as meridional flow or differential rotation
, one can rewrite equation as @xmath44 & = \iint { { \bf v}}(r,\theta ) \cdot \vec{\mathbb{k}}({{\mathbf r}}_1,{{\mathbf r}}_2;r,\theta ) \ ,
r \ , { { \mathrm d}}\theta \ , { { \mathrm d}}r , \label{eqintkernelgoal}\end{aligned}\ ] ] where the integrated kernel @xmath45 is defined as @xmath46 we thus have @xmath47 where @xmath48 is the integral over the @xmath43-coordinate of the variable @xmath33 in equation .
as the numerical evaluation of the integral of @xmath33 over @xmath43 is independent of the radial coordinate , for each latitude , the integration over @xmath43 and the loop over the radius become independent . as a consequence ,
the computation of an integrated kernel scales with the maximum of the number of radial and azimuthal grid points rather than with their product . for the example presented in section [ secopt3d ]
, the computation of an integrated 2d kernel is thus about 75 times faster compared to its 3d version .
a drawback of this approach , however , is that 2d integrated kernels can not be projected to different latitudes due to the nature of the coordinate system used .
therefore , this approach turns out to be advantageous if a smaller number of kernels is to be computed .
it is possible to extend the previous approach for computing 2d integrated kernels to the case of travel times which were averaged over multiple points , e.g. in a point - to - arc geometry , if all individual point - to - point measurements were obtained for the same travel distance @xmath49 .
indicating with @xmath50 and @xmath51 averages of @xmath52 and @xmath53 over all pairs of observation points , we have @xmath54
in the following , we use a numerical simulation of helioseismic wave propagation in the solar interior by @xcite , which includes a standard single - cell meridional flow profile and which has been used before for validation purposes ( see @xcite , @xcite , @xcite , and references therein ) .
as explained in @xcite , the flow was amplified by a factor of about 36 to a maximum amplitude of @xmath55 in order to increase the signal - to - noise ratio .
as the simulation is linear , this increase does not alter the physics .
the radial displacement of the oscillations in the simulation at @xmath56 are used as dopplergrams .
they are our input artificial data .
power spectra , cross - covariance functions , and travel times were obtained from this artificial data by @xcite using a time - distance measurement technique with phase - speed filters @xcite . in the following , these observational quantities
are compared to the zero- and first - order models to be validated in this paper .
cccccc l035 & 342.9 & 2.93 & 37.2 & 3.05 & 38.9 + l040 & 308.5 & 2.93 & 41.5 & 3.05 & 43.4 + l045 & 274.6 & 2.93 & 46.7 & 3.05 & 49.0 + l050 & 248.8 & 2.92 & 51.8 & 3.05 & 54.5 + l065 & 199.4 & 2.91 & 65.1 & 3.05 & 69.0 + l080 & 158.7 & 2.90 & 82.8 & 3.05 & 88.5 + l100 & 130.7 & 2.87 & 101.0 & 3.04 & 108.9 + l120 & 106.9 & 2.78 & 121.5 & 2.96 & 130.1 + l150 & 78.5 & 2.63 & 138.8 & 2.81 & 143.9 + l170 & 30.5 & 2.16 & 157.3 & 2.24 & 156.4 + filtered power spectra obtained from the simulation and from the zero - order model are compared in figures [ figpowercomp ] and [ figpowercomp2 ] .
figure [ figpowercomp ] shows cuts through the power spectra at the central harmonic degree of each filter ( top ) and power spectra summed over harmonic degree ( bottom ) .
figure [ figpowercomp2 ] shows power spectra integrated over frequency .
see table [ tabfilters ] for details on the filters used . in the following ,
the simulated data is compared to our model using two different sets of mode frequencies and damping rates . as a first - guess case ,
the zero - order model power spectrum , displayed as a green dotted line , was computed with frequencies from model s and damping rates from mdi ( `` unmatched '' in the legends appearing in this paper ) .
damping rates were provided by j. schou ( 2006 , private communication ) .
it can be seen that the peaks in the model power spectrum are systematically different from the ones in the simulation .
the widths of the peaks are systematically smaller in the model power spectrum and the central frequencies of the peaks differ , especially at frequencies above about @xmath57 ( best observable in the top right panel ) . as was discussed in (
* section 3.3 ) , the shape of a cut through the model power spectrum matches a lorentzian function centered at the input mode frequency and with the damping rate as its half width at half maximum .
the difference in the location and shape of the peaks between simulation and model thus shows that the mode frequencies and damping rates from the simulated data are different from model s frequencies and observed damping rates .
as the computation of sensitivity functions is strongly dependent on accurately modelling the data power spectra ( e.g. , @xcite ) , with real solar observations , one would use mode frequencies and damping rates as close to solar values as possible . in order to achieve a good match between model and data power spectra for the validation presented in this paper , we therefore fitted frequencies and damping rates to the simulated power spectrum .
these fitted values were used as parameters for the zero - order power spectrum shown as a red dashed curve in figures [ figpowercomp ] and [ figpowercomp2 ] ( `` matched '' in the legends ) .
it can be seen that both the locations and the widths of the peaks of the simulated and zero - order modelled power spectra now match well ( figure [ figpowercomp ] , top ) .
the amplitude of each peak , however , is not a free parameter in the zero - order model and therefore can not be adjusted to the simulation . additionally , the source correlation time , a free parameter modelling the sources in @xcite , was fine - tuned individually for each matched filtered zero - order model power spectrum in order to obtain a power - weighted mean frequency identical to the one from the simulation .
this is necessary for guaranteeing that the mean sensitivity of the kernel is correctly adjusted ( @xcite , @xcite ) .
in addition , the power spectra for the kernel were corrected by an @xmath30-dependent factor which accounts for a different behavior of the frequency - integrated power in the simulation and in the zero - order model , see figure [ figpowercomp2 ] .
see also table [ tabfilters ] for a summary of properties for the power spectra obtained with matched and unmatched frequencies and damping rates , for each individual filter used in this work . in the following , we will compare simulated data to the zero- and first - order models obtained using both sets of frequencies and damping rates . born approximation sensitivity functions in time - distance helioseismology as in @xcite , @xcite , and @xcite
are obtained for a travel - time fit which involves a reference cross - covariance function . on the data analysis side ,
it is advantageous to use some kind of disc- and time - averaged cross - covariance function as a reference .
this assures that travel - times measured at individual locations at a certain point in time correspond to local changes compared to the mean solar cross - covariance for a particular travel distance . on the modelling side ,
the zero - order cross - covariance function is usually used as a reference function .
figure [ figcrefmatched ] shows a comparison of these reference cross - covariance functions for different exemplary travel distances .
it can be seen that the model fits the data very well once the fitted frequencies and damping rates were used .
travel times fitted to the cross - covariances obtained from the simulated data can be compared to forward - modelled travel times , which can be obtained from the born kernels using equation or as the underlying flow field is known .
figures [ figfwdtts2d ] and [ figfwdtts1d ] show a comparison of such travel times .
the displayed measured travel - times were obtained by rebinning an original set of travel times for 598 central latitudes and 126 distances to 66 latitudes and 30 distances .
the corresponding averaging was applied to the kernels used for the forward travel times marked with `` avg . '' in figures [ figfwdtts2d ] and [ figfwdtts1d ] . for the forward - modeled travel times
marked with `` not avg . ''
, however , the kernels have not been averaged . instead
, we have just computed one center - to - arc kernel for each of the 66 times 30 distances .
furthermore , we also show forward travel - times from kernels obtained with an unmatched power spectrum . in figure [ figfwdtts2d ]
, it can be seen that the travel times from the kernels generally fit very well to the measured ones .
the forward - modelled travel times , which were obtained using properly averaged born kernels and a matched power spectrum ( top right panel ) , also reproduce some of the jumps visible in the measured travel times as a function of distance at mid - latitudes .
these jumps are introduced by a change of phase speed from filter to filter .
it can also be seen that there do not seem to be significant differences between the gabor and gb04 fits .
horizontal cuts through the panels displayed in figure [ figfwdtts2d ] are shown in figure [ figfwdtts1d ] , where the match between forward - modelled and measured travel times can be inspected in more detail .
it can be seen that forward - modelled travel times fit the measured ones within the measurement errors for most distances .
the agreement is particularly good for travel distances of about 8 - 20 degrees with measurement errors increasing with increasing travel distance . at very low travel distances ( about @xmath58 )
, it can be seen that the forward - modelled travel times do not fit the measured ones .
the reason for this effect lies in the fact that the simulated data only incorporates modes with harmonic degree @xmath59 .
this cut - off acts as an additional filter at a harmonic degree at which the filter with the lowest phase speed ( l170 ) is centered .
the modes selected by the filter thus do not form a proper wave packet , which may lead to artifacts in the cross - correlations ( see , e.g. , @xcite ) .
additionally , it is noteworthy that the quality of the match of the power spectrum ( compare matched vs. unmatched ) does not have a large effect on the travel times at higher travel distances ( @xmath60 ) . at smaller travel distances ( @xmath61 ) , however , one can observe a substantial effect of the quality of the match of the power spectrum on the travel times .
we note here that this conclusion may be different for other flow models .
we also note that the absence of a proper averaging of multiple kernels in distance and latitude has a considerable effect for smaller travel distances ( @xmath62 ) but seems to vanish for the largest distances in the case of the flow model considered here . for comparison ,
forward travel times using ray kernels are also shown in figures [ figfwdtts2d ] and [ figfwdtts1d ] .
it can be seen that the magnitude of the ray kernel travel times is , in general , very similar to those obtained using born kernels
. however , born kernels seem to better reproduce some features in the measured travel times , e.g. , jumps in the magnitude of the travel times from one filter to another such as from @xmath63 to @xmath64 ( see the middle panel in figure [ figfwdtts1d ] ) . as the given flow model varies on rather large length scales , a relatively good performance of the ray kernels
is expected .
this may be different for meridional flow models which vary on smaller length scales , see section [ secintro ] .
we also note here that the measured travel times obtained using the filter with the highest phase speed ( l035 ) encounter a spurious constant offset , see figure [ figfwdttsnooffset ] .
this offset was corrected for by substracting the mean travel time at each distance .
such an offset is not present in real solar data ( @xcite , @xcite ) . in @xcite ,
this offset was not noticed in the simulated data , as a few distances had been dropped from the analysis , which resulted in an equivalent correction .
the origin of this offset is not completely understood but may be connected to the relatively coarse spatial resolution of the simulated data ( 256 pixels on 180 degrees ) .
as the influence of the radial flow component on the travel times has not been taken into account by @xcite and @xcite , we evaluate its impact in the following using our born approximation model , see the green dashed line in figure [ figfwdtts1d ] . for small travel distances ,
the contribution from the radial flow to the total travel time is small and increases with travel distance .
e.g. , for a latitude of @xmath65 and a travel distance of @xmath66 , the impact of the radial flows relative to the horizontal flows is @xmath67 .
this ratio increases to @xmath68 for a travel distance of @xmath69 .
however , for all distances , the magnitude of the contribution of the radial flows is smaller than the measurement errors of the travel times .
this result is valid for the special case of the simulated data and the flow profile used in this study .
nevertheless , it indicates that it may be worthwhile to study the impact of including kernels for the radial flow component in meridional flow inversions .
in order to show that spherical born approximation kernels can be used for inferring solar meridional flows , a standard sola inversion procedure for the horizontal flow component is carried out following the approach of @xcite using only the horizontal component of the born kernels .
the travel times for the two smallest travel distances belonging to the lowest phase - speed filter ( l170 ) were excluded from the inversion as they can not be trusted , see the discussion in section [ sectts ] and figure [ figfwdtts1d ] . in figure [ figinvmeas ] ,
the inversion results are shown . the target flow profile ( first panel )
was obtained by convolving the target kernels with the flow profile from the simulation .
the second panel shows a flow map resulting from the inversion of noiseless forward - modelled travel times .
it can be seen to match qualitatively very well the target flow profile .
the inversion result for the noisy measured travel times ( third panel ) matches the target flow in a coarser sense , recovering some but not all of the flow pattern from the simulation .
figure [ figinvtargetks ] shows averaging and target kernels for two example target locations .
the misfit of the averaging kernels shown in the right panel of figure [ figinvtargetks ] was obtained with @xmath70 where @xmath71 is the averaging kernel and @xmath72 is the target kernel for a particular target location @xmath73 .
we note that the weights and thus the averaging kernels are identical for both inversions presented here .
in this paper , we have presented the validation of spherical born kernels for inferring the deep solar meridional flow with time - distance helioseismology .
we showed that it is possible to efficiently compute spherical born kernels for measuring the deep solar meridional flow .
to do so , we used the recently developed approach of @xcite , which was further optimized for computational efficiency , either for obtaining two - dimensional integrated or full three - dimensional kernels .
the numerical optimization was based on the horizontal variation of the eigenfunctions being separable from the radial dependence and the radial order of the mode .
compared to a recently developed method by @xcite , the numerical efficiency of our method is found to be similar , with some advantages in the case of filtered kernels or in the case of a fine frequency resolution as needed for deep flow measurements . using a spherical born approximation model , it is possible to accurately model observational quantities relevant for time - distance helioseismology such as the mean power spectrum , disc - averaged cross - covariances , and first - order travel times perturbed by a given flow field .
we also show that the match of the reference cross - covariance between model and observations depends on the match of the model power spectrum to the observed one .
the agreement is very good if the mode frequencies and damping rates entering the model are extracted from the measured power spectrum .
the match between observed and modelled travel times , however , does not seem to depend significantly on the match in the power spectrum for travel distances larger than @xmath74 for the flow model considered here . for travel distances
smaller than @xmath74 , we found a noticeable dependence of the forward travel times on the match of the power spectrum to observations . using a standard 2d sola inversion of travel times measured from the simulated data for the horizontal flow component
, we can recover most features of the input meridional flow profile in the inverted flow map .
the agreement is particularly good for the inversion of noiseless forward - modelled travel times .
when inverting noisy measured travel times , we obtain a coarse agreement between inverted and target flows .
this shows that born kernels can be used for inferring the deep solar meridional flow if the noise level in the data is small enough .
the born approximation is thus a promising method for inferring large - scale solar interior flows .
we note , however , that an extensive study is needed in order to compare the use of born and ray kernels in inversions of the deep meridional flow in more detail .
acknowledgements : the research leading to these results has received funding from the european research council under the european union s seventh framework programme ( fp/2007 - 2013 ) / erc grant agreement n. 307117 .
this work was supported by the solarnet project ( www.solarnet-east.eu ) , funded by the european commission s fp7 capacities programme under the grant agreement 312495 .
j.j . acknowledges support from the national science foundation under grant number 1351311 .
s.k . was supported by nasa s heliophysics grand challenges research grant 13-gcr1 - 2 - 0036 .
the authors thank thomas hartlep for providing the simulated data .
v.b . thanks kolja glogowski for computing eigenmodes for model s and helping with many science - related it issues .
the authors acknowledge fruitful discussions during the international team meeting on `` studies of the deep solar meridional flow '' at issi ( international space science institute ) , bern .
the authors thank the referee for constructive comments which improved the paper .
our starting point is from equations ( 38 - 40 ) in @xcite , @xmath75 where @xmath76 { \bm { { \bm \nabla } } } _ { { { \mathbf r } } } \bigg [ { { \mathcal o}}_k^{ln } ( { { \mathbf r } } ) \big [ p_{l}(\cos \delta_1 ) \big ] \bigg ] } , \label{eqzij } \\ { j_{ij}({{\mathbf r}}_1,{{\mathbf r}}_2 ) } & = ( 2l+1)(2\bar l + 1 ) r_{ln}(r_s ) r_{\bar l \bar n}(r_2)\nonumber \\ & \ , \ , \ , \times { \sum_{n ' } r_{ln'}(r_s ) r_{ln'}(r_1 ) \int_{-\infty}^\infty \frac{{{\rm i}}\omega^3 { { w^*_\text{diff}}}({{\mathbf r}}_1,{{\mathbf r}}_2,\omega ) m(\omega ) \ , f(l,\omega ) f(\bar l,\omega)}{4\pi \ , ( \sigma^{2}_{ln}-\omega^2 ) ( \sigma^{2*}_{ln'}-\omega^2)(\sigma^{2}_{\bar l \bar n}-\omega^2 ) } \ , { { \mathrm d}}\omega } .\label{eqjij}\end{aligned}\ ] ] it is now possible to separate the horizontal and radial dependence of the eigenfunctions imprinted in the operators @xmath77 in equation ( see @xcite , eq .
( 17 ) , for their definition ) so that one obtains @xmath78 using appropriate definitions for @xmath79 . from equations and we derive , bearing in mind @xmath80 , @xmath81 where we defined @xmath82 41ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty link:\doibase 10.1146/annurev - astro-081913 - 040012 [ * * , ( ) ] link:\doibase 10.12942/lrsp-2010 - 3 [ * * , ( ) ] link:\doibase 10.12942/lrsp-2005 - 6 [ * * , ( ) ] link:\doibase 10.12942/lrsp-2005 - 1 [ * * , ( ) ] link:\doibase 10.1088/0004 - 637x/725/1/658 [ * * , ( ) ] link:\doibase 10.1088/0004 - 637x/760/1/84 [ * * , ( ) ] link:\doibase 10.1038/36294 [ * * , ( ) ] link:\doibase 10.1086/342636 [ * * , ( ) ] link:\doibase 10.1086/381489 [ * * , ( ) ] link:\doibase 10.1086/339631 [ * * , ( ) ] link:\doibase 10.1086/432413 [ * * , ( ) ] link:\doibase 10.1007/s11207 - 008 - 9264-y [ * * , ( ) ] link:\doibase 10.1088/2041 - 8205/713/1/l16 [ * * , ( ) ] link:\doibase 10.1088/0004 - 637x/717/1/488 [ * * , ( ) ] link:\doibase 10.1086/311727 [ * * , ( ) ] link:\doibase 10.1002/asna.200610745 [ * * , ( ) ] link:\doibase 10.1088/2041 - 8205/774/2/l29 [ * * , ( ) ] , link:\doibase 10.1088/0004 - 637x/805/2/133 [ * * , ( ) ] , link:\doibase 10.1088/0004 - 637x/813/2/114 [ * * , ( ) ] , link:\doibase 10.1088/2041 - 8205/778/2/l38 [ * * , ( ) ] , link:\doibase 10.1038/362430a0 [ * * , ( ) ] link:\doibase 10.1086/309989 [ * * , ( ) ] in link:\doibase 10.1007/978 - 94 - 011 - 5167 - 2_26 [ _ _ ] , , vol . , ( ) pp .
link:\doibase 10.3847/0004 - 637x/824/1/49 [ * * , ( ) ] , http://adsabs.harvard.edu/abs/2016arxiv161101666g [ * * , ( ) ] \doibase 10.1023/a:1005283526062 [ * * , ( ) ] link:\doibase 10.1086/340015 [ * * , ( ) ] link:\doibase 10.1086/324766 [ * * , ( ) ] link:\doibase 10.1086/303680 [ * * , ( ) ] link:\doibase 10.1086/424961 [ * * , ( ) ] link:\doibase 10.1086/500103 [ * * , ( ) ] link:\doibase 10.1002/asna.200610724 [ * * , ( ) ] , link:\doibase 10.1051/0004 - 6361/201526971 [ * * , ( ) ] , link:\doibase 10.1088/0004 - 637x/762/2/132 [ * * , ( ) ] , link:\doibase 10.1086/423367 [ * * , ( ) ] http://adsabs.harvard.edu/abs/2010aste.book.....a[__ ] ( ) link:\doibase 10.1088/0004 - 637x/784/2/145 [ * * , ( ) ] , link:\doibase 10.1086/522914 [ * * , ( ) ] , link:\doibase 10.1088/0004 - 637x/788/2/127 [ * * , ( ) ] , link:\doibase 10.1088/0004 - 637x/794/1/18 [ * * , ( ) ] , in http://adsabs.harvard.edu/abs/2006esasp.617e..48z[__ ] , , vol . | accurate measurements of deep solar meridional flow are of vital interest for understanding the solar dynamo . in this paper , we validate a recently developed method for obtaining sensitivity functions ( kernels ) for travel - time measurements to solar interior flows using the born approximation in spherical geometry , which is expected to be more accurate than the classical ray approximation . furthermore , we develop a numerical approach to efficiently compute a large number of kernels based on the separability of the eigenfunctions into their horizontal and radial dependence .
the validation is performed using a hydrodynamic simulation of linear wave propagation in the sun , which includes a standard single - cell meridional flow profile .
we show that , using the born approximation , it is possible to accurately model observational quantities relevant for time - distance helioseismology such as the mean power spectrum , disc - averaged cross - covariance functions , and travel times in the presence of a flow field . in order to closely match the model to observations , we show that it is beneficial to use mode frequencies and damping rates which were extracted from the measured power spectrum .
furthermore , the contribution of the radial flow to the total travel time is found to reach 20% of the contribution of the horizontal flow at travel distances over @xmath0 . using the born kernels and a 2d sola inversion of travel times , we can recover most features of the input meridional flow profile .
the born approximation is thus a promising method for inferring large - scale solar interior flows . |
a finite charge can be pumped through a mesoscopic system in the absence of an applied bias voltage by changing periodically in time some parameters of the system . if the parameters are changed in time slowly with respect to the lifetime of the electrons in the system , pumping is _
adiabatic_. the idea of electron pumping through a mesoscopic system is based on a work of thouless.@xcite since then a large number of theoretical @xcite and experimental @xcite works has been dedicated to this field .
adiabatic pumping has a geometric nature : the pumped charge depends only on the area spanned by the pumping cycle in parameter space and not on its detailed timing .
it can therefore be related to geometric phases.@xcite in normal conductors transport is due to the transfer of individual quasiparticles . in superconducting devices ,
quasiparticle transport is hindered by the gap , and charge pumping is due to cooper pairs .
@xcite in hybrid normal - superconducting systems , andreev reflection at the interface between the normal and the superconducting region allows for sub - gap charge transport in the system . in the absence of coulomb interaction ,
adiabatic pumping in such hybrid systems can be studied by generalizing in nambu space the scattering approach to adiabatic pumping .
@xcite in the present paper we study a system consisting of a quantum - dot attached to a normal conducting lead and a superconducting lead ( n - dot - s system ) as shown in fig .
[ fig_system ] . in the absence of coulomb repulsion in the dot ,
adiabatic pumping through such a system has been discussed in refs . .
on the other hand , when coulomb repulsion is present , only dc transport through the n - dot - s system has been investigated.@xcite + and @xmath0 .
possible time - dependent parameters are : the strength of the tunnel couplings @xmath1 and @xmath0 , and the dot - level position @xmath2 . [ fig_system],width=268 ] we generalize a recently developed formalism to calculate the pumped charge through an interacting quantum dot@xcite to the case of the n - dot - s system .
first , we apply this formalism to a non - interacting quantum dot , studying systematically the ratio between the charge pumped through the n - dot - s system and the charge pumped through a system consisting of a dot attached to two normal conducting leads ( n - dot - n system ) .
this ratio was calculated in refs . for a non - interacting resonant system at zero temperature .
we consider finite temperature and extend the analysis to the off - resonant situation .
finally , we apply the formalism to the infinite - u anderson model in the kondo regime . we discuss the pumped charge as a function of the level position and the superconducting gap and compare to the interacting n - dot - n system .
in this paper we consider a quantum dot tunnel coupled to one normal and one superconducting lead , as shown in fig .
[ fig_system ] .
the system is described by the hamiltonian : @xmath3 .
the dot hamiltonian reads @xmath4 , where @xmath5 is a spin label , @xmath6 and we allow the dot level @xmath2 to depend on time @xmath7 .
the hamiltonian for the normal lead reads @xmath8 , and the one for the superconducting lead reads @xmath9 , with @xmath10 being the superconducting order parameter .
we take @xmath11 to be @xmath12-independent , which is appropriate for conventional bcs superconductors such as aluminum and niobium .
the operator @xmath13 ( @xmath14 ) creates ( annihilates ) a dot electron with spin @xmath5 , the operator @xmath15 ( @xmath16 ) creates ( annihilates ) an electron in lead @xmath17 with momentum @xmath12 and spin @xmath5 . the tunnel coupling between the dot and the leads is taken into account by the tunneling hamiltonians @xmath18 , where @xmath19 , and the tunneling amplitudes @xmath20 can be time - dependent .
we assume that the phase of @xmath20 does not depend on time , since a time - dependent phase would describe an applied bias voltage . from now on we work in nambu space , defining the fields : @xmath21 , and @xmath22 .
generalizing the approach of refs . , the pumped current can be related to the local green s function of the dot .
in particular , the first adiabatic correction to the charge current from the dot to the left ( normal ) conductor through the barrier is given by @xmath23 \left ( \hat{g}_{0}^{\text{r}}(\omega , t ) \right)^{-1 } \hat{g}_0^{\text{a}}(\omega , t ) \right]\right\}+j^{\text{corr}},\ ] ] where @xmath24 is the fermi function , the caret indicates a matrix in nambu space and @xmath25 is the pauli matrix given by @xmath26 . the first term in eq .
( [ adiabaticcurrent ] ) corresponds to the so called average - time approximation,@xcite while @xmath27 ( discussed below ) contains contributions due to vertex corrections@xcite and it is zero for all cases studied in this paper .
the instantaneous retarded green s function is defined as @xmath28 , and its fourier transform is @xmath29 .
the last argument ( @xmath7 ) of the green s function is the time with respect to which the adiabatic expansion has been performed.@xcite in particular , the instantaneous retarded green s function , @xmath30 is computed with all time - dependent parameters frozen at time @xmath7 .
the advanced green s function is related to the retarded one by @xmath31 .
the intrinsic line width @xmath32 is defined as @xmath33 with @xmath34 and @xmath35 .
the normal lead is supposed to have flat bands with a constant density of states @xmath36 .
the density of states @xmath37 is the one of the superconducting lead in its normal state and it is constant , too .
we now discuss the average - time approximation and give an expression for the correction term @xmath27 .
the average time approximation neglects terms due to vertex corrections@xcite in evaluating the adiabatic expansion of the self energy .
the self energy is defined by the dyson equation @xmath38 and it is a functional of the time - dependent hamiltonian @xmath39 , where @xmath40 $ ] .
the adiabatic expansion is performed by linearizing the time - dependence of the hamiltonian around a fixed time @xmath7 ( denoted as a third argument of the self energy ) . in order to solve the dyson equation for the adiabatic expansion of the green s function
, the average - time approximation is performed on top of the adiabatic expansion .
such approximation consists in replacing the linear time dependence on @xmath41 by the average time @xmath42 of the interval @xmath43 $ ] and thus it neglects the term @xmath44 where the time integral is calculated along the keldysh contour , @xmath45 is the time - ordering operator along the keldysh contour , and @xmath46 a functional derivative .
finally , the correction to the pumped current can be written as @xmath47.\nonumber\end{aligned}\ ] ] it is important to point out that the average - time approximation is exact whenever : i ) the dot is non - interacting , ii ) temperature is zero , the interaction is arbitrary , and the system can be effectively mapped to a fermi liquid , and iii ) when interaction is infinite and the self - energy @xmath48 is calculated up to linear order in @xmath49 .
therefore the correction term equals zero in all cases discussed in this paper . in the following we are interested in calculating the charge pumped through the system in the weak pumping regime ( bilinear response in the pumping fields @xcite ) .
the pumped charge @xmath50 is related to the pumped current by @xmath51 , where @xmath52 is the period of the cycle .
we can choose as pumping parameters , @xmath53 and @xmath54 , any two of the three quantities @xmath55 , @xmath56 and @xmath57
. the time - averaged part @xmath58 is denoted by a bar and the time - dependent part by @xmath59 . for the pumped charge , with pumping parameters indicated in the subscript , we find @xmath60 the prefactor @xmath61 accounts for the amplitude and the relative phase of the pumping parameters and is equal to the surface in parameter space enclosed in one pumping cycle .
it is defined as @xmath62\delta y(\tau)$ ] .
the order of the parameters in the argument of @xmath63 is important and has to be respected in the formulas above ( changing their order results in an additional minus sign . ) @xmath64 and @xmath65 are the retarded and advanced dot green s functions ( both matrices in nambu space although the caret has been omitted to simplify the notation ) with all parameters taken at their time - averaged value .
we start by considering the simple case of a non - interacting quantum dot .
the results which we obtain for the non - interacting dot , can equivalently be calculated by a generalization of brouwer s formula to the n - dot - s system.@xcite the instantaneous green s function of the dot is @xmath66 where the self energy due to the proximity of the superconducting lead reads @xmath67 we now focus on the limit of large superconducting gap @xmath68 , when @xmath69 and @xmath70 .
let us consider first the particle - particle spectral function , which is defined as @xmath71 .
the spectral function conveys important information on the proximity effect in the quantum dot .
furthermore , the dc transport properties are influenced by the energy distribution of spectral weight , for example the andreev - reflection probability @xmath72 has a similar structure as @xmath73 . due to the close relation of the spectral function with the retarded green s function appearing in the formula for the pumped charge
, the spectral function will also prove to be a useful quantity to analyze pumping in the n - dot - s system .
it turns out that @xmath74 is given by a linear combination of two lorentzians and reads @xmath75 with the lorentzian bell defined by @xmath76 .
the presence of two peaks in the particle - particle spectral function [ eq . ( [ eq_spectral_nonint ] ) ] is a signature of the proximity effect induced by the superconductor .
the width of the lorentzians appearing in eq .
( [ eq_spectral_nonint ] ) is equal to the coupling to the normal lead , while their positions are shifted , with respect to the dot - level position , by the coupling to the superconducting lead @xmath0 .
this fact can be referred to as a _ proximization _ of the dot - level position , which might be contrasted to the level _ renormalization _ due to coulomb interaction .
the repulsion of the two lorentzian peaks with increasing @xmath0 reflects the increasing coupling between electron and hole excitations in the dot [ represented by the off - diagonal terms in eq .
( [ eq_sigma_nonint ] ) ] . in the non - interacting case
( [ eqn_weak_pumping ] ) can be written in a compact form if the two pumping parameters , denoted by @xmath77 and @xmath78 , are chosen among the three quantities @xmath79.@xcite we denote by @xmath80 the parameter which is kept constant in time .
( [ eqn_weak_pumping ] ) simplifies to @xmath81 where @xmath82 is the denominator of the particle - particle spectral function with time - averaged parameters @xmath83\right\}^2\\ & + & 4\left(\frac{\bar{\gamma}_{\mathrm{n}}}{2}\right)^2\left[\left(\frac{\bar{\gamma}_{\mathrm{s}}}{2}\right)^2+\bar{\epsilon}^2\right ] \
.\nonumber\end{aligned}\ ] ] the quantity @xmath61 is the surface in parameter space enclosed in one pumping cycle .
the antisymmetric tensor @xmath84 has to be evaluated respecting the order of the parameters @xmath79 . in the case of weak - pumping with the two tunnel - barrier strengths , i.e @xmath85 , and @xmath86
, we can rewrite eq .
( [ eqn_charge_nonint ] ) as @xmath87 which relates the pumped charge to the probability of andreev reflection @xmath88 , computed with the time - dependent parameters frozen at their average value .
the probability of andreev reflection is given by@xcite @xmath89 ^ 2+\gamma^2_{\mathrm{n}}\left(\gamma^2_{\mathrm{s } } + 4\epsilon^2\right)}\\ & = & \frac{\pi}{4}\frac{\gamma_{\mathrm{n}}\gamma_{\mathrm{s}}^2}{\frac{\gamma_{\mathrm{s}}^2}{4}+\frac{\gamma_{\mathrm{n}}^2}{4}+\epsilon^2+\omega^2}\left [ l_{\gamma_{\mathrm{n}}}\left(\omega-\sqrt{\frac{\gamma_{\mathrm{s}}^2}{4}+\epsilon^2 } \right)\right.\nonumber\\ & + & \left .
l_{\gamma_{\mathrm{n}}}\left(\omega+\sqrt{\frac{\gamma_{\mathrm{s}}^2}{4}+\epsilon^2 } \right ) \right ] .
\label{andref}\end{aligned}\ ] ] + , as a function of @xmath90 .
the curve is the same for any couple of pumping parameters @xmath91 .
temperature is equal to zero and the bare dot level is at resonance .
[ fig_ratio_nonint],width=240 ] we are now interested in comparing our results to the ones obtained for an n - dot - n system ( see fig . [ fig_system ] with normal - conducting right lead ) .
the current pumped through a non - interacting dot attached to two normal conducting leads can be obtained using brouwer s
formula.@xcite for the case where the tunneling rates @xmath92 and @xmath93 are the pumping parameters , it has been found that the current in the presence of one superconducting lead is about a factor 4 bigger than in the case of two normal leads , if the dot level is at resonance ( @xmath94 ) and temperature is zero.@xcite we find exactly the same behavior for all three choices of pumping parameters , at zero temperature .
it turns out that the ratio @xmath95 ( equal for any pairs of pumping parameters ) has a maximum , as a function of the level position , at resonance.@xcite we plot in fig .
[ fig_ratio_nonint ] the ratio @xmath95 , for @xmath94 , as a function of @xmath90 : a maximum is present when @xmath96 . in the limit of small and large @xmath90
the charge ratio goes to zero .
this happens because , in the presence of a superconducting lead , pumping is mediated by the andreev reflection probability , eq .
( [ andref ] ) . on one hand , for @xmath97 ,
the ratio is suppressed since @xmath98 implies higher - order tunneling processes to the superconducting lead with respect to tunneling to the normal lead . on the other hand , for @xmath99 , the lorentzian contributions to @xmath98 get narrower with decreasing @xmath100 and more distant with increasing @xmath101 , thus suppressing the andreev reflection probability at @xmath102 ( relevant for zero temperature ) . in the case of finite temperature
the charge ratio depends on the parameter choice and can take values much bigger than @xmath103 .
[ fig_ratio_nonint_tfinite ] shows the ratio @xmath95 for the three different parameter choices at temperature @xmath104 , with the dot level being at resonance .
, as a function of @xmath90 .
temperature is equal to @xmath104 .
the pumping parameters are @xmath1 and @xmath0 ( solid line ) , @xmath1 and @xmath2 ( dashed line ) and @xmath0 and @xmath2 ( dotted and dashed line ) .
the bare dot level is at resonance .
[ fig_ratio_nonint_tfinite],width=240 ] in the following we discuss some qualitative features , in the high - temperature limit ( @xmath105 ) , of the ratio @xmath95 .
expressions for the pumped charge are reported in appendix [ a1 ] . while for the n - dot - s system
the leading order of the pumped charge is the first derivative of the fermi function , in the n - dot - n system this is the case only when the level position is one of the pumping parameters .
the first contribution to @xmath106 starts with the second derivative of the fermi function . on the other hand ,
the parameter @xmath107 enters in the spectral density of the n - dot - s system via the proximized peak positions .
the high - temperature limit of the charge ratio as a function of temperature is therefore a constant if @xmath108 are the pumping parameters and it is quadratic in temperature in the case that @xmath109 are the pumping parameters : [ eq_ratio_hight ] @xmath110 @xmath111 @xmath112 for all three cases described by eqs .
( [ eq_ratio_hight ] ) , the ratio of the pumped charge , as a function of the average dot level position , has a maximum when the average dot level is at resonance . at @xmath113 , with @xmath114 , we find that @xmath115 and @xmath116 . furthermore we note that due to the second term in eq .
( [ eq_ratio_hight_3 ] ) , the ratio @xmath117 is increasing in a monotonic way for increasing @xmath118 , while @xmath119 and @xmath120 have a maximum as a function of @xmath118 . this behavior is present in fig .
[ fig_ratio_nonint_tfinite ] , where we also observe a large value of the maximum of @xmath119 , which is due to the quadratic dependence on @xmath121 . the plots in fig .
[ fig_charge_nonint ] show the pumped charge , as a function of the average dot level position , through the n - dot - s system [ panel ( a ) ] and through the n - dot - n system [ panel ( b ) ] for different choices of pumping - parameter pairs at @xmath122 and for @xmath123 .
we point out several differences : in the n - dot - n case the pumped charge is the same for pumping with @xmath124 and with @xmath125 . in the presence of a superconducting lead the symmetry between the two leads
is absent as soon as temperature differs from zero .
furthermore we report that the width of the pumped charge is strongly temperature dependent for the n - dot - n system , while for the n - dot - s system the width is saturating for high temperatures to a value that depends solely on the coupling to the leads , @xmath1 and @xmath0 .
this is due to the fact that the gap of the superconductor is always bigger than temperature and no quasiparticle transport , which is sensitive to temperature , takes place between the superconductor and the dot .
the symmetry of the pumped charge as a function of the dot - level position is present for both systems . for the n - dot - s system
, when pumping with @xmath1 and @xmath0 , this happens despite the fact that , at least at high temperatures , pumping is dominated by the proximized level position . and in the n - dot - n [ panel ( b ) ] systems as a function of the average level position @xmath126 .
temperature is set to @xmath122 and @xmath123 .
the pumping parameters are @xmath127 ( solid line ) , @xmath128 ( dashed line ) and @xmath129 ( dashed - dotted line ) .
the units on the x - axes are the ones determining the width of the respective function .
[ fig_charge_nonint],title="fig:",width=240 ] and in the n - dot - n [ panel ( b ) ] systems as a function of the average level position @xmath126 .
temperature is set to @xmath122 and @xmath123 .
the pumping parameters are @xmath127 ( solid line ) , @xmath128 ( dashed line ) and @xmath129 ( dashed - dotted line ) .
the units on the x - axes are the ones determining the width of the respective function .
[ fig_charge_nonint],title="fig:",width=240 ] it might be interesting to contrast the pumped charge in the n - dot - s system with the dc linear conductance of the same system ( here all parameters are time - independent ) : @xmath130 in the high temperature limit , i.e. @xmath131 , the conductance is given by @xmath132
in the regime of strong coulomb repulsion double occupation is forbidden and coherent andreev scattering can occur only if @xmath11 is finite ( electrons can enter the superconductor within a time window of order @xmath133 ) . in the kondo limit , when temperature is zero , we can evaluate the retarded equilibrium dot green s function making use of the mean - field slave - boson technique.@xcite as the dot can only be empty or singly occupied we can describe the singly - occupied dot state by the fermion operator @xmath134 ( known as pseudo fermion ) and the empty dot by the boson operator @xmath135 .
the real dot operator @xmath14 is simply given by @xmath136 .
the operators @xmath135 and @xmath134 need to fulfill the constraint @xmath137 .
such a constraint is introduced in the hamiltonian using the lagrange multiplier @xmath138 which plays the role of a chemical potential . in the mean - field limit
, the boson operator @xmath135 is replaced by a real number @xmath139 .
the quantities @xmath140 and @xmath138 are determined by minimizing the free energy .
note that we are allowed to use an equilibrium formalism , since after performing the adiabatic approximation we are left with a current formula , eq .
( [ adiabaticcurrent ] ) , that contains only equilibrium quantities .
details of the derivation of the retarded dot green s function by means of the slave - boson technique , in the limit @xmath141 , are given in appendix [ a2 ] .
it is important to notice that the pumped charge in the deep kondo regime is zero .
therefore , in the following , we will be interested in the mixed - valence regime only .
[ panel ( a ) ] and @xmath142 [ panel ( b ) ] as a function of the average level position for different values of @xmath11 , solid lines , and @xmath143 and @xmath144 , dashed lines .
temperature is zero and we consider symmetric barriers ( @xmath145 ) .
the cut - off energy is @xmath146 .
[ fig_delta_dip],title="fig:",width=240 ] [ panel ( a ) ] and @xmath142 [ panel ( b ) ] as a function of the average level position for different values of @xmath11 , solid lines , and @xmath143 and @xmath144 , dashed lines .
temperature is zero and we consider symmetric barriers ( @xmath145 ) .
the cut - off energy is @xmath146 .
[ fig_delta_dip],title="fig:",width=240 ] the pumped charge in the weak - pumping limit for the aforementioned three pairs of pumping parameters can be calculated using eq .
( [ eqn_weak_pumping ] ) .
the pumped charge has a similar structure in all three cases .
we find , in particular , the general relation @xmath147 which is also valid for the noninteracting n - dot - s system at zero temperature . in analyzing the properties of the pumped charge as a function of the different system parameters we can profit from the fact that , in the kondo regime , the system behaves like a non - interacting system with renormalized parameters , denoted as @xmath148 , @xmath149 and @xmath150 .
we should however keep in mind that the renormalized parameters depend in a complex way on all the bare parameters .
panel ( a ) of fig .
[ fig_delta_dip ] shows the pumped charge @xmath151 ( solid lines ) as a function of the level position , for increasing values of the superconducting gap @xmath11 . the charge @xmath143 in the n - dot - n is also shown for reference ( dashed line ) . in panel ( b ) , the pumped charge is shown for @xmath1 and @xmath2 chosen as pumping parameters .
we note that for both the n - dot - n system and the n - dot - s system ( for any value of the superconducting gap @xmath11 ) the charge has a bell - like structure .
the width of the bell - like curves reflects the width of the respective particle - particle spectral functions , which for the n - dot - s system is given by @xmath148 while for the n - dot - n system it is given by @xmath152 .
for the values of parameters considered in fig .
[ fig_delta_dip ] ( @xmath153 ) , the curve relative to @xmath154 is still very far away from the curve relative to the n - dot - n system .
is shown as a function of the superconducting gap for different pairs of pumping parameters on a logarithmic scale .
the pumping parameters are @xmath155 ( solid line ) and @xmath124 ( dashed line ) .
the temperature is zero and @xmath153 .
[ fig_peakposition],title="fig:",width=240 ] + the striking similarity of the pumped charge for the two different pairs of pumping parameters [ compare figs .
[ fig_delta_dip ] ( a ) and ( b ) and eq .
( [ eq_similar_charge ] ) ] can be explained by the fact that pumping takes place via renormalized parameters .
indeed , in the zero - temperature regime with strong coulomb interaction , the green s function of the system depends on the bare parameters solely via renormalized parameters @xmath156 and @xmath150 .
therefore , the detailed behavior of the pumped charge is determined by the derivatives of the renormalized parameters with respect to the bare ones . in particular when the bare dot level is deep below the fermi energy , i.e. in the kondo regime , the renormalized parameters are independent of the bare parameters and therefore the pumped charge is zero .
now , we focus our attention on the dependence of the pumped charge on the superconducting gap . in a non - interacting dot ,
the pumped charge is always independent of the superconducting gap @xmath11 at @xmath157 , since the green s function of eq .
( [ eq_g_nonint ] ) , for @xmath102 , is independent of @xmath11 .
the situation is different in the presence of coulomb interaction , where the pumped charge depends on the superconducting gap via the renormalized parameters .
now , we investigate how the features of the bell curves , describing the pumped charge as a function of the average level @xmath126 , depend on the gap @xmath11 .
+ _ peak position .
[ fig_peakposition ] shows that the position of the peak of the pumped charge @xmath158 depends on @xmath159 .
the peak position is shifted towards zero ( from below ) for increasing gap .
this is consistent with the fact that the renormalized level acquires a logarithmic correction in @xmath11 proportional to @xmath0.@xcite for the pumped charge plotted as function of the superconducting gap @xmath11 .
the pumping parameters are @xmath155 .
the temperature is zero and @xmath153 .
[ fig_width],title="fig:",width=240 ] + _ peak width . _ fig .
[ fig_width ] shows the peak width @xmath160 ( half - height - width ) for the pumped charge @xmath151 as a function of the superconducting gap @xmath11 . as stated before , the peak width has an analogous behavior to the width of the spectral function .
the coupling to the superconducting lead influences the width of the spectral function via the renormalized coupling @xmath161 .
the decrease of @xmath161 with increasing superconducting gap induces the @xmath11-dependence shown in fig .
[ fig_width ] .
we have extended a green s function formalism to compute the pumped current through a quantum dot with coulomb interaction , to the case when the dot is tunnel - coupled to one normal and one superconducting lead .
first , we have applied our formalism to a non - interacting dot in an n - dot - s configuration , where we derived a compact formula for the pumped charge , and we calculated the ratio between the charge pumped through the n - dot - s system and through an n - dot - n system . at finite temperature , we found different characteristic behaviors of this ratio , depending on the choice of pumping parameters .
then , we have studied the pumped charge through a quantum dot with infinitely strong coulomb interaction at zero temperature using the mean - field slave - boson approach . in particular , we have focused our attention on the influence of the superconducting gap @xmath11 on the pumped charge .
we have explained features of the pumped charge exploiting the mapping to a non - interacting system with renormalized parameters . in the end
we wish to comment on the experimental realization of the pumping mechanism discussed theoretically in this article .
the system depicted in fig .
[ fig_system ] can be realized in analogy to an experiment reported recently by van dam _
et al . _
@xcite . here
a semiconductor quantum dot is contacted with superconductors , using indium arsenide nanowires combined with local gate voltages to create a quantum dot with tunable coupling to the leads .
superconductivity is induced in the wires by means of the proximity effect using aluminum contacts . in order to realize pumping , time - dependent control of the system parameters
is required .
in this appendix we report the expressions for the pumped charge through the non - interacting dot in the weak - pumping regime in the limit @xmath162 .
+ _ n - dot - s system _ from eq .
( [ eqn_charge_nonint ] ) , for the three possible choices of pumping fields , we obtain : @xmath163 \label{eq_hight_1}\end{aligned}\ ] ] @xmath164\label{eq_hight_2}\end{aligned}\ ] ] @xmath165 . \label{eq_hight_3}\end{aligned}\ ] ] _ n - dot - n system _ we compare the results for the n - dot - s system to the charge pumped through the n - dot - n system by varying the respective pumping parameters .
we find : @xmath166
in this appendix we report details of the slave - boson method for the n - dot - s system.@xcite we calculate instantaneous equilibrium green s functions where all parameters are always taken at their time - averaged value .
we therefore omit the bar and , e.g. , write @xmath167 instead of @xmath168 .
furthermore all green s functions considered in this appendix are @xmath169 matrices in nambu space and we omit the caret . the free energy of the system can be written as a function of the matsubara green s function @xmath170 [ related to the retarded green s function by the analytic continuation @xmath171:@xcite @xmath172 where the trace is performed in nambu space and @xmath173 is a fermionic matsubara frequency .
the pseudo - fermion green s function @xmath174 has the same functional form of the noninteracting matsubara green s function @xmath170 , but contains the renormalized parameters @xmath175 , @xmath176 and @xmath177 . minimizing the free energy with respect to the variables @xmath138 and @xmath140 we find the equations : the matrix in nambu space @xmath179 is given by @xmath180 equations ( [ eq_slave_boson]a ) and ( b ) are used to determine @xmath138 and @xmath140 . at zero temperature
we can replace the sum over the matsubara frequencies by an integral .
furthermore , we consider the limit @xmath181 and we perform the following approximation @xcite @xmath183\nonumber\\ & + & \frac{2 } { \pi } \left [ \arctan \left ( \frac{w+\frac{\tilde{\gamma}_{\mathrm{n}}}{2}+\frac{\tilde{\gamma}_{\mathrm{s}}}{2 } } { \tilde{\epsilon } } \right ) -\arctan \left ( \frac{\delta+\frac{\tilde{\gamma}_{\mathrm{n}}}{2}+\frac{\tilde{\gamma}_{\mathrm{s}}}{2 } } { \tilde{\epsilon } } \right ) \right]\\ \lambda & = & \frac{\gamma_{\mathrm{n}}}{2\pi } \mathrm{ln}\left ( \frac{\left(\delta+\frac{\tilde{\gamma}_{\mathrm{n}}}{2}\right)^2 + \tilde{\epsilon}^2+\frac{\tilde{\gamma}_{\mathrm{s}}^2}{4 } } { \frac{\tilde{\gamma}_{\mathrm{n}}^2}{4 } + \tilde{\epsilon}^2+\frac{\tilde{\gamma}_{\mathrm{s}}^2}{4 } } \right ) + \frac{\gamma_{\mathrm{n}}+\gamma_{\mathrm{s}}}{2\pi } \mathrm{ln}\left ( \frac{w^2 } { \left(\delta+\frac{\tilde{\gamma}_{\mathrm{n}}}{2 } + \frac{\tilde{\gamma}_{\mathrm{s}}}{2}\right)^2 + \tilde{\epsilon}^2 } \right)\nonumber\\ & + & \frac{\tilde{\gamma}_{\mathrm{s}}\gamma_{\mathrm{s}}}{2\pi\sqrt{\frac{\tilde{\gamma}_{\mathrm{s}}^2}{4}+\tilde{\epsilon}^2 } } \left [ \arctan \left ( \frac{\delta+\frac{\tilde{\gamma}_{\mathrm{n}}}{2 } } { \sqrt{\tilde{\epsilon}^2+\frac{\tilde{\gamma}_{\mathrm{s}}^2}{4 } } } \right ) -\arctan \left ( \frac{\frac{\tilde{\gamma}_{\mathrm{n}}}{2 } } { \sqrt{\tilde{\epsilon}^2+\frac{\tilde{\gamma}_{\mathrm{s}}^2}{4 } } } \right ) \right]\ , \end{aligned}\ ] ] ) .
the gap of the superconductor is half as big as the bandwidth cut - off ( @xmath188 ) .
the kondo - temperature of the n - dot - n and the n - dot - s system differ by a huge factor : @xmath189 and @xmath190 .
[ fig_spectral],width=240 ] from the retarded green s function of the n - dot - s system , we can calculate the spectral density @xmath191 and compare to the one for the n - dot - n system , in the same regime .
the result is shown in fig .
[ fig_spectral ] .
we find that the kondo temperature in the n - dot - s system is much smaller than in the n - dot - n system ( note that the width of the two curves is rescaled by the respective kondo temperature ) . nevertheless , in the deep kondo regime , the spectral weight at the fermi energy is the same for both systems .
the strength of the kondo - effect is therefore not affected by the presence of the superconductor in the limit of a large but finite gap @xmath11 .
n. e. fletcher , j. ebbecke , t. j. b. m. janssen , f. j. ahlers , m. pepper , h. e. beere , and d. a. ritchie , phys .
b * 68 * , 245310 ( 2003 ) ; j. ebbecke , n. e. fletcher , t. j. b. m. janssen , f. j. ahlers , m. pepper , h. e. beere , and d. a. ritchie , appl .
lett . * 84 * , 4319 ( 2004 )
. the condition for bilinear response is fulfilled if the pumped charge per infinitesimal area in parameter space depends little on the parameters , which vary within the area spanned by the pumping cycle .
this condition is necessarily fulfilled if the amplitude of the time - dependent part @xmath59 is small compared to the time - average @xmath58 .
when pumping with @xmath1 and @xmath0 , the pumped charge is equal to zero if the dot level is exactly at resonance ( in the n - dot - s system as well as in the n - dot - n system ) .
we therefore investigate the limit , in which the level is almost at resonance , i.e. where @xmath194 is much smaller than all other energy scales of the system . | we study adiabatic pumping through a quantum dot tunnel - coupled to one normal and one superconducting lead .
we generalize a formula which relates the pumped charge through a quantum dot with coulomb interaction to the instantaneous local green s function of the dot , to systems containing a superconducting lead .
first , we apply this formula to the case of a non - interacting , single - level quantum dot in different temperature regimes and for different parameter choices , and we compare the results with the case of a system comprising only normal leads
. then we study the infinite - u anderson model with a superconducting lead at zero temperature , and we discuss the effect of the proximity of the superconductor on the pumped charge . |
the electromagnetic force felt by a neutral atom near the surface of a substrate has been object of an intense investigation since the pioneering works by casimir and polder @xcite and lifshitz , dzyaloshinskii and pitaevskii @xcite .
in addition to the fundamental character of the force , these studies @xcite are presently motivated by the possibility of technological applications @xcite , by searching stronger constrains on hypothetical non newtonian forces @xcite as well as its role in biological systems @xcite .
new perspectives to study such a force are opened by the recent development in storing and manipulating ultra - cold atoms .
indeed experimental and theoretical research has been recently focused on the forces acting on ultra - cold atomic gases due to the presence of a nearby surface .
they include atomic beams @xcite , bose - einstein condensates @xcite and degenerate fermi gases @xcite .
the surface - atom force at thermal equilibrium @xmath0 can be in general separated in two parts @xmath1 the first one , @xmath2 , is related to zero - point fluctuations ( @xmath3 ) of the electromagnetic field . at short distances @xmath4 this force behaves like @xmath5 and is the analog of the van der waals - london inter - atomic force . at larger distances the inclusion of relativistic retardation effects gives rise to the casimir - polder asymptotic behaviour @xcite @xmath6 where @xmath7 and @xmath8 are the static polarizability of the atom and the static dielectric function of the substrate respectively . the function @xmath9 is defined , for example , in @xcite .
the second contribution to the force , @xmath10 , is due to the thermal fluctuations of the electromagnetic field .
this contribution was first considered by lifshitz @xcite who applied the theory of electromagnetic fluctuations developed by rytov @xcite . at large distances
the thermal contribution approaches the so - called lifshitz law @xmath11 such asymptotic behaviour is reached at distances larger than the thermal wavelength @xmath12 , corresponding to @xmath13 at room temperature .
thus it is the leading contribution to the total force .
the lifshitz force was originally evaluated at thermodynamic equilibrium .
a non - trivial issue is the study of such a force out of thermal equilibrium , characterizing configurations where the temperature of the substrate @xmath14 and environment @xmath15 , do not coincide .
for instance in typical experiments with ultra - cold atomic gases the environment temperature is determined by the chamber containing the substrate and the trapped atoms . in this paper
we describe the surface - atom force out of thermal equilibrium and how to recover its asymptotic behaviour at large distances .
we assume that the radiation surrounding the atom is not able to populate its internal excited states which are assumed to be at energies @xmath16 much higher than the thermal energy : @xmath17 this condition is very well satisfied at ordinary temperatures ( for example the first optical resonance of rb atoms corresponds to @xmath18k ) .
in the last part of the paper we analyze the effects of such a force on cold atoms , and in particular on the center - of - mass motion of a trapped bose - einstein condensate and on the bloch oscillations of ultra - cold fermionic atoms in an optical lattice .
in the calculation of the surface - atom force the main ingredient is clearly the electromagnetic field and its sources .
the latter , in our approach , are treated as point - like oscillating dipoles .
furthermore it is useful to write the fields using the green s function formalism , the green s function being the solution of the wave equation for a point - like source .
once this solution is known , the solution due to a general source can be obtained by the principle of linear superposition .
the dyadic green function @xmath19 describing the electromagnetic field in surface optics ( for isotropic , linear and non - magnetic media ) is the solution of the equation @xmath20-k^2\varepsilon(\omega;{\bf
r}){\bf \overline{g}}[\omega;{\bf r},{\bf r}']=4\pi k^2{\bf \overline{i}}\delta({\bf r}-{\bf r } ' ) , \label{elweqcfe}\ ] ] with the boundary conditions imposed by the geometry of the problem . in previous equation @xmath21
is the vacuum wavenumber , @xmath22 is the dielectric function and @xmath23 is the identity dyad . equation ( [ elweqcfe ] ) comes from the usual wave equation for the fourier transformed electric field @xmath24-k^2\varepsilon(\omega;{\bf r})\;{\bf e}[\omega;{\bf r}]=4\pi k^2{\bf p}[\omega;{\bf r } ] , \label{elweq}\ ] ] obtained from the macroscopic maxwell equations in which the sources are described by the effective electric polarization field @xmath25 $ ] related to the electric current by @xmath26=-i\omega{\bf p}[\omega;{\bf r}]$ ] .
the convolution of the solution obtained from eq.([elweqcfe ] ) and the effective electric polarization gives the electric field at the observation point @xmath27 @xmath28=\int { \bf \overline{g}}\left[\omega;{\bf r},{\bf r}'\right]\;\cdot { \bf p}\left[\omega;{\bf r}'\right]\;\rmd{\bf r } ' . \label{lafh}\ ] ]
let us consider the atom described by its complex dielectric polarizability function @xmath29 in a vacuum half space @xmath30 and placed at a distance @xmath4 from the surface of the dielectric half space @xmath31 .
let us choose an orthogonal coordinate system with the @xmath32 plane coincident with the interface and the @xmath4 axis such that the dielectric occupies the region with @xmath33 and the vacuum the region with @xmath34 .
the force acting on a neutral atom without a permanent electric dipole moment is @xcite @xmath35 where @xmath36 s are the atomic electric dipole components , we have used the einstein s summation convention for repeated indices and @xmath37 . in eq.([forcepert ] ) the average is done with respect to the state of the atom and of the field and the lowest order in perturbation theory has been considered .
the first term describes the ( spontaneous and thermal ) field fluctuations correlated with the induced dipole , and the second term involves ( spontaneous and thermal ) dipole fluctuations correlated to the field they induce .
the induced electric dipole for the atom at the position @xmath27 is @xmath38=\alpha(\omega)\;{\bf e}^{{\rm tot}}[\omega;{\bf r}]\approx\alpha(\omega)\;{\bf e}^{{\rm fl}}[\omega;{\bf r } ] , \label{diprelmks}\ ] ] where @xmath39 $ ] is the fluctuating field , and now @xmath40 is the atomic polarizability of the atom in an unbounded space . by modeling the atom as a point - like source dipole @xmath41 \rme^{-i\omega t}$ ] at @xmath27 ,
the corresponding polarization in the frequency domain is @xmath42={\bf d}[\omega]\;\delta({\bf r}''-{\bf r})$ ] , and the electric field at the position @xmath43 is @xmath44={\bf \overline{g}}[\omega;{\bf r}',{\bf r}]\cdot{\bf d}^{{\rm tot}}[\omega]\approx{\bf \overline{g}}[\omega;{\bf r}',{\bf r}]\cdot{\bf d}^{{\rm fl}}[\omega ] .
\label{kkso}\end{aligned}\ ] ] using eq .
( [ diprelmks ] ) and ( [ kkso ] ) , the fluctuating dipole and field contributions to the surface - atom force ( [ forcepert ] ) read @xmath45 e_j^{{\rm fl}\dag}[\omega';{\bf r'}]\right\rangle\big{|}_{{\bf r } } , \label{ffffhhhwiaa}\end{aligned}\ ] ] @xmath46\big{|}_{{\bf r}}\;\left\langle d_i^{{\rm fl}}[\omega ] d_k^{{\rm fl}\dag}[\omega']\right\rangle , \label{ffffhhhwiazza}\end{aligned}\ ] ] where the integrations are over the whole real frequency axis .
at thermal equilibrium , in order to calculate the average values in ( [ ffffhhhwiaa ] ) and ( [ ffffhhhwiazza ] ) , it is possible to use the fluctuation dissipation theorem @xcite .
one finds for the fluctuating dipoles @xmath47\;d_j^{{\rm fl}\dag}[\omega']\right\rangle= \frac{4\pi\hbar\;\delta(\omega-\omega')\;\delta_{ij}}{1-\rme^{-\hbar\omega / k_{\rm b}t}}\;\alpha''(\omega ) , \label{fdt11111}\ ] ] and for the fluctuating fields @xmath48\;e_j^{{\rm fl}\dag}[\omega';{\bf r } ' ]
\right\rangle= \frac{4\pi\hbar\;\delta(\omega-\omega')}{1-\rme^{-\hbar\omega / k_{\rm b}t}}\;{\rm im}\;g_{ij}[\omega;{\bf r},{\bf r } ' ] \label{fdt22222}.\ ] ] after substituting the previous equalities into eq.([ffffhhhwiaa ] ) and ( [ ffffhhhwiazza ] ) and using the reciprocity theorem @xmath49=g_{ji}[\omega;{\bf r}',{\bf r}]$ ] , the surface - atom force at thermal equilibrium becomes @xmath50\big{|}_{{\bf r}}\right ] .
\label{t=0mss}\ ] ] because of the relation @xmath51 , where @xmath52 is the bose factor , one can separate in eq.([t=0mss ] ) the zero - point fluctuations contribution @xmath2 from the thermal contribution @xmath10 .
the latter term is the sum of two contributions arising from the two terms of eq.([forcepert ] ) .
the first one is due to the field fluctuations and it is linear in @xmath53 .
the second one arises from the dipole fluctuations and it is linear in @xmath54 . as long as the condition ( [ condition ] ) is valid , the field fluctuations contribution is the leading term in @xmath10 .
a first important investigation of the surface - atom force out of thermal equilibrium was carried out by henkel _
@xcite who calculated the force generated by a dielectric substrate at finite temperature by assuming that the environment temperature is zero .
the principal motivation of that paper was the study of the force at short distances .
in this section we analyze the general case of an atom placed in vacuum at a distance @xmath4 from the flat surface of a substrate that we assume to be locally at thermal equilibrium at a temperature @xmath14 which can be equal or different from the environment temperature @xmath15 , the global system being in or out of thermal equilibrium respectively , but in a stationary regime @xcite . in this configuration
it is relatively easy to describe the radiation produced by the flat substrate , while it is less trivial to describe the radiation coming from the environment .
to face this problem we use the lifshitz trick @xcite for which the vacuum half space is assumed to be a dielectric locally at thermal equilibrium with temperature @xmath15 , by introducing an infinitesimal imaginary part of its dielectric function . using the fluctuation dissipation theorem and after integrating over an infinite volume the vacuum half space produces a radiation corresponding to the one that in a real systems is generated by the environment walls at @xmath15 .
we refer to the substrate as to the half space @xmath55 occupying the volume @xmath31 with @xmath33 , with dielectric function @xmath56 and in thermal equilibrium at the temperature @xmath14 .
the vacuum half space @xmath57 instead occupies the volume @xmath30 with @xmath34 and is characterized by a dielectric function @xmath58 and a temperature @xmath15 . only after calculating the electric fields in this configuration we set @xmath59 .
+ as well as for the thermal equilibrium case , the surface - atom force out of thermal equilibrium can be written as @xmath60 where the thermal contribution @xmath61 , provided the condition ( [ condition ] ) is satisfied , is dominated by the thermal part of the fluctuating fields correlation ( [ ffffhhhwiaa ] ) only , as at thermal equilibrium ) , in the calculation of the zero temperature force @xmath62 both dipole zero - point fluctuations ( [ fdt11111 ] ) and field zero - point fluctuations ( [ fdt22222 ] ) are needed . ] .
+ the physical origin of the electromagnetic field is @xcite the random fluctuating polarization field @xmath25 $ ] , whose correlations , at thermal equilibrium , are described by the fluctuation dissipation theorem @xmath63p_l^*[\omega';{\bf r}']\right\rangle=\frac{\delta(\omega-\omega')\delta({\bf r}-{\bf r}')\delta_{kl}\;\hbar\;\varepsilon''(\omega)}{1-\rme^{-\hbar\omega / k_{\rm b}t}}. \label{1fdttssscc}\ ] ] since the correlations of the source polarization field are _ local _ , the fluctuations of the sources at different points add up incoherently .
therefore we can assume that in the whole space the correlations of the sources are given by equation ( [ 1fdttssscc ] ) , valid for source dipoles in the the two half - spaces _ assumed to be locally _ at thermal equilibrium at two different temperatures @xcite . in order to calculate the field correlation function ( [ ffffhhhwiaa ] ) we express the electromagnetic field in terms of its source polarization field via eq .
( [ lafh ] ) , and using the eq .
( [ 1fdttssscc ] ) we write the thermal part of the surface - atom force out of thermal equilibrium as @xmath64 where the first thermal contribution @xmath65\partial_zg_{ik}^{*}[\omega;{\bf r},{\bf r}']\;\rmd^3{\bf r}'\right]}{\rme^{\hbar\omega / k_{\rm b}t_{\rm s}}-1}\label{tpcpf1ss2}\ ] ] arises from the sources in the substrate @xmath31 , while the second one @xmath66\partial_zg_{ik}^{*}[\omega;{\bf r},{\bf r}']\;\rmd^3{\bf r}'\right]}{\rme^{\hbar\omega / k_{\rm b}t_{\rm e}}-1 } \label{eqnapejc}\ ] ] is produced by the sources in the vacuum half space @xmath30 .
then reduces respectively to its transmitted component in eq.([tpcpf1ss2 ] ) @xcite and to its incident , reflected and local component in eq.([eqnapejc ] ) @xcite . ]
it is possible to show that the sum of eq.([tpcpf1ss2 ] ) and ( [ eqnapejc ] ) , at the same temperature , reproduce the thermal part of the force at thermal equilibrium @xcite .
indeed it is possible to apply to the whole space the green s functions property ( see , for example , @xcite ) @xmath67g_{jk}^*[\omega;{\bf r}_2,{\bf r}]=4\pi\textrm{im}\;g_{ij}[\omega;{\bf r}_1,{\bf r}_2 ] , \label{magicths}\ ] ] where the integration is on the volume @xmath68 such that on its surface the green function is zero . than we can express the complete surface - atom force out of thermal equilibrium in the convenient form @xmath69 where the equilibrium force @xmath0 is given by ( [ feq ] ) and @xmath70
is defined by eq.([tpcpf1ss2 ] ) . consistently with the assumption ( [ condition ] ) , in deriving the thermal part of eq.([fullnoneq ] ) we ignored terms proportional to the imaginary part of the atomic polarizability . for the same reason the _ wind _
contribution in eq.([tpcpf1ss2 ] ) and ( [ eqnapejc ] ) , related to @xmath54 , can be ignored and the real part @xmath71 , corresponding to the _ dispersive _ contribution , can be replaced with its static ( @xmath72 ) value @xmath7 . in this non - absorbing condition
the force of eq.([nonthermalequil ] ) can be also written in the form @xmath73 where @xmath74 is the thermal component of the electric energy density at the atom position . + after some lengthy algebra we find for eq.([tpcpf1ss2 ] ) the relevant result @xmath75 where we introduced the dimensionless variable @xmath76 , with @xmath77 the modulus of the electromagnetic wave - vector component parallel to the interface , and @xmath78 .
calculated from eq.([fullnoneq ] ) , for different thermal configurations.,scaledwidth=50.0% ] in figure [ fig:1 ] we show the explicit results for the full force obtained from eq.([fullnoneq ] ) as a function of the distance from the surface for different choices of @xmath14 and @xmath15 .
the calculations have been performed for a sapphire substrate ( @xmath79 ) and for @xmath80rb atoms ( @xmath81@xmath82 ) .
for @xmath0 we have used the predictions of @xcite .
the figure clearly shows that the thermal effects out of equilibrium are sizable ( solid lines ) , thereby providing promising perspectives for future measurements of the surface - atom force at large distances . to increase the attractive nature of the force it is much more convenient to heat the substrate by keeping the environment at room temperature ( lower solid line ) rather than heating the whole system ( dashed line ) .
when @xmath83 ( upper solid line ) the force exhibits a characteristic change of sign reflecting a repulsive nature at large distances ( see also discussion below ) . at short distances the thermal correction to the force becomes smaller and smaller and is determined by the temperature of the substrate .
the new effects are visible already at distances @xmath84 , where experiments are now becoming available @xcite .
in this section we discuss in details the large @xmath4 behaviour @xcite of the out of equilibrium force ( [ fullnoneq ] ) . after the substitution @xmath85 , eq.([explicit ] ) becomes @xmath86 where @xmath87 due to the presence of the exponential @xmath88 in eq.([oifwc ] ) , it is possible to show that only the region @xmath89 contribute to the large @xmath4 behaviour of the force that in such limit exhibits the non trivial asymptotic behaviour @xmath90 this force exhibits a slower @xmath91 decay with respect to the one holding at thermal equilibrium where it decays like @xmath92 ( see eq .
( [ neweqasymp ] ) ) . in the above equation
we have introduced the low @xmath93 expansion of eq.([oifjwc ] ) @xmath94 } \ ; \frac { 2+|\varepsilon(\omega)-1|}{|\varepsilon(\omega)-1|}. \label{f}\ ] ] result ( [ ld ] ) and ( [ f ] ) provide the large distance behaviour ( @xmath95 ) of the force ( [ explicit ] ) where the only assumption made was the condition ( [ condition ] ) . due to
the presence of the bose factor the force ( [ ld ] ) depends on the optical properties of the substrate at frequencies of the order of @xmath96 . for temperatures much smaller than @xmath97 , where @xmath98 is the lowest characteristic frequency of the dielectric substrate , only the static value of the dielectric function is relevant and so we can replace @xmath99 with its low frequency limit in eq.([oifwc ] ) .
in this limit @xmath100 is different from zero only for @xmath101 , and after the @xmath102 expansion eq.([oifwc ] ) becomes @xmath103 where we performed the change of variable @xmath104 and replaced @xmath100 with its @xmath102 expansion @xmath105 . for @xmath106 where @xmath107 is the thermal wavelength
, we can extend the upper limit of integration on @xmath108 to @xmath109 and so we obtain that the force ( [ fullnoneq ] ) felt by the atom approaches the asymptotic behaviour @xmath110 result ( [ levlimit ] ) holds at low temperature with respect to the first dielectric function resonance ( @xmath111 ) and at distances satisfying the condition ( [ pfei ] ) calculated at the relevant temperatures @xmath14 and @xmath15 . eq.([levlimit ] ) shows that , at large distances , the new force is attractive or repulsive depending on whether the substrate temperature is higher or smaller than the environment one .
furthermore , it exhibits a stronger temperature dependence with respect to equilibrium force ( [ neweqasymp ] ) , contains explicitly the planck constant and has a @xmath91 distance dependence ) of the force from the general equation ( [ explicit ] ) , it is possible to produce a more direct derivation assuming from the very beginning that one can neglect _ absorption _ and dispersion of the dielectric function of the substrate @xcite . ] .
the new dependence of @xmath112 on temperature and distance can be physically understood by noticing that the main contribution to the @xmath4 dependent part of the electric energy @xmath113 arises from @xmath102 .
such values of @xmath93 correspond to the component of the black - body radiation impinging on the surface from the dielectric side in a small interval of angles , of order of @xmath114 , near the angle of total reflection . this radiation creates slowly damping evanescent waves in vacuum . as
a result @xmath112 turns out to be , in accordance with eq.([levlimit ] ) , of order of @xmath115 , where @xmath116 is the energy density of the black - body radiation .
equation ( [ levlimit ] ) holds for a dielectric substrate where @xmath8 is finite . if we want to find the large distance limit for a metal we should use equation ( [ ld ] ) . in the limit of small values of @xmath117 we can use the drude model .
as only frequencies @xmath118 contribute , one can substitute in eq.([f ] ) @xmath119 , the real part @xmath120 remaining finite as @xmath121 .
than one finds @xmath122 , where @xmath123 is the electric conductivity , so that for a drude metal eq.([levlimit ] ) is replaced by @xmath124 where @xmath125 is the riemann function .
it is easy to show that eq.([levlimmetal ] ) is valid at the condition @xmath126
ultra - cold gases can provide a useful probe of the surface - atom force .
tool sensitive to the gradient of the surface - atom force is in fact the frequency shift of the center - of - mass oscillation of a trapped bose - einstein condensate @xcite . on the other hand , experiments based on bloch oscillations are _ interferometric _ tools sensitive to the force itself @xcite .
finally one could also think to interference experiments involving the macroscopic phase of bose - einstein condensates in a double well potential @xcite .
the position of the corresponding interference fringes are sensitive to the surface - atom potential . in the last part of this paper
we discuss the first two above mentioned experiments .
bose - einstein condensed gases @xcite are very dilute , ultra - cold samples characterized by unique properties of coherence and superfluidity .
the study of the collective oscillations @xcite of a bose - einstein condensate provides a useful probe of the surface - atom potential .
in fact it is possible to measure with great accuracy the frequency of the center - of - mass motion @xmath127 of a condensate . for a harmonically
trapped condensate the frequency @xmath127 corresponds to the harmonic trap frequency @xmath128 , where @xmath4 is the direction of the oscillations .
thus if a bec in a harmonic trap is placed at distance @xmath129 from the surface of a substrate , the surface - atom potential @xmath130 perturbs the trap potential and produces a shift in @xmath127 . in the limit of small oscillations ( for a complete analysis
see @xcite ) , such a shift is @xmath131 where @xmath132 is @xmath133 column density of the gas ( density integrated over the directions perpendicular to the direction of oscillation ) and @xmath134 is the thomas - fermi radius in the @xmath4 direction column density is easily evaluated in the so called thomas - fermi approximation where @xmath135 @xcite . ] .
therefore measuring @xmath127 it is possible to extract the surface - atom potential @xmath130 @xcite . in fig .
[ fig:2](*a * ) we plotted , as a function of the surface - condensate separation @xmath129 , the relative frequency shift @xmath136 for the center - of - mass oscillations of a @xmath80rb condensate close to a sapphire substrate .
in such a calculation we used the surface - atom potential corresponding to the force ( [ fullnoneq ] ) . )
of the center - of - mass oscillation of a be condensate ( @xmath137 , @xmath138hz ) and ( * b * ) relative shift of the bloch oscillation period of a degenerate fermi gas ( [ ojeopi ] ) , out of thermal equilibrium ( [ feqnnn]).,title="fig:",scaledwidth=49.0% ] ) of the center - of - mass oscillation of a be condensate ( @xmath137 , @xmath138hz ) and ( * b * ) relative shift of the bloch oscillation period of a degenerate fermi gas ( [ ojeopi ] ) , out of thermal equilibrium ( [ feqnnn]).,title="fig:",scaledwidth=49.0% ] when an external force @xmath139 is applied to a particle trapped in a periodical potential , the particle undergoes oscillations in momentum space ( the bloch oscillations ) . during this oscillations the particle quasi - momentum @xmath140 evolves according to @xmath141
this is what happens for example in a sample of ultra - cold atoms trapped in a @xmath133 optical lattice aligned along the vertical direction .
bloch oscillations produced by the effect of the gravity force @xmath142 have a period @xmath143 where @xmath144 is the lattice wave - length and @xmath145 is gravity acceleration .
if now a surface is brought close to the atomic sample the additional surface - atom force @xmath146 perturbs the gravitational potential and affects the dynamics of the bloch oscillations ( for a complete analysis see @xcite ) .
in particular it produces a shift of the period @xmath147 . in figure
[ fig:2](*b * ) we plotted the relative shift @xmath148 for different thermal configuration as the distance @xmath129 between the center of a cloud of @xmath149k fermionic atoms ( @xmath150@xmath82 ) and the surface of a sapphire substrate is varied .
we used also the approximation of a small cloud of fermi atoms , for which @xmath151 in the range of distances plotted in fig .
[ fig:2](*b * ) this approximation provides results in good agreement with the exact calculation @xcite that takes into account real experimental parameters of the gas . + it is worth noticing that both effects of the surface - atom force out of thermal equilibrium described in the last section , and plotted in figures [ fig:2](*a * ) and [ fig:2](*b * ) , are in the domain of the present experimental accuracy @xcite .
new results presented in this paper have been obtained in collaboration with l.p .
pitaevskii and s. stringari .
we are grateful to a. recati , i. carusotto , c. henkel , m.s .
toma@xmath152 , e. cornell , j. obrecht , d.m .
harber , j. mcguirk and s. reynaud for useful comments .
10 sukenik c i _ et al .
_ 1993 _ phys .
lett . _ * 70 * 560 ; shimizu f 2001 _ phys . rev . lett _ * 86 * 987 ; h.oberst _ et al .
_ 2005 _ phys . rev .
a _ * 71 * 052901 ; druzhinina v and dekieviet m 2003 _ phys .
lett . _ * 91 * 193202 | the surface - atom casimir - polder - lifshitz force out of thermal equilibrium is investigated in the framework of macroscopic electrodynamics .
particular attention is devoted to its large distance limit that shows a new , stronger behaviour with respect to the equilibrium case .
the frequency shift produced by the surface - atom force on the the center - of - mass oscillations of a harmonically trapped bose - einstein condensate and on the bloch oscillations of an ultra - cold fermionic gas in an optical lattice are discussed for configurations out of thermal equilibrium . |
cepheid variable stars play a central role in the calibration of the extragalactic distance scale and the hubble constant ( 0 ) .
the measurement of cepheid distances to a dozen galaxies with the has led to dramatic improvements in the calibration of the several secondary distance indicators , and narrowed the longstanding discrepancy between the long " and short " distance scales from @xmath6 km s@xmath7 mpc@xmath7 a few years ago to @xmath8 km s@xmath7 mpc@xmath7 currently ( e.g. , saha 1997 ; madore 1997 ) .
the completion of the key project on the extragalactic distance scale will double again the number of galaxies with cepheid distances , furnish firm calibrations of the tully - fisher ( t - f ) , surface brightness fluctuation ( sbf ) , planetary nebula ( pnlf ) and globular cluster luminosity functions , sn ii expanding parallax , and sn ia secondary distance indicators , and determine 0 to an accuracy of @xmath010% ( kennicutt , freedman , & mould 1995 ) . as the uncertainty in the distance scale as a whole
is reduced , systematic uncertainties in the cepheid distances themselves become increasingly important , because errors in the cepheid scale propagate with nearly full weight into the secondary distance ladder and 0 .
of particular concern is whether the cepheid period - luminosity ( pl ) relation varies systematically with metal abundance .
the metal abundances in the key project galaxy sample vary by nearly an order of magnitude ( zaritsky , kennicutt , & huchra 1994 ; hereafter zkh ) .
current calibrations of the pl relation are referenced to the large magellanic cloud , which has a lower metal abundance ( [ o / h ] @xmath9 ) than is often found in more luminous galaxies in the key project sample .
thus a significant @xmath4-dependence to the pl relation could introduce a systematic error in the overall distance scale , as well as an increased scatter in individual distances .
several theoretical and observational studies argue for a significant metallicity dependence , though the magnitude of the effect remains controversial .
theoretical models by chiosi , wood , & capitanio ( 1993 ; hereafter cwc93 ) predict a small change in the pl zeropoint at visual and near - infrared wavelengths , of order 0.1 mag per factor of ten increase in abundance , but earlier calculations predicted a stronger effect ( e.g. , stothers 1988 ) , and the theoretical models remain uncertain ( 2 ) .
evidence for larger effects have come from several recent empirical studies .
freedman & madore ( 1990 ) undertook the first direct observational test of the metallicity dependence of the pl relation , using @xmath10 photometry of cepheids in three fields in m31 .
they derived a change in distance modulus of @xmath11 , a difference which they did not regard as significant .
their data were subsequently reanalyzed by gould ( 1994 ) , who derived steeper dependences of @xmath12 and @xmath13 , depending on the bandpasses used in the fitting .
recently beaulieu ( 1997 ) and sasselov ( 1997 ) analyzed eros data for large samples of cepheids in the lmc and smc , and derived a change in inferred distance modulus of @xmath14 .
kochanek ( 1997 ) derived a similar dependence of @xmath15 , using published cepheid measurements of 17 galaxies .
most of the dependence arises from increased line blanketing and redder intrinsic colors at higher abundance , an effect that has been reported in galactic and magellanic cloud cepheids by several authors ( e.g. , caldwell & coulson 1986 ; laney & stobie 1994 , and references therein ) .
all of these results are consistent with a metallicity dependence that tends to cause the distances of metal - rich cepheids to be underestimated .
however the magnitudes of these claimed dependences range over an order of magnitude , so the significance of any @xmath4-dependence for the distance scale remains unclear .
a zeropoint change at the level of @xmath160.3 would have minimal effects on the ultimate distance scale , because the cepheid target galaxies observed with bracket a large metallicity range with a mean value close to that of the lmc cepheids that calibrate the pl relation ( 6 ) .
however dependences as large as those derived by gould ( 1994 ) would introduce systematic errors in 0 at the 1020% level , and constitute the dominant systematic error in the entire extragalactic distance ladder .
as pointed out by sasselov et al ( 1997 ) and kochanek ( 1997 ) , a strong metallicity dependence might account for part of the difference between low values of 0 determined from sn ia and other secondary indicators , if the cepheid abundances in the sn ia hosts are systematically lower .
such effects might also partly explain the longstanding difference in cepheid and rr lyrae distance scales in nearby galaxies .
it is imperative to test whether systematic errors at this level are present in the cepheid data . as part of the key project we undertook a further test of the metallicity dependence of the pl relation , by targeting two fields in the giant sc i spiral m101 ( kennicutt 1995 ) .
m101 is especially well suited for a metallicity test , because it combines an unusually steep radial abundance gradient with a large disk scale length , making it possible to identify cepheids with nearly to the center of the galaxy .
the fields we targeted differ in metal abundance by a factor of five , and bracket the abundance range between the lmc and the most metal - rich cepheid fields observed with . in this paper
we present a _ differential _
analysis of the cepheid pl relations in the two fields .
our approach differs from that taken in most of the previous studies , insofar as we direct our attention solely to the effects of abundance on the @xmath1 and @xmath2 pl relations used in most of the cepheid observations .
we apply the same pl fitting methods and abundance calibrations that are applied to other key project observations , so that we can place direct limits on the effects of abundance variations on the cepheid distance scale , independent of any possible cepheid metallicity dependence at other wavelengths .
we have also compared cepheid distances to 23 galaxies , spanning a 50-fold range in abundance , with distances derived from three other methods , the tip of the red giant branch method ( trgb ) , the planetary nebula luminosity function ( pnlf ) , and the tully - fisher ( tf ) method , to further constrain the magnitude of any metallicity dependence .
the paper is organized as follows . in
2 we briefly summarize the effects of abundance on cepheid magnitudes and colors as expected from theory , and present a series of simulations that quantify the effects on the @xmath1 and @xmath2 pl fits used in the observations .
3 describes the differential test in m101 and the derived metallicity dependence .
we compare the results to previous work in 4 . in
5 we use trgb , pnlf , and tf distances to further constrain the metallicity dependence .
finally , in 6 we assess the impact of the dependence on the overall distance scale , and discuss the prospects for a definitive calibration of the @xmath4 dependence .
the effects of metal abundance on the cepheid instability strip have been investigated by stothers ( 1988 ) , stift ( 1990 ; 1995 ) , and cwc93 .
the main predictions are that metal - rich cepheids at a given period are more luminous and redder .
the net effects on the observed pl relations are strongly wavelength dependent ; metal - rich cepheids will appear fainter in the blue due to line blanketing , while in the red the sense of the effect is to make the cepheids appear brighter .
the magnitudes of the metallicity effects are also strongly wavelength dependent . in this paper
we are concerned solely with the behavior of the pl relations in @xmath1 and @xmath2 , which is most relevant to observations .
the methodology used to fit the pl relations observed in the key project is described in freedman ( 1994 ) and ferrarese ( 1996 ) .
briefly , the cepheid target fields are observed at 12 epochs in @xmath1 ( f555w ) , and at @xmath17 epochs in @xmath2 ( f814w ) .
the observed pl relations at @xmath1 and @xmath2 are fitted with slopes fixed to those of the calibrating pl relations in the lmc ( madore & freedman 1991 ) : @xmath18 @xmath19 these calibrations assume an lmc true distance modulus of 18.50 @xmath0 0.10 and an average reddening @xmath20 , corresponding to @xmath21 . any difference in the apparent @xmath1 and @xmath2 moduli is assumed to be caused by reddening , and the true modulus is computed assuming @xmath22 = 1.666 , or @xmath23 = 2.50 ( cardelli , clayton , & mathis 1989 ) .
this is the same value found by stanek ( 1996 ) .
there are several ways in which metallicity effects could propagate into the cepheid distances derived in this way . aside from directly influencing the absolute magnitudes of the cepheids ,
any systematic change in @xmath24 color will propagate through the reddening correction and alter the inferred true modulus .
cwc93 used analytical fits to their theoretical models to predict the magnitude of these effects for pl relations observed in @xmath25 and @xmath26 . for a mean period of 10 days ,
their models predict @xmath27 = @xmath2822.2@xmath29 from fitting pl relations in @xmath30 and @xmath1 , but only @xmath311.7@xmath29 when fitted in @xmath1 and @xmath2 , where @xmath4 is the metal mass fraction ( @xmath32 = 0.02 ) .
the @xmath26 dependence corresponds to a systematic error of only 0.06 mag in true modulus ( 3% in distance ) between @xmath33 , approximately the full range found in the cepheid galaxies observed with .
such a weak dependence would be virtually impossible to measure in the data .
the predicted effects on distances derived from photometry in @xmath30 and @xmath1 would be much larger , 44% ( or 0.8 mag ) for the same abundance range .
however there are other , more subtle ways in which abundance effects might propagate , for example in the way that the width of the cepheid instability strip and the colors of the blue and red edges are affected ( cwc93 ) . in order to incorporate these effects , we constructed simulated cepheid data sets for different metal abundances and foreground reddenings , and analyzed them using the standard key project reduction procedures .
simulated pl relations for a given metallicity @xmath4 were derived using the equations given in cwc93 .
cepheid masses were then randomly drawn from a mass distribution given by @xmath34 over the mass range @xmath35 .
luminosities were calculated from equation ( 25 ) of cwc93 , and the red and blue edges of the instability strip were derived using equations ( 7 ) and ( 13 ) of the same paper .
stars were randomly positioned between the red and blue edges of the strip , and periods were then calculated from the p
l t@xmath36 relation given in equation ( 4 ) of cwc93 .
we used fits to the ( @xmath37,@xmath38 ) and ( @xmath24 , @xmath38 ) relations to derive the apparent @xmath1 and @xmath2 magnitudes .
reddening was then added assuming @xmath39 and @xmath40 .
nine samples of 100 cepheids each were constructed , with metal abundances @xmath4 = 0.01 , 0.02 , and 0.03 , and reddening @xmath41 = 0.1 and 0.2 mag .
the simulated data sets were then fitted to pl relations in @xmath1 and @xmath2 , following the standard procedures described earlier .
the results are summarized in table 1 .
the first three columns of the table are the input values to the simulation , while the last four columns are derived quantities .
we then determined the coefficient in equation ( 3 ) by minimizing the difference between @xmath42 and @xmath43 .
@xmath44 throughout this paper we will use the notation @xmath45 to denote the change in inferred true distance modulus per factor of ten in metal abundance ( cf .
beaulieu 1997 ; kochanek 1997 ) , with a subscript indicating the relevant wavelengths over which the pl relation is fitted ( @xmath1 and @xmath2 unless indicated otherwise ) .
thus for our simulations @xmath46 .
the metallicity effect derived in our simulations is slightly larger than the analytical approximation of cwc93 , but is still very slight , amounting to a distance error of less than 6% for an order of magnitude change in metal abundance .
the sign of the change is in the sense that the distance inferred at higher metal abundances is smaller ( cepheids appear brighter ) .
unfortunately the theoretical prediction is not robust .
for example , the theoretical prediction of the position of the red edge of the instability strip is poorly defined by theory , because of large uncertainties in the theory of convective transport .
more recent calculations by chiosi , wood , & capitanio ( 1997 ) , using new opacities , predict @xmath47 = @xmath280.06 , a smaller effect but one in the opposite sense .
although it is reassuring that the metallicity dependence predicted by theory is small , we can not yet use theory as the basis of a correction to observed distance moduli . moreover it would be very useful to have independent observational confirmation of the magnitude of the metallicity effects , if not an empirical calibration of the dependence .
the magnitude of the @xmath4-dependence of the pl relation can be measured directly , by observing cepheids with different metal abundances in the same galaxy .
this approach was introduced by freedman & madore ( 1990 ) , who applied the method to m31 using groundbased @xmath10 ccd photometry . as part of the key project , we observed cepheids in two fields in the giant sc galaxy m101 .
the inner " and outer " fields are centered at 14@xmath4803@[email protected] , @xmath28542135.7 and 14@xmath4802@[email protected] , @xmath28541758.3 ( 2000.0 ) , at galactocentric radii of 1.7 and 7.9 , respectively .
figure 1 shows a groundbased image of m101 with the inner and outer wfpc2 field locations superimposed .
the disk abundances in m101 are among the best studied of any galaxy , thanks to several hii region surveys ( kennicutt & garnett 1996 and references therein ) .
the hii region abundances are measured in terms of oxygen , and we will parametrize the cepheid metallicity dependence in terms of the nebular oxygen scale using the usual notation : [ o / h ] @xmath51 , and adopting @xmath52 .
we adopt this convention because metal abundances for most of the key project cepheid fields have been measured from hii region spectra ( zkh ) .
the abundances of the cepheids themselves could differ slightly from those of the hii regions , either because the stellar iron and oxygen abundances do not scale precisely , or from a small calibration offset between the stellar and interstellar abundances .
neither of these effects is likely to be important , however , because the long - period cepheids observed with ( @xmath53 days ) arise from relatively massive ( @xmath54 ) and shortlived ( @xmath55 yr ) progenitors ( cwc93 ) , and variations in [ o / fe ] are likely to be small in this young population ( wheeler , sneden , & truran 1989 ) .
there may be a larger absolute shift between the nebular and stellar abundance scales , but this is unimportant for the present application , as long as we calibrate and apply the abundance test in a self - consistent manner across the key project sample .
figure 2 shows the hii region abundance distribution over the relevant range of radii in m101 , with the locations of the cepheid fields indicated by the horizontal lines .
the solid points are empirical " abundances based on the strengths of the reddening - corrected [ oii]@xmath563727 , [ oiii]@xmath564959,5007 , and h@xmath57 lines , taken from the survey of kennicutt & garnett ( 1996 ) , and calibrated following zkh .
the inner cepheid field contains three hii regions measured in this survey , as indicated by the open circles .
the outer cepheid field contains two bright hii regions that were not included in the kennicutt & garnett ( 1996 ) survey .
spectra for these objects have been obtained subsequently using the blue channel spectrograph on the multiple mirror telescope , and their abundances are also shown with open circles in fig .
2 . the errors in the empirical abundances ( shown in fig .
2 ) are dominated by systematic uncertainty in the calibration of the absolute abundance scale , especially at high metallicities , where the calibration rests entirely on theoretical photoionization models . fitting to the empirical abundance distribution yields a difference between the inner and outer cepheid fields of @xmath58 dex .
more robust abundances based on measured forbidden - line electron temperatures are available for a handful of hii regions in m101 , and these data are shown as open triangles in fig .
2 . the innermost point comes from kinkel & rosa ( 1994 ) , with the other data coming from kennicutt & garnett ( in preparation ) .
these data show a significant offset from the empirical abundances over part of the radial range , but the difference between the inner and outer fields is very similar , @xmath59 dex , depending on whether the outermost points ( ngc 5471 ) are included .
we adopt the average difference of @xmath60 dex .
thus the inner and outer cepheid fields span nearly a factor of five in abundance .
observations of the m101 cepheid fields were carried out in cycles 35 for the outer field and cycles 45 for the inner field .
details of the respective data sets , including complete discussions of the cepheid photometry , variable identification , and period determinations can be found in separate papers by kelson ( 1996 ) and stetson ( 1997 ) for the outer and inner fields , respectively .
several technical obstacles prohibit a direct comparison of the inner and outer field photometry in their entirety .
the bulk of the observations of the outer m101 field were obtained before the refurbishment mission with the original wide field camera ( wfc ) , whereas the inner field was observed entirely in cycles 45 with wfpc2 . uncertainties in cross - calibrating wfc and wfpc2
could easily mask any real difference in the magnitudes and colors of the cepheids ( cf .
kelson 1996 ) .
fortunately we were able to obtain several exposures of the outer field with wfpc2 , and these form the basis of the differential photometry analyzed here . to complicate matters further ,
most of the wfpc2 observations were obtained early in cycle 4 , when the operating temperature of the camera was adjusted to reduce charge transfer efficiency problems ( holtzman et al .
the wfpc2 data set includes a mix of warm " and cold " observations , and the zeropoint calibrations of these data may differ .
the combined uncertainties in tying together the wfc vs wfpc2 data and the warm vs cold wfpc2 data could easily amount to several hundredths of a magnitude in @xmath61 color , and mask ( or mimic ) any signature of a metallicity effect . to avoid these problems we adopted a completely differential approach to the photometry of the inner and outer fields .
our wfpc2 observations include 5 pairs of visits , when exposures of both fields were taken with the same filter , focus , and camera temperature .
these epochs ( 3 in @xmath1 and 2 in @xmath2 ) were used to define a common photometric scale for both fields .
table 2 lists the journal of observations for this subset of the data .
the other ( extensive ) observations of each field were used to identify the cepheid variables and measure their light curves , periods , and phases ( kelson 1996 ; stetson 1997 ) , but the magnitude scales used in this paper were determined solely from the 5 wfpc2 epochs listed in table 2 .
the primary set of photometry was performed with the daophot / allframe package ( stetson 1994 ) .
details of the reduction and calibration procedures can be found in hill ( 1997 ) and stetson ( 1997 ) .
the inner and outer field data were reduced using identical point spread functions , aperture corrections , and zeropoint calibrations applied to the two data sets , to minimize the potential for any systematic photometric offset . as a check on our results we performed an independent reduction of the photometry using the dophot package ( schecter , mateo , & saha 1993 ) . to simplify the presentation
we will first describe the main allframe analysis , followed by a summary of the final results from both reductions .
cepheid variables were identified independently for the outer and inner fields by kelson ( 1996 ) and stetson ( 1997 ) , respectively .
of the 29 cepheids identified in the outer wfc field , 26 fell within the field of our wfpc2 images and were cleanly resolved in our allframe data .
the periods of these stars range from @xmath62 days . for the inner field we identified a sample of 50 high - quality cepheids with large amplitudes , well - determined light curves , and periods in the range @xmath63 days , and which appear to be free of crowding ,
as determined by the psf fits and from visual inspection of the wfpc2 images ( stetson 1997 ) .
the final product of the allframe run was 5 @xmath1 and 2 @xmath2 magnitudes for each star in the outer field , and 4 @xmath1 and 2 @xmath2 magnitudes for the inner field ( some epochs contained cosmic ray split exposures ) .
these were averaged to produce intensity - weighted mean magnitudes for each cepheid .
the average photometric errors returned by allframe for the stars in this magnitude range are @xmath64 mag in @xmath1 and @xmath2 , and as low as @xmath65 for the brighter cepheids .
the largest source of random error in the magnitudes is the variability of the cepheids themselves .
because the data were taken at a small number of random phases , our average magnitudes will show a considerable scatter about the true mean values .
for the inner field data , we were able to reduce this problem by phase matching the observations to the full light curves and determining true mean magnitudes ( but on the photometric zeropoint of the epochs in table 2 ) .
this was not done for the outer field cepheids , because of the long interval between the main wfc observations that define the light curves , and the larger photometric uncertainties in the wfc magnitudes .
we confirmed that this procedure does not introduce any systematic shift in the pl zeropoints .
the resulting pl relations for the inner and outer fields are shown in figures 3 and 4 , respectively . superimposed in each case
are the calibrating pl relations for the lmc , shifted to the best fitting distance modulus .
least squares solutions , constrained to the slopes of the lmc pl relations , yield @xmath1 and @xmath2 moduli for the inner field of @xmath66 mag and @xmath67 mag .
the corresponding solutions for the outer field are @xmath68 mag and @xmath69 mag .
note that these solutions are based solely on the subset of wfpc2 observations listed in table 2 , and are strictly applicable only for the differential comparison of the inner and outer fields presented here .
fully calibrated magnitudes and pl fits for the fields can be found in kelson ( 1996 ) and stetson ( 1997 ) . in the standard fitting procedure used in the key project , the difference in @xmath1 and @xmath2 moduli is assumed to represent the average reddening of the cepheids ; hence @xmath70 mag for the inner field and @xmath71 mag for the outer field .
applying the cardelli ( 1989 ) reddening curve yields for the true moduli @xmath72 mag for the inner field and @xmath73 mag for the outer field .
these errors include only the uncertainty introduced by the dispersion of the pl relations ( the errors in true modulus are comparable to those in @xmath1 and @xmath2 individually because the residuals in the two bandpasses are correlated ) .
the errors in the absolute distance to m101 would be larger , because of additional uncertainties in the photometric zeropoints , the lmc distance and reddening , and other systematics , but these cancel out in the differential comparison .
table 3 summarizes the results of these fits .
the results quoted here refer to the complete cepheid samples in each field , combined for the four wfpc2 detectors .
the fits for the complete cepheid samples yield a difference in inferred true distance modulus of @xmath74 mag , in the theoretically predicted sense ( metal - rich closer ) .
a possible source of concern in this comparison is the difference in the cepheid period distributions between the two fields .
most of the cepheids in the outer field have periods longer than 16 days ( log @xmath75 ) , whereas the inner field cepheids are more heavily weighted to the period range log @xmath76 .
the absence of short - period variables in the outer field is entirely an observational artifact , caused by interruptions in the observing sequence combined with the brighter magnitude limit in the wfc data .
this difference in periods could introduce a spurious difference in distance moduli between the two fields , if the slopes of the pl relations differ significantly from the canonical lmc values used in the fitting . to minimize any bias from this effect we refitted the pl relations , restricting the data to cepheids with log @xmath75 ;
this includes 24 of the 26 cepheids in the outer field and 30 of the 50 cepheids in the inner field . in addition
, we excluded 2 cepheids in the inner field with very red colors ( @xmath77 ) .
fits to these long - period subsamples yield true moduli for the inner and outer fields of 29.21 @xmath0 0.09 mag and 29.34 @xmath0 0.08 mag , respectively , or an inner @xmath31 outer difference of @xmath78 mag ( table 3 ) .
the average distance modulus we derive for the outer field , 29.36 @xmath0 0.08 mag , is in excellent agreement with the value 29.34 @xmath0 0.17 mag ( including all error terms ) derived from the original photometry of the m101 cepheids by kelson ( 1996 ) .
the outer field abundance is nearly the same as the lmc , which calibrates the pl relation , so its distance modulus is the appropriate one to apply to m101 .
the same analysis was performed using the dophot photometry package ( schechter , mateo , & saha 1993 ) .
a list of cepheid candidates was generated independently , and cross - checked with the final allframe list , yielding 34 cepheids in the inner field and 24 in the outer field .
the pl relations were then fitted following the same procedures .
fitting the complete cepheid sets ( @xmath79 days ) yielded a inner @xmath31 outer difference in true distance modulus of @xmath80 mag ( _ vs _ @xmath74 mag for allframe ) .
a comparison using only cepheids with @xmath81 days yields a difference of @xmath82 mag ( _ vs _ @xmath78 mag for allframe ) .
the two analyses yield nearly identical values for the metallicity effect in m101 . at first glance
it might appear that the agreement is _ too _ good , given the relatively large uncertainties , but recall that the same stars are being measured , and the quoted uncertainties largely reflect other factors , such as the intrinsic dispersion of the pl relation , which affect both data sets uniformly .
however the consistency of the dophot vs allframe comparison does offer some assurance that the apparent difference in pl relations between the inner and outer fields is not an artifact of field crowding errors in the data reduction .
we adopt for our final result the average of these fits , for a net difference of @xmath83 mag .
combining this with the observed difference in oxygen abundance of @xmath280.68 @xmath0 0.15 dex ( 3.1 ) gives a metallicity dependence : @xmath84 the quoted uncertainty includes errors in the cepheid photometry and in the abundance gradient .
note that the measured @xmath4-dependence is only significant at the 1.5 @xmath85 level .
theoretical models predict that the cepheid metallicity dependence is produced by a combination of brighter mean magnitudes and redder colors with increasing metallicity ( e.g. , cwc93 ) .
observational evidence for systematic changes in the magnitudes and/or colors of cepheids with abundance has been offered by several authors , including caldwell & coulson ( 1986 ) , gieren ( 1993 ) , laney & stobie ( 1994 ) , sasselov ( 1997 ) , and kochanek ( 1997 ) . with our @xmath1 and @xmath2 photometry alone , it is impossible to separate the effects of changes in intrinsic luminosity , intrinsic color , and reddening .
however we can use the m101 results along those from other cepheid data sets to place useful limits on these contributions .
the nature of the shift in the pl relation between the m101 inner and outer fields is illustrated in figure 5 , which shows the residuals of individual cepheids from the best fitting allframe @xmath1 and @xmath2 pl relations in each field .
the top and middle panels show the residuals for cepheids in the outer and inner fields , respectively , relative to the respective pl fits in those fields .
the solid line in each case shows the expected trajectory due to variations in temperature ( position in the instability strip ) , while the dashed line shows the ( nearly degenerate ) trend expected from reddening variations .
the behavior of the residuals in m101 is similar to that seen in all of the key project data sets , insofar as most of the dispersion in the pl relations can be attributed to the finite strip width and to reddening variations .
the remaining scatter about these correlations are presumably due to a combination of photometric errors and uncertainties in deriving the mean magnitudes from a small number of epochs .
the bottom panel of figure 5 shows the @xmath1 and @xmath2 residuals for stars in the inner field , but in this case with respect to the mean pl relations in the _
outer _ field .
the outer field cepheids have nearly the same abundances as the lmc cepheids that calibrate the pl relation , so any metallicity dependence in the pl relation will appear as a systematic residual in this diagram .
the inner field cepheids actually show two distinct differences .
first there is a general shift of the distribution to fainter magnitudes in both @xmath1 and @xmath2 , along the reddening / temperature trajectory .
this we tentatively attribute to a higher mean reddening in the inner field .
in addition there is a general shift of stars below the reddening line , which implies that the mean colors of the inner field cepheids are redder ( or else brighter ) than would be expected from reddening effects alone .
this represents the shift of @xmath86 mag in true distance modulus , transformed into the residual plane .
the clear appearance of this shift in figure 5 confirms that the difference in derived distance moduli between the inner and outer fields is significant , and is not merely an artifact of scatter in the pl relations .
we can not distinguish from @xmath1 and @xmath2 data alone whether the shift in distance modulus is produced because the metal - rich cepheids are brighter at both @xmath1 and @xmath2 ( with no color change ) , because the metal - rich stars are intrinsically redder , or from some combination of luminosity and color changes .
however the m101 data place bounds on the variation in either extreme .
the observed difference in true modulus of 0.16 mag could be produced at constant color by a shift in both @xmath1 and @xmath2 moduli of @xmath310.16 mag ( trivially ) , or by a shift in intrinsic color of @xmath280.08 mag .
since the abundance difference in the two fields is @xmath87 dex , the implied metallicity dependences would be @xmath88 in luminosity ( @xmath26 ) , or @xmath89 in @xmath61 color . in all likelihood some combination of these effects is involved .
note that the observed color excess in the inner field , @xmath90 mag , is roughly twice as large as any color difference that can be attributed to metallicity effects .
this means that part of the color difference between the cepheids in the inner and outer fields must be due to interstellar reddening .
another way to constrain the effects of @xmath4-induced color changes on cepheid distances is to test whether the observed cepheid colors correlate systematically with metal abundance .
kochanek ( 1997 ) found evidence for such a correlation , based on published photometry for cepheids in 17 galaxies .
we show an updated version of this test in figure 6 .
each point shows the average color excess for cepheids in a given galaxy ( or field ) , plotted as a function of metal abundance .
the color excesses are derived from pl fits in @xmath1 and @xmath2 ( equations [ 1 ] and [ 2 ] ) , with the galactic foreground reddening subtracted ( burstein & heiles 1984 ) .
the error bars signify the dispersion in the reddenings of individual cepheids .
data are taken from kochanek ( 1997 ) , along with data for ic 1613 ( freedman 1988 ) , the two m101 fields from this paper ( open circles ) , and the three fields in m31 ( freedman & madore 1990 ) , shown as crosses .
we excluded galaxies in the kochanek sample with color excesses derived at wavelengths other than @xmath1 and @xmath2 , and galaxies without measured hii region abundances .
this leaves a sample of 19 fields in 16 galaxies .
figure 6 shows a clear trend between cepheid color excess and metallicity , confirming kochanek ( 1997 ) .
the best fitting slope to our data is @xmath91 = 0.12 \pm 0.08 $ ] ( a linear fit in @xmath4 yields virtually the same slope over the relevant abundance range ) .
this color gradient is similar to what is inferred above from the m101 inner and outer fields , but the trend in figure 6 includes any changes in mean interstellar reddening with metallicity .
it is interesting that the color trend is restricted to galaxies with abundances higher than that of the lmc ( [ o / h ] = @xmath92 ) , where one might expect a higher dust - to - gas ratio and thus a higher reddening .
the absence of a color trend at lower abundances suggests that interstellar reddening may well be responsible for much if not most of the observed color trend . even if all of the trend in fig .
6 were due to metallicity effects , it would introduce at most a bias in the distance modulus @xmath93 .
the first application of the differential metallicity test for cepheids over a range of galactocentric distances was carried out for m31 by freedman & madore ( 1990 ; hereafter fm90 ) .
fm90 analyzed random - phase @xmath10 observations for 38 cepheids in three fields in m31 , located at radii of 3 , 10 , and 20 kpc .
no difference in true distance modulus was detected for the 3 and 10 kpc fields , but the derived true modulus for the 20 kpc field was higher by @xmath94 mag , when determined from the @xmath1 and @xmath2 data alone . combining this with an estimated abundance gradient of 0.75 dex over 320
kpc yielded a metallicity dependence @xmath95 and @xmath96 .
fm90 did not solve for @xmath45 explicitly , and the numbers listed here were derived from the pl solutions and abundance gradients given in their paper , following gould ( 1994 ) . in view of the low statistical significance of this result , fm90 interpreted it as a nondetection , and consistent with no measureable @xmath4 dependence of the pl zeropoint .
the m31 data were subsequently reanalyzed by gould ( 1994 ) , who used the correlations between magnitude residuals in @xmath10 in an attempt to separate the effects of metallicity and reddening .
his best solutions for the same data yielded values @xmath97 and @xmath98 , a much larger and statistically significant dependence .
these results can not are not directly comparable to ours , because the wavelength baseline extends to the blue , where the metallicity effects are expected to be larger .
however the analysis illustrates the large range of dependences which are consistent with the m31 data .
unfortunately both of these analyses were based on an erroneous value for the abundance gradient in m31 .
fm90 quoted an abundance range of 0.75 dex over the three cepheid fields ( 1.7 to 0.3 solar ) , citing blair , kirshner , & chevalier ( 1982 ) , but the abundance gradient given in the latter paper corresponds to a range of only 0.44 dex , when placed at the fm90 distance . adding data from dennefeld & kunth ( 1981 ) and recalibrating the oxygen abundances on the zkh scale
lowers the abundance range further to 0.31 @xmath0 0.16 dex over the 3 cepheid fields ( from 1.8 to 0.9 solar ) . applying this correction to the fm90 pl solutions changes their metallicity dependence to @xmath99 , with the larger fractional error reflecting the additional uncertainty in the abundance gradient .
likewise the revised gould ( 1994 ) solutions become @xmath100 and @xmath101 .
we conclude , in agreement with fm90 , that the existing m31 data do not offer very stringent constraints on the cepheid metallicity dependence . ongoing efforts to obtain more precise photometry over a wider wavelength baseline may improve the constraints ( kaluzny 1997 ; freedman 1997 ) . a different approach
to constraining the metallicity dependence has been employed by beaulieu ( 1997 ) , sasselov ( 1997 ) and kochanek ( 1997 ) .
beaulieu and sasselov have compiled a large database of photometry of lmc and smc cepheids from the eros microlensing project , and applied a differential multivariate analysis to determine the maximum likelihood change in distance moduli and color with metallicity .
they derive @xmath102 , this was derived by transforming photometry made at slightly shorter wavelengths ( corresponding approximately to johnson @xmath1 and @xmath103 ) , so the dependence might be slightly steeper than one would derive from @xmath1 and @xmath2 photometry .
kochanek ( 1997 ) has applied a similar approach to cepheid data for 17 galaxies measured from the ground and with , solving independently for residual extinction and metallicity - dependent colors and luminosities .
his maximum likelihood solution yields a mean change in magnitude at @xmath1 and @xmath2 of @xmath104 and a @xmath61 color change of 0.13 @xmath0 0.04 .
taken together these would introduce a metallicity dependence in inferred true modulus of @xmath105 .
these dependences are larger than the @xmath106 derived here from our m101 observations , but the results are consistent within their quoted uncertainties .
the metallicity sensitivity of the cepheid distance scale can be tested externally if independent distances are available for galaxies spanning a large range in abundance .
the tip of the red giant branch ( trgb ) method offers a strong test of the cepheid scale over a large abundance baseline .
two other secondary distance indicators , the planetary nebula luminosity function ( pnlf ) and the tully - fisher ( tf ) method , provide indirect constraints on the magnitude of the @xmath4 effects as well .
the trgb method utilizes the nearly constant luminosity of the red giant branch tip in the @xmath2 band ( @xmath107 ) .
globular cluster observations by da costa & armandroff ( 1990 ) indicate that the tip luminosity remains constant over a large range of metallicity ( @xmath108 < -0.7 $ ] ) , and theoretical isochrones show a similar constancy over the same abundance range , and for ages of 2 @xmath31 15 gyr ( lee , freedman , & madore 1993 ; madore , freedman , & sakai 1997 , and references therein ) .
the method has been applied to 9 galaxies with independently determined cepheid distances , as summarized in table 4 .
included are cepheid and trgb distances with their associated uncertainties and the metal abundances of the galaxies .
we list two sets of cepheid distances , the published values and those determined from the pl relations at @xmath1 and @xmath2 alone , if different .
testing for a @xmath4 dependence in the @xmath26 cepheid data alone is much cleaner , as it avoids problems introduced by a wavelength - dependent metallicity dependence , though the distance moduli determined from @xmath1 and @xmath2 alone are often less precise .
the abundances listed in table 4 are determined from hii regions ( skillman , kennicutt , & hodge 1989 ; zkh ) , and apply to the cepheids ( not the red giants ) , because we are testing for metallicity effects in the population i cepheids .
the galaxies span 7 magnitudes in absolute magnitudes and a factor of 50 in ( pop i ) metal abundance , thus providing us with a sensitive test for metallicity effects in the cepheids .
the red giant abundances of the galaxies all lie within the calibrated range of the trgb method , where metallicity effects in the trgb distances should be small .
we also include an indirect comparison for two spirals in the leo i group , ngc 3351 and ngc 3368 .
these galaxies have cepheid distances measured from graham ( 1997 ) and hjorth & tanvir ( 1997 ) , respectively , and a trgb distance determined for an elliptical member of the group , ngc 3379 , from sakai ( 1997b ) . in these cases
the comparison may be influenced by group depth , so the data should be accorded lower weight .
the trgb distances for these objects are marked in parentheses in table 4 .
the absolute distances measured with the two methods show good agreement , with an rms scatter of @xmath00.16 mag ( 8% in distance ) .
figure 7 shows the residuals plotted as a function of metal abundance ( ngc 3351 and ngc 3368 are indicated with open circles ) . there is a slight trend with abundance , and a least squares fit yields a best fitting slope @xmath109 , shown by the solid line in figure 7 . comparing to the more heterogeneous set of published cepheid distances (
not all at @xmath1 and @xmath2 alone ) yields a nearly identical fit ( @xmath110 ) .
one might worry that these trends could be influenced by a metallicity dependence in the trgb distances ( cf .
saleris & cassissi 1997 ) .
this is very unlikely , however , because the red giant abundances in these galaxies are virtually uncorrelated with the metallicites of the cepheid fields , and changing them does not appreciably influence the slope of the relation in figure 7 . as a check we recalculated the trgb distances using a steeper @xmath4 dependence for the rr lyrae luminosity scale , @xmath111 = 0.30 $ ] instead of 0.17 as adopted in table 4 . repeating the comparison in figure 7 with the new trgb distances yielded nearly the same result as before , @xmath112 , and if anything produces a slightly weaker cepheid @xmath4 dependence .
lee ( 1993 ) directly compared trgb and cepheid distances as a function of giant branch abundance , and found no significant dependence on ( pop ii ) metallicity .
although our analysis has focussed on the @xmath4 dependence of the pl relation in @xmath1 and @xmath2 , seven of the galaxies in table 4 have cepheid distances based on @xmath10 fits , and we can check whether there is evidence for a significant metallicity dependence in those distances . repeating the comparision yields
very similar results , with @xmath113 . given the small sample and the inhomogeneity of the data set , we regard this as a tentative result .
however it is interesting that there is no indication in the trgb comparison for a strong metallicity dependence , even in the blue .
to summarize , the comparison of cepheid and trgb distances yields another marginal detection of a cepheid metallicity dependence , @xmath114 , which is formally consistent with our m101 results ( @xmath115 ) , as well as with the theoretically predicted dependence ( @xmath116 ) .
the dashed and dotted lines in figure 7 show the dependences that are expected for values of @xmath45 between @xmath117 and @xmath118 ; the strong dependences suggested by previous studies are not confirmed here .
these results , together with our m101 analysis , offer the strongest evidence for a weak metallicity dependence to the pl relations in @xmath1 and @xmath2 .
we can perform a similar check for galaxies with cepheid and pnlf distances .
direct comparisons are available for 7 galaxies , and indirect comparisons via galaxies in the same group or cluster are available for 4 other cepheid hosts .
the data are listed in table 4 , with the indirect comparisons in parentheses .
most of these data were taken directly from comparisons in soffner ( 1996 ) and feldmeier , ciardullo , & jacoby ( 1997 ) , with the addition here of abundance values appropriate to the cepheid fields .
figure 8 shows the cepheid @xmath31 pnlf distance modulus residuals as a function of metal abundance .
again there is an excellent correlation between the absolute distances ( @xmath00.19 mag rms ) , as shown previously by soffner ( 1996 ) and feldmeier ( 1997 ) .
this comparison is not as clean as the trgb test , because the galaxies cover a smaller range in metal abundance , and the pnlf method itself may have a metallicity dependence at the @xmath00.2 mag level ( ciardullo & jacoby 1992 ) . given these large uncertainties , the pnlf distances do not impose tight constraints on the cepheid metallicity dependence , and if anything figure 8 is suggestive of a modest @xmath4 dependence in the pnlf distances . as a final check we can use the slope of the local tully - fisher ( tf ) relation to constrain the magnitude of any cepheid @xmath4 dependence .
distances from the tf method are currently available for 15 cepheid calibrating galaxies . unfortunately a residual test of the sort applied to the trgb and pnlf distances can not be applied here , because the intrinsic dispersion of the tf relation is much larger , and because one of the parameters in the tf method , galaxy luminosity , is itself strongly correlated with disk metallicity ( garnett & shields 1987 ; zkh ) . as a result
any metallicity dependence in the cepheids will primarily act to change the apparent slope of the tf relation , rather than affect the scatter of the relation ( gould 1994 ) .
we can exploit this effect , however , and test whether the slope of a tf relation calibrated from cepheid distances is significantly shallower than the slope derived from observations of distant spirals , where relative distances are derived independently of the cepheid scale .
the magnitude of the slope change for a given cepheid @xmath4-dependence can be easily derived .
consider for example the @xmath2-band tf relation : @xmath119 if we define the metallicity dependence @xmath45 in the usual way , @xmath120}}\ ] ] and define @xmath57 as the slope of the absolute magnitude vs metallicity relation , @xmath121}\ ] ] then the effect of the cepheid @xmath4 dependence will be to alter the tf slope : @xmath122 we used the @xmath2-band tf analysis of giovanelli ( 1997 ) to derive representative values of @xmath123 and @xmath57 .
these authors compiled observations of 555 spiral galaxies in 24 rich clusters and applied incompleteness corrections to define the slope of the tf relation independent of the cepheid scale , yielding @xmath124 ( bivariate fit ) .
giovanelli also compiled @xmath2 magnitudes and linewidths for 15 galaxies with cepheid distances , including 12 with measured hii region abundances .
a least squares fit of absolute magnitude vs [ o / h ] yields a luminosity - metallicity slope @xmath57 = 4.8 @xmath0 1.0 .
this is similar to @xmath57 = 4.6 from zkh ( but using @xmath30 magnitudes ) . applying @xmath125 and @xmath126 to equation ( 8) then allows us to predict the tf slopes for the cepheid calibrating galaxies , for various values of @xmath45 .
the predicted changes in slope range from 5% ( to @xmath127 ) for @xmath128 to 18% ( @xmath129 ) for @xmath130 .
figure 9 shows the tf relation for the 15 cepheid calibrating galaxies ( giovanelli 1997 ) , with the predicted slopes superimposed .
the slope of the cepheid - derived tf relation is actually _ steeper _ than the cluster - calibrated relation .
a bivariate fit to the 12 galaxies with reliable linewidths ( giovanelli 1997 ) yields @xmath131 .
fitting all 15 galaxies yields an even steeper slope .
the slopes of the cepheid - calibrated relations thus are consistent with no @xmath4 dependence , though the large uncertainty in the slope , not unexpected for a small sample , again limits the usefulness of this test .
our slope also agrees with that derived by pierce & tully ( 1988 ) for the ursa major cluster ( @xmath132 ) , but their linewidths and magnitudes were calibrated using slightly different methods , so we can not compare our slope directly to theirs .
nevertheless the tully - fisher data offer further evidence against a large abundance effect in the cepheids ( @xmath133 ) .
it should be possible to tighten these limits as more cepheid distances are accumulated over the course of the key project .
we begin by summarizing the results of the various tests presented in this paper .
the direct comparison of pl relations for the inner and outer cepheid fields in m101 imply a difference in inferred true distance moduli of 0.16 @xmath0 0.10 mag ( inner field closer ) , over an abundance baseline of 0.68 @xmath0 0.15 dex .
this implies a metallicity dependence of distance modulus inferred from @xmath1 and @xmath2 pl relations of @xmath106 .
an external comparison of cepheid distances with those obtained using the trgb method implies an even smaller dependence : @xmath109 , over a metallicity range of 1.7 dex .
comparisons with two sets of distance estimates , using the pnlf and tully - fisher methods , are consistent with no systematic metallicity dependence within large uncertainties , and they limit the magnitude of any dependence to @xmath134 .
a comparable upper limit is implied by the systematic dependence of cepheid color on metal abundance .
all of these results are consistent with a maximum range in @xmath135 , when measured from pl relations in @xmath1 and @xmath2 .
this range is consistent with the very weak sensitivity predicted by theory ( @xmath136 ) , or with no dependence at all .
it is important to emphasize that the results quoted here apply only to pl relations measured in the @xmath1 and @xmath2 bands .
there is theoretical and observational justification for suspecting that the metallicity dependence at bluer wavelengths is considerably stronger , but this is not relevant to the current measurements , which are carried out exclusively at @xmath1 and @xmath2 .
until the magnitude of the @xmath4 dependence at bluer wavelengths is well calibrated , it would be prudent to restrict extragalactic applications of cepheids to observations at @xmath1 and longer wavelengths .
what are the implications for the distance scale and 0 of a metallicity dependence in this range ? the abundance effects clearly are most important for galaxies with unusually high or low abundances relative to the lmc , which serves as the zeropoint calibrator for the pl relation . on the zkh calibration
the lmc has an oxygen abundance 12 @xmath28 log(o / h ) = 8.50 , or [ o / h ] = @xmath137 .
the most metal - rich galaxies in the key project sample reach [ o / h ] @xmath138 0.3 ( e.g. , ngc 3351 , m100 , m101 inner field ) . for @xmath128
the corresponding error in distance is @xmath319% ( @xmath310.18 mag ) , in the sense that distances to the metal - rich galaxies are systematically underestimated . for @xmath139
the distance change is as much as @xmath3118% , but this is a hard upper limit , applied to the most metal - rich galaxies observed with , and computed for the maximum allowable metallicity dependence .
it is worth noting that any strong cepheid metallicity dependence would tend to cause the cepheid distance to the lmc to be _ overestimated _ , and compensate partly for the distance effects on more distant metal - rich galaxies .
the cepheid @xmath4 dependence tends to cause the distances of metal - poor galaxies to be overestimated .
the abundances of the most metal - poor galaxies in the key project are only slightly below that of the lmc ( e.g. , ngc 3319 , with [ o / h ] @xmath140 ) . for those galaxies the predicted metallicity effects are less than 3% , even for the largest reasonable values of @xmath45
. the effects may be larger for some of the sn ia calibrators being observed by sandage and collaborators with ( e.g. , saha 1994 ) .
beaulieu ( 1997 ) and kochanek ( 1997 ) have suggested that the value of 0 derived from the sn ia method may increase by 16 @xmath31 24% , for an assumed @xmath141 .
our analysis suggests a much smaller difference , based on the weaker metallicity dependence observed in m101 , and the likelihood that the abundances of the sn ia calibrators are closer to the lmc than has been assumed by the other authors .
the most metal - poor sn ia calibrator is probably ngc 5253 , with [ o / h ] @xmath142 ( webster & smith 1983 ) , and its distance would have been overestimated by only 0 @xmath31 8% , for the range of @xmath45 considered here .
given the probable magnitude of these effects , should one apply metallicity corrections to cepheid distances measured with ? as illustrated above , the effects on individual distances are probably 10% _ at most _ , and for most galaxies the changes are at the few percent level .
cepheid distances currently measured with are subject to several random and systematic uncertainties at the 5% level , including uncertainty in the lmc distance , the absolute calibration of the @xmath1 and @xmath2 magnitudes measured with wfpc2 , and uncertainties in the reddening corrections applied to the wfpc2 photometry ( cf .
ferrarese 1996 ; hill 1997 ; madore 1997 ) . in view of these other errors and the uncertain magnitude of the metal abundance correction itself ,
we believe that it is most prudent at this time not to apply a metallicity correction to individual cepheid distances , but rather to include this source of uncertainty in the systematic error budget for the distance . at the end of the key project we hope to have better estimates of these uncertainties , and a much improved zeropoint calibration of the pl relation ( kennicutt 1995 ) . in the meantime , one can assess the importance of metal abundance effects on the overall distance scale .
beaulieu ( 1997 ) , sasselov ( 1997 ) and kochanek ( 1997 ) explored the consequences of a @xmath143 dependence on currently published measurements of 0 , and concluded that the key project 0 value could be reduced by up to 10% , based on the virgo cluster calibration of freedman ( 1994b ) and mould ( 1995 ) .
however the most recent key project 0 determination is based on a much larger sample of cepheid distances and secondary distance indicators , and is much less susceptible to this metallicity bias ( madore 1997 ) . as an illustration , figure 10 shows a histogram of oxygen abundances for the key project galaxies with hii region measurements .
these include the target galaxies , with abundances from zkh , and the groundbased calibrators m31 , m33 , ngc 300 , ngc 2366 , ngc 2403 , and ngc 3109 .
although there is a spread of over an order of magnitude in these abundances , the median value is [ o / h ] @xmath144 , nearly identical to the lmc abundance of [ o / h ] = @xmath92 .
thus the effect of even a substantial cepheid metallicity dependence will be negligible for any distance indicator that is calibrated using most or all of the key project cepheid sample ( e.g. , the tully - fisher relation ) .
the net abundance effect will differ for each secondary distance method , depending on the subset of cepheid calibrating galaxies , but for most methods the mean cepheid abundance will be solar or below , so the @xmath4 effects should be at the 0.10 mag level or less ( less than 5% in distance and 0 ) .
the situation is less certain for the sn ia distance scale , because abundances for most of the cepheid calibrators have not yet been measured .
we are obtaining hii region spectra for the galaxies in order to assess the importance of abundance effects on the sn ia distance scale .
systematic metal abundance effects may be more important for galaxies in the virgo and fornax clusters .
it is well established that spirals in the core of the virgo cluster are systematically more metal - rich than their field counterparts ( skillman 1996 ) , and this region contains several cepheid calibrating galaxies ( m100 , ngc 4548 , ngc 4535 , ngc 4639 , ngc 4571 ) .
less is known about the abundances in the fornax cluster , but the key project sample includes three fornax members ( ngc 1365 , ngc 1425 , ngc 1326a ) , and the one spiral with measured abundances ( ngc 1365 ) is relatively metal - rich .
it will be important when calibrating 0 via galaxies in virgo and fornax to consider the possible systematic abundance effects .
however from the comparisons given earlier the maximum magnitude of an abundance effect will be to underestimate the distances by @xmath14510% , and consequently lead to a decrease in 0 of 10% .
our @xmath1 and @xmath2 photometry of the two m101 fields provides tentative evidence for a small metallicity dependence of the pl relation at these wavelengths , and when combined with other tests , places firm upper limits on the magnitude of the metallicity effect .
these are sufficient to ensure that the overall consequences of a metallicity dependence on the ultimate 0 calibrations with are small ( at the level of a several percent or less ) , and comparable to a multitude of other known error sources in the distance scale calibration .
however the tests presented in this paper fall far short of establishing a definitive calibration of the pl metallicity dependence .
such a calibration would significantly improve the accuracy of individual cepheid distance determinations to individual galaxies , and provide valuable constraints on the theoretical understanding of the cepheids themselves . here
we briefly describe several observations that would address this need .
we also refer to reader to kochanek ( 1997 ) for a discussion of this subject .
as has been discussed previously by madore & freedman ( 1985 ) and kochanek ( 1997 ) , the primary limitation of the observations is the @xmath61 wavelength baseline .
this restriction was imposed primarily by the need to maximize observing efficiency on , and by the absence , until recently , of a near - infrared imaging capability on . a targeted program aimed at obtaining high precision cepheid photometry of a single galaxy such as m101 over a wide wavelength baseline and over
a wide range of abundances would enable one to break the degeneracy between reddening and metallicity effects , and accurately calibrate the @xmath4 dependence .
groundbased observations of this kind are being obtained by at least two groups for m31 .
the installation of nicmos on makes such a program very feasible in a more distant galaxy with a steeper abundance gradient , such as m101 .
the effects of metallicity and reddening on the pl relation are smaller at near - infrared wavelengths ( e.g. , mcgonegal 1982 ; cwc93 ) , which argues strongly for supplementing the existing wfpc2 @xmath1 and @xmath2 photometry of cepheid calibrating galaxies with nicmos imaging in @xmath146 and/or @xmath147 .
as the absolute uncertainty in 0 approaches the 10% level , the systematic uncertainties associated with metallicity and reddening corrections to the cepheid distances will become the dominant source of error in the entire extragalactic distance ladder .
thus a modest effort to assess and correct these systematic errors will have great leverage in improving the calibration of individual extragalactic distances , and ultimately the hubble constant as well .
we are pleased to thank george jacoby , chris kochanek , and dimitar sasselov for useful comments and discussions .
we also thank the referee for suggesting several improvements to the paper . the work presented in this paper
is based on observations made by the nasa / esa hubble space telescope , obtained by the space telescope science institute , which is operated by aura , inc . under nasa contract
5 - 26555 .
we gratefully acknowledge the support of the nasa and stsci support staff , with special thanks to peggy stanley and doug van orsow .
support for this work was provided by nasa through grant go-2227 - 87a from stsci .
this research has made use of the nasa / ipac extragalactic database ( ned ) which is operated by the jet propulsion laboratory , california institute of technology , under a contract with the national aeronautics and space administration .
2. radial abundance distribution in m101 , as derived from hii region spectra .
the round points and circles are empirical abundances based on the calibration of zkh .
open circles denote hii regions that are located in the inner and outer cepheid fields .
open triangles denote abundances based on hii regions with measured electron temperatures .
the horizontal lines indicate the radial coverage of the cepheid fields .
3. observed cepheid pl relations from allframe photometry of the m101 outer field in @xmath1 ( top ) and @xmath2 ( bottom ) .
the lines show the least squares fits , constrained to the slope of the lmc calibrating relations .
5. distributions of @xmath1 and @xmath2 residuals of individual cepheids from the pl fits shown in figs . 2 and 3 .
the top panel shows residuals of the outer field cepheids from the outer field fit .
the middle panel shows the residuals of inner field cepheids from the inner field fit .
the bottom panel shows the residuals of the inner field cepheids from the outer field pl fits . in each panel
the solid line shows the expected residuals from instability strip width effects , while the dashed line shows the expected trend from reddening variations .
6. correlation of the average color excess of the cepheids in a given galaxy or field with abundance . in each case
the galactic foreground reddening has been subtracted .
the solid line show the bivariate least squares fit to the correlation . fig .
7. difference between the distance moduli of nearby galaxies determined from cepheids and by the trgb method , plotted as a function of the cepheid abundances .
solid points indicate direct comparisons of galaxies with cepheid and trgb distances , while the open circles indicate indirect comparisons using cepheid and trgb distances of different galaxies in leo i group .
the solid line shows a bivariate least squares fit to the relation .
the dashed and dotted lines show trends expected for a different cepheid @xmath4 dependences .
8. similar comparison to fig . 7 , but in this case comparing cepheid distances with those derived from the pnlf method .
solid points indicate direct comparisons in galaxies with cepheid and pnlf distances , while open circles are indirect comparisons using different galaxies in the same group .
the lines show expected trends as in fig .
9. the @xmath2-band tully - fisher relation for cepheid calibrating galaxies , using data from giovanelli ( 1997 ) .
the open circles indicate galaxies with poorly determined linewidths , and which are not included in the fit .
the solid line shows the best fitting relation .
the dashed and dotted lines show the relations expected for different cepheid @xmath4 dependences @xmath45 , as in figs .
6 and 7 .
10. distribution of cepheid abundances in the key project sample , as determined from hii region measurements , on the calibration of zkh .
the vertical line indicates the abundance of the lmc , which defines the pl calibrations . | uncertainty in the metal abundance dependence of the cepheid variable period - luminosity ( pl ) relation remains one of the outstanding sources of systematic error in the extragalactic distance scale and the hubble constant .
to test for such a metallicity dependence , we have used the wide field planetary camera 2 ( wfpc2 ) on the _ hubble space telescope _ ( ) to observe cepheids in two fields in the nearby spiral galaxy m101 , which span a range in oxygen abundance of 0.7 @xmath0 0.15 dex .
a differential analysis of the pl relations in @xmath1 and @xmath2 in the two fields yields a marginally significant change in the inferred distance modulus on metal abundance , with @xmath3 = -0.24 \pm 0.16 $ ] .
the trend is in the theoretically predicted sense that metal - rich cepheids appear brighter and closer than metal - poor stars .
external comparisions of cepheid distances with those derived from three other distance indicators , in particular the tip of the red giant branch method , further constrain the magnitude of any @xmath4-dependence of the pl relation at @xmath1 and @xmath2 .
the overall effects of any metallicity dependence on the distance scale derived with will be of the order of a few percent or less for most applications , though distances to individual galaxies at the extremes of the metal abundance range may be affected at the 10% level . 0h@xmath5 |
we load our optical lattice with ultracold atoms originating from a spin - polarized @xmath69 bose - einstein condensate ( bec ) in the @xmath70 @xmath13 hyperfine ground state .
we produce small condensates with @xmath71 atoms ( such that the resulting lattice filling factor is near unity ) in a ioffe - pritchard magnetic trap ; to this trap we subsequently add a three - dimensional optical lattice , generated by intersecting beams from a ti : sapphire laser operating at @xmath72 nm . a lattice along @xmath73 divides the bec into a stack of independent 2d systems , and a separate , deformable double - well lattice formed from a single folded and retroreflected beam in the @xmath74 plane completes the confinement@xcite . since the total lattice depth in the lattice is a complicated function of the laser polarizations controlling the topology , we parametrize the intensity of the single @xmath74-lattice input beam for all lattice configurations in terms of the equivalent @xmath74-lattice depth in the @xmath50 configuration . in terms of this parametrization , the total light shift experienced by a trapped atom in the @xmath50 configuration will be approximately twice the @xmath74-lattice depth plus the depth of the lattice along @xmath73 . during loading ,
all lattice intensities follow an exponential profile , reaching their final values in 150 ms ( with a time constant of 50 ms ) .
typical final lattice depths are 20(1 ) @xmath55 for the vertical lattice and 20(1 ) to 40(2 ) @xmath55 for the @xmath74-lattice , where @xmath56 khz , with @xmath57 and m is the @xmath8rb atomic mass .
( unless otherwise stated , all uncertainties herein reflect the uncorrelated combination of statistical and systematic uncertainties . ) for all experiments , the initial lattice we load into is a state - independent square lattice of period @xmath75 nm . in fig .
3(c ) , the points at 41 @xmath55 were obtained at a lower depth of the lattice along @xmath73 , and have been corrected by 18 hz ( the estimated change in the differential shift from the lattice along @xmath73 ) to be consistent with the other data .
atoms loaded into this initialization lattice are deep into the mott - insulating phase with nominally one atom per site .
we then turn off the ioffe - pritchard trap , leaving a stable bias field of 322.9(1 ) @xmath17 t along @xmath76 .
this field is chosen to minimize the sensitivity of the hyperfine clock transition @xmath77 ( our storage qubit ) to external fields , while allowing for reasonably low sensitivity ( @xmath23 37 khz / mt ) in our working qubit , the well - known @xmath78 transition commonly used in atomic clocks .
we transform our lattice into the state - dependent configuration ( @xmath54 ) on a timescale ( 300 @xmath17s ) adiabatic with respect to vibrational excitation .
this transformation is effected through the use of high - voltage pockels cells which control both the input polarization of the @xmath74-lattice , and the relative phase of the in - plane ( @xmath79 , @xmath80 ) and out - of - plane ( @xmath73 ) polarization components @xcite .
this yields an a / b site dependent ellipticity which in concert with the atoms vector polarizability yields a spin - dependent vector light shift . to measure the state population of the a(b ) lattice sites , we dynamically adjust the topology of the lattice , converting b(a)-site atoms into high momentum states which spatially separate from the low - momentum atoms of interest in the a(b ) sites during time - of - flight @xcite .
we measure the atomic density distribution using resonant absorption imaging along @xmath73 and use stern - gerlach gradients @xcite to concurrently resolve differing @xmath81 components .
in addition to standard technical noise , uncertainty in the contrast measurements in the a(b ) site is due in part to imaging noise from background atoms from the b(a ) site not being intentionally measured .
depending on the depth of the lattice , the high - momentum atoms may spatially overlap with low - momentum atoms in the images ; the data in fig .
[ leftright ] ( for which this effect was slight ) a small correction was applied .
the lattice depth used in fig .
[ precision ] was large enough ( chosen for maximum a b isolation ) that the high - momentum states did not overlap , and no correction was necessary .
transitions between storage states are driven by a two - photon coupling comprising a microwave field at 6832.325 mhz and a rf field at 2.352975 mhz @xcite . for our field intensities and detunings
the two - photon rabi frequency is @xmath82 hz , given a @xmath83 khz detuning from the @xmath16 intermediate state . unlike single - photon transitions , the two - photon transition is sensitive to power - dependent shifts due to the radiofrequency and microwave coupling to the intermediate state .
this shift is only present during the application of the coupling , and must be considered when satisfying the the energy conserving condition of eq .
this shift is typically @xmath23 + 50 hz , however , and the @xmath84 can be adjusted to satisfy eq .
[ sum ] without compromising resonance , as the single - photon rabi frequencies @xmath85 are all @xmath86 4 khz .
we thank poul jessen and ivan deutsch for helpful discussion , and steven swift and eric huang for technical assistance with our direct - digital - synthesis hardware .
this work was partially supported by dto and onr . n.l .
acknowledges support from the national research council research associateship program . and @xmath2 , a field - insensitive clock - state pair .
* b * , application of a magnetic field gradient zeeman - shifts energy levels , spectrally selecting a specific qubit ( a ) from the register .
field - sensitive transitions can then be used to selectively map the storage - state superposition to a pair of working " states @xmath3 and @xmath4 .
* c * , after the site - selective mapping , arbitrary qubit rotation can be performed on the a qubit alone , as the working - state transition is off - resonant from the storage - state transition .
inverting the mapping process returns qubit a to a new storage - state superposition .
this scheme demonstrates crosstalk - free site - specific addressing , with both qubits almost always in field - insensitive superpositions . ] of the storage - state qubit .
the contrast of the resulting interference fringe as a function of delay is shown ; the storage - state qubit exhibits a coherence time of @xmath23 61(5 ) ms , assuming exponential decay .
the insertion of a spin - echo @xmath32-pulse in the ramsey sequence removes dephasing caused by spatial inhomogeneities , resulting in longer coherence times ( red ) . *
b * , a similar measurement , performed using the working states , yields a coherence time of @xmath23 21(2 ) ms , which also improves given a spin - echo @xmath32-pulse ( red ) .
* c * , we demonstrate the coherent mapping of a storage - state superposition to a working - state superposition .
a modified ramsey pulse sequence opens on the storage transition ( purple ) and closes on the working transition ( yellow ) , giving a fringe of observed contrast of 0.99(3 ) .
the mapping is effected with the use of appropriate @xmath32-pulses transferring the @xmath1 population to @xmath3 ( green ) and the @xmath2 population to @xmath4 ( orange ) .
we control undesired leakage " to a nearby state , as illustrated with the dashed grey transition . ] and the differential shift . * * a * , we apply an effective magnetic field @xmath45 to the b sites of a two - qubit register , zeeman - shifting the field - sensitive levels by an amount @xmath60 .
additionally , in the @xmath54 configuration the light shift at each site changes slightly by amounts @xmath87 and @xmath88 .
* b * , the calibration of @xmath60 is shown as a function of lattice - beam intensity , where the intensity is plotted in units of the equivalent @xmath50 lattice depth ( see methods ) .
a prediction of the frequency shift based on a model of our lattice is shown in pink . * c *
, measurements of the differential shift of the working - transition resonance frequencies in the @xmath50 lattice ( black ) and on the a sites of the @xmath54 lattice ( blue ) are shown as a function of intensity , relative to the expected free - space resonance , effectively measuring @xmath87 .
shown in green is an estimate of the differential shift caused by the lattice along @xmath73 , which is held at constant intensity .
* d * , ramsey fringes are shown illustrating the coherence of storage - state qubits after application of @xmath45 to the b sites during part of the ramsey delay .
we observe ramsey fringes for the a sites ( left graph , contrast 0.95(2 ) ) and b sites ( right graph , contrast 0.96(2 ) ) . ]
@xmath32-pulses vs @xmath64 corpse-@xmath32 pulses on a field - sensitive transition .
the @xmath32-pulses ( denoted as @xmath89 ) have arbitrary phase and standard pulse area , while the corpse pulses comprise three sequential pulses of differing area and phase , in this case @xmath90 .
the @xmath91 analogue of a corpse pulse is simple @xmath64-fold repetition of the corpse pulse . * a - d * , data representing @xmath32 , @xmath92 , @xmath93 , @xmath94-pulses and their respective corpse analogues .
the center frequency of the resonance drifts slightly between between graphs , illustrating the errors to which these experiments are vulnerable .
the dashed lines represent the expected spectrum of an @xmath91-pulse given the measured rabi frequency of the transition . ] | the establishment of a scalable scheme for quantum computing with addressable and long - lived qubits would be a scientific watershed , harnessing the laws of quantum physics to solve classically intractable problems .
the design of many proposed quantum computational platforms is driven by competing needs : isolating the quantum system from the environment to prevent decoherence , and easily and accurately controlling the system with external fields .
for example , neutral - atom optical - lattice architectures provide environmental isolation through the use of states that are robust against fluctuating external fields , yet external fields are essential for qubit addressing . here
we demonstrate the selection of individual qubits with external fields , despite the fact that the qubits are in field - insensitive superpositions .
we use a spatially inhomogeneous external field to map selected qubits to a _
different _ field - insensitive superposition ( optical mri " ) , minimally perturbing unselected qubits , despite the fact that the addressing field is not spatially localized .
we show robust single - qubit rotations on neutral - atom qubits located at selected lattice sites .
this precise coherent control is an important step forward for lattice - based neutral - atom quantum computation , and is quite generally applicable to state transfer and qubit isolation in other architectures using field - insensitive qubits .
the ability to address individual qubits is a vital component of most quantum computing architectures . in the case of neutral atom qubits held in an optical - lattice register @xcite , addressing
generally requires the interaction of specific atoms with a control field . of long - standing concern
is the difficulty of addressing only selected atoms amongst an ensemble of @xmath0 atoms in nominally identical lattice sites .
one approach is to use long - period lattices or arrays of independent single - atom traps sufficiently spaced to address with a focused optical beam @xcite ; in this case the optical diffraction limit sets a bound on the qubit register spacing ( and therefore register density ) .
alternatively , experiments using the mott insulator transition in optical lattices @xcite result in large arrays of subwavelength - separated single ground - state atoms , which are useful for collisional and exchange quantum gates @xcite @xcite .
schemes similar to magnetic resonance imaging for addressing subwavelength - separated qubits have been proposed and demonstrated , wherein an externally applied gradient field shifts local energies , mapping spectroscopic resolution to spatial resolution @xcite . however , these schemes require that the qubits be in field - sensitive states , a requirement which is at odds with the need for long coherence times . here
we demonstrate how to combine environmental insensitivity and site - specific subwavelength addressability , as illustrated in fig .
[ scheme ] .
our scheme is based on a register of qubits , each in a superposition of _ storage _ states @xmath1 and @xmath2 whose energy difference is insensitive to external magnetic fields , a significant source of environmental decoherence .
( such pairs of states are known colloquially as clock states " due to their utility as frequency standards . ) in addition , each qubit has a second pair of clock states , the _ working _ states @xmath3 and @xmath4 , which have a different transition frequency than the storage states .
while transitions between the storage states ( @xmath5 ) or between the working states ( @xmath6 ) are insensitive to external fields , the transitions between storage and working states ( e.g. @xmath7 ) _ are _ field - sensitive .
application of a field gradient thus spectrally selects a qubit from the register , allowing frequency - sensitive ( and therefore position - sensitive ) mapping of the selected qubit s coherent superposition between storage states and working states .
once the selected qubit has been transferred to the working states , we can then perform isolated arbitrary single - qubit operations on the selected qubit alone .
remaining qubits ( still in the storage states ) are unaffected by both the mapping process and the subsequent control operation on the selected qubit .
one can then map the selected qubit back to storage states , resulting in a qubit register with the addressed qubit in a new arbitrary superposition of storage states .
this idea may be applicable to other physical systems vulnerable to optical or electrical crosstalk ( e.g. , refs .
@xcite and @xcite ) , relaxing the required isolation for control fields used to address individual qubits in a spatially dense register .
we demonstrate this scheme in an optical lattice - based ensemble of registers , where each register is composed of two separately trapped @xmath8rb atoms acting as qubits a and b. the storage and working states of the qubit are encoded in four hyperfine sublevels of the ground - state manifold of @xmath8rb , which can be coupled with resonant microwave radiation .
both qubits are initialized in the storage - state superposition @xmath9 . after applying a localized effective magnetic field , we spectrally select qubit a and map it into the working - state superposition @xmath10 , where @xmath11 is a systematic phase depending on the details of the mapping process .
qubit b remains in the initial storage - state superposition .
we then apply a control operation on the working transition , transforming qubit a into the new state @xmath12 . using a modified ramsey technique
we verify the basic features of this scheme : namely , that storage - state coherence is unaffected by the application of the field gradient , and that the addressable mapping process is coherent .
the confinement for our register is provided by a double - well optical lattice ( see methods and ref .
@xcite ) .
we use the states @xmath13 and @xmath14 as storage states @xmath1 and @xmath2 , and the states @xmath15 and @xmath16 as working states @xmath4 and @xmath3 . at our operating field near 323 @xmath17 t , the linear magnetic - field dependence of the storage transition vanishes , making the storage - state qubit extremely insensitive to small field fluctuations or inhomogeneity @xcite .
the working transition has a slight linear sensitivity to magnetic field .
figure [ coherence](a - b ) shows the coherence properties of the storage and working states , measured using ramsey s method of separated oscillatory fields @xcite .
fig .
[ coherence](a ) shows the result of a standard ramsey sequence , where we open " with an initial @xmath18-pulse on the storage transition , placing each qubit in the equal superposition @xmath19 .
after a fixed delay , we close " the sequence with a final @xmath18-pulse of variable relative phase @xmath20 , and measure the populations in @xmath1 and @xmath2 as a function of @xmath20 .
the resulting fringe contrast as a function of the delay time decays with a dephasing time @xmath21 ms .
the residual dephasing is limited by inhomogeneity of the lattice - beam intensity since there is no known simple
magic wavelength " scheme @xcite for hyperfine transitions in @xmath8rb @xcite .
this inhomogeneity results in a small , lattice - induced differential light shift of the hyperfine states that is proportional to lattice intensity and inversely proportional to the lattice - beam detuning from atomic resonance .
we calculate a differential shift of @xmath22 hz for a typical total lattice light shift of 230 khz and estimate an inhomogeneity of @xmath23 10 hz for our experimental parameters , which is roughly consistent with the observed dephasing time .
we also measure the magnitude of the differential shift as a function of lattice intensity ( discussed below ) .
the inhomogeneity of the differential light shift is the main technical limitation on coherence in our ensemble of storage qubits , and can be improved by making the lattice beams more homogeneous and detuning the lattice farther from resonance , as with the @xmath24 @xmath17 m lattice used for precision spectroscopy in ref .
@xcite .
in addition , dynamical decoupling techniques have been shown to be useful in reducing the effects of inhomogeneous broadening @xcite . a simple spin - echo pulse sequence , which rephases time - independent inhomogeneous dephasing , gives a residual coherence time @xmath25 in excess of 300 ms . in an analogous measurement ,
fig .
[ coherence](b ) shows that the working - state coherence exhibits a somewhat shorter dephasing time of @xmath26 ms , dominated by background magnetic field gradients .
in contrast to the field insensitive qubit states , the transitions @xmath27 or @xmath28 _ are _ field sensitive .
this sensitivity enables spectroscopic qubit addressing , but requires that state transfers be performed fast enough to avoid the qubits vulnerability to field inhomogeneities while they temporarily occupy the field - sensitive superpositions @xmath29 or @xmath30 during the transfer from storage states to working states , as in fig .
[ scheme](b ) .
we measure this timescale using the standard ramsey method on the field - sensitive transition @xmath27 , giving @xmath31 @xmath17s , roughly 100 times shorter than that of the clock states . while the ramsey dephasing is linearly sensitive to weak field inhomogeneity , the transfer efficiencies of the @xmath32-pulses in our mapping sequence are only quadratically sensitive .
this suggests that the mapping is less sensitive to field inhomogeneities than simply given by @xmath33 .
to directly determine the impact of this field sensitivity on the mapping from storage to working states , we developed a modified ramsey sequence that is sensitive to decoherence in the mapping process .
this sequence is composed of the standard opening @xmath18-pulse on the storage transition , followed by @xmath32-pulses mapping storage states to working states , and closed by a final @xmath18-pulse on the working transition . while the standard ramsey sequence depends only on the relative phase of the two ( equal - frequency ) @xmath18-pulses , our modified sequence requires phase control of four different microwave signals , including the two mapping @xmath32-pulses . because all our pulses have different frequencies , the meaning and control of the relative phases involved is more subtle .
in particular , each modified ramsey sequence ( performed at time @xmath34 relative to a fixed origin ) involves four signals of the form @xmath35 , and the observed fringe depends on the relative phase @xmath36 , where @xmath37 if the four frequencies are phase - locked and satisfy the energy - conserving condition @xmath38 , the ramsey output is insensitive to the starting time @xmath34 of the sequence , depending only on @xmath39 .
we observe ramsey fringes by adjusting the phase of any of the microwave fields by a variable offset @xmath40 , with the @xmath41 and @xmath42 fringes out of phase with the @xmath43 and @xmath44 fringes .
a fringe resulting from varying @xmath42 is shown in fig .
2(c ) .
despite the fact that the total duration of the mapping process ( 200 @xmath17s ) is comparable to the dephasing time of the sensitive transitions , the fringe has high contrast , confirming that the qubit coherence is largely unaffected by the mapping process .
we obtain the necessary field gradients to make the mapping process addressable by generating an optically - induced effective magnetic field at every other lattice site , i.e. , on one site of each of our two - qubit registers , as illustrated in fig .
[ leftright](a ) .
this optically - induced effective magnetic field is proportional to the local ellipticity of the lattice light , which we control using electro - optic modulators ( see methods and ref .
@xcite ) .
the spatially varying effective magnetic field @xmath45 results from the atom s vector light shift ( which adds to the scalar light shift providing the lattice potential ) , where @xmath46 , @xmath47 is the vector polarizability , @xmath48 is the lattice electric field , and @xmath49 is the local ellipticity of the lattice field .
@xmath45 adds vectorially with the external bias field @xcite .
we can adiabatically transform the initial @xmath50 lattice into a lattice with a nonzero @xmath45 at the b sites of each quantum register , zeeman - shifting the resonant frequency of the b sites by @xmath51 .
the effective field remains zero at the a sites , resulting in an effective magnetic field gradient between a and b sites . for a given polarization configuration ,
this shift is proportional to the lattice - beam intensity as illustrated in fig .
[ leftright](b ) , and corresponds to a field gradient of up to @xmath23 8 t / m . in the presence of this gradient
, the a and b sites of the quantum register can be addressed using radiofrequency or microwave fields .
we now demonstrate a ) that our qubits are largely insensitive to @xmath45 , yet b ) we can nevertheless use @xmath45 to coherently address selected qubits in our quantum register .
the robustness of the storage - state qubit in the presence of the effective magnetic field is illustrated in fig .
[ leftright](d ) .
we show high - contrast ramsey fringes , measured independently for a and b sites , where we apply the addressing field @xmath45 to the b qubits between the @xmath18-pulses comprising the ramsey sequence .
as expected , the storage - state coherence on either site is unaffected by @xmath45 . during the application of @xmath45 ,
a phase offset of 19(2)@xmath52 develops between the ramsey fringes of the two sites , which corresponds to a small energy difference of @xmath5335 hz over the 1.5 ms ramsey delay .
this partially results from an intensity difference ( and associated difference in differential light shifts ) that exists between the a and b sites in the @xmath54 configuration ( see fig .
[ leftright](a ) ) . in order to understand this
, we measure the magnitude of the differential light shift for the working - state transition , first in the @xmath50 configuration , and then for the a sites in the @xmath54 configuration , both as a function of intensity , as illustrated in fig . [ leftright](c ) .
we parametrize the lattice - beam intensity in all lattice configurations in terms of the equivalent lattice depth of the @xmath50 configuration ( see methods ) .
given this parametrization , we measure the differential shift for atoms in the @xmath50 configuration to be 6.1(5 ) hz/@xmath55 , where @xmath56 khz , with @xmath57 and m is the @xmath8rb atomic mass .
this is near a calculated value of 4.9 hz/@xmath55 based on a model of the lattice and an atomic light - shift calculation .
the difference between these curves represents crosstalk : the extent to which the addressing process perturbs a - site atoms , which nominally experience no effective magnetic field .
a typical 25 hz difference in differential light shifts on the a sites ( between the @xmath58 and @xmath54 configurations ) combined with a typical 15 khz effective zeeman shift of the b sites suggests a crosstalk figure - of - merit for our system of @xmath23 0.002 .
for the two storage - state ramsey fringes of fig .
[ leftright](d ) , the estimated difference in differential shifts between the a and b sites of @xmath23 12 hz ( different than that of the case of the working transition ) combined with the @xmath45 application time of @xmath23 600 @xmath17s gives an expected phase shift of @xmath23 3@xmath52 , considerably smaller than that observed .
several effects could contribute to this discrepancy : drifts in the lattice intensity , peculiarities of beam alignment not included in our lattice model @xcite , and drifts in both microwave / rf power and background magnetic fields leading to shifts in the two - photon transition controlling the storage - state qubits ( see methods ) .
figure [ precision ] illustrates our full capability , where we combine coherent mapping between storage and working states with the addressing provided by the use of @xmath45 .
the combination is complicated by the need to simultaneously satisfy three possibly conflicting criteria : the prevention of leakage " of the mapped - site population into unwanted hyperfine states ( see fig .
[ precision](a ) ) , and the assurance that each of the two mapping pulses affects only one site ( a b isolation : see fig .
[ precision](b ) ) .
as our rabi frequencies are comparable to both the effective zeeman shift from @xmath45 as well as shifts of the hyperfine states due to the nuclear magnetic moment , some care is required .
we initialize both qubits in the register with a @xmath18-pulse on the storage - state transition , apply the effective magnetic field to the b sites , then apply the mapping pulses . since @xmath45 shifts the b - site mapping transitions from the a - site resonance ( by @xmath59 khz ) , conversion from storage states to working states
is performed _ only _ on the a sites .
we satisfy the above three isolation criteria by appropriately choosing three mapping parameters : the effective zeeman shift @xmath60 and the two microwave rabi frequencies of the mapping @xmath32-pulses . in particular ,
the rabi frequency of the second mapping pulse ( @xmath61 ) was chosen to eliminate leakage to undesired states , as in the global operation of fig .
[ coherence ] ; additionally , we chose @xmath60 such that the second mapping pulse exhibited zero response at a detuning of @xmath51 . finally , we chose the rabi frequency of the first mapping pulse ( @xmath62 ) so that it similarly exhibited zero response at a detuning of @xmath51 .
to verify the site - selective mapping of the a - site atoms to the working states , we a ) measure the state population in the a sites after closing the ramsey sequence with a @xmath18-pulse on the working transition , and b ) measure the state population in the b sites after a @xmath18 pulse on the storage transition .
the data shown in fig . [ precision ] proves the coherence and isolation of the transfer process and also demonstrates controlled single - qubit rotation performed on only one site of the register .
the high contrast of the resulting fringes confirms that the process is coherent , and that state pollution from imperfect a b isolation or intra - site leakage to undesired states is at most 3% ( comparable to our measurement uncertainty of about 3% ) .
any control scheme incorporating field - sensitive transitions ( as ours does ) is vulnerable to imperfections in the quality of the control , due to fluctuating or inhomogeneous background or control fields . while our observed global ramsey fringe contrasts are consistent with unity , scalable quantum computing places stringent limits on required control fidelities
this level of control can be achieved with composite pulses , providing a specific desired result ( such as that of the @xmath32-pulses used in our transfer process ) using a train of pulses of variable pulse area and phase .
composite pulses are designed to be robust against fluctuations and inhomogeneities over a given bandwidth @xcite ; pulses of particular interest to the quantum computing community have been discussed @xcite and explored experimentally @xcite . as a proof of principle , fig .
[ composite ] shows the results of applying the venerable corpse pulse sequence on the field - sensitive transition @xmath63 , along with the results of conventional @xmath32-pulses .
as expected , the @xmath64-corpse-@xmath32 spectra are significantly flatter about resonance than the equivalent @xmath32-pulse spectra .
these are not immediately applicable to our transfer process due to issues involving a b isolation , but demonstrate the inherent utility of the technique for neutral atom quantum computation .
the technique demonstrated here can be used with the effective field gradient of an individual focused laser beam @xcite to provide single site addressing .
the ultimate fidelity of such addressable control will be determined by a range of technical issues , such as the stability of the external and control fields , the spatial resolution of the addressing beam and its registration to the lattice position , and the lattice intensity .
the fundamental limit is set by the spontaneous photon scattering of the light providing the effective magnetic field gradient . in @xmath8rb atoms , for example
, one optimal choice for the addressing laser wavelength is found near 787 nm , detuned between the @xmath65 and @xmath66 transitions .
the total scattering probability during a site - selective @xmath32-pulse depends on the details of the system @xcite but we calculate that for a 1 @xmath17 m beam waist and 0.5 @xmath17 m lattice spacing , the scattering probability can be @xmath67 . in the context of our double - well lattice ,
future work will focus on implementing the transfer process with composite pulses and implementing benchmarking techniques @xcite to probe our control fidelities below the percent level .
we also seek to use our techniques to provide all possible inputs to the @xmath68 gate described in ref .
@xcite , where the relevant gate time is more than two orders of magnitude faster than the relevant coherence time . in a more general context ,
the approach we have demonstrated here is applicable to any quantum computing architecture where two appropriate pairs of states can be found , with the attendant ability to transform in and out of a storage - state quantum memory . |
in figure [ fig : dimer_lattice_representation ] we show on the left panel the schematic representation of our model hamiltonian .
the blue lines correspond the inter - dimer hopping matrix @xmath48 and the purple line the intra - dimer hopping @xmath49 . for simplicity
the model is depicted in 3d , but it is mathematically formulated in the limit of a large coordination lattice . the right hand side panel shows the corresponding dmft quantum impurity problem , where the dimer is embedded in a self consistent medium .
we numerically demonstrate that the ipt method exactly captures the atomic limit of the lattice model .
the fact that a perturbative calculation captures the atomic limit ( ie , @xmath50 ) is not to be expected , but , interestingly enough , is analogous to the well known property of the ipt solution of the one band hubbard model .
this is shown in figure [ fig : ipt_dimer_exact ] where we compare the exact self - energy of the atomic limit ( isolated dimer ) with the respective ipt solution .
the most stringent test for the comparison is done within the coexistence region , since there the structure of the green s functions and self - energies are very non - trivial .
figure [ fig : ct - qmc - ipt - coex ] shows the numerically exact ct - qmc green s functions and self - energies in the matsubara axis .
the data is shown along with fits obtained from the ipt solution at suitably close values of the parameters . the very good agreement between qmc and ipt is also found in the whole phase diagram .
this is shown in figure [ fig : ctqmc - ipt - benign ] where ipt and ct - qmc are compared at two parameter values away from the coexistence region . within dmft , the bandstructure is obtained as a function of the single particle energy @xmath51 , which in the semicircular dos lattice adopted here has a simple linear dispersion @xcite . hence at @xmath52 the non - interacting bandstructure are two parallel linear bands split by @xmath53 ( i.e. the bonding and anti - bonding bands ) . in figure
[ fig : electronic_structures ] we show the electronic dispersion ( ie , @xmath51 and @xmath38 resolved density of states @xmath54 ) of the metallic and insulating states obtained by the ipt , ct - qmc and ed methods , respectively . as seen in the figure , all the three methods provide the key qualitative features that are discussed in the text .
namely , the split quasiparticle bands in the metal and in the insulating states .
the dimer hubbard model has very few parameters inter - dimer ( or lattice ) hopping @xmath9 , intra - dimer hopping @xmath14 , local coulomb repulsion @xmath11 , and the temperature .
the lattice hopping provides the bandwidth of our model hamiltonian @xmath55 @xcite . in the case of vo@xmath56 reference lda calculations
agree that both , @xmath57 and @xmath58 bands have an approximate bandwidth of 2ev @xcite , hence we have set @xmath59 . the intra - dimmer hopping quoted in ref .
@xcite is @xmath2 0.68ev , which is about a factor of 2 larger than our adopted value 0.3ev .
however , that value concerns solely the @xmath58 ( d@xmath60 ) orbital .
there are in fact also two additional intra - dimer hopping amplitudes ( associated to e@xmath61 states ) , which are 0.22ev and a smaller value .
the values for the hopping amplitudes found in later works @xcite are consistent with this findings .
since we are considering a unit cell with two sites and one orbital each , we have a single intra - dimer hopping parameter .
therefore , and in the spirit of a mean field theory , we chose a value that is the approximate average of the 3 hopping amplitudes . in regard to the value of @xmath11 , we adopted the value of @xmath62 for the semi - quantitative comparison with experiments .
this is quite consistent with the values considered in @xcite , who systematically explored the range of @xmath63 to 3.5ev , and found the mit between 2.7 and 3ev at the lowest temperatures considered .
the values adopted in ref .
@xcite are @xmath64 and @xmath65 .
these values are in fact higher than ours , however , biermann et al dedicate the last paragraph of their letter to discuss their choice of the value of @xmath11 and @xmath66 .
they mention that smaller values , such as @xmath67 are also compatible with their findings .
this is also consistent with our choice of @xmath62 .
the mott insulating state is signaled by the divergence of the self - energy in the mott gap .
this divergence is clearly visible on the real axis in figure [ fig : ipt_dimer_mott_gap](middle and bottom rows ) . in the left column
we show the well known case of a single - band hubbard model ( @xmath68 ) with particle - hole symmetry . in this case
the self - energy diverges at zero frequency , and this is therefore visible in the matsubara axis too . eg . in figure
[ fig : ipt_dimer_exact ] above ( the @xmath68 blue squares on the top most right panel ) . in the case of the dimer lattice ( @xmath69 ) , with two atoms per unit cell , the dimerization splits the divergence of the self - energy into two poles , symmetric around zero frequency , as displayed on the right right column on figure [ fig : ipt_dimer_mott_gap ] . in this case
the divergence is no more visible on the matsubara axis , and in particular @xmath70 goes smoothly to zero as @xmath71 . ) .
the right column corresponds to the dimer lattice at @xmath69 . for both figures @xmath72[fig : ipt_dimer_mott_gap ] , scaledwidth=90.0% ] the dimer hubbard model is not a single band model despite being a single orbital per site system . since it has two sites ( the dimer ) in the unit cell , there are two bands .
the two sites are related by symmetry thus the two bands are degenerate in the atomic - site representation but are distinct in the bonding / antibonding basis .
the physics of the system is independent of the choice of the representation . to further clarify this point we show in figure [ fig : ipt_ab_spectra_character ] the dos from our model calculation from fig .
3 of the manuscript in the atomic - site basis , along with the bonding and antibonding dos .
the average of the latter two gives the former one ( which has degeneracy 2 ) .
[ fig : decomposed_optcond ] ) [ fig : ipt_ab_spectra_character ] ]
in order to identify the electronic structure of the insulator state of the model , we obtained the corresponding quantity in the atomic limit , which can be analytically solved in the real frequency axis ( the impurity model is an isolated dimer ) . in this limit the lattice green s function is obtained as , @xmath73 - \left[\begin{matrix}\sigma_{11 } & \sigma_{12}\\\sigma_{12 } & \sigma_{11}\end{matrix}\right]_{dimer}\ ] ] the electronic structure in this limit is shown in figure [ fig : dispersion_dimer ] at 2 different temperatures .
the left panel is at inverse temperature @xmath74 , where we see two highly dispersive bands along with two flatter ones . at high temperature @xmath75
more excitations are apparent as the first excited state of the dimer becomes thermally populated .
this dramatically enriches the electronic structure .
interestingly , these multiple excitations can be readily identified in the actual model results shown in the main text .
to calculate the optical conductivity one requires a geometrical definition of the lattice . as a common practice ,
one turns to the hypercube in infinite dimensions to keep on with the exact limit of the dmft approximation @xcite . using the peierls ansatz to find the current operator and aligning our dimer along the ,
say , @xmath76 direction in the hypercube one finds the current operator along the @xmath76 direction to be : where @xmath78 is the electron charge , @xmath79 is the imaginary unit , a is the lattice unit cell length , and @xmath80 is the separation between atoms of the dimer .
when one diagonalizes the lattice hamiltonian one can see it in terms of quasiparticles that form a bonding @xmath81 and an anti - bonding @xmath82 band , one uses then the operators @xmath83 to create ( annihilate ) quasiparticles in the bonding band with momentum @xmath84 and spin @xmath8 and analogously the operators @xmath85 for the anti - bonding band . in infinite dimensions to keep the kinetic energy constant one has to scale the hopping amplitude @xmath86 @xcite . then following the procedure established in @xcite we arrive to the expression @xmath87 \\ \nonumber & + 2 \pi e^2 \eta^2 t_\perp^2 \int d\omega ' \frac{f(\omega ' ) -f(\omega'+\omega)}{\omega}\times\\ & \int de \rho(e ) \left[a_b(e,\omega)a_a(e,\omega+\omega ' ) + a_a(e,\omega)a_b(e,\omega+\omega')\right ] \label{eq : optcond_gf_e_sum_ld}\end{aligned}\ ] ] where @xmath88 is the density of states of the hypercube , and @xmath89 are the spectral functions of the bonding and anti - bonding bands . from the previous equation we find that , in general , both intra - band and inter - band transitions are present .
notice that the latter are usually disregarded ( e.g. ref .
the two contributions are weighted with different factors that depend on the specific geometry . for the sake of simplicity , we set those geometric prefactors equal to unity , as they are expected to be of the same order ( a @xmath90 , @xmath91 and @xmath92 ) . in figure
[ fig : decomposed_optcond ] , we plot for the metal and the insulator case the individual contribution of the interband transitions from the bonding to antibonding bands(refer to figure [ fig : ipt_ab_spectra_character ] ) and of the intraband excitations . | we consider a minimal model to investigate the metal - insulator transition in vo@xmath0 .
we adopt a hubbard model with two orbital per unit cell , which captures the competition between mott and singlet - dimer localization .
we solve the model within dynamical mean field theory , characterizing in detail the metal - insulator transition and finding new features in the electronic states .
we compare our results with available experimental data obtaining good agreement in the relevant model parameter range .
crucially , we can account for puzzling optical conductivity data obtained within the hysteresis region , which we associate to a novel metallic state characterized by a split heavy quasiparticle band .
our results show that the thermal - driven insulator - to - metal transition in vo@xmath0 is compatible with a mott electronic mechanism , providing fresh insight to a long standing `` chicken - and - egg '' debate and calling for further research of `` mottronics '' applications of this system .
= 1 vanadium dioxide vo@xmath0 and vanadium sesquioxide v@xmath0o@xmath1 remain at the center stage of condensed matter physics as they are prototypical examples of systems undergoing a strongly correlated metal - insulator transition ( mit ) @xcite .
their unusual electronic behavior makes them very attractive materials for novel electronic devices @xcite .
in fact , they are intensively investigated in the emerging field of `` mottronics '' , which aims to exploit the functionalities associated to the quantum mott transitions .
a key goal is to create fast and ultra - low power consumption transistors , which may be downsized to the atomic limit @xcite .
vo@xmath0 and v@xmath0o@xmath1 have nominally partially filled bands , hence are expected to be metals .
however , they undergo a first order metal to insulator transition upon cooling at @xmath2 340k and 180k , respectively . this phenomenon has been often associated to a mott mit @xcite , namely a transition driven by the competition between kinetic energy and coulomb repulsion @xcite . yet , that point of view has been questioned as often@xcite . the case of vo@xmath0 , displaying a transition from a high-@xmath3 rutile ( r ) metal to a low-@xmath3 monoclinic ( m@xmath4 ) insulator ,
is emblematic @xcite .
the central issue is whether the transition is driven by a spin - peierls structural instability , or by the electronic charge localization of the mott - hubbard type .
this issue has been under scrutiny using electronic structure calculations @xcite based on the combination of density - functional theory in the local - density approximation with dynamical mean- field theory ( lda+dmft ) @xcite . in the pioneering work of ref .
@xcite , biermann et al . argued that the insulator should be considered as a renormalized peierls insulator .
namely , a band - insulator where the opening of the bonding - antibonding gap is driven by dimerization and renormalized down by interactions @xcite . on the other hand
, the calculations showed that within the metallic rutile phase , the coulomb interaction failed to produce a mit for reasonable values of the interaction .
more recently , the problem was reconsidered by brito et al .
@xcite and by biermann et al .
as well @xcite providing a rather different scenario .
brito et al . found a mit within a second monoclinic ( m@xmath0 ) phase of vo@xmath0 that only has half the dimerization of the standard m@xmath4 , for the same value of the coulomb interaction .
hence , they argued that mott localization must play the leading role in both mits .
nevertheless , they also noted that the mott insulator adiabatically connects to the singlet dimer insulator state , and therefore the transition should be considered as a mott - hubbard in the presence of strong inter - site exchange @xcite .
while those lda+dmft works provided multiple useful insights , the issue whether the first order mit at 340k in vo@xmath0 is electronically or structurally driven , still remains .
here we shall try to shed new light on this classic `` chicken - and - egg '' problem by adopting a different strategy .
we shall trade the complications of the realistic crystal structures and orbital degeneracy of vo@xmath0 for a model hamiltonian , the dimer hubbard model ( dhm ) , that captures the key competition between mott localization due to coulomb repulsion and singlet dimerization , i.e. peierls localization .
this permits a detailed systematic study that may clearly expose the physical mechanisms at play .
importantly , in our study the underlying lattice _ stays put_. therefore , we can directly address the issue whether a purely electronic transition , having a bearing on the physics of vo@xmath0 , exists in this model .
the specific questions that we shall address are the following : ( _ i _ ) does this purely electronic model predict a first order metal - insulator transition as a function of the temperature _ within the relevant parameter region _ ?
( _ ii _ ) what is the physical nature of the different states ?
( _ iii _ ) can they be related to key available experiments ?
these issues are relevant , since if this basic model fails to predict an electronic mit consistent with the one observed in vo@xmath0 , then it would be mandatory to include the lattice degrees of freedom . in the present study we shall provide explicit answers to these questions
we show that the solution of the dhm brings the equivalent physical insight for vo@xmath0 as the single band hubbard model for v@xmath0o@xmath1 , which is one of the significant achievements of dmft @xcite .
the dimer hubbard model is defined as @xmath5 \\ & + \sum_{i\alpha } u n_{i\alpha\uparrow } n_{i\alpha\downarrow}\end{aligned}\ ] ] where @xmath6 denotes n.n .
lattice sites , @xmath7 denote the dimer orbitals , @xmath8 is the spin , @xmath9 is the lattice hopping , @xmath10 is the intra - dimer hopping , and @xmath11 is the coulomb repulsion . for simplicity , we adopt a semicircular density of states @xmath12 .
the energy unit is set by @xmath9=1/2 , which gives a full bandwidth of 4@xmath9=2@xmath13=2 , where @xmath13 is the half - bandwidth .
this interesting model has surprisingly received little attention , and only partial solutions have been obtained within dmft @xcite .
the main results were the identification of the region of coexistent solutions at moderate @xmath11 and small @xmath14 at @xmath3=0 using the iterated perturbation theory ( ipt ) approximation @xcite and at finite @xmath3=0.025 by quantum monte carlo@xcite ( qmc ) @xcite .
here we obtain the detailed solution of the problem paying special attention to the mit and the nature of the coexistent solutions .
we solve for the dmft equations with hybridization - expansion continuous - time quantum monte carlo ( ct - qmc ) @xcite and exact diagonalization@xcite , which provide ( numerically ) exact solutions .
we also adopt the ipt approximation @xcite , which , remarkably , we find is ( numerically ) exact in the atomic limit @xmath15 , therefore provides reliable solutions of comparable quality as in the single - band hubbard model @xcite . furthermore , ipt is extremely fast and efficient to explore the large parameter space of the model and provides accurate solutions on the real frequency axis .
extensive comparison between ipt and the ct - qmc is shown in the supplemental material .
the dmft equations provide for the exact solution of the dhm in the limit of large lattice coordination and have been derived elsewhere @xcite . here
we quote the key self - consistency condition of the associated quantum dimer - impurity model , @xmath16 where @xmath17 and @xmath18 ( with @xmath19 ) are respectively the dimer - impurity green s function and self - energy . at the self consistent point
these two quantities become the respective local quantities of the _ lattice _ @xcite . an important point to emphasize
is that this quantum dimer - impurity problem is _ analogous _ to that in the above mentioned lda+dmft studies @xcite .
therefore , strictly speaking , our methodology is a cluster - dmft ( cdmft ) calculation ( cf supplemental material ) .
we start by establishing the detailed phase diagram , which we show in fig .
[ fig - phasediag ] .
we observe that at low @xmath3 there is a large coexistent region at moderate @xmath11 and @xmath14 below 0.6 @xcite .
this region gradually shrinks as @xmath3 is increased , and fully disappears at @xmath20 .
the lower panel shows the phase diagram in the @xmath11-@xmath3 plane at fixed @xmath14 . at @xmath14=0
we recover the well known single - band hubbard model result , where the coexistent region extends in a triangular region defined by the lines @xmath21 and @xmath22 @xcite .
the triangle is tilted to the left , which indicates that upon warming the correlated metal undergoes a first order transition to a finite-@xmath3 mott insulator .
this behavior was immediately associated to the famous 1st order mit observed in _ cr - doped _
v@xmath0o@xmath1 @xcite , which has been long considered a prime example of a mott hubbard transition @xcite .
it is noteworthy that this physical feature has remained relevant even in recent lda+dmft studies , where the full complexity of the lattice and orbital degeneracy is considered @xcite .
this underlines the utility of sorting the detailed behavior of basic model hamiltonians .
significantly , as @xmath14 is increased in the dhm , the tilt of the triangular region evolves towards the right .
this signals that @xmath14 fundamentally changes the stability of the groundstate .
in fact , as shown in the lower right panel of fig .
[ fig - phasediag ] , at @xmath14=0.3 we find that the mit is _ reversed _ with respect to the previous case .
namely , upon warming , an insulator undergoes a 1st order transition to a ( bad ) correlated metal at finite-@xmath3 .
we may connect several features of this mit to vo@xmath0 , both , qualitative and semi - quantitatively .
we first consider the energy scales and compare the parameters of the dhm to those of electronic structure calculations .
the lda estimate of the bandwidth of the metallic state of vo@xmath0 is @xmath22ev @xcite , which corresponds in our model to @xmath23 , hence @xmath9=0.5ev .
this is handy , since from our choice of @xmath9=0.5 , we may simply read the numerical energy values of the figures directly in physical units ( ev ) and compare to experimental data of vo@xmath0 .
we notice that the coexistence region ( with a 1st order transition line ) extends up to @xmath24 0.04 ( ev ) @xmath25 400k , consistent with the experimental value @xmath25 340k .
we then set the value of @xmath14=0.3ev , that approximately corresponds with lda estimates for the ( average ) intra - dimer hopping amplitudes ( cf supplemental material ) @xcite .
thus , the coexistence region is centered around @xmath26 , consistent with the values adopted in the lda+dmft studies @xcite
. -@xmath11 plane .
left one shows lower temperatures @xmath3=0.001 ( ipt ) and 1/200 ( ct - qmc ) , and right
one shows higher temperature @xmath3=0.03 ( ipt ) and 1/64 ( ct - qmc ) .
lower panels show the @xmath11-@xmath3 plane .
left one is for fixed @xmath14=0 ( i.e. single - band hubbard model ) , and right one for @xmath14=0.3 .,scaledwidth=45.0% ] we can make further interesting connections with experiments in vo@xmath0 .
the metallic state is unusual and it can be characterized as a _
bad metal_. namely , a metal with an anomalously high scattering rate that approaches ( or may violate ) the ioffe - regel limit @xcite . in fig .
[ fig - sigma ] we show the imaginary part of the diagonal self - energy , whose y - axis intercept indicates the scattering rate ( i.e. inverse scattering time ) . at @xmath24 0.04 (
i.e.@xmath2400k ) we observe a large value of the intercept , of order @xmath27=1/2 , which signals that the carriers are short lived quasiparticles .
in fact vo@xmath0 has such an anomalous metallic state @xcite .
this anomalous scattering is likely the origin of the surprising observation that despite the fact that the lattice structure has 1d vanadium chains running along the c - axis , the resistivity is almost isotropic , within a mere factor of 2 @xcite .
it is noteworthy that this lack of anisotropy observed in electronic transport experiments provides further justification for our simplified model of a lattice of dimers .
this bad metal behavior is a hallmark of mottness @xcite and also indicates that the mit in vo@xmath0 should be characterized as a mott transition .
additional insights on the mechanism driving the transition can be obtained from the behavior of the off - diagonal ( intra - dimer ) self - energy @xmath28 . from eq .
( [ eq : self ] ) , we observe that the intra - dimer hopping amplitude is effectively renormalized as @xmath29(0)$ ] . in fig .
[ fig - sigma ] we show the behavior of this quantity across the transition .
we see that in the metallic state it remains small , while it becomes large ( @xmath30 ) at low @xmath3 in the insulator @xcite .
the physical interpretation is transparent . in the correlated metal ,
the two dimer sites are primarily kondo screened by their lattice neighbors , as in the single band hubbard model each one forms a heavy quasiparticle band
. then these two bands get split into a bonding and anti - bonding pair by the small @xmath14 .
hence , the low energy electronic structure is qualitatively similar to the non - interacting one , with a larger effective mass . as @xmath3 is lowered , the dramatic increase in @xmath31(0)$ ] when the mott gap opens at the 1st order transition signals that the intra - dimer interaction is boosted by @xmath32 $ ] . unlike the one - band hubbard model , here the finite @xmath14 permits a large energy gain in the mott insulator by quenching the degenerate entropy .
this mechanism , already observed in other cluster - dmft models @xcite , stabilizes the insulator within the coexistence region , leading to the change in the tilt seen in fig.[fig - phasediag ] .
another way of rationalizing the transition is that at a critical @xmath33dependent @xmath14 the kondo screening in the metal breaks down in favor of the local dimer - singlet formation in the insulator . in this view
, the large gap opening may be interpreted as a @xmath33driven enhanced band splitting @xmath34 .
$ ] for the metal ( solid ) at fixed @xmath14=0.3 values of @xmath11 from 0 to 3 in steps of 0.5 ( upwards ) .
the experimentally relevant values @xmath11=2.5 and 3 , are highlighted with thick lines .
inset shows the @xmath11 dependence at fixed @xmath3=0.04 .
bottom : the effective intra - dimer hopping @xmath35= @xmath14+re@xmath36(0)$ ] ( bottom ) as a function of @xmath3 for the same parameters as the top panel .
metal states are in solid ( blue ) lines and the insulator in dashed ( red ) lines for @xmath11=2.5 and 3 .
the calculation are done with ipt.,scaledwidth=50.0% ] further detail is obtained from the comparison of the electronic structure of the metal and the insulator within the coexistence region . in the correlated metallic state shown in fig .
[ fig - arpes ] we find at high energies ( @xmath37 ) the incoherent hubbard bands , which are signatures of mott physics . at lower energies
, we also observe a pair of heavy quasiparticle bands crossing the fermi energy at @xmath38=0 .
consistent with our previous discussion , this pair of quasiparticle bands can be though of as the renormalization of the non - interacting bandstructure .
significantly , as we shall discuss later on , this feature may explain the puzzling optical data of qazilbash et al .
@xcite within the mit region of vo@xmath0 , which has remained unaccounted for so far . unlike the single - band hubbard model
, the effective mass of these metallic bands does not diverge at the mit at the critical @xmath11 , even at @xmath3=0 .
in fact , the finite @xmath14 cuts off the effective mass divergence as expected in a model that incorporates spin - fluctuations .
in fact the dhm may be considered @xcite the simplest non - trivial cluster dmft model .
it is interesting to note that the realistic values @xmath39 and @xmath14 = 0.3 lead to hubbard bands at @xmath40ev and a quasiparticle residue @xmath41 , both consistent with photoemission experiments of koethe et al .
@xcite . in fig .
[ fig - arpes ] we also show the results for insulator electronic dispersion at the same values of the parameters .
the comparison of the insulator and the metal illustrate the significant changes that undergo at the 1st order mit .
we see that the metallic pair of quasiparticle bands suddenly open a large gap .
more precisely , in contrast to the one - band case , here the hubbard bands acquire a non - trivial structure , with sharp bands coexisting with incoherent ones .
the coherent part dispersion can be traced to those of a lattice of singlet - dimers ( see sup . mat . ) .
hence , the insulator can be characterized as a novel type of mott - singlet state where the hubbard bands have a mix character with both coherent and incoherent electronic - structure contributions .
it is also interesting to note that the gap in the density of states is @xmath420.6ev , again consistent with the photoemission experiments @xcite .
= 0.3 , @xmath11=2.5 and @xmath3=0.01 .
right panels show the respective dos(@xmath38 ) .
the calculations are done with ipt(cf .
sup .
mat.),scaledwidth=45.0% ] in order to gain further insight and make further contact with key experiments , we now consider the optical conductivity response @xmath43 within the mit coexistence region .
a set of remarkable data was obtained in this regime by qazilbash et al .
@xcite , bearing directly on the issue of the driving force behind the transition .
they systematically investigated the @xmath43 as a function of @xmath3 using nano - imaging spectroscopy .
they clearly identified within the @xmath3 range of the mit the electronic coexistence of insulator and metallic regions , characteristic of a 1st order transition .
a crucial observation was that upon warming the insulator in the m1 phase , metallic puddles emerge with a @xmath43 that was significantly different from the signal of the normal metallic r phase .
thus , the data provided a strong indication of a purely electronic driven transition .
regarding this point we would like to mention also the works of arcangeletti et al .
@xcite and laverock et al .
@xcite that reported the observation of metallic states within the monoclinic phase under pressure and strain , respectively . coming back to the experiment of qazilbash et al .
, a key point that we want to emphasize here is that @xmath43 in the putative m1-metallic state was characterized by a intriguing mid - infra - red ( mir ) peak @xmath44 1800 @xmath45 ev , whose origin was not understood . from our results on the electronic structure within the coexistence region ,
we find a natural interpretation for the puzzling mir peak : it corresponds to excitations between the split metallic quasiparticle bands .
since they are parallel , they would produce a significant contribution to @xmath43 , which enabled its detection . in fig .
[ fig - ac ] we show the calculated optical conductivity response ( see sup . mat .
section 7 ) that corresponds to the spectra of fig .
[ fig - arpes ] . in the metal
we see that , in fact , a prominent mir peak is present at @xmath46 0.22 ev , in excellent agreement with the experimental value . on the other hand , the optical conductivity of the insulator shows a maximum at @xmath47 ev in both , theory and experiment .
moreover , we also note the good agreement of the relative spectral strengths of the main features in the two phases . of metal and insulator within the coexistence region for parameters
@xmath14=0.3 , @xmath11=2.5 and @xmath3=0.01 .
the calculations are done with ipt .
inset : the experimental optical conductivity adapted from ref .
.,scaledwidth=45.0% ] in conclusion , we showed that the detailed solution of the dimer model treated within dmft can account for a number of experimental features observed in vo@xmath0 .
the minimal model has an impurity problem which is analogue to that of lda+dmft methods , yet the simplicity of this approach allowed for a detailed solution that permitted a transparent understanding of many physical aspects of the electronic first order transition in this problem .
it exposes a dimer - mott - transition mechanism , where the effective intra - dimer exchange is controlled by correlations , it is weakened in the metal and strongly enhanced in the mott insulator . in the metal
, this leads to a pair of split quasiparticle bands , which then in the insulator further separate , to join and coexist with the usual incoherent hubbard bands . despite the simplicity of our model
, we made semi - quantitative connections to several experimental data in vo@xmath0 , including a crucial optical conductivity study within the 1st order transition , that remained unaccounted for .
our work , sheds light on the long - standing question of the driving force behind the metal - insulator transition of vo@xmath0 highlighting the relevance of the mott mechanism .
the present approach may be considered the counterpart for vo@xmath0 , of the dmft studies of the mott transition in paramagnetic cr - doped v@xmath0o@xmath1 .
_ acknowledgments _ we thank i. paul , s. biermann , g. kotliar and h - t kim for helpful discussions .
this work was partially supported by public grants from the french national research agency ( anr ) , project lacunes no anr-13-bs04 - 0006 - 01 , the nsf dmr-1005751 and dmr-1410132 . *
supplementary material * |
an important part of the visual system of vertebrate animals are the neural connections between the eye and the brain . at an initial stage of ontogenesis
the ganglion cells of the retina have random synaptic contacts with the tectum , a part of the brain which plays an important role in processing optical information . in the adult animal , however , neighboring retinal cells project onto neighboring cells of the tectum ( see figure [ fig1 ] ) .
further examples of these so - called _ retinotopic _ projections are established between the retina and the corpus geniculatum laterale as well as the visual cortex , respectively @xcite .
this conservation of neighborhood relations is also realized in many other neural connections between different cell sheets .
for instance , the formation of ordered projections between the mechanical receptors in the skin and the somatosensorial cortex is called somatotopy .
an even more abstract topological projection arises when the spatially resolved detection of similar frequencies in the ear are projected onto neighboring cells of the auditorial cortex .
a further notable neural map in the auditory system was discovered in the brain of the owl , where neighboring cells of the _ nucleus mesencephalicus lateralis dorsalis _ ( mld )
are excited by neighboring space areas , i.e. every space point is represented by a small zone of the mld @xcite .
the variety of examples suggest that there must be some underlying general mechanism for rearranging the initially disordered synaptic contacts into topological projections .
+ in the early 1940s , sperry performed a series of pioneering experiments in the visual system of frogs and goldfish @xcite .
fish and amphibians can regenerate axonal tracts in their central nervous system , in contrast to mammals , birds and reptiles .
sperry crushed the optical nerve and found that retinal axons reestablished the previous retinotopically ordered pattern of connections in the tectum .
then in the early 1960s sperry presented his chemoaffinity hypothesis which proposed that the retinotectal map is set up on the basis of chemical markers carried by the cells @xcite .
however , experiments over several decades have shown that the formation of retinotectal maps can not be explained by this gradient matching alone @xcite .
+ in the course of ontogenesis originally disordered mappings between retina and tectum evolve into ordered projections.,width=415 ] the group of von der malsburg suggested that these ontogenetic processes result from self - organization .
the basic notion in their theory is the following : once a fibre has already grown from the retina to the tectum , the fibre moves along by strengthening its contacts in some parts of its ramification and by weakening them in others .
it is assumed that these modifications are governed by two contradictory rules @xcite : on the one hand , synaptic contacts on neighboring tectal cells stemming from fibres of the same retinal region support each other to be strengthened . on the other hand ,
the contacts starting from one retinal cell or ending at one tectal cell compete with each other . in the case
that retina and tectum are treated as one - dimensional discrete cell arrays , extensive computer simulations have shown that a system based on these ideas of cooperativity and competition establishes , indeed , retinotopy as the final configuration @xcite .
this finding was confirmed by a detailed analytical treatment of hussler and von der malsburg @xcite where the self - organized formation of the synaptic connections between retina and tectum is described by an appropriate system of ordinary differential equations . applying the methods of synergetics @xcite for one - dimensional discrete cell arrays
, they succeeded in classifying the possible retinotopic projections and to discuss the criteria which determine their emergence .
the more complicated case of continuously distributed cells on a spherical shell was partially discussed in ref .
@xcite . + it is the purpose of this paper to follow the outline of ref .
@xcite and generalize the original approach by elaborating a model for the self - organized formation of retinotopic projections which is _ independent _ of the special geometry and dimension of the cell sheets .
there are three essential reasons which motivate this more general approach .
first , neurons usually do not establish 1-dimensional arrays but 2- or 3-dimensional networks .
hence the 1-dimensional model of hussler and von der malsburg can only serve as a simplistic approximation of the real situation .
secondly , we want to include cell sheets of different extent , which is a more realistic assumption than neural sheets with the same number of cells .
the third reason is that a general model is able to reveal what is generic , i.e. what is independent of the special geometry of the problem .
thus , here we generalize the hussler equations to continuous manifolds of arbitrary geometry . by doing so
, we proceed in a phenomenological manner and relegate a microscopic derivation of the underlying equations to future research .
+ it should be emphasized that our main objective is not the biological modelling of retinotopy . instead of that our considerations are devoted to the analysis of the dynamics of the nonlinear hussler equations by using mathematical methods from nonlinear dynamics and synergetics .
for the more biological aspects of retinotopy and the vast progress in modelling various retinotopically ordered projections during the last twenty years we refer the reader to the reviews @xcite .
+ in section [ secmo ] we present the general framework of our model and introduce the equations of evolution for the connection weights between retina and tectum .
we then perform in section [ secli ] a linear stability analysis for the equations of evolution around the stationary uniform state and discuss under which circumstances an instability arises . in section [ nonli ]
we apply the methods of synergetics , and elaborate within a nonlinear analysis that the adiabatic elimination of the fast evolving degrees of freedom leads to effective equations of evolution for the slow evolving order parameters .
they approximately describe the dynamics near the instability where an increase of the uniform growth rate of new synapses onto the tectum beyond a critical value converts an initially disordered mapping into a retinotopic projection .
finally , section [ summary ] and [ outlook ] provide a summary and an outlook .
in this section we summarize the basic assumptions of our general model .
we start with representing retina ( @xmath0 ) and tectum ( @xmath1 ) by general manifolds @xmath2 and @xmath3 , respectively . in the framework of an embedding of these manifolds in an euclidean space of dimension @xmath4 ,
the coordinates @xmath5 , @xmath6 of the corresponding cells can be represented by @xmath7 in the following we need measures of distance , i.e. metrics @xmath8 , @xmath9 on the manifolds .
the intrinsic coordinates of the @xmath10-dimensional manifolds @xmath3 , @xmath2 are denoted by @xmath11 , @xmath12 .
thus , the vectors ( [ xtxr ] ) of the euclidean embedding space can be parametrized according to @xmath13 , @xmath14 . with the covariant metric tensors @xmath15 the line elements on the manifolds
are given by @xmath16 the geodetic distances between two points of the manifolds read @xmath17 we define a measure for the magnitudes of the manifolds by @xmath18 where we integrate over all elements of @xmath2 , @xmath3 .
we characterize the neural connectivity within each manifold @xmath2 , @xmath3 by cooperativity functions @xmath19 , @xmath20 . in lack of any theory for the cooperativity functions we regard them as time - independent , given properties of the manifolds which are only limited by certain global plausible constraints .
we assume that the cooperativity functions are positive @xmath21 that they are symmetric with respect to their arguments @xmath22 and that they fulfill the normalization conditions @xmath23 furthermore , it is neurophysiologically reasonable to assume that the cooperativity functions @xmath19 , @xmath24 are larger when the distance between the points @xmath25 and @xmath26 is smaller .
this condition of monotonically decreasing cooperativity functions can be written as @xmath27 the neural connections between retina and tectum are described by a connection weight @xmath28 for every ordered pair @xmath29 with @xmath30 . in this paper we are interested in the temporal evolution of the connection weight @xmath28 which is essentially determined by the given cooperativity functions @xmath19 , @xmath20 of the manifolds @xmath2 , @xmath3 . to this end
we generalize a former ansatz of hussler and von der malsburg @xcite and assume that the evolution is governed by the following system of ordinary differential equations @xcite : @xmath31 \nonumber\\ & & -\frac{w(t , r)}{2m_r}\hspace*{1mm}\int\ ! dr'\left [ \alpha+w(t , r ' ) \,\int\ ! dt ' \ , \int\ !
dr '' c_t(t , t')\ , c_r(r',r'')\,w(t',r'')\right ] \ , . \label{haus}\end{aligned}\ ] ] here @xmath32 denotes the uniform growth - rate of new synapses onto the tectum , which will be the control parameter of our system .
these equations of evolution represent a balance between different cooperating and competing processes . to see this
, we define the growth rate between the cells at @xmath33 and @xmath34 @xmath35 so that the generalized hussler equations ( [ haus ] ) reduce to @xmath36 the cooperative contribution of the connection between @xmath37 and @xmath38 to the growth rate between @xmath33 and @xmath34 is given by the product @xmath39 as shown in figure [ fig2]a .
therefore , this cooperative contribution is integrated with respect to @xmath37 , @xmath38 and added to the uniform growth rate @xmath32 to yield the total growth rate ( [ gro ] ) between @xmath33 and @xmath34 .
apart from this cooperative term in the equations of evolution ( [ hausb ] ) , the remaining terms describe competitive processes .
the second term accounts for the fact that growth rates between @xmath33 and @xmath38 compete with the connections between @xmath33 and @xmath34 ( see figure [ fig2]b ) .
correspondingly , the third term describes the competition of the growth rates between @xmath37 and @xmath34 with the connections between @xmath33 and @xmath34 ( see figure [ fig2]c ) . a ) illustrations for the respective contributions to the generalized hussler equations ( [ hausb ] ) .
discussion see text.,title="fig : " ] b ) illustrations for the respective contributions to the generalized hussler equations ( [ hausb ] ) .
discussion see text.,title="fig : " ] c ) illustrations for the respective contributions to the generalized hussler equations ( [ hausb ] ) .
discussion see text.,title="fig : " ] now we show that the evolution of the system due to the generalized hussler equations ( [ haus ] ) leads to a lower bound for the connection weight . to this end
we assume the inequality @xmath40 to be fulfilled for some initial configuration .
then we conclude that the quantity @xmath41 is positive as both the cooperativity functions @xmath19 , @xmath20 and the connection weight @xmath42 are positive due to ( [ cpos ] ) and ( [ ine ] ) . on the other hand we read off from the normalization of the cooperativity functions ( [ norm ] ) that @xmath43 can not be larger than @xmath44 : @xmath45 . with this
we can find a lower bound for @xmath46 as follows .
the growth rate ( [ gro ] ) reads together with ( [ cg ] ) : @xmath47 .
it can be minimized by setting @xmath48 , i.e. @xmath49 whereas its maximum value follows from @xmath50 : @xmath51 to obtain a lower bound for @xmath46 in the hussler equations ( [ hausb ] ) , we insert the minimum ( [ min ] ) of the growth rate for the cooperative first term and its maximum ( [ max ] ) for the remaining competitive terms : @xmath52 hence a small but positive @xmath28 is prevented by a positive rate @xmath32 from becoming zero . in this way we can conclude that the connection weight @xmath28 is positive , when the inequality ( [ ine ] ) is valid in an initial configuration .
all further investigations will concentrate on solutions of the hussler equations ( [ haus ] ) with @xmath53 .
note that , in particular , the growth rates ( [ gro ] ) for such configurations are positive . to perform both a linear and a nonlinear analysis of the underlying hussler equations ( [ haus ] )
we need a complete orthonormal system for both manifolds @xmath2 and @xmath3 . with the help of the contravariant components
@xmath54 , @xmath55 of the metric introduced in section [ manifolds ] we define the respective laplace - beltrami operators on the manifolds @xmath56 where @xmath57 represent the determinants of the covariant components @xmath58 , @xmath59 of the metric .
the laplace - beltrami operators allow to introduce a complete orthonormal system by their eigenfunctions @xmath60 , @xmath61 according to @xmath62 here @xmath63 , @xmath64 denote discrete or continuous numbers which parameterize the eigenvalues @xmath65 , @xmath66 of the laplace - beltrami operators which could be degenerate . by construction , they fulfill the orthonormality relations @xmath67 and the completeness relations @xmath68 note that the explicit form ( [ lbo ] ) of the laplace - beltrami operators enforces the eigenvalues @xmath69 with the constant eigenfunctions @xmath70 because of ( [ m1 ] ) and the orthonormality relations ( [ ortho ] ) .
the cooperativity functions can be expanded in terms of the eigenfunctions according to @xmath71 in the following we assume for the sake of simplicity that the corresponding expansion coefficients are diagonal @xmath72 so we have @xmath73 thus , @xmath74 , @xmath75 are not only eigenfunctions of the laplace - beltrami operators as in ( [ lbo2 ] ) but also eigenfunctions of the cooperativity functions according to @xmath76 note that the normalization of the cooperativity functions ( [ norm ] ) and the orthonormalization relations ( [ ortho ] ) lead to the constraints @xmath77
now we employ the methods of synergetics @xcite and investigate the underlying equations of evolution ( [ haus ] ) in the vicinity of the stationary uniform solution . inserting the ansatz @xmath78 into the hussler equations ( [ haus ] ) , we take into account ( [ m1 ] ) as well as the normalization of the cooperativity functions ( [ norm ] ) . by doing so
, we deduce @xmath79 .
let us introduce the deviation from this stationary uniform solution @xmath80 and rewrite the hussler equations ( [ haus ] ) .
defining the linear operators @xmath81 the resulting equations of evolution assume the form @xmath82 here the linear , quadratic , and cubic terms , respectively , are given by @xmath83 to analyze the stability of the stationary uniform solution we neglect for the time being the nonlinear terms in ( [ gen43 ] ) and investigate the linear problem @xmath84 solutions of ( [ ls ] ) depend exponentially on the time @xmath85 , @xmath86 with @xmath87 and @xmath88 denoting the eigenfunctions and eigenvalues of the linear operator @xmath89 : @xmath90 now we use the complete and orthonormal system on the manifolds @xmath2 , @xmath3 , which have been defined in section [ cos ] , and show that the eigenfunctions of @xmath89 are products of the form @xmath91 indeed , when the operator ( [ c ] ) acts on ( [ ev ] ) , the expansion of the cooperativity functions ( [ cex ] ) leads , together with the orthonormality relations ( [ ortho ] ) , to @xmath92 thus , the operator @xmath93 has the eigenfunctions @xmath94 with the eigenvalues @xmath95 . in a similar way we obtain for the operator ( [ b ] ) : @xmath96 v_{\lambda_t \lambda_r}/2&\hspace*{0.5cm}&\lambda_t=0\,,\lambda_r\not=0;\lambda_r=0\,,\lambda_t\not=0\ , , \\*[2 mm ] 0&&\mbox{otherwise}\ , .
\end{array } \right .
\label{eig2}\ ] ] combining the eigenvalue problems ( [ eig1 ] ) , ( [ eig2 ] ) for @xmath93 and @xmath97 , we find @xmath98 thus , we conclude from ( [ eig1])([eig3 ] ) that the linear operator @xmath89 fulfills the eigenvalue problem ( [ lei ] ) with the eigenfunctions ( [ ev ] ) and the eigenvalues @xmath99 by changing the uniform growth rate @xmath32 in a suitable way , the real parts of some eigenvalues ( [ val ] ) become positive and the system can be driven to the neighborhood of an instability . which eigenvalues ( [ val ] ) become unstable in general depends on the respective values of the given expansion coefficients @xmath100 , @xmath101 .
the situation simplifies , however , if we follow ref .
@xcite and assume that the absolute values of the expansion coefficients @xmath100 , @xmath101 are equal or smaller than the normalization value @xmath102 : @xmath103 .
then the eigenvalue in ( [ val ] ) with the largest real part is given by some parameters @xmath104 with @xmath105 .
thus , the linear stability analysis reveals that the instability arises at the critical uniform growth rate @xmath106 and that its neighborhood is characterized by @xmath107 consequently , the absolute values of the eigenvalues of the unstable modes @xmath108 are much smaller than those of the stable modes @xmath109 : @xmath110 the resulting spectrum is schematically illustrated in figure [ fig3 ]
. schematic representation of the eigenvalues ( [ val ] ) at the instability .
the unstable part consists of those eigenvalues which nearly vanish whereas the stable part lies in a region separated by a finite distance from the stable part.,width=453 ]
in this section we perform a detailed nonlinear analysis of the hussler equations ( [ haus ] ) . using the methods of synergetics
@xcite we derive our main result in form of the order parameter equations which describe the emergence of retinotopic projections from initially undifferentiated mappings .
we return to the nonlinear equations of evolution ( [ gen43 ] ) for the deviation from the stationary uniform solution @xmath111 . as the eigenfunctions @xmath112 , @xmath61 of the laplace - beltrami operators @xmath113
, @xmath114 represent a complete orthonormal system on the manifolds @xmath2 , @xmath3 , we can expand the deviation from the stationary solution according to @xmath115 here we have introduced einstein s sum convention , i.e. repeated indices are implicitly summed over .
the sum convention is adopted throughout .
motivated by the linear stability analysis of the preceding section , we decompose the expansion ( [ vexp ] ) near the instability which is characterized by ( [ ins ] ) : @xmath116 we can expand the unstable modes in the form @xmath117 where the expansion amplitudes @xmath118 will later represent the order parameters indicating the emergence of an instability . correspondingly , @xmath119 denotes the contribution of the stable modes .
note that the summation in ( [ sexp ] ) is performed over all parameters @xmath120 except for @xmath121 , i.e. from now on the parameters @xmath122 stand for the stable modes alone . in
the following we aim at deriving separate equations of evolution for the amplitudes @xmath118 , @xmath123 . to this end
we define the operators @xmath124 which project , out of @xmath111 , the amplitudes of the unstable and stable modes , respectively : @xmath125 @xmath126 these equations follow from ( [ us])([proj2 ] ) by taking into account the orthonormality relations ( [ ortho ] ) . with these projectors the nonlinear equations of evolution ( [ gen43 ] )
decompose into @xmath127 note that we used the eigenvalue problem ( [ lei ] ) for the linear operator @xmath89 and its eigenfunctions ( [ ev ] ) to derive the first term on the right - hand side in ( [ u ] ) and ( [ s ] ) , where einstein s sum convention is not applied .
+ in general , it appears impossible to determine a solution for the coupled amplitude equations ( [ u ] ) , ( [ s ] ) . near the instability which is characterized by ( [ ins ] ) , however , the methods of synergetics @xcite allow elaborating an approximate solution which is based on the inequality ( [ rea ] ) . to this end
we interpret ( [ rea ] ) in terms of a _ time - scale hierarchy _
, i.e. the stable modes evolve on a faster time - scale than the unstable modes : @xmath128 due to this time - scale hierarchy the stable modes @xmath123 quasi - instantaneously take values which are prescribed by the unstable modes @xmath118 .
this is the content of the well - known _ slaving principle _ of synergetics : the stable modes are enslaved by the unstable modes . in our context
it states mathematically that the dynamics of the stable modes @xmath123 is determined by the center manifold @xmath129 according to @xmath130 inserting ( [ h ] ) in ( [ s ] ) leads to an implicit equation for the center manifold @xmath129 which we approximately solve in the vicinity of the instability below . by doing so
, we adiabatically eliminate the stable modes from the relevant dynamics
. then we use the center manifold @xmath129 in the equations of evolution ( [ u ] ) , i.e. we reduce the original high - dimensional system to a low - dimensional one for the order parameters @xmath118 .
the resulting order parameter equations describe the dynamics near the instability where an increase of the uniform growth rate @xmath32 beyond its critical value ( [ ins ] ) converts disordered mappings into retinotopic projections .
it turns out that the derivation of the order parameter equations contains integrals over products of eigenfunctions which have the form @xmath131 where @xmath132 , @xmath133 stand for the respective quantities @xmath63 , @xmath34 and @xmath64 , @xmath33 of the manifolds @xmath2 and @xmath3 .
examples for such integrals are : @xmath134 the first two integrals of ( [ i13 ] ) follow from the orthonormality relations ( [ ortho ] ) by taking into account ( [ coo ] ) : @xmath135 where @xmath136 corresponds to @xmath137 or @xmath138 , respectively .
note that we will later make frequently use of the following consequence of ( [ i13 ] ) and ( [ i1ab ] ) : @xmath139 integrals with products of more than two eigenfunctions can not be evaluated in general , they have to be determined for each manifold separately . at present
we can only make the following conclusion .
expanding the product @xmath140 in terms of the complete orthonormal system @xmath141 the latter integral of ( [ i13 ] ) is given by @xmath142 in addition , we will need also integrals of the type @xmath143 for instance , @xmath144 again we use the orthonormality relations ( [ ortho ] ) , the expansion ( [ wigner ] ) , and take into account ( [ coo ] ) to obtain @xmath145 where again @xmath136 corresponds to @xmath137 or @xmath138 , respectively .
now we approximately determine the center manifold ( [ h ] ) in lowest order . to this end
we read off from ( [ q ] ) , ( [ k ] ) , and ( [ s ] ) that the nonlinear terms in the equations of evolution for the stable modes @xmath123 are of quadratic order in the unstable modes @xmath118 .
thus , the stable modes can be approximately determined from @xmath146 with the nonlinearity @xmath147 using the definitions of the linear operators ( [ c ] ) , ( [ b ] ) and the decomposition of the unstable modes ( [ uexp ] ) as well as the projector for the stable modes ( [ proj2 ] ) , we see that the second and the third term in ( [ qn ] ) vanish due to ( [ cons ] ) @xmath148 whereas the first term yields @xmath149 and the fourth term leads to @xmath150\ , .
\label{fourth}\end{aligned}\ ] ] therefore , we read off from ( [ qn])([fourth ] ) the decomposition @xmath151 where the expansion coefficients are given by @xmath152 \ , .
\label{qres}\end{aligned}\ ] ] note that einstein s sum convention is not to be applied . to solve the approximate equations of evolution for the stable modes ( [ sn ] ) with the quadratic nonlinearity in the order parameters ( [ qnn ] )
, we assume that the center manifold ( [ h ] ) has the same quadratic nonlinearity : @xmath153 inserting ( [ hn ] ) in ( [ sn ] ) , we only need the linear term in ( [ u ] ) to determine the expansion coefficients of the center manifold : @xmath154 here , again , einstein s sum convention is not to be applied .
therefore , the eqs .
( [ qres])([hnn ] ) define the lowest order approximation of the center manifold . knowing that the center manifold depends in lowest order quadratically on the unstable modes near the instability
, we can determine the order parameter equations up to the cubic nonlinearity . because of ( [ q ] ) ,
( [ k ] ) , and ( [ u ] ) they read @xmath155 where the nonlinear term decomposes into three contributions : @xmath156 the first and the second term represent a quadratic and a cubic nonlinearity which is generated by the order parameters themselves @xmath157 whereas the third one denotes a cubic nonlinearity which is affected by the enslaved stables modes according to @xmath158 it remains to evaluate the respective contributions by using the definitions of the linear operators ( [ c ] ) , ( [ b ] ) and the decompositions ( [ uexp ] ) , ( [ sexp ] ) as well as the projector ( [ proj1 ] ) . we start by noting that the last three terms in ( [ u1 ] ) vanish due to ( [ cons ] ) , i.e. @xmath159 so the first term in ( [ u1 ] ) leads to the nonvanishing result @xmath160 correspondingly , we obtain for ( [ u2 ] ) @xmath161 furthermore , taking into account ( [ cons ] ) , we observe that four of the eight terms in ( [ u3 ] ) vanish : @xmath162 the nonvanishing terms in ( [ u3 ] ) read @xmath163 and @xmath164 as well as @xmath165 where we used @xmath102 in the last equation .
therefore , we obtain for ( [ u3 ] ) @xmath166 i^{\lambda_t^u}_{\lambda_t^u { } ' \lambda_t } \ , i^{\lambda_r^u}_{\lambda_r^u { } ' \lambda_r } \right .
\nonumber \\ & & \hspace{-1cm}\left
. - \frac{1}{2 } \ , \big[\,\frac{1}{\sqrt{m_t } } \,\delta_{\lambda_t 0 } \,\delta_{\lambda_t^u \lambda_t^u { } ' } \ , \left ( 1 + f_{\lambda_r } \right ) \ , i^{\lambda_r^u}_{\lambda_r^u { } '
\lambda_r } + \,\frac{1}{\sqrt{m_r}}\,\delta_{\lambda_r 0}\,\delta_{\lambda_r^u \lambda_r^u { } ' } \ , \left ( 1+f_{\lambda_t}\right ) \ , i^{\lambda_t^u}_{\lambda_t^u{}'\lambda_t } \big ] \right\ } \ , .
\label{k2}\end{aligned}\ ] ] taking into account ( [ hn ] ) , we read off from ( [ un ] ) , ( [ qnon ] ) , ( [ k1 ] ) , and ( [ k2 ] ) that the general form of the order parameter equations is independent of the geometry of the problem : @xmath167 the corresponding coefficients can be expressed in terms of the expansion coefficients @xmath100 , @xmath101 of the cooperativity functions ( [ cex ] ) and integrals over products of the eigenfunctions @xmath74 , @xmath75 which have the form ( [ abkurz ] ) or ( [ abkurzneu ] ) .
they read @xmath168 and @xmath169 i^{\lambda_t^u}_{\lambda_t^u { } ' \lambda_t } \ , i^{\lambda_r^u}_{\lambda_r^u { } ' \lambda_r } - \frac{1}{2 } \ , \left [ \,\frac{1}{\sqrt{m_t}}\,\delta_{\lambda_t 0}\ , \delta_{\lambda_t^u \lambda_t^u { } ' } \ , \left ( 1 + f_{\lambda_r } \right ) i^{\lambda_r^u}_{\lambda_r^u { } ' \lambda_r } \right.\right .
\nonumber \\ & & \left .
\hspace{-1.05cm}+\,\frac{1}{\sqrt{m_r}}\,\delta_{\lambda_r 0}\,\delta_{\lambda_r^u \lambda_r^u { } ' } \ , \left ( 1+f_{\lambda_t}\right ) \ , i^{\lambda_t^u}_{\lambda_t^u { } ' \lambda_t } \right ] \right\ } h^{\lambda_t
\lambda_r}_{\lambda_t^u { } '' \lambda_r^u{}'',\lambda_t^u { } ' '' \lambda_r^u { } ' '' } \ , \ , . \label{bb}\end{aligned}\ ] ] as is common in synergetics , the coefficients ( [ bb ] ) in general consist of two parts , one stemming from the order parameters themselves and the other representing the influence of the center manifold @xmath129 .
+ with ( [ ope])([bb ] ) we have derived the generic form of the order parameter equations for the connection weights between two manifolds of different geometry and dimension .
these equations represent the central new result of our synergetic analysis . specifying the geometry means inserting the corresponding eigenfunctions of the laplace - beltrami operators ( [ lbo ] ) into the integrals ( [ abkurz ] ) , ( [ abkurzneu ] ) appearing in ( [ aa ] ) and ( [ bb ] ) .
because the synergetic formalism needs not be applied to every geometry anew , our general procedure means a significant facilitation and tremendous progress as compared to the special approach in ref .
in this circular causality chain of synergetics for the order parameter equations of the generalized hussler equations ( [ haus ] ) . the control parameter
@xmath32 denotes the growth rate of new synapses onto the tectum . ]
paper we have proposed that the self - organized formation of retinotopic projections between manifolds of different geometries and dimensions is governed by a system of ordinary differential equations ( [ haus ] ) which generalizes a former ansatz by hussler and von der malsburg @xcite . the linear stability analysis determines the instability where an increase of the uniform growth rate @xmath32 beyond the critical value ( [ ins ] ) converts an initially disordered mapping into a retinotopic projection .
furthermore , it gives rise to a decomposition of the deviation from the stationary uniform solution @xmath111 near the instability in unstable and stable contributions . by inserting this decomposition in the nonlinear hussler equations ( [ haus ] ) ,
we obtain equations for the mode amplitudes of the unstable and stable modes , respectively . in the vicinity of the instability point the system generates a time - scale hierarchy , i.e. the stable modes evolve on a faster time - scale than the unstable modes .
this leads to the _ slaving principle _ of synergetics : the stable modes are enslaved by the unstable modes . in the literature
this enslaving @xmath170 is usually achieved by invoking an adiabatic elimination of the stable modes , which amounts to solving the equation @xmath171 .
however , the mathematically correct approach for determining the center manifold @xmath172 is to determine it from the corresponding evolution equations for the stable modes @xcite .
it can be shown that only for real eigenvalues this approach leads to the same result obtained by the approximation @xmath171 .
thus , it is possible to reduce the original high - dimensional system to a low - dimensional one which only contains the unstable amplitudes .
the general form of the resulting order parameter equations ( [ ope ] ) is independent of the geometry of the problem .
it contains typically a linear , a quadratic and a cubic term of the order parameters . as a general feature of synergetics , the coefficients ( [ ope ] ) , ( [ bb ] ) consist of two parts , one stemming from the order parameters themselves and the other representing the influence of the center manifold on the order parameter dynamics .
+ our results can be interpreted as an example for the validity of the circular causality chain of synergetics , which is illustrated in figure [ zirkaus ] . on the one hand
, the order parameters , i.e. the few amplitudes @xmath173 of the slowly evolving linear unstable modes @xmath174 , enslave the dynamics of the many stable mode amplitudes @xmath175 of the fast evolving stable modes @xmath176 through the center manifold . on the other hand ,
the center manifold of the stable amplitudes acts back on the order parameter equations .
the order parameter equations ( [ ope])([bb ] ) represent the central new result of this paper , and in the forthcoming publication @xcite they will serve as the starting point to analyze in detail the self - organization in cell arrays of different geometries . to this end
we assume that the manifolds are characterized by spatial homogeneity and isotropy , i.e. neither a point nor a direction is preferred to another , respectively .
this additional assumption requires the manifolds to have a constant curvature and their metric turns out to be the stationary robertson - walker metric of general relativity @xcite .
we therefore have to discuss the three different cases where the curvature of the manifolds is positive , vanishes , or is negative .
this corresponds to modelling retina and tectum by the sphere , the plane , or the pseudosphere . + a further intriguing problem concerns the question under what circumstances non - retinotopic modes become unstable and destroy the retinotopic order .
one could imagine that some types of pathological development in animals corresponds to this case .
+ as already mentioned , lacking any theory for the cooperativity functions , we have regarded them as time - independent given properties of the manifolds .
they are determined by the lateral connections between the cells of retina and tectum , respectively @xcite .
but neither a reason for their time - independence nor a detailed discussion of their precise mathematical form is available . to fill this gap
it will be necessary to elaborate a self - consistent theory of the cooperativity functions .
+ our generalized hussler equations are fully deterministic . in real systems , however , there are always fluctuations . to take into account such unpredictable small variations a stochastic force has to be added to the deterministic part of the equation .
such fluctuations are known to play an important role , especially in the vicinity of instability points @xcite .
+ finally , delayed processes could be included in our considerations .
synergetic concepts have been successfully applied to time - delayed dynamical systems in refs .
@xcite . in neurophysiological systems delays occur due to the finite propagation velocity of nerve signals @xcite as well as the finite duration of physiological processes such as the change of synaptic connection weights .
thus , it would be also worthwhile to expand the investigations to time - delayed hussler equations .
we thank r. friedrich , c. von der malsburg , and a. wunderlin for stimulating discussions at an initial stage of the work .
99 e. r. kandel , j. h. schwartz , and t. m. jessell ( eds . ) , principles of neural science , fourth edition ( mcgraw - hill , new york , 2000 ) .
e. i. knudsen and m. konishi , science * 200 * , 795 ( 1978 ) .
r. w. sperry , j. exper .
zool . * 92 * , 263 ( 1943 ) .
r. w. sperry , j. comp .
neurol . * 79 * , 33 ( 1943 ) .
r. w. sperry , proc .
50 * , 703 ( 1963 ) .
g. j. goodhill and l. j. richards , trends neurosci . *
22 * , 529 ( 1999 ) . c. von der malsburg and d. j. willshaw , proc .
74 * , 5176 ( 1977 ) .
d. j. willshaw and c. von der malsburg , philos . trans .
london b * 287 * , 203 ( 1979 ) .
a. f. hussler and c. von der malsburg , j. theoret
. neurobiol .
* 2 * , 47 ( 1983 ) .
h. haken , synergetics an introduction , third edition ( springer , berlin , 1983 ) .
h. haken , advanced synergetics ( springer , berlin , 1983 ) .
w. wagner and c. von der malsburg , private communication .
m. gmann , a. pelster , and g. wunner , a general model for the development of retinotopic projections between manifolds of different geometries ; in r. p. wrtz and m. lappe ( editors ) , proceedings of the 4 . workshop dynamic perception , bochum , germany , november 14 - 15 , 2002 ; akademische verlagsgesellschaft berlin , p. 253
g. j. goodhill and j. xu , network * 16 * , 5 ( 2005 ) .
n. v. swindale , network * 7 * , 161 ( 1996 ) .
m. gmann , self - organization between manifolds of euclidean and non - euclidean geometry by cooperation and competition , ph.d .
thesis ( in german ) , universitt stuttgart ( 2006);internet : www.itp1.uni-stuttgart.de/publikationen/guessmann_doktor_2006.pdf .
w. wischert , a. wunderlin , a. pelster , m. olivier , and j. groslambert , phys .
e , * 49 * , 203 ( 1994 ) .
m. gmann , a. pelster , and g. wunner , ann .
( leipzig ) * 16 * , 395 ( 2007 ) s. weinberg , gravitation and cosmology principles and applications of the general theory of relativity ( john wiley & sons , new york , 1972 ) .
c. von der malsburg , neural network self - organization ( i ) self - organization in the development of the visual system
. lecture notes ( 2000 ) ; internet : www.neuroinformatik.ruhr - uni - bochum.de / vdm/. h. risken , the fokker - planck equation .
methods of solution and applications , second edition ( springer , berlin , 1989 ) .
w. horsthemke and r. lefever , noise - induced transitions ( springer , new york , 1984 ) .
e. grigorieva , h. haken , s.a .
kashchenko , and a. pelster , physica d * 125 * , 123 ( 1999 ) c. simmendinger , a. pelster , and a. wunderlin , phys .
e * 59 * , 5344 ( 1999 ) m. schanz and a. pelster , phys .
e * 67 * , 056205 ( 2003 ) .
m. schanz and a. pelster , siam journ .
. dyn . syst .
* 2 * , 277 ( 2003 ) .
s. f. brandt , a. pelster , and r. wessel , phys .
e * 74 * , 036201 ( 2006 ) .
s. f. brandt , a. pelster , and r. wessel , eprint : physics 0701225 . | we generalize a model of hussler and von der malsburg which describes the self - organized generation of retinotopic projections between two one - dimensional discrete cell arrays on the basis of cooperative and competitive interactions of the individual synaptic contacts .
our generalized model is independent of the special geometry of the cell arrays and describes the temporal evolution of the connection weights between cells on different manifolds . by linearizing the equations of evolution around the stationary uniform state we determine the critical global growth rate for synapses onto the tectum where an instability arises . within a nonlinear analysis
we use then the methods of synergetics to adiabatically eliminate the stable modes near the instability .
the resulting order parameter equations describe the emergence of retinotopic projections from initially undifferentiated mappings independent of dimension and geometry . |
precision measurements allow us to probe physics at much higher energy scales than the masses of the particles directly involved in experimental reactions by exploiting contributions from quantum loops .
a prominent example is the global fit of the standard model ( sm ) to electroweak precision data , routinely performed by the lep electroweak working group and others ( for latest results see @xcite ) , which demonstrated impressively the predictive power of electroweak unification and quantum loop corrections . several theoretical libraries within and beyond the sm have been developed in the past , which allowed to constrain the unbound parameters of the sm and models of new physics .
however , most of these programs are relatively old , were implemented in outdated programming languages , and are difficult to maintain in line with the theoretical and experimental progress expected during the forthcoming era of the lhc .
these considerations led to the development of the generic fitting package _
gfitter _
@xcite , designed to provide a modular framework for complex fitting tasks in high - energy physics , like model testing and parameter estimation problems .
_ gfitter _ is implemented in c++ and relies on root functionality .
it consists of a core package providing the tools for data handling , fitting and statistical analyses and allows a consistent treatment of statistical , systematic and theoretical errors , possible correlations , and inter - parameter dependencies .
tools provided for statistical analyses include e.g. parameter scans , contours , monte carlo ( mc ) toy analyses and goodness - of - fit @xmath2-value evaluation .
more details on the framework can be found at @xcite . _
gfitter _ performs the minimisation of a @xmath6 test statistics quantifying the deviation of the experimental data from the predictions in a certain physics model .
the theoretical calculations are implemented via plug - in libraries for the _ gfitter _ framework . in this paper
we report results which are obtained using the first libraries implemented in the _ gfitter _ package : sm predictions of the electroweak precision observables and predictions of @xmath4 and @xmath5 physics observables in a model with two higgs doublets ( 2hdm ) .
in the global electroweak fit with _ gfitter _
state - of - the - art calculations are compared with the most recent experimental data to constraint the free parameters of the fit and to test the goodness - of - fit .
the sm parameters relevant for the global electroweak analysis are the coupling constants of the electromagnetic , weak and strong interactions , and the masses of the elementary bosons and fermions .
electroweak unification and simplifications arising from fixing parameters with insignificant uncertainties compared to the sensitivity of the fit allow to reduce the number of free fit parameters .
the remaining free parameters are the coupling parameters @xmath7 and @xmath8 , the masses @xmath9 , @xmath10 , @xmath11 , @xmath12 and @xmath13 . in addition , four free parameters enter to include the theoretical uncertainties of @xmath14 , @xmath15 and the electroweak form factors @xmath16 and @xmath17 . for the prediction of the electroweak precision observables
as measured by the lep , slc and tevatron experiments the most up - to - date calculations are implemented in the _ gfitter _
sm library using the oms scheme .
wherever possible the results have been cross - checked against the zfitter package @xcite .
the full two - loop and leading beyond - two - loop correction are available for the computation of @xmath14 and @xmath15 @xcite .
the partial and total widths of the @xmath18 are known to leading order , while for the second order only the leading @xmath19 corrections are available . among the new developments included
is the nnnlo perturbative calculation of the massless qcd adler function @xcite , contributing to the vector and axial - vector radiator functions in the prediction of the @xmath18 hadronic width ( and other observables ) .
it allows to fit the strong coupling constant with unique theoretical accuracy .
more details on the theoretical computations in _ gfitter _ can be found in @xcite . as a function of @xmath13 for the _ complete fit_.
the solid ( dashed ) lines give the results when including ( ignoring ) theoretical errors .
the minimum @xmath20 of the fit including theoretical errors is used for both curves in each plot to obtain the offset - corrected @xmath20 ; right : contours of 68% , 95% and 99% cl obtained from scans of fits with fixed variable pairs @xmath14 vs. @xmath12 for three sets of fits explained in the main text .
the horizontal bands indicate the @xmath21 regions of measurements ( world averages ) . , title="fig:",width=332 ] as a function of @xmath13 for the _ complete fit_. the solid ( dashed ) lines give the results when including ( ignoring ) theoretical errors .
the minimum @xmath20 of the fit including theoretical errors is used for both curves in each plot to obtain the offset - corrected @xmath20 ; right : contours of 68% , 95% and 99% cl obtained from scans of fits with fixed variable pairs @xmath14 vs. @xmath12 for three sets of fits explained in the main text .
the horizontal bands indicate the @xmath21 regions of measurements ( world averages ) .
, title="fig:",width=332 ] the following experimental measurements are used : the mass and width of the @xmath18 boson , the hadronic pole cross section @xmath22 , the partial widths ratio @xmath23 , and the forward - backward asymmetries for leptons @xmath24 , have been determined by fits to the @xmath18 line - shape measured precisely at lep ( see @xcite and references therein ) .
measurements of the @xmath25 polarisation at lep @xcite and the left - right asymmetry at slc @xcite have been used to determine the lepton asymmetry parameter @xmath26 .
the corresponding @xmath27 and @xmath28-quark asymmetry parameters @xmath29 , the forward - backward asymmetries @xmath30 , and the widths ratios @xmath31 and @xmath32 , have been measured at lep and slc @xcite .
in addition , the forward - backward charge asymmetry measurement in inclusive hadronic events at lep was used to directly determine @xmath33 @xcite . for the running quark masses @xmath10 and @xmath11 the world average values are used .
for @xmath7 we use the most recent phenomenological result @xcite .
results presented in this paper are obtained using the combined lep and tevatron results on the mass and the width of the @xmath34 boson @xcite , @xmath35 , @xmath36 , and the latest combined result on the top mass @xcite , @xmath37 , presented at this conference . the direct searches for the sm higgs boson at lep @xcite and the most recent results from the tevatron @xcite , leading to a 95% confidence level ( cl ) exclusion for @xmath38 and at @xmath39 respectively , are included using a gaussian approach that quantifies the difference between the observed test statistics ( the log - likelihood ratios ) and the expected values for the s+b hypothesis using the published values of the respective confidence level ( @xmath40 ) .
a contribution to the @xmath6 estimator of the fit is derived for each higgs mass .
we perform global fits in two versions : the _ standard ( `` blue - band '' ) fit _ makes use of all the available information except for the direct higgs searches ; the _ complete fit _ uses also the constraints from the direct higgs searches . due to the restricted space available
only example results of the electroweak fit are reported in the following .
more fit results and thorough studies of its statistical properties are given in @xcite where also the perspectives for the lhc , ilc and gigaz data are discussed .
the _ standard _
( _ complete _ ) _ fit _ converges at the global minimum value @xmath41 ( @xmath42 ) for 13 ( 14 ) degrees of freedom .
the estimation for @xmath13 from the _ standard fit _ without the direct higgs searches is @xmath43 and the @xmath44 and @xmath45 intervals are respectively @xmath46\,\rm gev$ ] and @xmath47\,\rm gev$ ] .
the _ complete fit _ represents the most accurate estimation of @xmath13 considering all available data .
we find @xmath48 .
the resulting @xmath49 curve versus @xmath13 is shown in fig .
[ fig : chi2lep ] ( left ) .
the shaded band indicates the influence of theoretical uncertainties .
the one , two and three standard deviations from the minimum are indicated by the crossings with the corresponding horizontal lines .
the @xmath44 and @xmath45 allowed regions of @xmath13 , including all errors , are @xmath50\ , \rm gev$ ] and @xmath51\;{\rm and}\;[180,225]]\ , \rm gev$ ] , respectively .
the inclusion of the direct higgs search results from lep leads to a strong rise of the @xmath49 curve below @xmath52 .
the data points from the searches at the tevatron , available in the range @xmath53 increases the @xmath49 estimator for higgs masses above @xmath54 beyond that obtained from the _ standard fit_. the strong coupling at the @xmath18-mass scale is determined by the _
complete fit _ with @xmath55 where the first error is experimental and the second due to the truncation of the perturbative qcd series
. figure [ fig : chi2lep ] ( right ) compares the direct measurements of @xmath14 and @xmath12 , shown by the shaded / green @xmath21 bands , with the 68% , 95% and 99% cl obtained for three sets of fits : the largest / blue ( narrower / yellow ) allowed regions are derived from the _ standard fit _ ( _ complete fit _ ) excluding the measured values in the fits . the inclusion of the lep and tevatron higgs searches significantly impacts the constraints obtained .
good agreement between direct measurements and indirect fit results is observed .
the third set of fits ( narrowest / green ) results from the _ complete fit _ including the measured values .
hence it uses all available information and leads to the narrowest allowed region .
the @xmath2-value of the global sm fit , quantifying the probability of wrongly rejecting the sm hypothesis , has been evaluated by means of toy mc experiments . for each mc experiment ,
the _ complete fit _ is performed yielding the @xmath56 distribution shown by the light shaded histogram in fig .
[ fig:2hdm - combined ] ( left ) .
the monotonously decreasing curves give the @xmath2-value of the sm fit as a function of @xmath56 obtained by integrating the distribution between @xmath56 and infinity .
the value of the global sm fit is given by @xmath57 , where the first error is statistical and the second accounts for the shift resulting from theoretical uncertainties .
as an example for a study beyond the sm we investigate models with an extended higgs sector of two doublets . in the type - ii 2hdm
, we constrain the mass of the charged higgs and the ratio of the vacuum expectation values of the two higgs doublets using current measurements of observables from the @xmath4 and @xmath5 physics sectors and their most recent theoretical 2hdm predictions , namely @xmath58 @xcite , the branching ratio ( br ) of @xmath59 @xcite , the br of leptonic decays of charged pseudoscalar mesons ( @xmath60 @xcite , @xmath61 @xcite and @xmath62 @xcite ) and the br of the semileptonic decay @xmath63 @xcite . for each observable , individual constraints have been derived in the ( @xmath64 ) plane .
figure [ fig:2hdm - combined ] ( right ) displays the resulting 95% _ excluded _ regions derived assuming gaussian behaviour of the test statistics , and one degree of freedom .
the figure shows that @xmath65 is mainly sensitive to @xmath66 excluding small values .
br(@xmath67 ) is only sensitive to @xmath66 for values below @xmath681 . for larger values
it provides an almost constant exclusion of a charged higgs lighter than @xmath68@xmath69 . for all leptonic observables
the 2hdm contribution can be either positive or negative since signed terms enter the prediction of the brs resulting in a two - fold ambiguity in the @xmath70 space .
in addition , we have performed a global fit combining the information from all observables . for the cl calculation in the 2-dim plane we performed toy mc tests in each scan point which allows to avoid the problem of ambiguities in the effective number of degrees of freedom .
the 95% cl excluded region obtained are indicated in fig .
[ fig:2hdm - combined ] ( right ) by the area below the single solid line .
we can exclude a charged higgs mass below @xmath71 independently of @xmath72 .
this limit increases towards larger @xmath72 , e.g. , @xmath73 are excluded for @xmath74 .
distribution of a toy mc simulation ( open histogram ) , the corresponding distribution for a _ complete fit _ ignoring theoretical uncertainties ( shaded / green histogram ) , an ideal @xmath6 distribution assuming a gaussian case with @xmath75 ( black line ) and the @xmath2-value as a function of the @xmath56 of the global fit ; right : 95% cl exclusion regions in the ( @xmath76 ) plane from individual 2hdm constraints and the toy - mc - based result ( hatched ) from the combined fit overlaid . ,
title="fig:",width=332 ] distribution of a toy mc simulation ( open histogram ) , the corresponding distribution for a _ complete fit _ ignoring theoretical uncertainties ( shaded / green histogram ) , an ideal @xmath6 distribution assuming a gaussian case with @xmath75 ( black line ) and the @xmath2-value as a function of the @xmath56 of the global fit ; right : 95% cl exclusion regions in the ( @xmath76 ) plane from individual 2hdm constraints and the toy - mc - based result ( hatched ) from the combined fit overlaid .
, title="fig:",width=332 ] | the global fit of the standard model to electroweak precision data , routinely performed by the lep electroweak working groups and others , has been revisited in view of ( _ i _ ) the development of the new generic fitting package , _ gfitter _ , ( _ ii _ ) the insertion of constraints from direct higgs searches at lep and tevatron , and ( _ iii _ ) a more thorough statistical interpretation of the results .
this paper describes the _
gfitter _ project , and presents state - of - the - art results for the global electroweak fit in the standard model , and for a model with an extended higgs sector .
example results are an estimation of the mass of the higgs boson ( @xmath0 ) and a forth - order result for the strong coupling strength ( @xmath1 ) . using toy monte carlo techniques
the @xmath2-value of the sm has been determined ( @xmath3 ) . as an example of a new physics model constraints
are derived for the two higgs doublet model of type - ii using observables from the @xmath4 and @xmath5 physics sectors . |
the traditional description of elastic fields produced by dislocations is based on the classical theory of linear elasticity
. however the classical dislocation theory breaks down in the dislocation core region .
the elastic fields due to dislocations which are calculated within the theory of elasticity contain singularities at the centre of the dislocation .
this is unfortunate since the dislocation core is an important region in metallic plasticity and dislocations and failure occur because of the high shear stress action in this region . clearly
, such singularities are unphysical and an improved model should eliminate them . on the other hand , in conventional plasticity theory no internal length scale enters the constitutive relations and no size effects are predicted . a field theory of elastoplasticity is proposed as a means of describing plastic and elastic deformation even in the dislocation core . in this theory
the elementary acts of plastic deformation at the microscopical level are investigated .
in fact , the dislocations are considered as elementary carriers of plasticity . from the field
theoretical point of view , dislocations bring new special degrees of freedom ( e.g. anholonomity ) and their presence leads to a specific response with the dimension of a moment stress . in straightforward manner , a new internal characteristic length scale enters the constitutive relation between the dislocation density tensor and the moment stress tensor . in this dislocation theory
it turns out that the force stresses , elastic and plastic strains , elastic and plastic distortions and displacement fields due to dislocations contain no singularities and they are finite .
in fact , they vanish at the centre of dislocations .
moreover , the extremum stress and strain occur at a short distance away from the dislocation line .
the plan of the paper is as follows . in section [ elastoplasticity ] , i present the basics of the elastoplastic field theory of dislocations .
the non - singular solutions of straight screw and edge dislocations calculated in the proposed theory of elastoplasticity are considered in section [ screw ] and [ edge ] , respectively .
section [ concl ] concludes the paper .
in this section we discuss how the dislocations can be described in the frame of field theory of elastoplasticity .
we shall in particular consider such plastic deformations which originate from the presence of dislocations . in elastoplasticity
the elastic incompatible distortion is given as @xcite @xmath0 it is an additive decomposition of the elastic distortion into compatible and purely incompatible distortion .
this decomposition can be justified by the help of the translational gauge theory of defects @xcite .
the displacement field @xmath1 gives rise to a compatible distortion and the tensor @xmath2 is the proper incompatible part of the distortion .
thus , it is not a formal additive decomposition .
note that the distortion is dimensionless . in elastoplasticity
the linear elastic strain tensor is given by means of the incompatible distortion tensor as @xmath3 the ( symmetric ) force stress is the response quantity ( excitation ) with respect to the strain and is given by the generalized hooke s law for an isotropic medium @xmath4 where @xmath5 , @xmath6 are shear modulus and poisson s ration , respectively .
the ( symmetric ) force stress has to satisfy the force equilibrium condition @xmath7 the skew - symmetric part of the distortion tensor defines the elastic rotation of a dislocation @xcite @xmath8 the rotation vector gives rise to a rotation gradient ( dewit s bend - twist tensor ) @xmath9 the total strain @xmath10 and distortion @xmath11 are defined in terms of the displacement field @xmath1 .
they must be compatible . in the presence of dislocations , the total strain is not completely elastic , but a part of it is plastic @xcite so that @xmath12 here @xmath13 is the plastic strain and @xmath14 the plastic distortion which are not derivable from the displacement field when the plastic strain is incompatible . in performing the differentiations of the displacement @xmath1
we obtain the elastic distortion @xmath15 plus excess terms which we identify with the plastic distortion @xmath14 .
we notice that this decomposition of @xmath16 is close to the green - naghdi decomposition ( see @xcite ) .
the dislocation density tensor is defined by means of the distortion tensor @xmath17 and @xmath18 physically , the incompatible distortion corresponds to dislocations .
the dislocation density tensor has the dimension of an inverse length .
it is a fundamental quantity in plasticity because the dislocation is the elementary carrier of plasticity .
the usual dislocation density tensor @xmath19 is recovered by ( see also @xcite ) @xmath20 here the index @xmath21 indicates the direction of the burgers vector , @xmath22 the dislocation line direction .
thus , the diagonal components of @xmath19 represent screw dislocations , the off - diagonal components edge dislocations . in this approach the dislocation density is not given a priori as a delta function .
it follows from the physics of the problem .
( [ torsion2 ] ) and ( [ alpha - ij ] ) justify the identification of @xmath23 with the plastic distortion .
the dislocation density tensor satisfies the translational bianchi identity @xmath24 which means that dislocations do not end inside the medium .
the moment stress which is the response quantity ( excitation ) to the dislocation density is given by ( see @xcite ) @xmath25 or with @xmath26 @xmath27 obviously , ( [ moment2 ] ) and ( [ moment3 ] ) are ( linear ) constitutive relations between dislocation density and moment stress in an isotropic medium .
the coefficient @xmath28 has the dimension of a force .
the field equation in elastoplasticity for the force stress in an isotropic medium is proposed as the following inhomogeneous helmholtz equation @xcite @xmath29 where @xmath30 is the stress tensor obtained for the same traction boundary - value problem within the `` classical '' theory of dislocations . in the field theory of elastoplasticity eq .
( [ stress - fe ] ) is obtained from the moment stress equilibrium condition ( see @xcite ) .
it is important to note that ( [ stress - fe ] ) agrees with the field equation for the stress field in eringen s nonlocal elasticity @xcite and in gradient elasticity @xcite .
the factor @xmath31 has the physical dimension of a length and therefore it defines an internal characteristic length ( dislocation length scale ) . using the inverse of the generalized hooke s law and ( [ stress - fe ] )
we obtain an inhomogeneous helmholtz equation for the strain fields ( see @xcite ) @xmath32 where @xmath33 is the classical strain tensor .
equation ( [ strain - fe ] ) is similar to the equation for strain of the gradient theory used by gutkin and aifantis @xcite if we identify @xmath34 with their corresponding gradient coefficient ( see , e.g. , equation ( 4 ) in @xcite ) . from the fact that the field equation of the force stress ( [ stress - fe ] ) has the same form in elastoplastic field theory , strain gradient elasticity and nonlocal elasticity some relations for the corresponding solutions will follow .
we assume that the stress and strain fields at infinity should have the same form for both the classical and elastoplastic field theory .
consider an infinitely long screw dislocation whose dislocation line coincides with the @xmath35-axis of a cartesian coordinate system . due to the symmetry of the problem
, we choose the burgers vector , which is parallel to the dislocation line , in @xmath35-direction : @xmath36 , @xmath37 .
we solve the force stress equilibrium condition ( [ feq ] ) identically by a so - called stress function ansatz . using the stress function ansatz @xcite the elastic stress of a volterra screw dislocation , in cartesian coordinates , reads @xmath38 or , in cylindrical coordinates , @xmath39 where @xmath40 and @xmath41 . here
, @xmath42 is the well - known stress function of elastic torsion , sometimes called prandtl s stress function .
it is given by ( see , e.g. , @xcite ) @xmath43 this is green s function of the two - dimensional potential equation @xmath44 here @xmath45 is the two - dimensional dirac delta function and @xmath46 denotes the two - dimensional laplacian @xmath47 .
obviously , the `` classical '' stress fields are singular at the dislocation line . for the modified stress we make the ansatz @xmath48 where @xmath49 is called the modified prandtl stress function .
substituting the stress functions into ( [ stress - fe ] ) , we get the following inhomogeneous helmholtz equation @xmath50 the solution of the modified stress function of a screw dislocation is given by @xmath51 where @xmath52 is the modified bessel function of the second kind and of order @xmath53 .
consequently , we find the force stresses @xmath54 and in cylindrical coordinates @xmath55 the appearance of the modified bessel function in ( [ t - cart ] ) and ( [ t - cyl ] ) leads to the elimination of classical singularity @xmath56 at the dislocation line . the modified stress field ( [ t - cart ] ) agrees with the stress field calculated by eringen @xcite within his version of nonlocal elasticity . additionally , it is interesting to note that the stress field ( [ t - cart ] ) is the same as the one obtained by gutkin and aifantis @xcite in their version of gradient elasticity .
the stress @xmath57 has its extreme value @xmath58 at @xmath59 , whereas the stress @xmath60 has its extreme value @xmath61 at @xmath62 .
we notice that the extremum stress may be identified with the theoretical shear strength .
the factor @xmath63 should be fitted by comparing predictions of the theory with experimental results . in general
, the parameter @xmath63 can be used to determine the width of a dislocation and the amplitude of the force stress .
let us now calculate the distortion of a screw dislocation . the distortion @xmath15 is given in terms of the stress function @xmath64 where the two functions @xmath65 and @xmath66 are used to express the antisymmetric part of the distortion ( @xmath67}\equiv \omega_1 $ ] and @xmath68}\equiv -\omega_2 $ ] ) .
eventually , @xmath65 and @xmath66 are determined from the conditions @xmath69 one finds for the distortion tensor of the screw dislocation @xmath70 and for elastic rotation vector @xmath71 the rotation vector is in agreement with the result calculated in the cosserat theory @xcite .
the incompatible elastic strain reads @xmath72 by means of the distortion tensor ( [ dist - screw ] ) the effective burgers vector , @xmath73 , for a circular circuit of radius @xmath74 is given by @xmath75 where @xmath76 is the burgers circuit .
it depends on the radius @xmath74 .
in fact , we find @xmath77 and @xmath78 .
this effective burgers vector differs appreciably from the constant value @xmath79 in the core region from @xmath80 up to @xmath81 .
thus , it is suggestive to take @xmath82 as the characteristic length ( dislocation core radius ) .
therefore , in the field theory of elastoplasticity the dislocation has a core in quite natural manner . outside this core region
the burgers vector reaches its constant value .
accordingly , the classical and the elastoplastic solution coincide outside the core region .
if we use the decomposition ( [ dist2 ] ) of the distortion into the compatible and purely incompatible distortion , the displacement field of a screw dislocation turns out to be ( see also @xcite ) @xmath83 where @xmath84 is multi - valued .
thus , the proper incompatible part of the distortion is the ( negative ) plastic distortion which is confined in the dislocation core region @xmath85 it fulfils eq .
( [ alpha - ij ] ) .
the plastic strain reads @xmath86 now we are able to calculate the dislocation density by means of the distortion tensor .
we obtain @xmath87 of course , this dislocation density satisfies the condition ( [ bianchi ] ) . in the limit as @xmath88 , the elastoplastic result ( [ alpha - zz ] ) converts to the classical dislocation density @xmath89 . by the help of ( [ moment3 ] ) the localised moment
stresses caused by the screw dislocation can be expressed in terms of the dislocation density as @xmath90 and @xmath91 accordingly , moment stresses of twisting - type occur at all positions where the dislocation density @xmath92 is non - vanishing . when @xmath88 , the moment stresses vanish .
the far - reaching rotation gradients of a screw dislocation read @xmath93 they are in agreement with the expressions calculated within the theory of cosserat media ( see @xcite ) .
consider an infinitely long edge dislocation whose dislocation line coincides with the @xmath35-axis while the burgers vector is parallel to the @xmath94-axis : @xmath95 , @xmath96 .
the extra half plane lies in the plane @xmath97 . in order to satisfy the equilibrium condition ( [ feq ] )
we use the second order stress function @xmath98 and specialize to the plane problem of an edge dislocation by setting @xmath99 .
the classical stress field of a straight edge dislocation can be given in terms of airy s stress function according to @xmath100 airy s stress function @xcite @xmath101 fulfils the following inhomogeneous bipotential ( or biharmonic ) equation @xmath102 for the modified stress we make the following stress function ansatz @xmath103 in addition , the strain is given in terms of the stress function as @xmath104 substituting ( [ stress - ansatz ] ) and ( [ stress - ansatz2 ] ) into ( [ stress - fe ] ) we get the inhomogeneous helmholtz equation @xmath105 the solution of the modified stress function of a straight edge dislocation is given by @xcite @xmath106 where the first piece is airy s stress function . by means of eqs .
( [ stress - ansatz ] ) and ( [ f_edge1 ] ) , the modified stress of a straight edge dislocation is given as @xmath107 the trace of the stress tensor @xmath108 produced by the edge dislocation in an isotropic medium is @xmath109 let us now discuss some details of the stress fields in the core region .
the stress fields have no artificial singularities at the core and the extremum stress occurs at a short distance away from the dislocation line .
in fact , when @xmath110 , we have @xmath111 and thus @xmath112 . it can be seen that the stresses have the following extreme values : @xmath113 at @xmath114 , @xmath115 at @xmath116 , @xmath117 at @xmath118 , and @xmath119 at @xmath62 .
the stresses @xmath120 , @xmath121 and @xmath122 are modified near the dislocation core ( @xmath123 ) .
note that @xmath124 can be identified with the theoretical shear strength .
the stresses @xmath125 and @xmath108 are modified in the region : @xmath126 .
far from the dislocation line ( @xmath127 ) the elastoplastic and the classical solutions of the stress of an edge dislocation coincide .
thus , the characteristic internal length @xmath31 determines the position and the magnitude of the stress extrema .
it is interesting and important to note that the solution ( [ t_zz ] ) agrees precisely with the gradient solution given by gutkin and aifantis @xcite ( with @xmath128 , @xmath129 is the gradient coefficient ) . for the elastic strain of an edge dislocation
we find @xmath130 which is in agreement with the solution given by gutkin and aifantis @xcite in the framework of strain gradient elasticity .
the plane - strain condition @xmath131 of the classical dislocation theory is valid .
the components of the strain tensor have the following extreme values ( @xmath132 ) : @xmath133 at @xmath134 , @xmath135 at @xmath136 , @xmath137 at @xmath138 , and @xmath139 at @xmath118
. it is interesting to note that @xmath140 is much smaller than @xmath141 within the core region ( see also @xcite ) .
the dilatation @xmath142 reads @xmath143 now we calculate the distortion of an edge dislocation .
the distortion @xmath15 is given in terms of the stress function ( [ f_edge1 ] ) : @xmath144 where @xmath145 is used to express the antisymmetric part of the distortion , @xmath146}$ ] .
eventually , @xmath145 is determined from the conditions : @xmath147 we find for the elastic distortion of the edge dislocation @xmath148 and for the rotation @xmath149 eq .
( [ rot_z ] ) is in agreement with the rotation vector calculated in the linear theory of dislocations in the cosserat continuum @xcite .
the far fields of ( [ dist_yy])([rot_z ] ) are identical to the classical ones given in @xcite .
now we decompose the elastic distortion ( [ dist_yy ] ) into its compatible and purely incompatible part by the help of eq .
( [ dist2 ] ) .
again , we may interpret the proper incompatible part as the negative plastic distortion . in components
it reads @xmath150 of course , it fulfils eq .
( [ alpha - ij ] ) .
this proper incompatible distortion is exactly the same as @xmath151 and @xmath152 of a screw dislocation ( see eq . ( [ plastic - dist - screw ] ) ) .
the appearance of such proper incompatible distortion is a typical result in elastoplasticity .
the plastic strain reads @xmath153 the corresponding compatible part can be given in terms of a multi - valued displacement field @xmath154 the core radius of the edge dislocation can be estimated as @xmath82 .
additionally , we find from the distortion the incompatible part @xmath155 and a single - valued displacement field @xmath156 this means that @xmath157 and @xmath158 are proper compatible distortions .
therefore , the field @xmath159 agrees ( up to a constant term ) with the corresponding formula in gradient elasticity ( see @xcite ) . the displacement field
@xmath159 differs from the classical one in the region @xmath123 . in the framework of elastoplasticity
the displacement fields ( [ u_x ] ) and ( [ u_y ] ) have no singularity .
therefore , these displacement fields can be used to model the dislocation core . in this way
, one can estimate the displacements , elastic strains and stresses near the dislocation cores and compare them with hrtem micrographs and atomic calculations .
the far fields of the displacements ( [ u_x ] ) and ( [ u_y ] ) are identical to the classical ones ( see , e.g. , @xcite ) .
it is worth noting that within the peierls - nabarro dislocation model the displacement @xmath159 is identically zero .
finally , the effective burgers vector of the edge dislocation can be calculated as @xmath160 again , we find @xmath161 and @xmath162 .
the effective burgers vector @xmath163 differs from the constant burgers vector @xmath79 in the region @xmath126 .
therefore , the core radius can be taken as @xmath82 .
for the value of @xmath164 the core radius is @xmath165 and for the value of @xmath166 the core radius is @xmath167 ( @xmath168 is the lattice constant ) . note that the effective burgers vector @xmath163 of an edge dislocation has the same form as the effective burgers vector @xmath169 of a screw dislocation ( see eq .
( [ burger - screw ] ) ) .
the proper incompatible part of the elastic distortion gives rise to a localized dislocation density and moment stress tensor .
we find for the dislocation density of an edge dislocation @xmath170 which satisfies the translational bianchi identity ( [ bianchi ] ) .
the dislocation density is short - reaching .
it is interesting to note that the dislocation density of an edge dislocation has the same form as the dislocation density @xmath92 of a screw dislocation ( see eq .
( [ alpha - zz ] ) ) . in the limit as @xmath88 ,
the elastoplastic result ( [ disl - den - edge ] ) converts to the classical dislocation density @xmath171 .
the localized moment stress of bending type is given by the help of ( [ moment3 ] ) according to @xmath172 this expression is very close to the moment stress of an edge dislocation given in @xcite ( see also the remarks in @xcite ) .
dewit s bend - twist tensor of an edge dislocation reads @xmath173 it is in agreement with the expression calculated in the theory of cosserat media ( see @xcite ) .
the field theory of elastoplasticity has been employed to consider straight screw and edge dislocations .
the dislocation densities of straight screw and edge dislocations obtained in elastoplastic field theory agree with eringen s two - dimensional nonlocal modulus ( nonlocal kernel ) used in @xcite .
consequently , every component of the dislocation density tensor is green s function of the helmholtz equation : @xmath174 the characteristic internal length in elastoplasticity is @xmath31 .
this length may be selected to be proportional to the lattice parameter @xmath168 for a single crystal , i.e. @xmath175 where @xmath176 is a non - dimensional constant which can be determined by one experiment @xcite .
for @xmath177 we recover classical elasticity . in @xcite the choice @xmath178 and in @xcite the choice @xmath179
are proposed .
that length specifies the plastic region and should be estimated by means of experimental observations and computer simulations of the core region .
exact analytical solutions for the displacements , elastic and plastic strain fields , and force stresses of dislocations have been reported which demonstrate the elimination of any singularity at the dislocation line .
it has been shown that the force stresses achieve their extreme values at a short distance away from the dislocation line .
therefore , we are able to obtain finite strain and stress in the ( linear ) field theory of elastoplasticity
. these maximum values may serve as measures of the critical stress level of fracture .
the stress fields are in agreement with the ones obtained by gutkin and aifantis @xcite .
additionally , we have calculated the moment stresses and dewit s bend - twist tensor for straight screw and edge dislocations .
the author acknowledges the support provided by the max - planck - institute for mathematics in the sciences .
in the expressions given by ( [ u - t ] ) , ( [ plastic - dist - screw ] ) , ( [ plastic - dist - edge ] ) and ( [ u_x ] ) we have considered @xmath84 as a multi - valued field . on the other hand , in the defect theory @xcite @xmath84
is usually used as single - valued and discontinuous function .
it is made unique by cutting the half - plane @xmath180 at @xmath181 and assuming @xmath84 to jump from @xmath182 to @xmath183 when crossing the cut .
if one uses the single - valued discontinuous form for @xmath84 , we obtain @xcite for the screw dislocation @xmath184 with @xmath185\big)\big\},\end{aligned}\ ] ] and for the edge dislocation @xmath186 with @xmath187\big)\big\},\nonumber\\ & \phi_{yx}=\phi_{yy}=0.\nonumber\end{aligned}\ ] ] for a more detailed discussion see @xcite .
99 m. lazar , ann .
( leipzig ) * 9 * ( 2000 ) 461 ; .
m. lazar , j. phys .
* 35 * ( 2002 ) 1983 .
m. lazar , ann .
( leipzig ) * 11 * ( 2002 ) 635 ; .
m. lazar , j. phys .
* 36 * ( 2003 ) 1415 .
edelen and d.c .
lagoudas , _ gauge theory and defects in solids _ , in : _ mechanics and physics of discrete system _ , vol . 1 ,
sih , ed . ,
north - holland , amsterdam ( 1988 ) .
r. dewit , j. res .
( u.s . )
* 77a * ( 1973 ) 49 .
r. dewit , j. res .
( u.s . )
* 77a * ( 1973 ) 607 .
green and p.m. naghdi , arch
* 18 * ( 1965 ) 251 .
e. krner , _ continuum theory of defects _ , in : _ physics of defects _
( les houches , session 35 ) , r. balian et al .
north - holland , amsterdam ( 1981 ) p. 215 . a.c .
eringen , j. appl .
* 54 * ( 1983 ) 4703 . a.c .
eringen , _ nonlocal continuum theory for dislocations and fracture _ , in : _ the mechanics of dislocations _ , eds . e.c
. aifantis and j.p .
hirth , american society of metals , metals park , ohio ( 1985 ) p. 101 .
gutkin and e.c .
aifantis , scripta mater .
* 40 * ( 1999 ) 559 .
gutkin and e.c .
aifantis , scripta mater .
* 35 * ( 1996 ) 1353 .
gutkin and e.c .
aifantis , scripta mater .
* 36 * ( 1997 ) 129 .
e. krner , _ kontinuumstheorie der versetzungen und eigenspannungen _ , erg .
* 5 * ( 1958 ) p. 1 .
hirth and j. lothe , _ theory of dislocations _ , john wiley , new york ( 1982 ) .
nowacki , bull .
* 21 * ( 1973 ) 585 .
edelen , int .
* 34 * ( 1996 ) 81 .
s. kessel , z. ang .
* 50 * ( 1970 ) 547 .
n. nowacki , arch .
* 26 * ( 1974 ) 3 .
s. minagawa , appl .
* 5 * ( 1977 ) 85 .
hehl and e. krner , z. naturforschg .
* 20a * ( 1965 ) 336 .
e. krner , _ the continuized crystal as a model for crystals with dislocations _ , in : _ materials modelling : from theory to technology _ , ed .
matthews , iop publishing ltd ( 1992 ) p. 13 .
altan and e.c .
aifantis , scripta metall .
* 26 * ( 1992 ) 319 . | by means of linear theory of elastoplasticity , solutions are given for screw and edge dislocations situated in an isotropic solid .
the force stresses , strain fields , displacements , distortions , dislocation densities and moment stresses are calculated .
the force stresses , strain fields , displacements and distortions are devoid of singularities predicted by the classical elasticity . using the so - called stress function method we found modified stress functions of screw and edge dislocations .
dislocations ; elastoplasticity ; moment stress ; stress functions . |
quarkonia are an essential probe to study the properties of the quark gluon plasma ( qgp ) .
the suppression of @xmath0 due to color - screening effects in the medium was initially proposed as a direct evidence of the qgp formation @xcite .
however , the interpretation of the @xmath0 suppression is still a challenge due to the contributions from the regenerated @xmath0 by the recombination of @xmath11 pairs in the medium and the cold nuclear matter effects .
therefore it is important to have more precise @xmath0 measurements over a broad kinematic range and even cleaner @xmath1 state measurements .
the latter do not suffer from the regeneration contribution due to the much smaller @xmath12 pair cross section , i.e. @xmath13 @xmath14b @xcite while @xmath15 @xmath14b @xcite at top rhic energy . the newly installed muon telescope detector ( mtd ) , which provides both the di - muon trigger and the muon identification capability at mid - rapidity , opens the door to measuring quarkonia via the di - muon decay channel at star . compared to the di - electron decay channel , the di - muon decay channel suffers much less from bremsstrahlung and thus provides much better invariant mass resolution to separate different @xmath1 states . using the mtd di - muon trigger , the star experiment recorded data corresponding to an integrated luminosity of 28.3 @xmath16 in p@xmath2p collisions at @xmath3 gev in the rhic 2013 run , and 14.2 @xmath17 in au@xmath2au collisions at @xmath6 gev in the rhic 2014 run . in these proceedings , we report ( 1 ) the measurements of @xmath0 production in p+p collisions at @xmath3 gev ; and ( 2 ) the measurements of the nuclear modification factor ( @xmath7 ) for @xmath0 and the production of different @xmath1 states in au+au collisions at @xmath6 gev .
@xmath18 scaling of @xmath0 cross section scaled by the branching ratio @xmath19 in the di - muon decay channel ( blue star ) and in the di - electron decay channel ( red circle ) . ]
@xmath18 scaling of @xmath0 cross section scaled by the branching ratio @xmath19 in the di - muon decay channel ( blue star ) and in the di - electron decay channel ( red circle ) . ]
figure [ fig : run13jpsicrosssection ] shows the cross section of @xmath0 in p@xmath2p collisions at @xmath3 gev in the di - electron and di - muon decay channels for @xmath4 gev/@xmath5 .
the di - muon decay channel extends @xmath8 reach down to 0 gev/@xmath5 .
the results in these decay channels are consistent in the overlapping @xmath8 range of @xmath20 gev/@xmath5 .
the experimental results can be well described by cgc+nrqcd calculations at low @xmath8 @xcite and nlo nrqcd calculations at high @xmath8 @xcite .
figure [ fig : run13xtscaling ] shows the @xmath21 scaling of @xmath0 cross section @xcite .
the @xmath0 cross section in p@xmath2p collisions at @xmath3 gev follows the common trend as a function of @xmath18 at high @xmath8 .
the breaking of the @xmath18 scaling at low @xmath8 can be attributed to the soft processes .
@xmath7 as a function of @xmath8 at rhic ( red star ) and at lhc ( blue square , black circle ) .
the lines and bands indicate transport model calculations for rhic and lhc energies . ]
@xmath7 as a function of @xmath8 at rhic ( red star ) and at lhc ( blue square , black circle ) .
the lines and bands indicate transport model calculations for rhic and lhc energies . ]
figure [ fig : jpsiyieldpt ] shows the invariant yield of @xmath0 in au@xmath2au collisions at @xmath6 gev for different collision centralities .
the new results in the di - muon decay channel are consistent with previous results in the di - electron decay channel @xcite within uncertainties . the nuclear modification factor @xmath22 of @xmath0 in 0 - 40% central au@xmath2au collisions is shown in fig . [ fig : jpsiraavspt ] , compared with lhc results @xcite .
the strong suppression at rhic at low @xmath8 indicates that dissociation plays a significant role in this @xmath8 range .
the hint of the increasing trend of @xmath7 at rhic at high @xmath8 can be explained by formation - time effects and feed - down of @xmath19 hadrons .
the less suppression of @xmath0 at lhc at low @xmath8 indicates larger regeneration contribution due to higher charm cross section , while more suppression of @xmath0 at lhc at high @xmath8 indicates larger dissociation rate due to higher temperature of the medium .
transport models from tsinghua @xcite and texas a&m university ( tamu ) @xcite , including dissociation and regeneration effects , can qualitatively describe the @xmath8 dependence of rhic and lhc data .
nuclear modification factor @xmath7 for @xmath23 gev/@xmath5 as a function of @xmath24 . ]
nuclear modification factor @xmath7 for @xmath23 gev/@xmath5 as a function of @xmath24 . ] centrality dependence of @xmath0 cross section is shown in fig .
[ fig : jpsiraanpartlowpt ] for integrated @xmath8 and in fig .
[ fig : jpsiraanparthighpt ] for @xmath23 gev/@xmath5 . for integrated @xmath8
, both models can describe centrality dependence at rhic , but tend to overestimate suppression at lhc . for @xmath23 gev/@xmath5
, there is tension among models and data .
new measurements in the di - muon decay channel provide a distinguishing power for these transport models .
@xmath25 ratio for world - wide p+p data , and for heavy - ion collisions at rhic , and lhc energies . ]
@xmath25 ratio for world - wide p+p data , and for heavy - ion collisions at rhic , and lhc energies . ]
figure [ fig : run14upssignal ] shows the di - muon mass spectrum in @xmath1 state mass range in au@xmath2au collisions at @xmath6 gev .
we observe signs of an indication of @xmath9 signals in the di - muon decay channel .
the raw yields of @xmath1 states are obtained by a simultaneous fit to the like - sign and unlike - sign distributions . in this fit ,
( i ) the @xmath1 state masses are fixed to the pdg values and their widths are determined by simulation ; ( ii ) the ratio of @xmath26 is fixed to the value in p@xmath2p collisions ; and ( iii ) the shape of @xmath12 and drell - yan background is estimated using pythia .
figure [ fig : upsilonratio ] shows the fitted @xmath25 ratio compared with the world - wide p+p data @xcite and cms data @xcite . there is a hint of less melting of @xmath9 relative to @xmath10 at rhic than at lhc
we present the first @xmath0 and @xmath1 measurements in the di - muon decay channel at mid - rapidity at rhic . in p@xmath2p collisions at @xmath3 gev
, inclusive @xmath0 cross section can be described by cgc+nrqcd and nlo nrqcd model calculations for @xmath4 gev/@xmath5 . in au@xmath2au collisions at @xmath6 gev ,
we observe ( i ) clear @xmath0 suppression indicating dissociation ; ( ii ) @xmath0 @xmath7 can be qualitatively described by transport models including dissociation and regeneration despite a tension at high @xmath8 ; and ( iii ) there is a hint of less melting of @xmath9 relative to @xmath10 at rhic compared to that at lhc .
these measurements in au@xmath2au collisions will have better statistical precision by combining the similar amount of data recorded in the rhic 2016 run .
20 t. matsui and h. satz , _ phys . lett . b _ * 178 * 416 - 422 ( 1986 ) a. adare _ et
( phenix collaboration ) , _ phys .
lett . _ * 103 * 082002 ( 2009 ) l. adamczyk _ et .
( star collaboration ) , _ phys .
d _ * 86 * 072013 ( 2012 ) yan - qing ma and raju venugopalan , _ phys .
lett . _ * 113 * 192301 ( 2015 ) h. shao _ et .
al . _ , _ jhep _ * 05 * 103 ( 2015 ) b. abelev _ et .
( star collaboration ) , _ phys .
c _ * 80 * 041902 ( 2009 ) l. adamczyk _ et .
( star collaboration ) , _ phys .
b _ * 722 * 55 - 62 ( 2013 ) l. adamczyk _ et
( star collaboration ) , _ phys .
c _ * 90 * 024906 ( 2014 ) y. liu _ et .
_ , _ phys .
b _ * 678 * 72 - 76 ( 2009 ) k. zhou _ et .
_ , _ phys . rev .
c _ * 89 * 054911 ( 2014 ) x. zhao and r. rapp , _ phys . rev .
c _ * 82 * 064905 ( 2010 ) x. zhao and r. rapp , _ nucl .
a _ * 859 * 114 - 125 ( 2011 ) b. abelev _ et .
( alice collaboration ) , _ phys .
b _ * 734 * 314 - 327 ( 2014 ) s. chatrchyan _ et .
( cms collaboration ) , _ jhep _ * 05 * 063 ( 2012 ) l. adamczyk _ et .
( star collaboration ) , _ phys .
lett . _ * 111 * 052301 ( 2013 ) w. zha _ et .
_ , _ phys .
c _ * 88 * 067901 ( 2013 ) s. chatrchyan _ et .
( cms collaboration ) , _ phys .
_ * 109 * 222301 ( 2012 ) s. chatrchyan _ et .
( cms collaboration ) , _ jhep _ * 04 * 103 ( 2014 ) | we present the first @xmath0 and @xmath1 measurements in the di - muon decay channel at mid - rapidity at rhic using the newly installed muon telescope detector . in p@xmath2p collisions at @xmath3 gev
, inclusive @xmath0 cross section can be described by cgc+nrqcd and nlo nrqcd model calculations for @xmath4 gev/@xmath5 . in au@xmath2au collisions at @xmath6 gev ,
we observe ( i ) clear @xmath0 suppression indicating dissociation ; ( ii ) @xmath0 @xmath7 can be qualitatively described by transport models including dissociation and regeneration with a tension at high @xmath8 ; and ( iii ) hint of less melting of @xmath9 relative to @xmath10 at rhic compared to that at lhc . |
to fully specify the quark structure of the nucleon at twist two level , three parton distribution functions are needed : the well known momentum and helicity distribution @xmath3 and @xmath4 , and the transversity distribution @xmath5 .
this latter describes the probability density of finding transversely polarized quarks in a transversely polarized nucleon , and it is currently receiving a lot of attention , both from the experimental and from the theoretical point of view . due to its chiral odd nature , the transversity pdfs must be coupled to another chiral - odd function in order to build an observable . for this reason , transversity can not be measured in dis experiments , but only in semi - inclusive dis experiments ( sidis ) , where at least one hadron in the final state is detected . in sidis ,
different channels allow to access transversity : the azimuthal distribution of single hadrons , the azimuthal dependence of hadron pairs and the measurements of the transverse polarization of @xmath1 hyperons .
all these channels have been investigated by the compass collaboration .
compass @xcite is a fixed target experiment at the cern sps , with a physics program focused on the study of the nucleon spin structure . from 2002 to 2004 , compass took data with a longitudinally polarized muon beam of 160 gev / c and a @xmath0lid target . for about the 20% of the time , the target nucleons have been polarized transversely with respect to the beam direction , in order to allow the measurements of transverse spin effects .
the @xmath0lid material is characterized by a high dilution factor ( @xmath60.38 ) , and the polarization values p@xmath7 reached during data taking are around 50% .
the transversity pdf can be accessed in single hadron production via the so called collins effect @xcite . in this mechanism ,
the fragmentation function of transversely polarized quarks into unpolarized hadrons is composed by two parts , an unpolarized and a polarized one , depending on the quark spin ( collins ff ) .
more precisely , the ff shows an azimuthal dependence with respect to the plane defined by the quark momentum and the quark spin , meaning that the event yield can be written as : @xmath8 where @xmath9 and p@xmath7 have been already introduced , and @xmath10 is the transverse spin transfer coefficient from the initial to the struck quark . the collins angle @xmath11 is defined as @xmath12 , where @xmath13 is the angle of the transverse momentum of the outgoing hadron and @xmath14 is the azimuthal angle of the struck quark spin ( @xmath15 is the azimuthal angle of quark before the hard scattering ) .
@xmath16 is the collins asymmetry , proportional to the convolution of the collins fragmentation function and the transversity distribution : @xmath17 where @xmath18 is the fraction of available energy carried by the hadron , and @xmath19 is the hadron transverse momentum with respect to the virtual photon direction . comparing the number of hadrons produced in sidis reactions on nucleons polarized transversely in opposite directions , the modulation given by the collins angle gives access to the asymmetry @xmath16 . in order to select dis events in the data , interaction vertices with one incoming muon ,
one scattered muon and at least one outgoing hadron have been selected . to select the dis regime , the photon virtuality @xmath20 is taken above 1 ( gev / c)@xmath21 , the fractional energy of the virtual photon @xmath22 between 0.1 and 0.9 , and the invariant mass of the final hadronic state @xmath23 above 5 gev / c@xmath21 .
moreover , the following cuts on the hadron variables have been applied : @xmath24 and @xmath25 0.1 gev / c .
after all the cuts applied , the total number of positive and negative hadrons is 7 and 8.5@xmath26 , respectively .
r0.55 the collins asymmetries of unidentified hadrons @xcite for the full statistics are shown in fig .
[ fig : nonid ] , as a function of @xmath27 , @xmath28 and @xmath19 .
the asymmetries are small , and compatible with zero within the very good statistical accuracy , which is reaching @xmath291% in the central @xmath27 bins .
the systematic errors , investigated with extensive studies , have been found to be well below the statistical precision .
a nave interpretation of the results can be achieved in the parton model : restricting to the valence region , the small asymmetries on deuterium are due to a cancellation between the transversity pdf for the u and d quark .
this result is valid even if the collins ff are opposite in sign and similar in size , as is suggested by the non zero asymmetries on a proton target measured by the hermes collaboration @xcite .
r0.55 in order to perform a more sophisticated flavor separation analysis , the hadrons in the final state have to be identified . in compass ,
charged kaons and pions are identified a rich @xcite . due to the different cherenkov thresholds , pions and kaons can be identified for momenta larger than @xmath293 gev / c and 10 gev / c , respectively , and up to 50 gev / c , that corresponds to a 1.5 @xmath30 @xmath31-k mass separation .
neutral kaons have been selected in the data from the invariant mass of the decay products .
after the selection , the invariant mass spectra show a peak at the @xmath32 mass , with a signal - to - background ratio around 15 . the final sample of @xmath32s for the analysis has been selected considering a @xmath3320 mev / c@xmath21 mass window around the @xmath32 mass .
the final statistics consists of 5.2 and 4.5@xmath26 @xmath34 and @xmath35 , 0.9 and 0.6@xmath26 @xmath36 and @xmath37 , and 0.25@xmath26 @xmath32 .
the results obtained for pions and kaons @xcite are presented in fig .
[ fig : pik ] . also here , the asymmetries are small and compatible with zero within the statistical errors .
the collins asymmetry is not the only possible single spin asymmetry that can be measured with a transversely polarized target . in the complete sidis cross section expression ,
8 structure functions depending on the transverse target polarization are present , yielding 8 asymmetries with independent azimuthal modulations .
the sivers asymmetries is the most famous one ; it is related to the sivers pdf , that gives a measurement of the correlation of the intrinsic transverse momentum of an unpolarized quark in a transversely polarized nucleon .
results for the sivers asymmetries have been provided for unidentified @xcite and identified hadrons @xcite .
the measured sivers asymmetries are very small and compatible with zero , indicating also in this case a cancellation between the u and d - quark contribution in an isoscalar target .
another mechanism sensitive to the transversity pdf is the two hadrons production .
the event selection for this analysis is the same as described for the single hadron analysis , apart for the lower cut on the hadron @xmath28 , that has been released to 0.1 .
all the events with at least two hadrons in the final state have been considered , and a cut on the sum of the relative energies , @xmath38 , has been applied to remove exclusive production . as for the collins mechanism , the fragmentation of a transversely polarized quark into a pair of hadrons is expected to depend on an azimuthal angle @xcite , yielding for the number of events : @xmath39 where @xmath40 is defined as @xmath41 ; @xmath42 is the azimuthal angle between the lepton scattering plane and the plane containing the virtual photon momentum * q * and the component @xmath43 of the relative hadron momentum * * r**@xmath44(*p@xmath45 * -*p@xmath46 * ) which is perpendicular to the summed hadron momentum * * p@xmath47**@xmath48**p@xmath45**+*p@xmath46*. the asymmetry @xmath49 is proportional to the transversity function convoluted with the fragmentation function describing the two hadrons production , @xmath50 : @xmath51 where @xmath52 is the invariant mass of the hadron pair .
two different analyses have been performed . in the first one ,
the hadron pair is composed of a positive and a negative hadron : the hadrons have been identified with the rich , so that four combinations have been considered : @xmath53 , @xmath54 , @xmath55 , @xmath56 . in the second analysis , the hadrons have been ordered considering the leading and sub - leading hadrons ( defined by the corresponding @xmath28 ) ; in this analysis , also combination of hadrons with the same charge are possible , leading to 16 combinations .
all these asymmetries have been measured as a function of @xmath27 , @xmath28 and @xmath57 .
no clear indication of a signal different from zero has been obtained from all the measurements .
in 2002 - 2004 compass has provided a complete set of measurement of transverse spin effects using a deuterium target .
all the measured asymmetries are small and compatible with zero , implying a cancellation between the u and d - quark contribution in a isoscalar target . 99 slides : + ` http://indico.cern.ch/contributiondisplay.py?contribid=277&sessionid=22&confid=24657 `
p. abbon _ et al .
_ [ compass collaboration ] , nucl .
instrum .
a * 577 * , 455 ( 2007 ) .
j. c. collins , nucl .
b * 396 * , 161 ( 1993 ) .
e. albrecht _ et al .
instr . and meth . a 553 , 215 ( 2005 ) , and references therein .
v. y. alexakhin _ et al . _ [ compass collaboration ] , phys .
lett . * 94 * , 202002 ( 2005 ) .
e. s. ageev _ et al .
_ [ compass collaboration ] , nucl .
b * 765 * , 31 ( 2007 ) .
m. diefenthaler [ hermes collaboration ] , proceedings of dis 2007 , page 579 ; arxiv:0706.2242 [ hep - ex ] . | the study of transverse spin effects is part of the scientific program of compass , a fixed target experiment at the cern sps .
compass investigates the transversity pdfs in semi - inclusive dis , using a longitudinally polarized muon beam of 160 gev / c impinging on a transversely polarized target . from 2002 to 2004
, data have been collected using a @xmath0lid target transversely polarized .
transversity has been measured using different quark polarimeters : the azimuthal distribution of single hadrons , the azimuthal dependence of the plane containing hadron pairs , and the measurement of the transverse polarization of baryons ( @xmath1 hyperons ) .
all the asymmetries have been found to be small , and compatible with zero , a result which has been interpreted as a cancellation between the u and d - quark contributions . in 2007
compass has taken data using a nh@xmath2 polarized proton target which will give complementary information on transverse spin effects . |
the structure of the @xmath0be nucleus has been considered as a prototype of cluster structure in nuclei .
it has been described theoretically by many different cluster models and several experiments have been performed in order to understand its structure and decay mechanisms .
much effort has been devoted to the @xmath7 state due to its astrophysical importance in the formation of @xmath0be , @xmath8c and heavier elements in stellar nucleosynthesis@xcite . @xmath0be and @xmath0b are mirror nuclei that decay into @xmath9 and @xmath10 respectively .
therefore their structures are expected to be very similar .
we present here the results of two - dimensional energy correlations after the three - body decay of several @xmath0be resonances .
moreover the same kind of results for the @xmath7-resonance of @xmath0b are shown in order to show the similarities .
we employ here the complex - scaled hyperspherical adiabatic expansion method to solve the faddeev equations which describe our three - body system@xcite .
the angular part of the hamiltonian is first solved keeping fixed the value of the hyperradius @xmath11 .
its eigenvalues serve as effective potentials while the eigenfunctions , @xmath12 are used as a basis to expand the total wave - function @xmath13 the @xmath11-dependent expansion coefficients , @xmath14 , are the hyperradial wave functions obtained from the coupled set of hyperradial equations@xcite .
we consider a two - body interaction able to reproduce the low - energy scattering properties of the two different pairs of particles in our three - body system .
ali - bodmer @xmath15 potential@xcite and coulomb potential between @xmath1-particles are considered .
the @xmath1-nucleon interaction is taken from cobis et al@xcite .
the @xmath16be- and @xmath0b - resonances are of three - body character at large - distances , where they decay into two @xmath1-particles and one neutron or proton , but this is not necessarily the case at short - distances.@xcite we use the three - body model at all distances because the decay properties only require the proper description of the emerging three particles .
a three - body short - range potential of the form @xmath17 is included to adjust the corresponding small - distance part of the effective potential .
the correct resonance energies , which are all - decisive for decay details as evident in the probability for tunneling through a barrier , are then correctly reproduced .
the eigenvalues of the angular hamiltonian for fixed hyperradius , serve as adiabatic potentials@xcite .
each of them correspond to a specific combination of quantum numbers , i.e. , partial - wave momenta between two particles in each jacobi system .
usually few angular eigenvalues are needed for achieving convergence . at small distances
the potentials have wells that support the bound states and resonances .
the energy distribution is the probability for finding a given particle at a given energy . it can be measured experimentally and is the only information that allows us to study the decay path . from the theoretical point of view , this information is contained in the large - distance part of the wave function , which must therefore be computed accurately .
the zeldovic regularized fourier transform of the wave function gives the energy distributions@xcite .
the resonance wave functions change sometimes substantially from small to large distances .
this dynamic evolution allows a better understanding of the decays details@xcite .
the decay mechanisms depend on the resonance properties and can be either sequential or direct or a mixture . in our case ,
all of the resonances can decay sequentially via @xmath18be(@xmath19 ) , i.e. it is allowed by angular momentum conservation .
one of the adiabatic components is related to the @xmath18be+@xmath2 structure and approaches the complex energy of the @xmath18be(@xmath19 ) resonance .
this component will give the sequential contribution@xcite . fig .
[ fig9be ] corresponds to the two - dimensional energy correlations ( or dalitz plots ) of the three fragments of @xmath0be after the decay .
the two possibilities , @xmath20 and @xmath15 , are shown and the @xmath21 of each state is labeled in the figure .
the energies are given in units of their maximum values for each case , i.e. @xmath22 for the @xmath1-particles and @xmath23 for the neutrons .
in all the cases we have removed the sequential decay via @xmath18be(@xmath19 ) .
it is important to remark that the results are directly comparable to measured distributions .
the first thing that we observe in fig .
[ fig9be ] is that all the plots are symmetric between @xmath1 s energies .
this symmetry is necessary since the @xmath1 s are identical particles .
the graphs corresponding to @xmath24 and @xmath25 , are very similar to each other .
first , we do not observe zeroes in the dalitz plots.@xcite this means that angular momentum conservation does not forbid any energy combination .
second , we see that the density increases towards higher @xmath1 energies . this is due to the coulomb repulsion , which enlarges the charged particles energies while reduces that of the neutron .
the other two cases show more structure and have zero probability points or curves . in our description the only difference between @xmath0be and @xmath0b
is the existing coulomb interaction between the @xmath1-particle and the proton in @xmath0b .
the structure of the resonances in both nuclei are expected to be very similar .
[ fig9b ] shows the comparison between the dalitz plots corresponding to the @xmath7-resonance of @xmath0be and @xmath0b .
the patterns are , in fact , almost indistinguishable .
we have described @xmath0be and @xmath0b resonances as three clusters by means of the complex - scaled hyperspherical adiabatic expansion method , including short - range and coulomb interactions .
the two - dimensional energy correlations of the decaying fragments are shown for the low - lying resonances of @xmath0be .
we compare one of them ( @xmath7 ) to the corresponding one in @xmath0b and find , as we expected , an almost identical distribution .
our distributions are open to experimental tests .
this work was partly supported by funds provided by dgi of mec ( spain ) under the contracts fis2008 - 01301 and fpa2007 - 62216 .
r.a.r . acknowledges support from ministerio de ciencia e innovacin ( spain ) under the `` juan de la cierva '' program . | the complex - rotated hyperspherical adiabatic method is used to study the decay of low - lying @xmath0be and @xmath0b resonances into @xmath1 , @xmath1 and @xmath2 or @xmath3 .
we consider six low - lying resonances of @xmath0be ( @xmath4 , @xmath5 and @xmath6 ) and one resonance of @xmath0b ( @xmath7 ) to compare with .
the properties of the resonances at large distances are decisive for the momentum distributions of the three decaying fragments .
systematic detailed energy correlations of dalitz plots are presented . |
the next major worldwide project in high energy particle physics will be a linear electron positron collider at the tev scale .
this machine will complement and extend the scientific results of the lhc currently operated at cern .
the most advanced proposal for this linear collider is the international linear collider ( ilc ) . here
, electron and positrons will be collided at centre - of - mass energies between 0.2 and 1 tev .
the detectors , which will be installed around the interaction point are required to achieve a jet energy resolution of @xmath0 , thus a factor two better than the energy resolution achieved for a typical detector at lep .
the reconstruction of the final state of the @xmath1 will be based on so - called particle flow algorithms ( pfa ) .
the goal is to reconstruct every single particle of the final state , which in turn demands a perfect association of the signals in the tracking systems with those in the calorimeters . as a consequence
this requires a perfect tracking of the particle trajectories even in the calorimeters . to meet these requirements
, the detectors have to cover the whole solid angle and have to feature an unprecedented high granularity .
the application of pfa requires a perfect reconstruction of the particle trajectories in the calorimeter . for this
the calorimeters have to be placed inside the coil of the super - conductive solenoid of the detectors .
this puts tight constraints on the available space for the installation of the detectors . the design of the detector components and notably that of the calorimeters has to take the following guidelines into account * optimisation of the number of calorimeter cells .
* choice of the absorber material and the infra - structural components such as cooling , power supplies , readout cables and the very front end electronics . for the electromagnetic calorimeter , which surrounds the tracking chambers these criteria has lead to the choice of tungsten with a radiation length of @xmath2=3.5 mm , a molire of @xmath3=9 mm and an interaction length of @xmath4=96 mm .
in the years 2005 to 2011 the calice collaboration has performed beam campaigns at desy , cern and fnal with a _ physics prototype _ of the electromagnetic calorimeter in order to demonstrate the principle of highly granular calorimeters and to confront the concept of particle flow with real data .
the first results of the data analysis have been published in three articles @xcite .
the next prototype , also called _ technological prototype _ , has been conceived during the year 2008 @xcite and enters now its construction phase .
it addresses , more than the physics prototype , the engineering challenges which come along with the realisation of highly granular calorimeters .
the key parameters of the new prototype are * size of an individual cell of 5.5x5.5@xmath5 . * a depth of 24@xmath2 .
* thickness of an individual layer of 3.4 mm and 4.4 mm according to the position within the calorimeter .
the figures [ fig : alveolar ] and [ fig : slab ] show the mechanical housing , constructed during 2010 , and a cross section through two calorimeter layers which form a slab .
the mechanical housing is realised by a tungsten carbon composite , which provides at the same time the absorption medium and the mechanical rigidity of the detector .
the silicon wafers are composed of high resistive silicon .
an example of a silicon wafer matrix is shown in fig .
[ fig : waf ] . while in principle
the manufacturing of these wafers is a well known technique , the objective is to produce these wafers at small price in order to reduce the cost since a surface as large as 3000@xmath6 will be needed for an ilc detector . on this contacts with major industrial partners are about to be forged . among others
the establishment of a co - operation with industry is supported within the _ aida _ project @xcite within the fp7 programme on research infrastructure in europe .
r0.5 the measurements with the physics prototype revealed a cross talk between the guard - ring , which surrounds the silicon wafers and neighboured silicon pads has been observed resulting in so - called _ square events _ of which the frequency increased with increasing energy of primary electrons @xcite .
the current r&d efforts target at the optimisation of the layout of the guard - ring . if feasible the guard - ring will even be suppressed .
this would also help to meet the requirement of minimal dead space between the wafers .
additionally , a clear view will have to be established on the tolerance in the quality of the wafers .
the wafers will be glued to the pcbs using a robotic system .
dedicated studies to realise this gluing are currently carried out .
the studies comprise the pressure which can be exerted onto the wafer , the type of glue and naturally the reliability of the glue . for the latter ,
the experience with the physics prototype have revealed no obvious problems .
it should however be noted that the constraints on the size of the glue dots were relatively relaxed .
the cells of the physics prototype for example were four times larger than those of the technological prototype .
the final calorimeter will comprise around @xmath7 channels in total . in order to reduce the non - equipped space in the detector ,
the front end electronics has to be integrated into the calorimeter layers , see fig .
[ fig : slab ] , which constitutes a major challenge for the construction of the calorimeter .
the space available for the readout circuits ( asics ) and the interface boards between the asics and the silicon wafers is about one millimetre .
this challenge is addressed in the current r&d programme .
the figure [ fig : pcb ] shows a picture of four asics bonded onto a thin pcb .
right to that the figure [ fig : asic ] shows a zoom into the area occupied by the asics .
it should be noted that one of the major challenges to be solved in the near future is the planarity of the pcb .
this issue is currently addressed in collaboration with industrial partners as well as by revising the entire assembly process of the detector .
note in passing that for protection purposes the asics will be encapsulated .
such an encapsulation is standard in industry under the keyword globtop and has been successfully applied to circuits implemented on a prototype pcb .
an alternative based on the epoxy araldite 2020 has been developed by the research groups of calice .
each of the asics of the new prototype , dubbed skiroc , will readout 64 calorimeter cells and realises the measuring of the analogue signal , the digitisation and the zero suppression such that only a limited number of channels are finally sent to the data acquisition . as an example of the excellent performance of the skiroc circuit ,
the threshold of 50% trigger efficiency as a function of the injected charge , in units of mips , is shown in fig .
[ fig : skitrig ] .
in addition it is indicated that the signal over noise ratio is about eight where ten is envisaged for the circuits of the linear collider . for more details the reader
is referred to @xcite .
a recent study reveals no undesired effect caused by embedding the circuits in the calorimeter volume as foreseen by the calorimeter design @xcite .
the data acquisition is based on custom made components in order to control cost and r&d effort for a daq at a future linear collider .
only the part closest to the detector , called _ detector interface card _
, is customised to the actual detector type .
an operational set - up containing prototype versions of the front end electronics and the daq system is shown in figure [ fig : vfe ] . and [ fig : asic ] the asics are still packaged _ _ ] due to the limited space available for cooling devices the heat dissipation of the asics has to be minimised .
in addition to the cooling , the heat dissipation will be reduced by a novel technique called _
power pulsing_. here , the front end electronics will only be fully enabled during the millisecond of a bunch train of particles .
such a bunch train contains about 3000 particle bunches separated by around 300ns . during
the around 200ms between these bunch trains , the bias currents of the front end electronics will be shut down .
clearly , this novel technique will require detailed studies in order to assure that the signal quality of each channel remains constant after each powering cycle .
power pulsing has been tested successfully in beam tests for asics with the a similar layout as skiroc @xcite .
a calorimeter layer will have a length of about 1.5 m and will be composed of several units which carry silicon wafers as well as the front end electronics .
great care is taken in the development of the technique to interconnect the individual units . apart from the reliability of the signal transfer along the slab
the interconnection must not exert mechanical or thermal stress to the wafers which are very close to the interconnection pads .
good progress has been made and a viable solution will be presented in the course of 2012 . the sensible ensemble has to be inserted into the alveolar structure which houses the calorimeter layers .
the integration cradles are under development and a first integration test with a demonstrator has been successfully conducted in october 2009 . for this demonstrator
a cooling system has been developed which in an upgraded form is already available for the large scale prototype .
the next prototype of a highly granular siw ecal is taking shape .
these proceedings report progress for a number of components .
it should be stressed that the r&d currently focusses on finding the technical solutions needed for an operation in a linear collider detector . starting from 2012 the first units of the detector
will be exposed to high energy particle beams .
these tests are a vital support for the ongoing r&d .
the schedule of the completion of the prototype with 30 short and one or two long asus is driven by the funding situation in the coming years which in turn depends also on the political situation of a future linear collider .
the solutions found for the prototype of the siw ecal can however also be applied in other fields of science .
this work is funded by the european union in the 6th framework program `` structuring the european research area '' , the 7th framework program `` capacities '' as well as by the cnrs / in2p3 budget for `` quarks and leptons '' , the french anr program and the japanese jsps program
99 calice collaboration ( j.repond ) , jinst * 3 * ( 2008 ) p08001 , arxiv:0805.4833v2 [ physics.ins-det ] .
calice collaboration ( j.repond ) , nucl .
instr . and meth .
a * 608 * ( 2009 ) 372 , arxiv:0811.2354v1 [ physics.ins-det ] .
calice collaboration , ( c. adloff ) , jinst * 5 * ( 2010 ) p05007 , arxiv:1003.2662v1 [ physics.ins-det ] .
m.anduze , eudet - report-2009 - 01 . for
an overview on the aida project please consult + https://espace.cern.ch/aida/default.aspx r. cornat , _ design of a siw electromagnetic calorimeter for the international linear collider and study of cross talk in silicon sensors _ , + proceedings of the nss - mic 2008 conference , dresden , germany .
s. callier , jinst * 6 * c12040 ( 2011 )
. calice collaboration , ( c. adloff ) , nucl .
instrum . and meth .
a * 654 * ( 2011 ) 97 , arxiv:1102.3454 [ physics.ins-det ] .
l. caponetto , arxiv:1111.5630v1 [ physics.ins-det ] . | the calice collaboration is preparing large scale prototypes for highly granular calorimeters for detectors to be operated at a future linear electron positron collider .
currently a prototype of a siw electromagnetic calorimeter is assembled which in terms of dimensions and layout meets already most of the requirements given by the linear collider physics program and hence the detector design . |
for almost 30 years the schechter luminosity function ( lf ; @xcite ) has been the standard tool for quantifying the galaxy population has three key parameters , @xmath1 the characteristic luminosity where the exponential cutoff cuts in , @xmath2 , the normalisation at this characteristic luminosity , and @xmath3 , the faint - end slope parameter .
a value of @xmath4 implies equal numbers of galaxies in magnitude intervals , a more negative ( or steep ) value implies numerous dwarf systems . ] .
the lf is loosely based on the press - schechter formalisation for the primordial halo distribution @xcite .
moreover the lf consistently provides a good formal fit to the observed luminosity distribution ( ld ; see for example @xcite ) .
this consistency , between the ld and lf , appears to hold regardless of environment ( @xcite ; @xcite ) .
the only departure from a pure schechter function appears to be in the central cores of rich clusters , where the galaxy ld is often seen to show a marked upturn at the giant - dwarf boundary ( @xmath5 mag ) .
perhaps the most well known example is the central ld of the coma cluster ( e.g. , @xcite ; @xcite ; @xcite and references therein ) .
the most plausible explanation is that the core contains an overdensity of giant and dwarf ellipticals bolstering both the bright and faint - end of the core cluster lf .
for example the more extensive coma survey by @xcite recovers a flat and invariant ld / lf ( @xmath6 ) to @xmath7 mag .
the phenomena of an upturn in the ld , has also been seen in virgo ( @xcite ; @xcite ) , a963 ( @xcite ) , a868 ( @xcite ) , a2554 ( @xcite ) and a2218 ( @xcite ) for example . however , apart from these `` active '' core environments , the overall lds from the field , to the local group , to the local sphere , and near & far rich clusters all consistently follow a smooth lf within the luminosity ranges probed
[ fig1 ] shows an ( incomplete ) summary of @xmath8 or @xmath9-band field and cluster lfs colour corrected to the johnson b filter .
the main point to take from fig .
[ fig1 ] is that the global and cluster lfs each show a broad but overlapping range of distributions .
clearly one can not reasonably argue for any significant variation between the global and overall cluster environment on the basis of these data .
studies based within the same survey data , for example the two - degree field galaxy redshift survey study by @xcite , generally find fairly subtle changes with environment .
hence it seems that the variations seen in fig .
[ fig1 ] indicates an unspecified systematic error in the various studies .
the most lauded of these is the unsavoury topic of surface brightness selection effects ( @xcite , @xcite ) .
the concern is that the galaxy population at each luminosity interval occupies a range in surface brightness ( or size ) .
surveys with shallow detection isophotes may miss both light from a galaxy s halo , as well as entire galaxies ( see for example @xcite and @xcite ) .
@xcite explored this possibility in detail and demonstrated that indeed surface brightness selection effects can play havoc with the recovered schechter function parameters and reproduce exactly the kind of variation seen in both the global and cluster lfs of fig .
[ fig1 ] .
more recently a number of papers have identified a clear luminosity - surface brightness ( or size)+36.57 $ ] where @xmath10 is the effective surface brightness , @xmath11 the absolute magnitude , and @xmath12 the semi - major axis half - light radius in kpc , hence the luminosity - surface brightness relation can be readily transformed to a luminosity - size distribution and we use the acronym lsp to indicate either . ] relation for field galaxies based on diverse datasets including : the hubble deep field @xcite ; the two - degree field galaxy redshift survey @xcite ; the sloan digital sky survey ( @xcite ; @xcite ) ; and a very local inclination and dust corrected sample of late - type disks @xcite .
these studies consistently show that low surface brightness is synonymous with low luminosity with a few notable exceptions as typified by malin 1 @xcite and the faint second disk surrounding ngc5084 . to fully resolve the potential impact of surface brightness selection effects one must consider the joint luminosity - surface brightness distribution .
this has been advocated in the past , not so much to compensate for selection bias , but to preserve the size ( or surface brightness ) information which may be of interest in its own right ( see @xcite and @xcite for instance ) .
this latter point is illustrated in plate 1 , where i show an example lf for a nearby volume limited sample and images of the actual galaxies contributing to the lf .
clearly much information is lost when one replaces these images with three simple numbers .
it is for these reasons the need to accommodate selection bias and the desire to explore additional parameter space coupled with the abundance of data that now moves us beyond the simple ld / lf to start exploring multivariate distributions . here
i introduce two such distributions , the luminosity - surface brightness plane ( for the reasons stated above ) and the colour - luminosity plane which is also of topical interest ( e.g. , @xcite ; @xcite and references therein ) .
to construct multivariate distributions requires an extensive wide area survey ( @xmath13 ) , to reasonable depth ( @xmath14 mag arcsec@xmath15 ) , with reasonable resolution ( fwhm @xmath16 ) , wavelength coverage ( e.g. , some of @xmath17etc ) and spectroscopic redshifts / distances .
the most notable catalogues for this purpose are : the two - degree field galaxy redshift survey ( 2dfgrs ; @xmath18 , 1800 sq deg , 250k z s ; @xcite ) , the sloan digital sky survey ( sdss ; @xmath19 , 10000 sq deg , 1 m z s ; @xcite ) , the millennium galaxy catalogue ( mgc ; @xmath0+sdss , 37 sq deg , 10k z s ; @xcite ) and the two micron all sky survey ( 2mass ; @xmath20 , all sky , sdss+2df+6df+mgc z s , @xcite ) . the mgc , although the smallest in area , is also the deepest ( @xmath21 mag arcsec@xmath15 ) , highest resolution ( fwhm @xmath22 ) and most complete survey ( see @xcite ; @xcite ; @xcite )
. it also overlaps with the other three surveys and hence provides a `` best of all worlds '' hybrid dataset for example 50 per cent of the @xmath23 mgc redshifts derive from the 2dfgrs or sdss , extensive optical colour coverage from sdss , and partial near - ir coverage from 2mass . the mgcjliske / mgc/ ( additional catalogues including redshift , morphological and structural parameters are available on request from [email protected] ) . ]
contains 10,061 resolved galaxies with @xmath24 mag with 95 per cent complete redshift coverage .
all galaxies have been analysed with a variety of software packages including sextractor ( @xcite ) , gim2d ( @xcite ) and eyeball classified to @xmath25 mag .
the luminosity - surface brightness plane ( lsp ) is of particular interest because it enables one to compensate for both luminosity ( malmquist bias ) _ and _ surface brightness selection effects ( aka `` disney bias '' ) . in @xcite
the lsp is derived for the mgc , which provides the most robust current estimate .
the mgc lsp analysis used the joint luminosity - surface brightness step - wise maximum likelihood method of @xcite and incorporate into this tracking of 5 selection boundaries relevant to each individual galaxy ( _ i.e. , _ maximum & minimum observable size & flux and minimum observable central surface brightness for detection , see @xcite ) .
an additional feature is the derivation of individual k - corrections using the combined mgc and sdss - dr1 colours ( @xmath26 ) .
[ fig2 ] shows the data as a series of gaussian fits across the lsp at progressive intervals of absolute magnitude .
the thicker weighted lines shows the surface brightness distribution for the most luminous galaxies and the fainter lines for the dwarf regime .
two facts leap out .
firstly that the distributions are bounded ( the gaussian fits have good @xmath27 s ) broadening towards lower luminosity .
secondly the peak of the distribution moves towards lower surface brightness for lower luminosity systems .
in other words low luminosity systems apparently show greater surface brightness diversity than giant systems . however this can also be interpreted in terms of the kormendy relation for spheriods ( @xcite ) and freeman s law for disks ( @xcite ) .
these two classic studies unveiled distinct relations for the structural properties of spheroid and disk components .
the kormendy study found that the more luminous the spheroid the lower its central surface brightness .
conversely freeman s study found that all disks , regardless of luminosity , have a constant central surface brightness of @xmath28 mag arcsec@xmath15 . the mgc result shown on fig .
[ fig2 ] are for the combined bulge+disk systems . around @xmath1
the effective surface brightness for spheriods and disks is fairly close a long - time nagging coincidence . however moving towards lower luminosity
the trends for spheriods and disks diverge leading to the broadening of the global surface brightness distribution . to investigate further
hence requires separating out these two structural components via 2d bulge - disk decomposition .
here we use gim2d ( @xcite ) and fig .
[ sball ] shows the data of fig . 2 subdivided by structural component .
the dotted line shows the original freeman distribution which remains relevant today , albeit with a far broader dispersion than originally reported ( see @xcite ) .
it would seem that galaxies consist of two principle components ( presumably formed via two mechanisms : merging and accretion / collapse ? ) and to unravel these two phases in detail must require robust bulge - disk decompositions of extensive samples over a variety of epochs .
the next most obvious key global parameter , after luminosity and surface brightness ( size ) is colour , and in particular the rest @xmath29 which straddles the 4000 - break and hence a crude indicator of the current star - formation rate . @xcite and @xcite have recently studied this plane extensively with sdss data and demonstrate clear bimodality of the colour distribution .
[ fig5 ] shows this trend for the 10k galaxies of the mgc ( using sdss colours ) .
[ fig5 ] also shows this trend for the bulge and disks separately .
to obtain the bulge colour we use the sdss psf magnitudes and to obtain the disk colour and we remove the bulge colour component from the global colour to reveal the disk colour - r_t -2.5 \log[10^{(1-b / t)}]$ ] where the filter subscript refers to total petrosian magnitude ( t ) , disk ( d ) or bulge psf magnitude ( b ) . ] .
we now see that the bimodal distribution can readily be explained in terms of predominantly red bulges and blue disks .
this component segregation implies distinct stellar populations with distinct evolutionary paths .
bulges must contain old stellar population and disks intermediate or young populations .
again this follows conventional wisdom but highlights yet further the important of bulge - disk decomposition and the need to study the component properties of galaxies rather than the global properties .
the two distributions outlined above , both suggest that the well known bulge and disk components of galaxies follow distinct trends in both the surface brightness ( size ) distribution and colour distribution .
this is of course not particularly new , however what is exciting is our ability to quantify these distributions and trends in detail for large statistical samples , and to extend this kind of structural analysis to higher redshift .
in particular the data resolution and signal - to - noise of the ground - based data discussed above is comparable to that available with the hubble space telescope ( @xcite , @xcite , @xcite , @xcite ) and the upcoming james webb space telescope .
there is nothing , other than hard diligent work , to prevent us from quantifying the evolution of these distributions across the entire path length of the universe .
however three further issues are worth raising : * 1 ) which wavelength is optimal for structural studies of galaxies ? *
\2 ) how might we push back the boundaries into the dwarf regime ?
\3 ) can we connect structural measurements to the properties of the dark matter halo ?
traditionally , almost all nearby galaxy catalogues have been based on flux - limited observations through an optical @xmath0 or blue bandpass filter ( @xmath30 nm ) , for example : the rc3 ; the two - degree field galaxy redshift survey ; the millennium galaxy catalogue etc .
this has been driven by technological / commercial necessity with flux detectors typically optimised to the spectral response of the human eye ( @xmath31 nm ) .
however the most physically meaningful bandpass in which to observe a galaxy is in the near - ir ( rest-@xmath32 band or @xmath33 m ) .
this is mostly because the stellar population that dominates the total stellar mass and therefore best traces a galaxy s gravitational potential is the long - lived low - mass population which emits in the near - ir ( see for example @xcite ) .
this is most clearly demonstrated by the obviously smoother appearance of a galaxy in the near - ir than in progressively bluer wavelengths ( see upper panel of fig . [ fig6 ] showing a montage of images for m51 in a variety of filters the near - ir images ( right most ) are significantly smoother .
the flux and shape of a galaxy in the near - ir is most dependent on the older relaxed stellar population and therefore a better tracer of the underlying potential .
conversely the flux and shape in the optical is linked to the young stars and therefore dependent on the current and possibly transient star - formation rate .
both optical and near - ir data are important if one wishes to understand galaxy formation and evolution .
the near - ir however appears to be the optimal filter for the investigation of the structural properties .
the other great advantage is of course the minimisation of the impact of dust obscuration .
this is illustrated in the main panel of fig .
[ fig6 ] which shows the location of the @xmath0 and @xmath32 band filters superimposed on the night sky spectrum ( dotted line ) , a galaxy s continuum before and after star - burst ( solid lines ) , and the dust attenuation curve ( long dashed line ) .
the impact of star - formation and dust is clearly less in the near - ir . the upcoming near - ir facilities and in particular the ukirt / wfcam and vista will have the capabilities to provide exactly the kind of wide , deep , high resolution data required for the comprehensive structural analysis of nearby galaxies .
the space density of dwarf galaxies remains elusive . figs .
[ fig1 ] & [ fig2 ] shows that the mgc can only sample with credibility to @xmath34 mag at which point both limiting statistics and the high and low surface brightness selection limits bite ( see @xcite ) .
[ fig7 ] illustrates this by showing the mgc galaxies on an absolute magnitude versus redshift plot .
the data are of course bounded by the @xmath35 mag limit which highlights the rapidly diminishing volume observed for low luminosity systems .
one way to overcome this is to simply conduct ever deeper redshift surveys ( as indicated on the figure ) .
however this has a diminishing return as the number of galaxies one must observe to find one low luminosity system becomes unreasonable .
one possible way forward is to use photometric redshifts to pre - select low - z candidates and then follow - up only these systems .
however the accuracy of photometric redshifts at low z is poor ( although improved by near - ir colours , see @xcite for example ) . to overcome the surface brightness selection limits ( both high and low ) the source data
must be improved to probe to very high resolution ( @xmath36 ) and very deep isophotes ( @xmath37 mag arcsec@xmath15 ) over wide areas ( @xmath38 sq deg ) .
no such survey exists but facilities such as subaru / suprime and magellan / imacs just about have the capability to achieve such a survey .
the alternative method is to observe the very local galaxy population ( i.e. , the local sphere of influence , defined as that within 10 mpc ) and obtain direct distance measurements rather than redshifts .
the lsp may have the potential to connect key observables ( luminosity and size ) to the fundamental underlying physical properties of bulge and disk systems ( mass and angular momentum ) .
in various studies of the formation of disk systems , ( _ e.g. , _ @xcite , @xcite , @xcite ) the dimensionless spin parameter ( @xmath39 , @xcite ) is directly related to the scale - length of the disk .
the spin parameter reflects how close the halo is to a rotationally supported system and is a key parameter monitored by the numerical simulations ( see @xcite ; @xcite ; @xcite ; maller , dekel & sommerville 2002 for example ) . the pivotal idea ( here echoing the toy model of @xcite )
starts with the premise that the baryons are coupled to the dark matter halo , because of this the luminosity ( generated by the baryons in the form of stars ) can be related to the systemic mass and the rotation of the stars / gas can be related to the systemic angular momentum . given this premise , which is intimated by the tully - fisher relation
, one can analytically relate @xmath40 to luminosity and surface brightness ( or size ) : @xmath41 ( from @xcite ) , where @xmath42 is the effective surface brightness , @xmath43 is the intrinsic luminosity in some filter and @xmath44 is the dependence of luminosity on the mass - to - light ratio ( equal to 0.69 in @xmath0 or 1.00 in @xmath32 @xcite ) .
numerical simulations consistently find that the distribution of the spin parameter is a log normal distribution which is globally preserved through hierarchical merging ( see for example @xcite ) this yields : @xmath45 or @xmath46 .
hence the gradient of any luminosity - surface brightness relation bears upon the relation between mass and light and the dispersion upon the breadth of the spin distribution .
[ fig8 ] shows the @xmath0-band lsp for a variety of samples as indicated ( lg , @xcite ; hdf , @xcite ; mgc , @xcite ; mw gcs , harris et al ( priv .
comm ) ; local sphere of influence , @xcite ; lsbgs , @xcite ) .
the solid lines show the approximate expectation as argued above and show remarkable agreement with the data - in detail the observed size distribution is marginally narrower than simulations predict ( see @xcite ) .
it is also worth noting that systems which form via merging ( i.e. , bulges ) and via accretion ( i.e. , disks ) are also predicted to show distinct @xmath40 distributions ( see for example @xcite ; @xcite ) . at the moment
far more data and detailed simulations are required however this connection is clearly promising and could ultimately result in a galaxy equivalent to the hertzsprung - russell diagram , allowing a meeting ground between numerical simulations and survey observations .
the galaxy luminosity function is not the only fruit , and with the many legacy datasets becoming available the time is now ripe to move beyond the lf and explore multivariate distributions . here
i ve presented two : the luminosity surface brightness ( size ) plane ( lsp ) ; and the colour luminosity plane ( clp ) .
both planes show that disks and bulges form distinct but overlapping distributions presumably indicating secular evolution of these components , i.e. , two mechanisms and two timescales .
this finding argues for the community to move away from global measurements and start to measure the properties of these distinct components independently .
we argue that this is best done in the near - ir and should be a key focus of upcoming ir facilities such as vista ( low - z ) & jwst ( high - z ) .
perhaps most important of all the lsp appears to provide a direct meeting ground to the numerical simulations .
this last point is by far the most important as its from the cross - talk between simulations and observations that real insight into the processes of galaxy evolution and formation will come .
i am indebted to my colleagues on the mgc and to the hard work of those that have slaved on the superb 2dfgrs , sdss and 2mass facilities and databases .
these data are changing the way we view galaxies in a profound way .
finally i d like to thank the organisers for a very enjoyable meeting and support from pparc as a visiting fellow to the university of bristol where this article was revised . | with the advent of large scale surveys ( _ i.e. , _ legacy surveys ) it is now possible to start looking beyond the galaxy luminosity function ( lf ) to more detailed statistical representations of the galaxy population , _
i.e. , _ multivariate distributions . in this review
i first summarise the current state - of - play of the @xmath0-band global and cluster lfs and then briefly present two promising bivariate distributions : the luminosity - surface brightness plane ( lsp ) ; and the colour - luminosity plane ( clp ) . in both planes galaxy bulges and galaxy disks
form marginally overlapping but distinct distributions , indicating two key formation / evolutionary processes ( presumably merger and accretion ) .
forward progress in this subject now requires the routine application of reliable bulge - disk decomposition codes to allow independent investigation of these two key components . [ firstpage ] galaxies : fundamental parameters galaxies : luminosity function , mass function galaxies : statistics surveys |
an interesting phenomenon that corresponds to spontaneous emissions ( as if a black body radiation ) from a black hole ( bh ) is the hr .
it is a semi - classical outcome of the quantum field theory @xcite .
hr dramatically changed our way of looking to the bhs ; they are not absolutely black and cold objects , rather they emit energy with a characteristic temperature : ht .
event horizon , where is an irreversible point ( in classical manner ) for any object including photons is the test - bed of the gedanken experiment for the hr .
the studies concerning this phenomenon have been carrying on by using different methods . in particular , the quantum tunneling @xcite of particles with different spins from the various bhs have gained momentum in the recent years ( the reader may be referred to @xcite and references cited therein ) . recently , it has been shown that hr of the bosons with spin-0 ( scalar particles ) and spin-1 ( vector particles ) from the tlwh @xcite , which is a bridge or tunnel between different regions of the spacetime is possible by using the pothg @xcite .
wormhole has been extensively studied in different areas @xcite .
however , ht of the tlwh appears to be negative because of the phantom energy ( exotic matter : the sum of the pressure and energy density is negative ) that supports the broadness of the wormhole throat @xcite .
in addition , it is a well - known fact that the virtual particle - antiparticle pairs are created near the horizon . in a bh spacetime the real particles with positive energy and temperature
are emitted towards spatial infinity @xcite .
however , in the pothg which is analog to the white hole geometry , the antiparticles come out from the horizon @xcite .
in other words , our analysis predicts that the energy spectrum of the antiparticles leads to a negative temperature for the tlwh . for the subject of the white hole radiation , the reader may refer to @xcite .
as it is shown by caldwell et al @xcite , the dark matter ( dm ) @xcite could have a phantom energy . in this regard
, the phantom energy can keep apart every bound object until the cosmos eventuates in the big - rip @xcite . on the other hand ,
dm does not emit , reflect or absorb light , making it not just dark but entirely transparent .
but if the dm particles strolling around a bh or a wormhole can produce gamma - rays would give a possibility to study the radiation of this mysterious matter @xcite .
dm has many candidates , and gravitino ( spin-@xmath0 ) is one of them @xcite .
gamma - ray decay of the gravitino dm has been very recently studied in @xcite .
so , hr of the gravitinos from the bhs and/or wormholes could make an impact on the production of the dm .
behaviors of the gravitino s wave function are governed by the rses @xcite .
so , our main motivation in this paper is to investigate the hr of the gravitino tunneling from the tlwh geometry . using the rses and hj method
, we aim to regain the standard ht of the tlwh .
the structure of this paper is as follows . in sec .
ii , we introduce the 3 + 1 dimensional tlwh @xcite and analyzes the rses for the gravitino particles in the pothg of the tlwh @xcite .
we show that the rses are separable when a suitable hj nsatz is employed .
then the radial equation can be reduced to a coefficient matrix equation that makes us possible to compute the probabilities of the emission / absorption of the gravitinos .
finally , we calculate the tunneling rate of the radiated gravitinos , and retrieve the standard ht of the tlwh .
we summarize and discuss our results in sec .
for the wave equation of the gravitino ( spin-3/2 ) particles , we start with the massless ( the mass has no remarkable effect in the computation of the quantum tunneling @xcite ) rses @xcite : @xmath1 where @xmath2 is a vector - valued spinor and the @xmath3 matrices satisfy @xmath4 .
the first equation is the dirac equation applied to every vector index of @xmath5 , while the second is a set of additional constraints to ensure that no ghost state propagates ; that is , to ensure that @xmath5 represents only spin-@xmath0 fermions , with no spin-@xmath6 mixed states @xcite .
the covariant derivative obeys @xmath7 where @xmath8 , \nonumber\\ \{\gamma^{\alpha},\gamma^{\beta}\ } & = 2g^{\alpha\beta}\times i. \label{4}\ ] ] the metric of tlwh in the generalized retarded eddington - finkelstein coordinates ( refcs ) , which is the pothg , is given by @xcite @xmath9 where @xmath10 @xmath11 .
misner - sharp energy is represented by @xmath12 which becomes @xmath13 on the trapping horizon ( @xmath14 ) @xcite .
marginal surfaces having @xmath15 are the past marginal surfaces in the refcs @xcite . for solving the rses ,
we use the following dirac @xmath16-matrices:@xmath17{cc}-i & -\sigma^{3}\\ -\sigma^{3 } & i \end{array } \right ) , \,\,\ \gamma^{r}=\sqrt{f}\left ( \begin{array } [ c]{cc}0 & \sigma^{3}\\ \sigma^{3 } & 0 \end{array } \right ) , \nonumber\\ \gamma^{\theta } & = \frac{1}{r}\left ( \begin{array } [ c]{cc}0 & \sigma^{1}\\ \sigma^{1 } & 0 \end{array } \right ) , \,\,\,\ \gamma^{\phi}=\frac{1}{r\text{sin}\theta}\left ( \begin{array } [ c]{cc}0 & \sigma^{2}\\ \sigma^{2 } & 0 \end{array } \right ) , \label{6}\ ] ] where the pauli matrices are given by @xmath18{cc}0 & 1\\ 1 & 0 \end{array } \right ) , \sigma^{2}=\left ( \begin{array } [ c]{cc}0 & -i\\ i & 0 \end{array } \right ) , \sigma^{3}=\left ( \begin{array } [ c]{cc}1 & 0\\ 0 & -1 \end{array } \right ) . \label{7}\ ] ] gravitino wave function ( @xmath19 ) has two spin states [ spin up ( i.e. positive @xmath20direction ) and spin down ( i.e. negative @xmath20direction ) ] : @xmath21 @xmath22 where @xmath23 denotes the gravitino action which is going to be expanded in powers of @xmath24 , and @xmath25 are the arbitrary constants .
here we shall only consider the spin up case , since the spin down case is fully analogous with it .
the action for the spin - up states can be chosen as follows @xmath26 therefore , the corresponding rses become @xmath27 + \sqrt{f}(-c_{0}\partial_{r}s_{\uparrow0 } ) & = 0,\label{11}\\ \frac{1}{r\text{sin}\theta}(-ic_{0}\partial_{\phi}s_{\uparrow0})+\frac{1}{r}(-c_{0}\partial_{\theta}s_{\uparrow0 } ) & = 0,\label{12}\\ \frac{1}{\sqrt{f}}\left [ \left ( -a_{0}-ic_{0}\right ) ( \partial _ { u}s_{\uparrow0})\right ] + \sqrt{f}(-a_{0}\partial_{r}s_{\uparrow0 } ) & = 0,\label{13}\\ \frac{1}{r\text{sin}\theta}(-ia_{0}\partial_{\phi}s_{\uparrow0})+\frac{1}{r}(-a_{0}\partial_{\theta}s_{\uparrow0 } ) & = 0 , \label{14}\ ] ] with the constraints equations : @xmath28 equations ( 15 - 18 ) are not important here . because these equations give an independent wave solution , so that they have no effect on the action @xcite .
afterwards , the separation of variables method is applied to the action @xmath29 : @xmath30 where @xmath31 and @xmath32 are energy and angular constants , respectively .
however , @xmath33 is an arbitrary complex constant .
thus , eqs .
( 11 - 14 ) reduce to @xmath34 equations ( 21 ) and ( 23 ) are about the solutions of @xmath32 , and they do not have contribution to the tunneling rate .
for this reason , we simply ignore them .
namely , the master equations for the tunneling rate are eqs .
( 20 ) and ( 22 ) . to analyze them , we first consider the case of @xmath35 @xcite . using eqs .
( 20 ) and ( 22 ) , we now have a solution for @xmath36 as @xmath37 the integrand has a simple pole at @xmath38 .
choosing the contour as a half loop going around this pole from left to right and integrating , one obtains @xmath39 where @xmath40 is the surface gravity at the horizon . on the other hand , if one sets @xmath41 , this time eqs .
( 20 ) and ( 22 ) admit the following solution for @xmath42 @xmath43 hence , we can derive the ingoing / outgoing imaginary action solutions as @xmath44 @xmath45 we can now set @xmath46 for the action of absorbed ( ingoing ) gravitinos .
we can tune their probability : @xmath47 to % 100 by letting @xmath48 consequently , the probability of the emitted ( outgoing ) gravitinos becomes @xmath49 recalling the definition of the tunneling rate : @xmath50 which is also equivalent to the boltzmann factor : @xmath51 , we read the ht of the tlwh as follows @xmath52 which is a negative temperature .
this result implies that if the trapping horizon remains in the past outer region , the wormhole throat would have a negative temperature @xcite
. the phantom energy , which is the special case of the exotic matter could be the reason of that negative temperature @xcite .
on the other hand , when @xmath53 in the action ( 19 ) , it is possible to obtain the positive temperature : @xmath54 .
although , the latter remark contradicts with the previous results @xcite ( and whence , one may easily get rid of the case of @xmath53 ) , however hong and kim @xcite showed that possibility of negative / positive temperature of the wormhole depends on the exotic matter distribution .
in this work , we have studied the hr of the gravitino particles from the tlwh in 3 + 1 dimensions .
tlwh has been introduced in the pothg .
we have analyzed the rses in the background of the tlwh with the help of hj method .
the probabilities of the emitted / absorbed gravitino particles from the trapped horizon of the tlwh have been computed . after comparing
the obtained tunneling rate with the boltzmann factor , we have recovered the standard ht of the tlwh , which is a negative temperature .
this is the special condition in which the high - energy states are more occupied than lower - energy states @xcite .
another possibility of the negative temperature may originate from the exotic matter distribution of the wormhole @xcite . meanwhile , very recently it has been claimed by @xcite that hr does not occur in the pothg .
in fact , the latter debatable remark is based on the study of @xcite stating that cosmic matter flux may turn the hr off . on the other hand ,
hayward show that switching off the radiation causes the wormhole to collapse to a schwarzschild bh @xcite . in summary
, gravitinos can tunnel through wormhole [ simply this can be thought as a wormhole with one entrance ( bh ) and one exit ( white hole ) ] .
in such a case , gravitinos tunnel from the bh with positive temperature , while they tunnel through the white hole with negative temperature .
thus our calculations are based on the exit of the wormhole , just as the white hole case . besides
, we have shown that positive temperature can be obtained by tuning the @xmath33-constant in the action ( 19 ) to zero .
however , the latter result is a debatable issue , and it demands much deeper analysis
. this will be our next venture in this line of study . | recent research shows that hawking radiation ( hr ) is also possible around the trapping horizon of a wormhole . in this article
, we show that the hr of gravitino ( spin-@xmath0 ) particles from the traversable lorentzian wormholes ( tlwh ) reveals a negative hawking temperature ( ht ) .
we first introduce the tlwh in the past outer trapping horizon geometry ( pothg ) .
next , we derive the rarita - schwinger equations ( rses ) for that geometry
. then , using both the hamilton - jacobi ( hj ) anstz and the wkb approximation in the quantum tunneling method , we obtain the probabilities of the emission / absorption modes .
finally , we derive the tunneling rate of the emitted gravitino particles , and succeed to read the ht of the tlwh . |
accretion in disks is generally thought to occur by the action of turbulence , for which the magnetorotational instability ( mri , * ? ? ? * ) is the most likely culprit . however , protoplanetary disks are cold ; the ionization level required to couple the gas to the ambient field is not always met @xcite , leading to zones that are `` dead '' to the mri @xcite .
so , the quest for hydrodynamical sources of turbulence continues , if only to provide accretion through this dead zone .
one such possible sources of hydrodynamical turbulence is the subcritical baroclinic instability ( sbi , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , a process shown to sustain large - scale vortices in the presence of a radial entropy gradient and thermal relaxation or diffusion . two - dimensional linear stability analysis and numerical simulations do not find instability if only seeded with linear noise @xcite , though it was shown that finite amplitude perturbations would trigger it , concluding that the instability is nonlinear in nature @xcite .
characterization of the instability through nonlinear numerical simulations shows that maximum amplification is found for thermal times in the range of 110 times the dynamical timescale @xcite .
although no criterion for a critical reynolds number was derived , @xcite show that as resolution is increased , ever smaller perturbations are necessary , as expected if the process is physical .
compressible simulations @xcite show that the spiral density waves excited by the vortices @xcite transport angular momentum at the level of @xmath1 , where @xmath2 is the shakura - sunyaev parameter @xcite .
if this process indeed occur in disks , it would provide not only accretion but also a fast route for planet formation in the dead zone , since vortices speed up the process enormously , by concentrating particles in their centers @xcite .
the appeal of the sbi , however , is severely hindered by its nonlinear nature . without the guide of analytics ,
nonlinear processes are difficult to characterize , and the accuracy of the numerics have to be well - established beyond reasonable doubts .
recently , klahr & hubbard ( 2014 , hereafter kh14 ) have claimed that , when considering the same equations that lead to sbi in 2d , linear growth exists if vertical wavelengths are considered .
the unstable mode is a slowly growing epicyclic oscillation , which led the authors to name the process `` convective overstability '' .
growth is powered by buoyancy and thermal relaxation in the same regime as the sbi , of cooling time of the order of the dynamical time .
we analyze this claim of linearity in more detail in this paper .
independent verification is desirable since unorthodox assumptions were made in the linear analysis of kh14 . in particular
, the authors assumed that the timescale for pressure equilibration is fast , and thus set the pressure perturbation to zero in the linear analysis .
because of this strong assumption , skepticism about the validity of the work naturally remains until a rigorous derivation of the dispersion relation is provided , unambiguously demonstrating that the eigenvector of the growing root has no appreciable pressure term . in this work ,
we provide such derivation .
another point raised by kh14 is the connection between this overstability and the sbi , if any . a priori
, the two processes have little to do with each other .
however , as the regimes of cooling time for both are similar , if the convective overstability exists , it may generate the finite amplitude perturbations that trigger the sbi . in this scenario
, the ( nonlinear ) sbi would simply be the saturated state of the ( linear ) convective overstability . since the difficulty on finding a source of finite amplitude perturbation in dead zones in the required range of cooling times had made the sbi look less attractive as a relevant disk process , a linear process that can spawn the sbi from arbitrarily low - level noise would be particularly interesting .
conversely , there is the possibility , of course , that the saturated state of the convective overstability may still be of too low amplitude to trigger the sbi .
we investigate these possibilities in the present study .
this paper is structured as follows . in sect 2
we perform a linear analysis calculating the full compressible dispersion relation . in sect 3
we take the anelastic limit to derive the instability criterion , finding the roots , the most unstable mode , and associated eigenvector . in sect 4
we perform numerical simulations in the shearing box to characterize the linear growth phase and nonlinear saturation in 2d and 3d .
we conclude in sect 5 .
.symbols used in this work [ cols="<,<,<",options="header " , ]
let us consider the compressible euler equations with thermal relaxation .
@xmath3 where @xmath4 is the density , @xmath5 is the velocity , @xmath6 is the pressure , @xmath7 is the adiabatic index , @xmath8 is the temperature , @xmath9 is a reference temperature , and @xmath10 is the thermal time .
we consider the cylindrical approximation , meaning that we omit the vertical component of the stellar gravity , as well as vertical stratification . in this approximation ,
the gravity is @xmath11 , with @xmath12 the keplerian angular frequency and @xmath13 the cylindrical radial coordinate .
a list of the mathematical symbols used in this work , together with their definitions , is provided in table [ table : symbols ] .
we linearize eqs .
( [ eq : continuity])([eq : energy ] ) into base state and perturbation ( the latter denoted by primes ) , as @xmath14 , @xmath15 , @xmath16 , @xmath17 , and @xmath18 . assuming the cylindrical approximation ( @xmath19 = 0 for the base state ) , eqs .
( [ eq : continuity])([eq : energy ] ) become @xmath20 in the above equations , @xmath21 , @xmath22 , @xmath23 , and the thermal relaxation term was linearized @xmath24 as per the equation of state , @xmath25 .
next we use the short - wave approximation , @xmath26 , and expand the perturbations in fourier modes , @xmath27 .
( [ eq : cont2])([eq : energy2 ] ) then become @xmath28 where @xmath29 , and we have also substituted @xmath30 .
the system is @xmath31 , where @xmath32^t$ ] , and the coefficient matrix is @xmath33.\label{eq : matrix}\ ] ] we have substituted @xmath34 so both @xmath35 and @xmath36 have dimension of frequency . in particular , @xmath37 , and @xmath38 , so @xmath39 is the square of the brunt - visl frequency .
the full dispersion relation @xmath40 is @xmath41 \nonumber\\ & + & \bar{\omega } \
c^2k_z^2(\kappa^2+n^2 ) + \frac{ic^2\kappa^2k_z^2}{\gamma\tau } = 0 , \label{eq : dispersion}\end{aligned}\ ] ] where @xmath42 is the square of the epicyclic frequency .
we consider now some limits of eq .
( [ eq : dispersion ] ) .
in the anelastic limit , @xmath43 , eq . ( [ eq : dispersion ] ) reduces to @xmath44 & & \nonumber\\ - \bar{\omega } \
k_z^2(\kappa^2+n^2 ) - \frac{i\kappa^2k_z^2}{\gamma\tau } & = & 0 , \label{eq : anelastic-1}\end{aligned}\ ] ] where @xmath45 and @xmath46 .
these terms are proportional to @xmath47 , so they are small and can be dropped .
the dispersion relation is thus @xmath48 where we have also substituted @xmath49 and @xmath50 . for adiabatic flow , @xmath51 , eq .
( [ eq : anelastic ] ) reduces to @xmath52 for @xmath53 ( in - plane incompressible motion ) , we retrieve @xmath54 , the solberg - hoiland criterion . for pure in - plane incompressible motions ( @xmath53 ) , eq .
( [ eq : anelastic ] ) reduces to @xmath55 which is the same as derived by kh14 ( their eq .
18 ) , using other assumptions .
substituting @xmath57 , growing solutions correspond to real positive @xmath58 .
the dispersion relation , real and imaginary , that need to vanish independently , are : @xmath59 @xmath60 - \varsigma(s^2 -\mu^2\kappa^2 ) = 0 .
\label{eq : gamma}\ ] ] substituting eq .
( [ eq : s ] ) into eq .
( [ eq : gamma ] ) , we get @xmath61 as we expect the growth to be small ( to be checked a posteriori ) , we take the limit @xmath62 , leading to @xmath63.\ ] ] this function has no extrema for finite @xmath64 . for @xmath65
, however , there is a maximum at @xmath66 , that is , maximum growth occurs for @xmath67 for which the growth rate is @xmath68 , i.e. @xmath69 recalling the definition of the brunt - visl frequency @xmath70 we can write it in terms of the power - law indices of the density and temperature gradients , @xmath71 , @xmath72 , and @xmath73 , resulting in @xmath74 where @xmath75 is the aspect ratio and @xmath76 is the scale height .
so , for keplerian disks , @xmath77 and @xmath78 .
it results from this that @xmath79 is of order @xmath80 , while the associated growth rate is of order @xmath81 , validating the assumption that @xmath82 .
notice that for @xmath83 , that is , @xmath84 , the dispersion relation ( eq .
[ eq : full - dispersion ] ) becomes @xmath85 for which the roots are @xmath86 , and @xmath87 , that is , no growth , and damped perturbations . for channel modes ( @xmath53 ) in keplerian disks ( @xmath88 ) ,
we find @xmath89 we plot in fig .
[ fig : stability ] the unstable range as a function of the density and temperature power law indices .
requires ( for @xmath90 ) that @xmath91 . for a power - law surface density @xmath92
, we have @xmath93 .
the requirement is then @xmath94 , which , for @xmath95 means @xmath96 . for @xmath97
the surface density has to be flat or increasing with distance in order to lead to instability , which is not reasonable .
for @xmath98 the onset of instability corresponds to @xmath99 ( also for @xmath95 ) , which is consistent with the range of @xmath100 $ ] ( with median -0.9 ) found in the observations of @xcite . ] to understand the most unstable mode , we check the eigenvector @xmath101 corresponding to this root , for which the eigenvalue is @xmath102 and the system is @xmath103 , where @xmath104.\label{eq : matrix}\ ] ] the 4th line is @xmath105 , which is only satisfied for the trivial solution @xmath106=0 . the reduced system becomes @xmath107 the solution is @xmath108 u_r^\prime.\end{aligned}\ ] ] since @xmath109 , the pressure perturbation is @xmath110 and , because @xmath111 and @xmath112 are or order @xmath113 , @xmath114 and @xmath115 are vanishingly small . that the pressure variation does not play a major role in the instability
justifies ( now a posteriori ) the @xmath116 approximation of kh14 .
the eigenvector is simply @xmath117^t , \label{eq : eigenvector}\ ] ] i.e. , an overstable epicycle .
we now turn to numerical simulations to check the evolution of the instability .
we use the shearing box model of @xcite , that includes the linearized pressure gradient .
we do so in order to benefit from shear - periodic boundaries , in contrast to the simulations in the appendix of kh14 , that are affected by radial boundaries .
the reader is referred to @xcite for the equations of motion , properties and caveats of the approximation . in particular , the density gradient is zero , and we drop the @xmath118-dependent term in the pdv work to keep shear - periodicity ( see appendix a of @xcite .
we solve the evolution equations with the pencil code @xcite which integrates the pdes with sixth order spatial derivatives , and a third order runge - kutta time integrator .
sixth - order hyper - dissipation terms are added to the evolution equations , to provide extra dissipation near the grid scale , explained in @xcite .
they are needed because the high - order scheme of the pencil code has little overall numerical dissipation @xcite .
we run a suite of 2d axisymmetric models ( @xmath118 and @xmath119 ) to understand the linear evolution and saturation properties of the instability .
the sound speed is @xmath120=0.1 , and the adiabatic index is @xmath121 .
the cooling time is @xmath122 .
our units are @xmath123 .
we initialize the simulations with the eigenvector corresponding to the epicycle oscillation , @xmath124 . because for @xmath125 the growth rate does not depend on @xmath126 , we arbitrarily choose @xmath127 for the channel mode .
the initial condition therefore is @xmath128 the fiducial model has resolution @xmath129 , box size @xmath130 , temperature gradient @xmath131 , and initial amplitude @xmath132 .
we vary these quantities to check convergence at saturation .
the evolution of the 2d axisymmetric box seeded with the channel mode is shown in the panels of fig .
[ fig:2d - saturation ] .
the linear phase matches the analytical prediction ( dashed black line ) for all models ran .
figure [ fig:2d - saturation]a shows the dependency on resolution .
convergence is achieved for 64 grid points per scale height .
there is also convergence for initial amplitude of perturbation , as seen in fig .
[ fig:2d - saturation]c .
the linear phase is identical in the three cases examined ( @xmath133 , and @xmath134 ) . in this figure we set @xmath135 as the time that saturation is achieved , to better compare the nonlinear evolution . in fig .
[ fig:2d - saturation]d we check how the instability depends on the pressure gradient .
again , the linear phase is reproduced for the different values of the brunt - visl frequency , and the amplitudes at saturation are similar , within a factor 23 .
difference in seen when we test the dependency on box size ( fig .
[ fig:2d - saturation]b ) .
the amplitude seemed to saturate at @xmath136 ( red line ) , since the model with @xmath137 ( green line ) shows a similar amplitude .
however , the model with @xmath138 ( cyan line ) shows a bifurcation at @xmath0150 orbits .
models with larger radial range ( @xmath139 and @xmath140 , purple and magenta lines , respectively ) show no convergence , even as the velocity dispersion increasingly approaches the sound speed .
interesting features are seen in this simulation , that help understand the behavior of the system .
we plot in fig .
[ fig : power ] the time evolution of the power in the first 5 large scale modes , in both @xmath118 ( upper middle panel ) and @xmath119 ( upper left panel ) .
the upper left panel shows the rms of the radial velocity .
four special / representative instants are labeled , and the @xmath141 field for these respective instants are shown in the lower panels .
the first instant , @xmath35 , corresponds to the first `` saturation '' seen at 50 orbits .
the power spectrum shows that the clean initial channel mode ( @xmath142=4 , @xmath143=0 ) persisted until this time , after which it saturates , exciting @xmath144 modes and other @xmath126 modes .
instant @xmath36 , at 67 orbits , corresponds to the local minimum in rms velocity .
the power spectrum shows that this happens when the @xmath142=1 mode becomes dominant .
subsequently , this mode keeps growing , at the same rate as the initial @xmath142=4 mode .
this is because the growth rate is independent of @xmath126 for @xmath125 , which at that time has similar power as the higher @xmath145 modes . from time
@xmath146=90 ( instant @xmath147 ) to 160 orbits the system settles into a steady state , with a dominant @xmath142=1 mode , and mixed @xmath143=0 and @xmath143=1 .
another bifurcation happens when the @xmath148 mode overtakes the @xmath142=1 mode .
simultaneously , it prompts @xmath143=1 to dominate over @xmath125 .
the final state ( labeled @xmath149 ) is thus vertically symmetric , with a box - wide radial wavelength .
this explains why we do not find convergence while increasing box vertical range from @xmath150=@xmath151 to @xmath152 to @xmath153 . in these boxes , because we kept the seed mode at @xmath154 , we initialized the instability with the @xmath142=1 , 2 , and 4 mode , respectively . in the last two simulations ,
the @xmath142=1 mode was growing , with less power , but eventually catching up as the seed mode saturates .
convergence with radial box size is never achieved in the 2d runs because the @xmath155 mode comes to dominate , no matter how wide we make the box .
the simulations with radial box size @xmath156 and @xmath157 show the same pattern , albeit with no intermediate phase of dominance of a @xmath142=1 mode .
next we turn to the 3d evolution of the instability .
we set a box of size @xmath158 , with resolution @xmath159 in @xmath118 , @xmath160 , and @xmath119 , respectively .
the cells thus have aspect ratio @xmath161 ( we have checked in @xcite that unit aspect ratio in @xmath118 and @xmath160 gave the same results for the twodimensional sbi )
. with the azimuthal direction present , vertical vorticity ( in - plane circulation ) can evolve unabridged .
we show in fig .
[ fig : figure3d ] ( left panel ) the evolution of the rms velocity ( red line ) and enstrophy ( black line ) .
when the initial @xmath126 mode saturates ( at 50 orbits , as in the 2d meridional models of fig 3 ) , a sharp rise in enstrophy occurs .
the situation is now very similar to the sbi , with high - amplitude perturbations ( @xmath162 ) , thermal relaxation , and an entropy gradient .
the nonlinear saturation state of this buoyant overstability should thus proceed very similarly to the evolution of the sbi . indeed , as the lower panels of fig .
[ fig : figure3d ] show , the saturated state develops into a large scale vortex .
the amount of angular momentum transport ( fig .
[ fig : figure3d ] , upper right ) is at the @xmath163 level , again , the typical level of the sbi .
it seems conclusive that the saturated state of the buoyant overstability is the sbi .
we conclude that indeed there is a linear overstability in the region of the parameter space of negative @xmath164 , finite cooling time @xmath10 , and non - zero @xmath126 perturbation .
the approximation @xmath165 done by kh14 is justified as @xmath165 ( and @xmath166 ) in the eigenvector of the most unstable modes is vanishingly small in comparison to the velocity amplitude ( eq . [ eq : eigenvector ] ) .
modeling the system numerically , we reproduce the linear growth rate in all cases . in the twodimensional meridional simulations , we find convergence in the saturated state with resolution , but not with box size , since a large - scale @xmath155 radial mode dominates the box .
however , in three dimensions this mode does not show up , as it gets sheared away .
we also show that the sbi is indeed the saturated state of the overstability .
saturation leads to a fast burst of enstrophy in the box , and a large - scale vortex develops in the course of the next @xmath0100 orbits after the convective overstability has built the finite amplitude perturbations .
the amount of angular momentum transport achieved is of the order of @xmath1 , as in compressible sbi models .
it remains to be shown if these processes ( both sbi and convective overstability ) operate in global models , i.e. , how they respond to boundary conditions and curvature terms .
the relation between this overstability and the goldreich - schubert - fricke instability @xcite should also be the subject of future work .
this work was performed in part at the jet propulsion laboratory , under contract with the california institute of technology funded by the national aeronautics and space administration ( nasa ) through the sagan fellowship program executed by the nasa exoplanet science institute .
this paper started from a https://www.facebook.com/groups/149092965243653/302839956535619/?notif_t=group_comment#[discussion ] between the author , alexander hubbard , matthew kunz , hubert klahr , henrik latter , geoffroy lesur , min - kai lin , george mamatsashvili , and orkan umurhan .
it further profited from input from anders johansen , mordecai - mark mac low , colin mcnally , neal turner , and andrew youdin .
natexlab#1#1 andrews , s. m. , wilner , d. j. , hughes , a. m. , qi , c. , & dullemond , c. p. 2009 , apj , 700 , 1502 balbus , s. a. & hawley , j. f. 1991 , apj , 376 , 214 barge , p. & sommeria , j. 1995 , a&a , 295l , 1 blaes , o. m. & balbus , s. a. 1994 , apj , 421 , 163 brandenburg , a. & dobler , w. 2002 , comp .
comm . , 147 , 471 fricke k. , 1968 , z. astrophys . , 68 , 317 gammie , c. f. , 1996 , apj , 457 , 355 goldreich p. , schubert g. , 1967 , apj , 150 , 571 heinemann , t. & papaloizou , j. c. b. 2012 , mnras , 419 , 1085 heinemann , t. & papaloizou , j. c. b. 2009 , mnras , 397 , 52 heinemann , t. & papaloizou , j. c. b. 2009 , mnras , 397 , 64 johnson , b.m .
, & gammie , c.f .
2005 , apj , 635 , 149 klahr , h. , & hubbard , a. 2014 , apj , accepted .
klahr , h. & bodenheimer , p. 2003 , apj , 582 , 869 klahr , h. 2004 , apj , 606 , 1070 klahr , h. , bodenheimer p. , 2006 ,
apj , 639 , 432 lesur , g. & papaloizou , j.c.b .
2010 , a&a , 513 , 60 lyra , w. & klahr , h. 2011 , a&a , 527a , 138 lyra , w. , johansen , a. , klahr , h. , & piskunov , n. 2008a , a&a , 479 , 883 lyra , w. , johansen , a. , klahr , h. , & piskunov , n. 2008b , a&a , 491 , l41 lyra , w. , johansen , a. , zsom , a. , klahr , h. , piskunov , n. 2009b , a&a , 497 , 869 mcnally , m. , lyra , w. , & passy , j .- c . 2012 ,
apjs , 201 , 18 .
nelson , r. p. , gressel , o. , umurhan , o. m. 2013 , mnras 435 , 2610 petersen , m. r. , julien , k. , stewart , g. r. 2007a , apj , 658 , 1236 petersen , m. r. , stewart , g. r. , julien , k. 2007b , apj , 658 , 1252 raettig , n. , lyra , w. , & klahr , h. 2013 , apj , 765 , 115 shakura , n. i. & sunyaev , r. a. 1973 , a&a , 24 , 337 turner , n.j . & drake , j.f .
2009 , apj , 703 , 2152 | recently , klahr & hubbard ( 2014 ) claimed that a hydrodynamical linear overstability exists in protoplanetary disks , powered by buoyancy in the presence of thermal relaxation .
we analyse this claim , confirming it through rigorous compressible linear analysis .
we model the system numerically , reproducing the linear growth rate for all cases studied .
we also study the saturated properties of the overstability in the shearing box , finding that the saturated state produces finite amplitude fluctuations strong enough to trigger the subcritical baroclinic instability .
saturation leads to a fast burst of enstrophy in the box , and a large - scale vortex develops in the course of the next @xmath0100 orbits .
the amount of angular momentum transport achieved is of the order of @xmath1 , as in compressible sbi models .
for the first time , a self - sustained 3d vortex is produced from linear amplitude perturbation of a quiescent base state . |
wz sagittae is a dwarf nova with extremely long time intervals between outbursts . during the vvlast outburst superhumps
were discovered , which led to the classification of wz sge as a su uma type dwarf nova ( patterson et al .
a review of investigations of wz sge in the past as well as a redetermination of the system parameters is given by smak ( 1993 ) .
the orbital period of 81 minutes is close to the period minimum of cataclysmic variables and by standard interpretation puts wz sge into a late state of evolution of dwarf nova systems . for these close binaries
the distance of secondary star and white dwarf primary decreases continously due to loss of angular momentum from the system by gravitational radiation until a period minimum is reached and it grows again for later times .
osaki ( 1996 , review ) showed that , in the framework of the thermal - tidal instability model , the decreasing number of normal outbursts of su uma systems between superoutbursts can be explained as an evolutionary sequence with decreasing mass overflow rate from the secondary star .
wz sge is an extreme case in only showing superoutbursts . due to its outburst behaviour
it was estimated that the viscosity parameter in the quiescent state should be as low as @xmath2 or even @xmath3 ( smak 1993 ) . recently
two different suggestions were made , by lasota et al .
( 1995 ) , hameury et al .
( 1997 ) and by warner et al .
( 1996 ) how the behaviour of wz sge could be modeled using a viscosity parameter in the cool state @xmath0 = 0.01 , which is only slightly smaller than what is usually chosen to describe ordinary dwarf nova outburst cycles .
we show with our computations that the viscosity has to be low , if we take the large amount of mass in the disk , 1 - 2 @xmath4 g , as a constraint .
this mass was deduced in two ways , from the uv outburst luminosity and from the hot spot brightness during quiescence ( smak 1993 ) .
also the optical outburst lightcurve shows the high mass flow rate in the beginning .
sin addition smak s value for the distance to wz sge is in the lower range of values given in the literature .
thus a high amount of mass in the disk seems unavoidable .
the evaporation of the inner disk is an important feature for the evolution of the disk , because it prevents the early onset of the instability .
we include the physics of the hole formation and find that the hole can either persist during the long quiescence for systems containing a white dwarf of large mass or the hole can be closed for small white dwarf mass .
but evaporation still continues at the inner disk edge close to the white dwarf .
important for the evolution is also to follow in detail how the disk grows due to the redistribution of angular momentum until , for some cases , the 3:1 resonance radius is reached and finally the onset of the disk instability occurs .
we found that during the long quiescence the disk can reach a quasi - stationary state .
x - rays originate from an inner disk corona and the thermal boundary layer of the white dwarf . we give a short description of observational data and model parameters in sect .
2 . in sect .
3 we describe the previous modelling of the long - term behaviour of wz sge . the computer code is modified to include the evaporation in the inner disk and the growth of the disk due to redistribution of angular momentum ( sect .
we show the results for the disk evolution and discuss the influence of parameters in sect .
the accretion of matter on to the white dwarf releases x - rays . in sect . 8
we evaluate how much one may expect to observe and compare with rosat and earlier observations . in sect .
9 we discuss modelling with higher values of @xmath0 in previous work .
we point out the consequences from the fact that the 3:1 resonance radius may be reached several years before the onset of the outburst .
we further present a new view of smak s ( 1993 ) interpretation of white dwarf luminosity variation , arising from our results . in the conclusion we sumarize what follows from our computations : ( 1 ) on the low viscosity , ( 2 ) on the importance of evaporation and ( 3 ) on the late evolutionary state of cataclysmic variables .
wz sg is the dwarf nova with the longest observed outburst period .
the three outbursts recorded in the past , november 22 1913 , june 29 1946 and december 1 1978 occurred after remarkably similar time intervals of 32.6 and 32.4 years , respectively ( mattei 1980 ) . wz sge is an su uma type dwarf nova , but unique in the way that no normal outbursts occur in between the superoutbursts .
smak ( 1993 ) argued , that the extremely long outburst cycle can only be explained by a very low viscosity in the disk during quiescence . from this work
we take the following parameters : mass of white dwarf @xmath5=0.45m@xmath6 and mass of secondary star @xmath7=0.058 m@xmath6 , mass overflow rate from the secondary star of about @xmath8= 2@xmath9 g / s .
the amount of matter in the disk @xmath10 accumulated until the outburst should be 1 - 2@xmath11 g. we also consider other parameters to see the dependence of the results on the assumed values .
the white dwarf mass is the most important one for the evolution .
sion et al .
( 1995 ) and cheng et al .
( 1997 ) analysed hst observations of wz sge and from spectra derived values for rotational velocity and gravity . together with the observed 28s period they inferred a white dwarf mass of 0.3m@xmath6 ( or higher , if the star is differentially rotating ) .
we add an independent estimate of about 0.7m@xmath6 from the observed period and assuming criticical rotation .
spruit and rutten ( 1998 ) found in their recent investigation [email protected]@xmath13 from an analysis of the stream impact region .
it is difficult to understand the long outburst recurrence time . in previous investigations
this behaviour was modelled with either a very small value of the viscosity parameter in quiescence @xmath0=0.001 or a moderate value @xmath0=0.01 , which is only slightly smaller than the values usually chosen to describe ordinary dwarf nova outbursts .
osaki ( 1995 ) simulated the outburst lightcurve with the the parameters @xmath0=0.001 , @xmath14=0.03 ( @xmath14 viscosity in the hot state ) , @xmath8=1.2 @xmath9g / s , @xmath15 and @xmath7=0.1m@xmath6 it was assumed that mass is accumulated in a torus at the circularisation radius ( `` lubow - shu '' radius ) .
this computation was based on the simplifying model of osaki ( 1989 ) .
the observed outburst amplitude and duration are then well described . in the thermal - tidal instability model for su uma
stars a decrease of the mass transfer rate leads to a longer superoutburst recurrence time and simultanously fewer normal outbursts in between ( for a review see osaki 1996 ) .
according to the standard scenario of cv evolution ( kolb 1993 ) mass transfer rates are expected to decrease with age . for the modelling of wz sge at the very end of this `` activity sequence ''
not only @xmath8 has to be low , also @xmath0 must be extremely low . to avoid the assumption of extremely small values @xmath0 two different attempts were made to understand the behaviour of wz sge , both using a moderate value in combination with a hole assumed in the inner disk .
lasota et al .
( 1995 ) and hameury et al .
( 1997 ) pointed out that a hole in the inner disk due to evaporation into a coronal layer above the cool disk ( meyer & meyer - hofmeister 1994 ) or due to the presence of a magnetosphere of the white dwarf ( livio & pringle 1992 ) prevents the onset of an outburst in the inner disk .
they observe that the x - ray flux from wz sge soon after the outburst was roughly the same as several years later and argue that this points to a quasi - stationary disk structure during quiescence .
their analysis shows that with an inner hole reaching to 2.5 - 5 @xmath16 cm , @xmath8 = @xmath9g / s and @xmath17=0.01 no outburst occurs , the disk is stable .
the mass contained in the disk is then much less than the value @xmath18 derived by smak . according to the estimate in hameury et al .
( 1997 ) it is below 3 @xmath19g / s . to solve the two problems how to get the instability triggered in the stable disk and how to supply a high amount of mass the authors suggest that a fluctuation in the mass overflow rate from the companion star might lead to an increase of surface density in the disk beyond the critical value and trigger the instability .
then a very high mass overflow is expected due to irradiation of the secondary star . in their scenario
this could happen any time after the disk is stationary .
warner and al .
( 1996 ) try to model wz sge with the same value @xmath0= 0.01 , a hole of 3 @xmath16 cm , @xmath8 = 6@xmath20g / s , @xmath21 and an outer disk radius of 1.1 @xmath22 cm .
the assumed hole makes it possible to accumulate mass over a long time until after 36 years the outburst occurs .
but due to the low accretion rate only 7 @xmath19 g / s matter were transferred from the compagnion to the disk .
the computed outburst lasts only 6 days .
close to the white dwarf a hot corona above the cool disk leads to evaporation ( meyer & meyer - hofmeister 1994 , f.k .
liu et al . 1995 ) of the inner disk .
liu et al.(1997 ) have shown in detail how the evaporation affects the disk for a typical dwarf nova .
the evaporation rate scales as @xmath23 and @xmath24 ( f.k .
et al . 1995 ) .
this allows to evaluate the coronal flow rate and the mass exchange between cool disk and hot corona ( evaporation and condensation ) as a function of radius . in a system like
vw hydri a hole is formed after the end of an outburst extending to about 2 @xmath16 cm and it remains that size during the quiescence until the next outburst starts .
the position of the inner edge of the disk is determined by the balance between evaporation of gas from the cool disk into the corona and the supply of new matter from outer disk regions .
the mass flow rate in the disk @xmath25 decreases with decreasing r. ( @xmath25 is related to the surface density ; for a description of disk structure see ludwig et al .
because of this there is always a position where the rates are equal .
the hole is larger the smaller @xmath25 is .
the cool disk in wz sge relaxes to a quasi - stationary inward flow of mass due to the long time available in quiescence and the hole can be closed after a few years depending on the white dwarf mass .
but evaporation is going on all the time and , at the inner edge , puts a significant fraction of this flow into the form of hot coronal gas .
the outer edge is determined by the redistribution of angular momentum .
the gas in the incoming stream from the secondary star has a certain specific angular momentum .
the radius at which the gas in the keplerian orbit around the white dwarf has the same specific angular momentum is called `` lubow - shu '' radius .
generally , in dwarf nova systems , there is no significant accretion on the white dwarf , all mass is kept in the disk during the comparatively short duration of quiescence . with the addition of low specific angular momentum gas from the secondary the outer disk radius shrinks towards the `` lubow - shu '' radius , only to expand again during outburst due to the flow of matter towards the white dwarf .
the situation in wz sge and comparable systems is different . during the long quiescence quasi - stationary accretion on to the white dwarf
is established .
the specific angular momentum of this accreted matter is stored in the disk and as a consequence the disk grows .
this goes on until it reaches the radius where the 3:1 resonance between kepler binary period and the kepler rotation occurs ( which for these low mass ratio @xmath5/@xmath7 systems is inside the tidal truncation radius ) .
thus the tidal instability ( whitehurst 1988 , lubow 1991 ) sets in and the excentric disk produces superhumps in the lightcurve as discussed by lubow ( 1994 ) . to follow the evolution of systems with a long lasting quiescence like wz sge
the radius change has to be taken into account in the computations .
ichikawa & osaki ( 1992 ) had already included in their numerical code radius changes for the evolution of u gem .
in our computations we solve the diffusion equation for mass and angular momentum flow and take the dependence of the critical values of surface density and the viscosity- surface density relation as described in ludwig et al .
in addition we take into account that the positions of both inner and outer edge of the disk change during the evolution .
we use the following formulae .
@xmath26 ( taken positiv inward ) , @xmath27 and @xmath28 are mass flow rate , surface density and kepler frequency in the disk , @xmath29 is the outer disk radius and index `` out '' designates values at @xmath29 , index `` ls '' at the lubow - shu radius @xmath30 .
conservation of mass and angular momentum at @xmath31 give @xmath32 @xmath33 where @xmath34 is the mass flow in the disk at @xmath31 ( taken positive inward ) .
further @xmath35 viscosity integral , @xmath36 the height above the midplane , @xmath37 the effective viscosity , parametrized proportional to the value @xmath38 ( shakura & sunyaev 1973 , compare ludwig et al .
1994 ) .
in the computer code we have an equidistant grid in @xmath39 and use @xmath40 instead of @xmath41 .
conservation of mass and angular momentum in the disk are guaranted with the proper determination of the position of the outer edge @xmath29 and the appropriate value @xmath42 there . @xmath42 and @xmath43 are then determined by eqs .
( 1 ) and ( 2 ) .
@xmath44 @xmath45\ ] ] @xmath46 is related to @xmath30 ( we took the values of @xmath30 from table 2 in the investigation of lubow & shu ( 1975 ) ) . if the 3:1 resonance radius is reached a further growth of the disk is effectively cut off since the tidal instability quickly increases to such a strength , that all surplus angular momentum is transferred back to the binary orbit .
we implement this by replacing eqs .
( 3 ) and ( 4 ) by @xmath47 and @xmath48
we follow the evolution of the cool disk from the early quiescence to the onset of the next outburst .
the orbital period is known .
the geometry of the binary system containing white dwarf and roche lobe filling secondary star is then determined , if we assume white dwarf mass .
further parameters are the mass overflow rate from the secondary star and @xmath0 .
we start our investigation with the parameters for @xmath5 and @xmath7 suggested by smak ( 1993 ) .
we want to model the evolution that means find the onset of an outburst after 32 years .
we take as constraints the amount of matter in the disk at this time , 1 - 2 @xmath4 g , and the mass accretion rate of 2 @xmath9g / s , also derived by smak .
note that this accretion rate will lead to the accumulation of the right amount of matter during the 32 years , but whether an outburst occurs at this time depends on the disk evolution , which is governed by the viscosity .
evaporation is an additional important feature , which influences the location where the instability is triggered . in the following
we show the results , ( 1 ) the time until an outburst arises and ( 2 ) the amount of matter accumulated in the disk until then .
we take @xmath17=0.001 fixed and vary the mass accretion rate .
then we discuss the variation of these quantities with @xmath17 and the influence of evaporation .
finally we show results for a larger white dwarf mass . [
[ ht ] we take an initial distribution of surface density in the disk according to the results for ordinary dwarf novae of ludwig & meyer ( 1998 ) , but scaled to the lower values of @xmath0 used here . to start the computation of the long - time evolution we have to assume the disk size in the beginning . to test the effect of this
we study two cases . in case ( a )
we take a relatively small disk .
such a disk then contains not much mass ( @xmath1 0.1 @xmath11 g ) and angular momentum . as an alternative , case ( b ) , we start with a disk reaching almost to the resonance radius .
then more mass is in the disk ( @xmath1 0.28 @xmath11 g ) .
consequently more matter should be in the disk at the onset of the outburst , so that the difference in mass @xmath49 corresponds to the amount 1 - 2@xmath11 g derived from the outburst luminosity .
the total amount accumulated should be higher in case ( b ) and the matter has to be packed more densily .
this demands a smaller viscosity @xmath50 .
all computations if not stated otherwise are case ( a ) .
the results for the computations case ( b ) ( @xmath51=2@xmath9g / s , @xmath0= 0.001 and 0.0008 ) are marked by the two asterisks in fig .
3 . comparing the evolution in case ( a ) and ( b ) we find , that the onset of the outburst happens for the same total amount of matter in the disk and at nearly the same time .
an exact initial distribution could only be determined by a detailed computation of the outburst .
due to the long quiescence the evolution is only slightly influenced by this uncertainty . in fig .
1 we show the evolution of the disk for our `` standard '' parameters . after 30.3 years
the critical surface density @xmath53 is reached at the distance r=1.86 @xmath16 cm from the white dwarf and the outburst starts . in the disk 1.44 @xmath4 g matter are accumulated . in the left panel
we show the surface density @xmath54 in the disk , in the right panel the corresponding rates @xmath25 every 3 years during the evolution .
the run of @xmath55(r ) during the first 3 years is due to the assumed initial distribution of surface density . in the early evolution
the mass flow rate in the inner disk is small and evaporation creates a hole .
the maximal extent of the hole r=3
@xmath16 cm , is already reached after less than 3 years .
the position of the inner edge depends on the balance between the rate of mass flow inward and evaporation .
the surface density increases continously during quiescence and with it @xmath25 .
the hole is closed after 9 years of disk evolution .
evaporation is still going on , then with the high rate of @xmath1 @xmath9g / s .
about 20% of this flow goes into wind loss , the remaining part accretes via the hot corona on to the white dwarf .
after the first decade we find an almost constant mass flow rate of 7 - 8 @xmath56 g / s over a large part of the disk .
the rate of evaporation , efficient in a ringlike region around the hole , can also be seen from the fig . 1 , left panel .
it is the difference between the nearly constant value and the lower values at distances r less than about 2.5 @xmath16 cm .
the values at the inner disk edge indicate how much mass flows via the cool disk on to the white dwarf .
differences resulting from accretion of hot versus cool gas are discussed in sect .
6 . the outer radius of the disk increases as can be seen from the extent of the lines at different evolutionary times . in fig .
2 we show the changes of the outer and inner disk radius for `` standard '' parameters , but an initially larger disk ( initial distribution case b),where the shrinking of the outer edge is more prominent .
the disk grows until the resonance radius r=1.6 @xmath22 cm ( `` standard '' parameters ) is reached .
the size of the disk is an interesting feature .
before the 3:1 resonance radius is reached roughly half of the newly accreted matter is needed to fill the larger and larger outermost disk areas .
the remaining part is too low to trigger an outburst .
after the resonance radius is reached the disk does not grow anymore , the mass flow in the disk suddenly is nearly doubled , and the surface density increases to the critical value . in fig .
1 , right panel , we see the higher mass flow rate after the resonance radius is reached .
we point out that evaporation prevents instability in the early evolution . without evaporation
the critical surface density would have been reached at r= 1.5 @xmath16 after 9.6 years ( compare fig .
the fact that the hole is closed does not change this situation .
[ [ ht ] [ [ ht ] if we follow the disk evolution for given parameters @xmath5 , @xmath7 and @xmath0 the accumulation of mass depends further on @xmath57 . to get a model appropriate for wz
sge the critical surface density has to be reached after @xmath1 32 years and
the amount of matter accumulated then has to be 1 - 2 @xmath4 g. fig .
3 shows this accumulation process for different rates of @xmath51 .
higher rates make the outburst happen too early .
the opposite happens for low rates . with @xmath51=@xmath9 g / s , which is half of what was suggested for wz sge ( smak 1993 ) the surface density in the disk never reaches the critical value , no outburst occurs .
the disk is stationary .
all matter accreted from the secondary star flows through the disk on to the white dwarf ( except the part lost by the wind ) .
this means , that mass is accreted continously via the corona and the disk on to the white dwarf . in fig .
4 we show how the disk reaches such a stationary state . for comparison
we have also performed computations neglecting evaporation . as can be seen from fig .
3 the outburst would happen much earlier , after 9.6 years instead of 32 years for the same parameters
. the amount of accumulated matter would be lower than what corresponds to the observed increase of 7 magnitudes in the outburst lightcurve and the duration longer than 1 month ( mattei 1980 ) .
such an outburst is prevented by the formation of a hole .
[ [ ht ] in the disk for a mass overflow rate from secondary @xmath57=@xmath9g / s , otherwise `` standard '' parameters.,title="fig:",width=332 ] a smaller viscosity means that the mass flow in the disk is reduced and the matter is packed more densily . for the evolution this results in a higher @xmath18 and a longer recurrence time .
figures 3 and 5 show the results for viscosity values @xmath0 = 5 - 8 @xmath3 instead of 0.001 . [
[ ht ] and mass transfer rate @xmath57.,title="fig:",width=264 ] we have computed one example of evolution for our `` standard '' parameters , but with an evaporation rate three times as high as the regular rate .
this means a larger hole , a later onset of the outburst and more matter accumulated ( see fig .
3 ) . the opposite case is neglecting evaporation .
we indicated already in fig .
3 ( squares ) at which time the instability would be triggered without a hole in the inner disk .
to fullfill the constraints , long recurrence time and large amount of matter in the disk , other parameters have to be taken .
we have performed a series of computations of disk evolution varying @xmath0 and @xmath57 .
the critical surface density is reached the earlier the higher the mass accretion rate .
one needs small values for @xmath0= 3 @xmath3 to 1 @xmath2 in combination with low rates @xmath57= 1.8 to 1.3 @xmath9 g / s to get the right recurrence time and amount of matter in the disk .
figure 5 shows the results in terms of the mass accumulated at the onset of outburst . only for the lowest rates @xmath57 is the resonance radius reached before the outburst starts . in the other cases the disk expands to the 3:1 resonance radius during the outburst
. then superhumps would only appear in the outburst .
figure 6 shows the mass flow rates in the disk for various times during the evolution for a low mass transfer rate , 1.3@xmath9g / s , from the seccondary star .
the disk is quasi - stationary . without evaporation
the matter is accreted via the cool disk on to the white dwarf .
for a rate a bit lower the critical surface density would never be reached .
so this case is similar to a quasi - stationary disk with evaporation ( see sect . 7.2 , evolution for @xmath57=1 @xmath9g / s in fig .
[ [ ht ] in the disk for a mass accretion rate from the secondary @xmath57=1.3 @xmath9g / s otherwise `` standard '' parameters , but without evaporation , title="fig:",width=332 ] for larger white dwarf mass the evaporation process is more important .
the higher rate keeps the hole open longer .
we computed the evolution for @xmath15 , @xmath58 and the mass ratio @xmath59=0.075 .
for the larger white dwarf mass a larger fraction of matter passes continously through the disk and in the inner part through the corona on to the white dwarf .
less matter can be accumulated in the disk .
we found the following results .
( 1 ) @xmath60 , @xmath0=0.001 , @xmath57=2 @xmath9g / s ( rate `` standard parameters '' ) no outburst .
( 2 ) @xmath15 , @xmath0=0.001 , @xmath57=2 @xmath9g / s , outburst after @xmath1100 years , this means this rate corresponds to an almost stationary disk .
if we increase @xmath57 to 3 @xmath9g / s an outburst arises after 50 years . in all three examples @xmath13
@xmath4 g of matter were accumulated . in agreement with our findings from parameter variations described earlier we get for the 1@xmath13 white dwarf about the right recurrence time 29.2 years and amount of matter 1.36 @xmath4 g with @xmath0=0.003 and @xmath57=2.5 @xmath9g / s
. the mass flow rate on to the white dwarf increases to values around 2 @xmath9 during the quiescence .
this and the deeper gravitational potential of the more massive white dwarf implies a higher x - ray flux from the system .
observations of x - rays from cataclysmic variables ( cvs ) with the einstein observatory also included wz sge ( for a review see cordova & mason 1983 , for the determination of temperatures eracleous et al . 1991 ) .
wz sge was observed in april 1979 , april and may 1980 .
it would be interesting to learn about changes shortly after the outburst .
the low number of counts had been at about the same level for the three times of observations , but the derived best fit temperatures variied between 2.9 and 5.8 kev .
the x - ray flux ( 0.1 - 3.25 kev ) , unabsorbed , was 0.46 to 0.62 @xmath61 @xmath62 .
mukai & shiokawa ( 1994 ) analysed the exosat me archival data on dwarf novae . from the brightest observation of wz sge in october 1985 they found the x - ray flux ( 2 - 10 kev ) of 0.39 @xmath61 @xmath63 .
richman ( 1996 ) analysed the x - ray flux from cvs observed with the rosat pspc . from the observations of wz
sge taken in april 1991 she found the best fit temperature 3.4 kev and the flux ( 0.1 - 2.4 kev ) 0.13 @xmath64 .
using a correlation between mass accretion rate and the ratio of x - ray to visual flux ( patterson & raymond 1985 ) the rosat observations indicate a mass accretion rate of @xmath65g / s ( richman 1996 ) .
this distant - independent estimate from the observations supports our theoretical picture , that accretion goes on all the time .
the amount of mass flow in quiescence of wz sge as suggested by our model agrees with this estimate
. accretion on to a white dwarf at a rate of @xmath66g / s releases energy , which , if mainly in x - rays , would predict significantly higher fluxes at earth than observed .
there are however a number of reduction factors , which together bring prediction and observation into approximate agreement .
we first discuss the release of accretion energy in the surroundings of the white dwarf . during the quasi - stationary evolution
the mass flow rate is around @xmath61@xmath13/y in the middle disk ( at distances larger than about @xmath67 cm , compare fig .
1 , right panel ) .
farther in , roughly @xmath68 of the mass flow remains in the optically thick disk , @xmath69 evaporate into the corona and of these , @xmath1 20% are lost in a wind .
if the white dwarf is not close to critically rotating , both flows , optically thick and thin , dissipate their rotational energy , about half of the accretion energy , in a frictional boundary layer close to the surface of the white dwarf ( pringle 1977 , pringle & savonije 1979 ) .
the boundary layer related to the coronal flow is always optically thin . at these low accretion rates also
the boundary layer related to the flow via the disk is probably optically thin ( narayan & popham 1993 , popham & narayan 1995 ) .
the other half of the accretion energy of the coronal flow is dissipated in the corona above the optically thick disk . what is not used up by the escaping wind
is conducted down and radiated away in the thermal boundary layer between corona and disk .
the energy dissipated in both , white dwarf and coronal boundary layer adds up to an efficiency of about 0.5 of the total gravitational energy release . with a hole
the situation is roughly the same .
another reduction is expected from the spectral distribution of x - rays .
liu et al . ( 1995 ) investigated how much x - rays and uv radiation are caused by the accretion of matter evaporating from the disk and flowing via the hot corona on to the white dwarf .
a theoretical spectrum was compared with rosat observations for vw hydri ( meyer et al .
the form of this spectrum is very general due to the scaling properties of such transition layers .
7 shows the luminosity contributions of the various temperature layers in the thermal boundary layer of an accreting white dwarf .
the multi - temperature character clearly can not be satisfacturely be described by a one - temperature spectrum ( see also richman 1996 ) .
7 shows that a fraction of about 0.4 of the radiation is below the sensitivity limit of 0.1 kev of the various x - ray observatories . for rosat also an upper cut - off might matter . [
[ ht ] 1.2 @xmath70k).,title="fig:",width=283 ] a further reduction of the observable flux is due to the fact that half of the optically thin emission layers are hidden by the white dwarf and accretion disk surfaces , respectively .
the reduction is 0.5 for a disk with an extended hole . if the disk reaches near to the surface of the white dwarf a further reduction results , because the white dwarf hemisphere towards the observer is partially covered by the disk and vice versa , depending on inclination .
the total observable fractions are then 0.32 for the white dwarf and 0.4 for the disk ( i=75@xmath71 for wz sge , smak 1993 ) .
these reductions add up to 0.15 and 0.10 with and without an extended hole .
this gives an observable x - ray luminosity of @xmath72 r radius of white dwarf .
for @xmath73 the flux density at earth ( unabsorbed ) is @xmath74 @xmath75 is the mass accretion rate in units of @xmath76/y .
this is about double the observed flux .
now we discuss the consequences of critical rotation on the observable x - ray flux and the white dwarf mass .
the photometric period of 28s discovered in uv light from wz sge by welsh et .
( 1997 and references to earlier work there ) appears to originate at the surface of the white dwarf and indicates a very rapid rotation of this star ( see also cheng et al . , 1997 ) . using the mass - radius relation and the variation of equatorial radius with rotation interpolated from hachisu ( 1986 ) by popham & narayan ( 1995 ) we can determine which white dwarf rotates critically at its equator with this period .
we find @xmath77 where @xmath78 is the equatorial radius of the critically rotating star , 1.6 times that of a non - rotating white dwarf with the same mass .
this mass is a lower limit since with this period any less massive white dwarf with its larger radius and smaller gravity would rotate supercritically . though the mass of the white dwarf in wz sge might be larger and
the star then rotate subcritically it is fairly probable that wz sge has already reached critical rotation during its long evolution
. one can evaluate how much mass transfer to the white dwarf is required to speed it up to critical rotation .
for example , if one assumes that mass is accreted with kepler angular momentum at the equator and that all accreted matter is lost again in consecutive nova explosions but with the mean angular momentum of the white dwarf surface then one obtains critical rotation after mass transfer of @xmath79 where the radius of gyration @xmath80 for a @xmath81 white dwarf , @xmath82 ( h. ritter , private communication ) was used .
if wz sge entered the cataclysmic variable evolution with an orbital period above 3 hours this amont of mass has already been transferred and the white dwarf will now be critically rotating .
the x - rays from accretion on to a critically rotating white dwarf are even more reduced since no frictional boundary layer forms .
material from an optically thick accretion disk then settles on the white dwarf surface without becoming optically thin and producing x - rays .
the only remaining contribution comes from the coronal gas that releases 1/2 of its gravitational energy as it spirals towards the white dwarf surface . again
roughly @xmath83 will be lost in a wind together with about twice its escape energy .
the remaining energy release is then @xmath84 .
the coronal accretion rate @xmath85 is 0.8 of the mass flow rate that has disappeared from the flow of the optically thick disk on approaching the white dwarf , in our example of fig .
1 , right panel , typically @xmath86 . of the energy released in x - rays 1/2
is covered by the disk surface ( or by the white dwarf if part is conducted to the white dwarf surface ) , about 0.2 is further hidden behind the white dwarf for wz sge s inclination , and 0.4 is out of spectral range ( compare sect .
alltogether this gives an observable x - ray luminosity of @xmath87 and a flux density at earth ( unabsorbed ) of @xmath88 for @xmath89 and r the equatorial radius .
+ these estimates show that the accretion rates in quiescence determined by our models lead to x - ray fluxes in good agreement with what has been observed , 1.3 to 6.2 @xmath90 , resolving an apparent puzzling discrepancy .
a slight preference for a critically rotating white dwarf is indicated .
the pulsations observed in various wavelength ranges might suggest accretion on the rotating poles of a magnetic white dwarf .
such a field would have to be weaker than @xmath91 gauss .
otherwise the disk / magnetosphere interface reaches beyond the co - rotation radius and matter entering the magnetosphere would be flung out instead of accreted . however , the only intermediate quality of the optical pulsations , @xmath92 , the fact that the mean period varied by 1.3 @xmath93 between 1995 and 1996 ( patterson et al .
1998 ) , and the difference between the uv - period ( 28.10s , welsh et al . 1997 ) and the x - ray/ optical period ( 27.86s , patterson et al .
1998 ) argues against an accretion spot frozen into the rotating white dwarf .
it might rather suggest accretion spots formed by magnetic fields brought in with the accreted matter and circulating around the nearly critically rotating white dwarf , similar to what has been suggested for dwarf nova oscillations ( meyer 1997 ) .
optically thin accretion at higher latitudes ( x - rays , also reprocessed into optical light ) could have a slightly shorter period than equatorial , optically thick disk accretion ( uv light ) .
the situation remains unclear and needs further study .
the magnetic pressure of a dipolar white dwarf field of surface strength @xmath91 gauss could just interfere with coronal evaporation pressures at @xmath94 and affect the coronal flow , possibly reducing the x - rays even below the observed level .
accretion along the magnetic funnel of accretion formed accretion spots would concentrate the thermal boundary layer to a smaller surface area and result in an increase of the maximum temperature @xmath95 .
the above estimates of x - rays observable will not be much affected by this .
we shortly discuss here the earlier modelling described in sect . 3 in the context of our results .
the simplifying model made already clear that low viscosity values are needed .
our more detailed computations allow to study two effects , which are important for the evolution , ( 1 ) the amount of matter not accumulated in the disk , because it is accreted on to the white dwarf or lost in the wind from the corona during quiescence ; ( 2 ) the growth of the disk with respect to the 3:1 resonance radius . in this model the disk , with an assumed inner hole of specified radius ,
is stationary and stable . a fluctuation in the mass overflow rate
is then thought to bring the disk to the point of instability and trigger the outburst .
one would then expect quite a fluctuation in repetition times between consecutive outbursts .
this contrasts with the nearly exactly ( only 2% variation ) regular repetition observed and the very regular build - up to outburst in the model presented in this paper .
a further problem might lie in the ( presumed irradiation caused ) large mass transfer required during outburst .
it becomes necessary since the assumed stationary disk contains too little mass to explain the observed outburst energy .
it would require an increase in the overflow rate by a factor of about 300 while observed increases in some dwarf novae are limited to only a factor 2 ( smak 1995 , hameury et al .
1997 ) . also in this model a hole is assumed .
we note that here as in the preceding model , a more detailed physical picture of the hole formation and change of its size can have a significant effect on the duration of the quiescence . the chosen low mass overflow rate from the secondary makes it possible to have the outburst occur after only 36 years .
but the low amount of matter provided for the outburst then shows up in the computed short duration .
the authors suggest that in su uma systems small , regular outbursts are followed by larger superoutbursts .
but for wz sge such regular outbursts were not observed .
it seems that the claim `` down with low @xmath38 '' can not be realized as suggested .
smak ( 1993 ) used iue observations of wz sge , taken 1979 july 11 and 1981 november 22 , 0.6 and 2.4 years after the december 1 1978 outburst to determine temperatures . together with observed fluxes he derived values for the product @xmath96 , where x is the fraction of the full `` stellar disk '' @xmath97 that is not occulted by the accretion disk and is radiating with the effective temperature determined .
his values are 2.72@xmath98 and 1.56@xmath98 , respectively .
smak assumed @xmath99=0.63 ( @xmath99 value in 1979 , @xmath100 in 1981 ) corresponding to a full disk whose lower half is partially occulted by the accretion disk ( i=75@xmath71 ) . from the ratio of the derived values @xmath96 one gets @xmath101=0.35
this was interpreted as shrinking of the white dwarf radiating area to an equatorial belt .
the long - term iue observations indicate a cooling of the white dwarf ( la dous 1994 ) .
our model allows another interpretation .
if in the 1979 observation an inner disk hole was present as expected theoretically in the early phase of quiescence , flux from one full hemisphere of the white is observable and , in smak s terminology , the fraction @xmath99 is equal to 1 . from the ratio of the values determined by smak then follows @xmath100=0.59 .
this agrees , within the errors with 0.63 corresponding to the fraction visible if the hole is closed .
this changes the estimate for the distance to wz sge from smak s value and error d=@xmath102pc to d=@xmath103pc for the @xmath104 white dwarf . for the critically rotating 0.69@xmath13 white dwarf
one may estimate ( see hachisu 1986 ) a radius @xmath1 0.77 @xmath105 of a `` stellar disk '' of the same area as the rotationally flattened cross section of the real star . with the equatorial radius of
about 1.2@xmath16 cm this yields the distance estimate d=@xmath106pc .
a not critically rotating star of larger mass would be correspondingly closer .
the observed change in the uv flux between 1979 and 1981 supports the prediction of formation and disappearance of an inner hole of the quiecent accretion disk . from our computations we found that the 3:1 radius can be reached already several years before the outburst ( in fig .
3 the bars at the evolutionary lines mark the time when the resonance radius is reached ) .
this means that the outer disk from this time on has an excentric shape and superhumps should be seen .
they dissipate the work done in transferring accreted angular momentum back to the orbit , @xmath107 , which is about twice the hot spot luminosity ( @xmath108 orbital frequency ) .
we note that the weak early superhumps in outburst do not differentially precess ( whitehurst 1988 , deyoung 1995 , patterson 1995 , kato et al .
1996 ) and we might expect the same stationarity in the lightcurve for such quiescent superhumps .
our detailed computations show that values @xmath0 around 0.001 are necessary to accumulate enough matter in the disk to explain outburst duration and amplitude .
our investigation shows new features which are essential for the understanding of wz sge .
the evaporation plays an important role in the way that it prevents a premature outburst .
as long as the disk size grows , more than half the matter transferred from the secondary star is used up to fill the newly added outer disk areas . only after the disk has reached the 3:1 resonance radius no further growth occurs and all accreted matter can flow inward .
this results in a higher mass flow rate and triggers the outburst .
evaporation goes on all the quiescence with the rate corresponding to the inner edge of the disk close to the white dwarf .
this means accretion of matter via the corona on to the white dwarf at a constant rate of about @xmath9g / s . without evaporation the outburst behaviour of wz sge
can also be modeled , but then requires small values @xmath0 for the cool state , @xmath1090.001 , and low mass overflow rate from the companion sta , r of order @xmath8=1.4 @xmath9g / s .
the finding that the value of @xmath0 has to be smaller than usually assumed for dwarf novae raises again the question of the cause .
a new interesting suggestion of gammie & menou ( 1998 ) relating low viscosity to low magnetic reynolds numbers does not seem to explain why the dwarf novae in late evolutionary phase should have low values with otherwise very similar quiescent disks .
most of our models predict superhumps before the outburst .
their appearance in late quiescence , observable as a change in the orbital lightcurve , would signal that the disk has reached its final radius at 3:1 resonance and that after the disk relaxation time the next outburst will occur .
models for wz sge are relevant for dwarf novae in a late stage of secular evolution .
wz sge is already a marginal system in that an outburst occurs only after a very long accumulation time . if the accretion rate would be only slightly smaller , no outbursts would arise .
instead the disk becomes stationary . since the mass transfer rate from secondary star to white dwarf decreases during the secular evolution of dwarf novae ( kolb 1993 ) we expect a large number of systems `` beyond wz sge '' .
such systems are not conspicuous optically , but the x - rays from all these binaries add up to quite an amount of radiation .
consequences arising from this will be investigated separately ( kolb et al . , in preparation )
cheng f.h . ,
, szkody p. , huang m. , 1997 , apj 484 , l149 cordova f.a , mason k.o . , in : accretion driven stellar x - ray sources , eds .
lewin , e.p.j .
van den heuvel , 1983 , cambridge univ .
press cambridge de young j.a . , 1995 ,
6157 eracleous m. , halpern j. patterson j. , 1991 , apj 382,290 gammie c.f .
, menou k. , 1998 , apj letters , in press hachisu , i. , 1986 , apjs 61 , 479 hameury j .-
, lasota j .-
, hur j .-
, 1997 , mnras 287 , 937 ichikawa s. , osaki y. , 1992 , pasj 44 , 15 kato t. , nogami d. , baba h. , 1996 , pasj 48 , l21 kolb u. , 1993 , a&a , 271 , 149 la dous c. , 1994 , space science reviews 67 , 1 lasota j .-
, hameury j .-
, hur j .-
, 1995 , a&a 302 , l29 liu f.k .
, meyer f. , meyer - hofmeister e. , 1995 , a&a 300 , 823 liu b.f .
, meyer f. , meyer - hofmeister e. , 1997 , a&a 328 , 247 livio m. , pringle j.e
. , 1992 , mnras 259 , 23p lubow s.h .
, 1991 , apj 381 , 268 lubow s.h . , 1994 , in : theory of acretion disks - 2 , ed .
duschl et al .
kluwer dordrecht , p. 109
lubow s.h .
, shu f.h . , 1975 , ap 198 , 383 ludwig k. , meyer f. , 1998 , a&a 329 , 559 ludwig k. , meyer - hofmeister e. , ritter h. , 1994 , a&a 290 , 473 meyer f. , 1997 , in : accretion disks - new aspects , eds .
meyer - hofmeister e. , spruit h. : lecture notes in physics 487 , springer verlag , berlin , heidelberg , new york , p 338 . meyer f. , meyer - hofmeister e. , 1994 , a&a 288 , 175 meyer f. , meyer - hofmeister e. , liu f.k . , 1996 , in : rntgenstrahlung from the universe , eds .
zimmermann h.u . , trmper , j.e.,yorke h. , mpe report 263 , max - planck - institut fr extraterrestrische physik , garching , p. 163 mattei j. , 1980 , j. r. astron .
can . , 74 , 53 mukai k. , shiokawa k. , 1994 , apj 418 , 863 narayan r. , popham r. , 1993 , nat 362 , 820 osaki y. , 1989 , pasj 41 , 1005 osaki y. , 1995 , pasj 47 , 47 osaki y. , 1996 , pasp 108 , 39 patterson j. , 1995 , iau circ .
6157 patterson j. , raymond r. , 1985 , apj 292 , 535 patterson j. , richman h. , kemp j. , mukai k.,1998 , pasp 110 , 403 patterson j. , mcgraw j.t . , coleman l. , africano j.l .
, 1981 , apj 248 , 1067 popham r. , narayan r. , 1995 , apj 442 , 337 pringle j.e . , 1977 , mnras 178 , 195 pringle j.e . , savonije g.i . , 1979 ,
mnras 187 , 177 richman h.r.,1996 , apj 462 , 404 shakura n.i .
, sunyaev r.a . , 1973 ,
a&a 24 , 337 sion e.m . ,
cheng f.h .
, long k.s .
et al . , 1995 , apj 439 , 957 smak j. , 1993 , acta astron .
43 , 101 smak j. , 1995 , acta astron . 45 , 355 spruit h.c . , rutten r.g.m . , 1998 , mnras in press warner b. , livio m. ,
, 1996 , mnras 282 , 735 welsh w.f . , skidmore w. , wood j.h . , cheng f.h .
, sion e.m .
1997 , mnras 291 , l57 whitehurst r. , 1988 , mnras 232 , 35 | we model the evolution of the accretion disk of wz sagittae during the long quiescence . we find that the large amount of mass in the disk derived from the outburst luminosity is a severe constraint and demands values of @xmath0 @xmath1 0.001 in contradiction to some recent suggestions .
we include in our computations the formation of an inner disk hole and the growth of the disk due to redistribution of angular momentum .
we find a new mode of disk evolution .
the disk is quasi - stationary .
only about half of the mass transfered from the companion star flows through the disk , the other half is needed to build up the steadily growing outer disk . when the 3:1 resonance radius is reached
the disk growth ends . from then on all transferred matter
flows inward , the surface density increases , leading to an outburst within a few years .
we predict superhumps at low luminosity during this last phase .
we discuss x - rays expected , the white dwarf mass and distance to wz sagittae . |
the highest resolution observations in astronomy are possible via very - long - baseline interferometry ( vlbi ) , which can be conducted routinely up to frequencies of 86ghz , yielding sharp images of the brightest objects in the universe resolved up to some tens of microarcseconds ( @xmath1as ; see , e.g. , the latest pioneering results by doeleman _ et al _ 2008 at even higher frequencies ) .
typical resolutions are of the order of 0.53mas at the frequencies where atmospheric and ionospheric effects do not severely affect the performance of the observations .
high resolution observations of active galactic nuclei ( agn ) provide access to the regions close to the central engine , believed to be a super - massive black hole ( e.g. , zensus 1995 ) .
those are complemented by single - dish flux density monitoring to follow the changes in the innermost part also from the spectra , which reveal variations of the optically - thin ( synchrotron ) and optically thick regions ( closer to the self - absorbed jet base ) . additionally , radio spectra turnover frequency imaging yields information on the jet physics , particle density , composition , etc .
( e.g. , lobanov 1998 ) .
polarisation observations at the higher frequencies can also reveal the direction and nature of the magnetic fields in the inner jet .
following the standard model ( e.g. , marscher 2006 ) the emission of a radio loud agn at different radio wavelengths originates at different distances from the base of the jet . at higher frequencies ,
the bulk emission comes from closer to the central engine , since the synchrotron - self - absorption of the inner part of the jet is reduced at shorter wavelengths . assuming a central black hole surrounded by an accretion disk , and perpendicularly to this , a jet with superluminal plasma accelerated , the radio emission is produced by synchrotron at a region where no self - absorption is present .
high - resolution radio observations are complemented by other regions of the spectra .
agn broadband spectra show absorption , reflection and emission lines as well as a power - law continuum radiation which can be studied by analyzing the spectral energy distribution .
their spectra show radio variability . in the x - rays , the observations of the iron k@xmath2 line at 6.4kev ( the most prominent fluorescent line ) probes the agn circumnuclear environment , from its variability and profile .
the combination of radio and x - ray has been successful , i.e. , in the case of 3c120 ( marscher _ et al _ 2002 ) , where x - ray flux density `` dips '' precede vlbi jet ejections .
the time delays are of the order of weeks to months caused by the offset of the core from the central engine and by the core shift between the observation frequency and the self - absorption turnover frequency .
these `` dips '' can be put in the context of the discoveries in x - ray binaries ( see fender & belloni 2004 ) .
flux - density selected samples of radio - loud agn are usually dominated by type 1 sources , with a one - sided jet morphology in high resolution radio images and the counter - jet usually not seen due to doppler relativistic boosting ( see e.g. , ros , these proceedings ) .
type 1 sources have an accretion disk which is seen face - on , and no absorption is seen in the soft region of the x - ray spectrum .
type 2 sources have an edge - on accretion disk , detected by strong absorption in the soft x - rays ( fabian _ et al _ 2000 ) , and free - free absorption is observed at the multi - wavelength radio structure imaging close to the central engine .
ngc1052 ( b0238@xmath3084 , j0241@xmath30815 ) , classified as low - ionization nuclear emission - line region object ( liner ) , is a key source between both types , since it is radio loud , oriented in the plane of the sky , and emits in x - rays . due to its closeness ( with a redshift of @xmath4=0.0049 ( knapp _ et al _ 1978 ) , ngc1052 , despite of its relatively low radio luminosity , is still very bright in radio wavelengths and can be detected at all observing vlbi bands in jet and counter - jet ( see , e.g. , lee _ et al _ 2008 for 86ghz for the highest available vlbi frequency results published , with just one bright component ) . at arcsecond scales ( kadler _
et al _ 2004a ; cooper _ et al _ 2008 ) it presents an east - west faint emission ( of 30@xmath5 in extension ) with a dominating core , which is also seen on mas - scales with a double - sided radio with a position angle of @xmath665@xmath7 .
the jet axis is considered to be nearly parallel to the plane of the sky , with values between a lower limit of 57@xmath7 and 72@xmath7 ( vermeulen _ et al _ 2003 ; kadler _ et al _ 2004b ) .
sawada - satoh _ et al _ ( 2008 ) determine values around 76@xmath780@xmath7 .
high opacities are measured toward the innermost component of the western jet ( kadler _ et al _ 2004b ; kameno _ et al _ 2001 ; vermeulen _ et al _ 2003 ) , and also at the inner edge of the eastern jet at centimetre wavelengths ( kadler _ et al _ 2004b ; sawada - satoh _ et al _ 2008 ) .
millimetre observations can see through this screen and reveal some structure at the innermost region .
emission from h@xmath8o spots has been detected both at jet and counter - jet ( sawada - satoh _ et al _ 2008 and references therein ) , redshifted by 50350kms@xmath9 with respect to the systemic velocity of the galaxy .
this water emission is interpreted to be located at the region where opacity dominates due to the thermal plasma in the obscuring torus via free - free absorption .
x - ray spectra further support the presence of a dense gas torus ( guainazzi & antonelli 1999 ; weaver _ et al _ 1999 ; kadler _ et al _ 2004a ) toward the nuclear x - ray source .
we started in 2005 a multi - mission campaign to track the ejection of new features at the base of the jet and counter - jets and compare with brightness monitoring and spectral studies in x - rays and radio , with the goal of establishing a relationship between changes in the high energy and the jet production in agn .
first results of the campaign were presented in kadler _ et al _ ( 2006 ) and ros _ et al _ ( 2007 ) .
vlba images of ngc1052 .
the images shown are spaced by their relative time intervals .
the contours are logarithmic with steps of factor 2 .
the alignment is relative to the emission gap . at the left panel
we show images from the 2 cm survey / mojave programme , registered over a decade . at the right panel
we show preliminary images from our monitoring programme at high frequencies , during three years since early 2005 .
the images have been convolved with common restoring beams , being @xmath10mas in p.a . for the @xmath112 cm images .
the beam for the shorter wavelengths are shown as grey ellipses next to the wavelength label .
the lowest contours are set accordingly to the noise level individually for each epoch , with values of some mjybeam@xmath9 .
, scaledwidth=95.0% ] ngc1052 has been observed with the vlba by the 2cmsurvey / mojave programme since 1995 , with typical resolutions of [email protected] in p.a . 0@xmath7 .
the images resulting from these measurements are shown in the left panel of fig .
[ fig : allmaps ] . to reproduced the structural variations in the jets of ngc1052 , we modelled the observed radio visibilities with gaussian circular functions .
we cross - identified the modelled features from epoch to epoch by criteria of continuity in the positions and flux densities .
a complex data base of functions results , with over 30 components along the more than 10 years of observations . in fig .
[ fig : kinematics ] we show the proper motions of these model - fitted features ( the observations and results are part of the complete kinematic analysis of the mojave/2cmsurvey program ( m l lister _ et al _ , in preparation ; ros , these proceedings ) .
the reference origin to align the images has been chosen as the middle point between the two innermost components , although we note that the opacity effects can shift this position .
this sets the question of the ejection epoch for the emerging components , since we can extrapolate the linear fit to be zero , but this can be considered as a lower limit .
notice that when computing kinematic epoch ejections for one - sided radio sources , the origin values are usually extrapolated to the core position ( @xmath13 surface ) .
we show the speeds and maximum flux densities for selected components in table [ table : kinematics2 cm ] . a more detailed analysis on the ejection , possible acceleration of components , non - linear motions , etc .
, is going to be presented in m kadler _ et al _ ( in preparation ) and elsewhere .
model fitting results to the parsec - scale twin - jet radio structure of ngc1052 at 15ghz .
the relative distances to the central region are shown as a function of time for the jet ( top ) and the counter - jet ( bottom ) .
the solid lines are linear regression fits to the data points , and their slopes give the reported proper motions .
the typical speeds observed are of @xmath140.26@xmath0 both in jet and counter - jet .
table [ table : kinematics2 cm ] shows the speed values for selected components . ,
scaledwidth=95.0% ] ccr@@xmath15lr@@xmath15lr@@xmath15l & no of & & + i d & epochs & & & + & & + e01 & 25 & 447&174 & 0.15&0.06 & 1971&13 + e03 & 26 & 595&26 & 0.196&0.009 & 1987.4&0.7 + e11 & 14 & 824&27 & 0.272&0.009 & 1999.1&0.1 + e17 & 12 & 1038&34 & 0.34&0.01 & 2002.8&0.1 + e18 & 10 & 889&27 & 0.294&0.09 & 2003.2&0.1 + w11 & 11 & 536&46 & 0.18&0.02 & 2003.3&0.3 + w06 & 14 & 813&53 & 0.27&0.02 & 1999.6&0.2 + w03 & 23 & 757&36 & 0.25&0.01 & 1994.1&0.1 + w01 & 20 & 763&30 & 0.25&0.01 & 1982.8&0.8 + we have been monitoring the structure of ngc1052 at 22ghz and 43ghz using the vlba every six weeks ( with some larger gaps in the time sampling due to scheduling reasons ) since early 2005 . the observations at these frequencies have different dynamic ranges and beams due to the different constraints during the observing epochs ( e.g. , weather constraints or failure of one or two stations of the vlba ) .
we present contour plots of the preliminary images convolved with a common beam in the right panel of fig .
[ fig : allmaps ] .
a detailed model fitting of the observed features is in progress , as well as the astrometric registration of the images , since they were observed in phase - referencing mode , including scans on the calibrator j0243@xmath30550 .
the imaging results show the free - free absorption screen to be still present at 22ghz and getting transparent at 43ghz .
new components can be traced when being ejected from the innermost part of the jet .
the light curves of the single - dish telescope measurements show a decay from 2005 to early 2008 , with some hints of a rising flux density for the latest months ( h aller and m aller , private communication ) .
our long - term 15ghz results confirm the morphology and kinematics reported in vermeulen _ et al _ ( 2003 ) , with prominent jet and counter - jet features moving downstream with mildly relativistic speeds of @xmath140.25@xmath0 .
the source has experienced a decrease in flux density at all frequencies since 2005 .
this is caused by the fading out of the features travelling outward the jet , which is not compensated by new , strong components at the base of the jet .
there are some hints that this situation is beginning to change from the early 2008 observations .
the results of the multi - band campaign , including _ rxte _ light curves and spectra , _ xmm - newton _ , _ suzaku _ , and _
chandra _ observations , will be published elsewhere .
our imaging campaign at the highest frequencies is planned to last at least until early 2009 .
the mojave program will be also continued beyond that and at least until 2010 , providing observations of ngc1052 at 15ghz every several months .
the vlba is operated by the national radio astronomy observatory , which is a facility of the usa national science foundation operated under cooperative agreement with associated universities , inc .
cooper n j , lister m l and kochanczyk m d 2008 _ apjs _ * 171 * 376388 | we have studied the inner regions of the liner galaxy ngc1052 since the mid 1990s at high resolution with 15ghz very - long - baseline interferometry observations . a compact , two - sided jet structure is revealed , with multiple sub - parsec scale features moving outward from the central region with typical speeds of 0.26@xmath0 .
complementary to this , since early 2005 we are performing a multi - mission campaign of observations of this source , including x - ray spectroscopy , x - ray , and radio flux density monitoring , and vlba observations at 22ghz and 43ghz .
x - ray variability is present at time scales of weeks , comparable with the structural changes observed by vlbi . here
we present first results of the high - resolution imaging observations and discuss these findings in the context of the multi - band campaign . |
state transfer has been almost universally studied for hamiltonians @xmath5 which have the very particular property of excitation preservation , @xmath10,\ ] ] which means that the number of qubits in the @xmath11 state are a constant of the motion .
if @xmath12 of the @xmath8 qubits are in the @xmath11 state , then we say the system is in the @xmath13 excitation subspace .
these studies then proceed by setting the initial state of every qubit not @xmath2 to @xmath14 so that when a state is placed on qubit @xmath2 , we are in a superposition of the 0 and 1 excitation subspaces , and the task becomes @xmath15 up to some phase @xmath16 which we could compensate for later ( here we distinguish between _ perfect _ transfer and arbitrarily accurate transfer , concentrating on the former ) .
the 0 excitation subspace is composed of a single state , and therefore does not evolve in time .
hence , we only have to concentrate on the evolution of the 1 excitation subspace , which is simply described by an @xmath17 matrix , @xmath18 .
there are many different hamiltonians that have the same @xmath18 such as the @xmath0 , heisenberg and coupled harmonic oscillator hamiltonians @xcite , not to mention more exotic varieties involving q - deformed oscillators @xcite
. we will continue to study transfer in the single excitation subspace of an excitation preserving hamiltonian . for coupled harmonic oscillators ( i.e. non - interacting bosons ) , this is no restriction since the single excitation subspace describes the dynamics of each individual boson no matter what other bosons are present in the system . while this is not true for spin systems hamiltonian to non - interacting fermions , for which a similar argument holds @xcite . ]
, it is a reasonable restriction in the sense that higher excitation subspaces can be described as a single excitation subspace of a system of more qubits @xcite .
the trade - off is that the necessary and sufficient conditions that we derive here are just necessary conditions in higher excitation subspaces for a given initial state on the rest of the system ( with sufficiency assured if transfer of excitations is possible from the same initial state on two neighbouring excitation subspaces in the same time ) . within the single excitation subspace
, we denote an excitation on qubit @xmath19 by @xmath20 consequently , the hamiltonian is expressed as @xmath21 there are two physically important restrictions that can apply to the hamiltonian .
the first is to make the hamiltonian real , i.e. @xmath22 .
it is not impossible that a hamiltonian would contain complex coefficients , but most naturally arising , or readily implemented , hamiltonian terms , such as @xmath23 only give rise to real hamiltonians .
the second is a geometry constraint .
consider a graph @xmath24 which is composed of edges @xmath25 and vertices @xmath26 , where we establish a correspondence between the @xmath8 vertices and the @xmath8 states @xmath27 .
the only allowed non - zero couplings @xmath28 are those for which the pair of vertices @xmath29 are an edge of the graph for all edges , giving a correspondence between the adjacency matrix of the graph and @xmath18 . while
just a special case of our more general formalism , we will also discuss this restriction in sec .
[ sec : uniform ] . ] .
the graph naturally induces a concept of transfer distance the minimum number of edges that one must follow to traverse the graph from the input node @xmath2 to the output node @xmath7 .
we consider short transfer distances uninteresting because they do not simplify the communication demands between distant qubits . in our analysis
, it will often help to work in the diagonal basis of @xmath18 , @xmath30 i.e. @xmath31 is an eigenvector of @xmath18 of eigenvalue @xmath32 .
note that for real hamiltonians , all the elements @xmath33 must also be real .
our primary goal is to prove necessary and sufficient conditions for perfect state transfer in the first excitation subspace of a spin - preserving hamiltonian .
these conditions can be expressed as the existence of a state transfer time @xmath4 , and transfer phase @xmath16 , in a condition on the eigenvectors @xmath34 for all @xmath19 , and on the eigenvalues , @xmath35 for all @xmath19 for which @xmath36 where @xmath37 is an integer , in close parallel to the equivalent results on chains @xcite .
however , we require the additional definition that @xmath38 to prove necessity , we start from the definition of state transfer in the single excitation subspace , requiring that there exists a @xmath4 and @xmath16 such that @xmath39 by taking the overlap with an eigenvector , @xmath40 one can immediately read off that @xmath41 by matching the weights .
the phases must also match @xmath42 up to a multiple of @xmath43 .
having proved necessity , we prove sufficiency .
assume that a suitable @xmath4 and @xmath16 exist .
so , @xmath44 we can now supply the conditions on @xmath32 , @xmath45 this yields a rather simple set of conditions which one use to verify that perfect transfer occurs in a network .
note that the task of testing the existence of a suitable time @xmath4 and phase @xmath16 can be largely neglected by taking differences and ratios of the eigenvalues .
one of the major advantages of such a characterisation for the spin chains was that this lead to the ability to calculate the required coupling strengths simply by specifying the desired spectrum .
this was due to a specific property of spin chains in that , for mirror symmetric ( real ) hamiltonians , after ordering the eigenvalues , @xmath46 so it was very easy to specify a suitable spectrum . for real hamiltonians ,
all the @xmath47 must be either @xmath48 or @xmath49 ( imposing that all the eigenvalues are integers , up to a scale factor and uniform shift ) , but as one varies the coupling strengths to adjust the eigenvalues , which @xmath47 take which values can change .
this was already observed as a practical problem in @xcite , making the task of designing perfect transfer networks far harder , though not impossible .
with these necessary and sufficient conditions in place , we can start to explore the general features of transfer in networks .
we want to know what is in principle possible without reference to specific hamiltonians .
our first observation is a consequence of the study in @xcite . for real hamiltonians , constrained to a bipartite coupling graph ( which also imposes that @xmath50 ) ,
the transfer phase @xmath51 is @xmath52 if the transfer distance is even and @xmath53 if the transfer distance is odd .
a bipartite graph is one whose vertices can be divided into two colourings , red and blue , such that edges only connect between a red and a blue vertex .
let us define @xmath54 for a hamiltonian @xmath18 which is connected via a bipartite coupling graph , it must be true that @xmath55 this means that for any eigenvector @xmath31 of @xmath18 with @xmath56 , @xmath57 must also be an eigenvector of @xmath18 , but with eigenvalue @xmath58 .
now , let s assume ( without loss of generality ) that @xmath2 is in the ` red ' partition
. we can write @xmath59 ( for simplicity of notation , we have assumed that there are no 0 eigenvalues , but recall that we only need to consider one zero eigenvector with non - zero overlap on @xmath2 , and it must satisfy @xmath60 , which allows us to treat this special case ) .
now let us evolve the state : @xmath61 and calculate the overlap with some vertex @xmath62 , remembering that for a real hamiltonian the overlaps are real . if @xmath62 is a red vertex , @xmath63 , then @xmath64 so the amplitude is always real .
since @xmath62 was a red vertex , it must be an even distance from @xmath2 .
on the other hand , if @xmath62 is a blue vertex , then @xmath65 and @xmath66 so the amplitude is always imaginary .
this provides another advantage when deciding if a network is capable of perfect transfer .
symmetries are an important tool in understanding any system .
indeed , the construction of perfect state transfer chains originally relied heavily on an assumption of symmetry @xcite , which was subsequently @xcite proven to be necessary .
we are thus interested in whether every perfect transfer hamiltonian @xmath18 has a symmetry operator @xmath67 which satisfies @xmath68 and @xmath69 .
the existence of a symmetry is proven by construction . by defining a unitary rotation that is diagonal in the basis of the hamiltonian
, it will clearly satisfy the commutation property .
specifying the phases as @xmath70 allows us to verify the desired transformation @xmath71 for a real hamiltonian @xmath18 , @xmath72 , so @xmath73 .
it is worth observing that there is still a continuous freedom in the definition of @xmath67 the phases that are applied to the eigenvectors for which @xmath74 which gives a way to see that @xmath67 is not necessarily a permutation ( which can not be continuous ) .
this manifests itself in the example of a chain below
if @xmath67 were a permutation , it would have to be the mirror symmetry operator . if one knows the symmetry operators of a system for some _ a priori _ reason , this identifies the values @xmath47 ( the eigenvalues of @xmath67 ) , and associates them with specific eigenspaces .
hence , for systems where @xmath67 can be identified , and the eigenvalues can be modified while preserving the symmetry , we should be able to construct perfect transfer networks .
this was the key insight for designing chains , and can hopefully now be applied in other scenarios .
we have mentioned at several points that a necessary condition on perfect end to end transfer chains is the presence of mirror symmetry @xcite .
it is also the case that any hamiltonian which achieves perfect transfer between opposite ends of a chain can equally achieve transfer between any mirror symmetric points .
one might be drawn into the expectation that all perfect transfers ( not just end to end ) on chains are hence governed by mirror symmetric coupling schemes .
this is not the case , as we will show by specific construction .
consider a matrix @xmath75 one can prove that to transfer between qubits 2 and 4 ( i.e. to create the evolution @xmath76 up to some phase @xmath16 , and for some time @xmath4 ) , one simply has to impose that @xmath77 and that the eigenvalues of @xmath18 are ( up to a scale factor ) alternately even and odd integers .
this eigenvalue condition is the same as for extremal transfer on the chain , but the coupling strengths are less restricted .
you can readily verify that @xmath78 is one class of non - symmetric examples which implement perfect transfer between qubits 2 and 4 with @xmath79 .
following on from some discussion of transfer rate in @xcite , @xcite examined the possibility of perfect transfer at high rate .
this involves inserting a second state at site @xmath2 before the first state has been removed at site @xmath7 , and yet requiring that the first state should still arrive perfectly .
some weak bounds were proven on possible rates for spin chains .
we will now prove stronger bounds for all networks described by a real hamiltonian ( one can prove identical bounds for arbitrary hamiltonians ) .
a necessary condition for the ability to insert a second quantum state into the spin network ( on the same input qubit ) at some time @xmath80 without disturbing the first quantum state is that @xmath81 for chains , this condition is sufficient , but for more general networks , this will not be the case .
ultimately , we will be interested in inserting many different states at different times .
again , for the chain , the only necessary condition is @xmath82 for all of the possible time intervals @xmath80 . for networks
, the evolution of the many - excitation state could be quite different to the evolution of the single excitation states , so it might be that there are further conditions imposed .
however , it is still a necessary condition , because our perfect transfer at high rate must work for all possible input states , which includes setting all previous inputs to @xmath14 except for one , from which one can extract that same condition .
thus , our question relates to whether , given there are @xmath83 unique time intervals @xmath84 at which @xmath85 , perfect transfer can occur to a site @xmath86 at a distance @xmath87 in a time @xmath4 . with @xmath83 time intervals
, one can have @xmath83 unique times @xmath88 by imposing fixed intervals .
we start by expressing our condition on the transfer distance as , for each integer @xmath89 , @xmath90 this can , instead , be written as @xmath91 where @xmath92 , which is readily transformed into a linear equation @xmath93 having resolved the possible degeneracies in the system , we have reduced from a system of size @xmath8 to @xmath94 , the number of unique eigenvalues .
each of the @xmath95 rows is linearly independent .
the next constraint that we must add is that of normalisation , @xmath96 now we need to add in the conditions corresponding to @xmath85 .
all our conditions so far have just been based on real values , and we will maintain this by dividing these conditions into real and imaginary parts .
the real parts give @xmath97 and , similarly , the imaginary components give @xmath98 given that all these times @xmath88 are less than @xmath4 , the half period of the system ( since we are assuming the hamiltonian is real and performs perfect transfer , it is periodic with a period @xmath99 ) , all of these rows must be linearly independent from each other . hence
, if a suitable set of @xmath100 is to possibly exist , it must be the case that @xmath101 in particular
, imagine we had @xmath94 conditions not including the normalization condition .
these would impose that all the @xmath102 , so it would be impossible to satisfy the normalization condition . ideally , we want the maximum transfer distance , which would be @xmath103 ( a chain ) , imposing that @xmath104 , as conjectured in @xcite . the only way to increase the perfect transfer rate is to reduce the transfer distance . however , you ca nt also lower the state transfer time ( as you would expect by shortening the transfer distance ) .
this is because the margolus - levitin theorem @xcite imposes a minimum time for evolving between two orthogonal states , such as a @xmath105 as an input state , and the @xmath106 required for the next input .
hence the transfer time is bounded from below by @xmath107 . in some sense , the ` standard ' perfect state transfer chains @xcite saturate the bound of eqn .
( [ eqn : ratebound ] ) in that for a chain of @xmath8 qubits , any state @xmath27 transfers a distance @xmath108 , but there are @xmath109 distinct times @xmath88 such that @xmath110 . unfortunately , however , these times are not equally spaced , so they are not useful for achieving a high rate of transfer .
it is worth noting that our analysis breaks down at the @xmath104 limit since @xmath4 is the length of the period , not the half period .
this means that half of the @xmath111 conditions can be the same as the other half , for suitably chosen values of @xmath88 .
we end up with @xmath112 , and this bound was saturated in @xcite for the sequential quantum storage solution . the idea of being able to choose which of several recipients , @xmath7 , @xmath113 etc .
, is to receive a quantum state was initially studied in @xcite , and some aspects have been further considered in @xcite , there may be gaps in some of the proofs , such as theorem 1 in @xcite . ] .
this task has since become known as routing @xcite and , by allowing some minimal control , it was achieved efficiently in a regular network of nearest - neighbour coupled spins .
routing is potentially an important property for a system to possess since this allows us to significantly alter the connectivity of an array of sites in a way that direct communication between pairs of sites does not .
however , we are now going to make a proof by contradiction that shows that for real hamiltonians , routing between multiple sites is impossible , and subsequently we will bound the number of possible recipients as a function of transfer distance for more general hamiltonians .
this is something that the constructions of , for instance @xcite , give no information about , or control over .
while they allow for the inversion from a desired unitary to a hamiltonian , this provides no control over any spatial limitations in the coupling patterns , variations of coupling strengths ( such as uniformity of coupling , or even real values ) and typical solutions couple every qubit to every other qubit in a completely arbitrary manner , entirely missing the point of state transfer , which is intended for use in systems of low connectivity .
we start by assuming that perfect transfer is possible between @xmath2 and @xmath7 , and the minimum time in which this occurs is @xmath114 .
so , we have @xmath115 recall that since the hamiltonian is real , all the @xmath47 are 0 or @xmath49 .
so , this means that if we evolve for twice the time , we have a perfect revival , @xmath116 demonstrating that the dynamics are periodic .
now let us assume that perfect routing is possible , meaning that there must exist a time @xmath117 such that @xmath118 however , by identical arguments , it must be the case that @xmath119 and hence @xmath120 this is just perfect transfer between @xmath7 and @xmath2 in time @xmath121 , which is impossible by assumption that @xmath114 is the shortest state transfer time .
hence the transfer from @xmath2 to @xmath113 ca nt exist , and if there is transfer to one site , there can not be transfer to any other sites . in order to break this restriction
, we have to take hamiltonians with complex entries .
this is exactly what happens in papers such as @xcite .
the preceding argument is rather powerful , revealing that many other intermediate states ca nt exist , since once can repeat it for any target state which is just a superposition of eigenvectors where , up to a global phase , all the amplitudes are real .
one obvious example comes from the bipartite systems we discussed previously a real hamiltonian on a bipartite lattice which is capable of perfect transfer can never , at any intermediate time , produce a state which is entirely localised on just one of the bipartitions of the graph . in the previous subsection
, we derived a trade - off between the maximum transfer rate and the distance of transfer .
we can do the same for a general case of routing , where we wish to transfer to @xmath122 different possible sites from @xmath2 , at locations @xmath123 and times @xmath124 .
as before , we have @xmath92 and @xmath125 .
now we have @xmath122 conditions for the perfect transfer @xmath126 ( the independence of these conditions is no longer imposed by periodicity , of which we are not assured , but by the assumption that the output vertices are distinct , yielding orthogonal states ) and there s the same normalisation condition .
now we also want to impose that all @xmath122 target vertices are at least a distance @xmath87 from @xmath2 .
hence , for @xmath127 to @xmath95 we have @xmath128 even by not restricting the @xmath100 to be real ( let alone positive ) , we arrive at the bound @xmath129 so , if you want to route between every possible vertex of a network , you must have transfer distance 1 .
this is exactly what happened in the examples of @xcite , but now we know that it s impossible to do better , making the results of @xcite all the more remarkable , achieving routing at a high transfer rate with only the addition of very modest controls .
the interpretation of the periodicity of the system also allows a minor insight into the state transfer time .
let us define the eigenvalue gaps between eigenvectors with support on @xmath2 as @xmath130 ( where the @xmath32 are ordered ) , and fix @xmath131 to be the largest real number such that @xmath132 is an integer for all @xmath19 .
then it must be the case that the state transfer time is given by @xmath133 .
this is because it allows @xmath134 to be @xmath52 , as would be required for state transfer , and yet @xmath135 , which corresponds to a perfect revival on the input spin . perhaps of most interest
would be finding graphs which are uniformly coupled ( to be defined momentarily ) , preferably maximising the growth of transfer distance with the total number of vertices , and keeping the degree of each vertex low .
there are two natural connections between hamiltonians restricted to the single excitation subspace and the underlying graph structure .
the first is the @xmath0 model , @xmath136 which has @xmath137 , the adjacency matrix of the graph @xmath24 with edge set @xmath25 .
the second is the heisenberg model , @xmath138 which has @xmath139 , the graph laplacian . our strategy for the two cases will be slightly different , and we start with the heisenberg case . from our necessary and sufficient conditions for state transfer , we know that the eigenvalues of @xmath18 can be written in the form @xmath140 for the eigenvectors @xmath31 which have support on the input vertex , where @xmath141 is a different integer for each @xmath19 , but @xmath131 and @xmath142 are fixed ( and relate to the transfer time and phase respectively ) .
in fact , @xmath143 because we know that a laplacian always has one eigenvector @xmath144 with eigenvalue 0 .
we are now going to assume that every unique eigenspace has support on the input vertex . under this assumption
, we can calculate @xmath145 where @xmath146 is the degree of vertex @xmath19 .
this instantly proves that @xmath131 is rational , and hence all the eigenvalues are rational .
however , it is well known @xcite that any hamiltonian with integer matrix elements and rational eigenvalues in fact has integral eigenvalues . hence , we can utilise the wide variety of results on laplacian integral graphs @xcite .
nevertheless , it is worth emphasising that it is only necessary that the graphs be integral .
it is in no way sufficient .
now we want to know about how the transfer distance is related to the maximum degree of the graph and the number of vertices , under this assumption about the support of the eigenvectors . for a connected graph , the diameter @xmath87 ( the maximum distance between any two points in the graph , which is an upper bound on the transfer distance )
is bounded by @xmath12 , the number of distinct eigenvalues : @xmath147 @xcite . however , since the minimum eigenvalue is 0 , and they are spaced by integers , the maximum eigenvalue must be larger than @xmath148 , and yet is upper bounded by @xmath149 , where @xmath150 is the maximum degree of any vertex .
we conclude that @xmath151 unfortunately , to have a scaling transfer distance , we must scale the maximum degree of the graph .
more general bounds are stated in @xcite , such that for any @xmath152 , @xmath153 this suggests that perhaps the logarithmic trade - off between transfer distance and number of qubits in the hypercube @xcite might be necessary
. however , we are still far from proving this .
instead , we only have the upper bound on @xmath87 of @xmath154 . for the @xmath0 model ,
any eigenvectors with support on the input vertex have eigenvalues @xmath155 where @xmath141 is a rational number .
again , we will assume that all eigenspaces have support on @xmath2 .
this is equivalent to the assumption made in @xcite , which imposed that every vertex in the graph be periodic ( i.e. have a perfect revival after @xmath99 ) , and , indeed , we will arrive at the same conclusions .
we also note that there are known instances where perfect transfer can be found without needing this assumption @xcite . as before ,
we proceed by calculating @xmath156 hence , @xmath142 is a rational number multiplied by @xmath131 ( thereby imposing that the transfer phase is a root of unity ) , which we incorporate into @xmath141 , so @xmath157 .
next consider the characteristic polynomial of @xmath18 , @xmath158 which contains integer coefficients @xmath100 ( because all the matrix elements of @xmath18 are integers ) .
each of the @xmath100 can be equated with a combination of the @xmath159 order products of the eigenvalues , which therefore appear as @xmath160 multiplied by a rational number , so @xmath160 is rational .
it is always true that @xmath161 , so @xmath162 is rational .
provided the graph is not bipartite , there is always a value of @xmath12 for which @xmath163 , so @xmath164 is rational , and hence @xmath131 is rational .
again , this imposes that , in fact , the graph is integral and we can use the many results on integral graphs @xcite .
however , if the graph is bipartite , these conclusions do not hold .
the simplest counter - example is the chain of 3 qubits which has eigenvalues @xmath165 and yet achieves perfect transfer . for these systems , we are not aware of any general bounds trading between the degree and transfer distance of the graph .
however , it is worth noting that if the graph is regular ( i.e. every vertex has the same degree , @xmath150 ) , the laplacian , @xmath166 , and adjacency matrix @xmath2 are related by @xmath167 and hence the previous bounds apply .
in @xcite , several ways of combining and manipulating integral graphs to give new integral graphs were proven : cartesian product of two integral graphs : : + @xmath168 has vertices @xmath169 with edges between @xmath170 if @xmath171 or @xmath172 .
this is the exclusive or .
conjunction of two integral graphs : : + @xmath173 has vertices @xmath169 with edges between @xmath170 if @xmath171 and @xmath172 ( also known as the tensor product of two graphs ) .
strong product of two integral graphs : : + @xmath174 has vertices @xmath169 with edges between @xmath170 if @xmath171 or @xmath172 or both .
join of two regular integral graphs : : + @xmath175 has vertices @xmath176 with edges @xmath177 if @xmath178 or @xmath179 or if one vertex is part of each graph , with the restriction that @xmath180 must be a perfect square .
complement of an integral graph : : + @xmath24 is the same as the original graph but the edge set is inverted .
one might therefore wonder if these same constructions can take perfect transfer graphs ( with the same transfer time ) and produce new perfect transfer graphs .
this study started in @xcite which showed how , by taking two graphs known to exhibit perfect transfer , one can construct a larger graph , via the graph product , that also exhibits perfect transfer .
while this was shown specifically for chains , it is easily generalised to all perfect transfer graphs @xcite . by way of contrast , we give examples in fig . [
fig : counter ] for which the conjunction , strong product and join do not generate perfect transfer graphs .
the complement is a more interesting case . in general
, it can not be true that there is perfect transfer , which one sees by considering the graph of two connected vertices .
this exhibits perfect transfer , but the complement , which is the unconnected graph , clearly does not perform transfer . however , there are some cases where the complement does give a perfect transfer graph , and these instances are readily verified .
let @xmath181 be the state specified in eqn .
( [ eqn : maxev ] ) .
since the graph is regular , @xmath181 is the maximum eigenvector , with eigenvalue @xmath150 . we can use this to write the complement @xmath182 of the adjacency matrix , @xmath183 so , if we perform a state transfer between the same input and output vertices in the same time @xmath4 , since @xmath184 commutes with @xmath2 , the condition on achieving state transfer with the complement is simply @xmath185 , which also applies to the laplacian of a regular graph .
in this paper , we have given necessary and sufficient conditions for the existence of perfect state transfer in a quantum network , using the single excitation subspace .
one should be aware , however , that the dynamics can be much richer in higher excitation subspaces , with the possibility of catalysing otherwise impossible transfers @xcite .
making use of these conditions allows us to easily decide if a system can perform perfect state transfer .
we have proven a bound on the maximum transfer rate .
the routing of quantum states between multiple different sites is impossible if the hamiltonian is real , although these results do not contradict existing schemes for the arbitrarily accurate scenario @xcite , when some degree of control is allowed @xcite , or when complex coupling coefficients are allowed @xcite .
we hope that the insights provided in this paper lead to progress in designing perfect state transfer hamiltonians in a wider class of systems .
using results from spectral graph theory , we have already been able to place bounds on many of the properties of uniformly coupled networks , such as the bound that the transfer distance can be no more than twice the degree of the graph for the heisenberg model or regular @xmath0 models ( provided the graph is periodic ) , and anticipate that much more is possible .
we also think it is important that the vague suggestion that networks should be more robust to perturbations should be put on a more rigorous footing .
99 s. bose , phys . rev .
lett . * 91 * , 207901 ( 2003 ) .
m. christandl , n. datta , a. ekert , and a. j. landahl , phys .
lett . * 92 * , 187902 ( 2004 ) .
g. m. nikolopoulos , d. petrosyan and p. lambropoulos , europhys .
lett . * 65 * , 297 ( 2004 ) ; j. phys . : condens .
matter * 16 * , 4991 ( 2004 ) .
m. christandl , n. datta , t. c. dorlas , a. ekert , a. kay , and a. j. landahl , phys .
a * 71 * , 032312 ( 2005 ) .
a. kay , int .
j. quantum inform . * 8 * , 641 ( 2010 ) .
p. karbach and j. stolze , phys .
a * 72 * , 030301(r ) ( 2005 ) .
j. p. keating , n. linden , j. c. f. matthews , and a. winter , phys .
a * 76 * , 012315 ( 2007 ) . c. k. burrell , j. eisert , and t. j. osborne , phys .
a * 80 * , 052319 ( 2009 ) .
p. j. pemberton - ross and a. kay , phys .
106 * , 020503 ( 2011 ) .
v. kostak , g. m. nikolopoulos , and i. jex , phys .
a * 75 * , 042319 ( 2007 ) .
a. bernasconi , c. godsil , and s. severini , phys .
a * 78 * , 052320 ( 2008 ) .
s. severini , j. phys . : conf .
ser . * 254 * , 012012 ( 2010 ) . c. godsil , arxiv:0806.2074 ( 2008 ) ; arxiv:1011.0231 ( 2010 ) .
a. kay , phys .
lett . * 102 * , 070503 ( 2009 ) ; f. pastawski , a. kay , n. schuch , and i. cirac , quantum inf .
* 10 * , 580 ( 2010 ) .
a. peres , phys .
a * 32 * , 3266 ( 1985 ) .
a. kay , phys .
lett . * 98 * , 010501 ( 2007 ) . c. di franco , m. paternostro , and m. s. kim , phys
. rev . lett .
* 101 * , 230502 ( 2008 ) .
m. b. plenio , j. hartley , and j. eisert , new j. phys .
* 6 * , 36 - 36 ( 2004 ) .
s. linnocente , c. lupo , and s. mancini , j. phys . a : math
. theor . * 42 * , 475305 ( 2009 ) . c. albanese , m. christandl , n. datta , and a. ekert , phys
. rev .
lett . * 93 * , 230502 ( 2004 ) .
t. j. osborne , phys .
b * 74 * , 094411 ( 2006 ) .
a. kay , d. k. k. lee , j. k. pachos , m. b. plenio , m. e. reuter , e. rico , optics and spectroscopy , * 99 * , 355 - 372 ( 2005 ) .
a. kay , phys .
a * 73 * , 032306 ( 2006 ) .
j. pemberton - ross , a. kay , and s. g. schirmer , phys .
a * 82 * , 042322 ( 2010 ) .
n. margolus and l. b. levitin , physica d : nonlinear phenomena * 120 * , 188 - 195 ( 1998 ) .
a. wjcik , t. uczak , p. kurzyski , a. grudka , t. gdala , and m. bednarska , phys .
a * 75 * , 022330 ( 2007 ) .
t. brougham , g. m. nikolopoulos , i. jex , phys .
a * 80 * , 052325 ( 2009 ) .
t. sander , http://www.math.tu-clausthal.de/studium/vorlesungen/ws0506/hsgraphenth/graph_ev_intro.pdf[an introduction to graph eigenvalues and eigenvectors " ] .
r. merris , linear algebra appl .
* 199 * , 381 ( 1994 ) .
r. merris , linear algebra appl .
* 197 * , 143 ( 1994 ) .
b. mohar and s. poljak , in combinatorial and graph - theoretical problems in linear algebra , r. a. brualdi , s. friedland , and v. klee , eds .
i m a volumes in mathematics and its applications , vol .
springer - verlag , 107 ( 1993 ) .
r. j. angeles - canul , r. norton , m. opperman , c. paribello , m. russell , and c. tamon , quant . inform .
* 10 * , 325 ( 2010 ) .
f. harary and a. j. schwenk , in graphs and combinatorics ( springer berlin heidelberg , 1974 ) * 46 * 45 . | perfect transfer of a quantum state through a one - dimensional chain is now well understood , allowing one not only to decide whether a fixed hamiltonian achieves perfect transfer , but to design a suitable one .
we are particularly interested in being able to design , or understand the limitations imposed upon , hamiltonians subject to various naturally arising constraints such as a limited coupling topology with low connectivity ( specified by a graph ) and type of interaction . in this paper , we characterise the necessary and sufficient conditions for transfer through a network , and describe some natural consequences such as the impossibility of routing between many different recipients for a large class of hamiltonians , and the limitations on transfer rate .
we also consider some of the trade - offs that arise in uniformly coupled networks ( both heisenberg and @xmath0 models ) between transfer distance and the size of the network as a consequence of the derived conditions .
the task of state transfer was introduced @xcite with the intention of reducing the control required to communicate between distant qubits in a quantum computer .
the perfect action is defined by starting with a single qubit state @xmath1 on some input node , @xmath2 , and @xmath3 is the state of the rest of the many - qubit system , and requiring that after evolution for some time @xmath4 under a fixed hamiltonian @xmath5 , the output state @xmath6 is produced , thereby transmitting the input qubit to another site , @xmath7 .
early attempts concentrated on a one - dimensional geometry .
transferring states perfectly by such a scheme requires the precise tuning of coupling strengths @xcite , and the necessary and sufficient conditions for the transfer are now well understood @xcite , allowing us both to recognise and design @xcite perfect transfer chains .
there are , however , a number of limitations that one might like to overcome . in the ideal case , we would have a uniformly coupled system which perfectly achieves high rates of transfer over large distances .
it would have the ability to route states to different recipients ( which is an important feature in increasing the connectivity within a quantum computer ) , and it would be less susceptible to static defects ( anderson localisation ) @xcite or dynamical fluctuations @xcite than the one dimensional systems . it has been proposed that spin networks could enable all of these properties , and certainly by allowing some small degree of control , they do @xcite . it is expected that in higher dimensional geometries , the effects of anderson localisation are diminished . also , if one allows a completely connected network , routing is possible between all the different nodes @xcite .
equally , perfect transfer is possible over long distances in a uniformly coupled network ( such as the hypercube , @xcite ) , although the best known scaling of the transfer distance with the number of vertices , @xmath8 , is only @xmath9 .
unfortunately , while we know that networks exhibit a lot of potential , our level of understanding is still far from that of chains in spite of efforts by godsil , severini and coauthors in proving some necessary conditions @xcite .
the purpose of this paper is to redress the balance by deriving necessary and sufficient conditions which will allow us to readily recognise perfect transfer networks i.e. the setting that we envisage is an experimentalist with a range of hamiltonians available to them that they can implement , and they want to know if perfect transfer can be realised .
we can try and predict what some of the properties of such a hamiltonian might be . for instance , the whole point of introducing state transfer @xcite was to reduce the difficulty of interacting distant sites with no direct coupling ( for which transfer is trivial ) , so we impose the fact that there is likely to be an underlying geometry of feasible couplings , and that the type of these couplings is likely to be limited . as a consequence , we prove the impossibility of routing within the single excitation subspace of a broad class of systems ( those whose hamiltonians are real ) , and bound the maximum transfer rate for excitation preserving hamiltonians . for uniformly coupled networks ,
we give the first upper bound on the transfer distance of a graph as a function of the number of vertices .
while the task of state transfer was proposed as a solution to a practical problem in quantum computers , it has provided a powerful technique for constructing other hamiltonian - driven evolutions , and has been used to understand adversarial hamiltonian perturbations @xcite , quantum computation @xcite etc . by moving beyond the study of chains to more general networks , we may be able to further extend the utility of these constructions . |
throughout the universe we observe powerful engines which accelerate particles to immense energies .
the precise details of how these engines function are still poorly understood , but polarization measurements of the high energy radiation are pivotal to gain a deeper insight into the physical environment in these systems .
+ highly ordered geometries are generally required to produce a net polarization signal in the emission from a source , and strong magnetic fields are usually thought to play a leading role in the emission processes .
the strength and direction of a polarization signal will lead back to an understanding of the physical mechanisms and their geometry at the emission site .
there have been few attempts at measuring hard x - ray polarization since measurements are hampered by large backgrounds and systematic effects within the detectors .
however , the advent of modern fast computing clusters has made large scale simulation of an instrument s response now possible . by combining instrument data and results from detailed monte - carlo mass - model simulations using geant4
, it is possible to put constraints on the polarization characteristics of the hard x - ray flux emitted by a source . using the compton scattered events in the spi instrument on _ integral _ ( 130 - 8000 kev ) ,
constraints have been already put on the percentage polarization in the prompt gamma - ray flux of gamma - ray bursts ( grbs ; ) and the crab pulsar @xcite .
the latter showing a remarkable alignment to the rotational axis of the spinning neutron star .
+ however , improvements on the methods and techniques employed in the search for polarization could be achieved .
this paper discusses new developments that have been made in the mass - model and in the analysis procedure for the spi instrument .
the signature of polarization in the source signal can be found through the interaction process of photons within detectors . + in the hard x - ray regime ,
the dominant process is compton scattering . in the case of a linearly polarized flux ,
the azimuthal angle @xmath2 of the scattered photon is no longer isotropic .
the probability for the photon to be scattered at a polar angle @xmath3 and an azimuthal angle @xmath2 is given by the klein - nishina differential cross - section @xmath4,\ ] ] where @xmath5 is the classical electron radius , @xmath6 is the ratio between the scattered and incident photon energies , and @xmath2 is the azimuthal scatter angle defined as the angle to the polarization unit vector .
an azimuthal anisotropy is observed in the scattering direction , producing a greater number of scatters in a direction perpendicular to the emission polarization angle ( p.a .
; see figure [ kn_plot ] ) .
this formula shows that as the energy increases the angular distribution related to the polarization becomes more isotropic .
+ using a suitable arrangement of detector elements , this anisotropy can be used to determine the direction and degree of polarization of any observed emission .
spi is a coded - aperture telescope on board the _ integral _ observatory operating between 20 kev and 8 mev .
the description of the instrument and its performances can be found in and .
spi consists of 19 hexagonal germanium detectors measuring 5.6 cm flat - to - flat and 7 cm deep , tessellated into a hexagonal shape ( see figure [ ge ] ) .
the center - to - center distance between detectors is 6 cm . originally designed for fine spectroscopy
, spi can be used as a compton polarimeter .
the measurement of polarization using compton events relies on the positions of the initial and final photon interaction in the detectors . in spi ,
there is no positional information available inside a detector so only the multiple events where the photon deposits its energy in more than one detector can be used .
this kind of event is identified by energy deposits occurring within a coincidence window of 350 ns and not detected by the anti - coincidence system .
the effective area of spi for multiple events is reported in and range from 10 @xmath7 to 50 @xmath7 between 150 kev and 1 mev . since _ _
integral__s launch , four detectors have failed : detector 2 ( 2003 ) , detector 17 ( 2004 ) , detector 5 ( 2009 ) , and detector 1 ( 2010 ) .
these failures have resulted in a decrease of the effective area of the instrument down to @xmath8 of the original area for multiple events .
+ besides its effective area , the quality of a polarimeter depends on the sampling of the compton scattered events distribution ( figure [ kn_plot ] ) by the detectors .
an usual figure of merit for polarimeters is the @xmath9-factor , defined as the ratio of the parallel and perpendicular components of the scattered distribution .
the @xmath9-factor , which is energy dependent , has been calculated for spi to be @xmath10 over the energy range of 0.1 - 1 mev @xcite .
+ when searching for polarization in a source emission , any scattered distribution on the detector plane will be modulated by the coded mask shadow and the dead detectors .
these effects can be mistaken for a polarization signal if they are not properly accounted for . instead of looking for an anisotropy in the scattered distribution ,
a much simpler method is to simulate the detector response for a polarized flux and compare it to the measured data . in this way ,
polarization analysis with spi is perform in the same way than imaging but using the instrument response with one additional dimension : the p.a .
thus , for each spi observation we need simulations for different p.a.s in the source emission .
the presented simulations are based on a model recoded from the geant3 timm ( for the integral mass model ; ) into the geant4 framework .
geant4 @xcite uses object orientated c++ language and includes extra physics such as polarization and material activation . the geant4 software package has been tested against calibration data from a prototype detector array @xcite and reliably reproducing the instrument response to a polarized beam . in the new model of _
integral _ , the spi and jem - x detectors have been both fully modeled .
however , at the current time , the ibis detector is only described with an approximate geometry ( figure [ g4model ] ) .
this is not a problem for our polarimetry studies , since the work revolves around the spi instrument .
masses situated away from the spi instrument will have little or no impact on the results .
conversely , the spi instrument has been modeled with a high degree of accuracy .
it comprises the active veto , the ge detector elements inside al capsules , surrounded by the be housing , and the mask assembly .
the geant4 incarnation of the _ integral _ model was specifically written to assess the background in the spi instrument .
this means that several components of the geometry and particle generation processes have not been considered but could become important when looking at the emission from an astronomical object .
for example , the exact geometry of the coded mask and the ability to generate photons with a given spectral shape had to be implemented .
+ a great deal of effort has been spent on updating the simulation to produce a much more accurate model of the spi instrument , including several missing components in the geometry . in the current status ,
our simulations take into account the effects of non - functioning ( dead ) detectors , off - axis illumination and mask shadowing .
compared to the model version used by @xcite , we have included the degraded transparency of the central mask pixel which is 66% at 130 kev and 80% at 600 kev .
we have also modified the anti - coincidence system .
in fact , the most important part surrounding the detectors was not active in the code and we introduced the anti - coincidence thresholds ( @xmath11100 kev ) .
these modifications have an important impact on the number of events in the outer detectors and drastically change the instrument response .
the individual germanium crystal dimensions have been slightly reconsidered ( resulting in a few @xmath12 smaller volumes ) modifying the instrument sensitivity . in particular ,
the misalignment of the spi telescope axis relatively to the _ integral _ satellite pointing axis has been corrected , which corresponds to a systematic of @xmath13 in the pointing accuracy .
this correction is essential for the simulation of the instrument response because it modifies the mask shadowing and thus , the event distribution over the detectors ( @xmath14 of the events ) .
these important changes in the instrument simulation lead to a good agreement with the spi standard response ( see section [ validation ] ) .
+ the polarimetry capability of the spi instrument is based on the compton efficiency of the detector plane , i.e. , the non - diagonal terms of the response matrix ( @xmath15 of the total efficiency ) . this geant4 model
allow us to carefully evaluate them , through their dependence on the p.a .
of the incident radiation .
it provides a robust technique , using the whole instrument response and thus giving the best sensitivity to the polarization characteristics .
the simulated source is characterized by its direction in equatorial coordinates ( r.a .
, decl . ) , an energy spectrum , and a p.a .
this angle corresponds to the position angle of the electric vector measured from north to east on the sky . for each pointing ,
simulations are carried out at 18 p.a.s ( 0@xmath1 - 170@xmath1 in @xmath16 steps ) and 1 unpolarized .
only the p.a.s between @xmath17 and @xmath18 need be modeled due to the @xmath19 symmetry of the compton scattering process .
+ to properly simulate the photons distribution , the data is pre - analyzed and fitted with a simple spectral shape .
this is performed with the classical spi spectral analysis using the xspec software @xcite .
the spectral models form the basis of a probability distribution for the fired photons so that a single photon may be simulated at any energy , but after millions of photons , the spectrum should resemble that of the observed data . in theory , it should be possible to simulate a flat input spectrum with a fixed number of photons in each energy bin and to normalize a posteriori the number of photons in each energy bin .
simulations would need to be carried out only once and spectral changes could be handled in the analysis .
however , this will needlessly simulate many photons at high energies and would drastically lengthen the running time of the simulations .
+ besides , a small difference in the input spectrum will not lead to a big effect in the simulation .
we have compared two simulations , with a power law slope of -3.0 and -1.5 , respectively .
this results in differences less than @xmath20 in the count rate distribution recorded in the detectors .
+ for each simulation , millions of photons are fired from the direction of the source .
their initial positions are randomly distributed on a @xmath21 cm surface , 500 cm away from the detector plane , so as to illuminate all the instrument .
this area has been chosen large enough to include the direct lighting from the source and the scattered events from the structure in the detected events . for a typical power law spectrum with a slope of -2.2 ,
50 million of photons are processed for each simulation resulting in @xmath22 single events and @xmath23 multiple events in the detector . for a source flux of 1 crab
observed during one pointing ( @xmath112 ks ) , this provides an oversampling of approximately 40 times .
+ for one photon fired ( event ) , all the energy deposits inside the detectors are stored if no deposit occurred in the anti - coincidence system .
+ the number of simulations needed makes this part of the analysis very time consuming .
one simulation takes @xmath24 hr to run on a single processor . to produce data corresponding to 200 spi exposures ( @xmath11400 ks ) , we need to run 3800 simulations ( @xmath25 )
using 32 10-core processors , the computing time is reduced from 950 to 3 days .
the original data produced by geant4 is a list of energy deposits within sensitive volumes .
these raw data require some selection and preparation before analysis . a low energy threshold of 20 kev is applied and the events occurring in dead detectors are removed .
these reductions affect the multiple events : some triple events will become doubles and some doubles will become singles . for the data analysis , we select the double events with energy deposits occurring in adjacent detectors .
there are 42 corresponding pseudo detectors for all the possible couples of adjacent detectors
. then the simulated data are energy binned and recorded in the same format as the spi data to simplify the analysis procedure .
+ for a single pointing , simulations are carried out for 18 p.a.s ( 100@xmath12 polarized ) and 1 unpolarized . however , the analysis requires simulated data for different polarization percentages for each angle .
these data are produced by mixing the polarized simulated distributions with the unpolarized one using the formula @xmath26 where @xmath27 represents the ( geant4 ) simulated data for a @xmath28 percent polarized flux at the angle @xmath2 , @xmath29 the @xmath30 polarized simulated data at angle @xmath2 , and @xmath31 the unpolarized simulated data .
the search for polarization in a source emission requires accurate simulations of the instrument response , any differences would lead to a spurious polarization signal .
figure [ 661 ] shows the spectral response of the geant4 simulation compared to the spi standard response matrix , for a mono - energetic source at 661 kev .
the single events are summed over all the detectors and the counts per 1 kev bin are plotted in fraction of the photo - peak . with the implemented mass model improvements , the two curves differ by only @xmath32 in counts per bin with compton to total counts ratio of @xmath33 and @xmath34 for the geant4 simulation and the spi response matrix , respectively .
this ratio is important as it drives the production of multiples events .
if the simulated detector geometry is too large or too heavy , this ratio would be smaller , i.e. , less multiple events and more single events will be produced .
figure [ pseudodetcounts ] shows the spatial distribution produced by the geant4 code compared to the measured spi response , for the same configuration .
compared to the spi response , the unpolarized theoretical curve ( blue ) differs by only @xmath14 in counts per pseudo - detector whereas the @xmath35 polarized one ( red ) differs by @xmath15 .
it means that the signature of a polarized signal is contained in the difference , i.e. , @xmath36 of the recorded counts . the spatial distribution is highly modulated by the mask pattern and the telescope pointing accuracy .
less obvious , the anti - coincidence system and especially its energy threshold has an important impact on the spatial distribution .
a too high threshold will miss multiple events , increasing the recorded double events in the outer pseudo - detectors .
in a single spi observation , the recorded flux is barely enough to produce an image or a spectrum of the source .
this means that for a more complicated analysis such as polarization measurement , many observations need to be summed together .
unfortunately , the instrument configuration changes along the time : the pointing axis is shifted by @xmath37 , every 30 - 40 minutes , while , on a longer timescale , detector failure must be accounted for together with other parameter ( i.e. , background ) evolution . for a given source ,
our study relies on simultaneous analysis of a great number of pointings .
considering one exposure , the recorded data are modeled as @xmath38 where @xmath39 is the source normalization , @xmath40 , the simulated count distribution , @xmath41 is the background normalization and @xmath42,the background spatial distribution .
the source model @xmath43 describes the number of counts for a science window @xmath44 , in the pseudo detector @xmath45 as a function of source polarization fraction , @xmath28 , and angle , @xmath2 .
the background spatial distribution is taken from an empty field observation temporally close to the observation . to be comparable with the data , the counts issued from the simulation are renormalized to the corresponding detector livetimes .
the @xmath39 and @xmath41 values are determined through a linear least squares resolution and the resulting @xmath0 value is stored . if the source and background flux can be considered constant on a long time scale , equation ( [ source_equ ] ) is solved for the corresponding set of science windows .
+ note that it is different from previous polarization studies made with the spi instrument , where the source and the background were not adjusted by a fitting procedure . for grbs ,
the background was estimated using appropriate time intervals before and after the source signal .
this is not feasible with most of the sources .
for the crab , @xcite considered a background with a constant pattern and its norm was calculated by subtracting the simulated counts from the data .
this is hazardous because it implies the perfect knowledge of the instrument sensitivity and the source strength during each science window . in our code , the background pattern is built every six months from relevant empty field observations and its normalization is allowed to vary on a timescale from one science window to a revolution . estimating the source and the background fluxes in the analysis procedure introduces additional degrees of freedom ( dof ) but provides a better determination of the source contribution and allows polarization measurements for any ( strong enough ) source .
the analysis process relies on the comparison between the recorded data and the theoretical ones .
it can be described by the following procedure . *
science windows with the source closer than @xmath46 from the pointing direction are selected .
this ensures a good illumination of the detectors .
* a quick standard analysis is run to filter the science window list from obvious data anomalies . *
the spi data are read into memory and the counts are summed over the selected energy range . *
the corresponding simulated data are read into memory , the counts are summed over the selected energy range and weighted by the appropriate detector livetimes . * if the source and the background fluxes can be considered as constant , the science windows are grouped per data sets ( with a common @xmath39 and @xmath41 parameters ) . * looping over the entire range of p.a.s @xmath2 ( 0@xmath1 - 170@xmath1 ) and polarization fractions ( p.f.s ) @xmath28 ( @xmath47 to @xmath30 ) , equation ( [ source_equ ] ) is solved simultaneously for each data set through a linear least - squares resolution . *
the corresponding @xmath0 are calculated , resulting in a @xmath48 @xmath0 map in the ( @xmath49 ) space ( see figure [ chi2_map ] ) . * using a cubic interpolation algorithm ,
the @xmath0 map is extended to @xmath50 . * the minimum @xmath0 is used as the indicator of the best parameters . *
a contour plot is displayed to visualize the confidence regions at the 1@xmath51 , 2@xmath51 , and 3@xmath51 levels ( see figure [ chi2_map ] ) .
some additional information are extracted from the @xmath0 map .
figure [ chi2_percent ] shows the p.f . giving the best fit for each simulated angle ( bottom ) and the associated @xmath0 ( top ) .
the best - fit solution appears in these curves at the lower @xmath0 value and at the maximum p.f .
, respectively .
+ note that only the p.f.s from @xmath52 to @xmath30 are physically correct .
however , the analysis is run over negative polarization percentages to find the best fits from a statistical point of view .
regarding the angular distribution predicted for polarized emissions ( from the differential compton cross - section ) , it is expected to have two mathematical solutions ( local minima ) : the first ( good ) one , with a positive fraction and the second , @xmath53 away , with a negative fraction ( understandable in terms of a pattern to be subtracted from a overestimated mean flux , see figure [ chi2_map ] ) .
this symmetry is also seen in the polarization modulation ( see figure [ chi2_percent ] ) . .
[ chi2_percent ] the errors can be estimated from the @xmath54 method depending on the number of parameters of interest @xcite .
the 1@xmath51 , 2@xmath51 , and 3@xmath51 confidence interval for one parameter of interest , without regard to the value of the second , is given by @xmath55 , @xmath56 , and @xmath57 .
the errors on the p.a . and p.f .
independently are calculated with these @xmath54 values .
the confidence region ( two dimensions ) , plotted above the @xmath0 map , is given by the @xmath0 criteria of two parameters of interest .
the @xmath58 , @xmath59 , and @xmath60 confidence regions are drawn at the level of @xmath61 , @xmath62 , and @xmath63 .
+ to be confident in our error analysis , we have verified this method by simulation using the bootstrap method .
it consists to run the same analysis thousands of times to quantify the errors .
we apply the technique where the simulated counts are randomized within their ( gaussian ) errors before to be re - analyzed .
the best fit values of the p.a . and
p.f . are stored for each trial . at the end , the pairs angle - fraction are spread around the real solution and form a normal distribution ( in two dimensions ) .
the analysis of this distribution provides the region ( pair angle - fraction ) or the interval ( angle or fraction ) containing @xmath58 , @xmath59 , and @xmath60 of the realizations .
we then identify on the @xmath0 map the @xmath54 which gives the same confidence levels .
we find that the @xmath58 , @xmath59 , and @xmath60 confidence intervals are obtained at the level of @xmath55 , @xmath56 , and @xmath57 , and the equivalent confidence regions at the level of @xmath61 , @xmath62 , and @xmath63 as expected .
the obvious candidate to search for polarization is the crab ( pulsar and nebula total emission ) since it is one of the brightest source in the hard x - ray domain and the pulsar emission is expected to be highly polarized . beyond that , a lot of polarization measurements exist at different wavelengths , giving a good set of results to compare with . + for this analysis , we used 239 spi science windows from revolutions 43 to 45 ( 2003 february ) , where the crab nebula was the main target .
the photons fired in the simulations follow a power law distribution with a spectral index of 2.2 ranging from 100 kev to 3 mev .
we have considered the detected counts in the energy range offering the best signal - to - noise ratio , 130 to 440 kev .
the low energy threshold is fixed to ensure that compton efficiency is known with enough precision .
the high energy one is related to the source signal - to - noise ratio decrease . +
the source and the background fluxes are considered as constant over one revolution ( @xmath113 days ) . the analysis process described in section [ process ] is applied , producing a @xmath0 map in the polarization angle - fraction space .
figures [ chi2_map ] and [ chi2_percent ] , previously quoted for illustration purposes , come from this crab observation analysis .
the corresponding contour plot with the confidence regions is highlighted in figure [ crab ] .
the lowest @xmath0 ( @xmath64 with dof=10031 ) indicates a p.a . of @xmath65 , measured from north to east , with a p.f . of @xmath66 ( errors given at 1@xmath51 ) .
+ the p.a .
, which indicates the direction of the photon electric vector , appears to align nicely with the spin axis of the crab pulsar , estimated at @xmath67 from x - rays imaging @xcite . extending the energy range to 130 - 650 kev , we find a p.f . of @xmath68 and @xmath69 in the energy range 130 - 1000
we did not find any significant dependence of the p.f . with the energy range .
this behavior is consistent with a synchrotron emission process , which is expected for the crab emission .
previous polarization measurements were made on the crab , using only the off - pulse emission which is expected to be much more polarized .
@xcite reported a p.a .
of @xmath70 in the optical domain and @xcite a p.a . of @xmath71 in the hard x - ray domain .
polarization measurements of the unpulsed emission has been already made with the spi instrument , based on a previous mass model version and a different analysis process by @xcite . with @xmath11600 science windows ( @xmath111.5 ms ) , covering the first three years of _ integral _ operations ( from revolution 43 to 422 in 2006 march ) , they found a p.a .
of @xmath72 with a p.f . of @xmath73 .
our work includes only 239 science windows ( 570 ks ) and consider the total ( pulsar and nebula ) emission , but we find a better constrained solution , showing that the mass model improvements and the analysis process have strongly reduced the systematics errors . to produce
a spurious polarization signal requires an effect able to favor one direction in the scattered events , i.e. , affecting more counts in some specific pseudo - detectors .
this kind of systematic error could come from the instrument itself or / and from the simulations .
polarimeters instruments are often designed to spin their detection plane during an observation so as to blend the potential detection artifacts .
spi is not spinning but its coded mask imaging capabilities require a specific observation strategy ( dithering ) during which the pointing direction shift around the source .
this means that the detectors , so the pseudo - detectors , lit by the source are not always the same .
moreover , when using many observations over a one year period , the detection plane azimuthal angle ( rotation around the pointing axis ) varies by 1 deg per day ( mean value ) .
these two characteristics drastically limit any instrumental systematic effect .
+ @xcite have investigated the potential intrinsic biases related to the detector with ground calibration data .
they could constrain them to be around 1@xmath12 , quite negligible for our study .
+ on the other side , we have to check that the simulations are free of any systematic effects . to estimate these systematics ,
we have applied our analysis procedure to the long distance imaging ground calibration data obtained at bruyres - le - chtel , france .
we have selected 10 exposures ( from 12 to 38 minutes ) performed with an unpolarized source of @xmath74cs at 125 m. we have produced the corresponding simulated observations with our model and selected the double events for which the total energy corresponds to the @xmath74cs line energy .
applying our analysis process we found a p.f .
@xmath75 whatever the angle .
this value is overestimated due to the uncertainties linked to the calibration setup such as the position and the finite distance of the radioactive source . as a second way
, we have used the standard spi response matrix which provides the response for a given source position at infinity . from this standard response matrix
, we have reproduced a set of simulated observations corresponding to the spi observations used in our crab analysis . as this standard response matrix does not include the effects of polarization in the compton scattering distribution ,
this simulated set is representative of observations of a non - polarized source .
we have analyzed these observations with the procedure described above and found a p.f .
@xmath76 whatever the angle .
consequently , this demonstrates that the imaging system of spi ( pixelated detector + coded mask + dithering ) and the use of a complete and validated instrument response ensure reliable polarization measurements with systematic effects smaller than @xmath77-@xmath78 .
we have presented specific analysis tools developed to study the polarization of emissions detected by the spi spectrometer , aboard the _ integral _ mission , in the hard x - ray domain ( 130 kev to 8 mev for the polarimetry studies ) .
polarization information being contained in the compton interaction parameters ( diffusion and azimuthal angles ) , we used the photons which undergo a compton interaction in one of the 19 germanium crystals with the diffused photons escaping to an adjacent crystal .
+ precise simulations using geant4 can provide the spi instrument response for these polarized events , for various p.a.s ( from @xmath17 to @xmath19 ) and p.f.s ( from @xmath52 to @xmath30 ) .
the observed data are compared with each of these simulations , through @xmath0 estimation in the same way than imaging is performed with spi data .
we produce a @xmath0 map in the two - dimensional polarization angle - fraction space and consider the lowest @xmath0 value to determine the most probable parameters of the incident flux .
+ we have first applied our method to @xmath11600 ks of observation on the crab and found immediately a p.a .
of @xmath65 , in perfect agreement with previously published value on the crab @xcite .
however , our analysis method is more precise than that presented in @xcite since we use an improved geant4 version of the spi/_integral _ mass model and develop more sophisticated algorithms for the background determination and source flux extraction . in particular
, we have included the spi alignment correction with the spacecraft axis to improve the source projection into the detector plane , a parameter essential for the simulations .
moreover , our analysis process adjusts the source and background contributions within the fitting procedure and permits their variation in time .
this allows us to better determine the source contribution and to apply the polarimetry analysis to other sources than strong grbs or constant sources .
+ in conclusion , the measurement of the polarization appears as a promising area in the hard x - ray domain , through the compton interaction process inside detection systems .
it requires a high statistic and thus strong sources .
nonetheless , pulsars , bright transients and grb are as many potential targets for this kind of studies , which bring crucial information , very complementary to the spectral and timing analyzes .
this promising and still unexplored field should provide key information on mechanisms at work inside the most powerful engines of the sky .
+ the _ integral _ spi project has been completed under the responsibility and leadership of cnes .
we are grateful to asi , cea , cnes , dlr , esa , inta , nasa , and ostc for support .
m. chauvin and d. j. clark gratefully acknowledge financial support provided by the cnes .
, g. , sowikowska , a. , kellner , s. , & steinle , h. 2005 , in aip conf .
, 801 , astrophysical sources of high energy particles and radiation , ed .
t. bulik , b. rudak , & g. madejski ( melville , ny : aip ) , 306 | we present recent improvements in polarization analysis with the _ integral _ spi data .
the spi detector plane consists of 19 independent ge crystals and can operate as a polarimeter .
the anisotropy characteristics of compton diffusions can provide information on the polarization parameters of the incident flux . by including the physics of the polarized compton process in the instrument simulation , we are able to determine the instrument response for a linearly polarized emission at any position angle .
we compare the observed data with the simulation sets by a minimum @xmath0 technique to determine the polarization parameters of the source ( angle and fraction ) .
we have tested our analysis procedure with crab nebula observations and find a position angle similar to those previously reported in the literature , with a comfortable significance . since the instrument response depends on the incident angle , each exposure in the spi data requires its own set of simulations , calculated for 18 polarization angles ( from 0@xmath1 to 170@xmath1 in steps of 10@xmath1 ) and unpolarized emission .
the analysis of a large amount of observations for a given source , required to obtain statistically significant results , represents a large amount of computing time , but it is the only way to access this complementary information in the hard x - ray regime .
indeed , major scientific advances are expected from such studies since the observational results will help to discriminate between the different models proposed for the high energy emission of compact objects like x - ray binaries and active galactic nuclei or gamma - ray bursts . |
most recent studies dealing with the magnetic structure of the solar corona above active regions use different force - free model approaches ( see recent review of * ? ? ?
* ) and base the modeling on photospheric vector magnetic field data from either the helioseismic and magnetic imager ( hmi ) on board the solar dynamics observatory ( sdo ) or the spectropolarimeter ( sp ) of the solar optical telescope ( sot ) on board the hinode spacecraft .
such model approaches are used to compensate the lack of routine direct measurements of the coronal magnetic field vector .
@xcite used sp data as lower boundary condition to an optimization approach to analyze the coronal magnetic field associated to a white light flare .
the modeling suggested that the flare originated from sheared and twisted field lines with low altitudes bridged by a set of higher magnetic field lines .
@xcite applied a mhd relaxation method based on sp data to investigate the buildup and release of magnetic twist and suspected the importance of the relative handedness of twisted field lines and the ambient field .
@xcite employed an optimization method based on hmi data to model the temporal evolution of the coronal field of an active region over five days .
the modeling displayed distinct stages of the build - up and release of magnetic energy and analyzed the association to changes in the magnetic field .
in a subsequent study , @xcite used the same method for a detailed analysis of the field topology during a series of eruptions observed from hmi . above the apexes of cusp - like loops observed in coronal images , the modeling result suggested the presence of a coronal null point .
comparisons of the outcome of _ different _ force - free model algorithms based on the _ same _ lower boundary conditions have been performed in the past too ( e.g. , * ? ? ?
* ; * ? ? ?
these studies revealed that order - of - magnitude estimates of these models based on large enough fields - of - view and high enough spatial resolution of the vector magnetic field data can be expected to be reliable .
the model outcome of the _ same _ reconstruction algorithm based on data from _ different _ instruments has been studied recently by @xcite , using data from hmi and the vector - spectromagnetograph ( vsm ) of the solis project @xcite .
they found agreements of the resulting force - free models in form of , e.g. , the relative amount of energy to be set free during an eruption but also found a considerable difference in the absolute model energy estimates .
in particular , they found the estimated energy content of the vsm model being about twice of that of the hmi model . in the present study we use vector magnetic field data of hmi and sp as an input to the same force - free model and
compare the results .
this is motivated due to the data of these two instruments being widely used recently and in the future expected to be frequently used as an input for the modeling of the coronal magnetic field .
we regard this as important in order to test the consistency of the model solutions and at the same time to give a feeling about the accuracy of the estimated physical quantities and magnetic field topology based on those models .
we also investigate the effect which a binning of the sp instrument data to a lower resolution has on the model outcome .
we , however , do not search for explanations of the differences of the hmi and sp inversion products itself , considering this as to be out of the scope of this study .
the hmi on board sdo @xcite obtains filtergrams at the photospheric fe 617.3 nm spectral line .
the full stokes vector is retrieved from filtergrams averaged over about 12 min and inverted using the milne - eddington ( me ) inversion algorithm of @xcite .
the 180@xmath0-azimuth ambiguity of the transverse field is resolved using a minimum energy algorithm @xcite and the resulting vector magnetograms have a plate scale of 0.5 arc - second . in its fast scan mode ,
sp as part of the sot @xcite on board the hinode spacecraft @xcite obeys a spatial resolution of @xmath10.32 arc - second .
it observes the fe spectral line couplet at 630.15 nm and 630.25 nm .
full stokes profiles are obtained with a spectral sampling of 2.25@xmath2 nm and a slit scan time of 1.6s .
the physical parameters from the full stokes profiles were obtained using the merlin me inversion algorithm @xcite .
the 180@xmath0-azimuth ambiguity is resolved in the same way as for hmi data . to perform a study as described above simultaneous observations from hmi and sp are required .
both , hmi and sp vector magnetic field data are available to analyze the magnetic structure of active region 11382 on 2011 december 22 .
sp scanned this active region from 04:46 ut to 05:29 ut ( i.e. , scanned @xmath1@xmath3/min from solar east to west ) .
the hmi vector map used for this study was retrieved at @xmath4@xmath505:00 ut , approximately at half of the scanning time of sp . given average photospheric conditions ( @xmath6@xmath1@xmath7 , @xmath8@xmath1@xmath9@xmath10 ) the alfvn speed is @xmath11@xmath110kms@xmath12 .
the alfvn travel time and thus the evolution of the photospheric magnetic field over a characteristic distance in the present study ( @xmath1@xmath13 ) is @xmath11h .
therefore , we assume that the temporal changes of the photospheric field over @xmath120 min , during which sp scanned before and after the time when hmi recorded , are negligible . regarding the longitudinal magnetic field we assume this as justified since , e.g. , also @xcite found that the average ratio of the longitudinal magnetic flux densities measured by sp and soho / mdi @xcite was not strongly influenced by the evolution of the photospheric magnetic field during the scanning time of sp .
the magnetic field vectors are transformed to heliographic coordinates @xcite .
we correct the sp data set for the effect of differential rotation where we use @xmath4 as reference time . given present computational capabilities high - resolution data is sometimes binned to a lower resolution in order to allow for near real - time magnetic field modeling and/or computational domains of feasible dimensions ( see e.g. , * ? ? ?
thus , in the course of co - alignment of the hmi and sp data , we also bin the original - resolution sp ( sp@xmath14 ) data to the resolution of hmi which involves a 2d linear interpolation . the binned sp data is hereafter denoted as sp@xmath15 data .
the field - of - view ( fov ) used to study active region 11382 is mainly determined by the area covered by the sp scan ( @xmath1140@xmath1685@xmath17 centered around s19w07 ; see figure [ fig : fig1 ] ) . within the hmi data which covers roughly @xmath1400@xmath16250@xmath17
, we define a window equally sized as the fov of sp and calculate the cross - correlation between its vertical field component and that of the sp data . by shifting the position of the hmi window we search for the highest cross - correlation to find the corresponding hmi sub - field
. data .
the gray - scale background reflects the vertical magnetic field , @xmath18 .
black / white arrows indicate the direction of the horizontal magnetic field , originating from positive / negative polarity where @xmath19@xmath20@xmath21 .
the dashed rectangle outlines the quiet - sun region used to calculate the 2@xmath22 uncertainty of the vertical and horizontal magnetic field .
the black arrow just below the dashed rectangle indicates the length of an arrow representing a horizontal field magnitude of @xmath7.,title="fig : " ] ( -55,205)*(a ) * ( -55,107)*(b ) * the noise level of the hmi data is on the order of @xmath23/@xmath24 for the longitudinal / transverse magnetic field ( x. sun , private communication ) .
the average uncertainty for both the longitudinal and transverse field , estimated from the sp inversion error maps is @xmath25@xmath23 . besides seeming rather low especially for the transverse field these standard error estimates may not be reliable ( b. lites , private communication )
thus , we employ a consistent measure of the uncertainty level for the data from the two instruments , as described in the following . when investigating the properties of the hmi and sp data in the course of the force - free modeling , we only consider ( ) pixels with values of the vertical field , @xmath19 , _ and _ horizontal field , @xmath26 , above the respective 2@xmath22 uncertainty levels ( 2@xmath27 and 2@xmath28 , respectively ) or ( ) pixels with values of @xmath19@xmath292@xmath27 _ and _ @xmath26@xmath302@xmath28 , the latter in order not to disregard the strong vertical fields in the center of the active region .
the respective 2@xmath22 uncertainty levels are calculated for the quiet - sun area , outlined by the dashed rectangles in figure [ fig : fig1 ] , where we find : 2@[email protected]/6.7@xmath31 and 2@[email protected]/12.2@xmath31 for the hmi / sp@xmath15 data .
the uncertainty levels for the sp@xmath14 data are calculated from a quiet - sun region , equivalent to that outlined in figure [ fig : fig1 ] . here
we find 2@[email protected]@xmath31 and 2@[email protected]@xmath31 , i.e. slightly higher than what was found for the sp@xmath15 data .
the latter estimates , in fact , conform with the findings of @xcite who had estimated the measurement uncertainties of sp for the quiet - sun inter network field strengths and fluxes as @xmath3015@xmath31 .
( -255,140)*(a ) * ( -125,140)*(b ) * ( -255,58)*(c ) * ( -125,58)*(d )
* photospheric polarization signals originate from atmospheric layers which are known not to be force - free .
for instance , @xcite showed , using vector magnetic field measurements of an active - region magnetic field , that it can be considered as to be force - free above 0.4 mm above a photospheric level .
modeling the interchanging dominance of plasma and magnetic pressure , @xcite was able to estimate the height regime of dominating magnetic fields above a sunspot / plage region as @xmath10.8
200 mm . using high - resolution sp vector maps
, @xcite found that umbral and inner penumbral parts of sunspots may nearly be force - free but that sunspots as a whole at a photospheric level might not entirely be so .
therefore , the inferred hmi and sp magnetic vector maps are not force - free consistent and need to be preprocessed to achieve suitable force - free consistent boundary conditions @xcite . from the vertical component of the preprocessed field vector , a potential field
is calculated and used as start equilibrium and to prescribe the boundaries of the cubic computational domain .
the bottom boundary is replaced by the preprocessed vector field and the set of force - free equations for the nonlinear case in cartesian coordinates solved @xcite .
a boundary layer of @xmath1@xmath32 is introduced towards the lateral and top boundaries where the nonlinear force - free ( nlff ) solution drops to the prescribed boundary field . for our analysis
we discard this layer and only consider the inner ( physical ) @xmath1120@xmath1660@xmath1670@xmath33 domain and the according bottom boundary field . since this method involves the relaxation of the magnetic field not only inside the computational domain but also on its bottom boundary we compute a potential field from the relaxed lower boundary based on the fast - fourier method described by @xcite .
hereafter , we refer to the 3d nlff field model based on the hmi vector map as an input to as `` hmi model '' . similarly , the `` sp@xmath15 model '' and `` sp@xmath14 model '' result from using the binned and original - resolution sp vector magnetic field , respectively , as input data to our preprocessing and force - free reconstruction algorithms .
we compare the hmi and sp@xmath15 model in [ ss : mfaff][ss : topo ] and summarize the effects of binning on the model outcome in [ ss : eof_binning ] .
after performing the data preparation as described in [ ss : event_selection_and_data_set ] , we are able to investigate how the vertical and horizontal field components of the hmi and sp@xmath15 vector maps compare to each other and to check the force - free consistency .
we only consider data points where the criteria outlined in [ ss : uncertainty_estimation ] are fulfilled .
we find on overall stronger vertical fields in the sp@xmath15 than the hmi data ( especially for @xmath19@xmath34@xmath7 ; see figure [ fig : fig2]a ) and stronger horizontal field ( except for @xmath26@xmath35@xmath24 ; see figure [ fig : fig2]b ) .
this can be seen also when comparing the area - integrated unsigned vertical flux , @xmath36 , and the average horizontal field , @xmath37 : the hmi data hosts @xmath154% of @xmath38 and @xmath181% of @xmath39 ( see table [ tab : forces ] ) .
hmi and sp@xmath15 have a comparable sensitivity on long scales , @xmath40 , i.e. at small wave numbers , @xmath41 , especially on scales @xmath40@xmath34@xmath42 ( figure [ fig : fig2]c , d ) . with decreasing scale the amount of detected field increasingly differs : sp data show considerably stronger fields on smaller scales .
( -255,110)*(a ) * ( -125,110)*(b ) * ( -255,58)*(c ) * ( -125,58)*(d ) * [email protected]@[email protected] & @xmath36 & @xmath37 & & & + & @xmath43@xmath44@xmath45 $ ] & @xmath46 $ ] & + hmi & 1.323 & 18.4 & 0&13 & 0&08 & 0&29 + sp@xmath15 & 2.440 & 22.7 & 0&03 & 0&06 & 0&81 + sp@xmath14 & 2.510 & 21.5 & 0&03 & 0&06 & 0&84 + + hmi & 1.280 & 19.2 & 0&01 & 0&01 & 0&03 + sp@xmath15 & 2.341 & 30.2 & 0&02 & 0&01 & 0&02 + sp@xmath14 & 2.427 & 30.1 & 0&02 & 0&02 & 0&02 + + hmi & 1.279 & 20.9 & 0&01 & 0&02 & 0&01 + sp@xmath15 & 2.338 & 33.5 & 0&01 & 0&03 & 0&01 + sp@xmath14 & 2.364 & 30.3 & 0&01 & 0&03 & 0&08 + listed are the total unsigned vertical flux , @xmath36 , the average horizontal field , @xmath37 as well as the net lorentz force components ( @xmath47 , @xmath48 and @xmath49 ) normalized to the magnetic pressure force , @xmath50 . only pixels with values
@xmath19@xmath292@xmath27 _ and _ @xmath26,@xmath292@xmath28 or @xmath19@xmath292@xmath27 _ and _ @xmath26@xmath302@xmath28 are taken into account .
a necessary condition for a magnetic field to be force - free is that the components of the net lorentz force are considerably smaller than a characteristic magnitude of the total lorentz force in case of a non force - free magnetic field .
the latter can be approximated by the magnetic pressure , @xmath51 , on the lower boundary @xcite .
the ratio @xmath52 with @xmath53@xmath5@xmath54 in table [ tab : forces ] shows that this conditions are met only to a certain degree .
the ratios found here agree with the values found by , e.g. , @xcite and @xcite and we also find @xmath49@xmath29[@xmath47 , @xmath48 ] .
preprocessing , however , certainly improves the force - freeness as it smooths the vertical field and , additionally , alters the horizontal field in order to minimize the net force and torque and to gain boundary conditions compatible with the force - free assumption ( showing ratios of clearly smaller than unity ) .
the effect of smoothing can be seen when comparing the spectral power of the raw and preprocessed hmi and sp@xmath15 data ( figure [ fig : fig3 ] ) : the signal on shorter scales ( i.e. , for large @xmath41 ) is reduced . on longer scales , the power of the vertical field remains the same ( see figure [ fig : fig3]a , c ) but that of the horizontal field is enhanced ( more pronounced for the sp@xmath15 data ; see figure [ fig : fig3]b , d ) .
the preprocessing leads to a slight reduction of @xmath36 : @xmath56 is reduced by @xmath13% and @xmath57 by @xmath14% .
the preprocessing also leads to an enhancement of @xmath37 : @xmath58 is enhanced by @xmath14% and @xmath39 is enhanced by @xmath130% ( see table [ tab : forces ] ) .
the preprocessed hmi data hosts @xmath155% of @xmath57 and @xmath164% of @xmath39 of the preprocessed sp@xmath15 data . in summary
, the preprocessing only slightly reduces the difference of the vertical unsigned flux between hmi and sp@xmath15 ( which is @xmath145% before as well as after preprocessing ) but enhances the difference of the average horizontal field , @xmath59 , ( from @xmath59@xmath119% _ before _ to @xmath59@xmath136% _ after _ preprocessing ) .
these preprocessed hmi and sp@xmath15 data are used as lower boundary condition for the nlff reconstruction . as mentioned above , our nlff modeling algorithm also relaxes the magnetic field on the bottom boundary of a cubic computational domain .
thus , besides during the preprocessing , the magnetic field vector on the lower boundary is altered also while iteratively seeking for the force- and divergence - free field solution in the volume .
it s modification to the preprocessed lower boundary data , as listed in table [ tab : forces ] , is that @xmath36 is reduced by @xmath350.1% and @xmath37 is increased by @xmath110% .
thus , the hmi nlff lower boundary hosts @xmath155% of @xmath57 and @xmath162% of @xmath39 of the sp@xmath15 nlff lower boundary data , comparable to the ratio we found for the preprocessed data . from the 3d model fields , we can estimate the magnetic energy content of the potential field ( @xmath60 ) and of the nlff field ( total energy ; @xmath61 ) .
an upper limit for the energy which can be released ( excess energy ) is given by @xmath62@xmath5@xmath61@xmath63@xmath60 .
we can estimate the statistical accuracy of our volume - integrated energy estimates by adding different artificial noise models to the hmi and sp magnetograms , consecutive application of the preprocessing and extrapolation algorithms and comparison of the resulting energy values .
this yields a statistical error of @xmath11% for both @xmath60 and @xmath61 and @xmath110% for @xmath62 .
the estimated _ absolute _ potential energy of the hmi model is @xmath148% of that of the sp@xmath15 model ( see table [ tab : energies ] ) .
this is a direct consequence of the hmi nlff bottom boundary hosting only @xmath155% of the unsigned vertical flux of the sp@xmath15 bottom boundary ( see table [ tab : forces ] ) .
a similar trend is found for the _ absolute _ total and excess energy of the hmi model .
however , the _ relative _ excess energy is about 20% of the total energy in both models ( given by the ratio @xmath64 in table [ tab : energies ] ) .
an excess energy on the order of @xmath65
@xmath66 is assumed to be sufficient for powering c - class flaring , which was actually observed for the active region analyzed here on the days before .
similar values related to c - class flaring activity were found by , e.g. , @xcite and lately by @xcite .
lcccc & @xmath61 & @xmath60 & @xmath62 & @xmath67 + & & + hmi & 3.45 & 2.80 & 0.65 & 0.19 + sp@xmath15 & 7.44 & 5.80 & 1.64 & 0.22 + sp@xmath14 & 7.31 & 5.85 & 1.46 & 0.20 + given are the total , potential and excess magnetic energy of the 3d magnetic model fields , listed as @xmath61 , @xmath60 and @xmath62 , respectively .
the ratio @xmath64 gives the relative amount of excess energy .
the statistical error of these estimates is @xmath11% for @xmath60 and @xmath61 and @xmath110% for @xmath62 .
( -435,220)*(a ) * ( -285,220)*(c ) * ( -135,220)*(e ) * ( -435,115)*(b ) * ( -285,115)*(d ) * ( -135,115)*(f ) * besides comparing a volume - integrated quantity like the magnetic energy , we are also interested how the modeled magnetic field configurations compare to each other .
to do so and to ensure that we are looking at the same topological structure we look for regions of strong gradients in the magnetic connectivity .
they are thought of being linked to the creation of strong current concentrations in the solar corona and believed to represent the footprint of quasi - separatrix layers @xcite .
we quantify the magnetic connectivity following @xcite and calculate the squashing degree @xmath68 at a height of @xmath15@xmath69 above the nlff lower boundary ( figure [ fig : fig4]a , b ) .
the squashing degree quantifies the eccentricity of an elliptical cross - section of a flux tube into which a flux tube of initially circular cross - section is transformed .
wherever @xmath68 is large , the magnetic field connectivity changes drastically over short distances .
according to this pattern , we choose a region of interest ( roi ) around a clearly distinguishable pattern of high values of @xmath68 ( rectangular outline in figure [ fig : fig4]a , b ) .
though clearly visible in both models , the @xmath68-ridge appears more diffuse in the sp@xmath15 model and its location differs up to @xmath70@xmath25@xmath42 for a given @xmath71 .
model in the range @xmath72@xmath30@xmath73 .
the vertical magnetic field is shown as gray - scale background .
black and white contours are drawn at [ @xmath7450@xmath31,@xmath74100@xmath31 ] and [ @xmath6350@xmath31,@xmath63100@xmath31 ] , respectively .
the locations where field lines originate from regions of @xmath68@xmath29100 within the roi ( rectangular outline ) re - enter the lower boundary are color - coded according to the vertical magnetic field there.,title="fig : " ] ( -140,225)*(a ) * ( -140,115)*(b ) * , to their apex height ( @xmath76 ) for the ( a ) hmi and ( b ) sp@xmath15 model .
field lines are calculated starting from every pixel location in the range @xmath77@xmath30@xmath73 where @xmath19@xmath292@xmath27 _ and _ @xmath26,@xmath292@xmath28 or where @xmath19@xmath292@xmath27 _ and _ @xmath26@xmath302@xmath28 and the color - code reflects the value of the absolute vertical field there.,title="fig : " ] ( -130,35)*(a ) * ( -24,35)*(b ) * ( -255,140)*(a ) * ( -125,140)*(b ) * ( -255,58)*(c ) * ( -125,58)*(d ) * in total , 92/83 magnetic field lines in the hmi / sp@xmath15 model ( ) start from locations where @xmath68@xmath29100 within the roi and ( ) connect back to the lower boundary ( see figure [ fig : fig4]c - f ) .
they qualitatively outline two neighboring connectivity domains which connect the negative polarity region to its neighboring positive polarity surrounding .
the locations where the field lines connect back to the lower boundary are shown color - coded based on the local vertical field in figure [ fig : fig5 ] .
we assume that the field lines we calculated represent thin flux tubes . for each flux tube , we choose its cross section at the location from where we started the field line calculation , @xmath78 , as the size of one pixel ( i.e. , @xmath78@xmath1360@xmath79@xmath80 ) .
furthermore , we assume the vertical field there determines the flux of the flux tube .
summation over the all considered thin flux tubes gives an estimate of the total absolute shared flux for the hmi / sp@xmath15 model , where we find @xmath81@xmath5@xmath82/@xmath83@xmath16@xmath84 .
we find a lower value of connected flux in the hmi model though more closed magnetic field lines are considered due to our selection criterion .
however , comparable is the _ relative _ amount of connected flux linked by these field lines : @xmath81 comprises @xmath11% of the unsigned vertical nlff lower boundary flux as listed in table [ tab : forces ] .
we also recognize differences in the connectivity pattern of the field lines as shown in figure [ fig : fig5 ] .
numerous sp@xmath15 model field lines connect to the weak - field regions ( @xmath19@xmath35@xmath85 ) between the two major positive polarity patches centered around ( @xmath86@xmath1@xmath87 , @xmath71@xmath1@xmath88 ) or to the weak - field surrounding at ( @xmath89@xmath35@xmath86 , @xmath32@xmath35@xmath71 ) .
however , none of the selected field lines of the hmi model does so , instead they re - enter the nlff lower boundary at locations of strong vertical fields ( @xmath19@xmath34@xmath85 ) . from figure
[ fig : fig4]c - f one recognizes that a larger number of field lines connects to the positive polarity at ( @xmath86,@xmath71)@xmath1(@xmath90 , @xmath42 ) and the highest field lines of the sp@xmath15 model reach up to greater heights than those in the hmi model . to investigate
if the latter represents a general trend or is biased due to our restrictive selection of start locations for field line calculation , we consider all field lines starting from a pixel location on the nlff bottom boundary in the range @xmath77@xmath30@xmath73 where @xmath19@xmath292@xmath27 _ and _ @xmath26,@xmath292@xmath28 or @xmath19@xmath292@xmath27 _ and _ @xmath26@xmath302@xmath28 .
we then find very similar relationships of the length of the calculated field lines , @xmath91 , to the height of their apex , @xmath76 for both models ( see figure [ fig : fig6 ] ) : the relation is well defined and to a first approximation linear ( see also , e.g. * ? ? ? * ) .
the sp@xmath15 model field lines follow a steeper distribution , i.e. , seem to be on overall higher .
it appears that the field lines carrying the strongest vertical fluxes are neither the shortest ones nor the longest ones .
instead , it seems that in both models , field lines with a length of 20@xmath35@xmath91@xmath35100@xmath69 and an apex height of 5@xmath35@xmath76@xmath3535@xmath69 carry most magnetic flux .
the longest and highest closed field lines in the hmi model are found to be @xmath1@xmath92 and @xmath93 , respectively .
this is lower than the values found for the longest / highest closed field lines of the sp@xmath15 model ( @xmath1@xmath94/@xmath95 ) .
it is not surprising that the sp@xmath15 model field lines tend to be longer and reaching higher up in the model atmosphere when looking at the magnetic field distribution on the lower boundary .
there , we find an average of @[email protected]/0.9 for the hmi / sp@xmath15 model , i.e. , the sp@xmath15 model field lines are on average more vertical .
( -435,115)*(a ) * ( -285,115)*(b ) * ( -135,115)*(c ) * model in the range @xmath72@xmath30@xmath73
. the vertical magnetic field is shown as gray - scale background .
black and white contours are drawn at [ @xmath7450@xmath31,@xmath74100@xmath31 ] and [ @xmath6350@xmath31,@xmath63100@xmath31 ] , respectively .
the locations where field lines originate from regions of @xmath68@xmath29100 within the roi ( rectangular outline ) re - enter the lower boundary are color - coded according to the vertical magnetic field there . ]
the binning of the original - resolution sp data ( sp@xmath14 ) to the resolution of hmi involves a 2d interpolation and the resulting changes are discussed in the following . as before , we only consider data values fulfilling the criteria outlined in
[ ss : uncertainty_estimation ] .
the binning causes a decrease of @xmath36 by @xmath13% ( see table [ tab : forces ] ) .
it also causes an increase of @xmath97 by @xmath16% .
the force - freeness remains basically the same , as do the relative occurrence of the vertical / horizontal field ( figure [ fig : fig7]a / b ) and the respective power distributions ( figure [ fig : fig7]c / d ) remain similar . now it is also evident that the binning is only to a minor degree responsible for the differences between the sp@xmath15 and hmi data .
this agrees with , e.g. , @xcite who found that the scaling the data to a lower resolution does not significantly alter the results of their particular analysis .
the hmi data hosts @xmath153% of @xmath98 and @xmath186% of @xmath97 ( see table [ tab : forces ] ) . naively , one would suspect that the difference in @xmath36 arises from the different resolution limits of the two instruments , i.e. , that the magnetic field is partially on scales which hmi can not resolve . in context with the relative occurrence of @xmath36 discussed in @xcite , however ,
this can not be concluded since it was found that hmi vertical fields are on overall weaker than those of vsm data ( which with a plate - scale of @xmath11 arc - second has a _ lower _ resolution than hmi ) .
therefore , the different occurrence rates must have reasons besides the different resolution limit of the instruments .
preprocessing the sp@xmath14 data leads to a decrease of @xmath98 by @xmath13% and and increase of @xmath97 by @xmath140% , which is comparable to the changes of the sp@xmath15 data due to preprocessing .
the modification to the sp@xmath14 lower boundary data during solving for the force- and divergence - free field , as listed in table [ tab : forces ] , is that @xmath98 decreases by @xmath13% and @xmath97 increases by @xmath351% .
summarizing , we find a similar behavior of the modifications to the sp@xmath14 data during the force - free modeling and very similar values as we found for the sp@xmath15 data ( compare the values listed in table [ tab : forces ] ) .
the potential energy of the sp@xmath15 model is @xmath11% lower and the total energy is @xmath12% higher than that of the sp@xmath14 model ( see table [ tab : energies ] ) .
hence , when basing the analysis on the sp@xmath15 data , we find and enhancement / reduction of the absolute energy estimates @xmath61/@xmath60 on the order of the statistical error of the energy estimates itself .
the excess energy , @xmath62 , is enhanced by @xmath110% , i.e. , also on the order of the statistical error .
the _ relative _ amount of @xmath62 , however , remains approximately the same .
repeating the analysis of the magnetic connectivity within the sp@xmath14 model , we find that in total 228 magnetic field lines originating from locations of @xmath68@xmath29100 within the roi in figure [ fig : fig8]a connect back to the lower boundary ( see figure [ fig : fig8]b , c ) .
again , we assume that those represent thin flux tubes with a cross section of @xmath78@xmath1220@xmath79@xmath80 and assume the vertical field at footpoint from which we started the field line calculation determines its flux .
summation over the all considered thin flux tubes gives an estimated total shared flux of @xmath81@xmath5@xmath99@xmath16@xmath84 which comprises about 1% of the unsigned vertical flux of the nlff lower boundary of the sp@xmath14 model .
this is almost identical to what was found for the model based on the sp@xmath15 data , as is the connectivity pattern ( compare figures[fig : fig9 ] and [ fig : fig5]b ) . considering all field lines starting from a pixel location on the nlff bottom boundary in the range @xmath77@xmath30@xmath73 where @xmath19@xmath292@xmath27 _ and _ @xmath26,@xmath292@xmath28 or @xmath19@xmath292@xmath27 _ and _ @xmath26@xmath302@xmath28
, we find an identical relationship of @xmath91and @xmath76 ( figure [ fig : fig10 ] ) as was found for the sp@xmath15 model ( compare figure [ fig : fig6]b ) .
the sp@xmath14 model field is found to be even more vertical ( @[email protected] ) and , again , the field lines carrying the strongest vertical fluxes are those with an intermediate length and apex height . , to their apex height ( @xmath76 ) for the sp@xmath14 model .
field lines are calculated starting from every pixel location in the range @xmath77@xmath30@xmath73 where @xmath19@xmath292@xmath27 _ and _ @xmath26,@xmath292@xmath28 or where @xmath19@xmath292@xmath27 _ and _ @xmath26@xmath302@xmath28 .
the color - code reflects the value of the absolute vertical field there . ]
non - identical photospheric vector magnetic fields inferred from polarization measurements of different instruments used as input for nonlinear force - free ( nlff ) coronal magnetic field models directly translate to substantial differences in the model outcome . to quantify these , we performed force - free magnetic field modeling using active - region vector magnetic field data based on measurements of polarization signals by the sdo / hmi and hinode sot / sp . the possible causes of the differences of the data products itself , including the , e.g. , different intrinsic nature and sensitivity of the instruments , the temporal evolution of the photospheric magnetic field during the ongoing scanning times or the inversion techniques used to infer the magnetic field vector from the measured polarization signals , were out of the scope of this work .
our aim was to compare force - free model results based on data from these two instruments so we applied all data preparation and modeling routines in exactly the same way to both data sets .
force - free coronal magnetic field modeling is computationally expensive and high - resolution data is sometimes binned to a lower resolution in order to shorten the computational time .
we , therefore , binned the original - resolution sp ( sp@xmath14 ) data to the resolution of hmi which allowed us , besides ( ) to quantify the deviations of the force - free modeling outcome due to the usage of data of _ different _ instruments ( hmi and sp ) , also ( ) to investigate the effect of binning the input data to a lower resolution on the model outcome , by subsequent comparison of the models based on the binned sp ( sp@xmath15 ) data and sp@xmath14 data .
we did not intend to mimic the different spatial resolution of the instruments by binning of the vector data since , as pointed out by @xcite , any kind of binning does not account properly for resolution effects . instead , in practice , a binning of high - resolution data prior to the nlff modeling is meant to result in feasible computational dimensions and to allow for near real - time modeling .
we used hmi and sp vector maps of active region 11382 on 2011 december 22 and found considerably higher vertical magnetic flux and average horizontal field in the sp data .
this difference of the vertical flux was found to be much larger than the modifications to the data due to application of our force - free modeling routines . also
the modifications to the vertical flux due to binning of the sp@xmath14 data to the resolution of hmi were found to be small compared to the inequality of detected flux by the two instruments .
unequal estimates of the magnitude and orientation of longitudinal and transverse fields based on the inversion of circular and linear polarization signals measured by the two different instruments , yield differing estimates of the vertical and horizontal magnetic field in a local coordinate system .
the unequal amount of vertical magnetic flux ( the projected sp data hosting about two times the unsigned vertical flux of hmi ) then directly translates to a considerable discrepancy in the _ absolute _ estimates of the energy content of the considered coronal volume : models based on the sp data hold about twice as much energy as does the hmi model . by tracing magnetic field lines in the half - space above the model
lower boundaries their connectivity was investigated .
the sp and sp@xmath14 models revealed a closely matching photospheric footprint of two neighboring coronal connectivity domains .
a similar footprint was found in the hmi model but locally shifted by up to several@xmath69 .
the same applies to the locations where field lines re - enter the nlff lower boundary : while the location of the footpoints of the sp and sp@xmath14 models coincide , those of the hmi model are displaced by up to ten@xmath69 .
moreover , on overall , the sp and sp@xmath14 model fields tend to be more vertical than the hmi model field which is a direct consequence of the relative vertical and horizontal field distribution given by the instrument data
. however , the models also showed great similarities : _ relative _ estimates like the fraction of energy in excess over a potential field ( about 20% of the total energy content ) or the fraction of vertical flux shared by two neighboring connectivity domains ( about 1% of the total vertical flux of the active region ) agree very well .
also the overall field - line geometry was found to be comparable : the length of all closed field lines in the model volumes was found to be more or less linearly related to the apex height .
common to the model outcomes is also that the shortest and lowest as well as the longest and highest field lines carry least vertical flux . in conclusion
, caution is needed when analyzing the coronal magnetic field and its connectivity with the help of force - free magnetic field models based on the vector magnetic field products of different instruments , made available to the community .
_ relative _ estimates and the overall structure of the model magnetic fields might indeed be reliable while _ absolute _ estimates might only be so concerning their order of magnitude
. moreover , binning of the magnetic vector data to a lower resolution prior to the force - free modeling results only in little differences in the model outcome , small compared to the remarkable deviations when basing the modeling on data from the two different instruments , hmi and sp .
we thank the anonymous referee for careful consideration in order to improve our manuscript .
we would also like to thank x. sun for support with the hmi vector magnetic field data and b. inhester for insightful discussions .
acknowledges supported by dfg grant wi 3211/2 - 1 , t.w .
is funded by dlr grant 50 oc 0904 .
sdo data are courtesy of the nasa / sdo hmi science team .
hinode is a japanese mission developed and launched by isas / jaxa , with naoj as domestic partner and nasa and stfc ( uk ) as international partners .
it is operated by these agencies in co - operation with esa and nsc ( norway ) .
hinode sp inversions were conducted at ncar under the framework of the csac . | photospheric magnetic vector maps from two different instruments are used to model the nonlinear force - free coronal magnetic field above an active region .
we use vector maps inferred from polarization measurements of the solar dynamics observatory / helioseismic and magnetic imager ( hmi ) and the solar optical telescope spectropolarimeter ( sp ) aboard hinode .
besides basing our model calculations on hmi data , we use both , sp data of original resolution and scaled down to the resolution of hmi .
this allows us to compare the model results based on data from different instruments and to investigate how a binning of high - resolution data effects the model outcome .
the resulting 3d magnetic fields are compared in terms of magnetic energy content and magnetic topology .
we find stronger magnetic fields in the sp data , translating into a higher total magnetic energy of the sp models .
the net lorentz forces of the hmi and sp lower boundaries verify their force - free compatibility .
we find substantial differences in the absolute estimates of the magnetic field energy but similar relative estimates , e.g. , the fraction of excess energy and of the flux shared by distinct areas .
the location and extension of neighboring connectivity domains differs and the sp model fields tend to be higher and more vertical .
hence , conclusions about the magnetic connectivity based on force - free field models are to be drawn with caution .
we find that the deviations of the model solution when based on the lower - resolution sp data are small compared to the differences of the solutions based on data from different instruments . |
half a lifetime has passed since human astronauts directly explored another planetary body .
currently , two robotic geologists are exploring the surface of mars , one scientific probe has just landed on saturn s moon titan , and a number of orbiters are studying several of the planets and moons in our solar system .
soon , we will be sending more robotic explorers to mars , more science stations to orbit our planetary neighbors , and perhaps new human explorers to the moon .
all of these exploration systems , human or robotic , semi - autonomous or remote - operated , can benefit from enhancements in their capabilities of scientific autonomy .
human astronaut explorers can benefit from enhanced scientific autonomy
astronauts with `` augmented reality '' visors could explore more efficiently and perhaps make more discoveries than astronauts who exclusively rely upon guidance from earth - based scientists and engineers .
remote - operated robotic rovers can benefit from enhanced scientific autonomy this autonomy could either be on - board the rover or in computers on the earth .
such scientific autonomy can enhance the scientific productivity of these expensive missions , but only if the autonomy measures are well tested , and therefore ` trustable ' by the controllers on the earth .
along these lines , we have constructed a field - capable platform ( at relatively low cost ) in order to develop and to test computer - vision algorithms for space exploration here on the earth , prior to deployment in a space mission .
this platform uses a wearable computer and a video camera , as well as the human operator .
since this astrobiological exploration system is part human and part machine , we call our system the ` cyborg astrobiologist ' .
recently , we have reported our first results at a geological field site with the cyborg astrobiologist exploration system ( mcguire _ et al . _
2004b ) . in this paper
, we discuss our results at a second geological field site , near riba de santiuste , in northern guadalajara , spain ( see the maps in figures 2 & 3 ) .
this second geological field site offers a different type of imagery than was studied in the previous paper .
this new imagery resembles some aspects of the imagery that mars exploration rover ( mer ) opportunity is currently studying on mars , and the new imagery has greater astrobiological implications than the imagery studied in the previous mission of the cyborg astrobiologist .
we show that the basic exploration algorithms that we introduced in our first paper also function rather well at this second field site , despite the change in nature of the imagery . to give context to our work
, we would like to remind the reader of two discoveries by the apollo 15 and the apollo 17 astronauts on the moon in the early 1970 s , namely of the ` genesis rock ' and of ` orange soil ' ( compton 1989 ) .
the apollo 15 astronauts were astronauts trained to be geologists " , and one of their missions was to find ` ancient ' rocks , in order to obtain information as to how the moon was formed .
given the ` bias ' of their mission , and given significant ` scientific autonomy ' by mission control on the second day of their mission at the hadley - appenines landing site , astronauts scott & irwin found an anorthosite .
scott & irwin dubbed this specimen ` the genesis rock ' , because it possibly recorded the time after the molten moon s surface first cooled down , over 4 billion years ago .
they initially found the genesis rock to be interesting partly because their mission requirements and their focused geological training biased them to look for such crystalline types of rocks .
this biased - search during the apollo 15 mission on the moon is not unlike one mission requirement of the mer opportunity rover .
one main objective for sending opportunity to the meridiani planum site on mars was to understand the mars odyssey orbiter s suggestion of abundant coarse - grained gray hematite at meridiani .
so the initial focus and bias of opportunity s study of meridiani was on determining the source and nature of the hematite , and based upon this bias , opportunity and the earth - based geologists and engineers have been highly successful in writing ` the story ' about hematite at meridiani ( squyres _ et al .
_ 2004 ; chan _ et al .
_ 2004 ) .
the apollo 17 astronauts had one astronaut who had been trained to be a geologist ( cernan ) and one geologist who had trained to be an astronaut ( schmitt ) .
one of schmitt s responsibilities ( as geologist / astronaut ) was to try to see if he could observe phenomena on the moon which the previous astronaut / geologists had not seen .
another responsibility for both cernan and schmitt was to try to find signs of recent volcanism , which was one reason their taurus - littrow landing site had been chosen , since it contained numerous craters of possible volcanic origin . despite the bias of the mission towards volcanic geology , schmitt discovered an unusual ` orange soil ' on the rim of one crater , which he initially thought could be due to some unusual oxidative process .
this orange soil later turned out to be composed of small orangish glass - like beads , probably of volcanic origin .
as with the rest of schmitt s geological observing on the moon in which he based his decisions on taking samples on visually detectable differences or similarities " ( compton 1989 ) , schmitt ( schmitt 1987 ) found the orange soil interesting because it was different from anything else he had seen on the moon . from our point of view , since schmitt was an expert geologist , he was able to go much beyond the naive mission goals and biases of looking for fresh signs of volcanism , to discover something else ( orange soil ) , which at the time seemed unrelated to volcanism . in summary , we point to the apollo 15 & 17 missions and the mer opportunity mission as examples of the need to have both biased and unbiased techniques for scientific autonomy during space exploration missions .
it is rather difficult to make a system that can reliably detect signatures of interesting geological and biological phenomena ( with an imager or with a spectrometer ) in a general and biased manner . in this report
, we describe the further testing of our _ unbiased _ `` cyborg astrobiologist '' system .
in section 2 , we discuss the hardware and software of our system , followed by summaries & results of the geological expeditions to riba de santiuste in sections 3 & 4 , and finishing with more general discussion & conclusions in section 5 .
our ongoing effort in the area of autonomous recognition of scientific targets - of - opportunity for field geology and field astrobiology is maturing . to date , we have developed and field - tested a `` cyborg astrobiologist '' system ( mcguire _ et al . _ 2004a ; mcguire _ et al . _
2004b ) that now can : 1 @xmath0 use human mobility to maneuver to and within a geological site ; @xmath0 use a portable robotic camera system to obtain a mosaic of color images ; @xmath0 use a ` wearable ' computer to search in real - time for the most uncommon regions of these mosaic images ; @xmath0 use the robotic camera system to repoint at several of the most uncommon areas of the mosaic images , in order to obtain much more detailed information about these ` interesting ' uncommon areas ; @xmath0 choose one of the interesting areas in the panorama for closer approach ; and @xmath0 repeat the process as often as desired , sometimes retracing a step of geological approach . in the mars exploration workshop in madrid in november 2003 , we demonstrated some of the early capabilities of our ` cyborg ' geologist / astrobiologist system ( mcguire _ et al . _
we have been using this cyborg system as a platform to develop computer - vision algorithms for recognizing interesting geological and astrobiological features , and for testing these algorithms in the field here on the earth ( mcguire _ et al . _
2004b ) . the half - human / half - machine
` cyborg ' approach ( see figure 1 ) uses human locomotion for taking the computer vision - algorithms to the field for teaching and testing , using a wearable computer .
this is advantageous because we can therefore concentrate on developing the ` scientific ' aspects for autonomous discovery of features in computer imagery , as opposed to the more ` engineering ' aspects of using computer vision to guide the locomotion of a robot through treacherous terrain .
this means the development of the scientific vision system for the robot is effectively decoupled from the development of the locomotion system for the robot .
after the maturation of the computer - vision algorithms , we hope to transplant these algorithms from the cyborg computer to the on - board computer of a semi - autonomous robot that will be bound for mars or one of the interesting moons in our solar system .
these algorithms could also work in analyzing remote - sensing date from orbiter spacecraft . with human vision , a geologist , in an unbiased approach to an outcrop ( or scene ) :
1 @xmath0 firstly , tends to pay attention to those areas of a scene which are most unlike the other areas of the scene ; and then , @xmath0 secondly , attempts to find the relation between the different areas of the scene , in order to understand the geological history of the outcrop .
the first step in this prototypical thought process of a geologist was our motivation for inventing the concept of uncommon maps .
see mcguire _
( 2004b ) for an introduction to the concept of an uncommon map , and our implementation of it .
we have not yet attempted to solve the second step in this prototypical thought process of a geologist , but it is evident from the formulation of the second step , that human geologists do not immediately ignore the common areas of the scene . instead , human geologists catalog the common areas and put them in the back of their minds for `` higher - level analysis of the scene '' , or in other words , for determining explanations for the relations of the uncommon areas of the scene with the common areas of the scene .
for example , a dark , linear feature transects a light - toned , delineated surface . at this specific scale , the dark feature is uncommon , an `` interest point '' , as it has specific relation to the surrounding light - toned material .
it can `` tell the story '' of the outcrop .
continued study may show how it cuts the delineation of the light - toned material , most likely indicating a younger age .
coupled with a capacity for microscopic analysis , or even better , spectrographical analysis of mineralogy , a continued study may show the dark feature to be of basaltic composition and the light - toned material to be of granitic composition .
this data compared with the information stored in the mind of the geologist ( knowledge ) may lead to the interpretation of the outcrop as a foliated granite ( gniess ) cut be a dolerite dike . prior to implementing the ` uncommon map ' , the first step of the prototypical geologist s thought process
, we needed a segmentation algorithm , in order to produce pixel - class maps to serve as input to the uncommon map algorithm .
we have implemented the classic co - occurrence histogram algorithm ( haralick , shanmugan & dinstein 1973 ; haddon & boyce 1990 ) . for this work ,
we have not included texture information in the segmentation algorithm nor in the uncommon map algorithm .
currently , each of the three bands of hue , saturation and intensity ( @xmath1 ) color information is segmented separately , and later merged in the interest map by summing three independent uncommon maps . in ongoing work
, we are working to integrate simultaneous color & texture image segmentation into the cyborg astrobiologist system ( e.g. , freixenet , muoz , mart & llad 2004 ) .
the concept of an ` uncommon map ' is our invention , though it indubitably has been independently invented by other authors , since it is somewhat useful . in our implementation
, the uncommon map algorithm takes the top 8 pixel classes determined by the image segmentation algorithm , and ranks each pixel class according to how many pixels there are in each class .
the pixels in the pixel class with the greatest number of pixel members are numerically labelled as ` common ' , and the pixels in the pixel class with the least number of pixel members are numerically labelled as uncommon. the ` uncommonness ' hence ranges from 1 for a common pixel to 8 for an uncommon pixel , and we can therefore construct an uncommon map given any image segmentation map .
rare pixels that belong to a pixel class of 9 or greater are usually noise pixels in our tests thus far , and are currently ignored . in our work , we construct several uncommon maps from the color image mosaic , and then we sum these uncommon maps together , in order to arrive at a final interest map . for more details on our particular software techniques , especially on image segmentation and uncommon mapping ,
see mcguire _
( 2004b ) . for this mission to riba ,
the non - human hardware of the cyborg astrobiologist system consisted of : 1 @xmath0 a 667 mhz wearable computer ( from via computer systems in minnesota ) with a ` power - saving ' transmeta ` crusoe ' cpu and 112 mb of physical memory , @xmath0 an indoor / outdoor sunlight - readable tablet display with stylus ( from via computer systems ) , @xmath0 a sony ` handycam ' color video camera ( model _ dcr - trv620e - pal _ ) , and @xmath0 a tripod for the camera .
the sony handycam provides real - time imagery to the wearable computer via an ieee1394/firewire communication cable .
the system as deployed to riba used two independent batteries : one for the computer , and the other for the camera .
the power - saving aspect of the wearable computer s crusoe processor is important because it extends battery life , meaning that the human does not need to carry spare batteries . a single lithium - ion battery for the wearable computer , which weighs about 1 kg , was sufficient for this 4 hour mission .
likewise , a single lithium - ion battery ( sony model np - f960 , 38.8wh ) was sufficient for the sony handycam for the 4 hour mission to riba , despite frequent use of the power - hungry fold - out lcd display for the handycam .
the main reason for using the tripod during the mission to riba is that it allows the user to repoint the camera or to zoom in on a feature in the previously - analyzed image . in the previous study at rivas ( mcguire _ et al . _
2004b ) , mosaicking and automated repointing was part of the study , so the stable platform provided by the tripod was essential .
we eliminated the pan - tilt unit from this mission to riba because it adds to the mobility of the cyborg astrobiologist system , by eliminating the extra bag containing a battery and communication & power cables to the pan - tilt unit .
another reason for eliminating the pan - tilt unit from the system for the mission to riba is that it saves time since there is less waiting around for the mosaicking and repointing to be completed .
the capacities of automated mosaicking and of automated repointing at interest points are , however , essential to the system in the long run , and will be re - introduced at a later stage when needed . ) .
if we had had the pan - tilt unit with us , then this would have been much more automatic , as the repointing could have been under computer control . ] during this mission , the cyborg astrobiologist system analyzed 32 images from 7 different tripod positions at 3 different outcrops over a 600 meter distance and over a 4 hour period . during the previous mission at rivas , with the pan - tilt unit enabled ,
the cyborg astrobiologist system only analyzed 24 mosaic images from 3 different tripod positions at only one outcrop over a 300 meter distance and over a 5 hour period .
see table [ missionparameters ] and figure [ missionmap ] . for this particular study ,
the mobility granted by the wearable computer was almost essential . using a more powerful
non - wearable computer could have restricted the mobility somewhat , and would have made it more difficult to study the third outcrop on the slopes of the castle - topped hill .
the head - mounted display that we used during the rivas mission ( 2004 ) was much brighter than the tablet display used during the riba mission .
together with the thumb - operated finger mouse , the head - mounted display was more ergonomic during the mission to rivas , than was the tablet display and stylus used during this riba mission .
however , the spatial resolution of the head - mounted display was somewhat less than the resolution of the tablet display . during the riba mission , we wanted to share and interpret the results interactively between the three investigators .
this would have been much more difficult with the single - user head - mounted display than it was with the multi - viewer higher - resolution tablet display .
so we used the tablet display during the mission to riba , with the intention of switching to the head - mounted display later in the day .
the wearable computer processes the images acquired by the color digital video camera , to compute a map of `` interesting '' areas .
what the system determines as `` interesting '' is based on the `` uncommon '' maps ( introduced in section 2.1 ) .
it is the relation between uncommon and common that eventually can tell the geological history of the outcrop .
the computations use two - dimensional histogramming for image segmentation ( haralick , shanmugan & dinstein 1973 ; haddon & boyce 1990 ) .
this image segmentation is independently computed for each of the hue , saturation , and intensity ( h , s , i ) image planes , resulting in three different image - segmentation maps .
these image - segmentation maps were used to compute ` uncommon ' maps ( one for each of the three ( h , s , i ) image - segmentation maps ) : each of the three resulting uncommon maps gives highest weight to those regions of smallest area for the respective ( h , s , i ) image planes .
finally , the three ( h , s , i ) uncommon maps are added together into an interest map , which is used by the cyborg system in order to determine the top three interest points to report to the human operator .
the image - processing algorithms and robotic systems - control algorithms are all programmed using the graphical programming language , neo / nst ( ritter _ et al .
_ 1992 , 2002 ) .
using such a graphical programming language adds flexibility and ease - of - understanding to our cyborg astrobiologist project , which is by its nature largely a software project .
we discuss some of the details of the software implementation in mcguire _
( 2004b ) .
after segmenting the mosaic image ( see figure [ explanation ] ) , we use a very simple method to find interesting regions in the , @xmath2 , & @xmath3 ) together to form an interest map ; and third , blurring this interest map .
this smoothing kernel effectively gives more weight to clusters of uncommon pixels , rather than to isolated , rare pixels . ] .
based upon the three largest peaks in the blurred / smoothed interest map , the cyborg system then shows the human operator the locations of these three positions , overlaid on the original image .
the human operator can then decide how to use this information : i.e. , whether to ignore the information or to zoom in on one of the interest points .
this step can be automated in future versions .
on the 8th of february , 2005 , three of the authors ( daz martnez , orm & mcguire ) tested the
cyborg astrobiologist " system for the second time at a geological site , with red - sandstone layers , near the village of riba de santiuste , north of sigenza in the northern part of the province of guadalajara ( spain ) .
.a description of some of the parameters of the missions to rivas vaciamadrid and to riba de santiuste .
[ cols="<,<,<,<",options="header " , ] these tests at riba are meant to be a complementary , confirmatory test of the methodology first tested in the spring of 2004 in rivas vaciamadrid .
the site in riba may have more direct relevance as an analog for the terra meridiani site on mars , which the mars exploration rover , opportunity , is now exploring . riba has red sandstone beds , with some outcrops showing local chemical bleaching due to reduction and mobilization of iron , as well as precipitation of oxidized iron impurities .
this process gives the sandstone both dark - red and white colored areas . within the red sandstone ,
there are concretions of dark - red oxidized iron , as well as some sites where concretion had only been partially complete .
furthermore , the coloring of the red sandstones at riba is much more brilliant " and saturated than the unsaturated white coloring of the clay- & sulfate - bearing cliffs in rivas .
the dichotomy and contrast of the colors at riba , between the oxidized red and the bleached white colors of the sandstones@xmath4@xmath4 certainly made the study of these red beds " into a relatively straight - forward , but highly - discriminating , test of the cyborg astrobiologist s current computer vision capabilities .
the rocks at the outcrops at riba are of triassic age ( 260 - 200 myrs before present ) , and consist of sandstones , gravel beds , and paleosols .
the rocks were originally deposited during the triassic in different layers by the changing depositional processes of a braided river system .
this river system consisted of active channels with fast transport of sand grains and gravel . during different millenia in the triassic , the river system shifted and evolved .
therefore , in the example of the paleosol outcrop , the deposition was only of fine - grained silt , which later was affected by soil processes and formed the paleosol layers .
the rock layers were folded by alpine tectonics in the cenozoic .
we arrived at the site at 12 noon on february 8 , 2005 , and in the next 30 mins the roboticist quickly assembled the cyborg astrobiologist system , and taught the two geologists how to use the system .
for the next 4 hours , the two geologists were in charge of the mission , deciding where to point the cyborg astrobiologist s camera , interpreting the results from the tablet display , and deciding how to use the cyborg astrobiologist s assessment of the interest points in the imagery .
often the geologists chose one of the top three interest points , and then either zoomed in on that point with the camera s zoom lens , or walked with the camera and tripod to a position closer to the interest point , in order to get a closer look at the interest point with the computer vision system .
the computer was worn on the geologist s belt , and typically took 30 seconds to acquire and process the image .
the images were downsampled in both directions by a factor of 2 during these tests , and the central area was selected by cropping the image ; the final image dimensions were @xmath5 .
we chose a set of 7 tripod positions at 3 geological outcrops at riba de santiuste in order to test the cyborg astrobiologist system ( see the map in figure [ missionmap ] ) .
this is an improvement upon the number of tripod positions and outcrops studied in the first mission to rivas vaciamadrid in the spring of 2004 ( see table 1 ) . the first geological outcrop at riba de santiuste consisted of layered deposits of red sandstone . in several of the images
( see figures 6a , 6b , & 6d for examples ) , the wearable computer determined aspects of white bleaching to be interesting . furthermore , in several of the images ( see figures 6c & 6d for examples ) , the wearable computer found the concretions to be interesting .
the second geological outcrop was partly covered by a crust of sulfate and carbonate minerals . at this outcrop ,
the computer vision system found the rough texture of the white - colored mineral deposits to be interesting ( see figure 7a ) . at the third geological outcrop
, we pointed the camera and computer vision system at a paleosol .
the computer vision system of the wearable computer found the calcified root structures within the paleosols to be of most interest ( see figure 7b ) .
without the problems with shadow and texture that we encountered during our first mission at the white gypsum - bearing cliffs of rivas vaciamadrid , the system performed admirably .
see figure 6b for such an example of an image without shadow or texture .
we show more details of the automated image processing by the wearable computer of this image in figure 5 .
this image contained an area with bleached domains ( in white ) in the red sandstone , as well as a few isolated iron - oxide concretions of dark red color .
processing this image by image segmentation and by uncommon - map construction was straightforward .
the wearable computer reported to the geologists that the three most interesting points were two areas of the white bleached domain , and a third point in the lower part of the image which was a somewhat - darker tone of red .
the geologists also found the smaller dark - red concretions in the upper part of the image to be interesting .
the wearable computer found these smaller dark - red concretions to be interesting , but only before the interest map was smoothed ( or blurred ) , as summarized in figure 5 .
the wearable computer does not report the top three points from the unblurred interest map to the user .
however , the full unblurred interest map , as well as the blurred interest map , are both displayable to the user in real time , and they are both stored to disk for post - mission analysis .
we regard the fact that the wearable computer had interest in the dark red concretions , even though it did not report this interest in its top 3 interest points , as a partial success .
perhaps , if the camera had zoomed in by a factor of 10 , it would have found these dark red concretions to be interesting , even in the blurred interest map .
hence , these features would be noticed when further approaching the outcrop . even at this distance ,
the small size of the dark red concretions was at the limits of the perception capabilities of the geologists themselves .
nonetheless , the computer did inform the geologists that it had interest in the white bleached spots and the darker red region at the bottom of the image , which we regard to be a higher level of success . in figures 6 & 7
, we present a number of other images that were acquired and processed by the cyborg astrobiologist at riba de santiuste . for each image , we show here the image segmentation for the saturation slice of the color image , since the saturation slice discriminated rather well the red and white colors .
we also show the final , blurred interest map , which was computed by summing the uncommon maps for hue , saturation & intensity as in figure 5 , and then smoothing the resulting map .
figure 6a shows the border between a large oxidized red - colored zone of the red bed and a large bleached white - colored zone of the red bed .
the wearable computer found the texture in the white zone to be most interesting .
figure 6c shows a highly - magnified view of some of the concretions in the red bed near tripod position 2 .
these concretions are not well - rounded like the `` blueberries '' ; rather , they are more like the `` popcorn '' concretions observed by opportunity at meridiani planum .
the computer vision system in the wearable computer found two of the darker - red concretions on the left side of the image to be interesting , and one bright white area on the right side of figure 6c to be interesting .
the discriminatory power of the image segmentation was not very clean at this high level of magnification .
figure 6d shows a zoomed - out view of some concretions and a bleached white zone near tripod position 2 .
the computer vision system found the area of the concretions to be most interesting , followed by the bleached area in the left side of the image .
as shown in figure 6e , in the same part of the outcrop were some quartzite rocks of green , gray and yellow colors . in an image containing these rocks
, the computer vision system found several of them to be more interesting than the surrounding red sandstone . likewise , at the second outcrop , we observed some highly - textured mineral deposits of whitish colors . in a zoomed - out view of these mineral deposits ( figure 7a ) , the wearable computer found the texture of the mineral deposits to be more interesting than the more - homogenous red color of the neighboring sandstone .
the wearable computer did not find the highly - textured region containing many dark holes to be as interesting as the white - textured mineral deposits . in an ideal computer vision system ,
one of the top three interest points probably should have been within the region with all of the dark holes .
these dark holes formed by a combination of physical ( wind erosion ) and chemical ( differential cementation ) processes .
based partly upon the guidance of the wearable computer , we acquired several samples of the mineral deposits for analysis in the laboratory .
based upon the geologists experience and upon acid tests of these mineral deposits at the field site , we suspected that the mineral deposits consisted largely of a white textured overlayer of 15 - 20 mm of a sulfate , with a thin gray underlayer of 1 - 10 mm of calcite .
the laboratory tests later confirmed that the overlayer is composed of largely of hydrated calcium sulfate ( gypsum ) . at the third outcrop ,
the cyborg astrobiologist system studied a layer of paleosols , which are `` affected '' silty deposits from a fluvial plain .
these paleosols contained partly - calcified root structures of plants that grew in this fluvial plain during the triassic .
at tripod position 6 , our computer vision system found the white - colored calcified - root structures to be more interesting than the surrounding matrix of red siltstone ( figure 7b ) .
however , at tripod position 7 , in a more complex image ( not shown here ) with more root structures and more diverse coloring , the computer vision system did not find the root structures to be particularly interesting .
we have compared the cyborg astrobiologist s performance in picking the top 3 interest points with a human geologist s quasi - blind classification .
geologist daz martnez verbally noted the interesting parts of the image .
then daz martnez was shown the cyborg astrobiologist s top three interest points , and he then judged how well the computer vision system matched his own image analysis .
we judged that there was concurrence between the human geologist and the cyborg astrobiologist on the interest points about 69 times ( true positives ) , with 32 false positives and with 32 false negatives , for a total of 32 images studied .
this was an average of 2.1 true positives ( tp ) for each image out of a possible ( usually ) 3 trials per image ; an average of 1.0 false positives ( fp ) per image out of a possible ( usually ) 3 trials per image ; and an average of 1.0 false negatives ( fn ) per image , with no _ a priori _ limitation on the number of false negatives per image .
there was double counting of positives since sometimes a localized physical feature corresponded to more than one interesting features in the image ( i.e. one image had some parallel lines at a geological contact , so the interest points corresponding to this region counted doubly ) .
one way to look at this data is to take the ratios @xmath6 , @xmath7 and @xmath8 . with our technique ,
it is difficult to assess the true negative rate @xmath9 . for the data from riba , we compute : a true positive rate of @xmath10 , a false positive rate of @xmath11 , and a false negative rate of @xmath12 . we did not attempt to compute the roc curves for these images , as a function of the number of interest points computed for each image .
a more careful analysis could be made with many more images from different field sites , but these numbers agree qualitatively with the results we were getting at rivas vaciamadrid , even though we have not computed the statistics .
qualitatively , we expect that for the types of imagery that we typically study , the cyborg astrobiologist will not do a near - perfect job ( with geologist - concurrence analysis giving @xmath13 and @xmath14 both less than 10% ) , nor will it do a poor job ( with geologist - concurrence analysis giving either or both @xmath13 and @xmath14 greater than 90% ) , it typically has mid - level false - positive and false - negative rates a mid - level in performance .
of course , we would like to improve the system to do a near - perfect job , but considering the simplicity of the image - segmentation algorithm and the uncommon - mapping technique , we believe that this work can provide a good basis for further studies . one approach for improving the false positive rate is to include further filtering of the interest points that are currently being determined from the uncommon maps . that way ,
errant positives could be caught before they are reported to the human operator . right now in the field , the human operator serves as a decent filter of false positives , so fixing the false positive rate is not so urgent ; fixing the false negative rate is somewhat more important .
one approach for improving the false negative rate would be to compute interest features in new and different ways ( i.e. , with edge detectors or parallel - line detectors ) , in order to ensure that all the interesting features are detected .
both approaches could be attacked , for example , by using context - dependent geologist knowledge , which would need to be coded into a future , more advanced cyborg astrobiologist .
currently , considering that we do not deploy such context - dependent geologist knowledge , our relatively - unbiased uncommon - mapping technique seems to be doing rather well .
we have shown that our cyborg astrobiologist exploration system performs reasonably well at a second geological field site . given similar performance at the first geological field site , we can have some degree of confidence in the general unbiased approach towards autonomous exploration that we are now using .
we can also have sufficient confidence in our general technique that we can use our specific technique of image segmentation and uncommon mapping as a basis for further algorithm development and testing . in the near future ,
we plan to : * upgrade our image - segmentation algorithm , in order to handle texture and color simultaneously ( cf .
this upgrade may give us more capabilities to handle shadow as well .
* test the cyborg astrobiologist system at field sites with a microscopic imager .
this will complete our mimicry of a stepwise approach by a human geologist towards an outcrop . at riba de santiuste ,
our system has been tested for exploration on imagery at a site that is not unlike meridiani planum , where the mer opportunity is now exploring . riba de
santiuste is similar to meridiani planum in its iron - oxide concretions and its sulfate mineral deposits .
the bleached - white zones of the red sandstones at riba de santiuste may be an analog for similar geochemical or perhaps biological phenomena on mars ( chan _ et al .
_ 2004 ; orm _ et al .
_ 2004 ; hofmann 1990 , 2004 ) .
the software behind the cyborg astrobiologist s computer vision system has shown its initial capabilities for imagery of this type , as well as for imagery of some other types .
this computer vision software may one day be mature enough for transplanting into the computer - vision exploration system of a mars - bound orbiter , robot or astronaut .
p. mcguire , j. orm and e. daz martnez would all like to thank the ramon y cajal fellowship program of the spanish ministry of education and science .
many colleagues have made this project possible through their technical assistance , administrative assistance , or scientific conversations .
we give special thanks to kai neuffer , antonino giaquinta , fernando camps martnez , and alain lepinette malvitte for their technical support .
we are indebted to gloria gallego , carmen gonzlez , ramon fernndez , coronel angel santamaria , and juan prez mercader for their administrative support .
we acknowledge conversations with beda hofmann , virginia souza - egipsy , mara paz zorzano mier , carmen crdoba jabonero , josefina torres redondo , vctor r. ruiz , irene schneider , carol stoker , jrg walter , claudia noelker , gunther heidemann , robert rae , jonathan lunine , ralph lorenz , goro komatsu , nick woolf , steve mojzsis , david p. miller , bradley joliff , raymond arvidson , and daphne stoner . the field work by j. orm was partially supported by grants from the spanish ministry of education and science ( aya2003 - 01203 and cgl2004 - 03215 ) .
the equipment used in this work was purchased by grants to our center for astrobiology from its sponsoring research organizations , csic and inta .
compton , w.d .
( 1989 ) . _ where no man has gone before : a history of apollo lunar exploration missions _ , chapters 13 - 14 . based on nasa history series _
sp-4214_. + http://www.apolloexplorer.co.uk/default.asp?libsrc=/books/sp-4214/cover.htm
freixenet , j. , muoz , x. , mart , j. & llad , x. ( 2004 ) .
color texture segmentation by region - boundary cooperation .
_ computer vision
eccv 2004 , eighth european conference on computer vision , proceedings , part ii , lecture notes in computer science _ springer , prague , czech republic , eds .
toms pajdla , jir matas , * 3022 * , pp .
250 - 261 . also available in the _ cvonline _
archive : + http://homepages.inf.ed.ac.uk/rbf/cvonline/local_copies/freixenet1/eccv04.html hofmann , b. ( 2004 ) .
redox boundaries on mars as sites of microbial activity .
_ iv european workshop on exo / astrobiology _ , held at the open university , milton keynes , united kingdom , abstract , + http://physics.open.ac.uk/eana/talks/redox%20boundaries%20on%20mars%20as%20sites%20of%20microbial%20activity.pdf mcguire , p. c. , rodrguez - manfredi , j. a. , sebastin - martinez , e. , gomez - elvira , j. , daz - martnez , e. , orm , j. , neuffer , k. , giaquinta , a. , camps - martnez , f. , lepinette - malvitte , a. , prez - mercader , j. , ritter , h. , oesker , m. , ontrup , j. & walter , j. ( 2004a ) .
cyborg systems as platforms for computer - vision algorithm - development for astrobiology , _ proceedings of the iii european workshop on exo / astrobiology _ , held at the centro de astrobiologia , madrid , _
european space agency special publication , esa sp-545 _ , pp .
141 - 144 .
http://arxiv.org/abs/cs.cv/0401004 mcguire , p. c. , orm , j. o. , diaz - martinez , e. , rodriguez - manfredi , j. a. , gomez - elvira , j. , ritter , h. , oesker , m. , & ontrup , j. ( 2004b ) .
the cyborg astrobiologist : first field experience , _ international journal of astrobiology _ , vol .
3 , issue 3 , pp .
189 - 207 .
http://arxiv.org/abs/cs.cv/0410071 orm , j. , komatsu , g. , chan , m. a. , beitler , b. , and parry , w. t. ( 2004 ) .
geological features indicative of processes related to the hematite formation in meridiani planum and aram chaos , mars : a comparison with diagenetic hematite deposits in southern utah , usa , _
icarus _ * 171 * , pp .
295 - 316 .
scott , j. r. , mcjunkin , t. r. , tremblay , p. l. ( 2003 ) .
automated analysis of mass spectral data using fuzzy logic classification , _ journal of the association for laboratory automation _ * 8 * ( 2 ) , pp .
61 - 63 .
scott , j.r . & tremblay , p.l .
highly reproducible laser beam scanning device for an internal source laser desorption microprobe fourier transform mass spectrometer , _ review of scientific instruments _ * 73 * ( 3 ) , pp .
1108 - 1116 . | the ` cyborg astrobiologist ' has undergone a second geological field trial , at a site in northern guadalajara , spain , near riba de santiuste .
the site at riba de santiuste is dominated by layered deposits of red sandstones .
the cyborg astrobiologist is a wearable computer and video camera system that has demonstrated a capability to find uncommon interest points in geological imagery in real - time in the field . in this second field trial ,
the computer vision system of the cyborg astrobiologist was tested at seven different tripod positions , on three different geological structures .
the first geological structure was an outcrop of nearly homogeneous sandstone , which exhibits oxidized - iron impurities in red and and an absence of these iron impurities in white .
the white areas in these `` red beds '' have turned white because the iron has been removed . the iron removal from the sandstone
can proceed once the iron has been chemically reduced , perhaps by a biological agent .
the computer vision system found in one instance several ( iron - free ) white spots to be uncommon and therefore interesting , as well as several small and dark nodules .
the second geological structure was another outcrop some 600 meters to the east , with white , textured mineral deposits on the surface of the sandstone , at the bottom of the outcrop .
the computer vision system found these white , textured mineral deposits to be interesting .
we acquired samples of the mineral deposits for geochemical analysis in the laboratory .
this laboratory analysis of the crust identifies a double layer , consisting of an internal millimeter - size layering of calcite and an external centimeter - size effluorescence of gypsum .
the third geological structure was a 50 cm thick paleosol layer , with fossilized root structures of some plants .
the computer vision system also found certain areas of these root structures to be interesting .
a quasi - blind comparison of the cyborg astrobiologist s interest points for these images with the interest points determined afterwards by a human geologist shows that the cyborg astrobiologist concurred with the human geologist 68% of the time ( true positive rate ) , with a 32% false positive rate and a 32% false negative rate .
the performance of the cyborg astrobiologist s computer vision system was by no means perfect , so there is plenty of room for improvement .
however , these tests validate the image - segmentation and uncommon - mapping technique that we first employed at a different geological site ( rivas vaciamadrid ) with somewhat different properties of the imagery .
* keywords : * computer vision , robotics , image segmentation , uncommon map , interest map , field geology , mars , meridiani planum , wearable computers , co - occurrence histograms , red beds , red sandstones , nodules , concretions , reduction spheroids , triassic period . 1 |
the quadratic assignment problem ( qap ) is a combinatorial optimization problem first introduced by koopmans and beckman ( 1957 ) .
it is np - hard and is considered to be one of the most difficult problems to be be solved optimally .
the problem was defined in the following context : a set of @xmath1 facilities are to be located at @xmath1 locations .
the distance between locations @xmath5 and @xmath6 is @xmath7 and the quantity of materials which flow between facilities @xmath5 and @xmath6 is @xmath8 .
the problem is to assign to each location a single facility so as to minimize the cost @xmath9 where @xmath10 represents the location to which facility @xmath5 is assigned
. it will be helpful to think of the @xmath1 facilities and the matrix of flows between them in graph theoretic terms as a graph of @xmath1 nodes and weighted edges , respectively .
there is an extensive literature which addresses the qap and is reviewed in pardalos et al .
( 1994 ) , cela ( 1998 ) , anstreicher ( 2003 ) , loiola et al .
( 2007 , and james et al .
( 2009a ) . with the exception of specially constructed cases ,
optimal algorithms have solved only relatively small instances with @xmath11 .
various heuristic approaches have been developed and applied to problems typically of size @xmath12 or less .
one of the most successful heuristics to date for large instances is _ robust tabu search _ , rts , ( taillard ( 1991 ) .
the use of tabu search for the quadratic assignment problem has been studied extensively ( drezner ( 2005 ) , hasegawa et al .
( 2000 ) , james et al , ( 2009a , 2009b ) , mcloughlin and cedeno ( 2005 ) , misevicius ( 2007 ) , misevicius and ostreika ( 2007 ) , skorinkapov ( 1994 ) , and wang ( 2007 ) ) .
some of the best available algorithms for the solution of the qap are the hybrid genetic algorithms that use tabu search as an improvement mechanism .
( see drezner ( 2008 ) ) .
here we will consider the robust tabu heuristic applied to _ sparse _ qap instances .
that is , the number of non - zero entries in the either the flow matrix and/or the distance matrix is of @xmath4 as opposed to @xmath0 . without loss of generality
we will assume the flow matrix is sparse .
many real world problems are sparse . in fact , this work was motivated by the study of random regular sparse graphs .
these graphs are very robust to partitioning and collapse due to removal of nodes or edges .
we are interested in the problem of determining how to assign the nodes of such a graph to locations in a metric space such that the total edge length of the graph is minimized ; this problem maps directly to a quadratic assignment problem .
there has been some previous work on sparse quadratic assignment problems .
milos and magirou ( 1995 ) developed a lagrangian - relaxation lower - bound algorithm for sparse problems and panos et al .
( 1997 ) developed a version of their grasp heuristic for sparse problems .
however , to the best of our knowledge , an efficient implementation of the robust tabu heuristic for sparse qap instances has not been proposed .
the tabu heuristic for the quadratic assignment problem consists of repeatedly swapping locations of two nodes . a single iteration of the heuristic consists of * determining the move which most decreases the total cost . under certain conditions ( see section 4 ) , if a move which lowers the cost is not available , a move which raises the cost is made .
so that cycles of the same moves are avoided , the same move is forbidden ( _ taboo _ ) until a specified later iteration ; we call this later iteration the _ eligible iteration _ for a given move .
this eligible iteration is traditionally stored in a _
tabu list _ or _
tabu table_. * making the move .
* recalculating the new cost of all moves .
the process is repeated for a specified number of iterations .
traditional implementations of robust tabu search require @xmath0 operations per iteration .
the complexity of @xmath0 is achieved by maintaining a matrix containing the cost @xmath13 of swapping @xmath14 and @xmath15 for all @xmath14 and @xmath15 , given a current assignment @xmath16 .
the complexity of the each step above is as follows : * @xmath0 - all possible @xmath17 moves are considered .
the cost of each move is retrieved from @xmath18 * @xmath19 - the locations of the two swapped nodes are simply transposed .
* @xmath0 - based on the following observations of taillard ( 1991 ) : + following taillard ( 1991 ) , starting from an assignment of facilities @xmath16 let the resulting assignment after swapping facilities @xmath20 and @xmath21 be @xmath22 .
that is : @xmath23 for a symmetrical matrix with a null - diagonal , the cost @xmath18 of swapping @xmath20 and @xmath21 is : @xmath24 + to calculate @xmath25 for any @xmath14 and @xmath15 with complexity o(n ) , we can use equation xx . for asymmetric matices or matrices with non - null diagonals , a slightly more complicated version of equation ( [ eqfull ] ) also of complexity o(n )
is given by burkhard and rendl ( 1984 ) .
+ to calculate @xmath25 in the case that the swapped facilities @xmath14 and @xmath15 are different from @xmath20 or @xmath21 , we use the value @xmath13 calculated in the previous iteration and find : @xmath26 * * the cost of moves which do not involve the two nodes in the previous move can be calculated in time @xmath19 .
there are @xmath0 of these moves . *
* the cost of moves which do involve the two nodes in the previous move must be calculated from scratch .
there are @xmath4 of these moves and the complexity of calculating each is @xmath4 .
to reduce the complexity of step ( a ) , instead of scanning all possible moves , we use multiple _ priority queues _ ( pqs ) to determine the best move .
a priority queue is a data structure for maintaining a set of elements each of which has an associated value ( priority ) .
a pq supports the following operations : * insert an item * remove an item * return the item with the highest value priority queues are used to efficiently find an item with the highest value without searching through all of the items .
the maximum complexity of pq operations is @xmath27 .
we will see below that there will be @xmath4 insertions and deletions in the pqs for each iteration so the asymptotic complexity of this step is reduced to @xmath28 .
furthermore , we will show that for problems of any practical size , pq operations are not the determinant of total complexity .
the complexity to recalculate the cost of moves in step ( c ) , can be reduced to @xmath4 as follows : * as in the traditional robust tabu implementation , the cost of moves which do not involve the two nodes in the previous move can be calculated in time @xmath19 . on average , there are @xmath29 nodes which are connected to the two nodes in the previous moves , where @xmath30 is the average degree ( average number of nodes adjacent to a given node ) of the graph corresponding to the flow matrix . for each of these @xmath31 nodes we must calculate the cost of @xmath32 possible moves .
thus , the cost is @xmath33 .
* the cost of moves which do involve the two nodes in the previous move must be calculated from scratch .
there are @xmath4 of these moves and the complexity of calculating each is @xmath34 since the cost of a node , @xmath14 , being in a specific location depends only on the on - average @xmath35 nodes adjacent to @xmath14 .
thus the complexity of step ( c ) is reduced to @xmath4 .
to describe our implementation , we must first describe the rules for determining the next move of taillard s robust tabu heuristic ( taillard ( 1991 ) ) .
the following definitions for the possible _ state _ of a potential move are useful : * if the current iteration is less than or equal to the eligible iteration , the move is _
ineligible_. * if the current iteration is greater than the eligible iteration , the move is _
authorized_. * if the current iteration minus an _ aspiration constant _ is greater than the eligible iteration the move is _
aspired_. the rules for determining the next move can then be stated as ( taillard ( 1991 ) ) : * if a move which decreases the lowest total cost found so far is available , the move which most decreases this total cost is chosen , independent of whether the move is ineligible , authorized or aspired . * if no move meets criterion ( 1 ) , the aspired move , if one is available , which most decreases the current total cost is chosen . *
if no moves meet criteria ( 1 ) or ( 2 ) , the lowest cost authorized move is chosen . to implement these rules for sparse problems , we use two types of pqs : _ delta _ pqs which contain the cost delta for a given move and _ tabu _ pqs which contain entries ordered by the eligible iteration for the move . the tabu pqs control the change of state of a move .
the delta pqs determine the lowest cost move in each state .
five pqs are used : * _ ineligible tabu pq _ - this pq contains moves , ordered by eligible iteration , which are in the ineligible state .
this pq allows us to efficiently determine when the state of a move can be changed to authorized . * authorized tabu pq - this pq contains moves , ordered by eligible iteration , which are in the authorized state .
this pq allows us to efficiently determine when the state of a move can be changed to aspired . * ineligible delta pq - this pq contains moves , ordered by the cost of the move , which are in the ineligible state .
this pq together with the two other delta pqs allows for efficient determination of the overall lowest cost move as required by rule 1 . * aspired delta pq - this pq contains moves , ordered by the cost of the move , which are in the aspired state .
this pq allows for efficient determination of the lowest cost aspired move as required by rule 2 . *
authorized delta pq - this pq contains moves , ordered by the cost of the move , which are in the authorized state .
this pq allows us to determine the lowest cost authorized move as needed by rule 3 . as illustrated in fig .
[ pqs ] , moves are inserted and removed in the pqs under the following circumstances : * at initialization all moves are inserted into the ineligible pqs .
* at the beginning of each iteration , any moves on the ineligible pqs which become authorized , because the iteration has increased by one , are moved from the ineligible pqs to the authorized pqs . * at the beginning of each iteration , any moves on the authorized pqs which become aspired , because the iteration has increased by one , are moved to the aspired pq . * after each move , any move for which the move cost or eligible iteration has changed is removed from the pq in which it is present and inserted in the appropriate pq based on move state and move cost .
( however , see lazy update discussion below . ) using these pqs we obtain _ exactly _ the same results as the traditional robust tabu implementation .
we minimize the time updating pqs by performing _
lazy updates_. after a change in the eligible iteration or move cost , if the state is changed or the value is increased , we update the pqs involved ; otherwise , we perform a lazy update and store the value in a data structure associated with the move and only do the update in the pq when and if the move becomes the move with the smallest value in the pq .
this use of lazy updates significantly decreases the time spent on pq operations .
= 7.0 cm .
[ ratio ] = 7.0 cm = 7.0 cm
we test our algorithm on instances with @xmath1 locations on a square grid with a euclidean metric .
the flow matrix for the @xmath1 facilities corresponds to an adjacency matrix for a @xmath35-regular random graph ; that is , each facility has flows of value one to @xmath35 other random facilities .
we run both the traditional non - sparse implementation and our new sparse implementation .
the non - sparse implementation is the code of taillard ( 1991 ) . to implement the priority queues in the sparse implementation , we use the complete balanced tree code of marin and cordero ( 1995 ) .
we perform a minimum of @xmath36 iterations for each instance . for @xmath37 ,
priority queue update operations consume just @xmath38 of the total time . assuming a @xmath39 behavior for pq operations , it is not until @xmath1 becomes astronomically large that the pq operations would take the majority of the time and the behavior of our implementation crosses over from @xmath4 to @xmath3 complexity .
we thus expect @xmath4 complexity for our sparse implementation for any practical sized problems . in fig .
[ ratio ] , for @xmath40 , we plot the ratio @xmath20 of the time per iteration of the non - sparse implementation to the time per iteration of the sparse implementation versus @xmath1 .
as expected , the plot is consistent with @xmath41 where @xmath42 .
this reflects the reduction in complexity from @xmath0 to @xmath4 . in fig .
[ comb ] we plot separately the time per iteration for the original and the sparse implementations .
consistent with fig .
[ ratio ] , the slopes on the log - log plot differ by 1 but the slopes are 2.5 and 1.5 , respectively , as opposed to the theoretical values of 2.0 and 1.0 . as explained in paul ( 2007 ) and saavedra and smith ( 1995 ) ,
this is due to the finite size of processor cache memory ; as the problem size ( and memory needed ) increases , there are a smaller percentage of cache hits causing slower operation . in fig .
[ tk ] we plot the time per iteration versus values of degree @xmath35 .
the plot is linear with a slight deviation with increasing @xmath35 .
this deviation from linear is due to the fact that as @xmath35 increases there is an increasing chance that a node @xmath14 will be connected to both nodes involved in the previous move ; however , the updated costs of moving node @xmath14 must be calculated only once .
we also performed numerical experiments on a class of problems known in the literature ( see drezner et al .
( 2005 ) ) .
these problems denoted as drexx are sparse and are designed to be very difficult to solve with heuristics .
we obtained results similar to those for described above ; performance is o(n ) .
for sparse quadratic assignment problems , we reduce the asymptotic complexity per iteration of robust tabu search to @xmath28 from @xmath0 ; for practical size problems , the complexity is reduced to @xmath4 .
central to achieving this reduction is the use of multiple priority queues and lazy updates to these queues .
the code which implements our approach and test qap instances used for this paper are available as supplementary material .
we thank jia shao for helpful discussions and the defense threat reduction agency ( dtra ) for support .
anstreicher , k. , 2003 .
recent advances in the solution of quadratic assignment problems . mathematical programming 97 , 27 - 42 .
drezner , z. , hahn , p.m .
, taillard .
d. , 2005 .
recent advances for the quadratic assignment problem with special emphasis on instances that are difficult for meta - heuristic methods .
annals of operations research 139 , 65 - 94 .
drezner z. and marcoulides g. , 2009 . on the range of tabu tenure in solving quadratic assignment problems recent advances in computing and management information systems . in : recent advances in computing and management information systems ,
157 - 169 , p. petratos and g. a. marcoulides ( eds ) , atiner .
james , t. , rego , c. , glover , f. , 2009a .
multistart tabu search and diversification strategies for the quadratic assignment problem .
ieee tran . on systems ,
man , and cybernetics part a : systems and humans 39 , 579 - 596 .
mcloughlin , j. f. , cedeno , w. , 2005 . the enhanced evolutionary tabu search and its application to the quadratic assignment problem .
gecco 2005 : genetic and evolutionary computation conference 1 , 975 - 982 .
pardalos , p.m. , rendl , f. , wolkowicz , h. , 1994 .
the quadratic assignment problem : a survey and recent developments . in : pardalos , p.m. , wolkowicz , h. , ( eds . ) , quadratic assignment and related problems .
dimacs series on discrete mathematics and theoretical computer science 16 , amer .
baltimore , md . 1 - 42 .
pardalos , p. m. , pitsoulis , l. s. , resende , m. g. c. , 1997 .
algorithm 769 : fortran subroutines for approximate solution of sparse quadratic assignment problems using grasp .
acm transactions on mathematical software ( toms ) 23 , 196 - 208 . | we propose and develop an efficient implementation of the robust tabu search heuristic for sparse quadratic assignment problems .
the traditional implementation of the heuristic applicable to all quadratic assignment problems is of @xmath0 complexity per iteration for problems of size @xmath1 . using multiple priority queues to determine the next best move instead of scanning all possible moves , and using adjacency lists to minimize the operations needed to determine the cost of moves , we reduce the asymptotic ( @xmath2 ) complexity per iteration to @xmath3 . for practical sized problems ,
the complexity is @xmath4 .
combinatorial optimization , computing science , heuristics , tabu search |
in 1982 calabi @xcite constructed , in local complex coordinates , a general @xmath1-parameter family of @xmath0-invariant extremal khler metrics , which he used to put an extremal khler metric on @xmath2 for all @xmath3 and any possible khler cohomology class . in particular , when @xmath4 , on all hirzebruch surfaces .
the main goal of these notes is to present calabi s general family in local symplectic action - angle coordinates , using the set - up of @xcite for toric khler geometry , and show that it actually contains other interesting cohomogeneity one khler metrics as particular cases ( see also @xcite ) .
these include : * the fubini - study , flat and bergman khler - einstein metrics of constant holomorphic sectional curvature ( positive , zero and negative , resp . ) .
* the complete ricci flat khler metric on the total space of @xmath5 for all @xmath6 and any possible khler cohomology class , constructed by calabi @xcite in 1979 . * the complete scalar flat khler metric on the total space of @xmath7 for all @xmath8 and any possible khler cohomology class , constructed for @xmath4 by lebrun @xcite in 1988 and for @xmath9 by pedersen - poon @xcite in 1991 ( see also simanca @xcite ) . * the complete khler - einstein metric with negative scalar curvature on the total space of the open disc bundle @xmath10 for all @xmath11 and any possible khler cohomology class , constructed by pedersen - poon @xcite . * the complete constant negative scalar curvature khler metric on the total space of the open disc bundle @xmath10 for all @xmath3 and any possible khler cohomology class , also constructed by pedersen - poon @xcite .
calabi s general family contains many other interesting cohomogeneity one special khler metrics . besides the bochner - khler orbifold examples presented in @xcite
, it contains for example a family of singular khler - einstein metrics on certain @xmath12 that are directly related to the sasaki - einstein metrics constructed by gauntlett - martelli - sparks - waldram @xcite in 2004 - see @xcite .
these notes are organized as follows . in section [ s : troicsympmflds ] we give a basic introduction to symplectic geometry and discuss some fundamental features of toric symplectic manifolds .
section [ s : torickmetrics ] is devoted to toric khler metrics .
after some relevant linear algebra prelimaries , we explain how these can be parametrized in action - angle coordinates via symplectic potentials for the associated toric compatible complex structures , and discuss some important properties of these symplectic potentials . in section [ s:2d - metrics ]
we write down the symplectic potentials that give rise to toric constant ( scalar ) curvature metrics in real dimension @xmath13 and identify the underlying toric symplectic surfaces .
this is a warm - up for section [ s : calabi ] , where we discuss in detail the above higher dimensional examples that arise in calabi s general family of local @xmath0-invariant extremal khler metrics .
i thank the organizers of the xi international conference on geometry , integrability and quantization , varna , bulgaria , june 510 , 2009 , where this material was presented as part of a mini - course . the work presented in section [ s : calabi ]
was carried out in january - june of 2001 , while i was at the fields institute for research in mathematical sciences , toronto , canada .
it was presented in seminar talks given at the workshop on hamiltonian group actions and quantization , fields institute , june 413 , 2001 , and at differentialgeometrie i m groen , mathematisches forschungsinsitut oberwolfach , germany , june 1016 , 2001 .
i thank the support and hospitality of the fields institute and the organizers of those meetings .
in this section we give a basic introduction to symplectic geometry and discuss some fundamental features of toric symplectic manifolds .
[ def : sympmfld ] a _ symplectic manifold _ is a pair @xmath14 where @xmath15 is a smooth manifold and @xmath16 is a closed and non - degenerate @xmath13-form , i.e. * @xmath17 is such that @xmath18 and * for any @xmath19 and @xmath20 , there exists @xmath21 such that @xmath22 .
the non - degeneracy condition ( ii ) implies that a symplectic manifold is always even dimensional . if @xmath15 has dimension @xmath23 , the non - degeneracy condition ( ii ) is equivalent to requiring that @xmath24 hence , a symplectic manifold @xmath14 is always oriented . [ ex : r2n ] the most basic example is @xmath25 with linear coordinates @xmath26 and symplectic form @xmath27 [ ex:2d ] any @xmath13-dimensional surface equipped with an area form is a symplectic manifold . for example , the sphere @xmath28 or any other compact orientable surface @xmath29 of genus @xmath30 .
[ ex : prod ] if @xmath31 and @xmath32 are symplectic manifolds , then @xmath33 is also a symplectic manifold .
here , @xmath34 means the sum of the pullbacks of the symplectic forms @xmath35 and @xmath36 from the factors @xmath37 and @xmath38 .
[ ex : kahler ] the imaginary part of the hermitean metric on any khler manifold is a symplectic form .
hence , any khler manifold is a symplectic manifold .
in particular , the complex projective space @xmath39 equipped with its fubini - study form @xmath40 is a symplectic manifold . when @xmath14 is a compact symplectic manifold we have that @xmath41\in h^{2n}(b,{\mathbb r } ) \rightarrow 0\ne [ \omega]\in h^2(b,{\mathbb r})\,.\ ] ] in particular , the spheres @xmath42 have no symplectic form when @xmath43 , since @xmath44 [ def : sympdiff ] let @xmath14 be a symplectic manifold .
a _ symplectomorphism _ of @xmath15 is a diffeomorphism @xmath45 such that @xmath46 .
these form the _ symplectomorphism group _ , a subgroup of @xmath47 that will be denoted by @xmath48 .
[ ex : hamdiff ] consider a symplectic manifold @xmath14 and let @xmath49 be a smooth function on @xmath15 .
the non - degeneracy of @xmath16 implies that there exists a unique vector field @xmath50 such that @xmath51 .
this vector field @xmath52 is called the _ hamiltonian vector field _ of the function @xmath53 and has the following fundamental property : @xmath54 this can be proved using cartan s formula to compute @xmath55 hence , on a symplectic manifold @xmath14 any smooth function @xmath56 gives rise , through the flow of the corresponding hamiltonian vector field @xmath57 , to a @xmath58-parameter group of symplectomorphisms .
one can use the symplectomorphisms constructed in the previous example to prove that : * the symplectomorphism group @xmath48 is always infinite - dimensional ; * the action of @xmath48 on the manifold @xmath15 is always @xmath59-transitive , for any @xmath60 ; * in particular , any point of a symplectic manifold @xmath14 looks locally like any other point of @xmath14 .
this last statement is made more precise in the following _ [ darboux ] _ let @xmath61 be a symplectic manifold of dimension @xmath23 .
then , any point @xmath19 has a neighborhood @xmath62 symplectomorphic to a neighborhood @xmath63 of the origin in @xmath64 , i.e. there exists a diffeomorphism @xmath65 in other words , @xmath66 which is in sharp contrast with what happens , for example , in riemannian geometry .
the lie algebra of the symplectomorphism group @xmath48 , viewed as an infinite - dimensional lie group , is naturally identified with the vector space @xmath67 of _ symplectic vector fields _ ,
i.e. vector fields @xmath68 such that @xmath69 , with lie bracket @xmath70 $ ] given by the usual lie bracket of vector fields . as before
, we can use cartan s formula to obtain @xmath71 hence , the vector space of symplectic vector fields is given by @xmath72 while its subspace of _ hamiltonian vector fields _ is given by @xmath73 in fact , as the following theorem shows , @xmath74 is a lie subalgebra of @xmath67 .
[ thm : hamvf ] if @xmath75 are symplectic vector fields , then @xmath76 $ ] is the hamiltonian vector field of the function @xmath77 , i.e. @xmath78 = x_{\omega(y , x ) } \in { \mathcal x}_h(b,\omega)\,.\ ] ] it suffices to compute @xmath76 \lrcorner\ , \omega$ ] , using standard formulas from differential geometry and the defining properties of @xmath79 , @xmath80 and @xmath16 : @xmath81 \lrcorner\ , \omega & = & { \mathcal l}_x ( y\lrcorner\,\omega ) - y\lrcorner\ , ( { \mathcal l}_x \omega ) \\
& = & d(x\lrcorner\,(y\lrcorner\,\omega ) ) + x\lrcorner\,(d(y\lrcorner\ , \omega ) ) - y \lrcorner\ , ( d(x\lrcorner\,\omega ) ) - y \lrcorner\ , ( x\lrcorner\ , d\omega ) \\ & = & d(\omega(y , x))\ , .
\end{array}\ ] ] [ rmk : hamgp ] @xmath82 is the lie algebra of a fundamental subgroup of the symplectomorphism group : the subgroup @xmath83 of _ hamiltonian symplectomorphisms _ of @xmath14 .
it follows from theorem [ thm : hamvf ] that this lie algebra can be naturally identified with the vector space @xmath84 , i.e. smooth functions on @xmath15 modulo constants , equipped with a bracket @xmath85 known as the poisson bracket : @xmath86 note that when @xmath87 we have that @xmath88 .
let @xmath14 be a symplectic manifold equipped with a symplectic action of @xmath89 i.e. with a homomorphism @xmath90 .
let @xmath91 be the vector fields generating the action of each individual @xmath92-factor .
then , since the action is symplectic , we have that @xmath93 i.e. @xmath94 [ def : hamabaction ] a symplectic @xmath95-action on a symplectic manifold @xmath14 is said to be _ hamiltonian _ if for every @xmath96 there exists a function @xmath97 such that @xmath98 , i.e. @xmath99 is the hamiltonian vector field of @xmath100 . in this case , the map @xmath101 defined by @xmath102 is called a _ moment map _ for the action .
suppose @xmath101 is a moment map for a hamiltonian @xmath95-action on @xmath14 .
then @xmath103 , for any given constant @xmath104 , is also a moment map for that same action .
[ rmk : isotorb ] the orbits of a hamiltonian @xmath95-action on a symplectic manifold @xmath61 are always _ isotropic _ , i.e. @xmath105 in fact , the tangent space to an orbit is generated by the hamiltonian vector fields @xmath106 , @xmath107 . using theorem [ thm : hamvf ] and
the fact that the torus @xmath95 is abelian , we have that @xmath108 \equiv 0 \rightarrow \omega(x_{h_k } , x_{h_l } ) \equiv \text{const.}\,,\ \forall\ , k , l\in \{1,\ldots , m\}\,.\ ] ] since @xmath95 is compact , there is for each @xmath107 and on each @xmath95-orbit a point @xmath109 where the function @xmath110 hence , the above constant is actually zero and each @xmath95-orbit is indeed isotropic .
this fact will be used below , in the proof of proposition [ prop : dimtorus ] .
[ ex : hamr2n ] consider @xmath64 , where @xmath111 as in example [ ex : r2n ] , and its usual identification with @xmath112 given by @xmath113 the standard @xmath114-action @xmath115 on @xmath25 , given by @xmath116 is hamiltonian , with moment map @xmath117 given by @xmath118 [ ex : hampn ] consider projective space @xmath119 , with homogeneous coordinates @xmath120 $ ] .
the @xmath114-action @xmath121 on @xmath122 given by @xmath123 = [ z_0 ; e^{-i y_1}z_1 ; \ldots ; e^{-i y_n}z_n]\,,\ ] ] is hamiltonian , with moment map @xmath124 given by @xmath125 = \frac{1}{\|z\|^2 } ( \|z_1\|^2 , \ldots , \|z_n\|^2)\,.\ ] ] note that the image of @xmath126 is the convex hull of the images of the @xmath127 fixed points of the action , i.e. the standard simplex in @xmath128 .
atiyah @xcite and guillemin - sternberg @xcite proved in 1982 the following convexity theorem .
[ thm : ags - convex ] let @xmath14 be a compact , connected , symplectic manifold , equipped with a hamiltonian @xmath95-action with moment map @xmath101
. then * the level sets @xmath129 of the moment map are connected ( for any @xmath130 ) ; * the image @xmath131 of the moment map is the convex hull of the images of the fixed points of the action .
the following proposition motivates the definition of a toric symplectic manifold .
[ prop : dimtorus ] if a symplectic manifold @xmath14 has an effective hamiltonian @xmath95-action , then @xmath132 .
@xmath133 [ def : toric ] a _ toric symplectic manifold _
is a connected symplectic manifold @xmath134 , equipped with an effective hamiltonian action of the @xmath135-torus , @xmath136 such that the corresponding _ moment map _ @xmath137 well - defined up to a constant , is proper onto its convex image @xmath138 .
[ rem : toric ] the requirement that the moment map be `` proper onto its convex image '' , something that is automatic for compact manifolds , makes the theory of non - compact toric symplectic manifolds analogous to the compact one ( see @xcite ) . [ ex : hamr2ntor ] @xmath64 , equipped with the standard hamiltonian @xmath114-action described in example [ ex : hamr2n ] , is a non - compact toric symplectic manifold .
[ ex : hampntor ] @xmath119 , equipped with the hamiltonian @xmath114-action described in example [ ex : hampn ] , is a compact toric symplectic manifold .
any toric symplectic manifold has an associated convex set , the image of the moment map of the torus action .
the convex sets that arise in this way are characterized in the following definition .
[ def : image ] a convex polyhedral set @xmath139 in @xmath128 is called _ simple _ and _ integral _ if : * there are @xmath135 edges meeting at each vertex @xmath140 ; * the edges meeting at the vertex @xmath140 are rational , i.e. each edge is of the form @xmath141 ; * the @xmath142 in ( 2 ) can be chosen to be a @xmath143-basis of the lattice @xmath144 .
a _ facet _ is a face of @xmath139 of codimension one .
a _ delzant set _ is a simple and integral convex polyhedral set @xmath145 .
delzant polytope _ is a _ compact _ delzant set .
two delzant sets are _ isomorphic _ if one can be mapped to the other by a translation . in 1988
delzant @xcite showed that any delzant polytope determines a unique compact toric symplectic manifold .
more precisely , if two compact toric symplectic manifolds have the same delzant polytope , then there exists an equivariant symplectomorphism between them .
this result can be generalized to the possibly non - compact setting of definition [ def : toric ] ( see @xcite ) .
[ thm : gendel ] let @xmath146 be a toric symplectic manifold , with moment map @xmath147 . then @xmath148 is a delzant set
. two toric symplectic manifolds are equivariant symplectomorphic ( with respect to a fixed torus acting on both ) if and only if their associated delzant sets are isomorphic .
moreover , every delzant set arises from some toric symplectic manifold .
[ rem : labels ] one can use the work of lerman and tolman @xcite to generalize theorem [ thm : gendel ] to orbifolds .
the outcome is a classification of symplectic toric orbifolds via _ labeled delzant sets _ , i.e. convex polyhedral sets , as in definition [ def : image ] , with `` @xmath143-basis '' in ( 3 ) replaced by `` @xmath149-basis '' and with a positive integer label attached to each facet .
each facet @xmath150 of a labeled delzant set @xmath145 determines a unique lattice vector @xmath151 : the primitive inward pointing normal lattice vector . a convenient way of thinking about a positive integer label @xmath152 associated to @xmath150 is by dropping the primitive requirement from this lattice vector : consider @xmath153 instead of @xmath154 . in other words ,
a labeled delzant set can be defined as a rational simple polyhedral set @xmath145 with an inward pointing normal lattice vector associated to each of its facets .
the proof gives an explicit symplectic reduction construction of a canonical model for each toric symplectic manifold , i.e. it associates to each delzant set @xmath139 an explicit toric symplectic manifold @xmath155 with moment map @xmath156 .
one can use these canonical models to derive general properties of toric symplectic manifolds .
for example , let @xmath157 denote the interior of @xmath139 , and consider @xmath158 defined by @xmath159 .
one easily checks that @xmath160 is an open dense subset of @xmath161 , consisting of all the points where the @xmath114-action is free .
it can be described as @xmath162 where @xmath163 are symplectic _ action - angle _ coordinates for @xmath164 , i.e. @xmath165 hence , one has a global equivariant darboux s theorem in this toric context .
note that in these action - angle coordinates the moment map is simply given by @xmath166 i.e. projection in the action coordinates .
[ def : lccs ] a _ compatible complex structure _ on a symplectic vector space @xmath167 is a complex structure @xmath168 on @xmath63 , i.e. @xmath169 with @xmath170 , such that @xmath171 is an inner product on @xmath63 .
this is equivalent to @xmath172 the set of all compatible complex structures on a symplectic vector space @xmath167 will be denoted by @xmath173 . the symplectic linear group @xmath174 acts on @xmath175 by conjugation : @xmath176 this action can be easily seen to be transitive and , if we fix @xmath177 and corresponding inner product @xmath178 , we have that @xmath179 where @xmath180 is the unitary group .
@xmath173 is a symmetric space and admits a beautiful explicit description due to c.l .
siegel @xcite .
[ def : suhs ] the _ siegel upper half space _
@xmath181 is the open _ contractible _ subset of the complex vector space of complex symmetric matrices defined by @xmath182 choose a symplectic basis for @xmath183 , i.e. an isomorphism @xmath184 , and identify the symplectic linear group @xmath174 with the matrix group @xmath185 consisting of @xmath186 real matrices @xmath187 such that @xmath188 where @xmath189 is the matrix form of @xmath190 written in @xmath191 blocks .
we will also write any @xmath192 in @xmath191 blocks : @xmath193 the following proposition is proved in @xcite .
@xmath194 acts on @xmath181 by linear fractional transformations : @xmath195 this action is transitive and the isotropy group of @xmath196 is @xmath197 : @xmath198 hence , @xmath199 given @xmath200 , define @xmath201 under the action of @xmath194 on @xmath181 , we have that @xmath202 let @xmath203 be given by @xmath204 for each @xmath205 , define @xmath206 by @xmath207 this defines a bijection @xmath208 which , up to @xmath209-conjugation , is equivariant with respect to the @xmath194-action on both spaces .
more precisely , if @xmath192 then @xmath210 in particular , @xmath211 is a contractible space and , for any symmetric @xmath191 real matrix @xmath212 , we have that @xmath213 this will be relevant below .
[ def : cacs ] a _ compatible almost complex structure _ on a symplectic manifold @xmath14 is an almost complex structure @xmath168 on @xmath15 , i.e. @xmath214 with @xmath170 , such that @xmath171 is a riemannian metric on @xmath15 .
this is equivalent to @xmath215 and @xmath216 .
the space of all compatible almost complex structures on a symplectic manifold @xmath14 will be denoted by @xmath217 . * the fact that @xmath211 is contractible implies that @xmath217 is non - empty , infinite - dimensional and contractible , for any symplectic manifold @xmath14 .
* a _ khler manifold _ is a symplectic manifold @xmath14 with an integrable compatible complex structure @xmath168 , i.e. one that is locally isomorphic to the standard complex structure @xmath209 on @xmath25 .
note that ( [ eq : r2n - cn ] ) gives the standard isomorphism @xmath218 . *
the space of integrable compatible complex structures on a symplectic manifold @xmath14 will be denoted by @xmath219 . * in general
, @xmath220 can be empty .
[ def : tccs ] a _ toric compatible complex structure _ on a toric symplectic manifold @xmath221 is a @xmath114-invariant @xmath222 .
the space of all such will be denoted by @xmath223 .
it follows from the classification in theorem [ thm : gendel ] , more precisely from the explicit symplectic reduction construction of the canonical model for any compact toric symplectic manifold @xmath221 , that @xmath224 is always non - empty .
it follows from the above bijection between @xmath211 and the siegel upper half space @xmath181 that any @xmath225 can be written in action - angle coordinates @xmath163 on @xmath226 as @xmath227 where @xmath228 and @xmath229 are real symmetric @xmath191 matrices , with @xmath230 positive definite . for _ integrable
_ toric compatible complex structures we have that : @xmath231 any real function @xmath232 is the hamiltonian of a @xmath58-parameter family @xmath233 of @xmath114-equivariant symplectomorphisms .
these are given in action - angle coordinates @xmath163 on @xmath234 by @xmath235 hence , it follows from ( [ eq : conj ] ) that the natural action of @xmath236 on @xmath237 , given by @xmath238 corresponds in the siegel upper half space parametrization to @xmath239 where @xmath240 this implies that , for any integrable @xmath241 , there exist action - angle coordinates @xmath163 on @xmath234 such that @xmath242 , i.e. such that @xmath243 with @xmath244 for some @xmath245 holomorphic coordinates for @xmath168 are given in this case by @xmath246 the corresponding riemannian ( khler ) metric @xmath247 on @xmath234 can the be written in matrix form as @xmath248 with @xmath249 [ def : symppot ] we will call such a potential function @xmath250 the _ symplectic potential _ of both the complex structure @xmath168 and the metric @xmath251 this particular way to arrive at the above local form for any @xmath241 is due to donaldson @xcite , and illustrates a small part of his formal general framework for the action of the symplectomorphism group of a symplectic manifold on its space of compatible complex structures ( cf .
@xcite ) .
[ ex : symppotr2n ] consider the standard linear complex structure @xmath252 given by ( [ eq : j0 ] )
. in action - angle coordinates @xmath163 on @xmath253 its symplectic potential is given by @xmath254 hence , in these action - angle coordinates , the standard complex structure has the matrix form @xmath255 while the standard flat euclidean metric becomes @xmath256 the proof of theorem [ thm : gendel ] associates to each delzant set @xmath257 , via an explicit symplectic reduction construction , a canonical khler toric manifold @xmath258 in @xcite guillemin gave an explicit formula for the symplectic potential of this canonical khler metric . to write it down one
just needs some simple combinatorial data that can be easily obtained directly from the polytope @xmath139 .
let @xmath259 denote the @xmath260-th facet ( codimension-@xmath58 face ) of the polytope @xmath139 .
the affine defining function of @xmath259 is the function @xmath261 where @xmath262 is a primitive inward pointing normal to @xmath259 and @xmath263 is such that @xmath264 .
note that @xmath265 .
[ thm : gui1 ] in appropriate action - angle coordinates @xmath163 , the canonical symplectic potential @xmath266 for @xmath267 is given by @xmath268 where @xmath269 is the number of facets of @xmath139 .
[ ex : symppotr2n-2 ] the symplectic potential presented in example [ ex : symppotr2n ] for the standard flat euclidean metric on @xmath25 is the canonical symplectic potential of the corresponding delzant set @xmath270 . [ ex : symppotpn ] for projective space @xmath39 the polytope @xmath257 can be taken to be the standard simplex , with defining affine functions @xmath271 where @xmath272 .
the canonical symplectic potential @xmath266 , given by @xmath273 defines the standard complex structure @xmath274 and fubini - study metric on @xmath39 .
theorem [ thm : abr2 ] below provides the symplectic version of the @xmath275-lemma in this toric context , characterizing the symplectic potentials that correspond to toric compatible complex structures on a toric symplectic manifold .
it is an immediate extension to our possibly non - compact setting of the compact version proved in @xcite . to properly state it we need the following definition .
[ def : jcomplete ] let @xmath146 be a symplectic toric manifold and denote by @xmath276 the hamiltonian vector fields generating the torus action .
a toric compatible complex structure @xmath277 is said to be _ complete _ if the @xmath168-holomorphic vector fields @xmath278 are complete .
the space of all complete toric compatible complex structures on @xmath146 will be denoted by @xmath279 .
[ thm : abr2 ] let @xmath168 be any complete compatible toric complex structure on the symplectic toric manifold @xmath155 .
then , in suitable action - angle @xmath163-coordinates on @xmath280 , @xmath168 is given by a symplectic potential @xmath281 of the form @xmath282 where @xmath283 is given by theorem [ thm : gui1 ] , @xmath53 is smooth on the whole @xmath139 , and the matrix @xmath284 is positive definite on @xmath157 and has determinant of the form @xmath285 with @xmath286 being a smooth and strictly positive function on the whole @xmath139 .
conversely , any such potential @xmath287 determines a ( not necessarily complete ) complex structure on @xmath288 , that extends uniquely to a well - defined compatible toric complex structure @xmath168 on the toric symplectic manifold @xmath155 .
[ rem : abr3 ] if one takes into account remark [ rem : labels ] , the word `` manifold '' can be replaced by `` orbifold '' in theorems [ thm : gui1 ] and [ thm : abr2 ] ( see @xcite ) .
there is no immediate relation between completeness of a toric compatible complex structure and completeness of the associated toric khler metric .
see remark [ rem : complete ] .
we now recall from @xcite a particular formula for the scalar curvature in action - angle @xmath163-coordinates .
a khler metric of the form ( [ metricg ] ) has scalar curvature @xmath289 given by @xmath290 which after some algebraic manipulations becomes the more compact @xmath291 where the @xmath292 , are the entries of the inverse of the matrix @xmath293 , @xmath294 symplectic potential .
see @xcite for an appropriate interpretation of this formula for the scalar curvature . because the delzant set @xmath257 of a symplectic toric manifold is only really well defined up to translations ( i.e. additions of constants to the moment map ) and @xmath295 transformations ( i.e. changes of basis of the torus @xmath296 )
, symplectic potentials should transform naturally under these type of maps . while the effect of translations is trivial to analyse , the effect of @xmath295 transformations is more interesting .
in fact : symplectic potentials transform quite naturally under any @xmath297 linear transformation .
let @xmath298 and consider the linear symplectic change of action - angle coordinates @xmath299 then @xmath300 becomes @xmath301 with @xmath302 and symplectic potentials transform by @xmath303 the corresponding hessians are related by @xmath304 and @xmath305 [ ex : hirzebruch ] figure [ fig : hirzebruch ] illustrates two equivalent descriptions of a toric symplectic rational ruled @xmath1-manifold or , equivalently , of a hirzebruch surface @xmath306 the linear map @xmath307 relating the two is given by @xmath308 the inward pointing normal that should be considered for each facet is specified .
the polytope on the right is a standard delzant polytope for the hirzebruch surface @xmath309 .
the polytope on the left is very useful for the khler metric constructions of section [ s : calabi ] and was implicitly used by calabi in @xcite .
in real dimension @xmath13 any orientable riemannian manifold is khler , since its area form is a symplectic form and oriented rotation by @xmath310 on each tangent plane is a compatible complex structure . in this section we write down the symplectic potentials that give rise to toric constant ( scalar ) curvature metrics in real dimension @xmath13 and identify the underlying toric symplectic manifolds @xmath15
. this will be a warm - up for the higher dimensional examples presented in section [ s : calabi ] . according to formula ( [ scalarsymp ] ) for the scalar curvature
, we are looking for symplectic potentials @xmath311 such that @xmath312 and @xmath313 where @xmath314 denotes the gauss curvature .
this implies that @xmath315 where @xmath316 is such that @xmath317 .
suppose that @xmath318 .
then @xmath319 this means that @xmath320 and @xmath321 can be written as @xmath322 hence @xmath323 and the metric is given in matrix form by @xmath324 i.e. we get a flat cylinder of radius @xmath325 .
suppose that @xmath326 and @xmath327 .
then , modulo a translation and possible sign change in the action variable @xmath328 , we can assume that @xmath329 , @xmath330 and @xmath331 this means that @xmath332 and @xmath3330,+\infty \right [ \to { \mathbb r}$ ] can be written as @xmath334 if @xmath335 this is the canonical symplectic potential giving the flat euclidean metric on @xmath336 ( cf .
example [ ex : symppotr2n ] ) . in general ,
as explained in @xcite , this is the symplectic potential of a cone metric of angle @xmath337 on @xmath336 , given in matrix form by @xmath338 when @xmath339 with @xmath340 , this corresponds to an orbifold flat metric on @xmath341 ( see @xcite ) .
suppose that @xmath342 .
then , modulo a translation in the action variable @xmath328 , we can assume that @xmath343 and @xmath344 this means that @xmath345 $ ] and @xmath333-\sqrt{c / k } , \sqrt{c / k } \right [ \to { \mathbb r}$ ] can be written as @xmath346\,.\ ] ] if @xmath347 this is the canonical symplectic potential giving the smooth round european football metric of total area @xmath348 and constant gauss curvature @xmath59 on @xmath349 ( cf .
example [ ex : symppotpn ] ) . in general , this is the symplectic potential of a singular american football metric of angle @xmath350 at the poles " .
suppose that @xmath351 .
then , modulo a translation in the action variable @xmath328 , we can assume that @xmath343 and @xmath352 if @xmath353 then @xmath354 , which means that @xmath320 and @xmath321 can be written as @xmath355 this is the symplectic potential of a metric of constant gauss curvature @xmath351 on @xmath356 i.e. an hyperboloid .
if @xmath357 then @xmath358 -\infty , - \sqrt{c / k } \right [ \cup \left ] \sqrt{c / k } , + \infty \right[\,.\ ] ] hence , up to a sign change in the action variable @xmath328 , we may assume that @xmath359 and @xmath360\sqrt{c / k},+\infty\right [ \to{\mathbb r}$ ] can be written as @xmath361 \,.\ ] ] if @xmath347 this is the symplectic potential of the hyperbolic metric of constant gauss curvature @xmath351 on @xmath336 .
in other words , in the action - angle coordinates @xmath163 of this symplectic model , the hyperbolic metric is given in matrix form by @xmath362 more generally , i.e. when @xmath363 , we get singular hyperbolic metrics on @xmath336 , with a cone singularity of angle @xmath364 at the origin .
[ rem : complete ] this case illustrates the fact that there is no immediate relation between completeness of a toric compatible complex structure and completeness of the associated toric khler metric . here
the metric is complete but the complex structure is not .
in fact , it easily follows from ( [ eq : holcoords ] ) that @xmath336 with this complex structure is biholomorphic to an open bounded disc @xmath365 .
if @xmath329 we have that @xmath366 and @xmath3600,+\infty\right [ \to { \mathbb r}$ ] can be written as @xmath367 this is the symplectic potential of a complete hyperbolic cusp metric on @xmath3680,+\infty\right [ \times s^1 $ ] , given in matrix form by @xmath369
in @xcite , calabi introduced the notion of extremal khler metrics .
these are defined , for a fixed closed complex manifold @xmath370 , as critical points of the square of the @xmath371-norm of the scalar curvature , considered as a functional on the space of all symplectic khler forms @xmath16 in a fixed khler class @xmath372 .
the extremal euler - lagrange equation is equivalent to the gradient of the scalar curvature being an holomorphic vector field ( see @xcite ) , and so these metrics generalize constant scalar curvature khler metrics .
moreover , calabi showed in @xcite that extremal khler metrics are always invariant under a maximal compact subgroup of the group of holomorphic transformations of @xmath370 .
hence , on a toric manifold , extremal khler metrics are automatically toric khler metrics , and one should be able to write them down using the previous action - angle coordinates framework . in this section ,
following @xcite , we will do that for the @xmath1-parameter family of @xmath0-invariant extremal khler metrics constructed by calabi in @xcite .
calabi used this family to put extremal khler metrics on @xmath373 for any khler class .
in particular , when @xmath4 , on all hirzebruch surfaces ( cf .
example [ ex : hirzebruch ] ) .
as we will see here , this family can be used to write down many other interesting extremal khler metrics , including the non - compact , cohomogeneity one , constant scalar curvature examples that were later constructed by lebrun @xcite , pedersen - poon @xcite and simanca @xcite .
using the action - angle coordinates set - up for toric sasaki geometry developed in @xcite , one can show @xcite that calabi s family also contains a family of khler - einstein metrics directly related to the sasaki - einstein metrics constructed in 2004 by gauntlett - martelli - sparks - waldram @xcite .
consider symplectic potentials @xmath374 of the form @xmath375 where @xmath376 and @xmath157 will be determined in each of the particular cases that we will consider .
a simple computation shows that @xmath377 where @xmath378 .
it then follows from ( [ scalarsymp ] ) that the scalar curvature of the corresponding toric khler metric is given by @xmath379 the euler - lagrange equation defining an extremal khler metric can be shown to be equivalent to @xmath380 i.e. the metric is extremal if and only if its scalar curvature @xmath289 is an affine function of @xmath328 ( see @xcite ) . requiring that the scalar curvature @xmath381 , given by ( [ eq : scalar - r ] ) , is an affine function of @xmath382
is easily seen to be equivalent to @xmath383 where @xmath384 are the @xmath1 parameters of the family . as shown by calabi in @xcite
, one can determine explicit values for the constants @xmath384 so that the corresponding symplectic potential , given by ( [ eq : sympot - r ] ) , gives rise to an extremal khler metric on @xmath385 in our framework , this can be seen as follows .
up to a @xmath297 transformation , generalizing to higher dimensions the one considered in example [ ex : hirzebruch ] , @xmath12 is determined by a moment polytope @xmath386 with defining affine functions @xmath387 where the real numbers @xmath388 determine the khler class , i.e. the cohomology class of the symplectic form @xmath389 .
hence , if follows from theorem [ thm : abr2 ] that , to determine a toric compatible complex structure on @xmath390 , the symplectic potential has to be of the form @xmath391 where @xmath392 is smooth on @xmath386 . comparing ( [ eq : sympot - r ] ) , ( [ eq : h ] ) and ( [ eq : sympot - p ] ) , one concludes that we must have @xmath393 where @xmath394 is a smooth function on @xmath386 .
this gives rise to a system of @xmath1 linear equations in the @xmath1 unknowns @xmath384 , which admits a unique explicit solution for any @xmath3 and @xmath395 such that @xmath388 ( see page 285 of @xcite or @xcite ) . by construction , all khler metrics in calabi s
@xmath1-parameter family are extremal .
a simple computation shows that their scalar curvature is given by @xmath396 hence , these metrics have @xmath397 and are @xmath398 moreover , one can show that these metrics are @xmath399 and @xmath400 we will now analyse in more detail these constant scalar curvature particular cases .
note that when @xmath401 we have @xmath402 which is the standard symplectic potential of the delzant set @xmath403 and determines the standard flat euclidean metric on @xmath25 ( cf .
example [ ex : symppotr2n ] ) .
assume that @xmath404 and @xmath405 with @xmath406 .
then @xmath407 where @xmath394 is a smooth function on the rational delzant set @xmath408 with defining affine functions @xmath409 the symplectic potential can be written as @xmath410 where @xmath392 is smooth on @xmath408 .
hence , for each @xmath411 , it defines a ricci - flat khler metric on the total space of the canonical line bundle @xmath412 ( as before , up to a @xmath297 transformation , the underlying non - compact toric symplectic manifold is determined by @xmath413 ) .
these are the metrics constructed by calabi in @xcite .
we will now show that calabi s family also contains the complete scalar - flat khler metrics on the total space of the line bundles @xmath414 constructed by lebrun @xcite and pedersen - poon @xcite ( see also simanca @xcite ) . up to a @xmath297 transformation
, these spaces are determined by the rational delzant sets @xmath415 , with @xmath406 and defining affine functions @xmath416 which means that the symplectic potential has to be of the form @xmath417 where @xmath392 is smooth on @xmath418 .
this implies that @xmath419 with @xmath394 smooth on @xmath418 .
since the scalar - flat condition is equivalent to @xmath420 , we get that @xmath421 this relation gives rise to a system of @xmath13 linear equations in the @xmath13 unknowns @xmath422 , which admits a unique solution for any @xmath3 and @xmath423 : @xmath424 note that when @xmath425 we get complete scalar - flat khler metrics on the total space of the line bundle @xmath426 i.e. on @xmath112 blown - up at the origin .
these were originally constructed by d. burns ( at least when @xmath4 ) . when @xmath430 we recover example [ ex : symppotpn ] , i.e. the standard complex structure and fubini - study metric on @xmath39 .
more generally , for any @xmath431 , this defines the standard complex structure and suitably scaled fubini - study metric on @xmath39 .
the corresponding moment polytope is the simplex in @xmath128 with defining affine functions @xmath432 when @xmath433 it follows from theorem [ thm : abr2 ] that the above symplectic potential determines a toric compatible complex structure @xmath434 on the toric symplectic manifold @xmath64 with delzant set @xmath435 .
the corresponding khler metric is a @xmath0-invariant khler - einstein metric of negative scalar curvature on the complex manifold @xmath436 . using the holomorphic coordinates given by ( [ eq : holcoords ] )
, one easily concludes that @xmath436 is biholomorphic to a ball @xmath437 , which implies in particular that @xmath434 is not complete .
moreover , the khler metric @xmath438 is , in fact , the well - known and complete bergman metric .
we will now show that calabi s family also contains the complete khler - einstein metrics with negative scalar curvature on the total space of the open disc bundles @xmath439 constructed by pedersen - poon @xcite .
as toric symplectic manifolds , and up to a @xmath297 transformation , these spaces are again determined by the rational delzant sets @xmath415 with defining affine functions @xmath416 which means that the symplectic potential has to be of the form @xmath417 where @xmath392 is smooth on @xmath418 .
this implies that @xmath419 with @xmath394 smooth on @xmath418 .
since the khler - einstein condition is equivalent to @xmath440 , we get that @xmath441 this relation gives rise to a system of @xmath13 linear equations in the @xmath13 unknowns @xmath442 , which admits a unique solution for any @xmath11 and @xmath423 : @xmath443 as remarked by pedersen - poon , these metrics are a superposition of calabi s ricci - flat metrics ( @xmath444 ) and bergman metrics ( @xmath433 ) .
the analogous superposition of calabi s ricci - flat metrics ( @xmath444 ) with fubini - study metrics ( @xmath431 ) gives rise to khler - einstein metrics on the projectivization of the above line bundles , with cone - like singularities in the normal directions to the zero and infinity sections .
as explained in @xcite , these metrics are directly related to the smooth sasaki - einstein metrics constructed in 2004 by gauntlett - martelli - sparks - waldram @xcite .
we will now show that calabi s family also contains the complete constant negative scalar curvature khler metrics on the total space of the open disc bundles @xmath445 constructed again by pedersen - poon @xcite . as before ,
we are interested in the rational delzant sets @xmath415 and symplectic potentials of the form @xmath417 with @xmath392 smooth on @xmath418 . assuming @xmath446 and @xmath447 , i.e.
@xmath448 , this implies that @xmath449 with @xmath394 smooth on @xmath418 .
this relation gives rise to a system of @xmath13 linear equations in the @xmath13 unknowns @xmath422 , which admits a unique solution for any @xmath3 and @xmath423 : @xmath450 abreu m. , _ khler geometry of toric manifolds in symplectic coordinates _
, in : symplectic and contact topology : interactions and perspectives , y.eliashberg , b.khesin and f.lalonde ( eds ) , fields institute communications 35 , american mathematical society 2003 , pp 124 . | in these notes , after an introduction to toric khler geometry , we present calabi s family of @xmath0-invariant extremal khler metrics in symplectic action - angle coordinates and show that it actually contains , as particular cases , many interesting cohomogeneity one examples of constant scalar curvature . partially supported by the fundao para a cincia e a tecnologia ( fct / portugal ) . . |
recently it becomes possible @xcite to compute the exact partition functions of four - dimensional @xmath4 supersymmetric gauge theories .
remarkably , the celebrated seiberg - witten solutions @xcite of the gauge theories emerge @xcite through the statistical models of random partitions .
@xmath4 supersymmetric gauge theories are realized @xcite by type iia strings on certain calabi - yau threefolds .
the topological vertex @xcite provides a powerful method to compute all genus topological @xmath5-model partition functions on local toric calabi - yau threefolds .
the string amplitudes on the relevant threefolds are computed @xcite by using this and shown to reproduce the gauge instanton contributions to the exact partition functions of the gauge theories .
it turns out surprisingly that the topological vertex is identified @xcite with the partition function of a model of random plane partitions @xcite .
this interpretation is further explored in @xcite .
the appearance of random plane partitions in string theory suggests their possible relation to the gauge theories .
nevertheless , the exact solutions of the gauge theories have not been studied from the viewpoint of random plane partitions . in this article , we investigate five - dimensional @xmath0 supersymmetric yang - mills theories from the perspective of random plane partitions .
a plane partition @xmath6 is an array of non - negative integers satisfying @xmath7 and @xmath8 , and identified with the three - dimensional young diagram as depicted in figure 1-(a ) .
the diagram is also regarded as a sequence of partitions @xmath9 , where @xmath10 .
see figure 1-(b ) . among the series of partitions , @xmath11 will be called the main diagonal partition of @xmath12 .
we consider the following model of random plane partitions .
@xmath13 where @xmath1 and @xmath14 are indeterminates . @xmath15 and @xmath16 denote respectively the total numbers of cubes and boxes of the corresponding diagrams . by an identification of @xmath1 and @xmath14 with the relevant string theory parameters , the partition function can be converted into topological @xmath5-model string amplitude on the local calabi - yau threefold @xmath17 .
[ 3d young diagram ] the above model has an interpretation as random partitions .
it can be seen by rewriting the partition function as @xmath18 where the summation over plane partitions in ( [ z(q , q ) ] ) are divided into two branches .
the partitions @xmath19 are thought as the ensemble of the model by summing first over the plane partitions whose main diagonal partitions are @xmath19 . as the model of random partitions
, it becomes a @xmath1-deformation of the standard plancherel measured random partitions .
it is well known that partitions are realized as states of two - dimensional free fermions . by using the folding @xcite of a single complex fermion into @xmath20-component ones ,
each partition can be expressed as a set of @xmath20 charged partitions @xmath21 , where @xmath22 are partitions and @xmath23 are the @xmath24 charges of the @xmath20-component fermions .
this correspondence is used @xcite to study the exact partition functions for four - dimensional @xmath4 supersymmetric gauge theories . among partitions , these coming from the charged empty partitions @xmath25 turn out to play special roles .
we call them the ground partitions . regarding the model as the @xmath1-deformed random partitions , we write the boltzmann weights for the ground partitions by @xmath26
. we will factor the partition function into @xmath27 this factorization turns out useful to find out the gauge theoretical interpretations .
the relevant field theory parameters are @xmath28 and @xmath29 , where @xmath30 are the vacuum expectation values of the adjoint scalar in the vector multiplet and @xmath31 is the scale parameter of the underlying four - dimensional theory .
@xmath29 is the radius of @xmath32 in the fifth dimension .
we identify these parameters with @xmath33 and @xmath23 in ( [ factored z(q , q ) ] ) as follows ; @xmath34 where @xmath35 . the parameter @xmath36 , which can be thought as a chemical potential for plane partitions , is often identified with the string coupling constant @xmath37 . via the above identifications we derive @xmath38 where the rhs is the exact partition function @xcite for five - dimensional @xmath0 supersymmetric @xmath39 yang - mills with the chern - simons term @xcite .
the five - dimensional theory is living on @xmath40 .
the chern - simons coupling constant @xmath41 is quantized to @xmath20 .
the above can be said more precisely as follows ; @xmath42 \sum_{\{\lambda^{(r)}\ } } z_q^{su(n)\,inst } ( \{\lambda^{(r)}\},\{p_r\},\,q)&= & z^{su(n)\,inst}_{\,5d\,\mbox{\tiny{sym } } } ( \{a_r\};\,\lambda , r,\hbar ) , \label{z 5dsym inst}\end{aligned}\ ] ] where the rhss are respectively the perturbative part and the instanton part of the exact partition function .
the gauge theory partition function is obtained as the @xmath43 limit of the above partition function .
the ground partition in ( [ z 5dsym ] ) becomes very large since its size is @xmath44 and the @xmath24 charges @xmath23 scale as @xmath45 at the field theory limit .
the equality ( [ z 5dsym pert ] ) shows that these ground partitions describe the perturbative regime of the coloumb branch .
@xmath22 in ( [ z 5dsym ] ) represent @xmath24 instantons with @xmath46 on a non - commutative @xmath47 @xcite .
they can be viewed as excitations from the ground partition .
these excitations emerge as the quantum ( non - perturbative ) deformation of the perturbative coloumb branch when @xmath22 are comparable to the ground partition .
we start section 2 with a brief review on the transfer matrix approach @xcite to random plane partitions . the perspective of the @xmath1-deformation is emphasized . in section 3
we examine the factorization ( [ factored z(q , q ) ] ) for the case of @xmath48 and confirm the equality ( [ z 5dsym inst ] ) .
section 4 is devoted to prove the equality ( [ z 5dsym pert ] ) for this case .
the discussion is also applicable to the cases of the higher rank gauge groups . in section 5
we treat the cases of @xmath39 .
the instanton parts can be converted to topological @xmath5-model string amplitudes on the relevant local calabi - yau threefolds .
these amplitudes are summarized in appendix .
a plane partition @xmath12 is an array of non - negative integers @xmath49 satisfying @xmath50 and @xmath51 for all @xmath52 .
plane partitions are identified with the three - dimensional young diagrams .
the three - dimensional diagram @xmath12 is a set of unit cubes such that @xmath53 cubes are stacked vertically on each @xmath54-element of @xmath12 .
the size of @xmath12 is @xmath55 , which is the total number of cubes of the diagram .
each diagonal slice of @xmath12 becomes a partition , that is , a sequence of weakly decreasing non - negative integers .
let @xmath9 be a partition along the @xmath56-th diagonal slice .
@xmath57 in particular @xmath58 is the main diagonal partition .
the series of partitions @xmath9 satisfies the condition @xmath59 where @xmath60 means the following interlace relation between two partitions @xmath61 and @xmath62 @xmath63 we have @xmath64 , where the size of a partition @xmath19 is also denoted by @xmath65 . a model of random plane partitions relevant to describe five - dimensional @xmath66 supersymmetric @xmath24 gauge theory
is defined by the following partition function . @xmath67
the boltzmann weight consists of two parts .
the first contribution comes from the energy of plane partitions @xmath12 , and the second contribution is a chemical potential for the main diagonal partitions @xmath11 . to contact with the @xmath24 gauge theory we will identify the indeterminates @xmath1 and @xmath14 with the following field theory parameters . @xmath68 where @xmath29 is the radius of @xmath32 in the fifth dimension and @xmath31 denotes the scale parameter of the underlying four - dimensional theory .
the parameter @xmath36 is often identified with string coupling constant @xmath37 .
the transfer matrix approach @xcite allows us to express the random plane partitions ( [ z u(1 ) ] ) in terms of two - dimensional conformal field theory ( 2@xmath69 free fermion system ) .
the partition function can be computed exactly by using the standard technique of 2@xmath69 cft .
it is well known that partitions are realized as states of 2@xmath69 free fermions by using the maya diagrams .
let @xmath70 and @xmath71 be complex fermions with the anti - commutation relations @xmath72 let @xmath73 be a partition .
the maya diagram is the set @xmath74 which is defined by @xmath75 the correspondence with the young diagram is depicted in figure 2 .
[ maya diagram ] by using the maya diagram the partition can be mapped to the following fermion state .
@xmath76 where @xmath77 is the length of @xmath19 , that is , the number of the non - zero @xmath78 . in ( [ |mu;n > ] ) the state @xmath79 is the ground state of the charge @xmath80 sector .
it is defined by the conditions @xmath81 we mainly consider the @xmath82 sector in the below .
it is also possible to regard @xmath36 as a chemical potential for plane partitions .
thermodynamic limit of ( [ z u(1 ) ] ) is obtained by letting @xmath43 .
let us recall that the mean values of @xmath15 and @xmath16 are respectively given by @xmath92 and @xmath93 .
it follows from ( [ macmahon ] ) that they behave @xmath94 and @xmath95 as @xmath43 ( @xmath96 ) .
therefore a typical plane partition @xmath12 at the limit @xmath43 is a plane partition of order @xmath97 , and its main diagonal partition @xmath11 becomes a partition of order @xmath98 .
we can interpret the random plane partitions ( [ z u(1 ) ] ) as a model of random partitions .
it is identified with a @xmath1-deformation of the standard plancherel measured random partitions . ( the deformation is different from @xcite . ) to see this , we rewrite the partition function by using the schur functions .
an insertion of the unity @xmath99 factorizes ( [ z u(1 ) transfer matrix ] ) into @xmath100 the matrix elements in the above turn to be @xmath101 where @xmath102 is the schur function @xmath103 specialized at @xmath104 ( @xmath105 ) .
therefore we obtain @xmath106 where the multiple index @xmath107 is used .
the expression ( [ z u(1 ) schur ] ) allows us to interpret ( [ z u(1 ) ] ) as a model of @xmath1-deformed random partitions .
it is also clear from ( [ factorized z u(1 ) transfer matrix ] ) that partitions @xmath19 are the main diagonal partitions @xmath11 .
the four - dimensional limit of the model is obtained by letting @xmath108 under the identification ( [ u(1 ) q and q ] ) . to obtain the limit of ( [ z u(1 ) schur ] ) the following product formula of the schur function @xcite becomes useful .
@xmath109 where @xmath110 is the hook length of the box @xmath54 in the young diagram , and @xmath111 . by using ( [ hook length formula 2 ] )
we can see from ( [ z u(1 ) schur ] ) that @xmath112 this is the random partitions @xcite which describes four - dimensional @xmath113 supersymmetric @xmath24 gauge theory on non - commutative @xmath47 .
the partition function ( [ z u(1 ) ] ) can be converted to topological string amplitude on a certain non - compact calabi - yau threefold by identifying @xmath1 and @xmath14 with the string theory parameters .
let us consider the topological @xmath5-model on @xmath17 .
all genus @xmath5-model partition function on this local geometry is computed @xcite from the @xmath114-theory viewpoint .
it is given by @xmath115 where the string coupling constant @xmath37 and the khler parameter @xmath116 of the @xmath117 appear as @xmath118 and @xmath119 .
the topological vertex @xcite makes it possible to compute all genus @xmath5-model partition functions on local toric calabi - yau threefolds by using diagrammatic techniques like the feynman rules , where the diagrams are the dual toric diagrams of the local geometries .
the diagram for the above local geometry is described in figure 3 .
a computation using the topological vertices presents the amplitude in the same form as ( [ z u(1 ) schur ] ) . in that expression , partitions
@xmath19 are attached to the @xmath117 .
[ u(1 ) geometry ]
the fock representation of a single complex fermion has @xcite an alternative realization by exploiting @xmath20 complex fermions : @xmath120 with the anti - commutation relations @xmath121 @xmath122 and @xmath123 are obtained from @xmath124 and @xmath125 by the following identifications . @xmath126 where @xmath127 .
it is convenient to consider partitions paired with the @xmath24 charges .
we denote such a charged partition by @xmath128 , where @xmath19 is a partition and @xmath80 is the @xmath24 charge .
the states @xmath129 constitute bases of the fock space of a single complex fermion .
thanks to the above realization of @xmath20 fermions , we can express @xmath128 uniquely by means of @xmath20 charged partitions @xmath21 and vice versa through the identification @xmath130 the conservation of the @xmath24 charges implies @xmath131 . the correspondence between the partitions @xmath19 and @xmath22 can be described explicitly by using the maya diagrams ( shifted by the @xmath24 charges ) .
it becomes @xmath132 where @xmath133 .
in particular , when @xmath82 , the following information on the partitions is obtainable from the above correspondence by applying the method of power - sums @xcite . @xmath134 the maya diagram @xmath135 is considered as the subset of @xmath74 by ( [ n-1 maya diagrams ] ) .
it is also a subset of @xmath136 ( @xmath137 ) .
we may colour the set @xmath136 just by attaching the number @xmath138 to all the elements .
@xmath135 gets coloured by @xmath138 .
( [ n-1 maya diagrams ] ) shows that @xmath74 has @xmath20 colours , where the number @xmath138 is attached to these elements coming from @xmath135 .
let us begin with the case of two - component fermions .
owing to the identification ( [ n-1 states ] ) any charged partition @xmath128 can be expressed by @xmath139 and @xmath140 .
we consider the neutral case , that is , @xmath82 .
@xmath141 and @xmath142 determine a partition in the neutral sector .
we denote such a partition by @xmath143 .
the corresponding maya diagram @xmath144 can be read from ( [ n-1 maya diagrams ] ) as @xmath145 in the present case , the quantities ( [ n-1 size k ] ) turn to be @xmath146 \kappa(\mu(\lambda^{(1)},\lambda^{(2)};p))&= & 4(\kappa(\lambda^{(1 ) } ) + \kappa(\lambda^{(2 ) } ) ) + 2(4p-1)(|\lambda^{(1)}|-|\lambda^{(2)}| ) .
\label{2 - 1 size k}\end{aligned}\ ] ] we regard the random plane partitions ( [ z u(1 ) ] ) as the @xmath1-deformed random partitions via the expression ( [ z u(1 ) schur ] ) and rewrite the partition function as @xmath147 subsequently we will further factor the above into the following form .
@xmath148 [ n=2 ground partition ] taking account of the contributions of the ground partitions to ( [ z u(1 ) two - component ] ) ,
we compute the following quantities .
@xmath154 the hook length formula ( [ hook length formula 2 ] ) becomes useful for the computation . by plugging the hook length of the boxes ( figure 4 ) into the formula ,
we evaluate ( [ def z pert ] ) as follows .
@xmath155 for @xmath153 .
the similar expression is also obtainable for @xmath156 .
we thus get @xmath157 1 & \
p=0 \\[0.5 cm ] \displaystyle { \prod_{k=1}^{2|p| } \left\ { \frac{q^{\frac{1}{2}}}{q_f^{\frac{1}{2}}q^k -q_f^{-\frac{1}{2}}q^{-k } } \right \}^k } & \
p\leq -1 \end{array } \right . \ \ , \label{z pert}\end{aligned}\ ] ] where we put @xmath158 .
we factorize the partition function ( [ z u(1 ) two - component ] ) in the following form .
@xmath159 where we define @xmath160 via the relations @xmath161 owing to ( [ def z pert ] ) we can rewrite the above as @xmath162 the expression ( [ z su(2 ) inst for gauge ] ) makes it possible to rewrite @xmath160 in a form convenient to compare with the topological string amplitude .
the first two infinite products have the same form . by noting the relation @xmath178 , the product formula ( [ product formula schur ] )
gives @xmath179 as regards the last infinite products in ( [ z su(2 ) inst for gauge ] ) , the following description is obtainable by the standard computation of 2@xmath69 cft .
@xmath180 & & = q^{\frac{1}{2 } \left(\kappa(\lambda^{(1)})-\kappa(\lambda^{(2)})\right ) } q_f^{\frac{1}{2}\left(|\lambda^{(1)}|+|\lambda^{(2)}|\right ) } \prod_{k=1}^{\infty}(1-q_f q^{2k})^{k } \sum_{\nu}q_f^{|\nu| } s_{\nu}(q^{2(\lambda^{(1)}+\rho ) } ) s_{\nu}(q^{2(\widetilde{\lambda^{(2)}}+\rho ) } ) , \nonumber \\ * \label{su(2 ) infinite product 2}\end{aligned}\ ] ] where @xmath181 denotes the partition conjugate to @xmath182 . finally , putting together ( [ su(2 ) infinite product 1 ] ) and ( [ su(2 ) infinite product 2 ] ) , we can rewrite ( [ z su(2 ) inst for gauge ] ) as follows ; @xmath183 we fix @xmath150 .
the relevant field theory parameters are @xmath184 and @xmath29 , where @xmath185 are the vevs of the adjoint scalar in the vector multiplet .
we identify @xmath186 and @xmath14 with @xmath184 and @xmath29 as follows .
@xmath187 since we have set @xmath158 the above implies @xmath188 where @xmath189 .
the identifications ( [ su(2 ) ft parameter ] ) convert @xmath160 to the instanton contributions in five - dimensional gauge theories .
this follows by rephrasing the expression ( [ z su(2 ) inst for gauge ] ) in terms of the field theory parameters .
in particular , we obtain @xmath190 the rhs is the instanton part of the exact partition function @xcite for five - dimensional @xmath66 supersymmetric @xmath48 yang - mills with the chern - simons term @xcite .
the chern - simons corrections come from the following factor in ( [ z su(2 ) inst for gauge ] ) .
@xmath191 which is also understood @xcite as a part of the so - called framing factor of the topological string vertices @xcite .
the chern - simons term is quantized with the coupling constant @xmath41 taking integral values . in the present case
, it becomes @xmath192 . in the next section
we will show that the square of @xmath193 is translated to the perturbative part of the partition function for the supersymmetric yang - mills . taking it for granted for a while , together with ( [ z su(2 ) sym inst ] ) we find @xmath194 the rhs is the exact partition function @xcite for the five - dimensional supersymmetric @xmath48 yang - mills with the chern - simons term .
we notice that , as confirmed @xcite for four - dimensional @xmath4 supersymmetric gauge theories , the gauge theory partition function is realized as the @xmath43 limit of the above partition function .
it is clear from ( [ z su(2 ) sym ] ) that the field theory limit is the thermodynamic limit with @xmath195 and @xmath29 fixed .
the identification ( [ su(2 ) a and p ] ) makes @xmath196 . the corresponding ground partition in ( [ z su(2 ) sym ] ) becomes very large .
the partitions @xmath22 represent @xmath24 gauge instantons with @xmath46 on a non - commutative @xmath47 . in the model of random partitions
, it follows from ( [ 2 - 1 maya diagrams ] ) that these gauge instantons become excitations from the ground partition .
they provide the deformation of the ground partition at the thermodynamic limit ( figure 5 ) .
these imply that the ground partitions describe the perturbative regime of the coloumb branch while their deformation becomes non - perturbative in the gauge theory .
[ pileup ] the relevant non - compact calabi - yau threefold is the local @xmath197 .
the geometric engineering @xcite dictates that four - dimensional @xmath4 supersymmetric @xmath48 yang - mills is realized by the canonical bundle over the hirzebruch surface @xmath198 .
it is an ale space with @xmath199 singularity fibred over @xmath117 .
type of the fibrations of the ale space is labelled by the integer @xmath56 , which is called the framing and taking values @xmath200 .
we choose the framing to be @xmath201 .
the geometrical data are the khler volumes of the compact two - cycles .
we denote the khler parameters of the base @xmath117 and the fibre @xmath117 ( the blow - up cycle of @xmath199 singularity ) respectively by @xmath202 and @xmath203 .
topological string amplitudes on the geometries dictated by the geometric engineering are computed in @xcite by using the topological vertices .
we summarize these amplitudes in appendix .
since they are generating functions of the world - sheet instantons , string coupling constant @xmath37 and the khler parameters appear in the amplitudes as certain combinations of @xmath118,@xmath204 and @xmath205 .
let us denote the topological string amplitude on the local @xmath197 by @xmath206 .
it is given in ( [ su(2 ) string amplitude ] ) . by comparing ( [ z su(2 ) inst for string ] ) with ( [ su(2 ) string amplitude ] ) , we see that the gauge instanton contribution ( [ z su(2 ) sym inst ] ) can be converted to the string amplitude as follows .
@xmath207 where we put @xmath208 .
this is in accord with the result @xcite that the topological string amplitude on the local @xmath197 leads to the instanton part of the partition function of five - dimensional supersymmetric @xmath48 yang - mills . due to the identification @xmath209 , the thermodynamic limit is in the perturbative regime of string theory .
the previous consideration about the role of the ground partitions in gauge theory could be translated in string theory .
the ground partitions at the thermodynamic limit must be interpreted as classical objects in string theory .
the relation @xmath210 , which follows from ( [ su(2 ) ins vs string ] ) , suggests that the ground partitions describe the resolution of @xmath199 singularity in the local @xmath197 .
the aim of this section is to show that the square of @xmath211 in ( [ z su(2 ) sym ] ) describes the perturbative part of the partition function for the five - dimensional supersymmetric yang - mills .
let @xmath212 be the partition @xmath213 , where @xmath80 is a positive integer .
the ground partition @xmath151 for @xmath214 is @xmath215 .
is @xmath216 .
the discussion presented below goes as well in these cases .
] let us consider the logarithm of @xmath217 .
it can be written as follows .
@xmath218 where @xmath1 and @xmath14 are translated to @xmath31 and @xmath29 by ( [ su(2 ) ft parameter ] ) .
we also note @xmath219 .
we further factor the logarithm as @xmath220 where we introduce @xmath221 the function ( [ g ] ) becomes identical to the function @xmath222 given in @xcite . to confirm this
, it should be tested first of all by the q - difference equation @xcite .
[ q - difference equation ] the function ( [ g ] ) satisfies the q - difference equation @xmath223 let us verify ( [ q - diff ] ) by using a recurrence relation among the young diagrams @xmath224 .
the young diagrams @xmath225 and @xmath226 are piled in figure 6-(a ) so that any two boxes at a common position have the same hook length . while in figure 6-(b ) , two @xmath215 are slightly shifted to cover the piled diagrams of ( a ) .
each box of the diagrams has the same hook length as that of @xmath225 at the same position . by noting ( [ ln pert ] ) ,
we exponentiate the lhs of ( [ q - diff ] ) to @xmath227 it follows from the above explanation of the figures that nearly all the products which come from @xmath225 and @xmath226 in ( [ q - difference equation diagrams ] ) cancel with the products coming from two @xmath215 , and that the net becomes only the contribution from the shaded box in figure 6-(c ) . hence , ( [ q - difference equation diagrams ] ) is equal to @xmath228 .
this turns to give the rhs of ( [ q - diff ] ) .
[ q - difference ] [ gamma n - o ] the function @xmath229 coincides with the function @xmath230 in @xcite , up to a linear function in @xmath231 . to prove the above we rewrite ( [ g ] ) as @xmath232
the first two terms become @xmath233 as regards the last term of ( [ hook ] ) we rewrite the finite sum as a subtraction between two infinite sums and then expand the logarithms @xmath234 the summations over @xmath235 in the above give rise to @xmath236 where the first term provides an analogue of the mcmahon function , and the last two terms are a linear function in @xmath231 . collecting ( [ polynomial of g ] ) and ( [ ln of g ] ) , we obtain the following expression for @xmath229 .
@xmath237 thus up to a linear function in @xmath231 , the function ( [ g ] ) is identified with the perturbative term given in @xcite .
we lastly remark that @xmath238 has the smooth @xmath239 limit , as is obvious from ( [ g ] ) . by the same argument as above we
can also express @xmath240 in a form analogous to ( [ vs.gamma1 ] ) .
it becomes as follows .
@xmath241 combining ( [ vs.gamma1 ] ) and ( [ vs.gamma2 ] ) , we arrive at @xmath242 where @xmath243 and @xmath244 are the collections of @xmath245 and @xmath246 in ( [ vs.gamma1 ] ) and ( [ vs.gamma2 ] ) . in terms of the vev of the adjoint scalar ( [ su(2 ) a and p ] )
the above becomes @xmath247 the expression ( [ vs.pert 2 ] ) shows that the square of @xmath248 can be identified with the perturbative part of the partition function for the five - dimensional supersymmetric yang - mills .
the previous discussions on the @xmath48 gauge theory could be generalized to the higher rank gauge groups .
owing to the identification ( [ n-1 ] ) of @xmath20-component fermions any charged partition @xmath128 can be expressed by a set of @xmath20 charged partitions @xmath21 .
we consider in the neutral sector , that is , @xmath249 .
@xmath21 determine a partition @xmath250 in this sector .
the corresponding maya diagram @xmath251 can be read from ( [ n-1 maya diagrams ] ) as @xmath252 in this setting it is convenient to colour - code a partition .
we @xmath20-colour the upper edges of the young diagram by attaching @xmath138 to the edge when the projection of the middle point belongs to @xmath253 . thereby rows and columns of the young diagram are @xmath20-coloured .
see figure 7 for the case of @xmath254 .
the box @xmath54 in the young diagram is bicoloured by @xmath255 , where @xmath256 and @xmath257 are respectively the colors of the @xmath258-th row and the @xmath259-th column .
the box bicoloured by @xmath260 is simply called @xmath260-box . we regard the random plane partitions ( [ z u(1 ) ] ) as the @xmath1-deformed random partitions via ( [ z u(1 ) schur ] ) and rewrite the partition function as @xmath261 similarly to the case of @xmath48 , we will further factor the above into the following form .
@xmath262 we consider the cases of the empty partitions @xmath263 for each @xmath264 , the corresponding partition @xmath265 is named the @xmath39 ground partition .
the case of @xmath254 is illustrated in figure 7 .
[ n=3 ground partition ] their contributions to the partition function lead us to introduce @xmath266 by @xmath267 in other words , we put @xmath268 this follows from ( [ def z su(n ) pert ] ) by applying the formula ( [ hook length formula 2 ] ) and noting measures asymmetry of the young diagram . ]
@xmath269 .
let us compute @xmath270 given in the above .
it is instructive to start with the case of @xmath254 . to simplify the discussion
, we order the three @xmath24 charges @xmath271 as @xmath272 .
they satisfy @xmath273 by the neutral condition . the ground partition for @xmath274 is bicoloured by @xmath275 and @xmath276 .
we factor the products in ( [ def z su(n ) pert hook ] ) according to the color - coding of the partition as follows .
@xmath277 where we depict @xmath278-box as @xmath279 , @xmath280-box as @xmath281 and @xmath276-box as @xmath282 .
each term of ( [ factor n=3 grand hook ] ) can be computed so that it acquires a form analogous to ( [ calculation of zpert in su(2 ) ] ) . for instance , the first term becomes @xmath283 these give the following expression for @xmath284 ; @xmath285 for the cases of general values of @xmath20 , we also order the @xmath24 charges as @xmath286 .
they satisfy @xmath287 . the @xmath39 ground partition for @xmath288 is now bicoloured by @xmath260 with @xmath289 .
see figure 8 for the illustration .
we factor the products in ( [ def z su(n ) pert hook ] ) according to the color - coding of the partition as in the case of @xmath254 .
@xmath290 the contribution of the @xmath260-boxes can be computed in a form similar to ( [ component n=3 grand hook ] ) .
we thus obtain the following expression for @xmath270 ; @xmath291 where we put @xmath292 for @xmath293 .
we factorize the partition function ( [ z u(1 ) n - component ] ) in the following form .
@xmath294 where @xmath295 are defined by the relations @xmath296 owing to ( [ def z su(n ) pert ] ) we can write the above as @xmath297 let us compute @xmath295 .
we order @xmath298 for simplicity .
the ratio of the schur functions in ( [ def z su(n ) inst schur ] ) is translated to infinite products by applying the formula ( [ product formula schur ] ) . by using the description ( [ su(n ) maya diagrams ] ) we can classify the ingredients of the products according as they are made of @xmath299 where @xmath293 , or @xmath300 where @xmath289 .
we then factor the rhs of ( [ def z su(n ) inst schur ] ) by this classification . taking ( [ n-1 size k ] ) into account , we obtain the following expression for @xmath295 .
@xmath301 the expression ( [ z su(n ) inst for gauge ] ) makes it possible to rewrite @xmath295 in a form convenient to compare with the topological string amplitude .
we first notice the identity @xmath302 which follows from the product formula ( [ product formula schur ] ) .
the identity ( [ su(2 ) infinite product 2 ] ) of the @xmath48 theory has the following generalization .
@xmath303 the substitution of ( [ su(n ) formula 1 ] ) and ( [ su(n ) formula 2 ] ) into ( [ z su(n ) inst for gauge ] ) gives rise to some simple powers of @xmath1 and @xmath304 in addition to a bunch of the schur functions .
the exponent of @xmath1 can be read as @xmath305 while the exponents of @xmath304 become @xmath306 together with the contributions of the schur functions from ( [ su(n ) formula 1 ] ) and ( [ su(n ) formula 2 ] ) we finally obtain @xmath307 we fix the @xmath24-charges @xmath23 .
the relevant field theory parameters are @xmath28 and @xmath29 , where @xmath30 @xmath308 are the vevs of the adjoint scalar in the vector multiplet .
we identify @xmath309 and @xmath14 with @xmath28 and @xmath29 as follows .
@xmath310 since we have set @xmath311 the above implies @xmath312 by the identifications ( [ su(n ) ft parameter ] ) , @xmath295 are translated to the instanton contributions in five - dimensional gauge theories .
this follows by rephrasing the expression ( [ z su(n ) inst for gauge ] ) in terms of the field theory parameters .
in particular , we obtain @xmath313 the rhs is the instanton part of the partition function @xcite for five - dimensional @xmath66 supersymmetric @xmath39 yang - mills with the chern - simons term @xcite .
the chern - simons corrections come from the following factor in ( [ z su(n ) inst for gauge ] ) .
@xmath314 which is also understood @xcite as a part of the so - called framing factor of the topological string vertices @xcite .
the chern - simons coupling constant becomes @xmath315 .
similarly to the @xmath48 case , the gauge theory is realized as the @xmath43 limit of the above partition function .
it is the thermodynamic limit with @xmath324 and @xmath29 fixed .
the corresponding ground partition in ( [ z su(n ) sym ] ) becomes very large and the @xmath24 instantons @xmath22 provide the deformation of the ground partition .
hence we can consistently say that the ground partitions describe the perturbative regime of the coloumb branch while their deformation is non - perturbative in the gauge theory .
the relevant non - compact calabi - yau threefold is an ale space with @xmath325 singularity fibred over @xmath117 .
type of the fibrations is labelled by the framing @xmath326 $ ] .
we choose the framing to be @xmath327 .
the geometrical data are the khler parameters @xmath202 and @xmath328 , where @xmath329 .
they correspond respectively to the khler volumes of the base @xmath117 and the @xmath330 blow - up cycles of @xmath325 singularity in the fibre .
the topological string amplitude on this local geometry is given in ( [ su(n ) string amplitude ] ) .
we denote the amplitude as @xmath331 , where @xmath332 and @xmath204 . by comparing ( [ z su(n ) inst for string ] ) with ( [ su(n ) string amplitude ] ) , we see that the gauge instanton contribution ( [ z su(n ) sym inst ] ) is converted to the string amplitude as follows .
@xmath333 where we put @xmath311 .
this is in accord with the result @xcite that the topological string amplitude on this local calabi - yau geometry leads to the instanton part of the partition function of five - dimensional supersymmetric @xmath39 yang - mills . due to the identification @xmath209
, the thermodynamic limit is in the perturbative regime of string theory .
the ground partitions at the thermodynamic limit should be interpreted as classical objects in string theory .
the relations @xmath334 , which follow from ( [ su(n ) ins vs string ] ) , suggest that the @xmath39 ground partitions describe the resolutions of @xmath325 singularity in this local geometry .
this identification allows us to interpret the perturbative chern - simons correction as the triple intersections of the four - cycles @xcite in the local geometry .
we present the topological string amplitudes related to four - dimensional @xmath4 supersymmetric gauge theories .
the geometric engineering @xcite dictates that four - dimensional @xmath4 supersymmetric @xmath39 yang - mills is realized by an ale space with @xmath325 singularity fibered over @xmath117 .
the fibration over @xmath117 is labeled by an integer @xmath335 $ ] , which is often called the framing .
we consider the case of @xmath327 . the geometric data turn to be related with the field theory parameters .
the khler parameter @xmath202 of the base @xmath117 is proportional to @xmath336 , where @xmath337 is the gauge coupling constant at the string scale .
the gauge instanton effect is weighted with @xmath338 .
this leads to @xmath339 .
the khler parameters @xmath328 of the blow - up cycles in the fibre are proportional to @xmath30 , the vevs of the adjoint scalar in the vector multiplet .
the topological vertex @xcite provides a powerful method to compute all genus topological @xmath5-model partition functions on local toric calabi - yau threefolds .
the computations are carried out by using diagrammatic techniques like the feynman rules .
the diagrams are the dual toric diagrams consisting of trivalent vertices .
topological string amplitudes on @xmath340 with open - string boundary conditions are attached to the vertices .
they are derived based on the large @xmath20 duality @xcite and the boundary conditions turn out to be fixed by three partitions . gluing the vertices by a certain rule ,
one obtains the topological string amplitudes on the threefolds .
the diagram which describes the above local geometry for the @xmath39 gauge theory is depicted in figure 9 .
the method of the topological vertex gives the amplitude in the following form .
@xmath341 & & = \sum_{\lambda^{(1)},\cdots,\,\lambda^{(n ) } } \left [ \begin{array}{l } q_b^{\frac{1}{2}\sum_{r=1}^{n}|\lambda^{(r)}|}\ , q^{-\frac{1}{2}\sum_{r=1}^{n } r\kappa(\lambda^{(r)})}\ { \displaystyle \prod_{r=1}^{n-1 } q_{f_r}^{r\sum_{s = r+1}^n|\lambda^{(s)}|}}\ { \displaystyle \prod_{r=1}^{n } s_{\lambda^{(r)}}(q^{\rho } ) } \\*[1.5 mm ] \times { \displaystyle \sum_{\nu^{(1)},\cdots,\,\nu^{(n-1 ) } } \bigl ( \prod_{t=1}^{n-1}q_{f_t}^{|\nu^{(t)}| } \bigr)}\ s_{\nu^{(1 ) } } ( q^{\lambda^{(1)}+\rho } ) s_{\nu^{(n-1 ) } } ( q^{\widetilde{\lambda}^{(n)}+\rho } ) \\*[2.0 mm ] \ \ \ \ \ \ \times { \displaystyle \sum_{\chi^{(1)},\cdots,\,\chi^{(n-2 ) } } \prod_{t=1}^{n-2}}\ s_{\nu^{(t)}/\chi^{(t ) } } ( q^{\widetilde{\lambda^{(t)}}+\rho } ) s_{\nu^{(t+1)}/\chi^{(t ) } } ( q^{\lambda^{(t)}+\rho } ) \label{su(n ) string amplitude } \end{array } \right]^2,\end{aligned}\ ] ] where the parameters @xmath342 and @xmath343 are given by @xmath344 ( [ su(n ) string amplitude ] ) for the case of @xmath48 becomes as follows ; ( @xmath345 ) @xmath346 & & = \sum_{\lambda^{(1)},\lambda^{(2 ) } } \left [ \begin{array}{l } q_b^{\frac{1}{2 } \left(|\lambda^{(1)}| + |\lambda^{(1)}| \right)}\ , q^{-\frac{1}{2 } \left(\kappa(\lambda^{(1 ) } ) + 2\kappa(\lambda^{(2)})\right ) } \ q_{f}^{|\lambda^{(2)}| } \\[1.5 mm ] \ \
\times s_{\lambda^{(1)}}(q^{\rho } ) \ , s_{\lambda^{(2)}}(q^{\rho } ) \ , \ , { \displaystyle\sum_{\nu}q_{f}^{|\nu| } } \ , s_{\nu}(q^{\lambda^{(1)}+\rho } ) s_{\nu}(q^{\widetilde{\lambda^{(2)}}+\rho } ) \end{array } \right]^2 .
\label{su(2 ) string amplitude}\end{aligned}\ ] ] n. seiberg and e. witten , _
`` electric - magnetic duality , monopole condensation , and confinement in n=2 supersymmetric yang - mills theory , '' _ nucl .
phys . * b426 * ( 1994 ) 19 , ` hep - th/9407087 ` ; erratum , ibid . *
b430 * ( 1994 ) 485 ; _
`` monopoles , duality and chiral symmetry breaking in n=2 supersymmetric qcd , '' _ ibid .
* b431 * ( 1994 ) 484 , ` hep - th/9408099 ` .
a. klemm , w. lerche , p. mayr , c. vafa and n. p. warner , _
`` self - dual strings and n=2 supersymmetric field theory , '' _ nucl . phys . *
b477 * ( 1996 ) 746 , ` hep - th/9604034 ` .
+ s. katz , a. klemm and c. vafa , _ `` geometric engineering of quantum field theories , '' _ nucl .
b497 * ( 1997 ) 173 , ` hep - th/9609239 ` .
a. iqbal , _ `` all genus topological string amplitudes and 5-brane webs as feynman diagrams , ''
_ ` hep - th/0207114 ` .
+ m. aganagic , a. klemm , m. marino and c. vafa , _ `` the topological vertex , '' _ ` hep - th/0305132 ` .
a. iqbal and a. k. kashani - poor , _ `` instanton counting and chern - simons theory , '' _ adv .
* 7 * ( 2004 ) 457 , ` hep - th/0212279 ` ; _
`` su(n ) geometries and topological string amplitudes , '' _ ` hep - th/0306032 ` .
t. eguchi and h. kanno , _
`` topological strings and nekrasov s formulas , '' _ jhep * 0312 * ( 2003 ) 006 , ` hep - th/0310235 ` ; _ `` geometric transitions , chern - simons gauge theory and veneziano type amplitudes , '' _ phys . lett . * b585 * ( 2004 ) 163 , ` hep - th/0312234 ` . a. okounkov and n. reshetikhin , _ `` correlation function of schur process with application to local geometry of a random 3-dimensional young diagram , '' _ j. amer . math .
* 16 * ( 2003 ) no.3 581 , ` math.co/0107056 ` .
k. intriligator , d. morrison and n. seiberg , _
`` five - dimensional supersymmetric gauge theories and degenerations of calabi - yau spaces , '' _ nucl . phys . * b497 * ( 1997 ) 56 , ` hep - th/9702198 `
. + a. iqbal and v. kaplunovsky , _
`` quantum deconstruction of a 5d sym and its moduli space , '' _ jhep * 0405 * ( 2004 ) 013 , ` hep - th/0212098 ` .
r. gopakumar and c. vafa , _
`` on the gauge theory / geometry correspondence , '' _ adv .
phys . * 3 * ( 1999 ) 1415 , ` hep - th/9811131 ` .
+ m. aganagic , m. marino and c. vafa , _
`` all loop topological string amplitudes from chern - simons theory , '' _ commun .
* 247 * ( 2004 ) 467 , ` hep - th/0206164 ` . | five - dimensional @xmath0 supersymmetric yang - mills theories are investigated from the viewpoint of random plane partitions .
it is shown that random plane partitions are factorizable as @xmath1-deformed random partitions so that they admit the interpretations as five - dimensional yang - mills and as topological string amplitudes .
in particular , they lead to the exact partition functions of five - dimensional @xmath0 supersymmetric yang - mills with the chern - simons terms .
we further show that some specific partitions , which we call the ground partitions , describe the perturbative regime of the gauge theories .
we also argue their role in string theory .
the gauge instantons give the deformation of the ground partition . `
hep - th/0412327 ` + ` ou - het 512 ` + december 2004 * five - dimensional supersymmetric yang - mills theories * * and random plane partitions * takashi maeda@xmath2 , toshio nakatsu@xmath2 , kanehisa takasaki@xmath3 and takeshi tamakoshi@xmath2 + |
the interior structures of planets other than the earth are generally poorly constrained .
although theoretical studies abound , our understanding is hampered by the lack of direct observational constraints ( see guillot 2005 , fortney & nettelman 2009 , and guillot & gautier 2014 for reviews ) . with thousands of recently discovered exoplanets / exoplanet candidates , a basic understanding of the internal structures of giant planets is more important than ever .
seismology offers the best hope for directly inferring interior structures of planets .
indeed , our understanding of the earth s interior owes its existence primarily to seismic measurements .
we advise the interested reader to consult dahlen & tromp ( 1998 ) , hereafter dt98 , for a comprehensive description of the techniques of earth seismology .
chaplin & miglio ( 2013 ) presents a review of recent developments in asteroseismology , while lognonne & mosser ( 1993 ) and stein & wysession ( 2003 ) discuss results in terrestrial seismology .
unfortunately , seismic measurements of other planets are much more difficult , and no unambiguous detections of oscillations in the outer solar system planets exist ( although there are tentative detections of pressure modes in jupiter via radial velocity techniques , see gaulme et al . 2011 and discussion in section [ modeamp ] ) .
saturn provides an amazing opportunity to indirectly detect global oscillation modes through their interaction with saturn s rings .
marley ( 1991 ) ( m91 ) and marley & porco 1993 ( mp93 ) predicted that some of saturn s oscillation modes ( in particular the prograde f - modes ) could be detected through waves in the rings launched at lindblad resonances with the gravitational forcing created by the modes ( see also pena 2010 ) .
this prediction was confirmed by hedman & nicholson ( 2013 ) ( hn13 ) , who used _
data to measure the azimuthal pattern numbers @xmath2 and pattern frequencies @xmath3 of several unidentified waves within saturn s c ring .
they found that the frequencies and pattern numbers matched marley & porco s predictions to within a few percent , and that the waves could not be explained by any other known process .
the puzzling aspect of hn13 s findings is that there are _ multiple _ waves of the same @xmath2 near the locations predicted by m91 and mp93 , whereas only one wave was expected .
the multiple waves , split by less than @xmath4 in frequency , appear to be generated by distinct oscillation modes within saturn whose frequencies are split by the same fractions . the observed splitting is not simple rotational splitting ( as this occurs between oscillation modes of different @xmath2 ) and suggests more complex physics is at play .
fuller et al . ( 2014 ) ( f14 ) investigated the effect of a solid core on the oscillation mode spectrum of saturn .
they found that if saturn has a large solid core that is relatively unrigid ( has a small shear modulus @xmath5 ) , the shear oscillation modes of the core can exist near the same frequencies as the f - modes that generate some of the observed waves in the rings .
modes very close in frequency to the f - modes can degenerately mix with them ( a process also known as avoided crossing ) , attaining large enough gravitational potential perturbations to generate waves in the rings .
however , f14 found that degenerate mixing was rare , and that only finely tuned models could qualitatively reproduce the observed mode spectrum .
the oscillations of rotating giant planets have also been examined in several other works ( vorontsov & zharkov 1981 , vorontsov 1981 , vorontsov 1984 , m91 , wu 2005 , pena 2010 , le bihan & burrows 2012 , jackiewicz et al .
2012 , braviner & ogilvie 2014 ) .
none of these works have extensively examined the effect of stable stratification and the resulting planetary mode spectrum ( although m91 does briefly consider the effect of stable stratification on the f - mode frequencies ) . in this paper , we examine saturn s oscillation mode spectrum in the presence of a stably stratified region deep within the planet . regions of stable stratification have been speculated to exist within giant planets due to the stabilizing effect of composition gradients ( leconte & chabrier 2012 , 2013 ) .
the composition gradients could be produced by dissolution of heavy core elements in the helium / hydrogren envelope ( wilson & militzer 2012a , 2012b ) or by gravitational settling of metals ( stevenson 1985 ) or helium ( salpeter 1973 , stevenson & salpeter 1977 ) .
recent simulations have sought to determine the large - scale time evolution of doubly diffusive convection produced by competing thermal / composition gradients ( rosenblum et al .
2011 , mirouh et al .
2012 , wood et al .
2013 ) , but the resulting global structure of giant planets is unclear .
figure [ satdiagram ] shows a simple schematic of the type of saturn models we consider .
it should not be interpreted too strictly , it is intended only to provide the reader with a general picture of our hypothesis for saturn s interior structure . if stably stratified regions exist , they allow for the existence of gravity modes ( g - modes ) in the oscillation mode spectrum .
for stable stratification deep within the planet , the g - modes can exist in the same frequency regime as the f - modes and can strongly mix with them .
this process is analogous to the mixed g - modes / p - modes observed in red giant stars , although somewhat complicated by saturn s rapid rotation .
our calculations reveal that g - mode mixing can naturally explain the observed splitting between the @xmath6 waves , but can not robustly reproduce the fine splitting between the @xmath1 waves .
we claim this is strong evidence for the existence of stable stratification within saturn , although some important physical ingredient ( e.g. , differential rotation ) may be required for a complete understanding .
_ cassini _ image of saturn and its rings , overlaid with a schematic cartoon of our hypothesis for saturn s interior structure .
the structure shown here is not quantitatively accurate .
it is meant only to illustrate the general features of saturn s interior structure that we advocate : a thick convective envelope ( which harbors f - modes , p - modes , and i - modes ) overlying a region of stable stratification near the core of the planet ( which harbors g - modes and r - modes ) .
we have also pointed out the c - ring , where all of the mode - driven waves of been observed . ]
our paper is organized as follows .
section [ models ] describes the toy saturn models we use in our calculations .
section [ modes ] summarizes our method of solving for oscillation modes in the presence of rapid rotation , and reviews the types of modes that exist in rotating planets . in section [ mix ] we examine the process of mode mixing induced by rotation , centrifugal , and ellipticity effects , and describe how this affects mode frequencies and eigenfunctions .
section [ results ] compares our results to observations , and we conclude with a discussion of these results in section [ discussion ] .
this section also addresses the issues of mode amplitudes , mode driving , and the prospects for observing saturnian and jovian p - modes via radial velocity techniques .
the interior structure of giant planets is not particularly well - constrained .
other than its mass @xmath7 and radius @xmath8 , the strongest observational constraint on saturn s interior structure is the measured value of the gravitational moment @xmath9 , which indicates that saturn must have a dense core of @xmath10 ( guillot 2005 ) .
we therefore create toy models which roughly match the measured values of @xmath7 , equatorial radius @xmath11 , polar radius @xmath12 , and @xmath9 .
we do not attempt to rigorously compare these models with any theoretical equations of state or microphysical models , although in section [ discussion ] we discuss how our models relate to recent theoretical developments in planetary interiors .
for the purposes of our adiabatic mode calculations , the only physical quantities of importance in saturn s interior are the density @xmath13 , brunt - vaisala frequency @xmath14 , sound speed @xmath15 , gravitational acceleration @xmath16 , and spin frequency @xmath17 . to create a toy model , we proceed as follows .
we first create a spherical model with a polytropic density profile of index @xmath18 , with a density profile @xmath19 .
we choose a sound speed @xmath20 such that the brunt - vaisala frequency [ brunt ] n_1 ^ 2 = -g_1 - is equal to zero everywhere .
we then choose inner and outer core radii @xmath21 and @xmath22 , and a core density enhancement @xmath23 .
we multiply the density of the inner core by @xmath23 , such that @xmath24 for @xmath25 .
the density of the outer core is calculated via ( r ) = _ 1(r ) r_in < r < r_out this form is somewhat arbitrary ; we use it to obtain a smooth density increase between the envelope and inner core . in the outer core , we readjust the soundspeed such that [ cs ] c_s^2(r ) = c_s1 ^ 2(r_out ) + r_in < r < r_out . once again , this sound speed profile is somewhat arbitrary .
this form ensures a positive value of @xmath26 in the outer core . because we focus only on f - modes ( for which @xmath13 is the defining quantity ) and g - modes ( for which @xmath26 is the defining quantity ) the precise value of @xmath27 has little effect on our results , except in so far as it affects the value of @xmath26 through equation [ brunt ] . after performing the above procedure
, we scale the mass and radius of our model to match that of saturn . from the profiles of @xmath13 and @xmath15 we calculate new values of @xmath16 and @xmath14 .
the resulting @xmath13 , @xmath27 , @xmath26 , and @xmath16 profiles of our reference model are shown in figure [ satmodel ] .
note the value of @xmath26 is positive only in the outer core ( @xmath28 ) .
consequently , g - modes are confined to the outer core in our models .
this type of model is physically motivated by a scenario in which stable stratification is generated within saturn by a composition gradient between the rocky / icy inner core and the hydrogen / helium envelope .
the maximum value of @xmath14 shown in figure [ satmodel ] is very typical for models containing a smooth density profile between the core and the envelope . in the stably stratified region ,
the brunt - vaisala frequency has a value of @xmath29 .
the dense core implied by saturn s @xmath9 entails typical values of @xmath30 and @xmath31 , while our models typically have @xmath32 .
typical peak values of @xmath14 in the stably stratified region thus naturally have order of magnitude @xmath33 .
radial profiles of density @xmath13 , sound speed squared @xmath27 , brunt - vaisala frequency @xmath14 , gravitational acceleration @xmath16 , and ellipticity @xmath34 in one of our saturn models .
all quantities are in units with @xmath35 .
this model has @xmath36 cm , @xmath37 cm , and @xmath38 , similar to the measured values of saturn . for reference
, this model has a central density of @xmath39 , and a core " mass ( defined as the mass within @xmath40/2=0.26 $ ] ) of @xmath41 . ] for each model we create , we perturbatively calculate the centrifugal flattening of the planet using clairaut s equation ( see eggleton 2006 ) . in this approximation ,
the radius of the spherically symmetric model is perturbed such that [ rac ] r_ac = r , where @xmath42 is the radius in the spherically symmetric model and @xmath43 is the perturbed radius .
the ellipticity of the model , @xmath44 , is proportional to the small parameter @xmath45 , where @xmath46 is the dynamical frequency of saturn .
the details of the calculation are described in [ ellipticity ] .
equation [ rac ] implies @xmath47 $ ] and @xmath48 $ ] , which in turn requires [ rav ] r = r_po/3 + 2 r_eq/3 .
the measured polar and equatorial radii of saturn are @xmath49 cm and @xmath50 cm .
equation [ rav ] then results in @xmath51 cm .
this is the radius of our spherical models .
we adopt a spin period of 10 hours , 34 minutes , which entails @xmath52.% , which is accurate enough for our purposes since we have excluded higher order rotational effects .
see guillot & gautier 2014 for discussion . ]
higher order corrections are of order @xmath53 , thus , we should expect quantities to deviate by a couple percent in a more accurate model .
it is important to note that the value of @xmath54 is somewhat insensitive to the density profile of a reasonable saturn model .
the value of @xmath9 , however , can vary considerably .
indeed , the measured value of @xmath9 has led to the conclusion that saturn must have a core which is much denser than saturn s average density , and therefore must be enriched in ices , rocks , or metals ( guillot 2005 ) .
we construct our models to have @xmath55 in accordance with the measured value of @xmath9 for saturn . will change with the addition of higher order centrifugal corrections , so we do not require our models to have a value of @xmath9 exactly equal to the measured value . ] in practice , this means choosing appropriate values of @xmath21 , @xmath22 , and @xmath23 to fulfill this constraint .
saturn rotates rapidly , with @xmath56 .
the rapid rotation distorts its shape away from sphericity , such that @xmath57 .
these two factors greatly complicate calculations of the oscillation modes of a realistic saturn model . in this section ,
we summarize our technique for calculating the mode frequencies and eigenfunctions in the presence of rapid rotation and non - sphericity .
the technique utilizes separation of variables in which the angular part of the eigenfunctions is projected onto spherical harmonics , and the radial part is projected onto pseudo - modes " which serve as basis functions for computing the normal modes . in the absence of rotation ,
the pseudo - modes are identical to the orthogonal oscillation modes of a planetary model . adding rotation modifies the characteristics of the pseudo - modes and introduces new classes of pseudo - modes .
rotation also causes the pseudo - modes to mix with one another via the coriolis / centrifugal forces .
the non - sphericity of saturn is treated perturbatively , causing additional mixing between the pseudo - modes . at all times
, we consider only linear , adiabatic , fluid oscillations in a uniformly rotating model .
the linear and adiabatic approximations are probably very good for the small - amplitude , low wave - number modes we consider .
the effect of a solid core was examined by fuller et al .
some degree of differential rotation and meridional circulation likely does exist deep within saturn , however , the amplitude and structure of such flows is not well constrained , and we do not investigate this complication in detail ( but see discussion in section [ discussion ] ) .
the derivation of the pseudo - mode oscillation equations , and our method for solving them , are described in [ pseudo ] . including rigid rotation effects to order @xmath45 allows pseudo - modes of angular degree @xmath58 to mix with pseudo - modes of angular degree @xmath59 .
we account for this mixing via the methods described in [ matrix ] . in our formalism ,
mode eigenfunctions have time and azimuthal dependence e^i ( t + m ) such that prograde modes have negative values of @xmath2 for positive mode frequencies @xmath60 .
mode eigenfunctions for an @xmath61 , @xmath62 pseudo f - mode ( black solid line ) , @xmath61 , @xmath63 pseudo g - mode ( blue dashed line ) , @xmath64 , @xmath65 pseudo g - mode ( green dashed line ) , and @xmath61 , @xmath66 pseudo i - mode ( red dot - dashed line ) .
all are @xmath1 ( prograde ) modes , and all quantities are plotted in units of @xmath35 , with pseudo - modes normalized via equation [ normpseudo ] . from top to bottom , we plot the radial displacement @xmath67 , horizontal poloidal displacement @xmath68 , horizontal toroidal displacement @xmath69 , and gravitational potential perturbation @xmath70 .
the modes are calculated for the planetary model shown in figure [ satmodel ] , and correspond to the boxed modes in figure [ modes ] . ] in the absence of rotation , pseudo - modes are identical to the usual stellar oscillation modes discussed in the literature
. there are pseudo p - modes , pseudo f - modes , and in the presence of stratification , pseudo g - modes .
rotation introduces two new classes of pseudo modes : pseudo rossby modes and pseudo inertial modes .
we do not discuss p - modes , as their frequencies are too large to have lindblad resonances within saturn s rings .
the f - modes , g - modes , and possibly the rossby modes and inertial modes are important for the problem at hand , and we discuss each of them below . in the discussion that follows , it is helpful to refer to a dispersion relation . to do this
, we use the wkb approximation on equations [ eqn1]-[eqn6 ] , ignoring the gravitational perturbations , and restricting ourselves to the low frequency limit ( @xmath71 , with @xmath72 the lamb frequency ) .
the result is [ dispersion ] k_r^2 = ^-1 . here
, @xmath73 and @xmath74 is a function of @xmath75 and @xmath2 and is of order unity ( see [ pseudo ] ) .
although this dispersion relation is not very transparent , it is very useful when evaluated in the appropriate limits . in particular , in the non - rotating @xmath76 limit
, we obtain the usual g - mode dispersion relation [ gdispersion ] k_r^2 = .
some other useful limits are discussed below .
spectrum of prograde @xmath1 pseudo - modes for the planetary model shown in figure [ satmodel ] , with surface gravitational potential perturbations @xmath77 plotted vs. mode frequency @xmath78 .
quantities are dimensionless and modes are normalized as described in figure [ modefunctions ] .
we have plotted pseudo - modes up to @xmath79 , and have only included pseudo - modes with frequencies in the vicinity of the @xmath61 pseudo f - mode .
the eigenfunctions of the boxed pseudo - modes are plotted in figure [ modefunctions ] ; the @xmath61 pseudo f - mode is the boxed mode with the largest value of @xmath80 .
pseudo - modes with the same number of nodes in the g - mode cavity are connected by lines .
mode mixing occurs most strongly between modes of like frequency ( i.e. , along vertical columns ) and between modes of the same radial order @xmath81 ( i.e. , along branches connected by lines ) . ] in the saturn models we examine , each angular degree @xmath58 family of pseudo - modes typically contains a single f - mode ( which are essentially the surface gravity modes discussed in braviner & ogilvie 2014 ) .
the f - modes have no nodes in their radial eigenfunctions ( in the convective envelope ) , and have frequencies of approximately [ fmodes ] _ f _ dyn .
the gravitational potential perturbations of the f - modes are large , and typically @xmath82 for all modes @xmath83 other than the f - mode ( under normalization convention of [ normpseudo ] ) . here
, the prescript notation indicates a pseudo - mode of angular dependence @xmath58 .
because the f - modes have the largest potential perturbations and also have lindblad resonances within the rings , we expect them to produce the highest amplitude waves in the rings .
figure [ modefunctions ] displays a plot of the radial displacement @xmath67 , horizontal poloidal component @xmath68 , horizontal toroidal component @xmath69 , and potential perturbation @xmath70 for the @xmath61 and @xmath1 pseudo f - mode of the model shown in figure [ satmodel ] .
the pseudo f - modes are not pure " f - modes because they have g - mode character in the g - mode cavity ( analogous to the mixed character of oscillation modes in red giant stars ) , but most of their inertia typically resides in the convective envelope .
since our models contain a stably stratified region , they support g - modes .
the pseudo g - modes are mostly localized to the stably stratified layer , and typically have small gravitational potential perturbations .
figure [ modefunctions ] shows the eigenfunctions of two representative pseudo g - modes . in the limit @xmath84 ,
pseudo g - modes obey the dispersion relation : [ gdisp ] k_r^2 = ^-1 .
note that g - modes ( of the same radial order ) have higher frequencies for larger values of @xmath14 , which can be produced by larger density gradients in a given model .
g - modes also have higher frequencies for g - mode cavities at small radii @xmath42 , and for large values @xmath75 .
the terms containing @xmath85 in equation [ gdisp ] modify the pseudo g - modes from the non - rotating g - mode dispersion relation of equation [ gdispersion ] . for @xmath86
they typically increase the radial wavenumber of pseudo - gmodes . when the second term in brackets is equal to zero , the wavenumber diverges .
this divergence occurs occurs at @xmath87 for @xmath88 , and at @xmath89 for @xmath90 . for even lower frequencies ,
the second term in brackets is negative , such that there are no very low frequency pseudo g - modes .
however , this does not imply that low frequency g - modes do not exist in rapidly rotating stars / planets , as the pseudo - mode mixing process ( see section [ mix ] ) allows the normal modes to have frequencies different from the pseudo - modes .
we therefore caution that our method for calculating ( and understanding ) very low frequency modes may not be optimal . in the super - inertial regime of interest ( @xmath91 ) ,
equation [ gdispersion ] translates to the approximate g - mode spectrum : [ gdispersion2 ] _
n ~ where @xmath81 is the number of nodes in the radial eigenfunction .
we find that in our saturn models , in which the value of @xmath14 is peaked in the outer core at at values @xmath92 , the spectrum of pseudo g - modes typically extends to _ higher _ frequencies than the f - modes .
low @xmath75 pseudo g - modes near the low @xmath75 pseudo f - modes typically have @xmath93 .
higher @xmath75 pseudo g - modes have even larger frequencies , meaning that their spectrum is dense in the vicinity of the low @xmath75 pseudo f - modes .
figure [ modes ] shows the pseudo mode spectrum up to @xmath79 .
we can see that our models contain many pseudo g - modes in the same frequency regime as low @xmath75 pseudo f - modes .
equation [ gdispersion2 ] can also be used to find g - mode frequency spacings : [ omspacen ] _
n ~ _ n and [ omspacel ] _
low @xmath75 pseudo g - modes of consecutive radial order have frequency spacings of @xmath94 at frequencies near the low @xmath75 pseudo f - modes .
pseudo g - modes of the same radial order @xmath81 but with angular number varying by @xmath95 have @xmath96 .
these frequency spacings are important for calculating the probability of finding g - modes with frequencies near the f - modes ( see section [ results ] ) and for understanding how modes can mix up to high values of @xmath75 ( see section [ mix ] ) .
in addition to f - modes and g - modes , our rotating models support inertial modes .
whereas f - modes and g - modes are restored by self - gravity and buoyancy , respectively , inertial modes are restored by the coriolis force .
the pseudo inertial modes are restricted to the sub - inertial frequency regime @xmath97 . like g - modes ,
they exhibit small pressure and gravitational perturbations .
figure [ modefunctions ] shows the eigenfunction of an inertial pseudo - mode . like the f - modes ,
the inertial modes are mostly restricted to the convective envelope , although they may have g - mode character within the g - mode cavity .
the inertial pseudo mode dispersion relation can be found by taking the @xmath98 limit of equation [ dispersion ] : [ idispersion ] k_r^2 = .
this dispersion relation is notable because it depends only on the frequency ratio @xmath85 , and the values of @xmath75 and @xmath2 , without any dependence on the material properties of the planet .
( see , e.g. , greenspan 1968 ) . like equation [ idispersion ]
, it shows that inertial modes are sensitive only to the spin frequency and their angle of propagation . ] note that the radial wavenumber diverges when the denominator is equal to zero , as discussed above .
it is possible that the divergence of the wave vector of the pseudo modes is related to the wave number divergence of inertial modes at critical latitudes ( discussed in ogilvie & lin 2004 and wu 2005 ) although we do not explicitly examine this phenomenon here .
because the pseudo inertial modes are restricted to the sub inertial frequency range , they all have lower frequency than saturn s f - modes , and can not undergo degenerate mixing ( avoided crossings ) with them . for this reason , we do not perform a detailed investigation of pseudo inertial modes .
we note , however , that the fundamental " ( only one node in its radial eigenfunction ) pseudo inertial mode shown in figure [ modefunctions ] mixes relatively strongly with the f - modes because they are both localized primarily in the convective envelope .
finally , we comment that our technique for solving for normal modes from a pseudo mode basis ( see [ matrix ] ) may not be tractable for solving for inertial modes , as an accurate calculation would likely require the inclusion of pseudo inertial modes of very high radial wave number @xmath81 and angular degree @xmath75 .
it may be simpler to use the techniques discussed in ogilvie & lin 2004 and wu 2005 .
rotation also introduces a class of modes not often explored in the literature , which we refer to as pseudo rossby modes ( these modes have been explored using the traditional approximation in lee & saio 1997 , fuller & lai 2014 ) .
these modes require stable stratification to exist , and are thus similar to g - modes . however , the rossby modes exist at lower frequency , and are purely _ retrograde _ modes .
the rossby pseudo - modes exist in the @xmath99 , @xmath100 limit of equation [ dispersion ] .
the result is [ rdispersion ] k_r^2 = .
the denominator of the fraction on the right is always positive , thus , for positive frequencies , we require @xmath101 to have @xmath102 ( i.e. , only retrograde modes can exist ) .
the pseudo rossby modes are retrograde modes , and thus can not undergo avoided crossings with the prograde modes we are interested in . however , because they have similar characteristics to g - modes , they mix somewhat strongly with them . moreover , because their frequencies are small ( @xmath103 ) , the rotational mixing between pseudo rossby modes of different @xmath75 is quite strong . like the inertial modes
, we do not include a more detailed analysis of the rossby modes since they do not exist in the super inertial regime . in reality
, the pseudo - modes are not strictly f - modes or g - modes or inertial modes , but share qualities of each . in the saturn model shown in figure [ satmodel ] ,
the @xmath61 pseudo g - mode from figure [ modefunctions ] exhibits some f - mode character due to its proximity in frequency to the @xmath61 pseudo f - mode , giving it an enhanced surface potential perturbation @xmath104 . in turn
, the @xmath61 pseudo f - mode has some g - mode character , causing its eigenfunction to have appear wave - like in the g - mode cavity .
additionally , the inertial mode from figure [ modefunctions ] has some g - mode character .
the latter mixing occurs because prograde pseudo modes in the sub - inertial frequency range behave like g - modes in stably stratified regions , and inertial modes in convective regions .
this feature is responsible for the spikes in @xmath104 in figure [ modes ] at @xmath105 , which are created by gravito - inertial mixed modes near the fundamental " pseudo inertial mode ( the inertial mode with only one node in its eigenfunction in the convective envelope ) .
as discussed above , the pseudo - modes are not normal oscillation modes because they mix with one another due to the coriolis and centrifugal forces , as well as the elliptical shape of the planet .
a useful analogy to understand planetary rotational mode mixing is the mode mixing between atomic electron energy levels induced by an electric field ( the stark effect ) or a magnetic field ( the zeeman effect ) . in the limit of a weak magnetic field ,
the zeeman effect splits degenerate energy levels into a multiplet .
the observed frequency splitting can then be used to infer the strength of the magnetic field .
similarly , in slowly rotating stars or planets , the coriolis force induces a small frequency splitting which can be used to measure the rotation rate .
however , for a strong magnetic field , atomic energy levels are split by so much that they overlap with adjacent energy levels .
the energy levels of different electron wave functions can not cross ( they can not be exactly degenerate ) ; instead they undergo an avoided crossing in which the modes mix with one another , exchanging character . near the avoided crossing the resulting electron wave functions are mixtures ( superpositions ) of the non - perturbed wave functions , and the energy levels ( eigenfrequencies ) of the states are similar but not exactly the same .
a very similar process occurs in rapidly rotating stars and planets . in our case ,
the rapid rotation of saturn perturbs f - mode eigenfrequencies enough that they overlap with neighboring g - mode eigenfrequencies . the f - modes and g - modes mix with one another , forming mixed modes that are superpositions of the f - mode and g - modes , and whose frequencies are similar but not equal .
the dynamics of rotational mode mixing is discussed extensively in fuller et al .
2014 , here we review the basic physics .
the mode mixing angle between two pseudo - modes @xmath83 and @xmath106 is ( 2 _ ) = . here
@xmath107 is the mode mixing coefficient , while @xmath108 and @xmath109 are the perturbed mode frequencies @xmath110 .
the effective value of @xmath107 is a combination of the mixing coefficients @xmath111 ( due to the coriolis force ) , @xmath112 ( due to centrifugal / ellipticity perturbations to the potential energy ) and @xmath113 ( due to ellipticity perturbations to the kinetic energy ) .
these mixing coefficients are discussed in more detail in [ matrix ] , and expressed explicitly in dt98 .
strong mode mixing occurs when @xmath114 .
the simplest way for this to occur is for two modes to be nearly degenerate , i.e. , for @xmath115 .
we refer to this case as degenerate mode mixing , which is equivalent to saying two modes are undergoing an avoided crossing .
however , strong mode mixing can occur at larger frequency separations if the value of @xmath116 is large .
this typically only occurs for modes of the same type ( g - mode , inertial mode , etc . ) and with the same number of nodes in their radial eigenfunctions .
mixed modes have eigenfunctions that are superpositions of the original pseudo - modes .
a pseudo g - mode mixed with a pseudo f - mode obtains a contribution of @xmath117 to its eigenfunction ( where @xmath118 is the pseudo f - mode eigenfunction ) .
since g - modes typically have very small intrinsic potential perturbations , their potential pertrubations are often dominated by the pseudo f - mode contribution such that they have @xmath119 .
our techniques only allow for mixing between pseudo - modes of the same @xmath58 , and pseudo - modes of @xmath59 . however , pseudo - modes with larger differences in @xmath75 may still mix indirectly through intermediary modes , as was discussed in fuller et al .
2014 . as the number of intermediary pseudo - modes increases
, we may expect that strong mixing becomes more difficult , although it remains possible .
so , while pseudo g - modes of large @xmath75 can mix with the low @xmath75 pseudo f - modes , they need to be very close in frequency for strong mixing to occur .
this idea motivates our proposal that the small frequency differences between the observed waves in saturn s rings are due to mode mixing between the low @xmath75 pseudo f - modes and larger @xmath75 pseudo g - modes . we do not attempt to provide a detailed investigation of all the different types of mixing ( coriolis , potential , kinetic ) between all the different types of pseudo - modes ( f - modes , g - modes , inertial modes , rossby modes )
here we only comment on the order of magnitude of the mixing .
the maximum possible values of the mixing coefficients are of order @xmath120 , @xmath121 , and @xmath122 .
typical values are even smaller , especially for modes whose radial orders differ , or which propagate in different cavities .
low @xmath75 f - modes have largest mixing coefficients with neighboring g - modes of the same @xmath75 , and with the fundamental inertial mode .
the g - modes usually mix most strongly with g - modes of @xmath123 with the same radial order @xmath81 .
we can now understand how mixing between the pseudo - modes shown in figure [ modes ] will proceed .
the pseudo g - modes will tend to mix with g - modes of the same radial order , i.e. , along the g - mode branches connected by lines in figure [ modes ] .
the pseudo modes will also mix with modes of nearly the same frequency , i.e. , along vertical columns in figure [ modes ] .
we also note that pseudo g - modes can mix with their negative frequency ( retrograde ) and rossby mode counterparts .
however , since these modes are further away in frequency , the degree of mixing should be smaller .
although the pseudo - modes do not experience any self - mixing due to the coriolis force of modes calculated for a non - rotating model is identical to the frequently discussed rotational frequency perturbation @xmath124 that is valid in the slow rotation @xmath125 limit . as a check on our numerical code
, we confirm that the frequencies of the pseudo - modes are shifted by @xmath124 compared to modes calculated for a non - rotating model , in the limit of slow rotation .
] , they do experience self - mixing due to the potential @xmath126 and kinetic @xmath127 mixing coefficients , causing their frequencies to shift .
the largest frequency shifts occur for the pseudo f - modes , and are due to the centrifugal component of @xmath126 .
this causes the computed normal f - modes to have frequencies roughly @xmath128 lower than their pseudo - mode counterparts .
the centrifugal correction is thus crucial for obtaining f - mode frequencies within @xmath0 of the observed values .
our ultimate goal is to compare our computed oscillation mode spectrum to the observed mode spectrum .
unfortunately , the observed spectrum is incomplete , as only modes with lindblad resonances that lie within the rings can excite observable waves .
moreover , only modes with large enough gravitational potential perturbations to subtantially affect the rings can be observed .
this likely means that only f - modes and g - modes undergoing avoided crossings with the f - modes can be observed .
we compare our results with the six waves observed by hn13 , keeping in mind that the observed mode spectrum is incomplete .
to estimate mode amplitudes , we use the same technique adopted in fuller et al .
2014 ( see section 5 and appendix d of that work ) .
the important idea is that the optical depth variation @xmath129 of the ring waves is proportional to the amplitude @xmath130 and gravitational potential perturbation @xmath77 of mode driving them .
if one can estimate the value of @xmath77 of the mode , one can use the measured optical depth variation to calculate the mode amplitude . to do this
, we assume the most prominent @xmath1 wave is generated by a mode with the same value of @xmath77 as the @xmath61 , @xmath1 pseudo f - mode of our model .
we then calculate the mode frequency@xmath131 and amplitude @xmath132 required to generate this wave , the latter of which is calculated from equation d.7 of fuller et al .
each mode amplitude of our model is then calculated assuming energy equipartition , i.e. , @xmath133 .
we re - examine the validity of energy equipartion in section [ discussion ] .
finally , we caution that the amplitude @xmath132 inferred from the ring observations is only approximate , as the precise amplitude of the waves in the rings is somewhat difficult to measure because of non - linear and wave damping effects . in what follows , it is helpful to recall the relation between rotating frame mode frequency @xmath78 , inertial frame mode frequency @xmath134 , and the wave pattern frequency @xmath135 : _ - m _ s = _ = -m _ p , .
wave amplitudes in saturn s rings ( characterized by their maximum optical depth variation @xmath136 ) as a function of wave pattern speed @xmath137 , for @xmath6 waves .
the blue diamonds are the waves measured by hn13 , while the red crosses are waves generated by the normal oscillation modes we have computed .
the modes are calculated for the saturn model shown in figure [ satmodel ] .
modes in the pink shaded region are not observed because their gravitational perturbations are too weak to generate detectable waves in the rings .
modes in the purple shaded region have lindblad resonances inside the inner edge of the c - ring where it is difficult to detect waves .
the vertical blue dashed line denotes the location of the maxwell gap , which may be opened by the @xmath138 , @xmath6 f - mode ( see section [ discussion ] ) . ]
same as figure [ m2 ] , but for @xmath1 modes . ]
same as figure [ m2 ] , but for @xmath139 modes . ]
same as figure [ m2 ] , but for @xmath140 modes .
no @xmath140 waves have been detected , likely because the mode amplitudes are too small ( see section [ discussion ] ) . ]
we begin by examining figure [ m2 ] and table [ table ] .
the two observed waves ( waves a and b in table [ table ] ) have pattern frequencies near the expected location for @xmath138 , @xmath6 f - modes ( m91 , mp93 ) , and are separated by @xmath141 in frequency .
our computed normal mode spectrum also contains two modes within @xmath142 of these observed pattern speeds . for comparison
, the frequencies @xmath134 of these two modes are listed in table [ table ] in rows a and b. the wave amplitude ratio of our two modes also approximately matches the observed amplitude ratio . in our analysis , these two modes are the @xmath138 , @xmath6 f - mode , and a neighboring @xmath138 , @xmath6 , @xmath143 g - mode ( i.e. , the @xmath144 mode ) .
the g - mode obtains a large potential perturbation due to mixing with the f - mode , allowing its effect on the rings to be observed .
both modes are mixed with other values of @xmath75 but are dominated by their @xmath138 components .
same as figure [ m3 ] . in this version
, we have included the locations of the pseudo - modes ( black asterisks ) .
where appropriate , we have drawn arrows from the location of the pseudo - modes to the normal modes , indicating which pseudo - modes are the dominant component of each normal mode .
the arrow extending from beneath the plot originates from an @xmath145 pseudo g - mode just below the bottom of the plot . ]
our saturn model was slightly fine - tuned to generate a frequency splitting of @xmath141 between the largest amplitude @xmath6 modes . however , spitting of this magnitude is very common .
figures [ m3 ] and [ m4 ] show very similar splitting , in which there is a @xmath146 g - mode with frequency split by a few percent from the f - mode frequency .
the models need not be fine - tuned to qualitatively reproduce this feature .
the reason for such splitting is as follows . [
cols="^,^,^,^,^,^,^ " , ] equation [ omspacen ] demonstrates that the frequency spacing between @xmath147 g - modes in the vicinity of the low @xmath147 f - modes is of order @xmath148 .
the average spacing between an f - mode and the nearest g - mode is one quarter of the g - mode frequency spacing , i.e. , @xmath149 .
the fractional frequency spacing of @xmath150 is smaller because the prograde modes are all shifted to higher frequencies in the rotating frame , such that typical frequency spacings are of order @xmath151 .
it is thus quite common to have a g - mode whose frequency is within several percent of the f - mode . in our models ,
the g - modes are more strongly mixed with the f - modes for smaller values of @xmath75 .
the reason is that higher @xmath75 f - modes are confined closer to the surface of the planet , away from the g - mode cavity deep in the interior .
therefore , low @xmath75 f - modes mix more strongly with the g - modes , which may explain why a frequency splitting of @xmath141 is only observed near the @xmath138 , @xmath6 f - mode .
such splitting also occurs in figures [ m3 ] and [ m4 ] , but only because this particular saturn model happens to have @xmath61 and @xmath152 g - modes whose frequencies are closer than average to the f - mode frequencies .
nonetheless , it is very possible that there exist additional waves near the observed @xmath1 and @xmath139 waves , which have smaller amplitudes , and which are split in frequency by a few percent .
let us now turn our attention to figure [ m3fancy ] .
the three observed @xmath1 waves have frequencies that are split an order of magnitude less than the @xmath6 waves , i.e. , their frequencies are split by less than a percent .
such splitting can not be explained by mode mixing between f - modes and neighboring g - modes of the same @xmath75 .
instead , the splitting may be due to avoided crossings between the f - mode and g - modes of much higher values of @xmath75 .
these avoided crossings could be very common , because the g - mode spectrum near the f - mode becomes very dense as we proceed to large values of @xmath75 ( see figure [ modes ] ) .
figure [ m3 ] shows that there is a mixed mode very near the f - mode , split by @xmath153 in frequency .
the frequencies @xmath134 of these two modes are listed in table [ table ] in rows c and d ( the mode in row e is the @xmath61 g - mode discussed above ) . in this case , the fine - splitting arises due to an avoided crossing between the @xmath61 f - mode and an @xmath145 g - mode . however , we find that strong mixing between the @xmath147 f - modes and high @xmath75 g - modes is uncommon .
it appears that the effective mixing coefficients are small , requiring very small frequency separations @xmath154 for strong mixing to occur .
the strong mixing shown in figures [ m3fancy ] occurs because the model has been fine - tuned to produce a resonance between the @xmath61 f - mode and an @xmath145 g - mode .
we find such strong mixing only occurs in about one tenth of our planet models .
simultaneous strong mixing between three modes , as appears to be observed , is even less common .
it is thus difficult to reconcile the observations with the notion of strong mixing between low @xmath75 f - modes and high @xmath75 g - modes , indicating that some un - included physical effects may be important ( see section [ discussion ] ) .
finally , let us examine figures [ m4 ] and [ m5 ] .
the @xmath152 and @xmath155 f - modes have higher frequencies , where the g - mode spectrum is less dense and where mode mixing coefficients are smaller .
it is therefore not surprising that no fine splitting is observed for the @xmath139 mode .
the @xmath155 , @xmath140 f - mode is not observed at all , even though our calculations suggest it should be observable .
this indicates that modes of higher @xmath75 and/or higher frequency are excited to lower amplitudes .
indeed , the observed @xmath139 wave is a factor of @xmath156 smaller in amplitude than our energy equipartition calculation predicts .
if the amplitude of the @xmath140 f - mode is smaller by a similar factor , it would not make an observable impact on the rings , in accordance with its non - observation .
we have examined the oscillation mode spectrum of saturn models with stable stratification deep in the interior .
the stably stratified region supports g - modes , and we find that the g - mode spectrum extends to higher frequencies than the f - modes that generate observed waves in saturn s rings .
g - modes close in frequency to the f - modes mix with them , allowing them to obtain large enough gravitational perturbations to generate waves in the rings .
we have proposed that this mechanism is responsible for the observed small frequency splittings between modes of the same azimuthal number @xmath2 inferred from ring observations .
if true , this requires the existence of a thick region of stable stratification deep within the interior of saturn , in contrast to the conventional notion that giant planets envelopes are convective throughout .
our models very naturally reproduce the observed frequency splitting of @xmath141 between the two @xmath6 waves observed by hn13 .
in fact , for g - mode cavities located deep within the planet near a `` core - envelope '' interface at @xmath157 , frequency splittings of this magnitude are difficult to avoid .
neither fully convective models nor models containing shear modes in a solid core can readily produce frequency splittings of this magnitude ( see f14 ) .
our results place some basic constraints on the interior structure of saturn . in order to reproduce the measured frequencies of the f - modes ,
the core of saturn @xmath158 must be substantially more dense than saturn s envelope .
a dense core is also required by the measured vale of @xmath9 , and it is reassuring that our seismic analysis is consistent with this constraint .
more quantitative constraints are difficult because f - mode frequencies are only slightly affected by the properties of the core ( f14 ) .
one must include rotational / ellipticity terms up to at least @xmath159 to calculate mode frequencies accurate enough to tightly constrain core properties .
we can also place some loose constraints on the characteristics of the stably stratified region within saturn . in order for
the g - modes to extend to high enough frequencies to mix with f - modes , the stable stratification must exist deep within the interior of saturn .
this is most naturally achieved for stable stratification at the large density gradient between the core and envelope at @xmath160 .
we also find the stable stratification must be confined to relatively small radii @xmath161 .
if the stable stratification extends to larger radii , mixing between @xmath162 f - modes and g - modes seems to be too strong . in this case
, we would expect the minimum frequency splitting between the @xmath6 modes to be larger than the measured value of @xmath141 .
we would also expect similar frequency splitting to be observed for both the @xmath1 and @xmath139 waves , in contrast with current observations .
the possible existence of stable stratification deep within giant planets has been speculated for decades ( stevenson 1985 ) and may be in accordance with recent theoretical investigations of giant planet interior structures .
wilson & militzer ( 2012a , 2012b ) have claimed that ices and rocks will dissolve in liquid metallic hydrogen at temperatures expected for saturn s core .
this dissolution and ensuing diffusion of heavy elements could set up a stabilizing composition gradient near the core - envelope interface , similar to the stable stratification in our saturn models .
leconte & chabrier ( 2012 ) have proposed that the counteracting effects of stabilizing composition gradients and destabilizing temperature gradients set up layered double - diffusive convective regions in giant planet interiors ( see also leconte & chabrier 2013 ) .
the stabilizing composition gradient leads to a real - valued brunt - vaisala frequency ( at least when averaged over many layers ) which could support g - modes .
alternatively , the layers may become unstable and merge ( rosenblum et al .
2011 , mirouh et al .
2012 , wood et al .
2013 ) , producing much thicker regions of stability / convection .
we have only investigated the effect of a single stably stratified region ; the effect of alternating layers of convection / stratification on the g - mode spectrum is not clear .
unfortunately , our models do not easily reproduce the frequency splittings of @xmath163 between the three observed @xmath1 waves .
although such splitting can occur , it is relatively rare , only occurring in roughly one tenth of our saturn models .
below , we investigate several ideas that could explain this discrepancy .
_ possibility 1 : the observed finely split waves are excited by a single mode .
_ there could be some dynamical mechanism that causes a single mode to excite multiple wave trains in the rings .
this seems unlikely because splitting of this nature is not observed for waves excited at lindblad resonances with saturn s moons .
alternatively , the splitting could be produced if there is some mechanism that periodically modulates the f - mode amplitudes or phases on a timescale of @xmath164 weeks .
there are no obvious dynamical / hydrodynamical mechanisms that could produce this sort of modulation .
_ possibility 2 : our saturn models are fundamentally different from saturn s interior structure . _
perhaps our models lack a feature necessary to reproduce the observed mode spectrum .
we have performed calculations with models containing a single density jump ( which produces a single associated interface mode , for each value of @xmath75 and @xmath2 ) and have found no improvement over simpler models .
however , it may be possible that models containing many density interfaces ( or models including alternating convective / stratified layers produced by double - diffusive convection ) could better match the observations .
alternatively , additional regions of stable stratification may exist within saturn , e.g. , at the molecular - metallic hydrogen phase transition or at a layer of helium accumulation due to helium rain out . _ possibility 3 : we have not included all the necessary pseudo - mode basis functions in our calculations .
_ it is possible that the inclusion of more basis functions in the matrix equation [ mat7 ] will lead to stronger mode mixing with the f - modes .
we have performed various calculations that have included large @xmath75 modes @xmath165 , high radial order g - modes , negative frequency ( retrograde ) modes , inertial modes , and rossby modes .
none of these additions appears to make a substantial qualitative difference in the mode spectrum near the prograde f - modes .
nonetheless , it may be possible that one must include a very large number of pseudo - modes to calculate an accurate normal mode spectrum . _
possibility 4 : a different mechanism causes mode splitting .
_ the inclusion of magnetic fields , differential rotation , meridional circulation , non - spherical corrections to the planetary structure , non - adiabatic effects or some other bit of un - included physics could cause mode splitting .
magnetic effects are unlikely to be important ( see discussion in fuller et al .
2014 ) . to the best of our knowledge
, the other effects listed above can perturb mode frequencies but do not cause mode splitting . _
possibility 5 : mode mixing is enhanced by excluded physics . _ the effects listed above may substantially increase the degree of mode mixing amongst pseudo - modes , allowing strong mixing between f - modes and high @xmath75 g - modes to be more common .
latitudinal differential rotation , in particular , will allow for direct mixing between pseudo - modes with @xmath166 and may make strong mixing more likely to occur .
we find this option to be the most appealing out of all the possibilities listed in this section .
we anticipate that future studies can either eliminate the possibilities listed above or demonstrate their validity .
the detection of additional mode - driven waves in the rings would also help narrow down the possibilities . to determine mode amplitudes ,
we have simply normalized the mode amplitudes such that the @xmath61 , @xmath1 f - mode reproduces the largest observed @xmath1 wave .
this entails dimensionless mode amplitudes of @xmath167 , radial surface displacements of @xmath168 cm , surface radial velocities of @xmath169 cm / s , and mode energies of @xmath170 .
these amplitudes are approximately the same as those calculated by mp93 .
the mechanism responsible for mode excitation remains unclear .
however , we note that energy equipartition amongst modes does not appear to be consistent with observations .
equipartition over - predicts the amplitude of the wave generated by the @xmath152 , @xmath139 f - mode , and predicts that the @xmath155 , @xmath140 f - mode should be observable ( but it has not been detected ) .
this implies that the excitation mechanism favors low frequency and/or low @xmath75 oscillation modes .
the low frequency convective motions near saturn s surface may be a good candidate for mode excitation .
although energy equipartition approximately predicts the correct amplitude of the largest @xmath6 wave ( wave a in table [ table ] ) , it is possible that this wave is generated by a mixed g - mode slightly split in frequency from the @xmath138 , @xmath6 f - mode .
the f - mode may instead create the maxwell gap , as originally postulated by mp93 ( see figure [ m2 ] ) .
this would require the f - mode be split in frequency by less than @xmath171 from the g - mode , similar to the splitting of the @xmath1 modes .
it would also imply the @xmath6 modes have amplitudes larger than suggested by energy equipartition , in accordance with the arguments above .
seismological constraints would be greatly improved with the detection of p - modes via radial velocity techniques .
unfortunately , if saturn s p - modes have energies similar to those of the f - modes , the surface radial velocity variation for a p - mode with angular frequency @xmath172 mhz is only @xmath173 cm / s .
the apparent decrease of mode energy with frequency amongst saturn s f - modes is not encouraging for this prospect .
nonetheless , we encourage searches for p - modes in saturn and jupiter , since they would provide valuable constraints on the interior structure ( jackiewicz et al .
2012 ) .
we can also compare the mode amplitudes calculated here to those claimed to be observed in jupiter via radial velocity techniques by gaulme et al .
( 2011 ) . to do this
, we construct a very simple jupiter model via the same method described in section [ models ] , which has a mass , radius , and @xmath9 that match those of jupiter .
we then calculate @xmath138 mode eigenfunctions for p - modes with frequencies @xmath174 mhz , similar to the frequencies of the gaulme et al .
( 2011 ) modes . finally , we calculate approximate mode amplitudes from the observed radial velocity variations ( of order @xmath175 cm / s ) via @xmath176 $ ] . this procedure results in p - mode amplitudes of @xmath177 and energies of @xmath178 .
as in fuller et al .
( 2014 ) , we find that the gaulme et al .
( 2011 ) measurements entail modes that are more than a thousand times more energetic than the saturn modes ( compared to the gravitational binding energy of each planet ) .
such large p - mode amplitudes seem unlikely , especially in light of the fact that the higher frequency saturn modes appear to have smaller energies .
we therefore remain skeptical that the gaulme et al .
( 2011 ) detections are truly due to jupiter s p - modes
. it may be possible for _ juno _ to detect the gravitational influence of jupiter s oscillation modes .
the gaulme et al .
( 2011 ) mode amplitudes imply potential perturbations of @xmath179 ( in units of jupiter s gravitational potential ) associated with jupiter s p - modes .
however , if jupiter s f - modes have similar energies , they will generate potential perturbations of @xmath180 at periods of @xmath181 hours .
juno _ may be able to detect these gravitational anomalies ( kaspi et al .
if , however , jupiter s f - modes have similar energies to those of saturn ( in units of the planetary binding energy ) , the modes only produce potential perturbations of @xmath182 . our claim that saturn contains stable stratification deep within its interior makes several testable predictions .
first , our models predict that the observed modes are only the tip of the ice berg " , and that saturn hosts a dense spectrum of g - modes in the f - mode frequency regime ( see figures [ m2]-[m4 ] ) that are currently unobserved .
observing these modes through their effect on the rings may be impossible because their gravitational perturbations are too weak to launch detectable waves in the rings .
their surface displacements are correspondingly small because the modes are localized in the g - mode cavity deep within the planet , so observing them with radial velocity measurements may also be very difficult . nonetheless , their detection would be consistent with our theory .
second , our models predict that there may be other , small amplitude waves in the rings excited by @xmath162 g - modes .
for instance , figure [ m2 ] shows an @xmath6 g - mode with a lindblad resonance at @xmath183 km that is on the border of detectability .
the exact frequencies of the detectable g - modes are not robust predictions of our model , we merely speculate that there could be other low amplitude @xmath184 waves separated from the observed waves by a few percent in frequency .
there are several unidentified waves in the c - ring ( see marley 2014 ) that could be excited by these g - modes .
finally , we find the arguments presented in the preceding subsection compelling enough to predict that the maxwell gap is generated by the @xmath138 , @xmath6 f - mode of saturn , as originally proposed in mp93 . indeed
, there is an observed wave train on the eccentric ringlet within the maxwell gap ( porco et al .
2005 , nicholson et al .
2014 , in prep ) , which may be excited by the @xmath138 , @xmath6 f - mode .
we predict this wave train has @xmath6 and that its pattern frequency is @xmath185 .
i thank matt hedman , phil nicholson , yanqin wu , dave stevenson , mark marley , and gordon ogilvie for helpful discussions .
i acknowledge partial support from nsf under grant no .
ast-1205732 and through a lee dubridge fellowship at caltech .
this research was supported in part by the national science foundation under grant no .
nsf phy11 - 25915 .
braviner , h. , ogilvie , g. , 2014 , arxiv:1404.2461v1 dyson , j. , schutz , b. , 1979 , roy soc .
series a , 368 , 389 dahlen , f. , tromp , j. , 1998 , theoretical global seismology , princeton university press eggleton , p. , 2006
, evolutionary processes in binary and multiple stars , cambridge university press fortney , j. , baraffe , i. , militzer , b. , 2009 , arxiv:0911.3154v1 fortney , j. , nettelmann , n. , 2009 , arxiv:0912.0533v1 fuller , j. , lai , d. , storch , n. i. , 2014 , icarus , 231 , 34 fuller , j. , lai , d. , 2014 , arxiv:1406.2717v1 gaulme , p , schmider , f. , gay , j. , guillot , t. , jacob , c. , 2011 a&a , 531 , 104 goldreich , p. , murray , n. , kumar , p. , 1994
, apj , 424 , 466 greenberg , r. , brahic , a. , 1984 , planetary rings , university of arizona press greenspan , h. , 1968 , the theory of rotating fluids , cambridge university press guillot , t. , 2005 , annu .
earth planet .
, 33 , 493 guillot , t. , gautier , d. , 2014 , arxiv:1405.3752v1 hedman , m. , nicholson , p. , 2013
, aj , 146 , 12 jackiewicz , j. , nettelmann , n. , marley , m. , fortney , j. , 2012 , icarus , 220 , 844 kaspi , y. , hubbard , w.b . ,
showman , a.p . ,
flierl , g.r .
, 2010 , georl , 37 , 1204 le bihan , b. , burrows , a. , apj , 764 , 18 leconte , j. , chabrier , g. , 2012 , a&a 540 , a20 leconte , j. , chabrier , g. , 2012 , a&a 540 , a20 lee , u. , saio , h. , 1997 , apj , 491 , 839 marley , m. , 1991 , icarus , 94 , 420 marley , m. , 2014 , icarus , 234 , 194 marley , m. , porco , c. , 1993 , icarus , 106 , 508 mirouh , g. , garaud , p. , stellmach , s. , traxler , a. , wood , t. , 2012 , apj , 750 , 61 nicholson , p.d . , french , r.g . , hedman , m.m . , marouf , e.a . ,
colwell , j.e .
, 2014 , icarus , in press ogilvie , g. i. , lin , d. n. c. , apj , 610 , 477 pena , f.g . , 2010 , tidal dissipation in extrasolar planets , phd thesis , university of toronto porco , c. , et al .
2005 , science , 307 , 1237 press , w.h .
, teukolsky , s.a .
, vetterling , w.t . , flannery , b.p .
, 1998 , numerical recipes ( cambridge univ . press ) rosenblum , e. , garaud , p. , traxler , a. , stellmach , s. , 2011 , apj , 731 , 66 salpeter , e. , apj , 181 , l83 stevenson , d. , 1985 , icarus , 62 , 4 stevenson , d. , salpeter , e. , 1977 , apjs , 35 , 239 vorontsov , s. , zharkov , v. , 1981 , atron . zh .
, 58 , 1101 vorontsov , s. , 1981 , atron .
zh . , 58 , 1275 vorontsov , s. , 1984 , astron .
, 61 , 700 wilson , h. , militzer , 2012 , apj , 745 , 54 wilson , h. , militzer , 2012 , prl , 108 , 111101 wood , t. , garaud , p. , stellmach , s. , 2013 , apj , 768 , 157 wu , y. , 2005 , apj , 635 , 674
here we summarize our method for calculating the effect of rotation on the structure of our saturn models .
we adopt a perturbative method , considering only terms of order @xmath45 .
our method follows eggelton 2006 . in the perturbative method ,
the radius of an elliptical shell is [ rac2 ] r_ac = r , where @xmath42 is the corresponding radius in the spherical model , and @xmath186 is the @xmath138 legendre polynomial .
the relative ellipticity @xmath187 is found by solving clairaut s equation : [ clairaut ] + ( + ) - = 0 .
the central boundary condition is @xmath188 at @xmath189 , while the definition of @xmath190 requires @xmath191 at @xmath192 .
the value of @xmath34 must be normalized according to the value of @xmath45 . to do this
, we first compute the integral [ qclair ] q = ^r_0 8 r^4 ( 5 + r d /dr ) dr .
then the surface ellipticity is ( r ) = and the gravitational moment @xmath9 is j_2 = .
note that both the ellipticity @xmath44 and the value of @xmath9 are proportional to the small number @xmath45 .
here we describe our method to solve for the pseudo - modes of the planet , which will be used as the basis functions for our final mode calculation .
our method uses similar techniques to the analysis of ogilvie & lin ( 2004 ) .
we begin by examining the linearized adiabatic fluid momentum equation [ momeq ] -^2 = - p - - g - 2 i _ s . for a perturbation with time dependence [ conv ] e^i(t +
m ) . equation [ momeq ] ( and all subsequent analysis ) applies in the rotating frame .
the time dependence of equation [ conv ] implies the prograde modes we will be interested in have @xmath193 .
the perturbation variables @xmath194 , @xmath195 , @xmath196 , and @xmath70 are the lagrangian displacement and eulerian perturbations in density , pressure and gravitational potential , respectively .
we consider solid body rotation such that the angular spin frequency is @xmath197 .
equation [ momeq ] applies to a spherical planetary model , we will introduce centrifugal / oblateness effects in a perturbative manner in section [ matrix ] .
we also use the continuity equation [ cont ] + ( r^2 _ r ) + _ _ = 0 , the adiabatic equation of state [ drho ] = p + _ r , and poisson s equation [ poisson ] ( r^2 ) + _ ^2 = 4 g . here
, @xmath15 is the adiabatic sound speed , @xmath14 is the brunt - vaisala frequency , and g is newton s gravitational constant .
we now project the pseudo - modes onto spherical harmonics .
we choose [ xir ] _ r ( r , , ) = u(r ) y_lm ( , ) , [ xit ] _
( r , , ) = v(r ) y_lm ( , ) + i w(r ) y_l+1 , m ( , ) [ xip ] _ ( r , , ) = v(r ) y_lm ( , )
- i w(r ) y_l+1 , m ( , ) , [ psi ] + ( r , , ) = ( r ) y_lm ( , ) , [ phi ] ( r , , ) = ( r ) y_lm ( , ) . in equations [ xit ] and [ xip ] , the horizontal displacements @xmath198 and @xmath199 represent the poloidal and toroidal parts of the horizontal displacement , respectively .
we have chosen the toroidal piece to have @xmath200 dependence so that it has the same symmetry as the poloidal component , therefore a pseudo - mode with an even value of @xmath201 is symmetric across the equator , whereas a pseudo - mode with an odd value of @xmath201 is anti - symmetric across the equator . also note that our definition of the toroidal displacement @xmath69 is different by a factor of @xmath202 from ogilvie & lin 2004 and dt98 .
our choice insures that the values of @xmath67 , @xmath68 , @xmath69 , @xmath203 , and @xmath70 are purely real . as we shall see below
, the pseudo - modes obtained from the projections [ xir]-[phi ] are not actual solutions of the momentum equation [ momeq ]
. they will satisfy equation [ momeq ] only after an integration over angle that eliminates coupling between spherical harmonics of different @xmath75 .
this coupling will be restored when we account for mixing between pseudo - modes of different @xmath75 in [ matrix ] .
equation [ momeq ] is , dropping the coordinate dependence of the variables for convenience @xmath204 @xmath205 @xmath206 where [ q ] q = is the rotation parameter .
additionally , the continuity equation is [ cont2 ] ( - ) y_lm + u y_lm + ( r^2 u ) y_lm - v y_lm = 0 . to simplify equations [ v ] and [ w ] , we exploit the fact that the divergence of a toroidal field is zero , while the curl of a poloidal field is zero . to do this ,
we operate by @xmath207 on equation [ v ] and add it to @xmath208 operated on equation [ w ] .
using @xmath209 and @xmath210 , we have @xmath211.\end{aligned}\ ] ] then , we operate on equation [ v ] by @xmath212 and combine it with @xmath207 operated on equation [ w ] to find @xmath213 \nonumber \\ & + q u \left [ 2 \cos \theta y_{lm } + \sin \theta \frac{\partial}{\partial \theta } y_{lm } \right ] .\end{aligned}\ ] ] we now use the identities [ ylm1 ] y_lm = s_lm y_l-1 , m + s_l+1 , m y_l+1 , m , with [ slm ] s_lm = , and [ ylm2 ] y_lm = l s_l+1,m y_l+1 , m - ( l+1 ) s_l m y_l-1 , m. then , multiplying equation [ v1 ] by @xmath214 and equation [ w1 ] by @xmath215 and integrating over angle , we find [ v2 ] l(l+1 ) v - m q v = + m q u - q l(l+2)s_l+1,m w , and [ w2 ] ( l+1)(l+2 ) w - m q w = q ( l+2)s_l+1,m u - q l(l+2)s_l+1,m v the angular integration has eliminated terms proportional to @xmath216 that cause mixing between pseudo - modes of different @xmath75 , and which will be accounted for in [ matrix ] . equations [ v2 ] and [ w2 ] provide algebraic relations for the values of @xmath68 and @xmath69 at any radius @xmath42
. inspection of equations [ v2 ] and [ w2 ] reveal that in the non - rotating limit @xmath76 we obtain @xmath217 , i.e. , the displacements are purely poloidal . furthermore , when @xmath76 , we recover @xmath218 $ ] which is the algebraic relation typically used in computations of modes in non - rotating bodies
. we can perform similar angular integrations on equations [ cont ] , [ poisson ] , and [ u ] .
we are left with a system of six equations , stemming from the three momentum equations , the second order poisson s equation , and the continuity equation .
these equations are composed of four differential equations and two algebraic relations for the six unknowns @xmath67 , @xmath68 , @xmath69 , @xmath203 , @xmath70 , and @xmath219 : @xmath220 = 0\\ \label{eqn5 } & \left [ l(l+1 ) - m q \right ] v - \frac{l(l+1)}{r \omega^2 } \psi - m q u + q l(l+2 ) s_{l+1 , m } w = 0 \\ \label{eqn6 } & \left [ ( l+1)(l+2 ) - m q \right ] w + ql(l+2)s_{l+1,m } v - q ( l+2 ) s_{l+1,m } u = 0\end{aligned}\ ] ] of course , equations [ eqn1]-[eqn6 ] also contain the additional unknown eigenfrequency @xmath60 , which satisfies a trivial seventh equation [ omeqn ] = 0 . in order to solve the seven equations [ eqn1]-[omeqn ] , we require seven boundary conditions . for pseudo - modes with @xmath221 , the usual boundary conditions apply : @xmath222 and @xmath223 the algebraic relations [ eqn5 ] and [ eqn6 ] constitute two more boundary conditions when applied at @xmath224 or @xmath225 . the final boundary condition is a normalization condition .
the most common choice is @xmath226 at @xmath192 .
however , the normalization is entirely arbitrary , and we use different normalizations to find different types of modes .
one must be cautious with this choice , as the modes have very different scales ( the g - modes , for instance , are localized in the g - mode cavity and have very small perturbations near the surface , whereas the opposite is true of inertial modes ) , and it is difficult to choose a single normalization that allows our relaxation code to reliably find all classes of modes . after we solve for the pseudo - modes , we renormalize them using the condition [ normpseudo ] dv ^ * - dv ^ * ( _ s ) = 1
the reason for this choice is that the pseudo - modes are orthonormal to one another under equation [ normpseudo ] for a given @xmath75 ( see [ matrix ] ) .
however , pseudo - modes of different @xmath75 are coupled to one another by the coriolis force , and we must use the technique described in [ matrix ] to solve for the normal modes .
here we describe our method for solving the mode eigensystem to determine mode eigenfrequencies @xmath60 and eigenfunctions @xmath194 , using the pseudo - modes as basis functions .
we follow the procedure outlined in dahlen & tromp 1998 ( dt98 ) .
an oscillation mode is a solution to the generalized eigenvalue problem [ mat1 ] * z * , where [ z ] * z * = .
the @xmath227 , @xmath228 , @xmath229 operators correspond to potential energy , kinetic energy , and coriolis force operators .
these operators are most conveniently expressed through their inner products with the pseudo - mode displacements : @xmath230 @xmath231 @xmath232 with @xmath233 given by equation 7.36 of dt98 . with our choice of the definition of @xmath69 ( equations [ xit ] and [ xip ] ) , all three operators are real and symmetric such that @xmath234 and likewise for @xmath233 and @xmath235 . in the non - rotating limit ,
equation [ mat1 ] reduces to the more familiar eigensystem , @xmath236 with the orthonomality requirement @xmath237 . including rotation
, the modes must satisfy @xmath238 \bxi=
\omega^2 \mathcal{t } \bxi$ ] with the modified orthonormality relation [ orthonorm ] t _
+ w_. the pseudo - modes we calculate in [ pseudo ] are not normal modes because they do not satisfy equation [ mat1 ] , nor do they satisfy the orthonormality relation [ orthonorm ] for @xmath239 .
however , the pseudo - modes do satisfy the relation [ pseudonorm ] _
| + 2 _ - _ ^2 | _ = 0 .
furthermore , for @xmath240 , the pseudo - modes are orthogonal in the sense that they satisfy [ pseudoorth ] _ | + 2 _ - _ ^2 | _ - _ | + 2 _ - _ ^2 | _ = 0 for . equations [ pseudonorm ] and [ pseudoorth ]
can be verified with a fair amount of algebra from equations [ u]-[w ] and [ eqn1]-[eqn6 ] .
the pseudo - modes therefore satisfy , using our chosen normalization of equation [ normpseudo ] , @xmath241 and @xmath242 the second lines of equations [ tpseudo ] and [ vpseudo ] can be easily verified from equations [ tmat ] and [ vmat ] because the operators @xmath228 and @xmath227 do not induce coupling between spherical harmonics of different values of @xmath75 .
however , the pseudo - modes are coupled across different values of @xmath75 because @xmath243 for @xmath239 .
* z*. the form of the inner products @xmath112 and @xmath113 are quite lengthy .
we refer the reader to equations d.80-d.97 of dt98 for explicit formulae . however , it is important to note some differences in notation .
the relationships between the variables in dt98 and our variables are @xmath244 , and @xmath245 . in short ,
these terms introduce mixing between a pseudo - mode of @xmath246 with other pseudo - modes of @xmath247 and @xmath248 .
since the axial symmetry is maintained , modes only mix if @xmath249 .
to solve equation [ matpert ] for @xmath60 , @xmath250 eigenvalue combinations , we decompose @xmath250 into our pseudo - mode basis : [ xidecomp ] * z * = _ a__. inserting the expansion [ xidecomp ] into equation [ matpert ] , we obtain [ mat2 ] _ a _
( .
we now take the inner product of equation [ mat2 ] with a basis mode conjugate @xmath251 to find @xmath252 \nonumber \\ & = \omega \sum_\beta a_\beta \left [ v_{\alpha \beta } + \delta v_{\alpha \beta } + \omega_\alpha \omega_\beta t_{\alpha \beta } + \omega_\alpha \omega_\beta \delta t_{\alpha \beta } \right ] .\end{aligned}\ ] ] using the pseudo - mode overlap equations [ tpseudo ] and [ vpseudo ] , we have @xmath253 \nonumber \\ & = \omega \left ( 2 \omega_\alpha^2 a_\alpha + \sum_\beta a_\beta \left [ \delta v_{\alpha \beta } + \omega_\alpha \omega_\beta \delta t_{\alpha \beta } \right ] \right).\end{aligned}\ ] ] letting @xmath254 , and dividing by @xmath255 , we have [ mat5 ] _ b_+ _ b_= ( b_+ _ b _ ) . equation [ mat5 ] is a generalized hermitian eigenvalue problem of form [ mat6 ] a * b * = , with the matrix elements of @xmath256 given by _
= _ _ + ( 1 - _ l_l _ ) w _ + v _ and @xmath257 given by _ = _ + v _
+ t_. the diagonal elements @xmath78 are the unperturbed pseudo - mode eigenfrequencies . the matrices @xmath235 , @xmath112 , and @xmath113 induce mixing between pseudo - modes of @xmath240 and @xmath258 . after calculating the matrices @xmath256 and @xmath257
, we numerically solve the matrix equation [ mat7 ] b^-1 a * b * = , for the eigenvalues @xmath60 and eigenvectors @xmath259 .
this method is prone to numerical inaccuracies because the matrix @xmath260 is not symmetric .
however , as long as the matrix @xmath261 is positive - definite ( which it must be for secularly stable planetary models ) , one can use a cholesky decomposition to solve the matrix equation [ mat6 ] ( see press et al .
this decomposition ensures that the matrix involved is symmetric ( making it more amenable for numeric solving algorithms ) and hence that the eigenvalues are all real . in practice , we use both the cholesky method and equation [ mat7 ] , and check that the results are identical .
we caution that the inclusion of very low frequency modes ( such as the rossby modes ) leads to numerical instability .
the reason is that the value of @xmath262 can become large for low frequency modes if the value of @xmath112 has numerical error .
this is nearly unavoidable on a finite radial grid since the value of @xmath112 is computed from radial integrals over oscillatory mode eigenfunctions .
although this issue does not appear to affect our calculations at higher frequencies , our techniques may be numerically ill - suited for reliable calculations of very low frequency modes . | seismology allows for direct observational constraints on the interior structures of stars and planets .
recent observations of saturn s ring system have revealed the presence of density waves within the rings excited by oscillation modes within saturn , allowing for precise measurements of a limited set of the planet s mode frequencies .
we construct interior structure models of saturn , compute the corresponding mode frequencies , and compare them with the observed mode frequencies .
the fundamental mode frequencies of our models match the observed frequencies ( of the largest amplitude waves ) to an accuracy of @xmath0 , confirming that these waves are indeed excited by saturn s f - modes .
the presence of the lower amplitude waves ( finely split in frequency from the f - modes ) can only be reproduced in models containing gravity modes that propagate in a stably stratified region of the planet .
the stable stratification must exist deep within the planet near the large density gradients between the core and envelope .
our models can not easily reproduce the observed fine splitting of the @xmath1 modes , suggesting that additional effects ( e.g. , significant latitudinal differential rotation ) may be important .
[ firstpage ] |
since the discovery of anderson localization in 1958 @xcite , significant efforts both analytical and numerical have been made to understand how localization is affected by interactions . in 2005 ,
basko , aleiner , and altshuler @xcite demonstrated that , under certain conditions , localization can survive in the presence of interactions .
this phenomenon was called many - body localization ( mbl ) .
the mbl state is a peculiar state of matter characterized by a number of counter - intuitive properties including ergodicity breaking @xcite , and it has been attracting a lot of attention recently ( see , e.g. , refs . [ ] and references therein ; for reviews , see refs . [ ] ) .
a different but closely related phenomenon to anderson localization is dynamical localization .
it was first introduced by casati , chirikov , ford , and izrailev @xcite for a prototypical dynamical model of quantum kicked rotor ( qkr)a quantum analog of the classical kicked rotor ( kr ) also known as the chirikov standard map @xcite .
experimentally , it was first observed by moore _
dynamical localization manifests itself in quantum suppression of the chaotic classical diffusion , which for kr occurs in the angular - momentum space when the kicking strength exceeds a critical value .
as opposed to anderson localization in disordered systems , dynamical localization is not related to genuine disorder or intrinsic randomness and is a consequence of deterministic system dynamics .
however , in 1982 fishman _
@xcite showed that the qkr model can be directly mapped onto the anderson model with quasidisorder , and that dynamical localization in qkr corresponds to localization in the anderson - type lattice model .
in particular , in the floquet formalism , the free rotor evolution between the kicks generates a lattice of angular - momentum states ( dimensionless angular momentum is integer due to quantization on a ring ) , and kicking embodies hopping between the `` sites '' of this lattice .
the role of interactions in both anderson and dynamical localization has been studied for a long time . during a few decades
, it was believed that interactions generally destroy localization due to the associated dephasing .
in particular , interactions were studied directly @xcite , and modeled by introducing noise @xcite , dissipation @xcite , and nonlinearity @xcite . some experimental probes
@xcite also tentatively suggested delocalization .
however , in some special cases of two interacting qkrs , dynamical localization was found to be preserved , although weakened specifically , for a single 2d qkr @xcite and for the interaction potential local in rotor angular - momentum space in 1d @xcite .
although dynamical localization for two coupled qkrs was predicted to disappear in the presence of coordinate - dependent interactions @xcite , in 2007 toloui _
@xcite reported localized regimes in two coupled qkrs .
furthermore , recently keser _ et al . _
@xcite showed that coupled many - body systems can possess dynamical localization , and the corresponding phenomenon was dubbed dynamical many - body localization ( dmbl ) .
however , dmbl has been found only for a specific integrable system of linear quantum kicked rotors ( lqkrs ) so far , and the existence of this phenomenon in more general , nonintegrable cases remains unclear . in this paper
, we propose a nonintegrable model of coupled quantum relativistic kicked rotors ( qrkrs ) .
we explicitly show dynamical localization for up to three coupled rotors ( see sec .
[ sec : results ] ) , and independently of these explicit calculations we argue that the dmbl state in a many - body ensemble of such systems is possible for a wide range of parameters without fine - tuning to integrability .
the many - body lqkr hamiltonian has the form @xmath0 where the single - particle part @xmath1 defined in eqs .
( [ generalhamiltonian ] ) , ( [ single_kicking ] ) , and ( [ single_lqkr ] ) depends on angular momentum linearly as @xmath2 .
and this many - body hamiltonian is a high - angular - momentum limit of the many - body qrkr hamiltonian given by @xmath3 where @xmath4 defined in eqs .
( [ generalhamiltonian ] ) , ( [ qrkr_freehamiltonian ] ) , and ( [ single_kicking ] ) depends on angular momentum as @xmath5 , since @xmath6 ( parameter definitions are given below ) .
therefore , any possible delocalization stops at sufficiently high angular momenta where this asymptotic dominates . on the other hand , outside of this asymptotic regime , the classical counterpart of the qrkr model exhibits chaotic behavior , and a quantum anderson - type mechanism is necessary to induce localization .
we show that this mechanism also works to some extent in the presence of interactions and conclude that in the general many - body case , localization can result from an interplay of both effects .
apart from the dynamical localization , we also address regimes where the single relativistic kicked rotor exhibits novel transport behavior and examine them in the interacting case .
we find that interactions facilitate this behavior and increase the number of such regimes .
new interesting transport effects can be found if the asymptotic behavior of the dispersion relation at low angular momentum is different from the behavior at high angular momentum . in 2003 , matrasulov _ et al .
_ @xcite suggested a model of qrkr , a quantum version of a classical relativistic kicked rotor ( rkr ) @xcite .
both rkr and qrkr models naturally possess this dispersion property .
it is important to note that from the viewpoint of the dispersion relation , qrkr interpolates between conventional qkr at low angular momenta and exactly solvable lqkr @xcite at high angular momenta .
recently , zhao _ et al .
_ @xcite discovered rich transport properties of rkr and qrkr that included various regimes ranging from localization to superballistic transport .
in general , the transport properties can be classified by the value of the index @xmath7 in the time dependence of the mean - squared generalized `` coordinate . '' for a rotor , the relevant choice is the angular - momentum space : @xmath8 in the case of pure localization @xmath9 .
the values of @xmath10 correspond to various types of delocalization .
@xmath11 corresponds to the standard diffusion @xmath12 , while the case @xmath13 is called anomalous diffusion . in particular
, @xmath14 is dubbed subdiffusion and @xmath15 is superdiffusion .
the regime with @xmath16 corresponds to ballistic transport .
there is also a special , less studied case of transient anomalous diffusion called superballistic transport that corresponds to @xmath17 .
only a few examples of this regime are known to date @xcite .
interestingly , both rkr and qrkr exhibit the superballistic transport regime @xcite . besides the angular - momentum dynamics
, qrkr also naturally possesses a spin - like degree of freedom , and dynamics in this `` spin '' space is quite peculiar ( strictly speaking , it is the particle - antiparticle space of the 1d dirac equation , but we will refer to it as spin for brevity ) .
the first spinful kicked rotor model
spin-@xmath18 qkr was suggested by scharf @xcite and later studied in refs .
[ ] , but the evolution of the spin in either of models did not receive much attention . in the present paper , first we review the qrkr model and introduce its spin dynamics properties that , to the best of our knowledge , have not been discussed in the literature
. then we numerically demonstrate robust localization upon driving for the model with up to @xmath19 interacting qrkrs .
most importantly , it means that in this model interparticle coupling that corresponds to infinite - range interaction in the respective lattice model does not always destroy few - body localization ( as opposed to the case in refs .
[ ] , but similarly to the one in refs .
more generally , we show that the coupled model inherits the transport regimes of the single qrkr model .
if generalized to a many - body system of qrkrs , this statement results in dmbl similar to that found in ref .
[ ] , but for a nonintegrable system . independently of the numerics , but in agreement with it
, we argue that this is the case because the difference between the dynamics of the many - body qrkrs model and the integrable model considered in ref .
[ ] vanishes as the angular - momentum terms increase and overwhelm the mass terms .
in this section , we review the qrkr model ( see also ref . [ ] ) and study the spin dynamics and spin - momentum entanglement in this model .
we find a number of unusual dynamic regimes involving the spin . from this point on
, we mostly refer to the rotors angular momenta simply as momenta for shortness . as for any kicked system
, the hamiltonian of the qrkr model reads @xmath20 where @xmath21 is a dimensionless time ( measured in the units of the kicking period , @xmath22 ) . throughout the paper
, we use the notation @xmath23 for qrkr , the free part @xmath24 is the dimensionless 1d dirac hamiltonian : @xmath25 where @xmath26 plays the role of an effective speed of light , @xmath27 is an effective mass , and @xmath28 is a vector of pauli matrices .
@xmath29 is a dimensionless angular momentum operator , @xmath30 , and @xmath31 is an angular coordinate of the particle , @xmath32 .
we assume that kicking has the following form : @xmath33 where @xmath34 is an effective kicking strength , and @xmath35 specifies the spatial period of the potential .
note that the kicking potential ( [ single_kicking ] ) is proportional to the unit matrix in the spin space .
spin dynamics with a large mass , zero average initial momentum , and small initial momentum - distribution width @xmath36 ( parameters are shown above the figure ) .
the blue marker indicates the initial point , and the color indicates time ( in the units of the kicking period ) .
the projection of the spin trajectory is shown on the @xmath37 coordinate plane . ]
spin dynamics with a medium mass equal to the average initial momentum , and small initial momentum - distribution width @xmath36 ( parameters are shown above the figure ) .
the blue marker indicates the initial point , and the color indicates time .
the projection of the spin trajectory is shown on the @xmath37 coordinate plane . ] in order to quantify the role of quantum and relativistic effects in qrkr as compared to rkr and qkr , respectively ( see sec . [
sec : results ] for the coupled rotors case ) , we make connection to the actual dirac equation for a kicked relativistic spin-@xmath18 particle of mass @xmath38 confined to a 1d ring of radius @xmath39 : @xmath40\psi,\hspace{-3pt}\ ] ] where @xmath41 is physical time , @xmath42 is the speed of light , @xmath43 is physical ( linear ) momentum , and @xmath44 is the amplitude of the kicking force . introduce a dimensionless `` effective planck constant '' : @xmath45 in the dimensionless dirac equation , we absorb @xmath46 into the other parameters , so that the hamiltonian ( [ generalhamiltonian ] ) , ( [ qrkr_freehamiltonian ] ) , ( [ single_kicking ] ) enters it as follows : @xmath47\psi,\ ] ] where @xmath48 and @xmath49 is the dimensionless kicking strength in the nonquantum rkr model [ see eq .
( [ rkr_ham ] ) below ] .
note that this straightforward interpretation is not related to the feasible physical realizations .
some of the latter are proposed in sec .
[ sec : proposals ] .
consider integer times @xmath21 only .
since the hamiltonian is periodic in time@xmath50and the external potential has a kicking form , the stroboscopic single - period evolution of the wave functions governed by eq .
( [ dirac_equation ] ) is given by the floquet operator @xmath51 as @xmath52 where @xmath53\exp\left[-iv(x)\right ] \\
\nonumber & = \exp\left[-i\left(2\pi\alpha p\hat\sigma^x + m\hat\sigma^z\right)\right]\exp\left[-ik\cos(qx)\right].\end{aligned}\ ] ] spin dynamics with a small mass , high average initial momentum , and medium initial momentum - distribution width @xmath36 ( parameters are shown above the figure ) .
the blue marker indicates the initial point , and the color indicates time .
the projection of the spin trajectory is shown on the @xmath37 coordinate plane . ]
entanglement entropy ( solid blue line , left axis ) and momentum variance @xmath54 ( dashed red line , right axis ) in the process corresponding to fig
. [ small_m_high_p0 ] .
parameters are shown above the figure . ]
entanglement entropy ( solid blue line , left axis ) and momentum variance @xmath54 ( dashed red line , right axis ) in the process corresponding to fig .
[ small_m_high_p0 ] .
parameters are shown above the figure . ]
an efficient way of calculating the evolution ( [ evol ] ) numerically is by switching from the coordinate representation to the momentum representation back and forth , applying each part of the floquet operator ( [ floquet ] ) in its eigenbasis .
this approach allowed us to reproduce the results for single qrkr obtained in ref .
the details of the numerical implementation of this method are given for coupled qrkrs in sec .
[ twoqrkrs ] . as we mentioned above , in the high - momentum region the qrkr model can be approximated by the lqkr model .
it is determined by hamiltonian ( [ generalhamiltonian ] ) with the free part : @xmath55 and kicking potential ( [ single_kicking ] ) .
this model has been proved to be integrable in any dimension by figotin and pastur @xcite . during the free evolution between the kicks , the local eigenspinors of @xmath56 ,
i.e. the eigenspinors at any given @xmath29 as a parameter , acquire the phases @xmath57 , where the quantized momentum @xmath29 takes only integer values .
as discussed in ref .
[ ] , the transport regime in qrkr is determined by the phases @xmath58 along with the kicking potential parameter @xmath59 and initial conditions .
specifically , in the low - momentum region ( @xmath60 ) , qrkr is always localized ( for the same reason as qkr and lqkr ) , because propagators @xmath61 $ ] act as quasi - random - number generators for a sequence of integers @xmath29 . in the high - momentum limit
, qrkr tends to lqkr , and the behavior of @xmath62 $ ] is determined by the rationality of @xmath63 .
namely , irrational values of @xmath63 give rise to the localized phase , and rational values@xmath64lead to delocalization ( in particular , to ballistic transport ) if @xmath65 . in the remaining case of @xmath66 , but @xmath67 , the dynamics is bounded , but this is not related to the anderson - type localization @xcite . in the general case of the wave function containing components with @xmath68 , an additional pattern
the superballistic transport arises due to the leakage of the wave function from the low - momentum region to the high - momentum one @xcite .
specifically , following the qualitative argument from ref .
[ ] , there are three contributions to the momentum - space transport .
one contribution is constant and comes from localization in the disordered region at small momenta .
another one is ballistic ; i.e. , the momentum variance grows quadratically in time .
it comes from the periodic nondisordered ( in terms of @xmath69 $ ] ) high - momentum region .
and the third contribution is superballistic .
it is related to the transfer of population from the moderate - momentum to the high - momentum region and can be qualitatively described by @xmath70 , where @xmath71 is a population transfer rate and @xmath72 is a coefficient of ballistic transport . in the simplest case of @xmath73 ,
this integral readily gives cubic growth of momentum variance .
besides the rich dynamics that qrkr shows in the momentum space , it also possesses very peculiar patterns in spin dynamics , even if the kicking is spin - independent , as in eq .
( [ single_kicking ] ) .
these patterns are related to the entanglement between the spin and momentum degrees of freedom that occurs at each step as a result of the combination of free evolution and kicking .
we performed a series of calculations of the spin evolution in the case of spin - independent kicking ( [ single_kicking ] ) . in figs .
[ large_m_q_3 ] [ small_m_high_p0 ] and [ resonancespin ] , we show trajectories of the tip of the spin vector more precisely , the vector @xmath74within the bloch sphere in four representative regimes . figs .
[ large_m_q_3 ] [ small_m_high_p0 ] correspond to the localized phase , while fig .
[ resonancespin ] describes the spin dynamics in the delocalized phase . as an initial state , we chose a gaussian @xmath75\chi_\sigma$ ] , where @xmath76 in fig .
[ large_m_q_3 ] and @xmath77 in all other cases . in fig .
[ large_m_q_3 ] , @xmath78 is much larger than the initial momentum spread centered around zero , and due to localization , the mass remains two orders of magnitude greater than @xmath79 for the highest populated momentum components . in this case
, the spin - tip trajectory is a flat disk that lies in the @xmath80 plane and constitutes rotation via @xmath81 $ ] .
the radius of the spin trajectory is determined by the degree of the spin - momentum entanglement and is oscillating in time within certain boundaries . in fig .
[ medium_m_med_p0 ] , the mass @xmath27 and the initial momentum @xmath82 are equal , which leads to the flat trajectory tilted at an angle@xmath83to the @xmath84 plane .
when the ratio between @xmath85 and @xmath27 is varied , the trajectory remains flat in a certain range of parameters , and only tilt angle changes accordingly .
spin dynamics in a resonant regime corresponding to delocalization ( @xmath63 is rational ) .
parameters are shown above the figure .
the blue marker indicates the initial point , and the color indicates time .
the projections of the spin trajectory are shown on the @xmath80 and @xmath37 coordinate planes . ] in fig .
[ small_m_high_p0 ] , @xmath86 , and the momentum spread @xmath87 . in this case , we have alternating regimes of dynamics .
when the majority of the momentum - space population is far away from @xmath88 , the spin - tip trajectory is flat in this case , it constitutes rotation in the @xmath84 plane due to the action of @xmath89 $ ] .
the wave function dynamics far away from @xmath90 consists of periodic splitting into two parts and recombining back .
one of these parts corresponds to classical acceleration due to in - phase kicking and another one to deceleration due to out - of - phase kicking . as a manifestation of localization
, these parts span only very limited vicinity of the initial wave packet ; fig .
[ small_m_high_p0_wf](a ) shows these parts at the largest separation alongside the initial state .
when the wave function components are split , the spin tip stays very close to some point @xmath91 at the @xmath92 axis . and
when the components recombine , the spin tip comes to the surface of the bloch sphere developing a flat part of the trajectory that is parallel to the @xmath84 plane and crosses the point @xmath91 .
however , if the initial wave function is centered close enough to @xmath90 , as is the case in fig .
[ small_m_high_p0 ] , one of the split components goes through this point , and the spin tip starts to move along the @xmath92 axis until that component leaves the vicinity of @xmath93 .
once the components of the wave function recombine , a new flat disk parallel to @xmath84 plane develops , and then this periodic pattern continues with the @xmath92-motion until the next disk is generated .
when the motion along the @xmath92 axis brings the spin tip to the surface of the bloch sphere , the @xmath92-motion reflects off it and continues back to the center of the bloch sphere . during this motion ,
each time a nonvanishing part of the wave function passes through @xmath93 , it gets modulated , split , and eventually becomes very noisy in that region [ see figs . [
small_m_high_p0_wf](b)[small_m_high_p0_wf](d ) ] . in fig . [ entent ] , we show the entanglement entropy as a function of time corresponding to this case . it is defined as @xmath94,\vspace{-10pt}\ ] ] where @xmath95 phase map of rkr in a strong kicking regime .
parameters : @xmath96 , @xmath97 , @xmath98 . at low momenta , chaotic trajectories span most of the phase space . at high momenta ,
all trajectories are periodic . ]
phase map of rkr in a strong kicking regime .
parameters : @xmath96 , @xmath97 , @xmath98 . at low momenta , chaotic trajectories span most of the phase space . at high momenta ,
all trajectories are periodic . ]
the sharply pronounced dips in the entanglement entropy correspond to the flat disk structures in the bloch sphere with their edges coming close to the surface of the sphere .
and the envelope of the entanglement entropy corresponds to the motion of the spin tip along the @xmath92 axis .
as the time goes , the wave function becomes more and more noisy , and no sharp disk structures are generated for some time .
this corresponds to the region between @xmath99 and @xmath100 kicks in figs .
[ small_m_high_p0 ] and [ entent ] . however , at some point , the revival of the disks structure in the bloch sphere occurs , and the corresponding revival of the dips structure in the entanglement entropy can be seen . as the wave function goes through the point @xmath93 many times ,
it splits again , and becomes randomized , so that eventually , the motion within the bloch sphere becomes less regular .
however , it retains the features described above for at least as long as @xmath101 kicks .
figure [ resonancespin ] shows the spin dynamics in the case of resonant value of @xmath63 , i.e. , delocalization in the momentum space . in this regime ,
the motion of the spin tip is continuously slowing down .
as time goes , more and more momentum components get populated , and the spin - tip trajectory tends to one limiting point inside the bloch sphere .
phase map of rkr in a very strong kicking regime .
parameters : @xmath96 , @xmath102 , @xmath98 . at low momenta ,
all trajectories are chaotic . at high momenta , all trajectories are still periodic . ]
before discussing coupled qrkrs , we need to address the difference between the classical models of rkr and nonrelativistic kr with respect to the notion of critical kicking strength .
rkr is described by the dimensionless hamiltonian : @xmath103 according to ref .
[ ] , as opposed to kr , in rkr , different kolmogorov - arnold - moser ( kam ) tori in the phase space are destroyed at different critical values of the kicking strength @xmath104 , which depend on the parameter @xmath105 .
most importantly , there are global limiting tori at high momentum that do not get destroyed at any finite value of @xmath49 if @xmath106 ( the latter condition is always satisfied in the quantum localized phase described in sec .
[ sec : qrkr ] , i.e. , when @xmath107 ) .
this behavior is illustrated in figs .
[ phasemapweak ] [ phasemapverystrong ] .
variegated regions of moderate momentum filled with chaotic trajectories are always bounded from both sides by global regular trajectories that span the rest of the phase space . however , although the existence of the limiting tori guarantees classically bounded trajectories , it is not the only source of localization exhibited by the qrkr , even when coupling is introduced .
the qrkr shows localization within both classically regular and classically chaotic regions .
therefore , in general , localization is caused by a combination of the classically bounded phase space and quantum anderson - type localization .
the same argument holds for coupled qrkrs .
we illustrate it in sec .
[ sec : results ] .
the many - body generalization of the lqkr model was considered in ref . [ ] , and it was shown analytically that the many - body lqkr model may exhibit the dmbl phase . in other words , it was shown that localization may survive in the presence of interactions .
this finding partially motivated the present study of the few - body generalization of the qrkr model , which is qualitatively distinct from the lqkr model due to the nonintegrability .
in this section , we consider the simplest interacting cases : a two - body and a three - body coupled - qrkr models .
specifically , we consider the models with the hamiltonian @xmath108 where @xmath24 is chosen in two different ways .
first , for the two - body case , we consider the sum of the dirac hamiltonians of the noninteracting qrkrs : @xmath109 here @xmath110 and @xmath111 are @xmath112 matrices , @xmath113 .
second , we use a spinless version of the qrkr to construct @xmath114 as mentioned in sec . [
sec : qrkr ] , it has been shown in ref . [ ]
that the spinless model @xmath115 possesses the same localization properties as the spinful qrkr . for three coupled particles
, we only use the spinless @xmath24 [ eq . ( [ eq : spinless_h0 ] ) ] to reduce computational complexity .
the kicking potential has a standard form : @xmath116 the interaction part is chosen in the same way as for coupled qkrs in refs .
[ ] , which generalizes the potentials considered in refs .
[ ] : @xmath117\},&\\ & l=2,\;3.&\end{aligned}\ ] ] we study these models numerically , and show that the localized phase is persistent with respect to the interaction in a certain range of parameters . the floquet operator @xmath118\exp\left[-i(v + h_{\rm int})\right].\ ] ] for @xmath24 in eq .
( [ eq : spinful_h0 ] ) , @xmath119 is a four - component function .
its four components correspond to the four possible spin configurations of two particles : @xmath120 , respectively . the free part of the floquet operator ( [ floquet - two ] ) in the spinful case is calculated using the properties of the pauli matrices : @xmath121 = \bigotimes\limits_{\ell=1}^2\left[\cos\left(\sqrt{c_\ell^2p^2_\ell + m_\ell^2}\right)\mathbb{i}\right .
\\ & & \left .- i\sin\left(\sqrt{c_\ell^2p^2_\ell + m_\ell^2}\right)\dfrac{c_\ell p_\ell \hat\sigma^x + m_\ell \hat\sigma^z}{\sqrt{c_\ell^2p^2_\ell + m_\ell^2 } } \right]\end{aligned}\ ] ] and can be efficiently applied numerically to a four - component wave function on a momentum - space grid at each step .
similarly , the kicking and interaction part of the evolution operator ( [ floquet - two ] ) can be efficiently applied numerically to a wave function on a coordinate - space grid .
we choose an initial wave function in the basis of momentum eigenstates : @xmath122 where @xmath123 or @xmath124 and , to make one step in time , we perform the discrete fourier transform of @xmath125 to go to the coordinate representation , where the potential part of the floquet operator ( [ floquet - two ] ) is diagonal , and we apply this part to the vector representing the coordinate - space wave function .
after that , we perform the inverse fourier transform to go back to the momentum representation and apply the free part [ operator ( [ mom_floquet ] ) for the spinful case ] to it
. then this cycle starts over for the next step .
this scheme allows us to achieve efficient numerical evolution that only requires application of diagonal operators and a fast fourier transform .
in figs . [ localizationstandard ] [ resonance ] , we present time dependence of the average mean - squared momentum per particle for two coupled spinful qrkr particles : @xmath126 . in ref .
[ ] , this quantity was shown to be a reliable indicator of dynamical localization ( as opposed to the average energy ) .
different values of the parameters determine various regimes that are exhibited by our model .
we start in a gaussian - shaped initial state in the momentum space centered around the point @xmath127 with both spins up : @xmath128 \left|\uparrow\uparrow\right>,\ ] ] average mean - squared momentum per particle in the localized regime .
oscillating red line shows the calculation result .
straight blue line is a linear fit that shows no slope up to a fitting error ( equation is given inside the plot ) .
parameters : @xmath129 ( @xmath130 ) .
insets : probability density at the final time @xmath131 kicks as a function of each momentum while integrated over the other one . ] and widths @xmath132 .
therefore , the initial value of the average mean - squared momentum per particle is @xmath133 .
figures [ localizationstandard ] and [ localization_heff_1 ] represent the dynamically localized state .
it is characterized by nonresonant values of the velocities ( @xmath134 ) .
saturation of the average mean - squared momentum per particle is verified by linear fits that have zero slope up to a fitting error ; the corresponding equations are shown inside the plots .
the insets in these figures show the probability density at the final time @xmath135 as a function of each of the momenta while integrated over the other one . as one can see from these insets , in the regime of localization , the wave functions decay exponentially with momenta and , in the vicinity of numerical boundaries , reach the values below @xmath136 .
this ensures that during the evolution , the population does not come close to the numerical boundaries , and there is no unphysical reflection from them .
stable exponential decay of a wave function with a constant bound on its width indicates localization .
average mean - squared momentum per particle in the localized regime .
oscillating red line shows the calculation result .
straight blue line is a linear fit that shows no slope up to a fitting error ( equation is given inside the plot ) .
parameters : @xmath137 .
insets : probability density at the final time @xmath131 kicks as a function of each momentum while integrated over the other one . ] in fig .
[ localizationstandard ] , we take parameters similar to those used in ref . [ ] for the single qrkr and add @xmath138 of interaction , i.e. , @xmath139 .
these parameters correspond to @xmath140 . in fig .
[ localization_heff_1 ] , we set @xmath141 and periodicity parameter @xmath142 and obtain more stable localization .
other parameters in this case are such that in the corresponding rkr model , many tori are destroyed giving way to a broad chaotic region . in particular , the kicking strength constant exceeds the first ( and , in this case , the only ) critical constant@xmath143corresponding to the single rkr .
figure [ resonance ] shows the delocalized phase described by the resonant values of the velocities : @xmath144 . even though @xmath145so that for single qrkr , such @xmath63 guarantees localization according to ref .
[ ] , in the many - body lqkr model , equal values of the velocities correspond to additional resonances due to the interaction that lead to divergence of the emergent momenta - containing integrals of motion ( ioms ) present in the many - body lqkr model @xcite .
average mean - squared momentum per particle in the superballistic regime .
black points show the calculation result .
red line is a power fit .
parameters : @xmath146 .
inset : probability density at the final time @xmath147 kicks as a function of each momentum while integrated over the other one . ] in the many - body qrkr model at large momenta , these ioms become approximate .
nevertheless , their divergence results in the divergence of the associated momenta , which is confirmed by our numerical results . in particular , in fig .
[ resonance ] we see a rapid transport in the momentum space that causes the fast growth of the probability density near the numerical boundaries of the momentum grid .
unfortunately , this complication makes further numerical analysis at large time scales inefficient .
however , we can clearly see the signs of superballistic transport in this plot .
in particular , in this example , @xmath148 until the wave function reaches the grid boundaries at the time beyond @xmath149 kicks , and we can not rely on the numerics after that point . the inset shows the probability density at final time @xmath135 as a function of each of the momenta while integrated over the other one ( @xmath150 and @xmath151 dependencies are the same in this case due to symmetry ) . in fig .
[ fig : two_vs_three_spinless ] , we compare localization in two - particle and three - particle spinless qrkr models at the same respective parameters . the plots are given in the lin - log scale to show details of saturation .
we should note that upon increasing the interaction strengths , there appear regimes of long - lasting logarithmic growth of the average mean - squared momentum per particle that may be generic for coupled nonintegrable dynamical systems but also satisfy the condition @xmath9 in eq .
( [ eq : transport ] ) . as one can see , in fig .
[ fig : two_vs_three_spinless ] , panel ( a ) , the localization length and time it takes the mean - squared momentum to saturate increase with the number of particles , as expected given the increased contribution of interactions
. however , this panel shows saturation well below the integrable region determined by @xmath152 .
in contrast to it , in panel ( b ) , the near - integrability threshold is @xmath153 , and the saturated value satisfies this condition to some extent .
notice that in this case , the saturated values of mean - squared momentum per particle are in the same range for two and three particles .
average momentum variance per particle in the localized regime for two ( blue lines ) and three ( red lines ) coupled spinless qrkrs .
panel ( a ) shows localization far from integrable limit ; parameters are @xmath154 .
panel ( b ) shows localization close to the integrable limit .
parameters are the same as in panel ( a ) except for @xmath155 being multiplied by a factor @xmath156 . ]
due to the structure of the floquet operators ( [ floquet ] ) and ( [ floquet - two ] ) for single and coupled qrkrs respectively , as well as for any kicked hamiltonian , quantum dynamics is almost invariant with respect to swapping the free and the kicked parts ; i.e. , the hamiltonians @xmath157 and @xmath158 generate the same floquet evolution .
more precisely , in order to get completely the same dynamics , when swapping , one should also change from considering evolution between the moments of time right after the kicks to those just before the kicks and vice versa .
in particular , the single qrkr is equivalent to a model with the hamiltonian @xmath159 let us put @xmath142 .
recall that the angular coordinate @xmath32 and the dimensionless angular momentum is quantized@xmath160 .
then one can establish a correspondence between the qrkr and a spin-@xmath161 particle hopping on a 1d lattice subject to a pulsed magnetic field .
this correspondence is summarized in the following table .
[ cols="^,^,^,^,^,^ " , ] then in the rotating wave approximation , @xmath162 where @xmath163 and @xmath164 are the ground and excited states of the atom in the rotating frame .
the same caveat regarding the quadratic kinetic term as in the previous setting applies here . as well as in the previous examples
, one could kick the first part of this hamiltonian instead of applying a pulsed lattice .
similar models can be constructed on the basis of the qkr and lqkr models . in particular ,
@xmath165 and @xmath166 correspond to a charged particle in a 1d lattice in the presence of a kicked electric field .
this field is linear in space for @xmath167 and uniform for @xmath168 .
extensions of the single - particle models ( [ qrkr_exp ] ) ( [ lqkr_exp ] ) to the case of many interacting particles can be mapped to corresponding many - body qrkrs , qkrs , or lqkrs in the same way .
hamiltonians ( [ qrkr_exp ] ) ( [ lqkr_exp ] ) might be realized in cold atoms in optical lattices .
interestingly , according to the mapping @xmath169 , for such systems , dynamical localization as well as other intriguing transport regimes such as superballistic transport , take place in real space rather than in momentum space , which makes these phenomena especially demonstrative in experiment .
starting with the single - particle qrkr model that possesses the rich variety of transport phases , we introduced its peculiar spin dynamics phenomenology and generalized it to the model of interacting qrkrs .
for the models of two and three coupled qrkrs , we showed that the transport regimes and , in particular , the localized phase can survive interactions .
we are not aware of any previous study of coupled qrkrs , but we point out that for the well - studied coupled qkrs and related static lattice models , most works predict delocalization at least for infinite - range interaction .
our calculations indicate the existence of the localized regimes for such a coupling of two and three qrkrs . unfortunately , exact numerical study of the many - body qrkr model is not feasible presently .
however , at high momenta , it can be approximated by the integrable many - body lqkr model , and this approximation works only better as the system goes to higher momentum states . in ref . [ ] , the many - body lqkr model was analytically shown to exhibit the dmbl phase .
besides that , as opposed to the case of qkr , the classical model behind qrkr is not chaotic at high momenta .
as we have shown , in the cases of two and three coupled qrkrs , localization has a quantum origin and does not rely on the existence of kam tori in the phase space of the corresponding classical problem .
however , for a large number of interacting rotors , if this localization happens to deteriorate completely , and the growth of the particles momenta recovers , at high enough momenta the system will enter the integrable regime and get localized .
this is our main argument in favor of dmbl in the nonintegrable system of the many - body qrkr model . in general
, the observable dynamical localization can represent a nontrivial interplay of both effects that may be difficult to disentangle . in summary , our argument supplemented by few - body calculations
provide a strong hint that the nonintegrable many - body qrkr model should exhibit dynamically localized many - body states .
in addition , we propose a class of kicked lattice models that map onto various kicked - rotor models and can be realized in the framework of cold atoms in optical lattices .
this realization might allow one to study dynamical localization including dmbl , and other anomalous transport phenomena exhibited by the qrkr and its many - body versions in experiment .
this work was supported by us - aro ( contract no .
w911nf1310172 ) , nsf - dmr 1613029 , and the simons foundation .
the authors are grateful to aydin cem keser , gil refael , trey porto , steve rolston , varun vaidya , borzumehr toloui , and victor yakovenko for valuable discussions . 88 ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1 '' '' @noop [ 0]secondoftwo sanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12 @startlink[1 ] @endlink[0 ] @bib@innerbibempty link:\doibase 10.1103/physrev.109.1492 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/j.aop.2005.11.014 [ * * , ( ) ] https://arxiv.org/abs/1312.0577 [ ( ) ] link:\doibase 10.1103/physrevlett.95.206603 [ * * , ( ) ] link:\doibase 10.1103/physrevb.77.064426 [ * * , ( ) ] link:\doibase 10.1103/physrevb.82.174411 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.114.140401 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.115.030402 [ * * , ( ) ] link:\doibase 10.1146/annurev - conmatphys-031214 - 014726 [ * * , ( ) ] , link:\doibase 10.1146/annurev - conmatphys-031214 - 014701 [ * * , ( ) ] in link:\doibase 10.1007/bfb0021757
[ _ _ ] , , vol . , ( , ) pp .
link:\doibase 10.1007/bf01029131 [ * * , ( ) ] , @noop * * , ( ) \doibase http://dx.doi.org/10.1016/0370-1573(79)90023-1 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.75.4598 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.49.509 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.61.659 [ * * , ( ) ] link:\doibase 10.1143/ptps.98.109 [ * * , ( ) ] http://inspirehep.net/record/381891 [ ( ) ] http://arxiv.org/abs/chao-dyn/9711021 [ ( ) ] link:\doibase 10.1143/ptps.116.393 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/0378-4371(94)00297-7 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.75.1598 [ * * , ( ) ] @noop * * , ( ) http://arxiv.org/abs/cond-mat/9905231 [ ( ) ] link:\doibase 10.1103/physrevb.61.4588 [ * * , ( ) ] link:\doibase 10.1007/s100510170222 [ * * , ( ) ] link:\doibase 10.1103/physreva.70.013401 [ * * , ( ) ] http://stacks.iop.org/0253-6102/51/i=3/a=17 [ * * , ( ) ] http://stacks.iop.org/0253-6102/54/i=2/a=09 [ * * , ( ) ] link:\doibase 10.1103/physreve.81.046201 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/0167-2789(83)90318-4 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.53.2187 [ * * , ( ) ] link:\doibase 10.1103/physreva.43.639 [ * * , ( ) ] link:\doibase 10.1103/physreva.44.2292 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.67.1945 [ * * , ( ) ] http://stacks.iop.org/0305-4470/27/i=14/a=011 [ * * , ( ) ] link:\doibase 10.1103/physreve.52.2497 [ * * , ( ) ] http://stacks.iop.org/0295-5075/35/i=7/a=517 [ * * , ( ) ] link:\doibase 10.1103/physreva.77.062113 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/0003-4916(90)90279-w [ * * , ( ) ] link:\doibase 10.1103/physreva.42.4647 [ * * , ( ) ] link:\doibase 10.1103/physreva.51.3136 [ * * , ( ) ] link:\doibase 10.1103/physreva.44.r3423 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.70.1787 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.100.094101 [ * * , ( ) ] http://stacks.iop.org/1751-8121/47/i=33/a=335101 [ * * , ( ) ] http://stacks.iop.org/0305-4470/37/i=44/a=l06 [ * * , ( ) ] link:\doibase 10.1103/physreve.79.026205 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.102.024101 [ * * , ( ) ] link:\doibase 10.1103/physreve.80.056212 [ * * , ( ) ] http://stacks.iop.org/0295-5075/96/i=3/a=30004 [ * * , ( ) ] http://dx.doi.org/10.1038/srep02359 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/0375-9601(96)00623-8 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.81.1203 [ * * , ( ) ] link:\doibase 10.1103/physreve.64.056233 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.92.054101 [ * * , ( ) ] link:\doibase 10.1103/physreva.70.041602 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.110.190401 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.60.867 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.73.2607 [ * * , ( ) ] http://stacks.iop.org/0295-5075/30/i=7/a=005 [ * * , ( ) ] http://stacks.iop.org/0295-5075/31/i=3/a=008 [ * * , ( ) ] http://stacks.iop.org/0951-7715/8/i=5/a=013 [ * * , ( ) ] \doibase 10.1051/jp1:1996149 [ * * , ( ) ] http://arxiv.org/abs/cond-mat/9603086 [ ( ) ] http://summit.sfu.ca/item/8122 [ ] http://arxiv.org/abs/0903.4632v2 [ ( ) ] link:\doibase 10.1103/physrevb.94.085120 [ * * , ( ) ] http://stacks.iop.org/1402-4896/68/i=4/a=004 [ * * , ( ) ] link:\doibase 10.1103/physreve.72.016213 [ * * , ( ) ] link:\doibase 10.1103/physreva.40.4072 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.49.833 [ * * , ( ) ] link:\doibase 10.1103/physrevb.29.6500 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/0167-2789(84)90185-4 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/0003-4916(85)90196-4 [ * * , ( ) ] link:\doibase 10.1103/physreve.90.022921 [ * * , ( ) ] link:\doibase 10.1103/physreve.64.012301 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.108.070603 [ * * , ( ) ] link:\doibase 10.1364/ol.38.004675 [ * * , ( ) ] http://stacks.iop.org/1367-2630/17/i=11/a=113030 [ * * , ( ) ] http://stacks.iop.org/0305-4470/22/i=19/a=016 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.72.72 [ * * , ( ) ] link:\doibase 10.1140/epjb / e2005 - 00002 - 2 [ * * , ( ) ] link:\doibase 10.1103/physrevb.84.115133 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.113.216802 [ * * , ( ) ] link:\doibase 10.1103/physrevb.93.075403 [ * * , ( ) ] http://projecteuclid.org/euclid.cmp/1103941644 [ * * , ( ) ] | a periodically driven rotor is a prototypical model that exhibits a transition to chaos in the classical regime and dynamical localization ( related to anderson localization ) in the quantum regime . in a recent work [ phys .
rev .
b * 94 * , 085120 ( 2016 ) ] , a. c. keser _ et al . _
considered a many - body generalization of coupled quantum kicked rotors , and showed that in the special integrable linear case , dynamical localization survives interactions . by analogy with many - body localization ,
the phenomenon was dubbed dynamical many - body localization . in the present work , we study nonintegrable models of single and coupled quantum relativistic kicked rotors ( qrkrs ) that bridge the gap between the conventional quadratic rotors and the integrable linear models . for a single qrkr ,
we supplement the recent analysis of the angular - momentum - space dynamics with a study of the spin dynamics .
our analysis of two and three coupled qrkrs along with the proved localization in the many - body linear model indicate that dynamical localization exists in few - body systems .
moreover , the relation between qrkr and linear rotor models implies that dynamical many - body localization can exist in generic , nonintegrable many - body systems . and localization can generally result from a complicated interplay between anderson mechanism and limiting integrability , since the many - body linear model is a high - angular - momentum limit of many - body qrkrs .
we also analyze the dynamics of two coupled qrkrs in the highly unusual superballistic regime and find that the resonance conditions are relaxed due to interactions .
finally , we propose experimental realizations of the qrkr model in cold atoms in optical lattices . |
algebras of hilbert space operators may be mapped onto algebras of ordinary functions on linear spaces , with an associative but non - commutative _ star product _ ( see , e.g. @xcite ) .
the images of the hilbert space operators are called _ operator symbols .
_ weyl maps @xcite , @xmath0-ordered operator symbols @xcite , their partial cases @xcite and tomograms @xcite are examples of this correspondence between hilbert space operator algebras and function algebras @xcite . in the case of tomograms ,
the operator symbols of the density operators of quantum mechanics are families of ordinary probability distributions @xcite .
a unified framework for operator symbols is presented in sect .
2 and their main properties are reviewed .
many of the results in sects .
2 and 3 are scattered in previous publications and are collected here to make the paper reasonably self - contained .
finite dimensional systems ( spin tomograms ) are studied in sect .
these operator symbols are then proposed as a framework for quantum information problems .
qudit states are identified with maps of the unitary group into the simplex .
the image of the unitary group on the simplex provides a geometrical characterization of the nature of the quantum states . in the remaining sections ,
generalized measurements , typical quantum channels , entropies and entropy inequalities are discussed in this setting .
in quantum mechanics , observables are selfadjoint operators acting on the hilbert space of states @xmath1 .
we map operators onto functions in a vector space @xmath2 in the following way : given the hilbert space @xmath1 and a trace - class operator @xmath3 acting on this space , let @xmath4 be a family of operators on @xmath1 , labelled by vectors @xmath5 . we construct the @xmath6-number function @xmath7 ( and call it the _ symbol of the operator _
@xmath3 ) by @xmath8 . \label{eq.a1}\ ] ] let us suppose that relation ( [ eq.a1 ] ) has an inverse , i.e. , there is a set of operators @xmath9 acting on the hilbert space such that @xmath10 equations ( [ eq.a1 ] ) and ( [ eq.a2 ] ) define an invertible map from the operator @xmath3 onto the function @xmath11 . multiplying both sides of eq.([eq.a2 ] ) by the operator @xmath12 and taking the trace ,
one obtains a consistency condition for the operators @xmath13 and @xmath9 @xmath14 = \delta \left ( \mathbf{x - x}^{\prime } \right ) .\ ] ] for two functions @xmath11 and @xmath15 , corresponding to two operators @xmath3 and @xmath16 , a star - product is defined by @xmath17 .
\label{eq.a5}\ ] ] since the standard product of operators on a hilbert space is associative , eq . ( [ eq.a5 ] ) also defines an associative product for the functions @xmath18 , i.e. , @xmath19 let us suppose that there is another map , analogous to the one in ( [ eq.a1 ] ) and ( [ eq.a2 ] ) , defined by the operator families @xmath20 and @xmath21
. then one has @xmath22 \label{eq.a21}\ ] ] and the inverse relation @xmath23 the function @xmath11 will be related to the function @xmath24 by @xmath25 \,d\mathbf{x } \label{eq.a23}\ ] ] with the inverse relation @xmath26 \,d\mathbf{y}. \label{eq.a24}\ ] ] the functions @xmath11 and @xmath27 , corresponding to different maps , are connected by the invertible integral transform given by eqs.([eq.a23 ] ) and ( [ eq.a24 ] ) with the intertwining kernels @xmath28 \label{eq.a28aa}\ ] ] and @xmath29 . \label{eq.a28bb}\ ] ] using formulae ( [ eq.a1 ] ) and ( [ eq.a2 ] ) , one writes a composition rule for two symbols @xmath11 and @xmath15determining their star - product @xmath30 the kernel in ( [ eq.a25 ] ) is determined by the trace of the product of the operators used to construct the map @xmath31 .
\label{eq.a26}\ ] ] equation ( [ eq.a26 ] ) can be extended to the case of the star - product of @xmath32 symbols of operators @xmath33 @xmath34 with kernel @xmath35 . \label{eq.a26''}\ ] ] the trace of an operator @xmath36 is determined by @xmath37 \,d\mathbf{x}_{1}\,d\mathbf{x}% _ { 2}\cdots \,d\mathbf{x}_{n}.\end{aligned}\ ] ] consider now a linear superoperator @xmath38 acting in linear space of operators .
the map of operators @xmath39 induces a corresponding map of their symbols @xmath40 the integral form of this map @xmath41 is determined by the kernel @xmath42 .
\label{ar3}\ ] ] as operator family @xmath4 , we take the fourier transform of the displacement operator @xmath43 @xmath44 where @xmath45 is a complex number , @xmath46 , and the vector @xmath47 may be interpreted as @xmath48 , with @xmath49 and @xmath50 being the position and momentum .
one sees that @xmath51 . the displacement operator ( creating coherent states from the vacuum ) may be expressed through creation and annihilation operators in the form @xmath52
@xmath53 the operator @xmath54 and its hermitian conjugate @xmath55 satisfy the boson commutation relation @xmath56=\hat{% \mathbf{1}}.$ ] the weyl symbol for an operator @xmath3 reads @xmath57 , \label{eq.a30}\ ] ] @xmath4 being given by eq.([eq.a27 ] ) .
the weyl symbols of the identity operator @xmath58 , the position operator @xmath59 and the momentum operator @xmath60 are @xmath61 the inverse transform , which expresses the operator @xmath3 through its weyl symbol , is @xmath62 that is , the operator @xmath9 in formula ( [ eq.a2 ] ) is related to @xmath4 by @xmath63 the star - product of the weyl symbols of two operators @xmath64 and @xmath65 , expressed through weyl symbols by @xmath66 with vectors @xmath67 and @xmath68 , is the operator @xmath3 with weyl symbol @xmath69\big\}. \label{eq.a36}\end{aligned}\ ] ] the @xmath0-ordered symbol @xcite @xmath70 of the operator @xmath3 is @xmath71 , \label{s1}\ ] ] with a real parameter @xmath0 , real vector @xmath47 and operator @xmath72 @xmath73 the displacement operator being @xmath74 also @xmath75 and @xmath76 the coefficient in eq .
( [ s2 ] ) leads to @xmath77 = 1 $ ] , meaning that the symbol of the identity operator equals @xmath78 .
the operator @xmath3 is obtained from @xmath79 this means that , for @xmath0-ordered symbols , the operator @xmath9 in the general formula ( [ eq.a2 ] ) takes the form @xmath80 if @xmath3 is a density operator @xmath81 @xcite , for the values of the parameters @xmath82 , the corresponding symbols are respectively the wigner , glauber sudarshan and husimi quasidistributions . for the explicit form of the kernel for the product of @xmath83 operator symbols we refer to @xcite .
density operators may be mapped onto probability distribution functions ( tomograms ) of one random variable @xmath2 and two real parameters @xmath84 and @xmath85 .
this map has been used to provide a formulation of quantum mechanics , in which quantum states are described by a parametrized family of probability distributions @xcite , alternative to the description of the states by wave functions or density operators .
the tomographic map has been used to reconstruct the quantum state , to obtain the wigner function by measuring the state tomogram , to define quantum characteristic exponents @xcite and for the simulation of nonstationary quantum systems @xcite . here
we discuss the tomographic map as an example of the general operator symbol framework .
the operator @xmath3 is mapped onto the function @xmath18 , where @xmath86 , which we denote as @xmath87 depending on the coordinate @xmath2 and the reference frame parameters @xmath84 and @xmath88 @xmath89 . \label{eq.53}\ ] ] the function @xmath87 is the symbol of the operator @xmath90 .
the operator @xmath91 is @xmath92 where @xmath59 and @xmath60 are position and momentum operators and the angle @xmath93 and parameter @xmath94 are related to the reference frame parameters by @xmath95 moreover , @xmath96 and @xmath97 is a projection density .
one has the canonical transform of quadratures @xmath98 @xmath99 using the approach of @xcite one obtains the relation @xmath100 in the case we are considering , the inverse transform determining the operator in terms of the tomogram symbol will be of the form @xmath101 where @xcite @xmath102 i.e. , @xmath103 the unitary displacement operator in ( [ eq.56q ] ) now reads @xmath104 where @xmath105 with @xmath106 and @xmath107 .
the trace of the above operator provides the kernel determining the trace of an arbitrary operator in the tomographic representation @xmath108 the operators @xmath109 and @xmath110 are creation and annihilation operators .
the function @xmath87 satisfies the relation @xmath111 , the tomographic symbol reads @xcite @xmath112 if one takes two operators @xmath64 and @xmath113 @xmath114 the tomographic symbol of the product @xmath115 is the star - product @xmath116 that is , @xmath117 with kernel given by @xmath118 .
\label{eq.59}\ ] ] the explicit form of the kernel reads @xmath119 x\right\ } \right ) , \label{kernel}\end{aligned}\ ] ] and the kernel for the star - product of @xmath83 operators is @xmath120 x\right\ } \right ) .
\label{kernelstar}\end{aligned}\ ] ]
of particular importance for quantum information purposes are finite - dimensional spin systems ( qubits , qutrits , etc . ) .
therefore , we describe here the tomographic operator symbols for spin systems .
further details may be obtained from refs .
@xcite . in this case , the physical interpretation of the symbol is as the set of measurable mean values of the operator in a state with a given spin projection in a rotated reference frame .
to set the notation , we describe here some standard operators used to discuss the properties of spin states . for arbitrary values of spin , let the observable @xmath121 be represented by a matrix in the standard basis of angular momentum generators @xmath122 , @xmath123 , @xmath124 as @xmath125 where @xmath126 the spin @xmath127 projector onto the @xmath128 component along @xmath129-axis is denoted @xmath130 and the same projector in a reference frame rotated by an element @xmath131 of @xmath132 is @xmath133 @xmath134 being a rotation operator of the @xmath135 irreducible representation with spin @xmath127 . since the projectors play an important role in constructing the tomographic map , we present several different expressions for these operators .
the projector can be given an alternative form in terms of the dirac delta - function @xmath136 and for the rotated projector @xmath137 or , in integral form @xmath138 \,d\varphi \ , . \label{a05}\ ] ] another form of the rotated projector is @xmath139 the matrix elements @xmath140 ( wigner @xmath141-functions ) are the matrix elements of the operator @xmath142 of @xmath135 group representation ( @xmath131 is an element of the @xmath135 group parametrized by euler angles ) .
the matrix elements have the explicit form @xmath143 with @xmath144 it is convenient to introduce the irreducible tensor operator for the @xmath135 group @xmath145 the irreducible tensors have the properties ( see @xcite ) @xmath146 in terms of the irreducible tensors , the operator @xmath147 is expressed as follows : @xmath148 this means that the irreducible tensors are a basis for the linear space of operators acting on the hilbert space of the @xmath135 irreducible representation .
the tomogram symbol of the observable @xmath121 is @xmath149 \nonumber \\ & = & \sum_{m_{1}^{\prime } = -j}^{j}\,\sum_{m_{2}^{\prime } = -j}^{j}\,d_{m_{1}m_{1}^{\prime } } ^{(j)}(\alpha , \beta , \gamma ) \,a_{m_{1}^{\prime } m_{2}^{\prime } } ^{(j)}\,d_{m_{1}m_{2}^{\prime } } ^{(j)*}(\alpha , \beta , \gamma ) \ , .
\label{equation05}\end{aligned}\ ] ] in view of ( [ equation05 ] ) , the tomogram depends only on two euler angles , i.e. , the tomogram depends on the spin projection and on a point on the bloch sphere .
the tomogram can be presented in another form using a kronecker delta - function , which is the general form for tomograms of arbitrary observables suggested in @xcite @xmath150 it is obvious that the tomogram of the identity operator is the unit . to derive the inverse of ( [ equation05 ] )
, we multiply by the wigner @xmath141-function @xmath151 and integrate over the volume element of the @xmath135 group , i.e. , @xmath152 where the known property of the wigner @xmath141-functions @xmath153 @xmath154 was used . in view of the symmetry relations and properties of the clebsch
gordan coefficients , we have that @xmath155 using the orthonormality property of clebsch gordan coefficients @xmath156 we have @xmath157 multiplying this equation by @xmath158 and summing over the indexes @xmath159 and @xmath160 we arrive at the result @xmath161 using eqs .
( [ eq.2 ] ) and ( [ equation16 ] ) we can write the observable operator @xmath121 in terms of unitary irreducible tensors as follows : @xmath162 substituting @xmath163 into ( [ equation24 ] ) , in view of the orthonormality of the clebsch gordan coefficients , we obtain the observable in terms of its tomogram : @xmath164 the density operator @xmath81 can be expanded in terms of irreducible tensors ( [ tlm ] ) as follows : @xmath165 one can express the operators determining the star - product of tomographic symbols in terms of irreducible tensors . by comparing the formulas defining the generic symbol of operators ( [ eq.a1 ] ) and its inverse ( [ eq.a2 ] ) with the formulae defining the observable tomogram ( [ new1 ] ) and its inverse ( [ equation25 ] )
, one can find the operators @xmath4 and @xmath9 explicitly .
the operators @xmath166 and @xmath167 can be expressed as follows : @xmath168 using formulae ( [ neweq37 ] ) and ( [ neweq38 ] ) , one can write down a composition rule for two symbols @xmath11 and @xmath169 determining the star - product of these symbols . the composition rule is @xmath170 the kernel in the integral of ( [ eq.25 ] ) is the trace of the product of the operators used to construct the map @xmath31 .
\label{eq.26}\ ] ] within this framework , according to ( [ a01 ] ) , ( [ a02 ] ) and ( [ a04 ] ) , one has two equivalent expressions for the operator @xmath4 @xmath171 or , due to the structure of this equation , @xmath172 the dual operator reads @xmath173 where @xmath174 is given in eq .
( [ tlm ] ) . inserting the expressions for the operators @xmath4 and @xmath175 in ( [ eq.26 ] ) and using the properties of irreducible tensors ( [ a ] ) and ( [ b ] )
, one obtains an explicit form for the kernel of the spin star - product @xmath176 one can extend the construction by introducing a unitary spin tomogram @xcite of the multiqudit state with density matrix @xmath177 . for this , one uses the joint probability distribution @xmath178 where @xmath179 is a unitary operator in the hilbert space of multiqudit states . for a simple qudit state , the tomogram unitary symbol is @xmath180 where @xmath181 is a @xmath182 matrix . since it is possible to reconstruct the density matrix using only spin tomograms , the unitary spin tomogram also determines the density matrix completely .
one can integrate in eq .
( [ equation29 ] ) the unitary spin tomogram @xmath183 using the haar measure @xmath184 instead of @xmath185 and adding the delta - function term @xmath186 .
this construction means that the spin quantum state is defined by a map of the unitary group to the simplex .
the following are the properties of the unitary spin tomograms of multiqudit systems : ( i)normalization @xmath187 ( ii)group normalization from the haar measure on the unitary group @xmath188 divided by the group volume @xmath189 , one obtains the measure @xmath190 with @xmath191 .
then , @xmath192 this property follows from the orthogonality condition for matrix elements of unitary matrices as elements of an irreducible representation of a compact group .
another property is @xmath193 where the tomogram @xmath194 is a tomogram for the subsystem density matrix @xmath195 .
an analogous unitary group integration property follows from the relation @xmath196 yielding @xmath197 that corresponds to @xmath198
the unitary spin symbol ( [ a86 ] ) defines , for each density matrix @xmath199 , a mapping from the unitary group @xmath200 , @xmath201 , to a @xmath202 dimensional simplex .
the nature of the image of @xmath200 on the simplex depends on the nature of the density matrix .
the unitary spin symbol image of @xmath200 on the simplex for most density matrices @xmath199 ( @xmath203s with at least two different eigenvalues ) has dimension @xmath202 . for pure states , it is the whole simplex and for mixed states a volume bounded by the hyperplanes @xmath204 where @xmath205 are the eigenvalues of the density matrix .
_ proof _ : @xmath206 such that @xmath207 is diagonal .
then by another @xmath208 @xmath209 if @xmath199 is a pure state , only one @xmath210 .
then @xmath211 that is , all points in the simplex are obtained .
therefore , for a pure state , the unitary tomographic symbol maps the unitary group on the whole simplex . to obtain the dimensionality of the image for a general ( mixed ) state , we consider the elementary @xmath200 transformations : @xmath212 consider these elementary transformations acting on the diagonalized matrix @xmath213 .
@xmath214 does not change the diagonal elements and both @xmath215 and @xmath216 have a similar action : @xmath217 a general infinitesimal transformation would be @xmath218 and the dimension of the simplex image of @xmath200 is the rank of the jacobian @xmath219 . if @xmath199 has at least two different eigenvalues , the rank is @xmath202 , this being the dimension of the simplex image .
the hyperplanes ( [ m1 ] ) bounding this simplex volume follow from the convex nature of the eigenvalues linear combination ( [ m2 ] ) .
the situation where all eigenvalues are equal is exceptional , the image being a point in this case .
@xmath220 figure 1 shows an example for a mixed state of a two - qubit state , when @xmath221 . for a bipartite system of dimension @xmath222
, the distinction between factorized and entangled states refers to the behavior under transformations of the factorized group @xmath223 .
we call a state _ factorized _ ,
if the density matrix is @xmath224 and _ classically correlated _ , if @xmath225 with @xmath226 .
the simplex symbol image under @xmath227 of a generic factorized or classically correlated state has dimension @xmath228 .
_ proof _ : for a classically correlated state , if @xmath229 it is not , in general , possible to find an element of @xmath230 diagonalizing @xmath199 .
therefore , one has to consider the action of the elementary unitary transformations ( [ m3 ] ) on a general matrix .
@xmath231 does not change the diagonal elements whereas the @xmath232 action for @xmath233 is @xmath234 and for @xmath235 is @xmath236 for generic @xmath199 matrices , @xmath237 and @xmath233 operate independently , therefore , infinitesimal transformations explore @xmath228 independent directions . @xmath220
the generalization to classically correlated multipartite systems is immediate , implying that the image dimension under @xmath238 is @xmath239 . as an example , we compute explicitly the equation for the two - dimensional surface image in the two - qubit case for a factorized state . in this case , one has to consider mappings from @xmath240 to the simplex .
let @xmath241 . here
, without loosing generality , @xmath242 may be considered as diagonal .
then @xmath243 is @xmath244 hence @xmath245 is @xmath246 implying @xmath247 figure 2 shows this two - dimensional surface in the 3-dimensional simplex . for a pure state
, this would be the image of the @xmath248 group . for a mixed state ,
the image is the intersection of the surface with the spanned volume , as in fig . 1
. theorem 4.2 suggests a notion of _ geometric correlation _ ,
namely , a state of a multipartite system is called geometrically correlated if the symbol image under @xmath249 has dimension less than @xmath239 . deviations from geometrical genericity occur when the systems are entangled or the density matrix has special symmetry properties . as an example , consider the entangled state @xmath250 a simple computation shows that the image under @xmath240 is defined by @xmath251 implying that the image is one - dimensional ( figure 3 ) however , the dimension reduction of the image of @xmath252 does not coincide with the notion of entanglement . as an example , consider the werner state @xmath253 @xmath254 , which is known to be entangled only for @xmath255 . in this case , because of the highly symmetric nature of the state , the orbit of the tomgraphic symbol is the same both for @xmath256 and @xmath257 , namely , @xmath258 implying that the image is always one - dimensional
. incidentally , peres separability criterium @xcite applied to the partial transpose @xmath259 expressed in tomographic operator symbols would be @xmath260 the state being entangled when there is an @xmath179 for which this identity is violated .
in the standard quantum formulation , measurements are realized by von neumann `` instruments '' which are orthogonal projectors @xmath261 onto eigenstates of the variables being measured .
the projectors applied to the pure state @xmath262 yield @xmath263 or in terms of density matrix @xmath264 , one has the result @xmath265 for a mixed state @xmath81 , the measurement provides the state density operator after measurement @xmath266 generalized measurements use positive operator - valued measures ( povm ) , that is , positive operators @xmath267 with the property @xmath268 the index @xmath269 being either discrete or continuous . in the latter case
, one has an integration in ( [ ai5a ] ) . within the framework of operator symbols and star - products , instead of ( [ ai4a ] )
, we have after the measurement a symbol for the density operator of the state @xmath270 @xmath271 being the symbol of the density operator of measurable state and @xmath272 the symbol of the instrument . in the tomographic probability representation ,
the result of measurements is described by a map of the probability distributions , namely , @xmath273 with the kernel of the star - product of tomograms given by eq . ( [ kernel ] ) . for the case of spin ( or unitary spin ) tomograms , the linear map of tomographic - probability distributions is realized by the formula @xmath274 the kernel of the star - product being given by eq .
( [ eq.26a ] ) .
one sees that , in the probability representation , the process of measurement , both with von neumann instruments and with povm , is described by a map of points in the simplex .
in the standard representation of quantum mechanics , states ( state vectors @xmath275 or density operators @xmath276 ) of closed system evolve according to unitary change @xmath277 or @xmath278 this evolution is a solution to schrdinger or von neumann equations @xmath279=0 , \label{v3}\ ] ] @xmath280 being the hamiltonian of the system .
the evolution can be cast into operator symbol form .
let @xmath281 be the symbol of an operator @xmath3 .
we do not specify at the moment what kind of symbols are used , considering them as generic ones with quantizer
dequantizer pair @xmath282 , @xmath283 .
then , the operator equation ( [ v3 ] ) for density operator reads @xmath284 we denote the symbol of the density operator @xmath276 by @xmath285 .
the solution of eq .
( [ v4 ] ) has a form corresponding to ( [ v2 ] ) @xmath286 one can rewrite the solution ( [ v2 ] ) as a superoperator @xmath38 acting in a linear space of operators , namely , @xmath287 in matrix form , eq .
( [ v6 ] ) reads @xmath288 for unitary evolution , the superoperator is expressed in terms of unitary matrix @xmath289 as a tensor product @xmath290 one rewrites the solution ( [ v5 ] ) for the symbol introducing the propagator @xmath291 for the unitary evolution ( [ v2 ] ) , the propagator reads @xmath292 where @xmath293 the kernels under the integral are given by eq .
( [ eq.26 ] ) . in the case of superoperators describing the evolution of an open system @xcite @xmath294 the propagator reads @xmath295 the propagator corresponds to a superoperator , which in matrix form reads @xmath296 for the case of continuous variables and symplectic tomograms , eq .
( [ v4 ] ) takes the form of a deformed boltzman equation for the probability distribution . for unitary spin tomograms
, one has @xmath297 the unitary evolution matrix being determined by an hamiltonian matrix @xmath298 this means that the unitary spin tomogram , a function on the unitary group , evolves according to the regular representation of the unitary group .
this means that the partial differential equation for the infinitesimal action is the standard equation for matrix elements of the regular representation , that is , @xmath299 @xmath300 being the hamiltonian hermitian matrix and @xmath301 the infinitesimal hermitian first - order differential operators of the left regular representation of the unitary group in the chosen group parametrization .
in this section , we consider the unitary spin representation of some typical quantum channels .
consider bit flip , phase flip and both with equal probability @xmath302 the kraus representation is @xmath303 for the unitary spin symbol representation , choosing the axis one has @xmath304 and with an arbitrary unitary group element @xmath305 one obtains @xmath306 when @xmath307 the image in the simplex contracts to a segment between @xmath308 and @xmath309 .
@xmath310 the kraus representation is @xmath311 with @xmath312 for the unitary spin symbol representation , consider the example @xmath313 @xmath314 then @xmath315 and when @xmath307 the image in the simplex contracts to a point .
@xmath316 the kraus representation is @xmath311 with @xmath317 for the unitary spin representation , consider an excited initial state @xmath318 @xmath319 then @xmath320 when @xmath50 varies from @xmath321 to @xmath78 , the image in the simplex first contracts to a point @xmath322 and then expands again to the whole simplex when @xmath307 .
the operator symbols being functions on the rotation or unitary groups are highly redudant descriptions of qudit states . as expected from the number of independent parameters in the density matrix ,
also here @xmath323 numbers are enough to characterize a @xmath324dimensional qudit .
this is easy to check .
consider the operator symbol ( [ a86 ] ) for an arbitrary @xmath324dimensional density matrix @xmath199 .
a general @xmath199 may be diagonalized by of @xmath325 independent unitary transformations and this , together with the @xmath326 independent diagonal elements , gives the desired result .
alternatively we may consider @xmath327 independent elements of the unitary group and compute the associated operator symbols .
then , the qudit state would be described by their diagonal elements .
therefore , a discrete quantum state ( qudit ) is coded by @xmath327 probability distributions . for each @xmath179 in the group ,
the elements in the operator symbol @xmath328 are the probabilities to obtain the values @xmath329 in a measurement of the quantum state @xmath199 by an apparatus oriented along @xmath179 .
therefore the problem of reconstructing the state @xmath199 from the set of @xmath330 operator symbol elements is identical to the reconstruction of the density matrix of a spin through stern - gerlach experiments , already discussed in the literature@xcite @xcite @xcite .
the tomographic operator symbols satisfy @xmath331 therefore , they are probability distributions @xmath332 . one defines the _ operator symbol entropy _ by @xmath333 and the _ operator symbol rnyi entropies _ by @xmath334 likewise , we may define the _ operator symbol relative q - entropy _ by @xmath335 with @xmath336 because the operator symbols @xmath337 are probability distributions , they inherit all the known properties of nonnegativity , additivity , joint convexity , etc . of classical information theory .
the relation of the operator symbol entropies to the von neumann and the quantum rnyi entropies is given by the following the von neumann @xmath338 and the quantum rnyi @xmath339 entropies are the minimum on the unitary group of @xmath340 and @xmath341 .
_ proof _ : from @xmath342 there is a @xmath343 such that @xmath344 is a diagonal matrix @xmath345 .
then @xmath346 for any other @xmath179 , the diagonal elements in ( [ tt1 ] ) are convex linear combinations of @xmath347 . by convexity of @xmath348
the result follows for the von neumann entropy .
the operator symbol rnyi entropy is not a sum of concave functions .
however , the following function is , @xmath349 called the tsallis entropy@xcite and related to the rnyi@xcite entropy by @xmath350 the minimum result now applies to @xmath351 by concavity and then one checks from ( [ tt2 ] ) that it also holds for @xmath341 .
therefore @xmath352 coincides with the quantum rnyi entropy @xmath353 @xmath220 the entropy @xmath340 varies from the minimum , which is von neumann entropy , to a maximum for the most random distribution . for each given state @xmath199 , one can also define the _ integral entropies _
@xmath354 where @xmath355 is the invariant haar measure on unitary group . in some cases
, the properties of the von neumann entropy may be derived as simple consequences of the classical - like properties of the operator symbol entropies .
for example : * subadditivity : * @xmath356 consider a two - partite system with density matrix @xmath357@xmath358 @xmath179 being a @xmath359 unitary matrix @xmath360 from the reduced symbols and density matrices @xmath361 @xmath362 one writes the reduced symbol entropies @xmath363 @xmath364 for each fixed @xmath179 , the tomographic symbols are ordinary probability distributions .
therefore , by the subadditivity of classical entropy , @xmath365 in particular , this is true for the group element in @xmath366 that diagonalizes the reduced density matrices @xmath367 and @xmath368 .
therefore , @xmath369 but , by the minimum property , @xmath370 thus subadditivity for the von neumann entropy is a consequence of subadditivity for the operator symbol entropies .
the situation concerning strong subadditivity is different .
strong subadditivity also holds , of course , for the operator symbol entropies for any @xmath179@xmath371 but the corresponding relation for the von neumann entropy is not a direct consequence of ( [ tt3 ] ) .
the strong subadditivity@xcite for the von neumann entropy @xmath372 expressed in operator symbol entropies would be @xmath373 where @xmath374 are the different group elements that diagonalize the respective subspaces .
therefore strong subadditivity for the von neumann entropy ( [ tt4 ] ) and strong subadditivity for the operator symbol entropies ( [ tt3 ] ) are independent properties . on the other hand , because of the invertible relation ( [ equation29 ] ) between the operator symbols and the density matrix , eq.([tt3 ] ) contains in fact a family of new inequalities for functionals of the density matrix .
to conclude we summarize the main results of this work : \(iii ) evolution equations for qudit operator symbols are written in the form of first - order partial differential equations with generators describing the left regular representation of the unitary group .
\(vi ) in view of the probability nature of the operator symbols , the corresponding entropies inherit the properties of classical information theory .
some of the properties of the von neumann entropy and quantum rnyi entropy are direct consequences of these properties . on the other hand
the properties of the operator symbol entropies also imply new relations for functionals of the density matrix . | hilbert space operators may be mapped onto a space of ordinary functions ( operator symbols ) equipped with an associative ( but noncommutative ) star - product . a unified framework for such maps
is reviewed .
because of its clear probabilistic interpretation , a particular class of operator symbols ( tomograms ) is proposed as a framework for quantum information problems .
qudit states are identified with maps of the unitary group into the simplex .
the image of the unitary group on the simplex provides a geometrical characterization of the nature of the quantum states .
generalized measurements , typical quantum channels , entropies and entropy inequalities are discussed in this setting . |
in this work we consider the problem of a spin - half fermion gravitationally bound by a static , massive body described by a non - singular metric .
we investigate the energy level spectrum of the particle in the limit that the metric exhibits a singularity .
previously , we calculated the spectrum of scalar ( spin - zero ) particles bound by such objects @xcite and found that it collapses to zero energy and becomes quasi - continuous in the singular limit .
importantly , bound states with zero energy ( where the binding energy is equal in magnitude to the rest mass of the bound particle ) were found not to exist when the metric is not singular . here
we calculate the spin-1/2 fermion energy level spectrum for three interior metrics describing the interior of the gravitating ( stationary ) body : the florides , schwarzschild , and soffel metrics , the latter so - named for its derivation and subsequent use in @xcite . for each metric
we calculate the energy spectrum of the particle numerically and using approximate analytic expressions in the near - singular limit of the metric . in the florides and
soffel spacetimes this limit is characterized by the approach of the schwarzschild radius @xmath0 towards the radius of the body @xmath1 , where @xmath2 is the gravitational constant and @xmath3 is the mass of the central body ( black hole limit ) . in the case of the schwarzschild interior metric
the singularity occurs for @xmath4 and corresponds to a infinite pressure at the origin , rather than the formation of an event horizon .
our results show that , as in the scalar case , in the limit that the interior metric becomes singular , the ground state energy tends to zero ; that is , the binding energy tends to the rest mass energy of the bound particle .
additionally , the energy spectrum becomes quasi - continuous in this limit .
this means that one can not increase the strength of the central potential ( by increasing @xmath3 ) to lower a bound state level from the positive continuum to the negative : one can not reach the dirac sea to facilitate pair production .
this suggests that pair - production can not occur in the field of such a static metric prior to the formation of a black hole .
additionally , we present a discussion of the semi - classical approach to analyzing the properties of the level spectrum .
it is shown that as the metric develops a singularity , all the semi - classical levels will tend to zero .
this behaviour can also be understood from classical quantities such as time dilation on the interior .
one may write a general static , spherically symmetric metric in the form ( units @xmath5 ) @xmath6 the wavefunction for a spin - half particle ( dirac spinor ) in a spherically symmetricl field may be separated using @xmath7 with @xmath8 here @xmath9 parametrises the angular momentum for a particle with well - defined values of orbital and total angular momentum , @xmath10 .
the coupled radial wave equations are then ( see , e.g. @xcite and appendix a ) : @xmath11 g({r } ) = 0 , \\
\label{coupledradialeqsb } \frac{dg({r})}{dr } -e^{\frac{\lambda}{2 } } \frac{\kappa}{r } g({r } ) + e^{\frac{\lambda}{2 } } \left [ e^{-\frac{\nu}{2}}\epsilon - m \right ] f({r } ) = 0 \,.\end{gathered}\ ] ] these equations can be decoupled to give a single second - order equation for the upper component @xmath12 @xmath13f(r)=0.\end{aligned}\ ] ] in order to determine the bound states of a spin-1/2 particle in the gravitational field of a finite - sized object we require a wave function that exists for all @xmath14 .
this necessitates an interior as well as an exterior metric to substitute into ( [ eq : secondordereqn ] ) and ( [ sepvarsubs ] ) .
the metric on the exterior of a spherically symmetric non - rotating body of mass @xmath3 is @xmath15 where @xmath16 is the schwarzschild radius of the gravitating object .
this , along with a suitable interior metric , allows us to construct interior and exterior wave equations of the form ( [ eq : secondordereqn ] ) which together span all @xmath14 .
these are solved both numerically and using approximate analytical formulae .
the bound states are found by numerically integrating interior and exterior versions of ( [ eq : secondordereqn ] ) using _ mathematica _ @xcite with the boundary condition for small @xmath14 determined by solving the zero order expansion of the interior wave equation around @xmath17 .
the interior wave function is propagated to the boundary where continuity of the functions @xmath18 and @xmath19 is imposed .
the solution is then propagated to large @xmath14 and the boundary condition @xmath20 is enforced by varying @xmath21 .
the s - wave bound states for a given @xmath1 and @xmath22 are computed as a function of @xmath23 .
these results are compared with analytic calculations in later sections . in the strong field limit ( as the metric becomes singular ) the wave function of a bound particle is contained entirely on the interior region of the body .
the assumption of near complete containment on the interior allows one to perform all calculations using only the interior wave equation .
the validity of this assumption is confirmed in subsequent sections . on the interior
it is convenient to switch to dimensionless variables .
we rescale the energy by the mass such that @xmath24 , and rescale all lengths , including the compton radius of the quantum particle , by the radius of the object @xmath1 : @xmath25 , @xmath26 , @xmath27 . one can re - write ( [ eq : secondordereqn ] ) in the form of a one dimensional schrdinger equation using the transformation @xmath28 where @xmath29 yielding @xmath30 where @xmath31 is an effective ( energy dependent ) potential
. analytic calculations of the bound state spectrum in the tight - binding limit can be performed by approximating the effective potential on the interior as that of a harmonic oscillator . for our interior metrics , @xmath32 , one
may taylor expand @xmath31 near the origin and match the resulting expression to that of a quantum harmonic oscillator , given by @xmath33 and use known solutions @xmath34 , where @xmath35 , @xmath36 , to approximate the energy levels of the system .
here @xmath37 is the effective energy of the harmonic oscillator system . in the florides case
this approach is valid all the way to the singular metric limit .
as will be seen in later sections , this is not true for the soffel and schwarzschild metrics where the harmonic approximation breaks down as the metric develops a singularity .
as the metric develops a singularity , the potential well becomes increasingly deep causing the wavefunction to oscillate many times on the interior . in this case one
may invoke the semi - classical approximation and calculate excited energy levels using bohr - sommerfield quantization .
when treating the problem semi - classically , all three singular metric limits may be taken .
to do this we investigate eqn .
( [ eq : geneqn ] ) for a metric possessing a singularity such that @xmath38 for some value of @xmath14 . in this limit eqn .
( [ eq : geneqn ] ) reduces to @xmath39 = 0\ ] ] allowing us to define the approximate semi - classical momentum @xmath40^{1/2}.\ ] ] assuming that the semi - classical approximation is valid , the total phase accumulated at the boundary will contain the integral @xmath41 d\rho \notag { } \\ & = x{\varepsilon } \mu-\frac{y \mu}{\eps}\end{aligned}\ ] ] setting this phase to @xmath42 yields the energy levels as @xmath43 this expression tends to zero in the singular limit if @xmath44 , which is satisfied in this limit as @xmath45 contains the most singular term .
thus all levels tend to zero and the spectrum collapses .
additionally , one may arrive at this conclusion by directly considering the metric . in the singular limit
, the coefficient of @xmath46 in the metric tends to zero giving rise to ( local ) time dilation that approaches infinity , with the classical period @xmath47 of a bound particle being similarly affected .
therefore in this limit the semi - classical level spacing becomes @xmath48 and we arrive at the previous result .
an interior metric that does not contain the singularity present in the schwarzschild interior solution , discussed in section ( [ sec : schwarzschild ] ) , is that developed by florides @xcite : @xmath49 which is valid for @xmath50 . for @xmath51 the horizon ( coordinate ) singularity exists at @xmath52 @xcite .
the florides metric corresponds to a spherically symmetric , constant density body with vanishing radial stresses . for our purposes this metric
is characterized by its effective potential @xmath31 which , near @xmath53 ( for @xmath54 ) , is given by @xmath55 equating the coefficients of the powers of @xmath56 between eqns .
( [ eq : qhov ] ) and ( [ eq : florideshoexp ] ) , we derive two equations for the two unknowns @xmath57 and @xmath58 and can therefore solve for both .
the energy solutions that corresponds to the energy eigenvalues are @xmath59 where @xmath60
the above expressions are valid under the assumption that the entire wavefunction is contained within the interior region @xmath61 ( @xmath62 )
. this condition can be satisfied when @xmath63 is large , forcing the particle to be bound near the origin .
formally we require @xmath64 .
the value of @xmath65 can be obtained using the virial theorem for our harmonic potential @xmath66 expanding @xmath57 about @xmath67 in ( [ eq : bindingcondition ] ) using ( [ eq : florepsilon ] ) to lowest order gives @xmath68 where our truncation of the expansion is valid for @xmath69 recalling that @xmath70 .
using this value of @xmath57 the condition ( [ eq : bindingcondition ] ) becomes @xmath71 which automatically satisfies eqn .
( [ eq : fc1 ] ) .
a second test is whether the expansion of @xmath72 as a harmonic potential is valid .
the ratio of the fourth order to second order term ( at the expected particle position ) is , for large @xmath63 , given by @xmath73 where @xmath74 is the coefficient of the @xmath75 term in the expansion ( [ eq : florideshoexp ] ) .
where the harmonic oscillator model is appropriate this should be much less than one , which leads to @xmath76 this condition is automatically satisfied by eqn .
( [ eq : interiorcondition ] ) .
furthermore , @xmath77 is satisfied automatically ( when ( [ eq : interiorcondition ] ) is satisfied ) as it must to be physically valid .
finally , we note that in the limit @xmath78 , we can expand the energy ( [ eq : florepsilon ] ) as @xmath79 this is always @xmath80 under condition ( [ eq : interiorcondition ] ) ( that is , the state is bound ) . figure ( [ fig : florideseplot ] ) shows the numeric ground state energy as a function of @xmath81 compared with eqn .
( [ eq : florepsilon ] ) . figure ( [ fig : floridesexcitedplot ] ) compares the analytics and numerics for the ground and first three excited states . ,
@xmath82 and @xmath83 .
circles : numeric ; solid line : analytic approximation ( [ eq : florepsilon]).,width=377 ] , and @xmath83 .
circles : numeric ; solid line : analytic approximation ( [ eq : florepsilon]).,width=377 ] the figures show that there is good agreement between our numerical and analytic spectrum . a large mass of @xmath84 was chosen for fig .
( [ fig : floridesexcitedplot ] ) in order to satisfy the condition ( [ eq : interiorcondition ] ) , which is better fulfilled as @xmath63 increases , ensuring the validity of the harmonic approximation for all @xmath85 displayed .
as an alternative to the approach adopted by florides , soffel _ et al .
_ extended the interior schwarzschild metric past the pressure singularity by analytic continuation @xcite .
we perform our analysis of the bound state spectrum using the same procedures and techniques employed in the previous two sections .
this ` soffel ' metric corresponds to @xmath86 , \\
e^\lambda & = \left(1-s \rho^2\right)^{-1}\notag{}.\notag { } \\\end{aligned}\ ] ] using the methods outlined previously , we find in the soffel metric the scaled energy , @xmath58 , is @xmath87}{\mu}\sqrt{\tau + \chi},\ ] ] where @xmath88 - 26\right ) . \ ] ] as in the previous case we examine conditions on @xmath63 ( taken to be large ) such that @xmath89 and the quadratic term in the expansion of the potential dominates ( using eqn.([eq : mainsoffelenergy ] ) as @xmath58 ) .
this yields the condition @xmath90 which is automatically satisified for @xmath54 ( a condition on @xmath63 w.r.t . @xmath91 and
@xmath92 can be generated in a similar fashion to eqn .
( [ eq : interiorcondition ] ) of the florides case , but is automatically satisfied by ( [ eq : soffelcond1 ] ) ) .
this is due to the presence of extra @xmath93 terms in the expansion of the effective potential corresponding to the soffel metric .
this also affects the ratio of the fourth order to second order coefficients .
we find that , unlike in the florides case , the effective potential changes in shape significantly near the origin as @xmath54 .
thus , in order to ensure that the potential where the particle is bound remains harmonic we require a more stringent condition on @xmath63 . specifically , we find that @xmath94 for comparison with results presented in fig .
( [ fig : soffelexcitedplot ] ) for @xmath95 this corresponds to @xmath96 .
as @xmath54 the energy expression ( [ eq : mainsoffelenergy ] ) becomes @xmath97\sqrt{1-s}\left(1+\frac{(n+3/2)\sqrt{s}}{\sqrt{2}\mu\sqrt{1-s}}\right)\ ] ] which tends to zero in the black hole limit for both positive and negative energy states ( eqn .
( [ eq : soffelmasscondition ] ) ensures the second term is @xmath98 in this limit ) .
figures ( [ fig : soffeleplot ] ) and ( [ fig : soffelexcitedplot ] ) show the ground state and excited state spectra respectively , with the numeric and analytic results displayed with each other as before . ,
@xmath82 and @xmath83 .
circles : numeric ; solid line : analytic approximation ( [ eq : mainsoffelenergy]).,width=377 ] , and @xmath83 .
circles : numeric ; solid line : analytic approximation ( [ eq : mainsoffelenergy]).,width=377 ] in the singular limit , @xmath99 , the spectrum ( [ eq : mainsoffelenergy ] ) is observed to exhibit the same limit as the florides spectrum , ( [ eq : florepsilon ] ) .
the positive energy spectrum collapses to zero from above as the gravitating body tends towards the black hole limit , and the negative energy levels ( dirac sea ) also collapse to zero .
( [ eq : soffelmasscondition ] ) indicates , and as was noted in sec .
( [ sec : harmonicsection ] ) , the harmonic approximation breaks down in the singular limit for this metric , i.e. only in the limit of infinite @xmath63 is the harmonic approximation valid . using the definitions in eqn .
( [ eq : bohrsommerfield ] ) we calculate @xmath100 , \notag { } \\ y & \stackrel{s\rightarrow1}{=}0\end{aligned}\ ] ] yielding @xmath101 for all levels as @xmath54 .
as previously noted , the standard schwarzschild interior metric can not be used to investigate the limit @xmath54 .
this is due to a pressure singularity forming at the origin for @xmath102 where the coefficient of @xmath46 is zero .
the interior solution for a constant density fluid sphere developed by schwarzschild @xcite is given by @xmath103 taking the singular limit ( @xmath104 ) of the effective potential yields @xmath105 by equating powers of @xmath56 between ( [ eq : qhov ] ) and ( [ mainschwarzexpansion ] ) one can solve for @xmath58 , giving the energies as @xmath106 where @xmath107 we can now check the validity of the approximations used . once again invoking the virial theorem and using eqn .
( [ eq : bindingcondition ] ) , the condition @xmath108 is satisfied for @xmath109 as with eqn .
( [ eq : interiorcondition ] ) in the florides case , a condition on @xmath63 w.r.t . @xmath91 and @xmath92 can be generated , but is automatically satisfied by ( [ eq : scond2 ] ) . as done
previously we require the quadratic terms in the expansion to dominate over higher order terms . in this case
the schwarzschild metric resembles the soffel metric considered previously , as it too has an additional condition on @xmath63 compared to the florides metric .
specifically , we find : @xmath110 by taking into account condition ( [ eq : scond2 ] ) we simplify the energy expression ( [ eq : mainschwarzenergy ] ) to @xmath111 which retains the behaviour @xmath112 for both positive and negative energy states
. figures ( [ fig : schwarzeplot ] ) and ( [ fig : schwarzexcitedplot ] ) show the ground state and excited state spectra respectively , with the numeric and analytic results displayed with each other as before . ,
@xmath82 and @xmath83 .
the metric becomes singular as @xmath81 approaches @xmath113 causing the binding energy to tend towards the rest mass of the fermion .
circles : numeric ; solid line : analytic approximation ( [ eq : mainschwarzenergy]).,width=377 ] , and @xmath83 .
circles : numeric ; solid line : analytic approximation ( [ eq : mainschwarzenergy]).,width=377 ] as in the florides case , our analytic formula reproduces the bound state energies well . in the singular limit
, the spectrum is observed to exhibit the same qualitative behaviour as that in the florides case .
the positive energy spectrum collapses to zero from above , but with one important difference : the gravitating body develops a pressure singularity , causing the spectrum to collapse , well before the back hole limit . as in the soffel case ,
the harmonic approximation breaks down when attempting to model the schwarzschild case in the @xmath114 limit ; one must take the infinite @xmath63 limit to ensure applicability as per eqn .
( [ eq : scond2 ] ) . the corresponding calculation in the semi - classical picture
is done using @xmath115 substituting this into eqn .
( [ eq : genen ] ) yields @xmath101 for all levels as @xmath114 .
we present calculations of the energy level spectrum for a massive spin-1/2 particle bound by a spherical static body .
this is done by matching a suitable interior metric to the standard schwarzschild exterior metric and solving the dirac equation .
numerical computation of the ground state energies of the particle shows that the existence of a zero energy bound state may only be possible in the limit that the interior metric develops a singularity . in the case of the florides and soffel metrics
this corresponds to the black hole limit ; for the schwarzschild interior metric this singularity represents infinite pressure at the centre .
this result is verified using an approximate analytical calculation . for the florides and soffel metrics
it is seen that the entire bound state spectrum collapses to @xmath116 as @xmath117 : the energies and the intervals between the energy levels are proportional to @xmath118 and @xmath119 $ ] respectively .
if we keep energy fixed the principal quantum number of the level @xmath85 tends to infinity . in the case of the schwarzschild interior metric
the singularity occurs for @xmath120 with the spectrum going as @xmath121 .
what is clear is that for any fixed @xmath22 , @xmath85 , @xmath91 , the general form of the equations for @xmath122 in the florides and soffel cases ensures that the bound state energy collapses to zero in the limit @xmath123 .
additionally , the form of the energy spectrum in the schwarzschild interior case results in the same collapse of the spectrum , except now the collapse occurs in the limit @xmath124 .
this means that for spin-1/2 particles , as in the case of scalar particles , the event horizon singularity is not the only singularity that results in collapse of the spectrum to a gapless state a pressure singularity gives the same result .
therefore the bound state energy spectrum of an object with a near - singular metric reproduce those of an object that is close to the black hole limit .
the existence of a bound state with zero energy is relevant to the phenomenon of particle pair production . due to quantum fluctuations ,
particle anti - particle pairs are produced around a black hole and it is possible that one of the pair escapes the gravitational field yielding hawking radiation @xcite .
other systems that give rise to pair production , also known as vacuum breakdown , include static and dynamic electromagnetic fields ( see , e.g. @xcite ) and time - varying gravitational fields ( see , e.g. @xcite ) . in the coulomb case
it is clear that one of the pair must be repelled , since the pair have opposite charge . in the gravitational case , however , both particles will possess the same ` gravitational charge ' and thus will be attracted by the potential . in the black hole case
the existence of the event horizon as a barrier facilitates one particle escaping to infinity as the other falls to the singularity . for gapless energy states to exist in the metrics we have considered
, the metric must become singular . for pair production with the ejection of one particle to infinity
it is actually necessary to have a negative energy level @xmath125 . as a concrete example , compare the gravitational field case with the electrostatic potential case where pair creation is possible in a strong coulomb field @xmath126 .
( for example , the ground state of an electron orbiting a finite - size nucleus reaches the lower continuum when @xmath127 @xcite ) . in the coulomb case
we have @xmath128 in the wave equation .
therefore , increasing the absolute value @xmath129 of the negative potential @xmath126 leads to the negative energy @xmath21 . in the gravitational field
we have the field dependent term @xmath130 appearing as a multiplicative factor of the energy , rather than as a subtractive function , a situation that can not introduce negative energy bound states .
+ we thank m. yu .
kuchiev and g. f. gribakin for useful discussions .
this work is supported by the australian research council .
g. h. gossel , j. c. berengut and v. v. flambaum , gen .
gravit . , * 43 * , 2673 ( 2011 ) .
m. soffel , b. muller and w. greiner , j. phys .
gen . , * 10 * , 551 , ( 1977 ) .
v. b. berestetskii , l. p. pitaevskii and e. m. lifshitz , _ quantum electrodynamics_. ( butterworth - heinemann , 1982 ) . p. s. florides , _ a new interior schwarzschild solution _ , proc .
lond . a * 337 * , 529 ( 1974 ) . to avoid misunderstanding , we should note that this metric does not correspond to any macroscopic object for @xmath131 .
this is further discussed by n. k. kofinti , _ on a new interior schwarzschild solution _ ,
* 17 * , 245 ( 1985 ) .
h. a. buchdahl , _ general relativistic fluid spheres _ , phys . rev . * 116 * , 1027 ( 1959 ) .
( wolfram research , inc . , champaign , il , 2008 ) .
v. f. mukhanov , s. winitzki , _ introduction to quantum effects in gravity _ , ( cambridge university press , cambridge , 2007 ) .
w. yongjiu , t. zhiming , _ energy levels of electron near kerr black hole _ ,
space sci . *
281 * , 689 ( 2001 ) .
w. g. unruh , _ absorption cross section of small black holes _
d * 14 * , 3251 ( 1976 ) .
c. doran , a. lasenby , _ perturbation theory calculation of the black hole elastic scattering cross section _ ,
d * 66 * , 024006 ( 2002 ) .
m. yu . kuchiev and v. v. flambaum , _ scattering of scalar particles by a black hole _ ,
d * 70 * , 044022 ( 2004 ) .
a. b. gaina , i. m. ternov , _ energy spectrum of the klein - gordon equation in schwarzschild and kerr fields _ , izvestiya vysshikh uchebnykh zavedenii , fizika * 10 * , 71 ( 1988 ) .
j. grain and a. barrau , _ quantum bound states around black holes _ ,
j. c * 53 * , 641 ( 2008 ) .
a. lasenby , c. doran , j. pritchard , a. caceres , s. dolan , _ bound states and decay times of fermions in a schwarzschild black hole background _
d * 72 * , 105014 ( 2005 ) .
m. s. pindzola , _
inelastic scattering of electrons by black hole atoms _ , j. phys .
b : at . mol .
42 * , 095202 ( 2009 ) .
k. schwarzschild , _ ber das gravitationsfeld einer kugel aus inkompressibler flssigkeit _ ,
. wiss . * 3 * , 189 ( 1916 ) .
s. w. hawking , _ black hole explosions " _
, nature * 248 * , 30 ( 1974 ) .
s. w. hawking , _ particle creation by black holes _ , commun .
phys . * 43 * , 199 ( 1975 ) .
h. kleinert , r. ruffini , s. xue , _ electron - positron pair production in space- or time - dependent electric fields _ , phys .
d * 78 * , 025011 ( 2008 ) .
a. hossenfelder , d. schwarz , w. greiner , _ particle production in time - dependent gravitational fields : the expanding mass shell _ , class .
* 20 * , 2337 ( 2003 ) .
v. s. popov , _ collapse to the centre " at z @xmath132 137 and critical nuclear charge _
, yad . fiz .
* 12 * , 429 ( 1970 ) [ sov .
* 12 * , 235 ( 1971 ) ] .
l. e. parker and d. j. toms , _ quantum field theory in curved spacetime_. ( cambridge , 2009 ) .
although the following discussion is already well represented in the literature ( see , for example @xcite ) , we restate the most important results here as a reference for the reader .
the dirac equation on a curved manifold ( with @xmath133 ) reads @xmath134 the dirac matrices @xmath135 satisfy the clifford algebra @xmath136 where @xmath137 is the reciprocal of the metric and @xmath22 is the rest mass of the spin-1/2 particle .
a tilde above a gamma matrix indicates that the matrix is to be taken in curved space - time , conversely , the absence of a tilde indicates that the gamma matrix is to be taken in flat minkowski space - time .
the covariant derivative is @xmath138 where the spin - affine connections , @xmath139 , may be determined from the relation @xmath140 the vierbein of vector fields , @xmath141 , are found form the equality @xmath142 where @xmath143 is the minkowski metric ( with signature @xmath144 here ) .
the christoffel symbols @xmath145 are @xmath146 this then yields the equations for @xmath12 and @xmath147 : @xmath148 g({r } ) = 0 , \\
\frac{dg({r})}{dr } - e^{\frac{\lambda}{2 } } \frac{\kappa}{r } g({r } ) + e^{\frac{\lambda}{2 } } \left [ e^{-\frac{\nu}{2}}\epsilon - m \right ] f({r } ) = 0.\end{gathered}\ ] ] | we consider the bound - state energy levels of a spin-1/2 fermion in the gravitational field of a near - black hole object . in the limit that the metric of the body becomes singular ,
all binding energies tend to the rest - mass energy ( i.e. total energy approaches zero ) .
we present calculations of the ground state energy for three specific interior metrics ( florides , soffel and schwarzschild ) for which the spectrum collapses and becomes quasi - continuous in the singular metric limit .
the lack of zero or negative energy states prior to this limit being reached prevents particle pair production occurring . therefore , in contrast to the coulomb case , no pairs are produced in the non - singular static metric .
for the florides and soffel metrics the singularity occurs in the black hole limit , while for the schwarzschild interior metric it corresponds to infinite pressure at the centre .
the behaviour of the energy level spectrum is discussed in the context of the semi - classical approximation and using general properties of the metric . |
consensus of multi - agent systems ( mass ) is a fundamental problem in collective behaviors of autonomous individuals , which has been extensively studied in the last decade @xcite .
the key problem for consensus is to design appropriate distributed protocols that each agent can only get information from its local neighbors , and the whole network of agents may coordinate to reach an agreement on certain quantities of interest eventually .
most of previous researches analyzed and solved the consensus problem based on the time - domain state - space model @xcite .
other studies utilized the nyquist stability criterion in the frequency domain @xcite .
one contribution of this paper is to provide new insights of the consensus problem by exploring the recent development of graph signal processing in spatial frequency domain .
graph signal processing has drawn great interests for analyzing high - dimensional data in recent years @xcite .
there are two different frameworks on graph signal processing .
one develops the discrete signal processing structure and concepts on graph based on the jordan normal form and generalized eigenbasis of the adjacency matrix@xcite . the other establishes the framework by merging algebraic and spectral graph concepts with classical signal processing which interprets the graph laplacian eigenvalues as graph frequencies and the orthonormal eigenvectors as graph fourier transformation @xcite . in this paper
, we will first reveal the explicit connection between filtering of graph signals and consensus of mass .
it is shown that the mas can reach its average consensus in finite time by designing an appropriate protocol filter to keep only the low frequency component of the graph ( corresponding to the zero eigenvalue of the graph laplacian matrix ) while surrogating other higher frequency components .
the designed graph filter can be implemented by a distributed consensus protocol derived from the closed - loop property of the mas and the laplacian matrix of the network graph .
viewing mass from graph signal processing perspective not only provides new insights , it also presents a new methodology to solve some challenging problems in mass on uncertain networks . for mass with estimated laplacian matrix
, we show that the graph signal processing perspective can help to design distributed consensus protocol gains as well as estimate the asymptotic consensus error , which is difficult to analyze by the existing time - domain state - space model based methods . for mass with completely unknown laplacian matrix except assuming the connectivity and the maximum degree of the network graph
, we provide a new design method of the consensus protocol gain .
the consensus error bound is also presented .
numerical examples are given to demonstrate the effectiveness of the proposed methods .
let @xmath0 be an undirected graph of order @xmath1 ( @xmath2 ) which consists of an agent set @xmath3 , an edge set @xmath4 , and a adjacency matrix @xmath5 \in { \re ^{n \times n}}$ ] .
an edge @xmath6 if and only if @xmath7 , that means there exist communications between agent @xmath8 and agent @xmath9 . moreover , self - edges are not allowed , i.e. , @xmath10 and @xmath11 .
the laplacian matrix of graph is defined by @xmath12 , where @xmath13 is the degree matrix , and @xmath14 .
the laplacian matrix @xmath15 is symmetric and all the eigenvalues are real .
* lemma 1 . * @xcite for an undirected graph @xmath16 , the laplacian matrix @xmath17 has at least one zero eigenvalue , and all the non - zero eigenvalues are positive , i.e. , @xmath18 .
furthermore , zero is a simple eigenvalue of @xmath17 with the associated orthonormal eigenvector @xmath19 , where @xmath20^t \in \re ^ n$ ] , if and only if the @xmath16 is connected . consider a graph @xmath0 , its laplacian matrix can be written as @xmath21 , where @xmath22 , @xmath23 \in \re ^ { n \times n}$ ] is a unitary matrix .
the graph signal @xmath24 is the collection of the signal values on all the agents , i.e. , @xmath25^t \in \re ^ n$ ] . similarly to classical signal processing , the fourier transform in the graph signal processing @xcite is defined on the graph spectra as @xmath26 , and the inverse graph fourier transform is given by @xmath27 , where @xmath28^t \in \re ^ n$ ] .
let @xmath29 be the transfer function of a filter , and @xmath30^t \in \re ^ n$ ] be the filtered graph signal in spatial frequency domain , then the graph spectral filtering can be defined as @xmath31 . taking the inverse graph fourier transform , the filtered graph signal in time domain
can be obtained as @xmath32 , where @xmath33^t \in \re ^ n$ ] is the filted graph signal in time domain .
consider the dynamics of the mas on a graph @xmath0 with @xmath1 agents described by @xmath34 where @xmath35 , @xmath36 is the state and the control input , respectively . a commonly used control protocol @xcite is as follows @xmath37 where @xmath38 is the control gain at time @xmath39 , and @xmath40_{n \times n}$ ] is the adjacency matrix of graph @xmath16 .
* definition 1 . * for a control protocol within time @xmath41 given by ( [ eq7 ] ) , the corresponding protocol filter is defined as @xmath42 the average consensus of multi - agent system ( [ eq6 ] ) under protocol ( [ eq7 ] ) is said to be achieved asymptotically if @xmath43 , @xmath44 . and
the average consensus is said to be reached at time @xmath41 if @xmath45 , @xmath46 , hold for all @xmath47 . from ( [ eq6 ] ) and ( [ eq7 ] ) , the dynamics of the multi - agent system can be calculated that @xmath48 let the initial state @xmath49 and the current state @xmath50 of the mas be the collective signal values of the original graph signal and the filtered one , respectively .
then , the mas plays a role of the filter for the graph signal with @xmath49 and @xmath51 , and the the transfer function of the filter is @xmath52 defined in definition 1 .
the following result gives the property of the protocol filter achieving average consensus .
* theorem 1 .
* for the mas ( [ eq6 ] ) on a connected graph @xmath16 , of which the laplacian matrix has eigenvalues as @xmath53 , assume the consensus protocol is in the form of ( [ eq7 ] ) .
then the mas reaches average consensus at time @xmath41 if and only if the corresponding protocol filter @xmath54 defined by ( [ eq07 ] ) satisfies @xmath55 and @xmath56 for @xmath57 .
the proof of theorem 1 is omitted due to space limitation .
it is easy to verify that the consensus state is @xmath58 .
theorem 1 shows that the protocol filter can be viewed as a low - pass filter with zero at high frequency components @xmath59 .
it follows from theorem 1 that the mas with @xmath1 agents on a connected graph can definitely reach the average consensus at time @xmath60 by properly choosing the control gain @xmath61 .
the following corollary shows that the consensus time can be smaller than @xmath60 .
* corollary 1 .
* for the mas ( [ eq6 ] ) on a connected graph with @xmath62 distinct nonzero eigenvalues ( @xmath63 ) , take the control gains as @xmath64 then the consensus protocol ( [ eq7 ] ) makes the mas reach average consensus at time @xmath62 , that is , @xmath65 .
* the fact that finite time consensus can be achieved by choosing the control gains equal to the reciprocal of nonzero laplacian eigenvalues is not new .
it has been obtained by using different methods , for example , matrix factorization method @xcite , minimal polynomial method @xcite . by defining the protocol filter @xmath54 , theorem 1
derives the consensus result from a graph signal processing perspective . in the next section
, we will show that this method is a powerful tool to solve the consensus of mass on uncertain networks , which is difficult ( sometimes unable ) to deal with by existing methods .
in the previous section , we assume that the graph structure of mass is completely known .
the exact average consensus in finite - time can be reached by designing a appropriate protocol filter .
this section will discuss the average consensus problem in two non - ideal cases : estimated laplacian matrix and unknown network topology .
denote the consensus error as @xmath66 , the consensus performance will be analyzed in both cases .
consider the mas on a connected graph @xmath67 with laplacian matrix @xmath68 .
denote the eigenvalues of @xmath68 as @xmath69 .
let @xmath70 be the estimated laplacian matrix of @xmath68 , where @xmath71
. for @xmath72 , denote @xmath73 * theorem 2 . * for the mas ( [ eq6 ] ) on the graph @xmath67 under the consensus protocol ( [ eq7 ] ) , take the control gain as @xmath74 where @xmath75 are eigenvalues of the estimated laplacian matrix @xmath76 . then the mas reaches average consensus asymptotically if @xmath77 and @xmath78 , where @xmath79 is defined in ( [ eq11 ] ) .
moreover , at time @xmath80 , the consensus error satisfies @xmath81 .
_ proof outline : _ the corresponding protocol filter at each control period can be written as @xmath82
. then the consensus error at time @xmath80 can be calculated by @xmath83 .
it is easy to see that @xmath84 since @xmath78 .
thus , the mass ( [ eq6 ] ) can reach the average consensus asymptotically by the protocol ( [ eq7 ] ) . for @xmath85 .
( b ) the trajectory of agent state @xmath86 for @xmath87 .
( c ) the trajectory of agent state @xmath86 for @xmath88.,width=302 ] for @xmath85 .
( b ) the consensus square error @xmath89 for @xmath87 .
( c ) the consensus square error @xmath89 for @xmath88.,width=302 ] _ example 1 : _ consider the mas on a connected graph @xmath67 with @xmath90 agents , the laplacian matrix is given as @xmath91 , where @xmath92 is the estimated laplacian matrix corresponding to an unweighted cycle network , and @xmath93 is the estimation error bound . from theorem 2 , the control gain in one control period can be designed as @xmath94 for @xmath95 .
let the simulation time be @xmath96 , respectively , then the evolution of agent states and the consensus error are shown in fig . 1 and fig
it can be seen that the mas ( [ eq6 ] ) under the designed protocol ( [ eq7 ] ) reaches average consensus asymptotically , and the consensus errors at each @xmath97 are @xmath98 , @xmath99 , and @xmath100 .
consider a connected graph @xmath101 with unknown network topology , the methods proposed in previous sections have no application to solve the average consensus problem in this case .
assume the maximum degree of the graph is given , i.e. , @xmath102 .
then , divide @xmath103 $ ] into @xmath104 uniform intervals , and the length of each interval is @xmath105 .
a general result for average consensus can be derived as follows . * theorem 3 .
* for the mas ( [ eq6 ] ) on an unknown connected graph @xmath101 with maximum degree @xmath106 , take the control gains as @xmath107 then the consensus protocol ( [ eq7 ] ) makes the mas reach average consensus asymptotically .
furthermore , assume the lower bound of the algebraic connectivity satisfies @xmath108 , then the consensus error at time @xmath109 satisfies @xmath110 , where @xmath111 .
_ proof outline : _ from the design of the control gain , the corresponding protocol filter at the end of each control period can be derived as @xmath112 .
it can be calculated that @xmath113 , where @xmath114 } \ { \prod\limits_{k = 0}^{{t_0 } - 1 } ( 1 - \frac{\lambda } { { ( { t_0 } - k)\phi } } ) \ } < 1 $ ] . and
the consensus error can be calculated as @xmath115 .
it is easy to see that @xmath116 since @xmath117 .
then the mas reaches the average consensus asymptotically under the protocol ( [ eq7 ] ) .
assume the lower bound of the algebraic connectivity satisfies @xmath108 , a more accurate upper bound of the protocol filter can be derived as @xmath118 .
then the consensus error at time @xmath109 satisfies @xmath119 .
it follows from theorem 3 that the algebraic connectivity of the graph plays an important role in reaching consensus , and the higher algebraic connectivity corresponding to a better consensus performance and a lower consensus error at each control period .
_ example 2 : _ consider a mas with @xmath90 agents on two different graphs , one is an unweighted cycle @xmath120 , the other is an unweighted path @xmath121 .
the maximum degrees of the two graphs are the same , i.e. , @xmath122 .
divide @xmath123 $ ] into @xmath124 uniform intervals , then @xmath125 and @xmath126 . from theorem 3 ,
the control gain in one period can be derived as @xmath127 , @xmath128 . for the mas ( [ eq6 ] ) on two graphs @xmath120 and @xmath121 respectively , the protocol ( [ eq7 ] ) with the designed periodic control gain can solve the average consensus asymptotically as shown in fig 3 . moreover , it is easy to verify that the algebraic connectivity of the cycle is higher than that of the path , thus the consensus performance of the graph @xmath120 is much better than the graph @xmath121 , and the consensus errors of the two graphs are shown in fig 4 . .
( b ) the agent state trajectory of the unweighted path @xmath121.,width=302 ] of the unweighted cycle @xmath120 .
( b ) the consensus error @xmath129 of the unweighted path @xmath121.,width=302 ]
this paper has established the explicit connection between filtering of graph signals and consensus of mass . it has been shown that the mas can reach its average consensus in finite time by designing an appropriate protocol filter , which can be implemented by a distributed consensus protocol . by using the concept of the protocol filter , we provide new methods to solve the average consensus problem in cases of estimated laplacian matrix and unknown network topology .
the asymptotic consensus error has been analyzed in both cases .
while the protocol filter is defined for mass on undirected graphs , it can be easily extended to direct graphs .
we only consider mass consisting of agents with first - order dynamics , it is interesting to extend our methods to mass with high - order dynamics .
s. zhu , m. hong and b. chen , `` quantized consensus admm for multi - agent distributed optimization , '' _ in proceedings of 2016 ieee international conference on acoustics , speech and signal processing ( icassp ) _ , pp . 4134 - 4138 , 2016 .
t. li and j. f. zhang , `` consensus conditions of multi - agent systems with time - varying topologies and stochastic communication noises , '' _ ieee transactions on automatic control _ , vol .
9 , pp . 2043 - 2057 , 2010 .
g. s. seyboth , d. v. dimarogonas , k. h. johansson , p. frasca and f. allgower , `` on robust synchronization of heterogeneous linear multi - agent systems with static couplings , '' _ automatica _ , vol .
392 - 399 , 2015 .
r. olfati - saber and j. s. shamma , `` consensus filters for sensor networks and distributed sensor fusion , '' _ in proceedings of the 44th ieee conference on decision and control ( cdc ) _ , pp .
6698 - 6703 , 2005 .
d. i. shuman , s. k. narang , p. frossard , a. ortega and p. vandergheynst , `` the emerging field of signal processing on graphs : extending high - dimensional data analysis to networks and other irregular domains , '' _ ieee signal processing magazine _ , vol .
3 , pp . 83 - 98 , 2013 .
j. mei and j. m. f. moura , `` signal processing on graphs : estimating the structure of a graph , '' _ in proceedings of 2015 ieee international conference on acoustics , speech and signal processing ( icassp ) _ , pp .
5495 - 5499 , 2015 .
a. sandryhaila , s. kar and j. m. f. moura , `` finite - time distributed consensus through graph filters , '' _ in proceedings of 2014 ieee international conference on acoustics , speech and signal processing ( icassp ) _ , pp .
1080 - 1084 , 2014 .
j. mei and j. m. f. moura , `` signal processing on graphs : performance of graph structure estimation , '' _ in proceedings of 2016 ieee international conference on acoustics , speech and signal processing ( icassp ) _ , pp .
6165 - 6169 , 2016 .
e. pavez and a. ortega , `` generalized laplacian precision matrix estimation for graph signal processing , '' _ in proceedings of 2016 ieee international conference on acoustics , speech and signal processing ( icassp ) _ , pp .
6350 - 6354 , 2016 .
s. k. narang , a. gadde and a. ortega , `` signal processing techniques for interpolation in graph structured data , '' _ 2013 ieee international conference on acoustics , speech and signal processing ( icassp ) _ , pp .
5445 - 5449 , 2013 .
a. anis , a. gadde and a. ortega , `` towards a sampling theorem for signals on arbitrary graphs , '' _ 2014 ieee international conference on acoustics , speech and signal processing ( icassp ) _ , pp .
3864 - 3868 , 2014 .
a. gadde and a. ortega , `` a probabilistic interpretation of sampling theory of graph signals , '' _ 2015 ieee international conference on acoustics , speech and signal processing ( icassp ) _ , pp . 3257 - 3261 , 2015 . | this paper revisits the problem of multi - agent consensus from a graph signal processing perspective . by defining the graph filter from the consensus protocol
, we establish the direct relation between average consensus of multi - agent systems and filtering of graph signals .
this relation not only provides new insights of the average consensus , it also turns out to be a powerful tool to design effective consensus protocols for uncertain networks , which is difficult to deal with by existing time - domain methods . in this paper
, we consider two cases , one is uncertain networks modeled by an estimated laplacian matrix and a fixed eigenvalue bound , the other is connected graphs with unknown topology .
the consensus protocols are designed for both cases based on the protocol filter .
several numerical examples are given to demonstrate the effectiveness of our methods .
multi - agent system , graph signal processing , average consensus |
baryogenesis through leptogenesis represents an attractive mechanism to explain the observed matter - antimatter asymmetry of the universe @xcite .
once the standard model ( sm ) is extended by including the see - saw mechanism in order to explain the strong suppression of the neutrino mass scale @xcite , the possibility of generating a cosmic lepton asymmetry via the lepton number and @xmath3 violating out - of - equilibrium decays of the seesaw singlet neutrinos arises as a natural possibility .
partial conversion of the lepton asymmetry into a baryon asymmetry then proceeds by means of electroweak sphaleron interactions @xcite that are non - perturbative sm processes .
qualitatively , it is almost unavoidable that a model that includes the seesaw mechanism will predict a certain amount of matter - antimatter asymmetry surviving until the present epoch , and then the question of whether leptogenesis is able to explain the puzzle of the baryon asymmetry of the universe is essentially a quantitative one . in recent years
, quantitative analysis of the standard leptogenesis scenario have become more and more sophisticated , taking into account many subtle but significant ingredients , such as various washout effects @xcite , thermal corrections to particle masses and @xmath3 violating asymmetries @xcite , spectator processes @xcite , flavor effects @xcite and the possible effects of the heaviest right handed majorana neutrinos @xmath5 @xcite ( for reviews of the most recent results see @xcite ) .
one assumption that is common to all these studies is that between the scale of the breaking of lepton number and the electroweak breaking scale , there are no additional sources of new physics that could affect the mechanism of leptogenesis .
this assumption is certainly justified both in terms of simplicity and also because it allows for a certain level of predictivity , that is mainly due to the fact that the same couplings that determine the @xmath3 asymmetries and the out - of - equilibrium conditions in the decays of the heavy majorana singlets are also responsible for the seesaw masses of the light neutrinos .
in particular , in the standard seesaw model , successful leptogenesis implies a lower bound on the values of @xmath0 that , even in the most favorable case of dominant @xmath4 initial abundance @xcite , puts a direct test of leptogenesis out of the reach of any foreseeable experiment.gev @xcite seem to be possible . ] in this paper we explore the implications of the presence below @xmath0 , of an additional energy scale related to the breaking of a new symmetry ( that could be for example a flavor symmetry ) .
as we will see , in some cases a large enhancement of the @xmath3 asymmetry in @xmath4 decays is easily obtained , while at the same time the scale of leptogenesis can be lowered by several orders of magnitude without conflicting with other conditions .
indeed , the scenario we will study can realize in a natural way some of the conditions needed to render leptogenesis viable down to the tev scale @xcite .
the main features of the model , that is directly inspired by @xmath1 models for flavor la froggatt - nielsen @xcite are outlined in section [ sec : model ] . in section [ sec
: cp - asymmetries ] , after reviewing the main results of standard leptogenesis , we highlight the most important new features that stem from the presence of the extra mass scale . in section [ sec
: conclusions ] we present the conclusions . in the appendix
we consider a minimal scenario where one light neutrino remains massless , and we derive a simple analytical expression for the ratio between the new mass scale and @xmath0 .
we assume that at some large scale close to the leptogenesis scale a horizontal @xmath6 symmetry forbids direct couplings between the lepton doublets @xmath7 and the heavy majorana neutrinos @xmath8 .
light neutrino masses can arise because of the presence of heavy vectorlike fields @xmath9 that are singlets with respect to @xmath10 and charged under the additional @xmath6 factor , and that couple to both the majorana fields and to the lepton doublets @xmath7 .
we use greek indices @xmath11 to denote the heavy majorana neutrinos and we write the lagrangian in terms of self - conjugate majorana spinors @xmath12 .
latin indices @xmath13 will denote the heavy dirac fields with @xmath14 , and latin indices @xmath15 will denote the sm left - handed lepton doublets @xmath16 .
the following lagrangian gives a simple realization of this scheme : @xmath17 where @xmath18 is the @xmath19 higgs doublet and @xmath20 is a complex scalar that is a singlet under @xmath10 and charged under @xmath6 .
the scalar @xmath20 is responsible for breaking the additional symmetry through a vacuum expectation value ( vev ) @xmath21 that we assume to be somewhat smaller than the mass scale of the vectorlike fields @xmath22 .
a simple @xmath6 charge assignment that forbids the @xmath23 coupling and yields the lagrangian ( [ eq : lag ] ) is for example @xmath24 , @xmath25 and @xmath26 . to put in evidence the main consequences of our scheme without complicating too much the discussion and the results
, we assume that to a good approximation the heavy dirac fields @xmath27 couple to the majorana neutrinos in a pure vectorlike way .
that is , we assume @xmath28 , and we neglect all the effects related to the pseudoscalar couplings . including also the pseudoscalar interactions would give rise to additional diagrams contributing for example to the total decay width of the majorana neutrinos and to the decay @xmath3 asymmetries , without changing our main conclusions .
let us note that all the interaction terms in eq .
( [ eq : lag ] ) also preserve a @xmath1 ( accidental ) global symmetry with assignment @xmath29 and @xmath30 , that we can readily identify with lepton number . then , as in the standard seesaw model , this symmetry is broken ( by two units ) only by the @xmath31 majorana mass term . as regards the @xmath6 symmetry , if for all the sm fields the charges @xmath32 were proportional to @xmath33 , this symmetry would leave unaffected all the charged leptons and quarks yukawa couplings .
in contrast , if the set of @xmath32 charges for the quarks and leptons is not a trivial one , @xmath6 could play a role as ( part of ) a flavor symmetry of the kind proposed long ago by froggatt and nielsen @xcite .
however , for the present discussion the issue if @xmath6 also contributes to determine the charged fermion mass pattern is to a large extent irrelevant , and accordingly we will not necessarily adopt all the assumptions that realize the naturalness conditions of flavor models based on abelian symmetries , and in particular : * we do not constrain the couplings in the lagrangian eq .
( [ eq : lag ] ) to be all of the same size and of order unity .
* we do not need to specify any precise value for the ratio @xmath34 .
* we do not constrain from below by means of fcnc considerations the scale of @xmath6 breaking ( this is consistent if e.g. we take @xmath35 ) .
after @xmath6 and electroweak symmetry breaking , the light neutrino mass matrix arising from ( [ eq : lag ] ) reads @xmath36_{ij } = \left[\tilde\lambda^t \frac{v^2}{m_n}\tilde\lambda\right]_{ij}\ ] ] where for convenience we have introduced the effective seesaw couplings @xmath37 that is , with respect to the standard seesaw mechanism the light neutrino masses have an additional suppression factor of the ratio @xmath38 and are of fourth order in the fundamental couplings ( @xmath39 and @xmath40 ) . an important point to note
is that in order to ensure that two light neutrino are massive , as is required by oscillation neutrino data , a minimum field content of two right - handed neutrinos @xmath41 and two vectorlike fields @xmath27 is needed .
a straightforward analysis then shows that even in this minimal scheme , both the matrices of the @xmath40 and @xmath39 coupling constants contain physical complex phases that can be relevant for leptogenesis . to summarize , in this model besides the electroweak breaking scale @xmath42 we have the following new mass scales : * the mass scale of the heavy vectorlike fields , @xmath43 .
* the lepton number breaking scale , @xmath44 . * the horizontal symmetry breaking scale , @xmath45 .
the consequences of the different hierarchies amongst these new scales is studied in detail in section [ sec : cp - asymmetries ] .
in this section we present the detailed results for the @xmath3 asymmetry in @xmath4 decays for the different cases as are determined by the hierarchy between the relevant scales @xmath0 , @xmath43 and @xmath45 , and we will explore qualitatively the implications of the different possibilities .
as already said , we take @xmath4 to be the lightest one of the majorana neutrinos , and @xmath46 to be the lightest of the vectorlike fields . for the different mass ratios we adopt the following notation : @xmath47 the @xmath3-asymmetry in the decay of the heavy majorana neutrinos @xmath4 is defined in the usual way as @xmath48 where @xmath49 and @xmath50 represent respectively the partial decay rates of @xmath4 into particles with lepton number @xmath51 and antiparticles with lepton number @xmath52 ( regardless of the fact that they are @xmath7 or @xmath2 states ) . a non - vanishing numerator in eq.([eq : cpdecay ] ) can arise from the interference between tree - level and loop - amplitudes , and the total decay rate in the denominator can be approximated with the tree level result .
when the ratio in eq.([eq : cpdecay ] ) is phase - space independent , as is the case in two - body decays , the @xmath3-asymmetry can be simply calculated in terms of products of the tree - level @xmath53 and loop @xmath54 amplitudes . however , when there is a dependence on the phase space as is the case for three - body decays discussed in section [ sec : case1 ] , the full decay widths have to be taken into account . in standard leptogenesis models
the lepton number breaking scale is constrained by the out - of - equilibrium condition on the decay rate of @xmath4 , and as the scale of lepton number violation is lowered , in order to satisfy this condition the size of the @xmath4 yukawa couplings must be accordingly reduced .
if we further require that the neutrino oscillation data are accounted for just by the ( type 1 ) seesaw mass matrix , and we forbid any additional source for the light neutrino masses , then it can be shown that a large suppression of the @xmath3 asymmetry in @xmath4 decays is unavoidable .
then the requirement that the final lepton asymmetry is large enough to account for the baryon asymmetry of the universe , implies a lower bound on the mass of @xmath4 that is several orders of magnitude larger than the electroweak breaking scale . as we will discuss below , in some cases
the presence of a new scale and of a new set of couplings associated with it can allow to satisfy the out - of - equilibrium condition and to account for the light neutrino mass scale , without necessarily implying any particular suppression of the @xmath3 asymmetries , even when @xmath0 is lowered down to the tev scale .
this ` decoupling ' of the size of the @xmath3 asymmetry from the decay width @xmath49 and from the scale of the light neutrino masses is rendered possible by the fact that , while the latter two quantities are mainly controlled by the @xmath39 parameters that couple the heavy vectorlike fields @xmath2 to the right handed neutrino @xmath4 , the @xmath3 asymmetry is essentially determined by the couplings @xmath40 between the fermions @xmath2 and the lepton doublets @xmath7 . when the masses of the heavy dirac fields and the @xmath6 symmetry breaking scale are both larger than the majorana neutrino masses ( @xmath56 ) there are no major differences from the standard fukugita - yanagida leptogenesis model @xcite .
after integrating out the @xmath2 fields one obtains the standard seesaw lagrangian containing the effective operators @xmath57 with the seesaw couplings @xmath58 given in eq .
( [ eq : tildelambda ] ) .
the right handed neutrino @xmath4 decays predominantly via 2-body channels as shown in fig .
[ fig : case0 ] .
this yields the standard results that for convenience we recall here .
the total decay width is @xmath59 and the sum of the vertex and self - energy contributions to the @xmath3-asymmetry for @xmath4 decays into the flavor @xmath60 reads @xcite @xmath61\right\},\ ] ] where @xmath62 at leading order in @xmath63 and after summing over all leptons @xmath60 , eq .
( [ eq:6 ] ) yields for the total asymmetry : @xmath64 where the sum over the heavy neutrinos has been extended to include also @xmath4 since for @xmath65 the corresponding combination of couplings is real . in the hierarchical case @xmath66 the size of the total asymmetry in ( [ eq:8 ] ) is bounded by the davidson - ibarra limit @xcite @xmath67 where @xmath68 ( with @xmath69 ) are the light neutrinos mass eigenstates and @xmath70ev@xmath71 is the atmospheric neutrino mass difference .
it is now easy to see that ( [ eq : di ] ) implies a lower limit on @xmath0 .
the amount of @xmath72 asymmetry that can be generated from @xmath4 dynamics can be written as : @xmath73 where @xmath74 accounts for the dilution of the asymmetry due to the increase of the universe entropy from the time the asymmetry is generated with respect to the present time , @xmath75 ( that can range between 0 and 1 , with typical values @xmath76 ) is the _ efficiency factor _ that accounts for the amount of @xmath77 asymmetry that can survive the washout process .
wmap data on the cosmic background anisotropy @xcite and considerations of big bang nucleosynthesis @xcite yield the experimental value @xmath78 , and therefore , assuming that @xmath79 is the main source of the @xmath33 asymmetry @xcite , eqs .
( [ eq : di ] ) and ( [ eq : etab ] ) yield : @xmath80 this limit can be somewhat relaxed depending on the specific initial conditions @xcite or when flavor effects are included @xcite but the main point remains , and that is that the value of @xmath0 should be well above the electroweak scale .
( -4,0.4)(6,3 ) ( -3.5,2)(-2.5,2 ) ( -2.5,2)(-2.2,2.3 ) ( -2.2,2.3)(-2.0,2.5 ) ( -2.5,2)(-2.2,1.7 ) ( -2.2,1.7)(-2,1.5 ) ( -3.5,2 ) ( -2,2.5 ) ( -2,1.5 ) ( -2.7,1 ) ( -1,2)(0,2 ) ( 0,2)(0.3,2.3 ) ( 0.25,2.25)(0.5,2.5 ) ( 0,2)(0.3,1.7 ) ( 0.15,1.85)(0.5,1.5 ) ( 0.5,2.5)(0.5,1.5 ) ( 0.5,1.5)(0.9,1.5 ) ( 0.9,1.5)(1.2,1.5 ) ( 0.5,2.5)(1,2.5 ) ( 0.9,2.5)(1.2,2.5 ) ( -0.8,2 ) ( 0.3,2.4 ) ( 0.3,1.5 ) ( 0.4,2 ) ( 1.1,1.5 ) ( 1.1,2.4 ) ( 0.5,1 ) ( 2.1,2)(2.8,2 ) ( 3.2,2)0.4cm090 ( 3.2,2)0.4cm90180 ( 3.2,2)0.4cm180270 ( 3.2,2)0.4cm270360 ( 3.6,2)(4.1,2 ) ( 4.1,2)(4.4,2.3 ) ( 4.3,2.2)(4.6,2.5 ) ( 4.1,2)(4.4,1.7 ) ( 4.4,1.7)(4.6,1.5 ) ( 3.2,2.4 ) ( 3.2,1.6 ) ( 2.3,2 ) ( 3.9,2 ) ( 4.6,1.5 ) ( 4.6,2.4 ) ( 3.2,1 ) interesting new possibilities arise when the @xmath6 symmetry breaking scale is lower than the leptogenesis scale , that is @xmath81 . in this case
@xmath4 can not decay directly into the light lepton doublets via the two body channel . in this regime
, we can distinguish three cases : * @xmath81 but all the masses @xmath82 are larger than @xmath0 : then @xmath4 can decay to @xmath60 only via the three body channel @xmath83 depicted in figure [ fig : case1 ] .
* @xmath0 is larger than all the scales related to the @xmath6 symmetry ( @xmath84 ) : then @xmath4 will decay via two body channels to @xmath27 and @xmath85 , ( see figure [ fig : case2 ] ) .
the heavy fermions @xmath2 will then transfer part of the asymmetry to the light leptons via lepton number conserving processes ( decays and scatterings ) . *
if @xmath44 arises from the same source than @xmath43 ( as for example from the vev of a singlet ) then some of the heavy fermions ( for example @xmath46 ) could be lighter than @xmath4 while the others can be heavier .
then @xmath4 will decay dominantly into @xmath86 via the two body channel in figure [ fig : case2 ] .
however , a new diagram contributing to the @xmath3 asymmetry is present in this case .
this diagram is interesting , since it yields the possibility of decoupling the lifetime of @xmath4 from the size of the @xmath3 violating asymmetry @xmath87 . in the next sections we will analyze in some detail these different possibilities . in the case
when all the vectorlike fermions @xmath2 are heavier than @xmath4 but the horizontal @xmath6 is still a good symmetry at @xmath89 , there is only one diagram contributing to the @xmath3 asymmetry , that is the wave function type of diagram depicted in figure [ fig : case1](b ) .
this is because on the one hand the loop correction to the @xmath90 vertex with @xmath91 ( @xmath92 ) and @xmath2 internal lines does not develop an imaginary part , and on the one other hand there is no correction at one loop to the vertex @xmath93 ( the first correction arises only at two - loops ) .
this is due to the fact that the standard vertex correction depicted in figure [ fig : case0]@xmath94 is of the majorana type ( with an inverted flow of fermion number ) while in the present case both @xmath2 and @xmath7 are dirac fermions .
( -4,-1.5)(4,2 ) ( -4.5,0)(-3,0 ) ( -4.3,0 ) ( -3,0)(-2,0 ) ( -3,0 ) ( -2,0)(-1.5,0 ) ( -1.5,0 ) ( -3.7,1)(-3.7,0.6 ) ( -3.7,0.6)(-3.7,0 ) ( -3.7,1 ) ( -2.5,1)(-2.5,0.6 ) ( -2.5,0.6)(-2.5,0 ) ( -2.5,1 ) ( -3,-1 ) ( 0,0)(1.3,0 ) ( 1.2,0)(1.7,0 ) ( 0.7,0)(0.7,0.6 ) ( 0.7,0.6)(0.7,1 ) ( 2.1,0)0.4cm090 ( 2.1,0)0.4cm90180 ( 2.1,0)0.4cm180270 ( 2.1,0)0.4cm270360 ( 2.5,0)(3.2,0 ) ( 3.1,0)(3.6,0 ) ( 3.5,0)(3.5,0.6 ) ( 3.5,0.6)(3.5,1 ) ( 3.5,0)(3.9,0 ) ( 3.8,0)(4.2,0 ) ( 0.2,0 ) ( 1.1,0 ) ( 0.8,1 ) ( 2.1,0.4 ) ( 2.1,-0.4 ) ( 3.1,0 ) ( 4.2,0 ) ( 3.5,1 ) ( 2.1,-1 ) unlike the standard case and the next two cases discussed in sections [ sec : case2 ] and [ sec : case3 ] , for three body final states the squared amplitudes and interference terms are not phase - space independent . our results for the total decay width and @xmath3-asymmetries given below correspond to the leading terms in the mass ratios @xmath95 .
for the @xmath4 total decay width we obtain @xmath96 the interference between diagrams [ fig : case1]@xmath97 and [ fig : case1]@xmath94 , summed over all the leptons @xmath98 running in the loop , yields the @xmath3-asymmetry : @xmath99\ , r_{a'}\,r_a\,r_b^2\ , } { \sum_{a\,a'}\>\left(h^\dagger h\right)_{a'a}\left(\lambda_{1a ' } \lambda_{1a}^*\right)\,r_a\,r_{a ' } } \\ & = & \frac{3}{128\pi } \frac { { \mathbb i}\mbox{m } [ \,\tilde\lambda\ , h\ , r^2\ , h^\dagger\,\tilde\lambda^\dagger]_{11 } } { ( \tilde \lambda\tilde \lambda^\dagger)_{11 } } = 0.\end{aligned}\ ] ] when written in terms of the effective couplings @xmath100 as in the second line in eq .
( [ eq : cpasymmetrythreebody ] ) , the vanishing of the @xmath3 asymmetry at leading order in the mass ratios @xmath95 is apparent .
however , this is true also beyond the leading order , and is related to the fact that the loop in diagram [ fig : case1]@xmath94 does not involve lepton number violation .
mathematically , one can see from the first equality in eq .
( [ eq : cpasymmetrythreebody ] ) that the vanishing of the @xmath3-asymmetry follows from the fact that by exchanging the two indices @xmath101 the combination of couplings within square brackets goes into its complex conjugate , and thus the sum over @xmath102 and @xmath103 yields a real term for any value of the index @xmath104 . since the full phase - space function ( that in eq .
( [ eq : cpasymmetrythreebody ] ) has been replaced with the leading term @xmath105 ) is symmetric under the exchange @xmath106 this holds also for the complete result . the fact that there is no source term for the lepton asymmetry implies that in the present case leptogenesis can occur only through the effects of lepton flavor dynamics combined with the violation of lepton number that is provided by the washout processes .
the single lepton - flavor @xmath3-asymmetries are indeed non - vanishing : @xmath107\,r_{a'}\,r_a\,r_b^2\ , } { \sum_{a\,a'}\;\left(h^\dagger h\right)_{a'a}\lambda_{1a ' } \lambda_{1a}^*\,r_{a'}\,r_a } \\ & = & \frac{3}{128\pi } \frac{\sum_i { \mathbb i}\mbox{m } \left[\left(h r^2 h^\dagger\right)_{ij } \tilde\lambda_{1i}\tilde\lambda^*_{1j}\right ] } { \left(\tilde \lambda\tilde \lambda^\dagger\right)_{11 } } \neq 0\,.\end{aligned}\ ] ] and will be a source of non vanishing asymmetry - densities in the different flavors .
these asymmetries will then suffer washouts processes ( like @xmath108 and @xmath109 channels scatterings @xmath110 , @xmath111 ) that in general are characterized by different rates for the different flavors . as is discussed e.g. in ref .
@xcite , under these conditions a net lepton asymmetry can result even if @xmath112 .
a quick inspection of equation ( [ eq : cpasymmetrythreebodyflavor ] ) shows that if the @xmath113 couplings are all of the same order of magnitude , the @xmath3 asymmetry @xmath114 is roughly proportional to @xmath115 .
that is , the dominant contribution to the @xmath3 asymmetry is determined by the @xmath40 couplings that are different from the effective couplings @xmath113 appearing in the neutrino mass matrix , and in particular larger by a factor of @xmath116 ( see eq .
( [ eq : tildelambda ] ) ) .
if we further assume the hierarchy @xmath117 between the fundamental couplings then , independently of the particular value of @xmath0 , leptogenesis will always occur in a regime when the @xmath27 interactions with the light leptons determine a ` flavor ' basis @xmath118 .
this ensures that the requirement that flavor dynamics participates in the generation of a lepton asymmetry is satisfied .
we will now address the following two interesting points : 1 ) in the present case is leptogenesis still compatible with a reasonable scale for the light neutrino masses ?
2 ) is the leptogenesis scale still bounded from below , or can it be lowered to values that are experimentally accessible ?
the out - of - equilibrium condition necessary to ensure that a macroscopic lepton asymmetry can be generated reads @xmath119 where the out - of - equilibrium parameter @xmath120 can normally lie in the range @xmath121 ( but values as large as @xmath122 are possible ) and the hubble parameter at decay time is @xmath123 with @xmath124 the number of relativistic degrees of freedom at @xmath125 and @xmath126 the planck mass . using ( [ eq : width - decay - treel ] ) this yields the condition @xmath127 from equation ( [ eq : numatrix ] ) we have @xmath128 where in the second relation we have assumed that @xmath0 dominates the seesaw matrix . from ( [ eq : out - of - eq ] ) and ( [ eq : trace ] ) we obtain the order - of - magnitude relation @xmath129 this ensures that if the ratio @xmath130 is not exceedingly small , the out - of - equilibrium condition can be satisfied for the correct scale of neutrino masses . when the value @xmath131 suggested by the previous equation is inserted into eq .
( [ eq : out - of - eq ] ) we obtain @xmath132 , that is the @xmath100 couplings should be of the order of the electron yukawa coupling when @xmath0 is at the tev scale . on the other hand
this limit does not constrain the size of the asymmetry in eq .
( [ eq : cpasymmetrythreebodyflavor ] ) , and for @xmath133 the @xmath3 asymmetry could be sufficiently large for successful leptogenesis .
for such a low leptogenesis scale , direct production of the @xmath2 states could be possible in collider experiments e.g. via off - shell @xmath32 boson exchange ( @xmath134 ) .
however , @xmath135 decays ( if kinematically accessible ) would be strongly suppressed by the small values of @xmath39 , and thus a direct detection of the majorana neutrinos would be a rather difficult task
. in this case @xmath4 decays proceed through the diagram in figure [ fig : case2]@xmath97 .
a lepton asymmetry is first generated in the @xmath2 states , and is transferred in part to the light leptons through the @xmath77 conserving interactions controlled by the couplings @xmath40 .
if we assume that the @xmath40-interactions are in equilibrium at @xmath125 ( as is the case if the couplings @xmath40 are larger than the couplings @xmath137 ) , then the reduction of the asymmetry in the @xmath2 states implied by chemical equilibrium with the light leptons @xmath7 also implies a reduction in the rates of the washout processes , with a corresponding enhancement of the efficiency @xmath138 .
this would favor the survival of a sizeable asymmetry .
neglecting in first approximation phase space suppressions from final state masses , the total decay width for this case reads : @xmath139 the @xmath3 asymmetry is determined by the interference between diagram @xmath97 and the loop diagrams @xmath94 and @xmath140 in figure [ fig : case2 ] .
( -3,0.4)(4,3 ) ( -3.5,2)(-2.5,2 ) ( -2.5,2)(-2.2,1.7 ) ( -2.2,1.7)(-2,1.5 ) ( -2.5,2)(-2.2,2.3 ) ( -2.2,2.3)(-2.0,2.5 ) ( -3.5,2 ) ( -2,2.5 ) ( -2,1.5 ) ( -2.7,1 ) ( -1,2)(0,2 ) ( 0,2)(0.3,2.3 ) ( 0.3,2.3)(0.5,2.5 ) ( 0,2)(0.3,1.7 ) ( 0.3,1.7)(0.5,1.5 ) ( 0.5,2.5)(0.5,1.5 ) ( 0.5,1.5)(0.9,1.5 ) ( 0.9,1.5)(1.2,1.5 ) ( 0.5,2.5)(1,2.5 ) ( 0.9,2.5)(1.2,2.5 ) ( -0.8,2 ) ( 0.4,2.5 ) ( 0.5,1.5 ) ( 0.4,2 ) ( 1.1,1.5 ) ( 1.1,2.4 ) ( 0.5,1 ) ( 2.1,2)(2.8,2 ) ( 3.2,2)0.4cm090 ( 3.2,2)0.4cm90180 ( 3.2,2)0.4cm180270 ( 3.2,2)0.4cm270360 ( 3.6,2)(4.1,2 ) ( 4.1,2)(4.4,2.3 ) ( 4.3,2.2)(4.6,2.5 ) ( 4.1,2)(4.4,1.7 ) ( 4.4,1.7)(4.6,1.5 ) ( 3.2,2.4 ) ( 3.2,1.6 ) ( 2.3,2 ) ( 3.9,2 ) ( 4.6,1.5 ) ( 4.6,2.4 ) ( 3.2,1 ) note that even if these diagrams are the same as in the standard case , the result for the asymmetry is different , since in contrast to the lepton doublets the vectorlike fields @xmath2 do not couple chirally to the majorana neutrinos . from the interference of diagrams _ ( a ) _ and _ ( b ) _ in figure [ fig : case2 ] we obtain : @xmath141 f_1(z_\beta)\ ] ] where @xmath46 given by @xmath142 \log\left ( \frac{1+z } { z } \right)\ \nonumber \\ & \stackrel{z \to \infty}{= } & \ \frac{1}{2\,\sqrt{z } } + \frac{1}{6\,z } + \dots \end{aligned}\ ] ] as regards the self - energy type of diagrams , like in the standard case besides the diagram depicted in figure [ fig : case2]@xmath140 there is another diagram @xmath143 of the majorana type with opposite fermion flux in the loop .
the contribution from the interference of these diagrams with diagram @xmath144 is : @xmath145\lambda_{\beta a}\lambda^*_{1a}\right\}\ , f_2(z_\beta)\ ] ] with @xmath146 adding the two contributions ( [ eq : case2-diagb-6 ] ) and ( [ eq : cas22-diagc-2 ] ) , summing over all the possible final states @xmath27 , and using the fact that @xmath147 = 0 $ ] to include in the leading order result also the sum over @xmath65 , we obtain : @xmath148 even if the two functions @xmath46 and @xmath149 are not the same as in the standard case with chiral leptons couplings @xcite , once the effective couplings @xmath100 are replaced with the fundamental @xmath39 s , at leading order eq .
( [ eq : epsilon - case2 ] ) coincides with the standard expression eq .
( [ eq:8 ] ) . however , the fact that the couplings in eq .
( [ eq : epsilon - case2 ] ) are not the seesaw couplings @xmath113 is rather important , because the fundamental @xmath39 s do not directly determine the light neutrinos masses .
this implies that for this case the bound of eq .
( [ eq : di ] ) does not hold .
as regards the consistency among different phenomenological requirements , an order of magnitude estimate suggests that there is some tension between the out - of - equilibrium condition ( [ eq : outofeq - general ] ) , that by means of eq .
( [ eq : width1 ] ) yields @xmath150 , and the light neutrino mass scale .
in fact the latter can get too much suppression from the small values of @xmath100 that are reduced by the factor @xmath151 with respect to the @xmath39 that determine the decay rate .
this yields the rough estimate @xmath152ev .
while it is always possible to assume a large hierarchy between the combination of couplings that control the out - of - equilibrium condition and the couplings that determine the light neutrino masses , the fair conclusion is that in the case under discussion successful leptogenesis can be ensured only by means of a careful choice of the relevant parameters .
this case corresponds to the situation when the value of @xmath0 lies in between different values of @xmath82 .
for definiteness we assume @xmath154 with @xmath155 . neglecting the contributions from three body
decays @xmath156 ( suppressed by @xmath157 and by phase space factors ) the @xmath4 decay rate reads @xmath158 for the @xmath3 asymmetry , besides the diagrams in figure [ fig : case2]@xmath97 the new type of diagram depicted in figure [ fig : case3]@xmath159 also contributes .
this contribution is qualitatively different from the previous ones in figure [ fig : case2 ] since it involves the couplings @xmath160 of the light leptons @xmath16 to the vectorlike fermions @xmath27 .
it is easy to see that in the case when the @xmath40 couplings dominate over the @xmath39 , diagram [ fig : case3]@xmath159 gives the leading contribution to the @xmath3 asymmetry .
the interference between diagrams [ fig : case2]@xmath97 and [ fig : case3]@xmath159 yields : @xmath161 f_2(\omega_{b } ) \,,\ ] ] where the function @xmath149 is given in eq .
( [ eq : cas22-diagc - f ] ) . after approximating @xmath162 ) the sum
can be extended over all @xmath104 since @xmath163 = 0 $ ] . in this case , while the constraint from the out - of - equilibrium condition is only slightly relaxed with respect to the previous case ( @xmath164 ) , it involves only the coupling @xmath165 .
it is then conceivable that some mechanism could suppress just this particular entry , without affecting the light neutrino mass scale ( e.g. a texture zero in the matrix of the @xmath39 s lifted by some higher order effect ) . as in the case discussed in section [ sec : case1 ] , small values of @xmath39 do not necessarily imply that the @xmath3 asymmetry is small , since the contribution in eq .
( [ eq : cp - asymm - case3 ] ) depends in a crucial way on the size of the couplings @xmath40 , and the factor of @xmath166 in the denominator can also enhance @xmath87 .
also in this case , the scale of leptogenesis could be much lower than the bound in eq .
( [ eq : m1limit0 ] ) without being in conflict with other conditions .
for example , by taking @xmath167 of the order of the electron yukawa coupling @xmath168 the out - of - equilibrium condition can be satisfied for values of @xmath0 as low as @xmath169tev . by assuming that the @xmath39 couplings different from @xmath165 are at least of the order of the @xmath170 yukawa @xmath171 , and taking for the seesaw suppression factor @xmath172
, we also obtain a reasonable value for the light neutrino mass scale . at the same time , the contribution to the @xmath3 asymmetry in eq .
( [ eq : cp - asymm - case3 ] ) can remain as large as @xmath173 .
( -4,2.3)(0,5 ) ( -4,3.7)(-2.7,3.7 ) ( -2.8,3.7)(-2.3,3.7 ) ( -3.3,3.7)(-3.3,4.1 ) ( -3.3,4.1)(-3.3,4.5 ) ( -1.9,3.7)0.4cm090 ( -1.9,3.7)0.4cm90180 ( -1.9,3.7)0.4cm180270 ( -1.9,3.7)0.4cm270360 ( -1.5,3.7)(-.8,3.7 ) ( -.9,3.7)(-.4,3.7 ) ( -3.8,3.7 ) ( -2.7,3.7 ) ( -3.3,4.5 ) ( -1.9,4.1 ) ( -1.9,3.3 ) ( -1.9,2.7 ) ( -.7,3.7 )
in this paper we have explored the consequences for leptogenesis of adding to the seesaw model a new scale related to the breaking of an additional @xmath6 symmetry .
our framework is inspired by @xmath1 models for flavor la froggatt - nielsen , but it differs from the usual schemes firstly in its simplicity , and secondly because we do not impose particular conditions for the values of the fundamental couplings , nor on the ratio between the symmetry breaking scale and the masses of the heavy vectorlike fermions . as a consequence , while the model represents an interesting playground to study variations of the standard leptogenesis scenario , it does not pretend to account also for the pattern of fermion masses and mixings .
( this could still be achieved in more complicated schemes in which @xmath6 appears as a component of the full flavor symmetry . )
we have found that in all the cases in which leptogenesis occurs while the @xmath6 symmetry is still unbroken , the expressions for the @xmath3 asymmetries in the decays of the lightest majorana neutrino @xmath79 differ from the standard case .
the most interesting situations occur when the @xmath4 lifetime and the @xmath3 asymmetry @xmath79 are controlled by two different sets of couplings , and are thus unrelated . in these cases
successful leptogenesis can be achieved even at a scale as low as a few tev , without conflicting with the requirement of @xmath4 out - of - equilibrium decays , and ensuring at the same time a reasonable value for the scale of the light neutrino masses .
in this section we present some analytic results for the minimal @xmath174 model with only two right - handed neutrinos @xmath175 and two pairs of vectorlike fermions @xmath176 and @xmath177 .
this is the minimal field content that ensures that two light neutrinos are massive , as is required by neutrino oscillations data . since in this case
one neutrino is exactly massless , the neutrino mass squared differences measured in oscillations experiments completely determine the light neutrinos masses , thus allowing for a certain level of predictivity . for a normal hierarchy spectrum we have @xmath178ev and @xmath179ev .
focusing on the case discussed in section [ sec : case1 ] , that is defined by the condition @xmath180 , the @xmath181 matrix of the effective couplings @xmath182 can be written as @xcite @xmath183 where r is a complex @xmath181 matrix satisfying r@xmath184r@xmath185 and r@xmath186r@xmath187diag@xmath188 @xcite , @xmath189 with @xmath190 the leptonic mixing matrix , and @xmath44 and @xmath191 are respectively the matrices of the heavy and light neutrinos mass eigenvalues . substituting ( [ eq : tildelam ] ) in the combination of couplings that determines the decay rate , eq .
( [ eq : width - decay - treel ] ) , we obtain @xmath192 then eq .
( [ eq : out - of - eq ] ) implies the following constraint on the ratio of the @xmath6 breaking scale and the lightest majorana mass @xmath0 : @xmath193\\ & \approx & 0.2\ , \left(|r_{13}|^2 + 0.17\ , |r_{12}|^2 \right).\end{aligned}\ ] ] this suggests that a mild hierarchy between @xmath45 and @xmath0 is certainly possible and , as was implicit in our scheme , allows for the possibility that the three scales @xmath43 , @xmath0 and @xmath45 can actually lie within a few orders of magnitude .
we thank y. nir for discussions and for pointing out an error in the first version of the paper .
we thank e. roulet and s. davidson for discussions .
d.a.s acknowledges m. hirsch for his help in some of the calculations .
d.a.s is supported by a spanish ph.d .
fellowship by m.c.y.t . the work of m.l . is supported in part by colciencias under contract number 1115 - 333 - 18739 .
the work of e.n . is supported in part by the istituto nazionale di fisica nucleare ( infn ) in italy , and by colciencias in colombia under contract 1115 - 333 - 18739 .
m. fukugita and t. yanagida , phys .
b * 174 * , 45 ( 1986 ) .
m. a. luty , phys .
d * 45 * , 455 ( 1992 ) .
p. minkowski , _ phys .
_ b * 67 * 421 ( 1977 ) ; t. yanagida , in _ proc . of workshop on unified theory and baryon number in the universe _ , eds . o. sawada and a. sugamoto , kek , tsukuba , ( 1979 ) p.95 ; m. gell - mann , p. ramond and r. slansky , in _ supergravity _
, eds p. van niewenhuizen and d. z. freedman ( north holland , amsterdam 1980 ) p.315 ; p. ramond , _ sanibel talk _
, retroprinted as hep - ph/9809459 ; s. l. glashow , in _ quarks and leptons _ , cargse lectures , eds m. lvy , ( plenum , 1980 , new york ) p. 707
; r. n. mohapatra and g. senjanovi , _ phys .
lett . _ * 44 * , 912 ( 1980 ) .
v. a. kuzmin , v. a. rubakov and m. e. shaposhnikov , phys .
b * 155 * , 36 ( 1985 ) .
w. buchmuller , p. di bari and m. plumacher , new j. phys .
* 6 * , 105 ( 2004 ) [ arxiv : hep - ph/0406014 ] ; annals phys . *
315 * , 305 ( 2005 ) [ arxiv : hep - ph/0401240 ] ; phys .
b * 547 * , 128 ( 2002 ) [ arxiv : hep - ph/0209301 ] ; nucl .
b * 643 * , 367 ( 2002 ) [ arxiv : hep - ph/0205349 ] ; w. buchmuller and m. plumacher , int .
j. mod .
a * 15 * , 5047 ( 2000 ) [ arxiv : hep - ph/0007176 ] ; phys .
rept . * 320 * , 329 ( 1999 ) [ arxiv : hep - ph/9904310 ] .
a. pilaftsis , phys .
lett . * 95 * , 081602 ( 2005 ) [ arxiv : hep - ph/0408103 ] ; a. pilaftsis and t. e. j. underwood , nucl .
b * 692 * , 303 ( 2004 ) [ arxiv : hep - ph/0309342 ] ; a. pilaftsis and t. e. j. underwood , phys .
d * 72 * , 113001 ( 2005 ) [ arxiv : hep - ph/0506107 ] .
t. hambye , y .
lin , a. notari , m. papucci and a. strumia , nucl .
b * 695 * , 169 ( 2004 ) [ arxiv : hep - ph/0312203 ] .
g. f. giudice , a. notari , m. raidal , a. riotto and a. strumia , nucl .
b * 685 * , 89 ( 2004 ) [ arxiv : hep - ph/0310123 ] .
w. buchmuller and m. plumacher , phys .
b * 511 * , 74 ( 2001 ) [ arxiv : hep - ph/0104189 ] ; e. nardi , y. nir , j. racker and e. roulet , jhep * 0601 * , 068 ( 2006 ) [ arxiv : hep - ph/0512052 ] .
r. barbieri , p. creminelli , a. strumia and n. tetradis , nucl .
b * 575 * , 61 ( 2000 ) ( for the updated version of this paper see [ arxiv : hep - ph/9911315 ] ) .
t. endoh , t. morozumi and z. h. xiong , prog .
* 111 * , 123 ( 2004 ) [ arxiv : hep - ph/0308276 ] ; t. fujihara , s. kaneko , s. kang , d. kimura , t. morozumi and m. tanimoto , phys .
d * 72 * , 016006 ( 2005 ) [ arxiv : hep - ph/0505076 ] .
a. abada , s. davidson , f. x. josse - michaux , m. losada and a. riotto , jcap * 0604 * ( 2006 ) 004 [ arxiv : hep - ph/0601083 ] ; a. abada , s. davidson , a. ibarra , f. x. josse - michaux , m. losada and a. riotto , jhep * 0609 * ( 2006 ) 010 arxiv : hep - ph/0605281 .
a. de simone and a. riotto , jcap * 0702 * , 005 ( 2007 ) [ arxiv : hep - ph/0611357 ]
. f. x. josse - michaux and a. abada , arxiv : hep - ph/0703084 . t. shindou and t. yamashita , arxiv : hep - ph/0703183 .
p. di bari , nucl .
b * 727 * , 318 ( 2005 ) [ arxiv : hep - ph/0502082 ] . o. vives , phys . rev .
d * 73 * , 073006 ( 2006 ) [ arxiv : hep - ph/0512160 ] .
g. engelhard , y. grossman , e. nardi and y. nir , arxiv : hep - ph/0612187 .
a. strumia , arxiv : hep - ph/0608347 ; e. nardi , arxiv : hep - ph/0702033 , arxiv:0706.0487 [ hep - ph ] ; y. nir , arxiv : hep - ph/0702199 ; m. c. chen , arxiv : hep - ph/0703087 ; s. davidson , arxiv:0705.1590 [ hep - ph ] .
t. hambye , nucl .
b * 633 * , 171 ( 2002 ) [ arxiv : hep - ph/0111089 ] . c. d. froggatt and h. b. nielsen , nucl .
b * 147 * , 277 ( 1979 ) .
+ [ -8pt ] d. n. spergel _ et al .
_ , arxiv : astro - ph/0603449 . g. steigman , int .
j. mod .
e * 15 * , 1 ( 2006 ) [ arxiv : astro - ph/0511534 ] .
j. a. casas and a. ibarra , nucl .
b * 618 * ( 2001 ) 171 [ arxiv : hep - ph/0103065 ] . | we study variations of the standard leptogenesis scenario that can arise if an additional mass scale related to the breaking of some new symmetry ( as for example a flavor or the b - l symmetry ) is present below the mass @xmath0 of the lightest right - handed majorana neutrino .
our scheme is inspired by @xmath1 models of flavor la froggatt - nielsen , and involves new vectorlike heavy fields @xmath2 .
we show that depending on the specific hierarchy between @xmath0 and the mass scale of the fields @xmath2 , qualitatively different realizations of leptogenesis can emerge .
we compute the @xmath3 asymmetries in @xmath4 decays in all the relevant cases , and we conclude that in most situations leptogenesis could be viable at scales much lower than in the standard scenario .
ific/07 - 24 + pi / uan - ft2007 - 172 + udea - pe-07/004 decays of heavy neutrinos , right - handed neutrinos , flavor symmetries 13.35.hb , 14.60.st , 11.30.hv |
the accurate knowledge of atomic masses is required for many areas of physics @xcite . in particular , the mass gives access to the nuclear binding energy which in turn reveals distinct features of nuclear structure , notably shell effects and their relative strength as a function of isospin . this paper reports on mass data in the vicinity of the doubly - magic @xmath0 nucleus .
strong nuclear structure effects are observed in particular at the _ neutron - deficient _ side from @xmath17 to @xmath18 .
they have been revealed by laser spectroscopic measurements of the optical isotope shift @xcite .
such measurements allow one to determine the changes of the root - mean - square charge radii @xmath19 relative to a reference nucleus .
characteristic irregularities were observed in the isotopic chains of iridium ( @xmath17 ) , platinum ( @xmath20 ) , gold ( @xmath21 ) , mercury ( @xmath22 ) , and thallium ( @xmath23 ) at approximately mid - shell ( @xmath24 ) .
the most prominent example is the odd - even shape staggering effect within the mercury isotopes @xcite .
these phenomena were first explained for mercury as two different minima in the potential energy surface at a weakly oblate and a stronger prolate deformation ( shape coexistence ) @xcite .
+ in addition , several rotational bands were observed in the nuclear excitation spectra of these nuclei .
an overview on nuclear shape coexistence of even - mass nuclides is given in @xcite .
in @xmath25 , even triple shape coexistence has been observed @xcite . here , the three lowest energy states having spin / parity @xmath26 correspond to spherical , oblate and prolate shapes , all within an energy spanning only @xmath27 .
these phenomena are described by particle - hole excitations of proton pairs across the closed shell @xcite .
the interaction of additional proton - hole pairs and valence neutrons will lead to a minimum in energy around mid - shell .
if the energies of the ground state and the intruder state are close , tiny changes in the nuclear binding energy are decisive for the nuclear ground state shape .
+ such minute nuclear binding energy effects in this mass range ( @xmath28 ) are only accessible by mass determinations with mass uncertainties @xmath29 of @xmath30 .
this has been demonstrated by isoltrap measurements on neutron - deficient mercury isotopes @xcite . in that work ,
isomeric states , a 54-kev level in @xmath31 and a 128-kev level in @xmath32 were even identified .
this was the first example of nuclear spectroscopy by mass spectrometry .
the present work complements the information on high - precision masses in this area of the chart of nuclei with data on bismuth , lead , and thallium isotopes .
the interconnection of the nuclides studied in this work by @xmath33-decay chains results in an improvement of further mass values up to @xmath34 .
+ in case no direct experimental data existed , previous mass values were obtained by a combination of up to four @xmath35 links with results from mass spectrometry . for example , the @xmath33-decay of neutron - deficient bismuth was studied at the lisol facility for odd @xcite and even isotopes @xcite .
all direct data being available in this region are either isoltrap measurements @xcite or mass determinations by the schottky technique at the experimental storage ring esr @xcite .
a detailed comparison to the latter data sets is made in sec .
[ esrcomp ] .
mass values of neutron - rich cesium isotopes were previously determined by triplet measurements with a mattauch - herzog spectrometer @xcite and @xcite .
+ a particular challenge while addressing the neutron - deficient nuclides experimentally is caused by the existence of two or even three isomeric states with both short half - lives and low excitation energies of sometimes less than @xmath36 .
this requires an extremely high mass resolution of up to @xmath37 .
+ masses on the _ neutron - rich _ side of the doubly magic @xmath0 nucleus are of great importance for predictions on the stability of superheavy elements , which are stabilized by shell corrections @xcite .
a variation of this correction by about @xmath38 leads to changes in the calculated half - lives for spontaneous fission or @xmath33 decay by several orders of magnitude .
since no experimental mass values are available in this region , model calculations of the shell correction can be verified at the adjacent shell closure of @xmath39 .
+ furthermore , accurate mass values are required in order to provide reliable reference masses for the calibration of mass data obtained in measurements at the experimental storage ring esr @xcite . in an esr mass determination on neutron - deficient nuclides
@xcite 117 reference masses were used in order to calibrate the 466 mass values studied in that experiment . in the neutron - rich region
the number of reference masses is particularly low .
the penning trap mass spectrometer isoltrap @xcite is installed at the on - line isotope separator isolde @xcite at cern . here
, mass - separated beams of short - lived nuclides are provided for different experiments at an energy of @xmath40 .
they are produced by bombarding a thick target with high - energy ( @xmath41 ) proton pulses ( @xmath42 protons / pulse ) which induce spallation , fission , or fragmentation reactions .
reaction products diffuse from the heated target into an ion source region . here
, they are ionized either by surface ionization , by electron impact in a hot plasma , or a resonant photo ionization in a resonance ionization laser ion source ( rilis ) @xcite .
+ in this experiment nuclides around the shell closure of @xmath43 were produced by proton - induced ( 1.4-gev ) spallation and fragmentation reactions on a @xmath44 target ( @xmath45 ) . for lead and bismuth a resonant laser ionization
was applied , whereas cesium , thallium , francium , and radium were surface - ionized .
finally , the ions were mass - selected with a resolving power of @xmath46 in the high - resolution separator ( hrs ) .
+ figure [ a_kap52_is_setup ] shows a schematic of the isoltrap mass spectrometer .
the radiofrequency quadrupole ( rfq ) cooler and buncher @xcite decelerates and accumulates the continuous 60-kev isolde beam and delivers a cooled , low - energy ion bunch for efficient injection into the preparation penning trap @xcite . here ,
a mass - selective buffer - gas cooling scheme @xcite is employed in order to separate isobars .
a resolving power , @xmath47 , of up to @xmath48 can be achieved in this process . in the precision penning trap
, the mass @xmath49 of a stored ion species with charge @xmath50 is determined via a measurement of its cyclotron frequency @xmath51 , where @xmath52 denotes the strength of the magnetic field .
a cyclotron resonance spectrum is obtained by the time - of - flight detection method @xcite : the extra radial kinetic energy resulting from the resonant excitation with an azimuthal quadrupolar rf - field at @xmath53 is detected by a reduction in the time of flight of the ejected ions towards a detector ( mcp5 in fig .
[ a_kap52_is_setup ] ) .
the resolving power is approximately given by the product of the cyclotron frequency and the observation time : @xmath54 @xcite .
a more detailed description of a typical measurement cycle at isoltrap can be found in a recent publication @xcite .
+ an illustrative example of a cyclotron resonance curve is shown in fig .
[ a_kap54_zyklres_133 ] . in such a resonance curve ,
the mean time of flight of the ions is displayed as a function of the applied radiofrequency , and a resonant excitation can be recognized by a shorter flight time .
the diagram comprises 16 frequency scans containing a total amount of 1820 detected ions .
the linewidth of the central resonance , as given by the fourier limit @xmath55 , results in a resolving power of @xmath56 ( mass resolution @xmath57 ) with an excitation time of the quadrupolar rf - field @xmath58 .
this is sufficiently high to resolve ground and isomeric states of @xmath59 .
the theoretical lineshape @xcite was used to fit the data points . in order to calibrate the strength of the magnetic field @xmath52
, the cyclotron frequency of a reference nuclide with well - known mass value is determined in regular time intervals .
this allows for the accurate mass determination of short - lived nuclides far from stability with relative mass uncertainties @xmath60 @xcite .
examples of recent measurements on @xmath61 @xcite as well as on several isotopes of potassium @xcite , nickel , copper , gallium @xcite , and krypton @xcite cover a wide mass range from @xmath62 .
the primary result of an isoltrap measurement is the cyclotron frequency ratio @xmath63 of a reference ion with well - known mass relative to the ion species to be determined .
thus , these experimental ratios are independent from the uncertainty of the reference mass @xmath64 . from the ratios @xmath65 , the atomic mass of the nuclide under study can be deduced by @xmath66 here , the atomic mass of the reference ion is indicated by @xmath67 and @xmath68 is the mass of the electron . a correction for the binding energy of the missing electron in the order of ev
can be neglected . + extensive studies using carbon cluster ions @xcite helped to improve the assessment of systematic effects .
for example , systematic errors like mass - dependent frequency shifts or the effect of a variation in the magnetic field strength were quantified .
in addition , these investigations result in a better estimate of the final uncertainty of the mass determination .
the residual systematic error of the isoltrap mass measurements , estimated previously to be @xmath69 , was determined in @xcite to be @xmath70 the data analysis was performed as described in @xcite .
some aspects that are essential in the analysis of the data presented here are summarized in the following .
simultaneous storage of ions of different mass values may result in a shift of the line center of the cyclotron frequency resonance .
such possible contaminations depend on the ion production process . in the case of measurements at isoltrap , isobars or nuclides in isomeric states
might be present in addition to the nuclide under investigation .
systematic experimental studies using 10 to 70 ions stored in the trap as well as simulations with two ions interacting by coulomb forces are discussed in @xcite and @xcite : if the two ions have different masses , the resulting frequency shifts depend on the relative distance of the line centers divided by the linewidth that is chosen during the measurement process .
if both lines can not be resolved , only _ one _ cyclotron resonance is observed at the frequency of the center of gravity . if the cyclotron resonances of both ion species can be resolved , both line centers are shifted to lower frequencies .
the size of the shift has been found to be proportional to the number of stored ions .
+ to be able to exclude the possible presence of ions of different mass in these measurements , the line center of the cyclotron resonance is analyzed with respect to the number of ions registered by mcp 5 ( see fig . [ a_kap52_is_setup ] ) in a so - called count - rate - class analysis .
the number of ions counted by the microchannel plate detector after each irradiation of the stored ion cloud with radiofrequency and ejection from the trap is subdivided into at least three count - rate classes with comparable average ion counts @xmath71 . a possible shift induced by a _ contamination _
is hence corrected by extrapolating the cyclotron frequency to a vanishing number of detected ions .
two examples of a count - rate - class analysis , both with the same average number of ions per count - rate class @xmath71 , are shown in fig .
[ a_kap54_z - class ] .
the upper plot shows a measurement done with stable @xmath72 ions from the off - line alkali ion source ( see fig .
[ a_kap52_is_setup ] ) , which delivers only surface - ionized alkali elements .
the resulting fit of the slope of @xmath73 shows no count - rate dependent effect .
the lower one shows a measurement of the radionuclide @xmath74 as delivered by isolde . with an increasing number of detected ions per cycle a significant slope of @xmath75 of the cyclotron frequency towards lower values
is observed .
this might be due to the presence of ions in the isomeric state .
see also the discussion on the nuclide @xmath74 in sec .
[ evena ] . combining several measurements ,
the final value of a frequency ratio is obtained from the weighted mean of individual results .
an ion species @xmath76 experiences a frequency shift which can originate for example from electric field imperfections or a possible misalignment of the magnetic and electric field axis versus each other @xcite .
since these shifts are to first order _ mass - independent _ , this results in a systematic shift in the determination of the frequency ratios , depending on the mass difference @xmath77 between the mass of interest _ m _ and the reference mass @xmath78 .
the upper limit of this effect has been previously estimated to @xmath79 @xcite .
frequency ratios between carbon clusters whose values are exactly known were quantified by a linear dependance @xcite , @xmath80 where @xmath81 is the difference between the measured and the true frequency ratio . in the experiments reported here ,
covering the mass region between @xmath82 and @xmath83 , the reference mass of @xmath72 ( @xmath84 ) differs by about 70 mass units from the mass to be determined . after applying the correction given by eq .
( [ 54 - 8 ] ) an uncertainty of the same size , @xmath85 is added quadratically to the uncertainty . due to the large mass difference @xmath77 , this yields an additional error on the order of 1 kev to the deduced mass excess value . in the end , the residual systematic uncertainty of @xmath86 ( eq . ( [ 54 - 4 ] ) )
is added quadratically to obtain the final uncertainty .
an unambiguous mass assignment to a specific state , ground state or isomer , is not possible if a non - resolved mixture of nuclei in their ground and isomeric states is present .
if the production ratio @xmath87 ( isomer to ground state ) , and the isomeric excitation energy @xmath88 are known , the ground state mass _ m _ can be extrapolated from the mass @xmath89 , corresponding to the observed cyclotron resonance , by @xcite @xmath90 if the production ratio of two ion species in ground and isomeric state is not known , the presence of the so - called contamination can be observed in the data analysis by studying the behavior of the resulting cyclotron frequency as a function of the count rate class ( see fig .
[ a_kap54_z - class ] ) . only in the rare case of very similar half - lives of the ion under investigation and contaminating ion
, contaminations might remain undiscovered .
+ table [ a_kap5_nubase ] lists all relevant data of the nuclides that have been studied : half - life @xmath91 , spin state @xmath92 and the isomeric excitation energy @xmath88 . in case
several isomeric states are known to exist , those having a half - life longer than @xmath93 are listed are not considered as contaminants in the studies reported here , since the typical storage time of @xmath94 exceeds by far @xmath91 . ] .
the assignment of a measured mass value to an isomeric state is performed under the assumption that isomeric states with a given @xmath92 and excitation energy might appear as listed in tab .
[ a_kap5_nubase ] .
the chosen rf - excitation time results in a width of the cyclotron resonance @xmath95 , which corresponds to a width @xmath96 in energy units .
this value is also called mass resolution and is listed in the last column , since it serves as an important criterion for the observation of a contamination and hence for the identification of the states .
lrrlccc + nuclide & & & spin / parity & excitation & @xmath97 + & & & @xmath92 & energy @xmath88 [ kev ] & [ kev ] + + nuclide & & & spin / parity & excitation & @xmath97 + & & & @xmath92 & energy @xmath88 [ kev ] & [ kev ] + @xmath98 & & 582 & ms & @xmath99 & & 220 + @xmath100 & & 225 & ms & ( @xmath99 ) & & 660 + @xmath101 & & 3.4 & s & @xmath102 # & & 340 + @xmath103 & & 2.7 & ms & @xmath104 # & 850(30 ) & 340 + @xmath105 & & 6.9 & s & @xmath102 # & & 340 + @xmath106 & & 60 & ms & @xmath104 # & 625(17 ) & 340 + @xmath74 & * & & 40 # & s & ( @xmath107 ) & & 210 + @xmath3 & * & & 27.5 & s & ( @xmath108 ) & 40(190 ) & 210 + @xmath109 & & 2.9 & s & ( @xmath110 ) & 420(190 ) & 210 + @xmath4 & & @xmath11151 & s & ( @xmath102 ) & & 540 + @xmath112 & & 15.6 & s & ( @xmath104 ) & 332(4 ) & 540 + @xmath113 & & 1.84 & h & @xmath107 & & 120 + @xmath5 & & 1.41 & h & ( @xmath108 ) & 394.2(5 ) & 120 + @xmath6 & & & @xmath102 & & 430 + @xmath114 & & 8 & min & @xmath115 & & 120 + @xmath7 & & 43 & min & @xmath116 & 319.3(7 ) & 120 + @xmath0 & & & @xmath117 & & 270 + @xmath118 & * & 6.3 & s & ( @xmath119 ) & & 560 + @xmath120 & * & 6.2 & s & ( @xmath110 ) & 150(190 ) & 560 + @xmath121 & & @xmath122 500 & ns & @xmath108 # & 420(190 ) & 560 + @xmath123 & & 12.3 & s & ( @xmath104 ) & & 110 + @xmath124 & & 128 & ms & ( @xmath102 ) & 242(7 ) & 110 + @xmath125 & & 34.6 & s & ( @xmath119 ) & & 110 + @xmath126 & & 39.6 & s & ( @xmath110 ) & 210(50 ) # & 110 + @xmath127 & & 67 & s & ( @xmath104 ) & & 110 + @xmath128 & & 3.2 & s & ( @xmath102 ) & 308(7 ) & 110 + @xmath129 & * & 95 & s & ( @xmath119 ) & & 580 + @xmath130 & * & 125 & s & ( @xmath131,@xmath108 ) & 100(70 ) # & 580 + @xmath132 & * & 115 & s & ( @xmath110 ) & 230(310 ) & 580 + @xmath59 & & 183 & s & @xmath104 # & & 120 + @xmath133 & & 87 & s & @xmath102 # & 399(6 ) & 120 + @xmath134 & & 5.1 & min & ( @xmath119 ) & & 590 + @xmath135 & & 0.6 & s & ( @xmath108 ) & 167(3 ) & 590 + @xmath136 & & 4 & min & ( @xmath110 ) & 270(4 ) & 590 + @xmath137 & & 9.3 & min & ( @xmath104 ) & & 600 + @xmath138 & & 5.04 & min & ( @xmath102 ) & 350(160 ) & 600 + @xmath139 & & 19 & ey & @xmath104 & & 270 + @xmath140 & & 7.6 & min & ( @xmath104 ) & & 470 + @xmath141 & & 36.4 & min & ( @xmath142 ) & 1347.5(2.5 ) & 470 + @xmath143 & & 2.17 & min & @xmath144 # & & 480 + @xmath145 & & 550 & ms & @xmath104 # & & 420 + @xmath146 & & 3.85 & s & ( @xmath104 ) & & 260 + @xmath147 & & 50.2 & s & @xmath102 # & & 810 + @xmath148 & & 2.46 & s & @xmath117 & & 470 + @xmath149 & & 4 & min & @xmath150 & & 810 + @xmath151 & & 93 & min & @xmath117 & & 330 + in the measurements reported here possible contaminations are mostly isomeric states , which could not be separated in the preparation penning trap .
exceptions are the heavy isobars @xmath147 and @xmath149
. the spallation process does not transfer large angular momentum and hence low - spin states are predominantly produced .
fragmentation reactions lead to a production of higher spin states .
nuclides in states with half - lives shorter than few 10 ms are suppressed due to the slow release time of the reaction products from the isolde target and the time needed for the isobaric separation in the preparation penning trap ( see fig .
[ a_kap52_is_setup ] ) . + a further selection of an atom in a specific nuclear state can be obtained by resonance laser ionization ( ris ) with the isolde laser ion source and a proper choice of laser wavelengths and bandwidths @xcite . due to different hyperfine splittings and isotope shifts in the ground and isomeric state
, ris can enhance the selectivity .
+ in the case of bismuth investigated in the present work , the lasers were used in the broad - band mode .
the production ratios in this experiment can not be deduced from previous measurements with the conventional ion source at isolde . for lead ,
the applied broad - band mode ( laser bandwidth = 10 ghz ) does not result in a selection of isomeric states , since the isomeric shift and hyperfine splitting are smaller or of the same size as the laser bandwidth .
table [ a_kap5_ratios ] lists all frequency ratios that were determined relative to the cyclotron frequency of @xmath72 ions . if possible , the assignment to a particular isomeric state is indicated and will be discussed in sec .
[ assign ] .
the last column lists all mass excess values deduced from the determined frequency ratios @xmath65 .
llcr + nuclide & frequency ratio & @xmath152 & mass excess + & @xmath153 & & [ kev ] + + nuclide & frequency ratio & @xmath152 & mass excess + & @xmath153 & & [ kev ] + @xmath98 & 1.090516444(91 ) & @xmath154 & -60052(11 ) + @xmath100 & 1.10562975(48 ) & @xmath155 & -52011(60 ) + @xmath101 & 1.361768475(82 ) & @xmath156 & -12802(10 ) + @xmath105 & 1.376786217(81 ) & @xmath157 & -16592(10 ) + @xmath3 & 1.399332234(69 ) & @xmath158 & -19874.4(8.6 ) + @xmath159 & 1.40683800(18 ) & @xmath160 & -22154(23 ) + @xmath5 & 1.474515617(98 ) & @xmath161 & -27103(12 ) + @xmath6 & 1.542259739(82 ) & @xmath162 & -23818(10 ) + @xmath7 & 1.482061386(45 ) & @xmath163 & -24429.5(5.5 ) + @xmath0 & 1.564849038(42 ) & @xmath164 & -21742.2(5.2 ) + @xmath165 & 1.42950438(20 ) & @xmath166 & -10535(25 ) + @xmath123 & 1.437006669(64 ) & @xmath167 & -13244.9(8.0 ) + @xmath126 & 1.444529605(72 ) & @xmath158 & -13398.5(9.0 ) + @xmath127 & 1.452033781(84 ) & @xmath168 & -15875(10 ) + @xmath130 & 1.45955793(12 ) & @xmath169 & -15878(50 ) is increased in the ame2003 due to a possible contamination by the @xmath170 and @xmath171 states . ]
+ @xmath59 & 1.467064775(45 ) & @xmath172 & -18023.7(5.6 ) + @xmath135 & 1.47459021(12 ) & @xmath173 & -17868(50 ) @xmath174 + @xmath175 & 1.482099666(89 ) & @xmath176 & -19706(25 ) in the ame2003 for possible contamination from @xmath138 . ]
+ @xmath139 & 1.572401384(35 ) & @xmath177 & -18254.9(4.3 ) + @xmath140 & 1.61770721(12 ) & @xmath178 & 1648(15 ) + @xmath143 & 1.625265520(93 ) & @xmath179 & 5873(11 ) + @xmath145 & 1.52741073(13 ) & @xmath180 & 861(16 ) + @xmath146 & 1.542441561(73 ) & @xmath181 & -1308.6(9.0 ) + @xmath147 & 1.72332168(30 ) & @xmath182 & 35816(37 ) + @xmath148 & 1.61017053(17 ) & @xmath183 & 100(20 ) + @xmath149 & 1.72329550(16 ) & @xmath184 & 32575(19 ) + @xmath151 & 1.73083537(10 ) & @xmath168 & 34518(12 ) +
the atomic mass evaluation ame is an adjustment procedure giving mass data for all those nuclides where relevant experimental information , like for example mass values or half - lives , are known .
such an evaluation was performed in december 2002 before these isoltrap data were included .
this atomic mass evaluation is referred to in the following as intermediate atomic mass evaluation 2002 ( iame2002 ) .
it is an up - to - date adjustment after the published issue of ame1995 @xcite and the nubase compilation @xcite .
the assignment of the observed cyclotron resonances to nuclear states was not straightforward at all since this area of the chart of nuclei is rich in low - lying isomers .
subsequently , the isoltrap results , reported in this publication , were included into the atomic mass evaluation 2003 .
a detailed description of the adjustment procedure is given in the particular issues of the ame @xcite . here ,
only some general features that are relevant for this work are described . as a result of the ame , adjusted mass values
are obtained from different types of experimental input data . even though a possible existence of more than 6000 different combinations of protons and neutrons is predicted , data from only approximately 2830 nuclides are available @xcite . if direct experimental data are not existing for nuclides near to the region of known masses , mass values are extrapolated from systematic , experimental trends .
all experimental results can be represented as a relation between two ( and more ) nuclides .
they span a net of interconnections between nuclides across the whole nuclear chart as it can be seen in the diagram for @xmath185 to @xmath186 ( fig .
[ connect ] ) . for the present isoltrap data giving a frequency ratio to a nuclide with precisely known atomic mass value
, these connections are linked to the mass of @xmath72 , which has been determined with a relative uncertainty of only @xmath187 @xcite . + all input data are transformed into a set of linear equations . for the case of experimental frequency ratios
this is explained in @xcite .
each datum @xmath188 with a standard deviation @xmath189 represents a link between two , three , or four unknown masses @xmath190 .
the over - determined set of _ x _ equations for _ y _ mass values @xmath191 is represented as a set of linear equations with _
y _ parameters and is finally solved by the least - squares method as described in @xcite .
most of the isoltrap frequency ratios that are published in the present work were introduced already into the most recently published atomic mass evaluation , _
ame2003 @xcite .
the assignment of a measured mass value to an isomeric state is performed under the assumption that isomeric states with a given @xmath92 and excitation energy might appear as listed in tab .
[ a_kap5_nubase ] . in this section
those odd - mass nuclides having a long - lived isomeric state are discussed . according to tab .
[ a_kap5_nubase ] the ground states of @xmath193 and @xmath114 have always a low spin value , whereas in the bismuth isotopes @xmath194 the ground state has a high spin . + the comparison of the results of tab .
[ a_kap5_ratios ] with those of the intermediate adjustment ( iame2002 ) @xcite , helps to indicate which spin states have most probably been produced during the experiment .
shown in fig .
[ odd ] is such a comparison of the isoltrap masses to the iame2002 masses with and without the added isomeric energy .
the zero line depicts the experimental isoltrap data with their uncertainties .
the differences to the literature values are indicated by open symbols , whereas the position of excited nuclear states is taken from the current nubase data ( tab .
[ a_kap5_nubase ] ) @xcite .
the uncertainty of the excited state is derived by adding the uncertainties of the ground state mass value and the excitation energy quadratically .
+ * @xmath101 * and * @xmath105 * : here , the half - lives of the excited isomeric states ( @xmath195 ) of @xmath196 for @xmath101 and @xmath197 for @xmath105 , are much shorter than the duration of a measurement cycle ( approximately @xmath198 ) .
therefore , the measured cyclotron frequency ratios are clearly assigned to the respective ground states .
the measurements of isoltrap agree with the values from the intermediate adjustment iame2002 ( compare fig . [ odd ] ) , but the uncertainties got significantly improved . + * @xmath4 * : the measurement of the nuclide @xmath4 has been performed with a mass resolution of @xmath199 , which is insufficient to resolve both ground and isomeric states , differing by @xmath200 .
hence , an eventually present admixture of either of the states can not be excluded .
the determined isoltrap value suggests a predominant production of the isomer . + * @xmath7 * : the mass values of both ground and isomeric states in @xmath114 are listed in iame2002 with uncertainties of only @xmath201 .
the value resulting from the isoltrap measurement with a final uncertainty of @xmath202 agrees within the error bars to the one of the excited @xmath203 state ( @xmath204 ) and hence is assigned to it . + * @xmath194 , @xmath140 * : a series of bismuth isotopes with odd mass numbers @xmath205 and @xmath206 has been studied . with the exception of the case of @xmath137 the mass resolution @xmath207 ( see tab . [ a_kap5_nubase ] ) is at least a factor of two smaller than the mass difference between both states .
therefore , a contamination of the respective second state is definitely excluded in the count - rate - class analysis . the remaining question which of the two states has been measured
can not be solved easily .
yield measurements for lead and bismuth isotopes at isolde reveal that states having higher spin values are predominantly produced @xcite .
these are for @xmath208 the @xmath209 ground states of bismuth .
indeed , the comparison of the isoltrap results shows a good agreement with the values of the intermediate evaluation iame2002 for the ground states of @xmath194 .
the measurements of @xmath137 were performed with a resolution of @xmath210 which is not sufficient to resolve ground and isomeric state , differing by @xmath211 .
measurements of the neighboring odd-@xmath192 isotopes @xmath127 and @xmath59 with approximately @xmath212 resolution did not show any hint of a contamination by an excited state .
a systematic comparison of individual measurements with low ( @xmath213 ) and high ( @xmath212 ) resolution was used in order to quantify a correction by a possible contamination by the excited @xmath138 state .
this correction of @xmath214 increases the uncertainty of the isoltrap value in the atomic mass table ame2003 from @xmath215 to @xmath216 . + considering the isomeric states of all odd - mass nuclides studied here a consistent picture is observed . except for cases where
the presence of a state can be excluded right away due to its short half - life , higher spin states are predominantly produced . in @xmath4 and @xmath114
these are the _ excited _ states with spins of @xmath217 and @xmath218 . in bismuth ( @xmath219 )
the ground states have a higher spin value of @xmath217 . for the nuclide @xmath140 with an excited state of @xmath220 energy and a spin @xmath221 ,
the @xmath217 ground state mass value has been determined . in order to facilitate a spin - to - state assignment ,
the present isoltrap data are plotted in fig .
[ even ] and compared relative to ground state mass values and excitation energies from the intermediate evaluation iame2002 .
these nuclides have sometimes _ two _ excited , long - lived states ( see also tab .
[ a_kap5_nubase ] ) .
excitation energies of states with half - lives longer than @xmath222 are shown .
+ apart from the @xmath222-state of @xmath118 all nuclides have half - lives that are sufficiently long ( @xmath223 ) in order to be studied with the penning trap mass spectrometer isoltrap .
due to the lack of any quantitative information on production ratios , a conclusion can be only based on a systematic comparison within the current data . in analogy to the nuclides of odd mass number , it is expected that lower spin states will be produced with lower yields .
+ * @xmath3 * : an unambiguous assignment of the isoltrap mass value to one of the long - lived states in @xmath74 is not straightforward .
this is also reflected by the different spin assignments given in nubase ( tab .
[ a_kap5_nubase ] ) and in the evaluated nuclear structure data files ( ensdf ) .
mass values of both excited states are given with uncertainties of @xmath224 .
the energy difference between @xmath3 and @xmath109 is @xmath225 @xcite .
the attempt to employ a very long rf - excitation of @xmath226 , corresponding to a mass resolution of @xmath227 , failed . for excitation times beyond @xmath228 ,
a decay of the stored ions has been observed ( see half - lives in tab .
[ a_kap5_nubase ] ) .
+ figure [ tl186 ] shows the results of the four cyclotron frequency measurements . with a mass resolution of @xmath229 ( measurement # 2 - 4 ) a mixture between @xmath3 and @xmath109
is resolved and the presence of either of the states should have been detected .
measurement # 2 and # 4 yield count - rate - class corrected frequencies .
the slope @xmath230 in [ mhz / ion ] of the linear extrapolation of the cyclotron frequency of @xmath231 is indicated for each measurement .
the third data point ( # 3 ) , which has been obtained after an additional waiting time of @xmath232 in the upper precision trap , does not show a significant count rate effect , since this measurement allowed for a decay of the @xmath233-state . to avoid any undetected contamination from the low - resolution measurement # 1 , it was not considered in the final weighted average of the frequency ratio . after a comparison to the spin systematics , as given in tab .
[ a_kap5_nubase ] , and the agreement of the isoltrap value with the intermediate evaluation iame2002 this result is assigned to the @xmath3 state .
it is not excluded that a small contamination by the ground state ( @xmath234 ) might be present .
+ * @xmath5 * : the mass resolution , being @xmath235 during this measurement , was sufficiently high to separate both states of @xmath113 .
the resulting isoltrap value agrees with the mass value of the excited , medium spin state @xmath5 ( @xmath236 , @xmath237 ) though the uncertainties of the iame2002 values are rather large . + * @xmath238 * : these nuclides have three long - living states similar to almost all even-@xmath192 bismuth isotopes up to @xmath239 .
the excitation energies of about @xmath240 are all known with uncertainties of around @xmath241 , except for @xmath134 ( fig . [ even ] ) .
the resolving power employed during those cyclotron frequency measurements of approximately @xmath242 corresponds to a mass resolution between @xmath243 and @xmath244 . in order to obtain a mass resolution of @xmath224 an excitation time of @xmath245
is required . + the mass value of isoltrap for @xmath134 agrees well to the one of the @xmath246-state ( @xmath135 ) .
however , a contamination can not be excluded .
accounting for a possible admixture of either the @xmath170 or the @xmath171 state the original uncertainty of @xmath247 had been increased to @xmath224 in the ame2003 .
+ due to the good agreement for @xmath134 , it is concluded that the medium spin state has been predominantly produced for @xmath129 as well .
since the possible presence of both isomeric states , @xmath170 and @xmath171 , whereas each of them is about @xmath36 different in energy , can not be excluded with a resolution of @xmath213 , the original error of @xmath247 is increased in the atomic mass evaluation 2003 to @xmath224 . + in @xmath118 the medium spin state exhibits a half - life in the nanoseconds regime .
+ the measured isoltrap value with @xmath216 uncertainty agrees within @xmath248 to the value of the @xmath249-state ( @xmath118 ) . due to the minor resolution during the measurement ,
a mixture of both states @xmath118 and @xmath120 differing by @xmath250 might be present .
+ _ * @xmath125 * _ : in contrast to other even-@xmath192 bismuth isotopes from @xmath118 to @xmath251 , no medium nuclear spin of , for example , @xmath252 or @xmath253 has been observed .
however , the general trend in this experiment points to a favored production of medium spin states , @xmath254 or 7 .
the question on the production ratio between the @xmath170 ground state and the excited @xmath171 state remains unsolved . from the general trend throughout all these measurements it is observed , that it is the higher spin state that is produced with a higher weight and measured in the trap .
consequently the mass value has been tentatively assigned to @xmath126 in the atomic mass evaluation 2003 . in the mass measurement of this nuclide
two observations are made : first , the counting class analysis of two individual measurements with a resolution of @xmath255 indicate the presence of a contamination and give a shifted , extrapolated value .
secondly , the characteristic pattern for an in - trap decay with short half - life is observed in the time - of - flight spectrum although the half - lives of both known states are more than ten times longer than the excitation time of @xmath256 .
this might indicate a third , so far unknown , isomeric state with shorter half - life or any other contamination .
the nuclides @xmath147 and @xmath149 were stored simultaneously in the precision trap since the applied mass resolving power @xmath257 in the preparation penning trap was not sufficient to separate the @xmath186-isobars .
figure [ fr229_ra229 ] shows the cyclotron resonance curves of the isobars , which differ in mass by @xmath258 . the center frequencies including the first sidebands are clearly separated with a resolving power of @xmath259 , using an excitation time of @xmath260 .
a superposition of two individual cyclotron resonance curves , weighed by their relative abundance , was fitted to the data points . assuming a detection efficiency of the microchannel plate of 30 %
, the average number of detected ions of 2.5 corresponds to about eight ions that were simultaneously stored .
therefore , the possibility of a frequency shift dependent on the number of ions ( see sec .
[ counti ] ) was investigated in a detailed count - rate - class analysis .
+ the result of such an analysis is plotted in fig .
[ z - class_229 ] for the cyclotron frequencies of @xmath147 ( left ) and @xmath149 ( right ) . here
, a total number of 2340 registered ions was subdivided into three ( upper diagrams ) as well as into four ( lower diagrams ) count - rate classes . _
different _ subdivisions were made in order to exclude any possible count - rate dependance produced within the statistics . in both cases no significant count - rate dependance is observed .
figure [ rest ] shows the mass excess values of other nuclides studied in this work , including the stable isotopes @xmath6 , @xmath0 , and @xmath139 .
the masses of these stable isotopes , which are known with high accuracy , agree within one standard deviation with iame2002 .
the largest discrepancies are observed for @xmath98 ( @xmath261 ) , @xmath100 ( @xmath262 ) , and @xmath149 ( @xmath263 ) .
some of the mass values which are reported in this work have also been determined by schottky mass spectrometry in the experimental storage ring esr at gsi @xcite .
figure [ trap - esr ] compares the results of both experiments and the values of the iame2002 .
a larger set of data is obtained by including also the indirectly determined results at esr using @xmath264 values from literature @xcite . in most of the cases
the data from both measurement techniques agree within their uncertainties .
an exception is the nuclide @xmath101 , where both values differ by more than @xmath265 .
since the esr value has been indirectly determined , this discrepancy could be attributed as well to the @xmath266 value .
table [ a_kap5_ma8_input ] compiles the mass excesses of those radionuclides discussed here and compares the data with values given in nubase 1997 @xcite and @xcite .
in addition , the results of the atomic mass evaluation published in 2003 ( ame2003 ) are given , which include the isoltrap results .
due to the manifold interconnections , the present isoltrap data improve not only the directly measured nuclides , but also a large region from @xmath267 to @xmath268 interconnected via @xmath33 chains in particular .
[ linked nuclides ] shows the respective section of the nuclear chart from tantalum to neptunium isotopes , indicating all related nuclides in this region .
+ for primary nuclides , which directly contribute to the adjustment procedure , the total influence of each input value in the ame is given in tab .
[ a_kap5_ma8_input ] by the _ significance _ in per cent .
it is the sum of all influences on individual mass values @xmath190 .
the quantity v / s denotes the difference between both values , divided by the uncertainty of the isoltrap value .
the last column marks those data which were not used in the adjustment procedure .
this is the case , if , for example , a contamination of the isoltrap measurement by a further isomeric state could not be excluded , or if , for the stable nuclides , another high - accuracy mass determination already exists , which is at least three times more precise than the isoltrap value .
+ in addition , the mass values of around 100 other nuclides changed after the atomic mass evaluation with the isoltrap data reported here .
the largest modifications in the adjusted mass values is observed for @xmath269 ( @xmath270 ) .
+ lrrrcrc + nuclide & & sign . &
v / s & remark + & isoltrap & nubase1997 and @xcite & ame2003 & % & & + + nuclide & & sign . & v / s & remark + & isoltrap & nubase1997 and @xcite & ame2003 & % & & + @xmath98 & -60052(11 ) & -60180(50 ) & -60057(11 ) & 94 & -0.4 & + @xmath100 & -52011(60 ) & -52290(150 ) & -52020(50 ) & 79 & -0.1 & + @xmath101 & -12802(10 ) & -12490#(300 # ) & -12801(9 ) & 92 & -0.2 & + @xmath105 & -16592(10 ) & -16290#(190 # ) & -16587(10 ) & 91 & 0.4 & + @xmath3 & -19874.4(8.6 ) & -19980(50 ) & -19874(9 ) & 100 & & + @xmath112 & -22154(23 ) & -22085(22 ) & -22109(8 ) & & 2 & f + @xmath271 & -27103(12 ) & -27470#(100 # ) & -27497(12 ) & 100 & & e + @xmath6 & -23818(10 ) & -23820.8(1.4 ) & -23820.6(1.3 ) & & -0.3 & u + @xmath7 & -24429.5(5.5 ) & -24423(15 ) & -24429(6 ) & 100 & & + @xmath0 & -21742.2(5.2 ) & -21749.6(1.3 ) & -21748.5(1.2 ) & & -1.2 & u + @xmath120 & -10535(25 ) & -10590(50 ) & -10483(10 ) & & 2.1 & f + @xmath123 & -13244.9(8.0 ) & -13211(23 ) & -13240(7 ) & 86 & 0.6 & + @xmath126 & -13398.5(9.0 ) & -13220#(120 # ) & -13399(9 ) & 100 & & + @xmath127 & -15875(10 ) & -15870(80 ) & -15873(10 ) & 100 & 0.2 & + @xmath130 & -15878(50 ) & -15970#(440 # ) & -15880(50 ) & 100 & & + @xmath59 & -18023.7(5.6 ) & -18039(21 ) & -18024(6 ) & 100 & & + @xmath135 & -17868(50 ) & -17856(30 ) & -17842(25 ) & 100 & 0.5 & + @xmath137 & -19706(25 ) & -19689(15 ) & -19688(8 ) & 100 & 0.8 & + @xmath139 & -18254.9(4.3 ) & -18259.1(1.5 ) & -18258.5(1.4 ) & & -0.8 & u + @xmath140 & 1648(15 ) & 1710(100 ) & 1649(15 ) & 100 & & + @xmath143 & 5873(11 ) & 5780#(100 # ) & 5874(11 ) & 100 & & + @xmath145 & 861(16 ) & 870(70 ) & 861(16 ) & 100 & & + @xmath146 & -1308.6(9.0 ) & -1310(50 ) & -1310(8 ) & 100 & -0.1 & + @xmath147 & 35816(37 ) & 35790#(360 # ) & 35820(40 ) & 100 & & + @xmath148 & 100(20 ) & 100(10 ) & 101(9 ) & 100 & 0 & + @xmath149 & 32575(19 ) & 32440(60 ) & 32563(19 ) & 91 & -0.6 & + @xmath151 & 34518(12 ) & 34550(30 ) & 34518(12 ) & 100 & & +
the mass values and the corresponding nuclear binding energies represent the underlying forces acting in a nucleus . in analogy to the ionization potential of atoms , plots of separation energies for the removal of two protons , @xmath272 or two neutrons , @xmath273 reveal the details of nuclear shell structure . in this representation , odd - even effects caused by the usually much larger binding energy for even proton or
neutron numbers are removed .
+ figures [ bbb ] and [ ddd ] show the behavior of separation energies for protons and neutrons versus the respective nucleon number .
the results from ame1995 and ame2003 are compared . a global trend , a continuous decrease in the separation energies , is observed .
discontinuities from this general trend appear at shell closures or in regions of strong deformations .
the @xmath274 shell closure is clearly visible in fig .
the much less pronounced strength of the proton shell closure @xmath275 is larger at the magic neutron number @xmath276 than at the half - filled shell @xmath24 ( fig .
[ ddd ] ) .
the main differences between the appearance of the mass surface in 1995 and 2003 are substantially reduced uncertainties , due to improved as well as numerous new experimental input data .
these are isoltrap measurements in the isotopic chains of mercury , platinum , lead , polonium , radon , and radium @xcite as well as a large number of new results obtained at the esr by the schottky technique @xcite . in particular the small uncertainty ( @xmath277 ) of data from isoltrap along the extended isotopic chain from @xmath278 to @xmath279
enables the observation of finer nuclear structure effects and to study , for example , nuclear shape coexistence and the influence of low - excitation intruder states @xcite .
note the pronounced irregularities in the thallium isotopic chain at neutron numbers @xmath280 and 114 .
+ the two - proton separation energies are much more regular and smooth in this region of the nuclear chart than those for two neutrons .
however , a peculiar irregularity in the two - proton separation energies appears for the isotopes of thallium ( @xmath281 ) with even neutron numbers @xmath282 from 98 to 108 : some separation energies are nearly equal or even higher than those of the isotopes with one additional neutron .
such an effect is very uncommon in the chart of nuclei . in order to investigate this anomalous behavior further
, the two - proton separation energies are plotted in fig .
[ stagg ] against the neutron number and the isotopes of an element are connected by lines .
this presentation shows very clearly in the region of the lead isotopes at @xmath43 the steadily increasing shell gap from mid - shell @xmath24 towards the magic neutron number @xmath274 .
also the above mentioned irregularities in the thallium and ( less pronounced ) in the mercury isotopes are revealed distinctly as an odd - even effect .
the reason for this effect is unclear .
three possible explanations are discussed in the following : + 1 .
the masses of @xmath101 and @xmath105 , determined with isoltrap cyclotron resonances and included in ame2003 , were wrongly assigned to the ground states of these isotopes .
however , an assignment to the isomeric states ( excitation energy @xmath283 or @xmath284 ) would increase the staggering even further . also the mass of @xmath4 is determined to more than 60 % by isoltrap , but in this case indirectly via the @xmath264 value of the @xmath33 decay and the directly measured mass of @xmath123 . also here , a possible problem in the assignment can be excluded due to the sufficiently high resolving power for separating ground and isomeric states and because of the short half - life of the isomer .
some mass measurements might be wrong .
this explanation is contradictory to the fact that the mass values in question stem from different sources and mass spectrometric techniques ( isoltrap , esr , @xmath264 values , etc . ) .
furthermore , the odd - even staggering is not an isolated phenomenon but appears also ( less pronounced ) in the neighboring isotopic chain of mercury .
this is not easily explicable since the mercury and the thallium isotopes are not interconnected by @xmath266 values .
the odd - even staggering of the two - proton separation energies in the thallium isotopic chain is factual and represents a nuclear structure effect .
this would imply a stronger pairing effect for neutron numbers around mid - shell ( @xmath24 ) than for nearly double - magic thallium nuclei with @xmath23 and @xmath285 .
similar to the drastic odd - even staggering observed in the charge radii of the mercury isotopes ( caused by shape transitions ) @xcite these phenomena depend very critically on the neutron and proton configurations and very elaborated calculations are required to reproduce these effects .
nuclear masses , as determined in this work , and nuclear charge radii , as obtained by optical spectroscopy from isotope shift measurements , illustrate global and collective properties of nuclear matter .
the mass is a measure of the binding energy and the charge radii reflect static and dynamic deformations . the nuclear system , as a function of protons and neutrons , maximizes the binding energy according to the nuclear shape .
the pairing energy , which is about @xmath38 , plays an important role in this context and becomes visible as odd - even effect in the masses as well as in the charge radii .
+ due to the complex interplay of single - particle and collective degrees of freedom , one observes in the region of the nuclear chart investigated in this work both a spherical and deformed shape in the same nucleus at low excitation energies .
this phenomenon , first observed at isolde / cern in the seventies by optical spectroscopy @xcite is known as `` shape coexistence or shape isomerism '' and nowadays an extensive bulk of data has been collected throughout the chart of nuclei @xcite .
shape coexistence is visible in the vicinities of the shell closures @xcite .
most clearly it is observed in the slope of charge radii in the mercury isotopic chain : it manifests itself in a huge odd - even shape staggering near the @xmath286 mid - shell region @xcite and a large difference in charge radii and therefore nuclear deformation between the ground and isomeric state in @xmath287 @xcite ( fig .
[ figrad3 ] ) .
since the nuclear states of very different configuration are ( almost ) degenerate in energy , one can not expect to observe any major trend in the slope of the masses or binding energies as a function of neutron number .
in fact , only now the high mass resolving power @xmath288 of up to @xmath37 of the penning trap mass spectrometer isoltrap enables one to resolve ground and isomeric states in this region of the nuclear chart where shape coexistence occurs .
the implementation of many new accurate mass data from isoltrap ( @xcite and this work ) to the data base of the atomic mass evaluation and a large data set of charge radii @xcite stimulated an attempt to reveil a correlation between masses and charge radii @xcite .
+ in order to uncover finer details in the binding energies regarding odd - even staggering and general trends @xcite , the absolute values of the deduced pairing gap parameters for neutrons are plotted instead of the experimental masses .
+ the three - point indicator for the shell gap parameter @xmath289 is given by @xmath290,\ ] ] which is often interpreted as a measure of the empirical pairing gap @xcite .
another commonly used relation is the four - point indicator , which averages the @xmath291 values and is given by @xmath292 .
\vspace*{-0.045cm}\ ] ] figures [ figrad1 ] - [ figrad5 ] compare the results from mass measurements with those from isotope shift determinations for the isotopic chains of platinum , gold , mercury , thallium , and lead .
for these elements , extensive data are available . in order to allow an easy comparison of pairing gap parameters with charge radii for the different elements ,
the scales of the x- and y - axis are made identical in figs .
[ figrad1 ] - [ figrad5 ] .
the pairing gap parameters of the isotopic chains of the even-@xmath293 elements lead , mercury and platinum are nearly identical in the regions of @xmath294 and @xmath295 , respectively .
the pairing gap parameters for the isotopes of the odd-@xmath293 elements thallium and gold show a similar behavior but the pairing gap is smaller in magnitude . around
mid - shell neutron number @xmath24 a minimum develops which is barely visible in the case of lead .
it becomes sharper and is shifted to @xmath296 for thallium , distinct and narrow for mercury but very much pronounced and broader for gold and platinum .
+ the slopes of the isotopic chains of lead , thallium and mercury are almost identical in the region where data are available for the same neutron number .
only the odd - even staggering is much more pronounced for the thallium isotopes as compared to those of lead and mercury . +
down to neutron number @xmath297 , the slope of the charge radii for the isotopes of gold and platinum is a little bit smaller than for the other isotopic chains which indicates increasing deformation . for @xmath296 in the case of mercury and @xmath298 for gold
a sudden change of the charge radii is observed which is due to a sharp change in deformation .
+ comparing the slopes of pairing parameters with those of the charge radii , one observes that the extremely sharp discontinuity at the shell closure at @xmath274 is visible in the charge radii only as a small kink in their slope , that the pronounced increase of deformation at mid - shell @xmath24 is reflected as a local minimum in the pairing parameters , and that the size of the odd - even staggering in the pairing parameter @xmath291 is not reflected in the size of the one of the charge radii .
there is an interesting observation : for thallium an unusual odd - even staggering between the isomers below mass number @xmath299 is observed . +
the new data obtained in this region of the nuclear chart near @xmath300 will hopefully stimulate theoretical work to develop nuclear models further which are able to describe simultaneously experimental mean square charge radii as well as experimental masses of spherical and deformed nuclei in a satisfactory way .
up to now , only little attention has been paid to such investigations which concern model - independent gross properties of nuclear matter .
for example , the charge radii are only poorly described by macroscopic - microscopic mass models which are , however , quite successful in describing nuclear masses @xcite .
even more demanding is a description of the odd - even staggering of masses and of the charge radii as attempted in ref .
@xcite for spherical nuclei .
the fine structure of the mass surface is important for the elucidation of the dramatic shape changes that appear so suddenly in this region of the nuclear chart . for most of the nuclides that were investigated in this work
a clear mass - to - state assignment was possible .
few exceptions due to an insufficient resolving power during the measurement are the nuclides @xmath4 , @xmath118 , @xmath301 , and @xmath137 .
spectroscopic data giving production ratios could assist these identifications using eq .
( [ 54 - 10 ] ) .
the excitation energies of bismuth isotopes ( odd - odd nuclei ) suffer from particulary high uncertainties .
hence , these cases should be addressed in a future mass measurement requiring the following conditions : ( i ) if the half - lives of the states are sufficiently long in order to allow an extremely high - resolution measurement : a mass resolution of , for example , @xmath302 can be obtained with an excitation time of @xmath232 . in these cases
the count - rate - class analysis can give a meaningful interpretation , _
i.e. _ a possible contamination is detected .
( ii ) similar to the measurements of both states of @xmath303 @xcite , a narrow - band laser ionization with a bandwidth of @xmath304 can be used to selectively ionize one of the states . as a prerequisite
, the element of interest has to be accessible for laser ionization and the hyperfine structure splitting of the isomeric components with different nuclear spin has to be known .
( iii ) in cases where the half - life is not sufficiently long to perform a measurement with extended excitation time , the mass determination should be supported by nuclear spectroscopy in order to enable a clear identification of states . + * acknowledgements * + we gratefully acknowledge the isolde technical group and the rilis group for assistance during the experiments .
this work was supported by the german ministry for education and research ( bmbf contracts 06mz962i , 06mz21s , and 06lm968 ) , by the european commission ( contracts hpri - ct-1998 - 00018 ( lsf ) and hpri - ct-2001 - 50034 ( nipnet ) ) , and by the association of helm- holtz research centers ( contract vh - ng-037 ) .
77 d. lunney , j.m .
pearson , and c. thibault , rev .
phys . * 75 * ( 2003 ) 1021 .
k. blaum , phys . rep . * 425 * ( 2006 ) 1 .
otten , treatise on heavy ion physics * 8 * ( 1989 ) 517 , and references therein .
kluge and w. nrtershuser , spectrochim .
acta b * 58 * ( 2003 ) 1031 , and references therein .
j. bonn _ et al .
_ , phys .
b * 38 * ( 1972 ) 308 .
p. dabkiewicz _ et al .
_ , phys .
lett . b * 82 * ( 1979 ) 199 . s. frauendorf and v.v .
pashkevich , phys .
* 55b * ( 1975 ) 365 .
wood _ et al .
. rep .
* 215 * ( 1992 ) 101 .
andreyev _ et al .
_ , nature * 405 * ( 2000 ) 430
. k. heyde _ et al .
. rep . * 102 * ( 1983 ) 291 .
k. heyde , nucl .
a. * 507 * ( 1990 ) 149c .
s. schwarz , ph.d .
thesis , university of mainz ( 1998 ) , isbn 3 - 8288 - 0735 - 6 .
s. schwarz _ et al .
_ , nucl . phys .
a * 693 * ( 2001 ) 533 .
r. fossion _ et al .
_ , nucl .
phys . a * 697 * ( 2002 ) 703 .
e. coenen _ et al .
_ , phys .
lett . * 54 * ( 1985 ) 1783 .
van duppen _ et al .
_ , nucl .
phys . a * 529 * ( 1991 ) 268 .
g. bollen _ et al .
_ , j. mod . opt .
* 39 * ( 1992 ) 257 .
f. herfurth _ et al .
_ , eur .
j. a * 25 * ( 2005 ) 17 .
t. radon _ et al .
_ , nucl .
phys . a * 677 * ( 2000 ) 75 .
litvinov , ph.d .
thesis , university of giessen ( 2003 ) .
litvinov _ et al .
_ , nucl .
phys . a * 756 * ( 2005 ) 3 .
m. epherre _
et al . _ ,
c * 19 * ( 1979 ) 1504 .
et al . _ ,
phys . a * 378 * ( 1982 ) 443 .
z. patyk _ et al .
_ , nucl .
phys . a * 491 * ( 1989 ) 267 . g. bollen _ et al . _ ,
nucl . instr .
meth . a * 368 * ( 1996 ) 675 .
f. herfurth _ et al .
_ , nucl . instr
. meth . a * 469 * ( 2001 ) 254 .
m. mukherjee _ et al .
_ , submitted to eur .
j. a. e. kugler , hyp .
int . * 129 * ( 2000 ) 23 .
mishin _ et al .
. instr .
meth . b * 73 * ( 1993 ) 550 .
u. kster _ et al .
_ , spectrochim .
acta * 58 * ( 2003 ) 1047
. h. raimbault - hartmann _ et al .
_ , nucl .
b * 126 * ( 1997 ) 378 . g. savard _ et al .
_ , phys .
lett . a * 158 * ( 1991 ) 247 .
g. grff , h. kalinowski , and j. traut , z. phys .
a * 297 * ( 1980 ) 35 .
m. knig _ et al .
_ , int .
j. mass .
spectrom .
ion process .
* 142 * ( 1995 ) 95 .
g. bollen , nucl .
phys . a * 693 * ( 2001 ) 3 .
g. sikler _ et al .
_ , nucl .
phys . a * 763 * ( 2005 ) 45 .
a. kellerbauer
_ et al . _ ,
j. d * 22 * ( 2003 ) 53 .
k. blaum _ et al .
_ , j. phys . b * 36 * ( 2003 ) 921 .
s. george _ et al .
_ , phys .
lett . * 98 * ( 2007 ) 162501 . c. yazidjian
_ et al . _ ,
c * 76 * ( 2007 ) 024308 . c. gunaut _ et al .
_ , phys . rev .
c * 75 * ( 2007 ) 044303 .
p. delahaye _ et al .
_ , phys .
c * 74 * ( 2006 ) 034331 .
k. blaum _ et al .
_ , eur .
j. a * 15 * ( 2002 ) 245 .
m. knig , diploma thesis , university of mainz ( 1991 ) .
g. bollen _ et al .
_ , phys .
c * 46 * ( 1992 ) r2140 .
d. beck _ et al .
_ , nucl .
meth . b * 126 * ( 1997 ) 374 .
g. audi _ et al . _ , nucl .
a * 624 * ( 1997 ) 1 .
g. audi , compilation of an intermediate atomic mass evaluation in 2002 .
this evaluation did not contain the isoltrap data presented in this work but all other mass data published up to this date .
iame is used for comparison with our mass values , since those are already included in the ame 2003 @xcite .
bradley _ et al .
_ , phys .
lett . * 83 * ( 1999 ) 4510 . g. audi and a.h .
wapstra , nucl .
phys . a * 595 * ( 1995 ) 409 .
g. audi and a.h .
wapstra , nucl .
a * 565 * ( 1993 ) 1 .
g. audi , a.h .
wapstra , and c. thibault , nucl .
phys . a * 729 * ( 2003 ) 337 .
g. audi , hyp .
132 * ( 2001 ) 7 .
wapstra , g. audi and c. thibault , nucl .
phys . a * 729 * ( 2003 ) 129 .
d. beck _ et al .
_ , eur . phys . j. a * 8 * ( 2000 ) 307 .
a. kohl , ph.d .
thesis , university of heidelberg ( 1999 ) .
r. julin _ et al .
_ , j. phys .
g * 27 * ( 2001 ) r109 .
g. ulm _ et al .
_ , z. phys .
a * 325 * ( 1986 ) 247 .
j. bonn _ et al .
_ , phys .
lett . b * 36 * ( 1971 ) 41
. h. de witte _ et al .
_ , phys .
* 98 * ( 2007 ) 112502 . c. weber _ et al .
j. a * 25 * supplem . 1 ( 2005 ) 201 .
w. satula _
et al . _ ,
* 81 * ( 1998 ) 3599 .
j. dobaczewski _
et al . _ ,
phys . rev . *
63 * ( 2001 ) 024308 .
f. buchinger _
et al . _ ,
c * 72 * ( 2005 ) 057305 .
fayans and d. zawischa , int . j. modern phys .
b * 15 * ( 2001 ) 1684 . c. weber _ et al . _ ,
lett . a * 347 * ( 2005 ) 81 . | accurate atomic mass measurements of neutron - deficient and neutron - rich nuclides around the doubly - magic @xmath0 and of neutron - rich cesium isotopes were performed with the penning trap mass spectrometer isoltrap at isolde / cern .
the masses of @xmath1 , @xmath2 , @xmath3 , @xmath4 , @xmath5 , @xmath6 , @xmath7 , @xmath0 , @xmath8 , @xmath9 , @xmath10 , and @xmath11 were determined .
the obtained relative mass uncertainty in the range of @xmath12 to @xmath13 is not only required for safe identification of isomeric states but also allows mapping the detailed structure of the mass surface . a mass adjustment procedure was carried out and the results included into the atomic mass evaluation .
the resulting separation energies are discussed and the mass spectrometric and laser spectroscopic data are examined for possible correlations . , , , , , , , , , and atomic mass , binding energy , penning trap , radionuclide , isomer , cesium , thallium , lead , bismuth , radium , francium + 07.75.+h mass spectrometers , 21.10.dr binding energies and masses , 27.70.+q @xmath14 , 27.80.+w @xmath15 , 27.90.+b @xmath16 |
amongst the many kinds of sources in the variable x - ray sky , x - ray transients have been observed since the first experiments in the late 60 s . according to their duration ,
it is possible to distinguish between long - duration x - ray transients ( lasting from weeks to few months ) and short - duration x - ray transients ( lasting from hours to very few days ) .
long - duration x - ray transients are mainly related to be - neutron star systems ( be x - ray transients ) and to soft x - ray transients ( a subclass of low - mass x - ray binaries ) .
the latter include the best black hole candidates found so far ( tanaka & shibazaki 1996 ) .
but not too much is known with respect to most of short - duration x - ray transients .
the main reason is the lack of counterparts at other wavelengths .
one subclass are the so - called fast x - ray transients , that have been observed with many detectors since the launching of the _ vela _ satellites ( heise et al .
1975 , cooke 1976 ) until the advent of the _ bepposax _ satellite ( heise et al . 1998 ) .
durations range from seconds ( belian , conner & evans 1976 ) to less than few hours .
normally , they have been seen once , and never seen in quiescence , implying high peak - to - quiescent flux ratios ( 10@xmath5 - 10@xmath6 ) ( ambruster et al .
spectral characteristics vary substantially , from a hard spectrum in mx 2346 - 65 ( kt @xmath0 20 kev ; rappaport et al . 1976 )
to soft spectra with blackbody temperatures from kt = 0.87 to 2.3 kev ( swank et al .
1977 ) . in two cases , precursors to the main event were observed by _
sas 3 _ ( hoffman et al . 1978 ) .
the precursors rose and felt in brighteness in less than 0.4 s. in the ariel v database , 27 sources were discovered ( pye & mchardy 1983 ) , and 10 more were detected in the _ heao 1 _
a-1 all sky survey ( ambruster & wood , 1986 ) , implying a fast transient all - sky event rate of @xmath7 for fluxes f@xmath1 @xmath8 3 @xmath2 10@xmath9 erg @xmath10 s@xmath4 in the 2 - 10 kev energy band . in the _ heao 1 _ a-2 survey ,
52 events were found ( connors 1988 ) , but 37 of them were related to four of the brightest x - ray sources in the lmc . due to the large difference of observational characteristics
, it seems that these events are caused by more than one physical mechanism . in several cases ,
there have been tentative optical identifications on the basis of known sources in the transient error boxes .
one suggestion has been that many of the fast x - ray transients are related with stellar flares originated in active coronal sources , like rs cvn binaries or dme - dke flare stars .
rs cvn systems are binaries formed by a cool giant / subgiant ( like a k0 iv ) with an active corona and a less massive companion ( a late g - dwarf ) in a close synchronous orbit , with typical periods of 1 - 14 d. peak luminosities are usually l@xmath1 @xmath0 10@xmath11 erg s@xmath4 .
the rs cvn system lds 131 was identified with the x - ray transient detected by _
heao 1 _ on 9 feb 1977 ( kaluzienski et al . 1978 , griffiths et al . 1979 ) .
the highest peak luminosity was recorded by _
ariel v _ for the flare observed from dm uma in 1975 ( pye & mc hardy , 1983 ) . the hardest flare yet observed was for the system hr 1099 , on 17 feb 1980 , which was detected by _
heao 2 _ at energies up to 20 kev ( agrawal & vaidya 1988 ) . the most energetic x - ray flare was observed by _ ginga _ from ux ari .
its decay time was quite long ( @xmath0 0.5 days ) .
[ cols="^,^,^,^,^,^,^,^ " , ] x - ray flares can be also observed from m or k dwarfs with balmer lines in emission ( these are the active cool dwarf stars dme - dke ) . in the _
ariel v _ sky survey , rao & vahia ( 1984 ) suggested seven dme stars as responsibles of x - ray flares that reached peak luminosities l@xmath1
@xmath12 10@xmath11 erg s@xmath4 ( 2 - 18 kev ) . at mic
is the dme star with the largest number of recorded events ( four ) , with a big flare in 1977 ( kahn 1979 ) reaching a peak luminosity l@xmath1 = 1.6 @xmath2 10@xmath13 erg s@xmath4 ( 2 - 18 kev ) , which is @xmath0 100 times larger that the strongest solar flares .
one of the most energetic flares was the flare observed by _
exosat _ in yy gem on 15 nov 1984 .
pallavicini et al .
( 1990 ) estimated a total energy flare e@xmath1 = 10@xmath3 erg ( 0.005 - 2 kev ) , and a decay time t@xmath14 = 65 min ( one of the longest decay times ever mesured for such events ) .
one of the longest flare ever reported was observed for more than 2 hours for the the x - ray source exo 040830 - 7134.7 ( van der woerd et al .
1989 ) .
x - ray flares have been observed from algol - type binaries ( schnopper et al .
1976 , favata 1998 ) , w uma systems and young stellar objects ( ysos ) .
most of the ysos are deeply embedded young stars and t tauri stars .
t tauri stars are pre - main sequence stars ( ages 10@xmath15 - 10@xmath16 yr ) that may exhibit x - ray flaring activity via thermal bremsstrahlung from a hot plasma ( see linsky 1991 ) .
montmerle et al .
( 1983 ) reported a `` superflare '' in the t tauri star rox-20 , in the @xmath17 oph cloud complex , which reached a peak luminosity l@xmath1 = 1.1 @xmath2 10@xmath11 erg s@xmath4 ( 0.3 - 2.5 kev ) , and an integrated flare x - ray energy e = 10@xmath3 erg . an _ asca _
observation of a flare in v773 tau implied l@xmath1 @xmath0 10@xmath18 erg s@xmath4 ( 0.7 - 10 kev ) , and a total energy release of @xmath0 10@xmath19 erg , which is among the highest x - ray luminosities observed for t tau stars ( tsuboi et al .
recently , x - ray activity in protostars has been detected .
these are objects closely related to t tau stars , with ages 10@xmath20 - 10@xmath21 yr , which are known to be strong x - ray sources when they enter the t tauri phase .
see montmerle et al .
( 1986 ) and montmerle & casanova ( 1996 ) for comprehensive reviews .
koyama et al .
( 1996 ) reported a extremely high temperature ( kt @xmath0 7 kev ) _ quiescent _ x - ray emission from a cluster of protostars detected by _
asca _ in the r cra molecular cloud .
preibisch ( 1998 ) also reported a _
observation of the source ec 95 within the serpens star forming region , for which a _ quiescent _ ( dereddened ) soft x - ray luminosity of @xmath0 ( 6 - 18 ) @xmath2 10@xmath11 erg / s is derived , which is at least 60 times larger than that observed for other 7 ysos .
two of these ysos have displayed flaring activity .
in particular , irs 43 in the @xmath17 oph cloud showed an extremely energetic superflare with a _
luminosity of l@xmath1 = 6 @xmath2 10@xmath18 erg s@xmath4 ( 0.1 - 2.4 kev ) ( grosso et al . 1997 ) .
theoreticians have suggested other sources like dwarf novae ( stern , agrawal & riegler 1981 ) or bizarre type i x - ray bursts ( lewin & joss 1981 ) as origin of some fast x - ray transients .
the watch instrument is the all - sky x - ray monitor onboard the _ granat _ satellite , launched on 1 dec 1989 .
it is based on the rotation modulation collimator principle ( lund 1985 ) .
the two energy ranges are approximately 8 - 15 and 15 - 100 kev . usually , the uncertainty for the location of a new and short - duration source is 1@xmath22 error radius ( 3-@xmath23 ) .
more details can be found in castro - tirado ( 1994 ) and brandt ( 1994 ) . during 1990 - 1992
, watch discovered three bright fast x - ray transients .
a small quantity when we compare with the other instruments mentioned above .
this is mainly due to the higher low energy cut - off for watch ( @xmath0 8 kev ) , implying that only the harder events are detected .
the events were immediately noticeable as an increase in the low energy count rate ( fig .
1 ) lasting up to a few hours .
no positive detections were made for the higher energy band .
their observational characteristics are summarized on table 1 .
three objects were also discovered by watch during 1990 - 92 , although it is likely that some others are present in the whole watch data base ( 1990 - 96 ) .
the sources reported here were observed to peak at x - ray fluxes up to 3 @xmath2 10@xmath24 erg @xmath10 s@xmath4 in the 8 - 15 kev band , and were discovered by analysis of the corresponding modulation pattern data .
their observational characteristics are summarized on table 2 . a fourth event , _ grs 1133 + 54 _ , lasting for @xmath0 1 day on 19 - 20 nov 1992 was reported by lapshov et al .
( 1992 ) .
lccccccc source & date & u.t .
( onset ) & duration ( _ days _ ) & @xmath25(2000.0 ) & @xmath26(2000.0 ) & b & f@xmath27(erg @xmath10 s@xmath4 ) + grs 0817 - 524 & 10 oct 1990 & 14:00 & @xmath0 1 & 08h 19m.4 & [email protected] & [email protected] & 3.1 @xmath2 10@xmath24 + grs
1133 + 540@xmath28 & 19 nov 1992 & 12:00 & @xmath0 1 & 11h 35m.7 & + [email protected] & + [email protected] & 2.2 @xmath2 10@xmath24 + grs 1148 - 665 & 10 oct 1990 & 12:00 & @xmath0 1 & 11h 50m.4 & [email protected] & [email protected] & 3.1 @xmath2 10@xmath24 + grs 1624 - 375 & 24 sep 1992 & 03:00 & @xmath0 1 & 16h 27m.7 & [email protected] & + [email protected] & 2.2 @xmath2 10@xmath24 + + we have searched for x - ray flares in the watch data base that could have occurred in simultaneity to observed flares at other wavelengths reported elsewhere .
the results were negative both for @xmath29 eridani ( a flare was observed by _
rosat _ on 21 feb 1991 , smith et al .
1993 ) and au mic ( a flare was observed by _
euve _ on 15 july 1992 , cully et al . 1993 ) . in the case of ev lac ,
no watch data were available for an extraordinary optical flare that occurred on 15 sep 1991 ( gershberg et al .
the three events quoted on table 1 imply a rate of @xmath0 5 year@xmath4 . if the sources of the three events are galactic , their high lattitude would suggest that they are close to us , although no nearby known flare stars ( pettersen 1991 ) have been found in the corresponding watch error boxes . _
grs 1100 - 771 .
_ only four variable stars are catalogued in the watch error box : cs cha , ct cha , cr cha and tw cha .
all are orion - type variables .
tw cha undergoes rapid light changes with typical amplitudes of 1@xmath30 .
the whole error box lies in the chamaeleon i star - forming cloud , and the position of 41 ( out of @xmath0 75 ) soft x - ray sources detected by _
rosat _ in the cloud ( alcal et al .
1995 ) , are compatible with the position for grs 1100 - 771 derived from the watch data .
the denis near - ir survey of the region has revealed 170 objects in the cloud , mostly t tau stars ( cambrsy et al .
feigelson et al .
( 1993 ) also found that @xmath080 % of the x - ray sources are identified with t tau stars with x - ray luminosities ranging from 6 @xmath2 10@xmath31 to 2 @xmath2 10@xmath13 erg s@xmath4 .
the watch detection implies a _ flaring _ luminosity l@xmath1 = 0.6 @xmath2 10@xmath32 ( d/1 pc)@xmath5 erg s@xmath4 ( 8 - 15 kev ) , or l@xmath1 = 1.2 @xmath2 10@xmath32 ( d/ 1 pc)@xmath5 erg s@xmath4 ( 0.1 - 2.4 kev ) for a thermal spectrum with kt = 10 kev and no absorption ( greiner 1999 ) .
if the x ray source lies in the chamaeleon i cloud ( at @xmath0 150 pc ) , l@xmath1 = 1.35 @xmath2 10@xmath3 erg s@xmath4 ( 8 - 15 kev ) , or l@xmath1 = 2.6 @xmath2 10@xmath3 erg s@xmath4 ( 0.1 - 2.4 kev ) . with an integrated flare energy of @xmath0 10@xmath33 erg across the x - ray band ( 0.1 - 10 kev ) . taking into account that kt @xmath0 10 kev , a value higher than that observed in the t tau star v773 ( tsuboi et al .
1998 ) , and that the energy release would be @xmath0 100 times larger than that seen in the yso irs 15 in the @xmath17 oph cloud ( grosso et al . 1997 ) , we tentatively suggest a superflare arising from one of the ysos in the chamaeleon i star - forming cloud as the most likely candidate for grs 1100 - 771 .
if this is indeed the case , it will be difficult to explain how this exceptionally high x
ray luminosity can be explained by a solar - like coronal emission mechanism .
however , a large amount of energy , stored in large magnetic structures that it is set free by magnetic reconnection events ( see grosso et al .
1997 ) , can account for the watch detection .
but we notice that the highest quiescent luminosity for any of the x - ray sources detected by _
rosat _ in the watch error box is l@xmath1 @xmath0 10@xmath13 erg s@xmath4 ( 0.1 - 2.4 kev ) at the utmost ( alcal et al .
1997 ) , i.e. 100 times lower than that seen in irs 43 for which a steady supply of energy ( like a large number of reconnection events ) was proposed ( preibisch 1998 ) . in any case , the possibility of grs 1100 - 771 being an isolated neutron star unrelated to the cloud can not be excluded , given the relatively large error box provided by watch .
_ grs 2037 - 404 .
_ it was discovered on 23 sep 1992 ( castro - tirado , brandt & lund 1992 ) , and lasted for 110 min , reaching a peak intensity of @xmath0 0.8 crab . on our request ,
a schmidt plate was taken at la silla on 29 sep , and based on this plate , it was reported the presence of the mira star u mic near maximum brightness at 7.0 mag ( della valle & pasquini , 1992 ) .
u mic is a mira star and we do not consider this star as a candidate due to the different timescales of the involved physical processes .
moreover it is too far away from the error box centre ( @xmath0 2@xmath22 ) .
hudec ( 1992 ) proposed another variable object inside the watch error box as a candidate to be further investigated .
this is the star rv mic ( b @xmath0 10.5 ) discovered in 1948 ( hoffmeister 1963 ) .
although is classified as a mira cet type star , it is apparently a highly variable object .
another four variables are within the 1@xmath22 radius watch error box : uu mic ( a rr lyr type ) , uw mic ( a semi - regular pulsating star ) , sz mic ( an eclipsing binary ) and uz mic ( an eclipsing binary of w uma type ) . in any case , we carefully examined the schmidt plate and found no object inside the 1@xmath22 radius watch error box that would have varied by more then 0.5@xmath30 when comparing with the corresponding eso sky survey plate .
_ grs 2220 - 150 . _
no variable stars are catalogued within the watch error box .
the high flux ( 2.1 crab ) is similar to the peak of the fast transient at 20h14 m + 30.9 discovered by _
oso-8 _ in 1977 ( selermitsos , burner & swank 1979 ) .
the peak luminosity for grs 2220 - 150 was l@xmath1 = 1.5 @xmath2 10@xmath32 ( d/1 pc)@xmath5 erg s@xmath4 ( 8 - 15 kev ) , or l@xmath1 = 6 @xmath2 10@xmath32 ( d/1 pc)@xmath5 erg s@xmath4 ( 2 - 15 kev ) if we assume a crab - like spectrum .
considering the high peak luminosities , a flare of a dme or a dke star may be excluded in these two latter cases .
no nearby rs cvn stars from the list of lang ( 1992 ) were found in the error boxes .
it can be possible that these events would be associated to old isolated neutron stars accreting interstellar matter with unstable nuclear burning ( ergma & tutukov 1980 ) . according to zduenek et al .
( 1992 ) , when the accretion rate is higher than 10@xmath34 g s@xmath4 , the hydrogen burning triggered by electron capture becomes unstable .
as the mass of the accreted envelope should be 10@xmath35 g , several hundred years will be required for accreting this ammount of matter . assuming that the number of isolated neutron stars with such high rate is 10@xmath15 ( blaes & madau 1993 )
, one should expect 10 - 100 fast x - ray transients per year . with the exception of grs 1133 + 54 ,
the other three sources quoted in table 2 are concentrated near the galactic plane , suggesting that they could be more distant than the three fast x - ray transients mentioned above . in this case , no nearby dwarf star or rs cvn systems have been identified within the corresponding error boxes . _ grs 0817 - 524 . _
no variable stars are catalogued in the error box .
however , we note the presence of an x - ray source detected by _ rosat _ : 1rxs j081938.3 - 520421 is listed in the all - sky bright source catalogue ( voges et al .
1996 ) , with coordinates r.a.(2000 ) = 8h19m38.3s , dec(2000 ) = -52@xmath22 04 21 " .
_ grs 1148 - 665 .
_ two variable stars lie within the error box : cy mus ( a rr lyr type ) and tw mus ( an eclipsing binary of w uma type ) .
we also note here the existence of a _ rosat _ all - sky bright source : 1rxs j115222.9 - 673815 , at r.a.(2000 ) = 11h52m22.9s , dec(2000 ) = -67@xmath22 38 15 " .
_ grs 1624 - 375 .
_ no variable stars are reported within the error box .
a 300-s optical spectrum taken in march 1997 at the 2.2-m eso telescope of the possible candidate suggested by castro - tirado ( 1994 ) reveals that this is a m - star unrelated to grs 1624 - 375 ( lehmann 1998 ) .
however there are three sources in the _ rosat _ catalogue : 1rxs j162730.0 - 374929 , 1rxs j162751.2 - 371923 and 1rxs j163137.0 - 380439 .
taking into account the _ a priori _ probability of finding such bright x - ray sources in the typical watch error boxes ( @xmath0 1.5 sources per error box ) , we conclude that none of the 1rxs sources is related to the longer duration x - ray transients detected by watch . in the _
ariel v _ catalogue of fast x - ray transients , pye & mchardy ( 1983 ) described 10 x - ray transients with durations ranging 0.5 - 4 days .
some of them were associated with known be - neutron stars systems , like 4u 0114 + 65 , and other with rs cnv systems , like @xmath23 cen or dm uma ( for which a flare lasting 1.5 days was observed by heao-2 ) .
the _ rossixte _
satellite has recently detected other two x - ray transients : xte j0421 + 560 and xte j2123 - 058 .
xte j0421 + 560 lasted for @xmath0 2 days ( smith et al .
1998 ) and it is probably related to a black hole in a binary system .
xte j2123 - 058 lasted for @xmath0 5 days and is presumably related to a low - mass binary in which a neutron star undergoes type - i bursts ( levine et al .
1998 , takeshima and strohmayer 1998 ) .
another short - duration event , lasting @xmath0 2 days has been observed in x - rays and radio wavelengths in the superluminal galactic transient grs 1915 + 105 ( waltman et al .
we consider that the three long duration events reported in this paper , with fluxes in the range 0.350.5 crab assuming a crab - like spectrum , are likely to have originated from compact objects in low - mass or high - mass x - ray binaries , similarly to xte j0421 + 560 and xte j2123 - 058 .
amongst the 3 fast x - ray transients discovered by watch in 1990 - 92 , grs 1100 - 771 might be related to a superflare arising from a yso in the chamaeleon i cloud , that would make -to our knowledge- the strongest flare ever seen in such an object . but the possibility of grs 1100 - 771 being an isolated neutron star unrelated to the cloud can not be excluded , given the relatively large error box provided by watch . amongst the other 3 longer duration x - ray transients , none of them seem to be related to known objects , and we suggest that the latter are likely to have originated from compact objects in low - mass or high - mass x - ray binaries .
we would like to thank the referee , thirry montmerle , for his very constructive comments .
we are grateful to j. greiner for useful discussions and for providing the expected fluxes in the 0.1 - 2.4 kev by means of the exsas package .
we also thanks r. hudec for valuable comments , and g. pizarro , a. smette and r. west for the schmidt plates taken at eso ' s la silla observatory , and to i. lehmann for the optical spectrum taken at the 2.2-m eso telescope in the grs 1624 - 375 field .
we are indebted to the staff of the evpatoria ground station in ukraine and those of the lavotchkin and babakin space center in moscow , where _ granat _ was built .
this search made use of the simbad database , at the centre de donnes astronomiques de strasbourg .
this work has been partially supported by spanish cicyt grant esp95 - 0389-c02 - 02 .
agrawal , p. c. & vaidya , g.r .
1988 , mnras 219 , 777 alcal , j. m. et al .
1995 , a&as 114 , 109 alcal , j. m. et al .
1997 , a&as 319 , 184 ambruster , c. , wood , k. s. , meekins , j. f. , et al . 1983 ,
apj 269 , 779 ambruster , c. w. & wood , k. s. 1986 , apj 311 , 258 belian , r. d. , conner , j. p. & evans , w. d. 1976 , apj 207 , l33 blaes , o. & madau , p. 1993 , apj 403 , 590 cambrsy , l. et al .
1998 , a&a 338 , 977 castro - tirado , a. j. , brandt , s. & lund , n. 1992 , iau circ .
5617 castro - tirado , a. j. 1994 , ph .
d. thesis , univ . of copenhagen connors ,
a. 1988 , ph .
d. thesis , univ . of maryland cooke ,
b. a. 1976 , nat 261 , 564 cully , s. c. , siegmund , o. h. w. , vedder , p. w. & vallerga , j. v. 1993 , apj 414 , l49 della valle , m. & pasquini , l. 1992 , priv .
ergma , e. & tutukov 1980 , a&a 84 , 125 favata , f. 1998 , nucl .
b 69/1 - 3 , 23 feigelson , e : d. , casanova , s. , montmerle , t. & guibert , j. 1993 , apj 416 , 623 .
gershberg , r. a. , et al .
1992 , priv .
greiner , j. 1999 , priv .
griffiths , r. e. , et al .
1979 , in : baity , w. , peterson , l. ( eds . ) , x - ray astronomy ( cospar ) , pergamon , new york , p. 93
grosso , n , montmerle , t. , feigelson , e. et al .
1997 , nat 387 , 56 heise , j. , brinkman , a. c. , schrijver , j. et al . 1975 ,
apj 202 , l73 heise , j. , in ' t zand , j. , ricci , r. , d ' andreta , g. & muller , j. 1998 , iau circ . 6892 .
hoffman , j. a. , lewin , w. h. g. , doty , j. , jernigan , g. , haney , m. & richardson , j. a. 1978 , apj 221 , l57 hoffmeister , c. 1963 , vernder , sterne sonneberg 6 , heft 2 .
hudec , r. 1992 , priv .
kaluzienski , l. j. , et al .
1978 , iau circ .
3714 koyama , k. et al .
1996 , pasp 48 , l87 lang , k.r .
1992 , in : astrophysical data , springer verlag , berlin levine , a. , swank , j. & smith , e. 1998 , iauc 6955 lapshov , i. , sazonov , s. & sunyaev , r. 1992 , priv .
lehmann , i. 1998 , priv .
lewin , w. h. g. & joss , p. c. 1981 , space sci .
28 , 3 linsky , j. l. 1991 , mem . s. a. it .
62(2 ) , 307 lund , n. 1985 , in : culhane , j. l. ( ed . ) , x - ray instrumentation in astronomy , spie 597 , 95 montmerle , t. , koch - miramond , l. , falgarone , e. , & grindlay , j.e . 1983 ,
apj 269 , 182 montmerle , t. , feigelson , e. , bouvier , j. & andr , ph .
1986 , in : protostars and planets iii , univ . of arizona press , p. 689 montmerle , t. & casanova , s. 1996 , in : magnetodynamic phenomena in the solar atmosphere - prototypes of stellar magnetic activity , eds : y. uchida et al . ,
kluwer , p. 247 pallavicini , r. , tagliaferri , c. & stella , l. 1990 , a&a 228 , 403 pettersen , b. r. 1991 , mem .
62(2 ) , 217 preibisch , t. 1998 , a&a 338 , l25 pye , j. p. & mchardy , i. m. 1983 , mnras 205 , 875 rao , a.r . , & vahia , m.n .
1987 , a&a 188 , 109 rappaport , s. , buff , j. , clark , g. , et al .
1976 , apj 206 , l139 schnopper , h. w. , delvaille , j. p. , epstein , a. et al . 1976 ,
apj 210 , l75 selermitsos , p. , burner , a. , & swank , j. h. 1979 , bull . of the aps
24 , 584 smith , d. , remillard , r. , swank , j. , takeshima , t. & smith , e. 1998 , iau circ .
6855 smith , m. a. , grady , c. a. , peters , g. j. & feigelson , e. 1993 , apj 409 , l49 stern , r. a. , agrawal , p. c. & riegler , g. r. 1981 , nat 290 , 573 swank , j. h. , becker , r. h. , boldt , e. a. , et al .
1977 , apj 212 , l73 takeshima , t. and strohmayer , t. e. 1998 , iauc 6958 tanaka , y. & shibazaki , n. 1996 , ara&a 34 , 607 m.g . , a&a 341 , l79 tsuboi , y. , koyama , k. , murakami , h. et al .
apj 503 , 894 van der woerd , h. , tagliaferri , g. , thomas , h.c .
& beuermann , k. , 1989 , a&a 220 , 221 waltman , b. et al .
1998 , iauc 6930 voges , w. et al .
1996 , iauc 6420 ( see also the mpe www page ) .
zduenek et al .
1992 , apj 384 , 129 | during 1990 - 92 , the watch all - sky x - ray monitor on _ granat _ discovered six short - duration x - ray transients . in this paper
we discuss their possible relationship to peculiar stars .
only one of the fast ( few hours ) x - ray transients ( grs 1100 - 771 ) might be tentatively ascribed to a superflare arising from a young stellar object in the chamaeleon i star - forming cloud . at the distance of @xmath0 150 pc , l@xmath1 = 1.35 @xmath2 10@xmath3 erg s@xmath4 ( 8 - 15 kev ) , or 2.6 @xmath2 10@xmath3 erg s@xmath4 ( 0.1 - 2.4 kev ) assuming a thermal spectrum with kt @xmath0 10 kev , a temperature higher than those previously seen in t tauri stars ( tsuboi et al . 1998 ) .
the _
peak _
x ray luminosity is at least 2 times higher than that derived for the protostar irs 43 ( grosso et al . 1997 ) which would make -to our knowledge- the strongest flare ever seen in a yso .
however , the possibility of grs 1100 - 771 being an isolated neutron star unrelated to the cloud can not be excluded , given the relatively large error box provided by watch . regarding the longer duration ( @xmath0 1 day )
x - ray transients , none of them seem to be related to known objects .
we suggest that the latter are likely to have originated from compact objects in low - mass or high - mass x - ray binaries , similarly to xte j0421 + 560 . |
a number of high - redshift star - forming galaxies have been identified through the lyman break surveys ( steidel et al .
1996a , b ; lowenthal et al .
since the number density of these lyman break galaxies ( lbg ) is comparable to or somewhat higher than that of the local @xmath2 galaxies , they are thought to be natural candidates of the progenitors of nearby normal galaxies ( lowenthal et al .
1997 ) . from the observed uv continuum emission , a typical star - formation rate ( sfr ) of lbgs
is estimated to be moderate , @xmath3 m@xmath4 yr@xmath5 ( @xmath6=50 km s@xmath5 mpc@xmath5 and @xmath7=0.5 ) . however , there are increasing evidences of the presence of significant amount of extinction in the uv continuum by interstellar dust in lbgs ( @xcite ; @xcite ) .
if the inferred uv extinction is corrected , their sfr would be more than @xmath8 yr@xmath5 ( dickinson 1998 ) , which may imply that high - redshift lbgs are in their major star - formation period . on the other hand , deep surveys in sub - mm wavelength with sub - millimeter common - user bolometer array ( scuba ; holland et al .
1998 ) revealed a population of fir - luminous galaxies possibly at high redshift and have a large sfr , @xmath9 m@xmath4 yr@xmath5 ( smail et al 1997 ; hughes et al .
1998 ; barger et al . 1998 ; lilly et al .
although the optical identification of the scuba sources has not been well established in some cases , they are also candidates of high - redshift forming galaxies .
it is then natural to ask what the relationship between these two populations is , or , how the high - redshift lbgs appear in sub - mm wavelength . in the deepest sub - mm imaging
so far at the hubble deep field ( hdf ; williams et al .
1996 ) by hughes et al .
( 1998 ) , none of the optically - selected lbgs ( steidel et al .
1996b ; lowenthal et al .
1997 ) was detected .
the upper - limit flux value at 850 @xmath10 m , however , can be used to put some important constraints for the spectral energy distribution ( sed ) of the lbgs . in this letter
, we investigate sub - mm emission properties of the lbgs in hdf to constrain their physical properties such as dust temperature by using the results of the scuba observations by hughes et al .
our sample of lbgs is the same as that used in sawicki and yee(1998 ) .
they are the 17 galaxies at z@xmath11 with firm spectroscopic redshifts and with photometric data from optical to near - infrared ( nir ) wavelength , taken from the 41 high - redshift lbg candidates originally provided by steidel et al .
( 1996b ) and lowenthal et al .
redshifts and magnitudes are listed in table 1 of sawicki and yee ( 1998 ) .
hughes et al . ( 1998 ) detected five sources at 850 @xmath10 m with scuba in hdf .
we found none of the optically - selected lbgs is detected within the error circles of the scuba sources .
hughes et al .
evaluated their detection limit @xmath12 mjy ( 5@xmath13 ) .
hereafter we quote this value as the upper - limit flux density at 850 @xmath10 m for all the lbgs although the true detection limit may change across the field slightly due to the position - dependent confusion noises and detector sensitivities .
the upper limit of sub - mm flux density gives a constraint on the spectral shape from the rest - frame uv to fir wavelength of the lbgs .
we compare the observed optical - nir flux densities and the upper limit of sub - mm flux density of lbgs with the three template spectra of local starburst galaxies . for the template spectra
, we adopt the sed of arp220 ( klaas et al .
1997 ) , as an example of heavily dust enshrouded starburst galaxies , and the averaged sed of high - reddening starburst galaxies ( sbh ) and low - reddening starburst galaxies ( sbl ) compiled by schmitt et al .
these templates are redshifted and then fitted to the optical - nir photometric data .
figure [ fig1]a and [ fig1]b show the best and the worst case in the results of our least - square fitting procedure . in many cases ,
the template of sbl gives a better fit but those of arp220 and sbh are always steeper than the spectra of lbgs ( figure [ fig1]a ) . for several objects , spectra of lbgs are even flatter than the sbl template ( figure [ fig1]b ) .
reduced @xmath14 values ( degree of freedom is four ) ranges 300 - 550 for arp220 , 50 - 250 for sbh , and 1.5 - 30 for sbl .
the absorption at the rest - frame uv wavelength in the lbgs thus seems much smaller than in arp220- or sbh - like galaxies .
if we assume the templates of arp220 and sbh ( fitted at optical - nir wavelength ) , the expected flux density at @xmath0 m exceeds 2 mjy for 11 and 9 of the 17 objects , respectively .
if the sbl template is applied , none of the objects has a @xmath0 m flux density larger than 2 mjy , which is consistent with the results of the scuba observation .
thus , not only the rest - frame uv - optical seds but also the fir - uv seds of lbgs are better represented by the sbl template . on the other hand ,
scuba - selected bright sub - mm sources show more fir excess .
for example , barger et al . (
1998 ) detected a sub - mm source with @xmath15 mjy by scuba in the lockman - hole field and found a plausible optical counterpart with @xmath16 .
they argued that only the sed of ultra - luminous infrared galaxies like arp220 can match this flux ratio .
the seds of the lbgs are thus much shallower than those of the scuba sources detected above a few mjy level at 850 @xmath10 m . the observed star - forming regions of lbgs are not likely to be heavily enshrouded by dust compared with the scuba - selected bright sub - mm sources or the local ultra - luminous infrared galaxies .
we first estimate the amount of dust extinction of lbgs by fitting the model uv / optical spectra ( at 900 - 10000 ; salpeter imf is assumed ) with an extinction to the observed seds following the similar manner as in sawicki and yee ( 1998 ) but using the different evolutionary - synthesis model developed by kodama & arimoto ( 1997 ) .
we assumed the reddening law derived by calzetti for the local starburst galaxies .
the estimated extinction , @xmath17 , ranges 0.3 - 0.5 mag with ages of @xmath18 0.03 gyr , which agrees with the results in sawicki & yee ( 1998 ) . for 13 of the 17 objects ,
acceptable fittings are obtained with finite amounts of extinction and we will discuss the 13 objects in this section .
the @xmath19 m flux densities of lbgs are then calculated , with the assumption that all the energy absorbed in uv / optical wavelength is re - emitted by the modified black - body radiation with single temperature .
the observed 850 @xmath10 m flux density , @xmath20 , is given by @xmath21 where @xmath22 is absorbed flux , @xmath23 is a dust temperature , @xmath24 is a power of the emissivity law , @xmath25 is the planck function , and @xmath26 .
figure [ fig2 ] shows the dependence of the estimated flux density on the dust temperature and emissivity for a typical case .
we assume @xmath27 in the following discussion , in order to compare with the properties of the local galaxies which are obtained with @xmath27 .
for the given @xmath24 value , the upper limit of the flux density at 850 @xmath10 m ( 2 mjy ) imposes the lower limit of the dust temperature , @xmath28 . in figure [ fig3 ]
( the fourth panel from the top ) , the distribution of the obtained @xmath28 is shown along with those of the dust temperature of local galaxies from the samples of the optically - selected cfa galaxies classified as e - s0/a and sb - sbc ( sauvage and thuan 1994 ) and the @xmath29-selected infrared - bright galaxies ( young et al .
these dust temperature of the local galaxies are measured by fitting the @xmath29 60 and 100 @xmath10 m flux densities with single - temperature modified - planck function with @xmath30 .
the distribution of @xmath28 of the lbgs has a peak at around 45 k and significantly higher than those of the local galaxies .
the median value of the temperature _ lower limit _ is 43 k while those of cfa and infrared - bright samples are @xmath31 k ( e / s0 ) , 30 k ( sb - sbc ) , and 36 k ( infrared bright ) . according to the peto - prentice generalized wilcoxon test
, the temperature distribution of lbgs is different from those of the local galaxies with a confidence level larger than 99.9999 % .
it is possible that the high dust temperature of the lbgs is introduced by our assumption of single dust temperature .
most of the nearby star - forming galaxies have a hot dust component ( @xmath32 k ) in addition to a cold one ( @xmath33 k ) ( eales , wynn - williams , and duncan 1989 ; klaas et al .
iras 60 and 100 @xmath10 m flux densities of the local galaxies may be dominated by only the cold component .
so as to estimate the energy drained to the hot dust , we fitted the seds of the three template spectra used in the previous section by the two component models .
the results are @xmath34 for the sbl , @xmath35 for the sbh , and @xmath36 for the arp220 template ( the result for arp220 is consistent with that of klaas et al . ) .
if we consider the contribution by the possible hot component in calculating the fir luminosity of the lbgs , adopting the ratio @xmath34 which gives the lowest limiting temperature , the peak of @xmath28 of the lbgs is shifted to @xmath37 k. according to the peto - prentice generalized wilcoxon test , they are still different higher than 99.9999 % level .
true @xmath28 must be between these two extreme cases , and it may be @xmath38 k. in any case , the dust temperature of the lbgs are likely to be still higher than those of the nearby normal galaxies .
the difference of local galaxies and high - redshift lbgs can be in dust extinction properties instead of dust temperature .
the reddening law in the lbgs is not understood very well ( e.g. , pettini et al .
if we use the smc - like reddening law instead of the calzetti law , the total amount of absorbed uv luminosity becomes smaller and the expected sub - mm flux density becomes several times smaller .
in such a case , only a weak constraint for dust temperature is obtained .
uv spectra of the local starburst galaxies are well approximated by a power - law of index @xmath39 ( 1250 @xmath40 ) , @xmath41 ( @xcite , @xcite ) , and there is an empirical relation between @xmath39 and @xmath42 values ( @xcite ) . in the range of @xmath43 , approximately , @xmath44 where @xmath45 $ ] @xmath46 @xmath47 , and @xmath48 ( @xmath49 ) .
hence , @xmath50 if we assume the modified black - body radiation with the power of emissivity law @xmath51 , the dust temperature @xmath23 implies the flux ratio @xmath52
. then we obtain , @xmath53 the sub - mm flux is predicted by the spectrum from eq.[eq of f60 ] .
again , the upper limit of @xmath54 mjy gives @xmath28 values .
the bottom panel of figure [ fig3 ] shows a distribution of @xmath28 obtained by this empirical method .
it peaks at @xmath55 , which is similar to or somewhat lower than the value calculated from the model uv extinction .
it is not surprising that the empirical method gives wider @xmath28 distribution since there is fairly large scatter in the empirical relation ( eq.[eq of b - fir ] ) itself ( meurer et al 1997 ) .
the average dust temperature of lbgs can also be examined with the observed intensity of cosmic far - infrared background ( cfirb ) at @xmath0 m , although the field of view of hdf ( 4.7 arcmin@xmath56 ) is very small and statistical uncertainty may not be negligible ; a typical fluctuation of cfirb at this scale is still unknown . the total @xmath0 m flux in hdf evaluated from the average intensity of the cfirb radiation , @xmath57 w m@xmath58 sr@xmath5 ( fixsen et al .
the five scuba sources contribute about 45 % of this flux .
the contribution of the lbgs in hdf should not significantly exceed the remaining 55% of the cfirb flux . in figure [ fig4 ]
, we plot the contribution of the lbgs studied in the previous section to the cfirb with respect to the assumed dust temperature .
three curves are those given by the model calculations with single and double temperature components and the prediction using the empirical relation .
depending on temperature , lbgs contribute a significant fraction ( @xmath59 ) . with the constraint not to exceed the cfirb flux , we obtain fairly high limiting temperature , @xmath38 k , which is consistent with the limit obtained by using the scuba upper - limit flux density .
we showed that the lbgs in hdf have uv - to - fir seds different from those of the ultra - luminous fir galaxies in the local universe and of the scuba - selected bright sub - mm sources .
it may imply that the birght sub - mm sources detected above a few mjy at 850 @xmath10 m are rather extremely dust - rich objectes among the high - redshift star - forming galaxies selected by the lyman break technique although it does not have to mean that the entire parent populations of sub - mm sources and lbgs are fundamentally different .
we also claimed that dust temperature in the lbgs should be higher than @xmath1k , if we assume the reddening law appropriate for local starburst galaxies .
what causes the relatively high dust temperature in the lbgs ?
high dust temperature may be results of some intrinsic properties of star formation in high - redshift galaxies .
there are some possible explanations for effective dust heating .
for example , a flatter initial mass function ( imf ) which produces a larger number of massive stars for a given mass of dust may provide the effective dust heating .
high star - formation efficiency ( a higher sfr for a given mass of dust regardless of imf ) brings the similar situation
. the difference may also be in properties of dust grains .
for the case of low metal abundance ( possibly true in young galaxies ) , dust may be dominated by small - size grains which are easily heated .
finally , we discuss the detectability of lbgs in sub - mm observations .
assuming an appropriate spectral shape like sbl or dust temperature of @xmath60 , we predict that sub - mm flux at @xmath19 m of the lbgs in hdf is typically 0.1 - 1 mjy .
so within the reasonable amount of observational time , only the brightest lbgs could be detected with scuba . in the coming decade , however , new sub - mm observational facilities such as large millimeter and sub - millimeter array ( naoj ) or millimeter array ( nrao ) , etc . are planned .
our results presented in this letter provides rather optimistic future prospects ; only a few hour exposure time will be needed to detect typical lbgs with such new facilities and then we will learn about true star - formation properties of typical high - redshift galaxies . | we investigate expected sub - mm emission and dust properties of the lyman break galaxies ( lbgs ) in the hubble deep field ( hdf ) .
the scuba deep survey ( @xcite ) provides an upper limit of the @xmath0 m flux densities of the lbgs . with this constraint
, we argue that a typical ultra - violet to far - infrared spectral shape of the high - redshift lbgs is rather close to a template spectrum of low - reddening starburst galaxies in the local universe but different from that of heavily dust - enshrouded ultra - luminous far - infrared galaxies like arp220 .
we also evaluate the lower - limit temperature of dust in lbgs assuming single- and two - component modified black - body spectrum . to estimate the total amount of energy re - emitted in fir wavelength , we take two different approaches ; model fitting of the uv spectra of lbgs and an empirical method using the relationship between uv spectral shape and uv / fir flux ratio observed for local starburst galaxies .
both methods give lower - limit temperature of the lbgs as @xmath1k , which is higher than the typical dust temperature of local optical- and infrared - selected galaxies .
this result is also supported by the comparison of the expected sub - mm flux of the lbgs with the cosmic fir background radiation .
the high dust temperature may indicate the effective massive - star formation or different dust properties in the high - redshift lbgs . |
in superdense nuclear matter that constructs the core of a neutron star , a great amount of energy is carried away by the neutrino .
for the standard composition of superdense matter ( neutrons , with an admixture of protons and electrons ) , the main neutrino emission mechanisms are powerful direct urca reactions , @xmath2 and much weaker modified urca reactions , @xmath3 the direct urca reactions can proceed only if the ratio of the proton number density to the total baryon number density @xmath4 exceeds a certain threshold value which allows simultaneous energy and momentum conservation @xcite .
otherwise , direct urca reactions are forbidden .
the most powerful neutrino energy losses are produced by modified urca reactions .
furthermore , the @xmath0 reactions just mentioned bring the constituents into the state of chemical equilibrium , @xmath5 , which determines the relative concentrations of particles . however , if any external or macroscopic phenomenon changes the density of a matter element , the @xmath0 equilibrium state will change , resulting in a departure from equilibrium quantified by the chemical imbalance @xmath6 .
several authors have investigated many astrophysical situations in which the nonequilibrium @xmath0 processes occur , such as gravitational collapse of neutron star @xcite , pulsar spin - down @xcite , a hypothetical time variation of the gravitational constant @xcite , the existence of even relatively slow hydrodynamic flow @xcite or of millisecond oscillations @xcite in the neutron star interiors .
the departure from equilibrium , @xmath7 , increases the phase space for nonequilibrium @xmath0 process and strongly enhances corresponding neutrino emissivity with respect to the chemical equilibrium value @xcite .
it is well known that below a certain critical temperature nucleons in neutron star matter will be in superfluid states .
the appearance of energy gaps @xmath8 and @xmath9 reduces the particle momentum space , which contributes to the reaction rates ; that is , the `` effective '' widths of momentum space of the reacting particles near the fermi surfaces become thinner .
this suppresses the reaction rate and the neutrino emissivity . in @xmath0-equilibrium matter ,
superfluid reductions in reaction rate and neutrino emissivity are dependent on the ratio of the gap to the temperature @xmath10 . for strong superfluidity , the reductions decay exponentially , while for intermediate values of @xmath10 , the situation is more sophisticated .
the detailed expressions are given in a review@xcite .
the departure from equilibrium @xmath11 increases the neutrino emissivity , while the superfluid energy gaps behave in the opposite way .
indeed , the nonequilibrium @xmath0 reaction with nucleon superfluidity is a new case that is worth studying .
reisenegger @xcite first analyzed the mechanism to produce a faster reaction owing to nonequilibrium with neutron and proton superfluidity . using a crude model
, he assumed that for @xmath12 , @xmath0 reactions were completely suppressed , while for @xmath13 the effect of superfluidity could be neglected .
villain and haensel broke through reisenegger s steplike modeling and calculated the precise reduction factors of the net reaction rates for nonequilibrium direct and modified urca processes in the presence of various types of superfluidity by means of sophisticated numerical methods @xcite . however , neutrino emissivity is also an important quantity .
we need to calculate the relevant neutrino emissivities , which , along with the net reaction rates , can be used to simulate the evolution of superfluid neutron star cores that are off @xmath0 equilibrium .
our primary goal here is to obtain the total neutrino emissivities of nonequilibrium @xmath0 processes in superfluid circumstances .
the paper is organized as follows . in sec .
ii , we recall the nonequilibrium condition and basic features of nucleon superfluidity . in secs .
iii and iv , we detail the numerical calculations of neutrino emissivities for durca and murca reactions and show our results .
section v includes a short summary of our conclusions and discussion of possible applications .
we consider a simple @xmath1 model of neutron star matter .
each of the constituents is strongly degenerate , with fermi momenta @xmath14 and chemical potential @xmath15 ( @xmath16 ) .
they are in thermodynamic equilibrium but not necessarily in chemical equilibrium . in the absence of @xmath0 equilibrium
, there is a finite difference in the chemical potentials , @xmath17 , following the conventions of ref .
@xcite .
nucleon superfluidity occurs via cooper pairing of particles owing to an attractive component of their interaction , with the appearance of a gap @xmath18 in the particle energy spectrum near the fermi level @xcite .
it is widely accepted that there are likely three types of nucleon superfluidity in @xmath1 matter , @xmath19@xmath20@xmath21 , and , denoted a , b , and c , respectively ( table [ tab : abc ] ) @xcite .
we introduce two dimensionless quantities , @xmath22 and @xmath23 , to describe the chemical deviation and gap amplitude , for convenience , and use natural units with @xmath24 throughout this paper .
to guarantee the integrity of this paper , we recall fundamental properties and characteristic quantities @xcite .
the onset of superfluidity is accompanied by the appearance of the energy gap @xmath18 . near the fermi surface ( @xmath25 )
, we have @xmath26 here @xmath27 , @xmath28 and @xmath29 are the fermi momentum and fermi velocity , respectively ; @xmath30 is the chemical potential ; and @xmath31 , where @xmath32 is an amplitude that determines the temperature dependence of the gap and @xmath33 describes the dependence of the gap on the angle @xmath34 between the quantization axis and the particle momentum .
the quantities @xmath35 and @xmath36 are determined by the superfluidity type ( table [ tab : abc ] ) . in case
a the gap is isotropic , and @xmath37 . in cases
b and c , the gap depends on @xmath34 .
note that in case c the gap vanishes at the poles of the fermi sphere at any temperature : .
+ ' '' '' & superfluidity type & @xmath38 & @xmath33 & @xmath39 + a & @xmath40 & 1 & 1 & 0.5669 ' '' '' + b & @xmath41 & 1/2 & @xmath42 & 0.8416 + c & @xmath43 & 3/2 & @xmath44 & 0.4926 + [ tab : abc ] the gap amplitude @xmath32 is determined by the bcs theory which can be written as @xmath45 where @xmath46 , @xmath47 is a solid angle element in the direction of particle momentum @xmath48 , @xmath49 is the fermi - dirac distribution , @xmath38 is a numeric coefficient ( table [ tab : abc ] ) , and @xmath50 note that in the absence of nucleon superfluidity , @xmath51 . according to table [ tab : abc ] , the dimensionless gap @xmath52 can be written as : @xmath53 in the following calculations that involve nucleon superfluidity , we assume that only one type of nucleon is superfluid , which is equivalent to assuming that the larger gap prevails .
this approximation seems to be quite reasonable @xcite .
our goal in this paper is to investigate the influence of nucleon superfluidity on the neutrino emissivity of nonequilibrium @xmath0 processes , which is described by the reduction factor @xmath54 .
@xmath55 and @xmath56 refer to the neutrino emissivities in two cases : one is nonequilibrium @xmath0 process with nucleon superfluidity ; the other is nonequilibrium @xmath0 process without nucleon superfluidity .
furthermore , @xmath57 labels the type of reactions , @xmath58 denotes the superfluidity type ( a , b , or c ) , and @xmath59 the type of superfluid nucleon ( @xmath60 or @xmath61 ) with gap @xmath62 .
note that for @xmath63 ( the critical temperature of the superfluid nucleon ) , @xmath64 , and for @xmath65 , @xmath66 .
as shown in ref .
@xcite , the direct urca process is allowed by the momentum conservation when @xmath67 . for pure @xmath1 matter where @xmath68 , it corresponds to @xmath69 .
this happens if the density is several times higher than the standard nuclear matter density , @xmath70 g @xmath71 . in the absence of @xmath0 equilibrium ,
the neutrino emissivities of two direct urca processes with nucleon superfluidity are in the following forms : @xmath72\ ; \delta(\mathbf{p}_f-\mathbf{p}_i)|m_{fi}|^{2}\prod_{j=1}^3 p_{f_j } m_j^\ast \int_0^\infty d x_\nu \ ; x_\nu^3 \nonumber \\ & & \int_{-\infty}^{+\infty } d x_1 \ ; f(z_1 ) \int_{-\infty}^{+\infty } d x_2 \;(1-f(z_2 ) ) \int_{-\infty}^{+\infty } d x_3 \;(1-f(x_3))\delta(z_2+x_3-z_1+x_\nu + \xi ) , \label{eq : qdn}\end{aligned}\ ] ] @xmath73\ ; \delta(\mathbf{p}_f-\mathbf{p}_i)|m_{fi}|^{2 } \prod_{j=1}^3 p_{f_j } m_j^\ast \int_0^\infty d x_\nu \ ; x_\nu^3 \nonumber \\ & & \int_{-\infty}^{+\infty } d x_2 \ ; f(z_2 ) \int_{-\infty}^{+\infty } d x_3 \;f(x_3 ) \int_{-\infty}^{+\infty } d x_1 \;(1-f(z_1))\delta(z_1-z_2-x_3-x_\nu - \xi ) .
\label{eq : qdp}\end{aligned}\ ] ] where @xmath74 corresponds to @xmath75 respectively , @xmath76 is the dimensionless energy of the neutrino , @xmath77 , @xmath62 , and @xmath78 have been defined in sec .
[ 2 ] , and @xmath79 is the fermi - dirac function of nucleons and electron , @xmath80 is the corresponding fermi momentum and @xmath81 is the effective particle mass .
furthermore , @xmath82 is the solid angle element in the direction of the particle the momentum @xmath83 , and the @xmath18 functions describe momentum and energy conservations of the particles in the initial and final states .
finally , @xmath84 is the squared reaction amplitude , and for nonrelativistic nucleons , it is independent of particle momenta and can be taken out of the integral . in this paper , we focus on the total neutrino emissivity @xmath85 .
let us start with the total neutrino emissivity @xmath86 of the direct urca processes without nucleon superfluidity under @xmath0 equilibrium .
under the condition of @xmath0 equilibrium and without nucleon superfluidity , we set @xmath87 and replace @xmath88 with @xmath89 .
the direct and inverse reactions have the same neutrino emissivity .
then @xmath86 can be written as ( for details see ref .
@xcite ) : @xmath90\delta(\mathbf{p}_f-\mathbf{p}_i ) , \label{eq : adur } \\ & & i_{0}^{(d ) } = \int_0^\infty d x_\nu \ ; x_\nu^3 \left [ \prod_{j=1}^3 \int_{-\infty}^{+\infty } d x_j \ ; f_j \right ] \delta \left ( \sum_{j=1}^3 x_j - x_\nu \right ) , \label{eq : idur}\end{aligned}\ ] ] here the integrals @xmath91 and @xmath92 are standard ( @xcite ) : @xmath93 in this section , we recall the neutrino emissivity of nonequilibrium durca processes without nucleon superfluidity .
when @xmath0 processes are off chemical equilibrium , @xmath94 , the direct and inverse reactions of urca processes have different neutrino emissivities .
the expressions for the neutrino emissivities of the two reactions are similar to eqs .
( [ eq : qdn ] ) and ( [ eq : qdp ] ) , but the variable @xmath88 is replaced with the corresponding @xmath89 .
the total neutrino emissivity can be presented in the forms @xcite @xmath95 with @xmath96 , \\ & & j(x_\nu - \xi)= \left [ \prod_{j=1}^3 \int_{-\infty}^{+\infty } d x_j \ ; f_j \right]\delta(x_1 + x_2 + x_3 - x_\nu + \xi ) , \label{eq : jminus } \\ & & j(x_\nu + \xi)= \left [ \prod_{j=1}^3 \int_{-\infty}^{+\infty } d x_j \ ; f_j \right]\delta(x_1 + x_2 + x_3 - x_\nu -\xi ) .
\label{eq : jplus}\end{aligned}\ ] ] according to ref .
@xcite , @xmath97 we obtain @xmath98 consider suppression of nonequilibrium direct urca processes by proton or neutron superfluidity .
the superfluidity affects the dispersion relation of nucleons under the integrals in eqs .
( [ eq : jminus ] ) and ( [ eq : jplus ] ) , in accordance with eq .
( [ eq : gap ] ) .
the neutrino emissivity can be written as @xmath99 @xmath100 is the superfluid reduction factor .
the total neutrino emissivity @xmath101 is in the following form : @xmath102 where @xmath103\ ; \delta(\mathbf{p}_f-\mathbf{p}_i)\int_0^\infty dx_\nu x_\nu^3 \ ,
\nonumber \\ & & \int_{-\infty}^{+\infty } dx_1 dx_2 dx_3 \ , \left [ f_1(1-f_2)(1-f_3)\delta(z_1 -z_2-x_3-x_\nu+\xi ) \ , + f_2f_3(1-f_1)\delta(z_1 + x_\nu - z_2-x_3+\xi)\right ] , \qquad \label{eq : idsf}\end{aligned}\ ] ] and @xmath82 is the element of the solid angle in the direction of the momentum of superfluid nucleon @xmath59 . here
, we consider the case where only one type of nucleon is superfluid , then the @xmath104 variable for the nonsuperfluid nucleon has to be replaced with the corresponding @xmath105 variable , while @xmath62 is `` included '' in the @xmath104 variable for the superfluid nucleon .
thus , the reduction factor of nucleon superfluidity for a nonequilibrium durca process can be presented by @xmath106 where @xmath107 .
\qquad \label{eq : hdsf}\end{aligned}\ ] ] in eq .
( [ eq : hdsf ] ) , integration over the electron variable can be done using the formula for fermi integrals , similar to the treatment in ref .
@xcite , @xmath108 which enables us to integrate over the @xmath105 variable for the nonsuperfluid nucleon , giving @xmath109.\end{aligned}\ ] ] we assume only one type of nucleon superfluidity and numerically evaluate @xmath100 for the wide range of interest with algorithmic methods ( refer to the appendix of @xcite ) . in fig .
[ 1fig ] we present the reduction factor for durca reaction with isotropic superfluidity of protons or neutrons .
the variable is dimensionless chemical departure @xmath110 .
the curves correspond to different values of dimensionless gap amplitude @xmath111 . from this figure
we can get that there are at least two features : ( i ) for a given curve , the change of the chemical departure value almost does not affect the reduction factor until it increases at a value ; ( ii ) for huge gaps @xmath112 , as soon as @xmath113 , and until the reduction factor becomes close to 1 , the curves are almost parallel .
this means that for small chemical departure , the reduction factor can be written as the function of the variable @xmath114 only , while for huge gaps it can be written as @xmath115 .
these are similar to the net reaction rates presented by villain and haensel @xcite . additionally , the behavior of the superfluid suppression on the neutrino emissivity here is different from that proposed by reisenegger @xcite .
he suggested that for @xmath12 the beta reactions were completely suppressed ( @xmath116 ) , while for @xmath117 the effect of superfluidity could be neglected ( @xmath118 ) .
however , as can be seen from this figure , the reduction factor @xmath119 does nt equal to @xmath120 or @xmath121 .
in other words , we find that in a proper range of the value @xmath122 , the neutrino emissivity will be suppressed by the nucleon superfluidity , but it is not completely suppressed .
if taking the onset of rising r as the threshold point , we find that in the region of @xmath123 , @xmath124 ( approximately 0.78 in fig .
[ 2fig ] ) , while for @xmath125 , @xmath126 . here
@xmath127 is the value of @xmath30 at the threshold point .
eventually , when and only when the value of @xmath128 approaches infinity , @xmath129 , our results are fully same with that in reisenegger @xcite .
these are easily illustrated by plotting @xmath130 as function of @xmath122 in fig .
[ 2fig ] , where @xmath130 is the partial derivative of @xmath119 with respect to @xmath110 .
it is clearly that the curves in this figure can confirm our above discussions .
the above behavior of superfluid suppression on the neutrino emissivity can also be seen from all the other comparable figures in the following . figs .
[ 3fig ] and [ 4fig ] show the dependence of the reduction factor of durca reaction on dimensionless chemical departure and dimensionless gap amplitude with anisotropic superfluidity of b and c , respectively .
compared with fig .
[ 1fig ] , we can see that the reduction factors can either increase ( type c ) or decrease ( type b ) in respect to that of type a , which means that the maximal value of the gap on the fermi surface is important since the value for case c is larger than for case a , while for case b is smaller . .
direct reaction with isotropic superfluidity of protons or neutrons . , scaledwidth=50.0% ] ( @xmath131 ) on the chemical imbalance for various gap s amplitude .
the numbers next to the curves give value of @xmath111 .
direct reaction with isotropic superfluidity of protons or neutrons .
, scaledwidth=50.0% ] ) but with anisotropic superfluidity type b of the neutrons . ,
scaledwidth=50.0% ] ) but with anisotropic superfluidity type c of the neutrons .
, scaledwidth=50.0% ]
as mentioned in the previous section , the direct urca process in @xmath1 matter is allowed at densities several times the nuclear density .
then in smaller densities the modified urca process becomes the main neutrino reaction .
reaction ( [ eq : murca ] ) differs from reaction ( [ eq : durca ] ) by the presence of an additional nucleon to ensure momentum conservation .
the modified urca process will be labeled by upperscripts ( mn ) , where @xmath132 indicates the neutron branch of reaction ( [ eq : murca ] ) and @xmath133 indicates the proton branch .
similar to above section , we give the expressions of the neutrino emissivities for the direct and inverse murca processes out of equilibrium in superfluid nuclear matter , @xmath134\delta(\mathbf{p}_f-\mathbf{p}_i)|m_{fi}|^{2 } \prod_{j=1}^5 p_{f_j } m_j^\ast\int_0^\infty \dd x_\nu \ ; x_\nu^3 \nonumber \\ & & \int_{-\infty}^{+\infty } \dd x_1\dd x_2 \dd x_3 \dd x_4\dd x_5\ ; f(z_1)f(z_2 ) \;(1-f(z_3 ) ) \;(1-f(z_4))(1-f(x_5))\delta(z_3+z_4+x_5-z_1-z_6+x_\nu + \xi ) , \label{eq : qmnn}\end{aligned}\ ] ] @xmath135\delta(\mathbf{p}_f-\mathbf{p}_i)|m_{fi}|^{2 } \prod_{j=1}^5 p_{f_j } m_j^\ast \int_0^\infty \dd x_\nu \ ; x_\nu^3 \nonumber \\ & & \int_{-\infty}^{+\infty } \dd x_1\dd x_2 \dd x_3 \dd x_4\dd x_5\;(1-f(z_1))(1-f(z_2 ) ) f(z_3 ) \;f(z_4)f(x_5 ) \delta(z_3+z_4+x_5-z_1-z_2-x_\nu - \xi ) .
\label{eq : qmnp}\end{aligned}\ ] ] where @xmath136 correspond to @xmath137 ( @xmath138 and @xmath139 refer to the initial and finial spectator nucleons ) , while @xmath140 corresponds to the electron .
the remaining parameters are the same as in durca case .
the total neutrino emissivity of one branch of murca processes is @xmath141 .
if none of the nucleons is superfluid , we have no necessary to make distinction between the two branches in the phase - space integral and @xmath142 respectively . under the condition of chemical equilibrium , @xmath143 , the neutrino emissivity of direct and inverse murca processes without nucleon superfluidity are the same .
the total neutrino emissivity of one branch of murca processes have been calculated by @xcite , @xmath144\delta(\mathbf{p}_f-\mathbf{p}_i ) , \label{eq : amur } \\ & & i_{0}^{(mn ) } = \int_0^\infty \dd x_\nu \ ; x_\nu^3 \left [ \prod_{j=1}^5 \int_{-\infty}^{+\infty } \dd x_j \ ; f_j \right ] \delta \left ( \sum_{j=1}^5 x_j - x_\nu \right ) , \label{eq : imn}\end{aligned}\ ] ] here @xmath145 is the fermi - dirac distribution function .
the quantity @xmath84 appear to be constant and can be taken out the integral .
the detailed calculations refer to @xcite .
we have @xmath146 the neutrino emissivity of non - equilibrium modified urca process without nucleon superfluidity has been studied by some authors ( @xcite ) .
when the matter is out of beta equilibrium , @xmath7 , it opens additional volume in the phase - space which results in an increase of the neutrino emissivity .
it can be written as : @xmath147 with @xmath148 , \label{eq : imnnosf}\\ & & j(x_\nu - \xi)= \left [ \prod_{j=1}^5 \int_{-\infty}^{+\infty } \dd x_j \ ; f_j \right]\delta(\sum_{j=1}^5 x_j - x_\nu + \xi ) = \frac{\pi^2 + ( x_\nu - \xi)^2}{2(e^{x_\nu - \xi } + 1 ) } , \label{eq : jxminus}\\ & & j(x_\nu + \xi)= \left [ \prod_{j=1}^3 \int_{-\infty}^{+\infty } \dd x_j \ ; f_j \right]\delta(\sum_{j=1}^5 x_j - x_\nu -\xi ) = \frac{\pi^2 + ( x_\nu + \xi)^2}{2(e^{x_\nu + \xi } + 1)}. \label{eq : jxplus}\end{aligned}\ ] ] then the effect of the equilibrium departure on the neutrino emissivity of murca process can be presented as : @xmath149 according to @xcite , @xmath150 thus @xmath151 the analysis of the modified urca process is quit similar with the direct urca process .
the appearance of nucleon superfluidity reduces the available phase - space which suppresses the neutrino emissivity , @xmath152 .
we label @xmath153 as the superfluid reduction factor . we can write @xmath154 notice that for @xmath63 ( the critical temperature of superfluid nucleon ) @xmath155 , and for @xmath156 @xmath157 .
the total neutrino emissivity @xmath158 is in the following form : @xmath159 where @xmath160 is the same with eq .
( [ eq : amur ] ) and @xmath161\delta(\mathbf{p}_f-\mathbf{p}_i)\int_0^\infty \dd x_\nu \ ; x_\nu^3 \left [ \prod_{j=1}^5 \int_{-\infty}^{+\infty}\dd x_j \ ; f_j \right ] \ , [ \delta ( \sum_{j=1}^5 z_j - x_\nu+\xi ) + \delta ( \sum_{j=1}^5 z_j - x_\nu-\xi ) ] .
\label{eq : imnsf}\ ] ] thus , the reduction factor of nucleon superfluidity for non - equilibrium murca process can be presented by @xmath162 in the following , we numerically analyze some special cases that may occur in dense matter and give the diagrams to illustrate the relationship of superfluid reduction factor with the equilibrium departure and the superfluid gap amplitude .
case 1 . the neutron branch with isotropic superfluidity of protons since the singlet - state pairing proton gap is isotropic , the angular and energy integrations in eq .
( [ eq : imnsf ] ) are separated .
the angular integrals are the same as in the non - superfluid case , and we get @xmath163 with @xmath164.\ ] ] this result is also valid for the case when the neutron superfluidity is of singlet type in the proton branch of murca processes .
the proton branch with any superfluidity of neutrons let the protons be normal , but the neutrons be superfluid due to @xmath165 cooper pairing .
the integration over solid angles @xmath82 of all particles but the neutron in eq .
( [ eq : imnsf ] ) is the same as for non - superfluid case .
we get @xmath166 where @xmath167 , @xmath34 is the angle between the quantization axis and the momentum of the superfluid neutron and @xmath168 is the same as in eq .
( [ eq : rpamn ] ) .
the neutron branch with isotropic superfluidity of neutrons now consider the neutron branch with isotropic superfluidity of neutrons .
let @xmath169 in eq .
( [ eq : imnsf ] ) refer to neutron . integrating over @xmath170 and @xmath171
, we get @xmath172 @xmath173 employing the numerical calculations similar with that done in @xcite , we give the results for case 1 and case 3 in figs . [ 5fig ] and [ 6fig ] respectively .
we need not give the similar graph for case 2 .
the result of case 2 with isotropic superfluidity of neutrons are same with case 1 .
comparing figs .
[ 5fig ] and [ 6fig ] , we can see that the impact with superfluid spectator nucleons is much larger than that with superfluid non - spectator nucleons .
based on the similar reasons with that presented in @xcite , the cases with triplet - state nucleon superfluidity have been neglected . .
modified neutron branch with isotropic superfluidity of protons . , scaledwidth=50.0% ] )
but for modified neutron branch with isotropic superfluidity of neutrons .
, scaledwidth=50.0% ]
we apply the methods of villain and haensel in paper @xcite to calculate the effects of nucleon superfluidities on the neutrino emissivities of direct and modified urca processes out of equilibrium , while they calculated the corresponding net reaction rates .
although the numerical methods are similar , there are two differences in calculating neutrino emissivities and net reaction rates which can be seen easily by camparing eq .
( [ eq : imnsf ] ) and eq . ( 25 ) of paper @xcite , for example .
firstly , the integral @xmath174 takes the place of @xmath175 .
secondly , the two delta functions add together in eq .
( [ eq : imnsf ] ) while subtract each other in eq .
( 25 ) of previous paper .
so the calculations of the neutrino emissivities are necessary .
we here give the graphical overview of those results for - : : durca reaction with superfluidity of each type ; - : : murca reaction with isotropic superfluidity of the non - spectator nucleon ( e.g. proton for the neutron branch ) ; - : : murca reaction with isotropic superfluidity of the spectator nucleon ( e.g. neutron for the neutron branch ) . as shown in calculations , the appearance of nucleon superfluid
reduces the neutrino emissivity by a factor that depends on the value of @xmath10 and of @xmath122 , the types of the beta process ( direct or modified urca ) and of superfluidity . when the quantity @xmath122 is relatively small
, the superfluid reduction factors do nt change with the value of @xmath122 and become the function of only one variable @xmath10 .
it should be similar with that under the condition of beta equilibrium .
however , the violation of superfluid suppression on neutrino emissivity occurs when the chemical departure reaches a threshold value .
the threshold tends to be the result presented by reisenegger ( @xcite ) only when the energy gap of superfluidity is large enough .
it s well - known that neutrino emissivity plays a key role on the cooling of neutron stars ( @xcite ) .
we recognize that although the superfluid gap suppresses dramatically the neutrino emissivity in dense matter , it is possible that the chemical imbalance of beta reaction large enough will break the suppression . meanwhile , the imbalance contributes chemical heat to neutron star core . along with the results of reaction rates calculated by villain and haensel ( @xcite ) ,
our results are required for numerical simulation of neutron star cooling .
we have the primary investigations on this problem .
it will be reported in our another paper .
the authors are grateful to k.s .
cheng for discussions , to l. villain for sending us their subroutines which calculate reaction rates .
this research was supported by nfsc under grants no 10773004 .
j. m. lattimer , c. j. pethick , m. prakash and p. haensel , phys .
lett . * 66 * , 2701 ( 1991 ) .
p. haensel , astron . astrophys . * 262 * , 131 ( 1992 ) .
a. reisenegger , astrophys .
j. * 442 * , 749 ( 1995 ) .
a. reisenegger , astrophys .
j. * 485 * , 313 ( 1997 ) .
r. fernndez , and a. reisenegger , astrophys .
j. * 625 * , 291 ( 2005 ) .
p. jofr , a. reisenegger , and r. fernndez , phys .
lett . * 97 * , 131102 ( 2006 ) .
v. a. urpin , and d. a. shalybkov , mon . not .
. soc . * 281 * , 145 ( 1996 ) .
a. reisenegger , and p. goldreich , astrophys .
j. * 395 * , 240 ( 1992 ) .
d. g. yakovlev , a. d. kaminker , o. y. gnedin , and p. haensel , phys . rep . * 354 * , 1 ( 2001 ) . l. villain , and p. haensel , astron . astrophys . * 444 * , 539 ( 2005 ) .
u. lombardo and h .- j .
schulze , in _ physics of neutron star interiors _ , springer lecture notes in physics ( springer , berlin ) , eds .
d. blaschke , n. k. glendenning , and a. sedrakian , v. * 578 * , p. 30
( 2001 ) k. p. levenfish , and d. g. yakovlev , astron . lett . *
20 * 43 ( 1994 ) d. g. yakovlev , k. p. levenfish , and yu .
a. shibanov , physics - uspekhi * 42 * , 737 ( 1999 ) .
shapiro , and s.a .
teukolsky , 1983 , in _ black holes , white dwarfs and neutron stars _ , ( wiley - interscience , new - york , 1983 ) b. l. friman , and o. v. maxwell , astrophys .
j. * 232 * , 541 ( 1979 ) .
d. g. yakovlev , and k. p. levenfish , astron .
astrophys . * 297 * 717 ( 1995 ) d.page , j. h.applegate , astrophys .
j. lett .
* 394 * l17 ( 1992 ) d. g. yakovlev and c. j. pethick , ann .
. astrophys . * 42 * , 169 ( 2004 ) . | we investigate the influence of nucleon superfluidity on the neutrino emissivity of nonequilibrium @xmath0 processes .
calculations of the reduction factors for direct and modified urca processes with three types of nucleon superfluidity in @xmath1 matter are performed .
the numerical results are given because the analytical solution is impossible .
we find that the superfluid influence is closely related to the chemical departure from @xmath0 equilibrium .
for a small chemical departure , the superfluid reduction factor depends almost only on the gap and is hardly affected by the departure , while for a large enough departure , it rapidly enhances neutrino emissivity .
the onset of `` enchanced '' emission has some corresponding thresholds that seem to be linked to the ratio of the energy gap to the chemical departure . |
derived equivalences have become increasingly important in representation theory , lie theory and geometry .
examples are ranging from mirror symmetry over non - commutative geometry to the kazhdan - lusztig conjecture and to brou s conjecture for blocks of finite groups . in all of these situtations , and in many others , derived equivalences
are used that involve finite or infinite dimensional algebras . derived equivalences between algebras , or rings ,
exist if and only if there exist suitable tilting complexes , as explained quite satisfactorily by rickard s morita theory for derived categories of rings ( see @xcite ) ) .
derived equivalences have been shown to preserve many significant algebraic and geometric invariants and often to provide unexpected and useful new connections .
a crucial question in this context has , however , not yet received enough answers : _ how to construct derived equivalences between rings in a general setup ? _ a good answer - certainly not unique - to this question should be general , flexible and systematic and apply to a multitude of algebraic and geometric situations .
one well - developed approach is based on the theory of tilting modules , building upon results by happel @xcite .
other answers use ring theoretic constructions , such as trivial extensions @xcite .
the aim of this article is to provide a rather different approach .
the input of the technology developed here is a triple of objects @xmath3 in a triangulated category .
these objects are required to be related by certain universal maps ( cohomological approximations - a new concept introduced here , continuing approximation theory of auslander , reiten and smal @xcite ) and some cohomological orthogonality conditions in degrees different from zero only .
the output is a derived equivalence between cohomological endomorphism rings of @xmath4 and of @xmath5 .
the flexibility of the construction lies in the following features : we enhance endomorphism rings by higher extensions to produce cohomological endomorphism rings , broadening the classical concept of yoneda extension algebras . here
, we can choose a set of cohomological degrees to define the cohomological endomorphism ring .
choosing degree zero only gives endomorphism rings in the usual sense - and then no orthogonality assumption is needed .
choosing all integers , or a suitable subset thereof ( satisfying an associativity constraint ) , amplifies the concept of yoneda extension algebras @xmath6 .
there is also some flexibility in the choice of @xmath7 . a special case of such a triple is given by any auslander - reiten triangle @xmath8 in a derived module category ; this already indicates generality of the construction .
our assumptions are actually much more general and not limited to objects in derived categories of algebras .
a particular feature of the derived equivalences constructed by this method is that they also provide a very general _ mutation procedure _ , turning one ring into another one in a systematic way .
tilting theory has arisen as a far reaching extension of reflection functors for quivers . under some assumptions , but not in general , it provides mutation procedures between two given quivers or algebras , both of which are endomorphism rings of tilting modules ; in the case of quivers one may reflect at sink or source vertices .
mutations similar in style also have come up in various geometric situations .
the theory of cluster categories , or more generally of calabi - yau categories , has extended reflections to a mutation procedure , which works for representations of quivers at all vertices .
such mutations fit into the present framework as well .
there is , though , a new feature introduced by our approach : reflection does not work in general in derived categories ( of quivers or algebras ) .
therefore cluster theory passes to the cluster category , a quotient of a derived category modulo the action of some functor ; endomorphism rings are taken there .
in contrast to this , the current approach always produces equivalences on the level of derived categories , not just of quotient categories ; throughout we are considering derived equivalences between ( cohomological ) endomorphism rings or quotients thereof . in the case of quivers , this possibility of passing to quotient algebras allows mutation at an arbitrary vertex .
more generality and flexibility is added by extending the concept of higher extensions , that is of shifted morphisms ; it is possible to replace the shift functor by any other auto - equivalence of the ambient triangulated category .
there is even a version using two such functors . the main result of this article provides a construction of derived equivalences in a setup that is very general in several respects .
in the following explanation we start with a special case and then add generality step by step , finally arriving at the main result .
the setup always is a triangulated category @xmath9 , which is an @xmath10-category for some commutative artinian ring @xmath10 , with identity ; so , morphism sets in @xmath9 are @xmath10-modules . 1 . to start with , we choose any object @xmath7 in @xmath9 and a triangle @xmath11 $ ] , where @xmath12 and @xmath13 are @xmath14-approximations , that is universal maps from @xmath15 to objects in @xmath14 or from @xmath14 to @xmath16 , respectively ; in particular , @xmath17 is in @xmath14 .
for instance , auslander - reiten triangles ( over algebras ) provide such situations .
if the triangle is induced by an exact sequence in an abelian category , then the theorem implies a derived equivalence between the two endomorphism rings @xmath18 and @xmath19 .
this can be seen as a mutation procedure relating the two endomorphism rings .
the derived equivalence has already been established in @xcite .
2 . in the second step , recasting an idea of @xcite , endomorphism rings are replaced by _ cohomological endomorphism rings _ in the following sense : higher extensions between modules @xmath20 and @xmath21 are shifted morphisms in the derived category , @xmath22)$ ] .
using yoneda multiplication of extensions , this defines an algebra structure on the cohomological endomorphism ring , or generalised yoneda algebra , @xmath23)$ ] .
when @xmath20 is a complex , or any object in a triangulated category @xmath9 , negative degrees @xmath24 may occur .
the main theorem provides derived equivalences between such generalised yoneda algebras .
the construction works , however , not only for these yoneda algebras , but also for perforated ones in the following sense : choose a subset @xmath25 .
then , under some associativity constraint requiring @xmath1 to be admissible ( see subsection [ 2.3 ] ) , the space @xmath26)$ ] is an associative algebra , that in general is neither a subalgebra nor a quotient algebra of the yoneda algebra @xmath27)$ ] .
this algebra is called a _ @xmath1-yoneda algebra _ or a _ @xmath1-perforated yoneda algebra_. we will use the notation @xmath28 for the algebra @xmath29)$ ] , where @xmath30 is any object in @xmath9 .
+ the assumptions of the first step get modified by using cohomological approximations , in the degrees specified by @xmath1 , instead of approximations in degree zero only .
auslander - reiten triangles still satisfy these properties . adding higher extensions requires also to add an orthogonality assumption without which the result would be wrong : assume @xmath31 ) = 0 = { { \rm hom } } ( y , m[j])$ ] for all @xmath32 .
for the sake of exposition also assume for a moment that the above triangle @xmath33 $ ] is in a derived module category and it is induced from an exact sequence with corresponding properties .
then there are derived equivalences between @xmath1-yoneda algebras @xmath34 .
3 . this result needs to be modified , if the triangle is not induced by an exact sequence any more .
then some annihilators have to be factored out of the degree zero parts of the cohomological endomorphism rings , and the derived equivalences are connecting the quotient algebras @xmath35 and @xmath36 . here , the ideals @xmath37 and @xmath38 can be described as follows : let @xmath39 and @xmath40 the idempotent element in @xmath41 corresponding to the direct summand @xmath7 .
then @xmath38 is the submodule of the left @xmath41-module @xmath42 , which is maximal with respect to @xmath43 .
let @xmath44 , and @xmath45 the idempotent in @xmath46 corresponding to the direct summand @xmath7 .
then @xmath37 is the submodule of the right @xmath46-module @xmath47 , which is maximal with respect to @xmath48 .
+ another , equivalent , description of @xmath37 and @xmath38 is that @xmath37 consists of all elements @xmath49 such that @xmath50 for @xmath51 and @xmath52 factorises through @xmath14 and @xmath53 , and @xmath38 consists of all elements @xmath54 such that @xmath55 for @xmath51 and @xmath56 factorises through @xmath14 and @xmath57 , where @xmath58 is the diagonal morphism diag@xmath59 , and @xmath60 is the skew - diagonal morphism skewdiag@xmath61 .
4 . the fourth level of generalisation allows to replace the shift functor by any auto - equivalence of the triangulated category @xmath62 , thus providing a new and versatile meaning of higher extensions in terms of morphisms with one variable shifted by powers of the auto - equivalence .
the additional datum @xmath63 gets mentioned , when necessary , in the notation as an additional superscript , as in @xmath64 . in this general form ,
the main theorem is as follows : let @xmath1 be an admissible subset of @xmath65 , and let @xmath9 be a triangulated @xmath10-category and @xmath7 an object in @xmath9 .
assume that @xmath63 is an invertible triangle functor from @xmath9 to itself .
suppose that @xmath66\ ] ] is a triangle in @xmath62 such that @xmath67 the morphism @xmath12 is a left @xmath68-approximation of @xmath15 and @xmath13 is a right @xmath69-approximation of @xmath16 , @xmath70 @xmath71 for all @xmath51 .
then @xmath72 and @xmath73 are derived equivalent , where @xmath37 and @xmath38 are the above ideals of the @xmath1-yoneda algebras @xmath74 and @xmath75 , contained in @xmath76 and @xmath77 , respectively .
[ thm1 ] a fifth level of generalisation , using two functors @xmath63 and @xmath78 , will be discussed in the appendix .
a further generalisation of some results in this paper to @xmath0-angulated categories introduced in @xcite will be considered in @xcite .
the second level of generality , where @xmath63 is the shift functor and both @xmath37 and @xmath38 are zero , is already widely applicable .
this case happens frequently for the derived category @xmath79 of an @xmath10-algebra @xmath80 .
let @xmath1 be an admissible subset of @xmath81 , and let @xmath80 be an @xmath10-algebra and @xmath7 an @xmath80-module . if @xmath82 is an exact sequence in @xmath83 such that @xmath12 is a left @xmath84-approximation of @xmath15 and @xmath13 is a right @xmath85-approximation of @xmath16 in @xmath79 , and that @xmath86 for all @xmath51 , then the @xmath1-yoneda algebras @xmath87 and @xmath88 are derived equivalent .
[ cor2 ] these results partly generalise some results of @xcite .
the setup here , and the main result , covers , combines and extends several classical concepts : auslander algebras - endomorphism rings of direct sums of all modules of an algebra of finite representation type - are the ingredients of the celebrated auslander correspondence , characterising finite representation type via homological dimensions .
auslander algebras of derived equivalent algebras are , in general , not derived equivalent ; positive results in this direction - for self - injective algebras of finite representation type - previously have been obtained in @xcite . in the current approach
new results can be obtained by appropriate choices of @xmath89 .
another intensively studied class of algebras is that of yoneda algebras , that is , algebras of self - extensions of a semisimple module , or more generally of any module .
apparently , the constructions in corollary [ cor2 ] and in @xcite provide the first general class of derived equivalences for yoneda algebras .
perforated yoneda algebras first have been defined in @xcite , under the name @xmath1-auslander - yoneda algebras .
the approach developed there has been based on the existence of particular kinds of derived equivalences for algebras , which then have been used to construct derived equivalences for perforated yoneda algebras .
the main novelty of the present approach is the systematic use of cohomological data , such as cohomological approximations and perforated yoneda algebras .
this relates smoothly with a wide variety of concepts , such as auslander - reiten sequences and triangles , dominant dimension , calabi - yau categories and frobenius categories .
the article is organised as follows . in section [ sect2 ] , we first fix notation , and then recall definitions and basic results on derived equivalences as well as on admissible sets and perforated yoneda algebras . also , we extend the notion of @xmath90-approximation to what we call cohomological @xmath91-approximation with respect to @xmath92 , where @xmath63 is a functor and @xmath1 is a subset of @xmath81 . in section
[ sect3 ] , the main result , theorem [ thm1 ] , is proven and various easier to access situations are described , for which the assumptions of theorem [ thm1 ] are satisfied .
section [ sect4 ] explains how theorem [ thm1 ] applies to a variety of situations : derived categories of artin algebras , frobenius categories and calabi - yau categories . also , the connection to the concept of dominant dimension is explained . in section [ sect5 ] ,
two examples are given to illustrate the results and to show the necessity of some assumptions in theorem [ thm1 ] .
in the appendix , a more general formulation of theorem [ thm1 ] is stated , which involves two functors , in order to add more flexibility with a view to potential future applications .
the authors are grateful to rundong zheng and yiping chen at bnu for carefully reading the first version of the manuscript . the corresponding author c.c .
xi thanks nsfc for partial support .
w. hu is grateful to the alexander von humboldt foundation for a humboldt fellowship .
much of this work has been done during visits of xi and hu to the mathematisches institut , universitt zu kln , in 2010 .
in this section , we shall recall basic definitions and facts which will be needed in the proofs later on . throughout this paper , @xmath10 is a fixed commutative artinian ring with identity . given an @xmath10-algebra @xmath80 , by an @xmath80-module we mean a unitary left @xmath80-module ; the category of all ( respectively , finitely generated ) @xmath80-modules is denoted by @xmath80-mod ( respectively , @xmath83 ) , the full subcategory of @xmath80-mod consisting of all ( respectively , finitely generated ) projective modules is denoted by @xmath80-proj ( respectively , @xmath93 ) .
there is a similar notation for right @xmath80-modules .
the stable module category @xmath94 of @xmath80 is , by definition , the quotient category of @xmath83 modulo the ideal generated by homomorphisms factorising through projective modules in @xmath93 .
an equivalence between the stable module categories of two algebras is called a _ stable equivalence_. an @xmath10-algebra @xmath80
is called an _ artin @xmath10-algebra _ if @xmath80 is finitely generated as an @xmath10-module .
for an artin @xmath10-algebra @xmath80 , we denote by @xmath95 the usual duality on @xmath83 , and by @xmath96 the nakayama functor @xmath97 . for an @xmath80-module @xmath7
, we denote the first syzygy of @xmath7 by @xmath98 , and call @xmath99 the _ heller loop operator _ of @xmath80 .
the transpose of @xmath7 , which is an @xmath100-module , is denoted by @xmath101 .
let @xmath102 be an additive @xmath10-category , that is , @xmath102 is an additive category in which the set of morphisms between two objects in @xmath102 is an @xmath10-module , and the composition of morphisms in @xmath102 is @xmath10-bilinear . for an object @xmath15 in @xmath103 ,
we denote by @xmath104 the full subcategory of @xmath102 consisting of all direct summands of finite direct sums of copies of @xmath15 .
an object @xmath15 in @xmath102 is called an _ additive generator _ for @xmath102 if @xmath105 = add@xmath106 . for two morphisms @xmath107 and @xmath108 in @xmath102 , we write @xmath109 for their composition . for two functors @xmath110 and @xmath111 however , we write @xmath112 for the composition instead of @xmath113 . if @xmath114 is a map between two sets @xmath15 and @xmath16 , we denote the image of @xmath45 by im@xmath115 .
moreover , if @xmath45 is a homomorphism between two abelian groups , we denote the kernel and cokernel of @xmath45 by @xmath116 and @xmath117 , respectively .
recall that a functor @xmath118 is invertible if there is a functor @xmath119 such that @xmath120 and @xmath121 . in this case
we write @xmath122 for @xmath78 .
if @xmath102 = @xmath123 and if @xmath63 is invertible , we say that @xmath63 is an auto - equivalence . if @xmath63 is a functor from @xmath102 to @xmath102 , then we write @xmath124 , and @xmath125 for @xmath126 if @xmath122 exists , and @xmath127 otherwise .
let @xmath9 be a triangulated @xmath10-category with a shift functor [ 1 ] . for two objects @xmath15 and
@xmath16 in @xmath9 , we write sometimes @xmath128 for @xmath129)$ ] .
let @xmath1 be a subset of @xmath65 .
an object @xmath7 ( or a full subcategory @xmath130 ) of @xmath131 is called _ @xmath1-self - orthogonal _ provided that @xmath132 ( or @xmath133 ) for all @xmath134 , where @xmath135 means that @xmath136 for all @xmath137 . in case @xmath138 , we say that @xmath7 is _ self - orthogonal_. for @xmath139 , we say that @xmath7 is @xmath0-self - orthogonal , which is sometimes , perhaps less suggestively , referred to as @xmath0-rigid . replacing the shift functor by a triangle auto - equivalence @xmath63 ,
one may also define the notion of @xmath92-self - orthogonality , but we refrain from introducing this notion here .
let @xmath102 be an additive @xmath10-category . by a complex @xmath140 over @xmath102 we mean a sequence of morphisms @xmath141 between objects @xmath142 in @xmath102 : @xmath143 such that @xmath144 for all @xmath145 ; we write @xmath146 .
for a complex @xmath140 , the _ brutal truncation _
@xmath147 of @xmath140 is a quotient complex of @xmath140 such that @xmath148 is @xmath149 for all @xmath150 and zero otherwise .
similarly , we define @xmath151 . for a fixed @xmath152 ,
we denote by @xmath153 $ ] the complex obtained from @xmath140 by shifting degree by @xmath0 , that is , @xmath154)^0=x^n$ ] .
the category of all complexes over @xmath102 with chain maps is denoted by @xmath155 .
the homotopy category of complexes over @xmath103 is denoted by @xmath156 .
when @xmath102 is an abelian category , the derived category of complexes over @xmath102 is denoted by @xmath157 .
the full subcategories of @xmath158 and @xmath157 consisting of bounded complexes over @xmath103 are denoted by @xmath159 and @xmath160 , respectively . as usual , for an algebra @xmath80 , we simply write @xmath161 for @xmath162 , @xmath163 for @xmath164 and @xmath165 for @xmath166 .
similarly , we write @xmath167 and @xmath79 for @xmath168 and @xmath169 , respectively . for an @xmath10-algebra @xmath80 ,
the categories @xmath163 and @xmath167 are triangulated @xmath10-categories . for basic results on triangulated categories ,
we refer the reader to @xcite and @xcite . the following result , due to rickard ( see ( * ? ? ?
* theorem 6.4 ) ) by a direct approach , and to keller by working in the more general setup of differential graded algebras , is fundamental in the investigation of derived equivalences .
let @xmath170 and @xmath171 be two rings .
the following conditions are equivalent : @xmath172 @xmath173 and @xmath174 are equivalent as triangulated categories ; @xmath175 @xmath176 and @xmath177 are equivalent as triangulated categories ; @xmath178 @xmath179 and @xmath180 are equivalent as triangulated categories ; @xmath181 @xmath182 and @xmath183 are equivalent as triangulated categories ; @xmath184 @xmath171 is isomorphic to @xmath185 , where @xmath186 is a complex in @xmath182 satisfying : @xmath67 @xmath186 is self - orthogonal , that is , @xmath187)=0 $ ] for all @xmath188 , @xmath70 @xmath189 generates @xmath190 as a triangulated category .
[ derequ ] two rings @xmath170 and @xmath171 are called _ derived equivalent _ if the above conditions ( a)-(e ) are satisfied . a complex @xmath186 in @xmath190 as above
is called a _ tilting complex _ over @xmath170 . for artin algebras ,
the above equivalent conditions can be reformulated in terms of finitely generated modules : two artin @xmath10-algebras @xmath80 and @xmath191 are said to be _ derived equivalent _ if their derived categories @xmath79 and @xmath192 are equivalent as triangulated categories . in this case , there is a tilting complex @xmath186 in @xmath193 such that @xmath194 .
let @xmath195 be the set of natural numbers , and let @xmath65 be the set of all integers .
for a natural number @xmath0 or infinity , let @xmath196 .
recall from @xcite that a subset @xmath1 of @xmath197 containing @xmath198 is called an _ admissible subset _ of @xmath197 if the following condition is satisfied : _ if @xmath199 and @xmath200 are in @xmath1 such that @xmath201 , then @xmath202 if and only if @xmath203 . _ any subset @xmath204 of @xmath81 is an admissible subset of @xmath65 .
moreover , for any subset @xmath1 of @xmath81 containing zero and for any positive integer @xmath205 , the set @xmath206 is admissible in @xmath65 ( for more examples , see @xcite ) .
nevertheless , not every subset of @xmath81 containing zero is admissible , for instance , @xmath207 is not admissible .
in fact , this is the smallest non - admissible subset of @xmath81 .
admissible sets were used to define @xmath1-yoneda algebras in @xcite , under the name of @xmath1-auslander - yoneda algebras. the formulation there works more generally for monoid graded algebras . for our purpose in this paper , we restrict to the case of an object in a triangulated category .
let @xmath1 be an admissible subset of @xmath197 , and let @xmath62 be a triangulated @xmath10-category with a shift functor [ 1 ] .
suppose that @xmath63 is a triangle functor from @xmath9 to @xmath208 .
recall that we put @xmath209 for @xmath210 if @xmath122 does not exist .
let @xmath211 be the bi - functor @xmath212 @xmath213 @xmath214 suppose that @xmath215 and @xmath30 are objects in @xmath62 .
let @xmath216 and @xmath217 .
we define a composition as follows : @xmath218 @xmath219 since @xmath1 is admissible , this composition is associative .
thus @xmath220 is an @xmath10-algebra .
it is called the _ @xmath1-yoneda algebra _ or , when @xmath1 is fixed , the _ perforated yoneda algebra _ of @xmath15 with respect to @xmath63 .
then @xmath221 is a left @xmath222-module . when @xmath223 , the algebra @xmath224 is the orbit algebra of @xmath15 under @xmath63 ( see @xcite ) . for convenience
we write @xmath225 for @xmath226 . in case @xmath227 where @xmath80 is a ring with identity , we write @xmath228 for @xmath229 , and @xmath230 for @xmath231 . when @xmath63 coincides with the shift functor , we omit the upper index @xmath63 , and call @xmath232 the @xmath1-yoneda algebra of @xmath15 , without referring to the shift functor .
this is the algebra introduced in @xcite and there called an auslander - yoneda algebra .
the following lemma is essentially taken from ( * ? ? ?
* lemma 3.5 ) , where a variation of it appears .
the proof given there carries over to the present situation .
let @xmath9 be a triangulated @xmath10-category with a triangle endo - functor @xmath63 , and let @xmath233 be an object in @xmath9 .
suppose that @xmath234 , @xmath235 and @xmath236 are in @xmath237 , and that @xmath1 is an admissible subset of @xmath65 .
then , @xmath67 there is a natural isomorphism @xmath238 which sends @xmath239 to the morphism @xmath240 for @xmath241 .
moreover , if @xmath239 and @xmath242 , then @xmath243 .
@xmath70 the functor @xmath244 is faithful .
@xmath245 if @xmath246 for all @xmath247 , then the functor @xmath248 induces an isomorphism of @xmath10-modules : @xmath249 [ extprop ] the properties described in lemma [ extprop ] will be frequently used in the proofs below .
the class of @xmath1-yoneda algebras with respect to a functor includes a large class of algebras , for example the following : ( * a * ) the endomorphism algebra of a module , in particular , the auslander algebras of representation - finite algebras .
here we choose @xmath250 .
( * b * ) the generalised yoneda algebra of a module if we take @xmath251 .
this includes the preprojective algebras ( see @xcite ) and the hochschild cohomology rings of given algebras .
choosing @xmath252 , we get for instance the even hochschild cohomology rings of algebras .
( * c * ) certain trivial extensions : for an artin algebra @xmath80 and an @xmath80-module @xmath7 we choose @xmath253 for @xmath254 an arbitrary natural number .
then @xmath255 is the trivial extension of @xmath256 by the bimodule @xmath257 .
such rings appear naturallly in the ( bounded ) derived category @xmath258 of coherent sheaves of a smooth projective variety @xmath259 over @xmath260 . indeed , if @xmath15 is a @xmath261-spherical object in @xmath258 , then its cohomological ring @xmath262 is @xmath263 , this is a graded ring isomorphic to @xmath264/(t^2)$ ] with @xmath265 of degree @xmath261 . for further information on spherical objects ,
we refer the reader to ( * ? ? ?
* section 3c ) . in general , if @xmath266 such that @xmath267 for @xmath268 , then @xmath269 is the trivial extension of @xmath270 by the bimodule @xmath271 .
note that @xmath272 is admissible . in this case
, we also get a trivial extension .
( * d * ) the polynomial ring @xmath273 $ ] : if we take @xmath274 for @xmath275 , then the perforated yoneda algebra @xmath276/(x^2)}^{\phi}(r)}$ ] is isomorphic to @xmath277 $ ] with @xmath265 a variable . if @xmath278 , then @xmath276/(x^2)}^{\phi}(r)}\simeq r[t]/(t^n)$ ] .
@xmath123-split sequences have been defined in @xcite in the context of constructing derived equivalences between certain endomorphism algebras .
let us recall the definition and a result in @xcite .
let @xmath102 be an additive category and @xmath123 a full subcategory of @xmath102 .
a sequence @xmath279 in @xmath102 is called an _
@xmath123-split sequence _
if @xmath67 @xmath280 , @xmath70 @xmath45 is a left @xmath123-approximation of @xmath15 , and @xmath281 is a right @xmath123-approximation of @xmath16 , and @xmath245 @xmath45 is a kernel of @xmath281 , and @xmath281 is a cokernel of @xmath45 .
typical examples of @xmath123-split sequences are auslander - reiten sequences .
every @xmath282-split sequence provides a derived equivalence ( see ( * ? ? ?
* theorem 1.1 ) ) . here
are some details , for later reference .
@xcite let @xmath102 be an additive category , and @xmath7 an object in @xmath102 .
suppose that @xmath283 is an _
_ add__@xmath284-split sequence in @xmath102 .
then the endomorphism ring @xmath285 of @xmath286 is derived - equivalent to the endomorphism ring @xmath287 of @xmath288 via a tilting module of projective dimension at most @xmath289 .
[ deadss ] now , the question arises whether theorem [ deadss ] can be extended to @xmath1-yoneda algebras .
the second example in the final section demonstrates that this is no longer true if we just replace the endomorphism algebras in theorem [ deadss ] by @xmath1-yoneda algebras .
nevertheless , we shall show that under certain orthogonality conditions , there still is a positive answer .
this will be discussed in detail in the next section .
the condition ( 3 ) of a @xmath123-split sequence are substitutes in this general setup for requiring the short exact sequence to be exact .
since triangles in triangulated categories are replacements of short exact sequences , we may reformulate the notion of @xmath90-split sequences in the following sense for triangulated categories .
let @xmath9 be a triangulated category with a shift functor [ 1 ] , and let @xmath123 be a full additive subcategory of @xmath9 .
a triangle @xmath290\ ] ] in @xmath9 is called a _
@xmath123-split triangle _
if @xmath291 , the map @xmath12 is a left @xmath123-approximation of @xmath15 and the map @xmath13 is a right @xmath123-approximation of @xmath16 .
thus , for an artin @xmath10-algebra @xmath80 , every @xmath123-split sequence in @xmath80-mod extends to a @xmath123-split triangle in @xmath79 .
next , we introduce the left and right cohomological @xmath90-approximations with respect to @xmath92 , which generalise the notions of left and right @xmath123-approximations , respectively .
suppose that @xmath102 is a category with an endo - functor @xmath292 .
let @xmath123 be a full subcategory of @xmath102 , and let @xmath1 be a non - empty subset of @xmath81 .
if @xmath63 has an inverse , then @xmath1 may be chosen to be a subset of @xmath65 .
suppose that @xmath15 is an object of @xmath102 .
a morphism @xmath293 in @xmath102 is called a _ left cohomological @xmath90-approximation _ of @xmath15 with respect to @xmath92 ( or shortly , a left @xmath294-approximation of @xmath15 ) if @xmath295 , and for any morphism @xmath296 with @xmath297 and @xmath298 , there is a morphism @xmath299 such that @xmath300 . here
similarly , we have the notion of a right @xmath301-approximation of @xmath15 in @xmath9 , that is , a morphism @xmath302 with @xmath95 in @xmath123 is called a right @xmath301-approximation of @xmath15 if , for any @xmath298 and any morphism @xmath303 with @xmath304 in @xmath123 , there is a morphism @xmath305 such that @xmath306 .
note that if @xmath307 and @xmath250 , then we get the original notion of approximations in the sense of auslander and smal .
( in ring theory , such approximations are called pre - envelope and precover , respectively ) . moreover ,
if @xmath308 , then every left @xmath301-approximation of @xmath15 is also a left @xmath309-approximation of @xmath15 , and every right @xmath310-approximation of @xmath15 is also a right @xmath123-approximation of @xmath15 . if @xmath311 $ ] and @xmath9 = @xmath79 for an artin algebra @xmath80 , then @xmath312 for all @xmath313-mod and all @xmath314 .
for this reason , a @xmath315-approximation has been called a _
cohomological _ approximation . in this paper ,
we are mainly interested in the case where @xmath102 is a triangulated @xmath10-category @xmath9 with an endo - functor @xmath63 , and @xmath123 is a full subcategory of @xmath9 .
thus , a morphism @xmath316 with @xmath295 and @xmath317 is a left @xmath318-approximation of @xmath15 if and only if the canonical map @xmath319 , defined by @xmath320 , is surjective for all @xmath297 .
similarly , a morphism @xmath321 with @xmath295 and @xmath317 is a right @xmath301-approximation of @xmath15 if and only if the canonical map @xmath322 is surjective for every @xmath297 and @xmath323 .
if , moreover , @xmath63 is an invertible triangle functor , then a morphism @xmath321 with @xmath295 and @xmath324 is a right @xmath301-approximation of @xmath15 if and only if the canonical map @xmath325 is surjective for all @xmath297 .
note that here we need the minus sign for @xmath1 and that @xmath122 exists .
if @xmath63 coincides with the shift functor [ 1 ] , we simply speak of @xmath326-approximations , without mentioning @xmath63 .
note that if @xmath1 contains zero and if @xmath327 for all @xmath51 and @xmath328 , then @xmath45 is a left @xmath294-approximation of @xmath15 if and only if @xmath45 is a left @xmath123-approximation of @xmath15 .
a dual statement is also true for a right @xmath301-approximation of @xmath15 .
here is a source of examples of @xmath329-approximations .
suppose that @xmath9 = @xmath79 for @xmath80 an artin @xmath10-algebra and that @xmath1 is a subset of @xmath65 .
let @xmath330 $ ] be an auslander - reiten triangle in @xmath9 .
if neither @xmath15 nor @xmath16 belongs to @xmath331)$ ] for every @xmath51 , then @xmath12 is a left @xmath332-approximation of @xmath15 , and @xmath13 is a right @xmath332-approximation of @xmath16 . finally , we note the difference of a left @xmath310-approximation of @xmath15 from a left @xmath333-approximation of @xmath15 in the sense of auslander and smal , where @xmath334 is the full subcategory of @xmath9 with all objects in @xmath335 for all @xmath298 .
suppose @xmath308 .
then a @xmath310-approximation is a @xmath336-approximation , but the converse is not true in general .
if @xmath337 , then the two concepts are independent .
so , roughly speaking , a cohomological @xmath123-approximation with respect to @xmath92 emphasises not only the factorisations but also that the object belongs to the given subcategory @xmath123 ( and not to @xmath335 for @xmath51 ) .
in this section , we shall prove theorem [ thm1 ] and derive some consequences and some simplifications in special cases .
suppose that @xmath9 is a triangulated @xmath10-category with a shift functor [ 1 ] , and @xmath7 is an object in @xmath9 .
suppose that @xmath63 is an auto - equivalence of @xmath9 , which may be different from the shift functor .
for a subset @xmath1 of @xmath65 , we define @xmath338 , and @xmath339 @xmath340 let @xmath0 be a positive integer . for simplicity
, we write @xmath341 for @xmath342 , and @xmath343 for @xmath344 if @xmath9 is clear in the context .
similarly , the notations @xmath345 and @xmath346 are defined . as usual , @xmath63 is omitted in notation when it coincides with the shift functor . given a triangle
@xmath347 $ ] in @xmath62 with @xmath348 , we define @xmath349 , \quad \bar{w}=(0 , w)^t : m\oplus y{\longrightarrow}x[1],\ ] ] where @xmath350 stands for the transpose of the matrix @xmath351 , and @xmath352 \big\},$ ] @xmath353 sets @xmath37 and @xmath38 are indeed independent of @xmath63 and @xmath354 , and contained in @xmath355 and @xmath356 , respectively .
the main result of this paper is the following theorem which is a reformulation of theorem [ thm1 ] .
let @xmath1 be an admissible subset of @xmath65 , let @xmath9 be a triangulated @xmath10-category with an auto - equivalence @xmath63 , and let @xmath7 be an object in @xmath9 .
suppose that @xmath66\ ] ] is a triangle in @xmath62 such that the morphism @xmath12 is a left @xmath357-approximation of @xmath15 , that the morphism @xmath13 is a right @xmath358-approximation of @xmath16 and that @xmath359 and @xmath360 .
then the algebras @xmath72 and @xmath73 are derived equivalent .
[ verygthm1 ] _ proof .
_ let @xmath361 and @xmath362 .
set @xmath363;\ ] ] @xmath364.\ ] ] then there are two triangles in @xmath9 : @xmath365 , \end{cd}\ ] ] @xmath366@>{-\widetilde{w}[-1 ] } > > v@>{\widetilde{\alpha } } > > m_1\oplus m@>{\widetilde{\beta } } > > y. \end{cd}\ ] ] since @xmath63 is a triangle functor , there is a natural isomorphism @xmath367{\rightarrow}[1]f$ ] .
that is , for any object @xmath15 in @xmath9 , there is an isomorphism @xmath368){\rightarrow}(fx)[1]$ ] , which is natural in @xmath15 .
the isomorphism @xmath369){\longrightarrow}(f^ix)[j]$ ] is denoted by @xmath370 . in part
ii of this article , there will be further discussion of these natural transformations .
@xmath67 for any morphism @xmath371 with @xmath298 , there is a morphism @xmath372{\rightarrow}(f^iy)[-1]$ ] such that @xmath373)x_i = t_i\delta(f , i , y,-1)^{-1}\big(f^i(\widetilde{w}[-1])\big).$ ] @xmath70 for any morphism @xmath374 with @xmath298 , there is a morphism @xmath375{\rightarrow}(f^ix)[1]$ ] such that @xmath376 = @xmath377 [ easy ] _ proof . _
( 1 ) note that @xmath58 is a left @xmath378-approximation of @xmath379 .
thus , given @xmath371 , there is a morphism @xmath380 such that @xmath381 .
since @xmath63 is a triangle functor , the second triangle implies that there is a triangle ( see @xcite ) @xmath382 @>{\delta(f , i , y,-1)^{-1}\big(-f^i(\widetilde{w}[-1])\big)}>>f^iv @>{f^i{\widetilde{\alpha}}}>>f^i(m_1\oplus m)@>{f^i{\widetilde{\beta}}}>>f^iy . \end{cd}\ ] ] thus there is a morphism @xmath372{\rightarrow}(f^iy)[-1]$ ] such that @xmath373)x_i = t_i\delta(f , i , y,-1)^{-1}\big(f^i(\widetilde{w}[-1])\big).$ ] \(2 ) the proof of ( 2 ) is similar to that of ( 1 ) , using the following triangle @xmath383.\end{cd}\ ] ] @xmath384 now we prove that the quotient rings in theorem [ verygthm1 ] are well - defined .
the @xmath37 and @xmath38 appearing in theorem [ verygthm1 ] are ideals of @xmath385 and @xmath386 , respectively .
[ ideal ] _ proof .
_ by definition , a morphism @xmath387 factorises through @xmath14 if and only if there is an object @xmath388 in @xmath14 and there are two morphisms @xmath389 and @xmath390 such that @xmath391 . a morphism @xmath392 factorises through @xmath393 : y[-1]{\rightarrow}v$ ] if and only if there is a morphism @xmath394 $ ] such that @xmath395\big)$ ] . in the following , we shall prove that @xmath37 is an ideal in @xmath396 .
the set @xmath37 is closed under addition in @xmath385 . to show that @xmath37 is a two - sided ideal in @xmath385 , we pick an @xmath397 and a @xmath398 , and calculate the products @xmath399 and @xmath400 in @xmath385 .
note that @xmath401 and @xmath402 since @xmath50 for @xmath51 .
we write @xmath403 for @xmath404 and @xmath405 , where @xmath388 is an object in @xmath14 , and @xmath406)$ ] for a morphism @xmath407 $ ] .
we first show that @xmath37 is a right ideal .
\(1 ) let @xmath408 .
the map @xmath409 factorises through an object in @xmath14 .
since @xmath52 factorises through @xmath393 $ ] , it follows from lemma [ easy ] ( 1 ) that @xmath409 factorises also through @xmath393 $ ] .
\(2 ) let @xmath51 . in this case , @xmath410 by the assumption @xmath411 .
let @xmath412 and @xmath413 be the projections of @xmath379 onto @xmath15 and @xmath7 , respectively
. then the composition @xmath414 belongs to @xmath415 .
thus @xmath416 . by lemma
[ easy ] ( 1 ) , there is a morphism @xmath372{\rightarrow}f^iy[-1]$ ] such that @xmath373)y_i= t_i\delta(f , i , y,-1)^{-1}f^i(\widetilde{w}[-1])$ ] .
hence @xmath417)y_i(f^ip_m)= st_i\delta(f , i , y,-1)^{-1}f^i(\widetilde{w}[-1])(f^ip_m)\\ & \\ & = st_i\delta(f , i , y,-1)^{-1}f^i\big(\widetilde{w}[-1]p_m\big ) \\ & \\ & = st_i \delta(f , i , y,-1)^{-1}f^i\big((w[-1],0){\tiny\left({{0}\atop{1_m}}\right)}\big)=0.\end{array}\ ] ] altogether , @xmath418 for @xmath51 .
hence @xmath419 , and @xmath37 is a right ideal in @xmath396 .
next , we show that @xmath37 is a left ideal , that is , we check @xmath420 .
\(3 ) the map @xmath421 factorises through an object in @xmath14 and through @xmath393 $ ] .
\(4 ) let @xmath51 .
note that @xmath422 is a left @xmath357-approximation of @xmath379 .
thus there is a morphism @xmath423 such that @xmath424 . by assumption ,
this implies that @xmath426 , and therefore @xmath427 . since @xmath428)(f^ip_m)=0 $ ]
, we get @xmath429 .
thus @xmath430 for @xmath51 . hence @xmath431 , and @xmath37
is a left ideal in @xmath432 .
thus @xmath37 is an ideal in @xmath385 .
similarly , @xmath38 is an ideal in @xmath386 .
@xmath384 we know that @xmath433 is a @xmath396-module for any object @xmath30 in @xmath9 .
the next lemma shows that the ideal @xmath37 of @xmath385 may annihilate some modules of this form .
keep the notations as above . then @xmath67 @xmath434 .
@xmath70 @xmath435 \}.$ ] for @xmath436 with @xmath437 , we have @xmath438 if and only if @xmath50 for all @xmath51 and @xmath52 factorises through @xmath14 and @xmath439 $ ] , where @xmath440 is defined in lemma [ extprop ] ( 1 ) .
@xmath441 let @xmath442 with @xmath443 .
then @xmath444 if and only if @xmath45 factorises through @xmath439 $ ] .
[ vanish ] _ proof . _
( 1 ) we denote by @xmath445 the canonical inclusion .
let @xmath446 and @xmath447 .
then @xmath448 since @xmath50 for @xmath51 .
since @xmath37 is an ideal in @xmath385 , it follows that @xmath449 . by the definition of @xmath37
, we have @xmath450 for all @xmath51 and @xmath451 factorises through @xmath393 $ ] . moreover , @xmath452p_m\big)\lambda_m= s\cdot 0 \cdot \lambda_m=0 $ ] , where @xmath453 is a morphism from @xmath379 to @xmath454 $ ] .
hence @xmath450 , and @xmath455 for all @xmath456 . thus ( 1 ) follows .
\(2 ) let @xmath457 be the canonical inclusion .
as in case ( 1 ) , it follows that , for @xmath446 and @xmath458 , we have @xmath448 , and that @xmath459 factorises through @xmath393 $ ] and @xmath14 .
hence @xmath460)p_x)=s(w[-1])$ ] , where @xmath453 is a morphism from @xmath379 to @xmath454 $ ] .
conversely , let @xmath461 , and suppose that @xmath50 for all @xmath134 and that @xmath52 factorises through @xmath14 and @xmath439 $ ] . for @xmath462 in @xmath131 ,
we denote by @xmath463 the element of @xmath464 concentrated only in degree @xmath308 .
then it is straightforward to check that @xmath465 belongs to @xmath37 .
thus , @xmath466 .
\(3 ) first , suppose @xmath467 and @xmath468 .
then @xmath469 , the image of @xmath470 under @xmath471 , belongs to @xmath472 .
thus , by ( 2 ) , we have @xmath50 for all @xmath473 and that @xmath52 factorises through @xmath14 and @xmath439 $ ] .
conversely , suppose that @xmath474 . then , for any @xmath475 , the image of @xmath476 under @xmath471 is @xmath477 .
since @xmath478 is a @xmath385-submodule of @xmath479 , we have @xmath480 .
secondly , suppose that @xmath481 is a direct sum of @xmath0 copies of @xmath379 , and @xmath482 .
we identify @xmath483 with @xmath484 , and write @xmath485 , a column matrix with @xmath486 .
then the image of @xmath471 is the sum of the image of @xmath487 for @xmath488 .
now the conclusion follows from the first case .
finally , suppose that @xmath481 is a direct summand of @xmath0 copies of @xmath379 , that is , @xmath489 . if @xmath490 , then we may consider @xmath491 as an element in @xmath492 .
then the proof is reduced to the second case .
\(4 ) follows from ( 3 ) because of @xmath493 .
let @xmath494 be the complex @xmath495 where the term @xmath479 is in degree zero .
then it is the direct sum of the following two complexes @xmath496 let @xmath497 , and let @xmath498 be the canonical surjection . then , by lemma [ vanish ] ( 1 ) , we may write @xmath499 with @xmath500 . the complex @xmath501 in @xmath502 is the direct sum of the complexes @xmath503 each term of @xmath186 is a finitely generated projective @xmath504-module . before proceeding further
, we need to introduce some more notation .
set @xmath505 where @xmath37 and @xmath38 are defined just before theorem [ verygthm1 ] .
@xmath186 is a tilting complex over @xmath506 .
[ tiltcpx ] _ proof .
_ it is clear that @xmath507 ) = 0 $ ] for @xmath508 and for @xmath509 .
we have to check that @xmath510 ) = 0 $ ] and @xmath511 ) = 0.$ ] let @xmath512)$ ] .
consider the following diagram : @xmath513 since both @xmath15 and @xmath514 are in @xmath515 , lemma [ extprop ] ( 1 ) provides an isomorphism @xmath516 and an element @xmath517 such that @xmath518 . by assumption
, @xmath519 is a left @xmath68-approximation of @xmath15 .
this yields for each @xmath298 a morphism @xmath520 such that @xmath521 . clearly , @xmath522 , and @xmath523
now , we have to check the following diagram is commutative : @xmath524 in fact , if @xmath525 , then it is sent to @xmath526 by @xmath527 , and further sent to @xmath528 by @xmath529 .
an easy calculation shows that @xmath530 , the image of @xmath531 under @xmath532 thus the diagram is commutative , and @xmath533 this means that @xmath534 - since @xmath535 is surjective - and that @xmath536 in @xmath537 .
therefore @xmath510 ) = 0 $ ] .
let @xmath538)$ ] .
consider the following diagram : @xmath539 since @xmath535 is surjective and @xmath540 is projective in @xmath170-mod , @xmath541 can be lifted along @xmath535 , say @xmath542 with @xmath543 . by assumption , we have @xmath359 , and , by lemma [ extprop ] ( 3 ) , there is a homomorphism @xmath544 such that @xmath545 .
thus @xmath546 lemma [ extprop ] ( 2 ) implies @xmath547 .
therefore @xmath548 factorises through @xmath549 $ ] . by lemma [ vanish ] ( 4 )
, the image of @xmath550 is contained in @xmath478 .
it follows that @xmath551 and @xmath536 .
hence @xmath511 ) = 0.$ ] now , @xmath189 generates @xmath552 as a triangulated category .
thus @xmath186 is a tilting complex over @xmath506 .
@xmath384 to get a tilting complex from @xmath494 , one may consider the ideal @xmath553 of @xmath554 consisting of all endomorphisms @xmath555 which are of the form @xmath109 with @xmath556 and @xmath557 such that @xmath558 and @xmath559 .
then it is easy to show that the quotient complex of @xmath494 modulo @xmath560 is a two - term tilting complex over @xmath561 .
we will not use this complex because its endomorphism algebra can not be described in a nice way . note that the ideal @xmath553 of @xmath562 is properly contained in @xmath37 in general .
the two rings @xmath563 and @xmath564 are isomorphic .
_ since @xmath506 is a quotient algebra of @xmath170 , the category @xmath565 can be viewed as a full subcategory of @xmath566 , and it follows that @xmath567 can be viewed as a full subcategory of @xmath568 .
thus , we have an isomorphism @xmath569 . to prove the lemma , we shall construct an isomorphism from @xmath570 to @xmath563 .
let @xmath571 .
since @xmath572 is an epimorphism and @xmath573 is a projective @xmath170-module , there is a @xmath170-module homomorphism @xmath574 such that @xmath575 .
let @xmath576 and @xmath577 for all @xmath578 .
then it follows from @xmath579 that @xmath580 is an endomorphism in @xmath581 . by lemma [ extprop ] ( 1 )
, we can assume that @xmath582 and @xmath583 with @xmath584 and @xmath585 .
now , it follows from @xmath586 that @xmath587 for each @xmath588 , we can form the following commutative diagram in @xmath131 : @xmath589 \\
@v{x_i}vv @vv{y_i}v @vv{h_i}v @vv{x_i[1]}v\\ f^ix @>{f^i\bar{\alpha}}>>f^i(m_1\oplus m ) @>{f^i\bar{\beta } } > > f^iw @>{(f^i\bar{w})\delta(f , i , x,1 ) } > > ( f^ix)[1].\\ \end{cd}\ ] ] for some morphism @xmath590 . thus , for each @xmath571 , we get an element @xmath591 which is @xmath386 by definition .
this leads us to defining the following correspondence : @xmath592 _ claim 1 .
_ @xmath593 is well - defined .
+ _ proof .
_ suppose that @xmath571 is null - homotopic , that is , there is a map @xmath594 such that @xmath595 and @xmath596 .
since @xmath535 is surjective and @xmath540 is projective in @xmath566 , there is a map @xmath597 such that @xmath598 .
hence @xmath599 and @xmath600 . by the assumption @xmath411 , lemma [ extprop ] ( 3 )
yields a map @xmath601 such that @xmath602 .
therefore , @xmath603 and @xmath604 .
consequently , @xmath605 and @xmath606 .
thus @xmath55 for all @xmath134 and @xmath607 . by lemma [ vanish ] ( 3 )
, we have @xmath50 for all @xmath134 and @xmath608 for some morphisms @xmath609 and @xmath610 with @xmath558 . since @xmath519 is a left @xmath14-approximation of @xmath15 ,
there is a morphism @xmath611 such that @xmath612 .
it follows that @xmath613 now we consider the commutative diagram @xmath614 .
suppose @xmath615 .
then we have shown that @xmath616 .
hence @xmath617 .
this implies that @xmath618 factorises through @xmath619 , and , consequently , that @xmath620 since @xmath621 .
it follows from @xmath622)=0 $ ] that @xmath623 factorises through @xmath624 . since @xmath625
, we get @xmath626 .
altogether , we have shown that @xmath627 for all @xmath134 .
now consider the diagram @xmath614 in case @xmath408 .
first , we have @xmath628 , which means @xmath629 factorises through @xmath619 .
second , since @xmath630)=\bar{w}(\bar{\alpha}[1])(cb+t)[1]=0 $ ] , the morphism @xmath629 factorises through @xmath514 which is in @xmath14 .
thus , @xmath631 and @xmath632 is zero in @xmath563 .
this shows that @xmath593 is well - defined .
_ claim 2 .
_ @xmath593 is injective .
+ _ proof .
_ suppose that @xmath633 .
then @xmath631 , that is , @xmath627 for all @xmath134 , and @xmath629 factorises through both @xmath619 and @xmath14 .
suppose @xmath634 for a morphism @xmath635{\longrightarrow}w$ ] .
for each @xmath134 , since @xmath636 , the morphism @xmath637 factorises through @xmath638 , and consequently @xmath55 for all @xmath134 since @xmath639 . for each @xmath134
, it follows from @xmath640)=h_i(f^i\bar{w})\delta(f , i , x,1)=0 $ ] that @xmath641 $ ] factorises through @xmath642 $ ] , or equivalently , the morphism @xmath643 factorises through @xmath514 .
hence @xmath50 for all @xmath134 since @xmath639 .
now we consider the case @xmath408 .
first , we have @xmath644 , which implies @xmath607 for a morphism @xmath601 .
second , @xmath645 .
it follows that @xmath646 , and therefore @xmath647 factorises through @xmath549 $ ] .
since @xmath648 factorises through @xmath14 and since @xmath649 is a right @xmath14-approximation of @xmath650 , we see that @xmath629 factorises through @xmath60 , say @xmath651 for some @xmath652 .
thus , @xmath653 ) = h_0\bar{w } = r\bar{\beta}\bar{w } = 0 $ ] , or equivalently , @xmath654)x_0=0 $ ] .
it follows that @xmath52 factorises through @xmath514 .
since @xmath655 also factorises through @xmath514 , we see that @xmath647 factorises through @xmath14 .
thus we have shown that @xmath647 factorises through both @xmath14 and @xmath549 $ ] .
now , by lemma [ vanish ] ( 3 ) , we have @xmath656 .
hence @xmath657 this implies that @xmath658 since @xmath535 is surjective . moreover
, one can check that @xmath659 hence @xmath660 is null - homotopic , and consequently @xmath593 is injective .
_ claim 3 .
_ @xmath593 is surjective .
+ _ proof .
_ let @xmath661 with @xmath623 for @xmath298 .
since @xmath60 is a right @xmath358-approximation of @xmath650 , we have a morphism @xmath662 such that @xmath663 for @xmath298 .
this means that there is a commutative diagram @xmath664 \\
@v{x_i}vv @vv{y_i}v @vv{h_i}v @vv{x_i[1]}v\\ f^ix @>{f^i\bar{\alpha}}>>f^i(m_1\oplus m ) @>{f^i\bar{\beta } } > > f^iw @>{(f^i\bar{w})\delta(f , i , x,1 ) } > > f^ix[1].\\ \end{cd}\ ] ] now , define @xmath665 , @xmath666 ; @xmath667 , @xmath668 and @xmath669 for @xmath670 . then @xmath671 belongs to @xmath581 .
since @xmath672 takes elements in @xmath478 to elements in @xmath472 , the image of @xmath673 under the map @xmath674 is zero , and consequently , there is a unique map @xmath675 such that @xmath676 .
now we have @xmath677 hence @xmath678 since @xmath535 is surjective . defining @xmath679 and @xmath680 for all @xmath681
, we see that @xmath682 is an endomorphism in @xmath570 and @xmath683 .
thus @xmath684 is surjective .
_ claim 4 .
_ @xmath593 is an @xmath10-algebra homomorphism .
+ _ proof . _
the map @xmath593 is @xmath10-linear , so it preserves addition . for multiplication , we take @xmath660 and @xmath685 in @xmath570 .
let @xmath686 and @xmath687 be in @xmath581 such that @xmath575 , @xmath688 , @xmath689 and @xmath690 .
suppose that @xmath691 and @xmath692 with @xmath693 and @xmath694 .
let @xmath695 and @xmath696 be in @xmath171 making the diagram @xmath614 commutative , that is , @xmath697)=h_i(f^i\bar{w})\delta(f , i , x,1 ) , } { \bar{\beta}h'_i = y'_if^i\bar{\beta } , \quad \bar{w}(x'_i[1])=h'_i(f^i\bar{w})\delta(f , i , x,1)}\ ] ] for all @xmath588 .
then , by definition , we have @xmath683 , @xmath698 and @xmath699 now we calculate @xmath700 .
let @xmath701 .
then @xmath702 , @xmath703 , and @xmath704 , where @xmath705 , and @xmath706 .
for each @xmath707 , one has to check that @xmath708 however , this follows from @xmath709 similarly , for each @xmath707 , we have @xmath710\big).\ ] ] this means @xmath711 .
thus @xmath593 is a ring homomorphism , and the proof of theorem [ verygthm1 ] is finished .
@xmath384 before proceeding ,
we comment on the conditions in theorem [ verygthm1 ] .
( * a * ) let @xmath712 $ ] be a triangle in @xmath9 with @xmath713 , @xmath411 and @xmath714 . if @xmath12 is a left @xmath68-approximation of @xmath15 , then @xmath715 for @xmath51 .
similarly , if @xmath13 is a right @xmath716-approximation of @xmath16 , then @xmath717 for @xmath473 . in particular
, if @xmath7 is an @xmath92-self - orthogonal object of @xmath9 , that is , @xmath718 for every @xmath473 , and if @xmath12 is a left @xmath357-approximation of @xmath15 and @xmath13 is a right @xmath716-approximation of @xmath16 , then @xmath719 and @xmath720 .
( * b * ) under the conditions of theorem [ verygthm1 ] , there are isomorphisms @xmath721 for every @xmath51 .
in fact , this follows from the following general statement : let @xmath9 be a triangulated category with a shift functor [ 1 ] .
suppose that @xmath63 is a triangle functor from @xmath9 to itself , and that @xmath123 is a full subcategory of @xmath9 .
let @xmath722 be a positive integer .
suppose that @xmath723\ ] ] is a triangle in @xmath9 , such that @xmath724 is a left @xmath725-approximation of @xmath726 , and that @xmath727 is surjective for every @xmath728 and @xmath729 .
if @xmath730 for @xmath731 , then @xmath732 . from
the given two triangles the following exact commutative diagram can be formed : @xmath733)@>>>{{\rm hom } } _ { \cal t}(y_1 , f^id_2[1 ] ) \\
@vvv @vvv @vv{\qquad\qquad \quad(*)}v @vvv\\ { { \rm hom } } _ { \cal t}(d_1 , f^id_2 ) @>{(d_1,f^i(\beta_2))}>>{{\rm hom } } _ { \cal t}(d_1 , f^iy_2)@>0>>{{\rm hom } } _ { \cal t}(d_1 , f^ix_2[1])@>>>{{\rm hom } } _ { \cal t}(d_1 , f^id_2[1 ] ) . \\ \end{cd}\ ] ] since @xmath734 by assumption and since @xmath735 and @xmath736 are surjective by the property of approximation , the conclusion follows from the commutative square ( @xmath737 ) .
@xmath384 ( * c * ) let @xmath738 $ ] be an @xmath14-split triangle in @xmath9 . define @xmath739 , @xmath740 , @xmath741 , and @xmath742 .
let @xmath37 and @xmath38 be as defined in theorem [ verygthm1 ] .
then the ideals @xmath37 and @xmath38 in theorem [ thm1 ] have the following characterisation : \(i ) let @xmath40 be the idempotent in @xmath41 corresponding to the direct summand @xmath7 of @xmath650 . then @xmath38 is the submodule of the left @xmath41-module @xmath42 , which is maximal with respect to @xmath43 .
\(ii ) let @xmath45 be the idempotent in @xmath46 corresponding to the direct summand @xmath7 of @xmath379 .
then @xmath37 is the submodule of the right @xmath46-module @xmath47 which is maximal with respect to @xmath48 . by lemma [ ideal ] , the sets @xmath37 and @xmath38 are ideals of @xmath46 and @xmath41 , respectively .
\(i ) let @xmath743 and @xmath744 be the canonical projection and injection , respectively . by definition , @xmath745 .
the set @xmath42 is precisely the set of all endomorphisms of @xmath650 that factorise through @xmath14 .
the endomorphisms of @xmath650 factorising through @xmath619 are those endomorphisms @xmath469 that satisfy @xmath746 , and consequently @xmath747 . hence @xmath38 is a submodule of @xmath748 with @xmath43 .
suppose that @xmath749 is another submodule containing @xmath38 with @xmath750 .
then @xmath750 implies @xmath751 , and consequently @xmath752 for all @xmath558 . for each @xmath753 , the image of the morphism @xmath754
is contained in @xmath755 since @xmath755 is a left ideal of @xmath41 .
thus , the morphism @xmath756 is a @xmath41-module morphism from @xmath757 to the image of @xmath758 .
hence @xmath759 , and consequently @xmath746 .
this implies @xmath760 .
this proves ( i ) .
\(ii ) the proof is similar to that of ( i ) .
@xmath384 a special case of theorem [ verygthm1 ] is the following corollary .
let @xmath1 be an admissible subset of @xmath65 , and let @xmath9 be a triangulated @xmath10-category with an auto - equivalence @xmath63 , and let @xmath7 be an object in @xmath9 .
suppose that @xmath761 $ ] is an @xmath14-split triangle in @xmath62 , and suppose that @xmath15 and @xmath16 both are in @xmath762 .
then @xmath72 and @xmath763 are derived equivalent .
[ cortoverygthm1 ] the following special case of theorem [ verygthm1 ] is useful to construct explicit examples .
let @xmath9 be a triangulated @xmath10-category with @xmath764 $ ] the shift functor , and let @xmath7 be an object in @xmath9 .
suppose that @xmath712 $ ] is a triangle in @xmath62 such that @xmath348 , and suppose that @xmath765 and @xmath766 .
then , for any admissible subset @xmath1 of @xmath767 , the algebras @xmath35 and @xmath768 are derived equivalent . [ 3.8 ] _ proof . _
we show that @xmath13 is a right @xmath85-approximation of @xmath16 .
note that , for @xmath298 , we always have @xmath769 . hence @xmath770)=0 $ ] for @xmath456 . now apply @xmath771,-)$ ] with @xmath298 to the triangle
@xmath712 $ ] : @xmath772,m_1){\longrightarrow}{{\rm hom } } _ { \cal t}(m[-i],y){\longrightarrow}{{\rm hom } } _ { \cal t}(m[-i ] , x[1]){\rightarrow}\cdots\ ] ] because of @xmath771,x[1])= { { \rm hom } } _ { \cal t}(m , x[i+1])=0 $ ] , the map @xmath13 is a right @xmath773-approximation of @xmath16 .
similarly , it follows from @xmath774 for @xmath456 that @xmath12 is a left @xmath332-approximation of @xmath15 .
now corollary [ 3.8 ] follows from theorem [ verygthm1 ] .
@xmath384 another special case of theorem [ verygthm1 ] is that @xmath775 and @xmath776 .
here is a condition when the ideals @xmath37 and @xmath38 in theorem [ verygthm1 ] vanish .
let @xmath777 $ ] be an @xmath14-split triangle in @xmath9 .
define @xmath739 , @xmath740 , @xmath741 , and @xmath742 .
let @xmath778 be the ideal of @xmath46 consisting of all @xmath387 that factorises through @xmath393 : y[-1]{\rightarrow}v$ ] , and let @xmath779 be the ideal of @xmath41 consisting of all @xmath780 that factorises through @xmath781 $ ] .
@xmath67 suppose that @xmath46 is an artin algebra . if @xmath782 , then @xmath783 .
@xmath70 suppose that @xmath41 is an artin algebra . if @xmath784 , then @xmath785 . [ bigideal ] _ proof .
_ we prove ( 1 ) .
the proof of ( 2 ) is similar to that of ( 1 ) , and we omit it .
we have a triangle @xmath454{\stackrel{-\widetilde{w}[-1]}{{\longrightarrow}}}v{\stackrel{\widetilde{\alpha}}{{\longrightarrow}}}m_1\oplus m{\stackrel{\widetilde{\beta}}{{\longrightarrow } } } y$ ] , apply @xmath786 to this triangle , and get the following exact sequence of right @xmath46-modules : @xmath787 where @xmath788 is the cokernel of @xmath789 now , applying @xmath790 to the above exact sequence , we get another exact sequence which is isomorphic to the following exact sequence : @xmath791 since @xmath58 is a left @xmath14-approximation of @xmath379 , the map @xmath792 is surjective , and consequently @xmath793 so , the right @xmath46-module @xmath788 has no composition factors in @xmath794 , and that @xmath788 has composition factors only in @xmath795 .
this is equivalent to saying that the @xmath46-module @xmath796 has composition factors only in @xmath797 which is isomorphic to @xmath798 .
let @xmath799 be an element in @xmath800 .
then @xmath469 factorises through @xmath801 $ ] , or equivalently , @xmath802 .
this implies that @xmath803 .
thus the image of @xmath804 is contained in the kernel of @xmath805 , which is isomorphic to @xmath796 .
therefore , if @xmath806 , then the top of the image of @xmath804 is contained in @xmath782 , this is a contradiction .
thus we must have @xmath807 . since @xmath786 is a duality from @xmath515 to @xmath808
, we obtain @xmath809 .
thus @xmath783 .
@xmath384 _ remark_. ( 1 ) if we substitute @xmath14-split " for left @xmath810-approximation " and
right @xmath811 , @xmath812-approximation " in proposition [ bigideal ] , and if we consider @xmath554 and @xmath813 instead of @xmath46 and @xmath41 , then proposition [ bigideal ] is still true .
the proof is almost the same .
\(2 ) by definition , there are inclusions @xmath814 and @xmath815 .
sometimes it is easy to verify that @xmath778 and @xmath779 vanish if the algebras @xmath46 and @xmath41 are described by quivers with relations . for the derived category of an abelian category , the following result provides an explicit example for @xmath816 .
let @xmath817 be an abelian category , and let @xmath7 be an object of @xmath817 .
suppose that @xmath818 is an exact sequence in @xmath817 with @xmath348 .
consider the induced triangle @xmath777 $ ] in @xmath819 .
then the ideals @xmath37 and @xmath38 defined in theorem [ verygthm1 ] vanish .
[ vanish2 ] _ proof .
_ every exact sequence @xmath820 in @xmath821 gives rise to a triangle @xmath822 $ ] in @xmath823 .
now we show that the exactness of the given sequence in @xmath821 implies that the two ideals @xmath37 and @xmath38 in theorem [ verygthm1 ] are equal to zero . since @xmath37 is contained in @xmath824 , it is sufficient to show that if a morphism @xmath825 factorises through @xmath14 and @xmath393 $ ] , then @xmath809
. let @xmath469 be such a morphism .
then we see immediately that @xmath802 in @xmath823 .
since @xmath817 is fully embedded in @xmath819 , we also have @xmath802 in @xmath817 .
consequently , @xmath809 since @xmath58 is injective in @xmath817 .
thus @xmath775 .
dually , we can show @xmath776 .
hence proposition [ vanish2 ] holds true .
@xmath384 as an immediate application of the proof of theorem [ verygthm1 ] together with a result on derived equivalences in @xcite , we have the following corollary .
we keep all assumptions of theorem [ verygthm1 ] . if @xmath506 and @xmath563 both are left coherent rings @xmath826for example , if @xmath1 is finite and @xmath9= @xmath79 with @xmath80 a finite dimensional algebra over a field@xmath827 , then @xmath828 , where @xmath829 stands for the finitistic dimension of @xmath506 .
recall that , given a ring @xmath20 with identity , the _
finitistic dimension _ of @xmath20 is defined to be the supremum of the projective dimensions of finitely generated @xmath20-modules of finite projective dimension .
since the map @xmath830 in the proof of theorem [ verygthm1 ] is not always injective , the tilting complex @xmath186 is not , in general , isomorphic in @xmath502 to a tilting module .
thus the derived equivalence presented in theorem [ verygthm1 ] is not given by a tilting module in general ( in contrast with the situation of theorem [ deadss ] ) .
in fact , it is easy to see that the derived equivalence in theorem [ verygthm1 ] is given by a tilting module if the kernel of @xmath831 is @xmath478 . moreover , a small additive category may be embedded into an abelian category of coherent functors ( see ( * ? ? ? * chapter iv , section 2 ) ) .
this will , however , not in general turn a @xmath90-split sequence in the additive category into an exact sequence in the abelian category since otherwise the sequence would split , and therefore can not provide a triangle in the derived category of the abelian category . consequently , theorem [ deadss ] can not be obtained from theorem [ verygthm1 ] by taking @xmath250 and embedding an additive category into an abelian category .
finally , we mention that theorem [ verygthm1 ] generalises the result ( * ? ? ?
* proposition 5.1 ) by choosing @xmath250 . indeed , under the conditions of (
* proposition 5.1 ) , the ideals @xmath37 and @xmath38 in theorem [ verygthm1 ] vanish .
theorem [ verygthm1 ] covers various other situations , some of which will be discussed in the next section .
in this section , we shall describe some natural habitats for theorem [ verygthm1 ] and relate it to several widely used concepts that fit with or simplify the assumptions of theorem [ verygthm1 ] . throughout
, we choose @xmath63 to be the shift functor of the triangulated category considered .
we note that alex dugas , in independent work @xcite that also is motivated by @xcite , has constructed derived equivalent pairs of symmetric algebras . as explained in @xcite ( remark ( 3 ) in section 4 ) his examples appear in our framework , too .
a first consequence of theorem [ verygthm1 ] is the following result for @xmath832 with @xmath80 an artin @xmath10-algebra .
let @xmath1 be an admissible subset of @xmath81 , let @xmath7 be an @xmath80-module , and let @xmath82 be an exact sequence in @xmath83 with @xmath12 a left @xmath833-approximation of @xmath15 and @xmath13 a right @xmath773-approximation of @xmath16 in @xmath79 such that @xmath834 and @xmath835 . then the perforated yoneda algebras @xmath87 and @xmath88 are derived equivalent .
[ gthm1 ] _ proof .
_ this is a consequence of theorem [ verygthm1 ] and proposition [ vanish2 ] if we take @xmath9 = @xmath79 .
@xmath384 under the assumptions of theorem [ gthm1 ] , the higher cohomology groups @xmath836 of @xmath15 is isomorphic to the higher cohomology groups @xmath837 of @xmath16 for each @xmath51 .
this follows from the comment ( b ) before corollary [ cortoverygthm1 ] . when requiring additional orthogonality conditions on @xmath15 and @xmath16 in theorem [ gthm1 ] , we get the following corollary .
let @xmath1 be an admissible subset of @xmath81 , let @xmath7 be an @xmath80-module , and let @xmath82 be an @xmath14-split sequence in @xmath83 such that @xmath838 . then the perforated yoneda algebras @xmath87 and @xmath88 are derived equivalent .
[ thm1adss ] _ proof .
_ this follows immediately from corollary [ cortoverygthm1 ] and proposition [ vanish2 ] .
@xmath384 if the orthogonality conditions in corollary [ thm1adss ] hold for @xmath767 or @xmath81 , then we get the following consequence .
suppose that @xmath7 is an @xmath80-module .
let @xmath82 be an @xmath14-split sequence in @xmath83 such that @xmath839 for @xmath0 a positive number or infinity .
then , for any admissible subset @xmath1 of @xmath767 , the perforated yoneda algebras @xmath87 and @xmath88 are derived equivalent .
[ corthm1adss ] the following result shows that the orthogonality conditions are related to the concepts of _ short cycle _ and _ short chain _ in @xmath80-mod ( see ( * ? ? ?
* chapter ix , p.313 ) ) . recall that a short cycle of length @xmath840 from an indecomposable module @xmath15 to @xmath15 is a sequence of non - zero radical homomorphisms @xmath841 with @xmath7 indecomposable ; and a short chain is a sequence of non - zero radical homomorphisms @xmath842 with @xmath15 indecomposable .
let @xmath80 be an artin algebra , and let @xmath843 be an auslander - reiten sequence in @xmath83 .
suppose neither @xmath15 nor @xmath16 lies on a short cycle of length @xmath840 , nor on a short chain .
then the trivial extension of @xmath844 by the bimodule @xmath845 is derived equivalent to the trivial extension of @xmath846 by the bimodule @xmath847 .
[ shortcycle ] _ proof .
_ an auslander - reiten sequence @xmath843 is always an @xmath14-split sequence .
since @xmath16 does not lie on a short cycle , the auslander - reiten formula @xmath848 ( see @xcite ) implies @xmath849 .
moreover , @xmath15 not lying on a short cycle implies @xmath850 .
similarly , the auslander - reiten formula yields that @xmath851 - since @xmath15 does not lie on a short chain - and that @xmath852 - since @xmath16 does not lie on a short chain .
thus corollary [ shortcycle ] follows from corollary [ corthm1adss ] when @xmath853 .
@xmath384 the next corollary is a consequence of corollary [ corthm1adss ] .
let @xmath80 be an artin algebra , and let @xmath15 be an @xmath80-module such that @xmath854 for all @xmath855 with @xmath0 a fixed positive integer or infinity .
then , for any admissible subset @xmath1 of @xmath767 , the perforated yoneda algebras @xmath856 and @xmath857 are derived equivalent .
[ gnc ] _ proof .
_ if @xmath854 for a fixed @xmath254 , then @xmath858 is an @xmath859-split sequence in @xmath80-mod , where @xmath860 is a projective cover of @xmath861 .
using this fact , corollary [ gnc ] follows immediately from corollary [ corthm1adss ] .
@xmath384 the condition @xmath854 on @xmath15 in corollary [ gnc ] is related to the context of the _ generalised nakayama conjecture_. this states that if an @xmath80-module @xmath21 satisfies @xmath862 for all @xmath863 then @xmath21 should be projective .
the above corollary [ gnc ] ( or ( * ? ? ? * theorem 1.1 ) ) describes the shape of the syzygy modules @xmath861 : if @xmath15 is indecomposable and non - projective and satisfies @xmath854 for all @xmath863 , then , for each @xmath864 , there is an indecomposable non - projective module @xmath865 such that @xmath866 for an integer @xmath867 . in corollary [ gnc ] , there are isomorphisms @xmath868 for all @xmath254 .
thus the algebras @xmath856 and @xmath857 are the extensions of @xmath869 and @xmath870 by the same ideal @xmath871 , respectively . the algebras @xmath87 and @xmath88 in corollary [ corthm1adss ] , however , are the extensions of @xmath844 and @xmath846 by possibly different ideals @xmath872 and @xmath873 , respectively . recall that a module @xmath874-mod
is called _ reflexive _ if the evaluation map @xmath875 is an isomorphism of modules .
let @xmath7 be a reflexive @xmath80-module . then , for any subset @xmath876 , the perforated yoneda algebras @xmath877 and @xmath878 are derived equivalent , where @xmath879 is the co - syzygy operator .
[ reflexive ] _ proof .
_ by ( * ? ?
* iv , proposition 3.2 ) , the kernel and cokernel of the evaluation map @xmath880 are @xmath881 and @xmath882 , respectively . as @xmath883 for any @xmath80-module @xmath233 , corollary [ reflexive ] follows from corollary [ gnc ] for right modules .
@xmath384 a special case of corollary [ gnc ] , is a result on self - injective algebras that has been obtained in ( * ? ? ?
* corollary 3.14 ) ) : if @xmath80 is a self - injective artin algebra , then , for any admissible subset @xmath1 of @xmath884 , the perforated yoneda algebras @xmath856 and @xmath857 are derived equivalent.[selfinjalg ] another concept related to the generalised nakayama conjectures and to modules being projective and injective , is the _ dominant dimension _ of an algebra or a module .
suppose that @xmath80 is an artin @xmath10-algebra . by definition ,
the _ dominant dimension _ of @xmath80 is greater than or equal to @xmath0 if in the minimal injective resolution of @xmath885 : @xmath886 the first @xmath0 injective @xmath80-modules @xmath553 , @xmath887 are projective . in this case
we write @xmath888 .
let @xmath889 be the @xmath722-th cosyzygy of @xmath80 , that is , the cokernel of the map @xmath890 .
for an @xmath80-module @xmath15 , we define @xmath891 to be the number of non - isomorphic indecomposable direct summands of @xmath7 .
the _ self - injective measure _ of @xmath80 is defined to be the number @xmath892 , where @xmath553 is an injective hull of @xmath80 .
thus , if @xmath80 is self - injective , then @xmath893 . if @xmath894 , then @xmath80 is self - injective if and only if @xmath893 .
so the nakayama conjecture can be reformulated as : if @xmath895 , then @xmath893 .
let @xmath80 be an artin algebra , and let @xmath21 be the direct sum of all non - isomorphic indecomposable projective - injective @xmath80-modules .
@xmath67 if @xmath896 , then @xmath897 is derived equivalent to @xmath80 for @xmath898 .
@xmath70 if @xmath899 , then @xmath900 .
[ dom ] _ proof .
_ since the sequence @xmath901 is an @xmath902-split sequence ( or an @xmath903-split sequence ) , the orthogonality conditions in corollary [ corthm1adss ] are trivially satisfied . derived equivalence preserves the number of non - isomorphic simple modules .
therefore , corollary [ dom ] follows now from corollary [ corthm1adss ] .
here we also use the observation that @xmath904 for all @xmath905 .
alternatively , one can also use lemma [ deadss ] to prove this corollary . @xmath384
examples of algebras of dominant dimension at least @xmath0 can be obtained in the following way : let @xmath80 be a self - injective algebra and @xmath15 an @xmath80-module . if @xmath906 for all @xmath488 , then dom.dim@xmath907 .
finally , we turn to auslander - regular algebras .
let @xmath170 be a @xmath200-algebra over a field @xmath200 .
recall that @xmath170 is called _ auslander - regular _ if @xmath170 has finite global dimension and satisfies the gorenstein condition : if @xmath908 are non - negative integers and @xmath7 is a finitely generated ( left or right ) @xmath170-module , then @xmath909 for every submodule n of @xmath910 . here , if @xmath7 is a right @xmath170-module , then @xmath911 is a left @xmath170-module .
let @xmath912 be the minimal number @xmath913 such that @xmath914 .
then for any submodule @xmath911 of @xmath915 , we have @xmath916 for @xmath917 thus : let @xmath170 be an auslander - regular @xmath200-algebra , and @xmath7 a finitely generated right @xmath170-module . then ,
for any submodule @xmath15 of @xmath915 , and any admissible subset @xmath1 of @xmath918 , the algebras @xmath919 and @xmath920 are derived equivalent .
[ aregular ] let @xmath921 be a _
frobenius _ abelian category , that is , @xmath922 is an abelian category with enough projective objects and enough injective objects such that the projective objects coincides with the injective objects .
we denote by @xmath923 the stable category of @xmath922 modulo projective objects .
it is shown in @xcite that @xmath923 is a triangulated category , in which the shift functor @xmath764 $ ] is just the co - syzygy functor @xmath879 , and the triangles in @xmath923 are all induced by short exact sequences in @xmath922 .
for each morphism @xmath924 in @xmath922 , we denote by @xmath463 the image of @xmath45 under the canonical functor from @xmath922 to @xmath923 .
note that the objects of @xmath923 are the same as those of @xmath817 .
let @xmath1 be an admissible subset of @xmath925 , and let @xmath7 , @xmath15 , and @xmath16 be objects in @xmath922 .
then @xmath67 for arbitrary @xmath926 and @xmath927 , there is an isomorphism @xmath928)\simeq { { \rm hom } } _ { \underline{\cal a}}(u , u'[i]),\ ] ] which is functorial in @xmath233 and @xmath929 ; @xmath70 a monomorphism @xmath930 in @xmath922 is a left @xmath332-approximation of @xmath15 in @xmath819 if and only if @xmath931 is a left @xmath833-approximation of @xmath15 in @xmath923 ; @xmath245 an epimorphism @xmath932 in @xmath922 is a right @xmath773-approximation of @xmath16 in @xmath819 if and only if @xmath933 is a right @xmath773-approximation of @xmath16 in @xmath923 .
[ lemfrobenius ] _ proof .
_ ( 1 ) for @xmath934 , the isomorphisms @xmath928)\simeq { { \rm ext}}_{\cal a}^i(u , u')\simeq { { \rm hom } } _ { \underline{\cal a}}(u , \omega^{-i}u')={{\rm hom } } _ { \underline{\cal a}}(u , u'[i]).\ ] ] are functorial in @xmath233 and @xmath929
. thus ( 1 ) follows .
\(2 ) first , let @xmath935 be in @xmath1 .
by ( 1 ) , there is a commutative diagram @xmath936)\ar[r]^{(\alpha , m[i])}\ar[d]^{\simeq } & { { \rm hom } } _ { { \mathscr{d}^{\rm b}(\cal a)}}(x , m[i])\ar[d]^{\simeq}\\ { { \rm hom } } _ { \underline{\cal a}}(m_1 , m[i])\ar[r]^{(\underline{\alpha } , m[i ] ) } & { { \rm hom } } _ { \underline{\cal a}}(x , m[i ] ) . } \ ] ] thus , the map @xmath937)$ ] is surjective if and only if @xmath938)$ ] is surjective .
now we consider the case @xmath408 .
if every morphism from @xmath15 to @xmath7 in @xmath922 factorises through @xmath12 , then every morphism from @xmath15 to @xmath7 in @xmath923 factorises through @xmath931 .
conversely , assume that every morphism from @xmath15 to @xmath7 in @xmath923 factorises through @xmath931 .
let @xmath939 be a morphism in @xmath922 .
then @xmath940 for some @xmath941 in @xmath922 .
thus @xmath942 in @xmath922 factorises through a projective object @xmath943 , say @xmath944 for some @xmath945 and @xmath946 in @xmath922 .
since @xmath943 is also injective and @xmath12 is a monomorphism , there is some morphism @xmath947 such that @xmath948 .
altogether , @xmath949 factorises through @xmath12 .
thus the statement ( 2 ) follows .
the proof of ( 3 ) is similar to that of ( 2 ) .
@xmath384 let @xmath1 be an admissible subset of @xmath925 .
suppose that @xmath922 is a frobenius abelian category , that @xmath7 is an object in @xmath922 , and that @xmath950 is a short exact sequence in @xmath922 with @xmath951 such that the induced triangle @xmath952 $ ] in @xmath923 satisfies the conditions in theorem [ verygthm1 ] .
then the algebras @xmath953 and @xmath954 are derived equivalent .
[ frobprop ] _ proof .
_ this follows from lemma [ lemfrobenius ] and proposition [ vanish2 ] .
@xmath384 suppose that @xmath922 is a frobenius abelian category and @xmath7 is an object in @xmath922 .
let @xmath820 with @xmath951 be a short exact sequence in @xmath922 such that the induced triangle in @xmath923 is an @xmath14-split triangle
. then @xmath955 and @xmath956 are derived equivalent .
[ frobcor1 ] _ proof .
_ taking @xmath957 , the corollary follows from proposition [ frobprop ] .
@xmath384 .
if @xmath817 is a frobenius ( not necessarily abelian ) category , then corollary [ frobcor1 ] is still true . for the precise definition of a fronenius category ,
we refer the reader to @xcite .
the module category of a self - injective artin algebra is a frobenius abelian category . in this case
, we have the following corollary .
let @xmath80 be a self - injective algebra , and let @xmath7 be an @xmath80-module .
suppose @xmath958 $ ] is an @xmath14-split triangle in @xmath94 .
then @xmath959 and @xmath960 are derived equivalent . _ proof .
_ since all triangles in @xmath94 are induced by short exact sequences in @xmath83 , there is a short exact sequence @xmath961 in @xmath80-mod with @xmath943 projective such that the induced triangle is isomorphic to the given triangle @xmath958 $ ] in @xmath94 .
the triangle @xmath962 $ ] also is an @xmath963-split triangle in @xmath94 .
the corollary then follows from corollary [ frobcor1 ] .
@xmath384 the theory of calabi - yau and cluster categories provides very natural contexts for our construction of derived equivalences .
let @xmath200 be a field , and let @xmath131 be a @xmath200-linear triangulated category which is hom - finite , that is , the hom - space @xmath964 is finite dimensional over @xmath200 for all @xmath15 and @xmath16 in @xmath131 .
recall that @xmath131 is called _ @xmath965-calabi - yau _ for some non - negative integer @xmath0 if there is a natural isomorphism between @xmath966 and @xmath967)$ ] for all @xmath15 and @xmath16 in @xmath131 , where @xmath968 is the usual duality .
it follows that @xmath969 for @xmath970 .
( see @xcite for more information on calabai - yau categories . )
note that if @xmath139 , then @xmath971 for each @xmath588 .
let @xmath972 .
suppose that @xmath131 is an @xmath965-calabi - yau triangulated category , and that @xmath7 is an object in @xmath131 .
let @xmath973 $ ] be a triangle in @xmath131 with @xmath951 . then : @xmath67 the morphism @xmath12 is a left @xmath84-approximation of @xmath15 if and only if the morphism @xmath13 is a right @xmath773-approximation of @xmath16 ; @xmath70 if @xmath12 is a left @xmath833-approximation of @xmath15 and if @xmath7 is @xmath0-self - orthogonal , then @xmath974 and @xmath975 . [ propcy ] _ proof .
_ we will abbreviate @xmath976 by @xmath977 .
first we assume that @xmath12 is a left @xmath84-approximation of @xmath15 .
now , for each @xmath588 , there is a commutative diagram with exact rows @xmath978 , m_1)\ar[r]^{(m[-i ] , \beta)}\ar[d]^{\simeq } & ( m[-i ] , y)\ar[d]^{\simeq}\\ ( m , m_1[i])\ar[r]^{(m , \beta[i ] ) } & ( m , y[i])\ar[r ] & ( m , x[i+1])\ar[r]^{(m , \alpha[i+1])}\ar[d]^{\simeq } & ( m , m_1[i+1])\ar[d]^{\simeq}\\ & & d(x , m[n - i])\ar[r]^{d(\alpha , m[n - i ] ) } & d(m_1 , m[n - i ] ) . } \ ] ] since @xmath979 is in @xmath1 , and since @xmath12 is a left @xmath84-approximation of @xmath15 , the map @xmath980)$ ] is surjective , and consequently @xmath981)$ ] is injective . hence @xmath982)$ ]
is injective , and therefore @xmath983 , \beta)$ ] is surjective .
this shows that @xmath13 is a right @xmath85-approximation of @xmath16 .
the proof of the other implication in ( 1 ) can be done similarly .
\(2 ) it follows from ( 1 ) and the comment before corollary [ cortoverygthm1 ] that @xmath984 and @xmath985 . since @xmath131 is @xmath965-calabi - yau , we have @xmath986)\simeq d(x , m[n+1-i])=0 $ ] , and @xmath987)\simeq d(y , m[n+1-i])=0 $ ] for all @xmath134 .
thus @xmath988 and @xmath989 .
@xmath384 let @xmath278 , and let @xmath131 be an @xmath965-calabi - yau triangulated category .
suppose that @xmath7 is @xmath0-self - orthogonal and @xmath990 .
let @xmath991 $ ] be a triangle in @xmath131 with @xmath13 a right @xmath14-approximation of @xmath16 .
then the algebras @xmath992 and @xmath993 are derived equivalent , where @xmath37 and @xmath38 are defined as in theorem [ verygthm1 ] .
[ cycor1 ] _ proof .
_ since @xmath985 , for each @xmath134 , the map @xmath983 , m_1){\longrightarrow}(m[-i ] , y)=0 $ ] induced by @xmath13 is surjective .
taking into account that @xmath13 is a right @xmath14-approximation of @xmath16 , we see that @xmath13 is , in fact , a right @xmath773-approximation of @xmath16 . by proposition [ propcy ]
( 1 ) , the map @xmath12 is a left @xmath84-approximation of @xmath15 .
since @xmath7 is @xmath0-self - orthogonal , the proof can be finished by applying proposition [ propcy ] ( 2 ) and corollary [ cortoverygthm1 ] to the triangle .
@xmath384 corollary [ cycor1 ] is related to mutations in a calabi - yau category . here
are some definitions from @xcite .
let @xmath9 be an @xmath965-calabi - yau category .
an object @xmath21 in @xmath9 is called an _ @xmath0-cluster tilting object _ if @xmath21 is @xmath0-self - orthogonal , and if any @xmath994 with @xmath995 for @xmath488 is in @xmath903 .
the object @xmath21 is called _ basic _ if the multiplicity of each indecomposable direct summand of @xmath21 is one .
let @xmath21 be an @xmath0-cluster basic tilting object in an @xmath965-calabi - yau category @xmath9 , and @xmath16 a direct summand of @xmath21 , that is , @xmath996 .
let @xmath997 be a minimal right @xmath14-approximation of @xmath16 , and let @xmath998\ ] ] be a triangle containing @xmath13 .
note that we allow @xmath16 to be decomposable , and that @xmath15 is indecomposable if and only if @xmath16 is indecomposable .
the object @xmath999 is called the _ left mutation _ of @xmath21 at @xmath16 . in the case of tilting modules , @xmath15 is called a _ tilting complement _ to @xmath7 in the literature ( see , for example , @xcite ) .
it was pointed out in @xcite that the left mutation of @xmath21 at @xmath16 is again an @xmath0-cluster tilting object ( for some special cases , see @xcite , and also @xcite ) .
in fact , this can be seen in the following way : the proof of corollary [ cycor1 ] and comment ( b ) on the conditions of theorem [ verygthm1 ] imply that @xmath1000 is @xmath0-self - orthogonal .
morover , let @xmath1001 and consider a triangle @xmath1002 $ ] with @xmath1003 a left @xmath14-approximation of @xmath1004
. then @xmath1005 by lemma [ propcy ] and the comment ( b ) .
thus @xmath1006 , @xmath1007 , and @xmath1008 is again an @xmath0-cluster tilting object in @xmath9 .
the notion of a right mutation of @xmath21 at @xmath16 is dual .
usually , @xmath76 and @xmath356 are not derived equivalent .
when they are derived equivalent may be an interesting question . here is a sufficient condition .
let @xmath1009 and @xmath1010 .
then @xmath67 @xmath1011 and @xmath1012 are derived equivalent .
@xmath70 suppose that @xmath16 is indecomposable .
let @xmath1013 be the simple @xmath170-module corresponding to @xmath15 , and let @xmath1014 be the simple @xmath171-module corresponding to @xmath16 .
suppose that @xmath1014 is not a submodule of @xmath171 , and @xmath1013 is not a quotient of @xmath1015 .
then @xmath170 and @xmath171 are derived equivalent .
[ mutation ] _ proof .
_ statement ( 1 ) is a direct consequence of corollary [ cycor1 ] , and ( 2 ) follows from ( 1 ) and proposition [ bigideal ] .
@xmath384 consider a @xmath840-calabi - yau category , and assume that @xmath1016 .
then we re - obtain the result ( * ? ? ?
* theorem 5.3 ) from corollary [ mutation ] ( 2 ) .
first , we present an explicit example which satisfies all conditions in theorem [ verygthm1 ] .
let @xmath200 be an algebraically closed field of characteristic @xmath840 , and let @xmath1017 be the group algebra of the alternating group @xmath1018 .
then there are three simple @xmath80-modules , which are denoted @xmath1019 , and @xmath1020 , respectively .
their projective covers are @xmath1021 , @xmath1022 and @xmath1023 , respectively .
it is well known that @xmath1024 is morita equivalent to the following algebra given by quiver @xmath1025^{\alpha_1}^(0){\omega}^(1){\bar{\omega}}\ar@<2pt>[rdd]^{\beta_1}_(1.1){k } & & \bullet\ar@<2pt>[ldd]^{\alpha_2}\ar@<2pt>[ll]^{\beta_2}\\ & & \\ & \bullet\ar@<2pt>[luu]^{\alpha_3}\ar@<2pt>[ruu]^{\beta_3 } & } \ ] ] and relations @xmath1026 , where the subscripts are considered modulo @xmath1027 . as
this algebra is symmetric , the auslander - reiten translation @xmath1028 is just the second syzygy @xmath1029 .
the auslander - reiten quiver of this algebra is well - known to have a component of the following form : ( 330 , 260 ) ( 105 , 230)(1,-1)20 ( 70 , 210)(1,1)20 ( 85 , 233)@xmath1030 ( 70 , 195)(1,-1)20 ( 105 , 175)(1,1)20 ( 85 , 165)@xmath1031 ( 70 , 140)(1,1)20 ( 105 , 160)(1,-1)20 ( 70 , 125)(1,-1)20 ( 105 , 105)(1,1)20 ( 85 , 95)@xmath1032 ( 70 , 70)(1,1)20 ( 105 , 90)(1,-1)20 ( 70 , 55)(1,-1)20 ( 105 , 35)(1,1)20 ( 85 , 25)@xmath1030 ( 15 , 200)@xmath1033 ( 45 , 200)@xmath1034 ( 15 , 130)@xmath1033 ( 45 , 130)@xmath1035 ( 15 , 60)@xmath1033 ( 45 , 60)@xmath1031 ( 215 , 210)(1,1)20 ( 250 , 230)(1,-1)20 ( 230 , 233)@xmath1036 ( 290 , 210)(1,1)20 ( 290 , 140)(1,1)20 ( 290 , 70)(1,1)20 ( 290 , 195)(1,-1)20 ( 290 , 125)(1,-1)20 ( 290 , 55)(1,-1)20 ( 35 , 230)(1,-1)20 ( 35 , 160)(1,-1)20 ( 35 , 90)(1,-1)20 ( 35 , 175)(1,1)20 ( 35 , 105)(1,1)20 ( 35 , 35)(1,1)20 ( 215 , 195)(1,-1)20 ( 250 , 175)(1,1)20 ( 230 , 165)@xmath1037 ( 215 , 140)(1,1)20 ( 250 , 160)(1,-1)20 ( 215 , 125)(1,-1)20 ( 250 , 105)(1,1)20 ( 230 , 95)@xmath1038 ( 215 , 70)(1,1)20 ( 250 , 90)(1,-1)20 ( 215 , 55)(1,-1)20 ( 250 , 35)(1,1)20 ( 230 , 25)@xmath1036 ( 310 , 200)@xmath1033 ( 265 , 200)@xmath1039 ( 310 , 130)@xmath1033 ( 265 , 130)@xmath1040 ( 310 , 60)@xmath1033 ( 265 , 60)@xmath1041 ( 165 , 235)@xmath200 ( 140 , 210)(1,1)20 ( 175 , 230)(1,-1)20 ( 120 , 200)@xmath1042 ( 143 , 203)(1,0)13 ( 157 , 200)@xmath1022 ( 179 , 203)(1,0)13 ( 196 ,
200)@xmath1043 ( 140 , 195)(1,-1)20 ( 175 , 175)(1,1)20 ( 165 , 165)@xmath1020 ( 140 , 140)(1,1)20 ( 175 , 160)(1,-1)20 ( 120 , 130)@xmath1044 ( 143 , 133)(1,0)13 ( 157 , 130)@xmath1021 ( 179 , 133)(1,0)13 ( 196 , 130)@xmath1045 ( 140 , 125)(1,-1)20 ( 175 , 105)(1,1)20 ( 165 , 95)@xmath1046 ( 140 , 70)(1,1)20 ( 175 , 90)(1,-1)20 ( 120 , 60)@xmath1047 ( 143 , 63)(1,0)13 ( 157 , 60)@xmath1023 ( 179 , 63)(1,0)13 ( 196 , 60)@xmath1048 ( 140 , 55)(1,-1)20 ( 175 , 35)(1,1)20 ( 165 , 25)@xmath200 consider the auslander - reiten sequence @xmath1049 let @xmath1050 , @xmath1051 , and @xmath1052 .
this sequence provides an auslander - reiten triangle in the triangulated category @xmath94 : @xmath1053.\ ] ] we shall check that this triangle satisfies the conditions of theorem [ verygthm1 ] .
we choose @xmath1054 and @xmath311 $ ] .
since this is an auslander - reiten triangle in @xmath94 , the map @xmath1055 is a left @xmath833-approximation of @xmath15 , and the map @xmath1056 is a right @xmath773-approximation of @xmath16 ( see the example at the end of section [ sect2 ] ) .
it follows from the above auslander - reiten quiver of @xmath80 that @xmath1057 and @xmath1058 thus the above triangle in @xmath94 satisfies all conditions in theorem [ verygthm1 ] , and therefore , by proposition [ frobprop ] , the algebras @xmath1059 and @xmath88 are derived equivalent .
furthermore , we have @xmath1060 .
there is an epimorphism from @xmath1061 to @xmath1020 and an epimorphism from @xmath1062 to @xmath200 .
the latter can not factorise through a projective module , we get dim@xmath1063 . moreover , there is an epimorphism from @xmath1061 to @xmath1046 and an epimorphism from @xmath1062 to @xmath1046 .
this implies dim@xmath1064 .
similarly , dim@xmath1065 .
note that all the indecomposable modules appearing in the auslander - reiten triangle are @xmath289-self - orthogonal .
a more precise calculation shows that @xmath1066 and @xmath1067 the following example shows that the ext - orthogonality conditions in corollary [ thm1adss ] and therefore in theorem [ verygthm1 ] can not be dropped .
let @xmath80 be the algebra ( over a field @xmath200 ) given by the following quiver with relations : = 0.5 cm ( 20,3.5 ) ( 5,1.3)@xmath289(7.3,1.3)@xmath840 ( 5,2)@xmath1068(7.3,2.2)(-1,0)1.8(7.5,2)@xmath1068(8.7,2.3 ) ( 9.7,2.2)(-1,-2)0.2 ( 6.5 , 2.5)@xmath13(10 , 2)@xmath1069(11.5,2)@xmath1070 this example is in a class of examples constructed by small @xcite .
the algebra @xmath80 is of finite representation type , its finitistic dimension equals one , while the finitistic dimension of the opposite algebra @xmath100 is zero .
we denote by @xmath1071 and @xmath1072 the simple and projective modules corresponding to the vertex @xmath722 , respectively .
let @xmath1073 be the quotient module of @xmath1074 by @xmath1071 , and @xmath1075 , where @xmath95 is the usual duality .
then there is an auslander - reiten sequence @xmath1076 this is an @xmath14-split sequence in @xmath80-mod .
if we take @xmath1054 , then @xmath1077 .
an easy calculation shows that @xmath1078 is a quasi - hereditary algebra , and thus has finite global dimension .
the algebra @xmath88 contains a loop which is given by the short exact sequence induced by the loop @xmath12 at the vertex @xmath840 .
thus it has infinite global dimension by the no loops theorem. it follows that @xmath87 and @xmath88 can not be derived equivalent since derived equivalences preserve the finiteness of global dimensions . also , one can see that @xmath1079 and @xmath1080 for @xmath254 .
this example shows that the orthogonality conditions in corollary [ thm1adss ] can not be omitted .
moreover , it shows that the result in ( * ? ? ?
* theorem 1.1 ) can not be extended from endomorphism algebras to @xmath1-yoneda algebras without any additional conditions .
in theorem [ verygthm1 ] , there is only one functor @xmath63 involved .
when working with the derived category of a hereditary algebra , or the stable category of a self - injective algebra , or the derived category of coherent sheaves of a projective variety over @xmath260 , apart from the shift functor there are other prominent functors , for example the auslander - reiten translation dtr . to have available a general statement of construction of derived equivalences , which is similar to theorem [ verygthm1 ] , we define @xmath1-perforated yoneda algebras for two functors over a triangulated category , and formulate a two - functor version of theorem [ verygthm1 ] .
in this appendix , we summarise the ingredients for a possible generalisation of theorem [ verygthm1 ] .
the proof of this generalisation is analogous to that of theorem [ verygthm1 ] , but more technical and tedious .
so we omit it here .
let @xmath1 be a subset of @xmath1081 , which we consider as a semigroup with ordinary addition .
let @xmath9 be a triangulated @xmath10-category with shift functor [ 1 ] , and let @xmath15 be an object in @xmath9 .
suppose that @xmath63 and @xmath78 are two triangle functors from @xmath9 to itself , such that @xmath113 is naturally isomorphic to @xmath112 .
for @xmath15 in @xmath9 , let @xmath1082 be an isomorphism induced from the natural transformation @xmath1083 .
then we define a general model for the above definition is : given a bi - graded algebra @xmath1090 , we define @xmath1091 , and a multiplication by @xmath1092 if @xmath1093 , and zero otherwise . if @xmath1 is admissible , for example , @xmath1 is the cartesian product of two admissible sets in @xmath65 , then @xmath1094 is an associative algebra .
so , we have to check that @xmath1095 is an associative algebra with respect to the above multiplication .
this can be based on the following lemma .
suppose @xmath63 and @xmath78 are two triangle functors from @xmath9 to itself such that @xmath113 is naturally isomorphic to @xmath112 .
for any triangle functor @xmath1096 from @xmath9 to itself , there is a natural isomorphism @xmath1097 for all @xmath1098 such that , for @xmath1099 , _ proof . _ for functors @xmath1102 and @xmath1103 from @xmath131 to itself , we define @xmath1104 to be the induced natural isomorphism from the functor @xmath1105 to the functor @xmath1106 .
so , @xmath1107 is just the given natural isomorphism from @xmath113 to @xmath112 .
now we shall construct inductively a natural isomorphism @xmath1108 from @xmath1109 to @xmath1110 for all non - negative integers @xmath722 and @xmath24 and functors @xmath1096 from @xmath9 to itself .
if @xmath408 or @xmath1111 , then @xmath1112 , and we define @xmath1108 to be the identity natural transformation . for each positive integer @xmath1113 , we assume that @xmath1114 is defined .
now we define @xmath1115 for each positive integer @xmath1116 , assume that @xmath1117 is defined .
we define @xmath1118 \(2 ) we first prove ( 2 ) for @xmath1123 .
if @xmath1124 , then ( 2 ) is clearly true .
now suppose @xmath1125 .
we shall show ( 2 ) by induction on @xmath1126 .
in fact , if @xmath1127 , it is straightforward to check ( 2 ) .
now we assume that @xmath1128 .
then we have @xmath1129 this proves ( 2 ) for @xmath1125 .
now assume @xmath1130 .
since @xmath1132 is a natural transformation from @xmath1133 to @xmath1134 , the following diagram of natural transformations is commutative : @xmath1135^{\delta(1 , s , f^rg^{p-1}l)}\ar[d]_{f^sg\delta(p-1 , r , l ) } & gf^sf^rg^{p-1}l\ar[d]^{gf^s\delta(p-1 , r , l)}\\ f^sgg^{p-1}f^rl \ar[r]^{\delta(1 , s , g^{p-1}f^rl ) } & gf^sg^{p-1}f^rl.\\ } \ ] ] hence @xmath1136 this proves ( 2 ) .
@xmath384 let @xmath123 be a full subcategory of @xmath9 , and @xmath15 an object of @xmath9 .
a morphism @xmath293 with @xmath1138 is called a left @xmath1139-approximation of @xmath15 if @xmath1140 is surjective for every object @xmath297 and @xmath1141 .
dually , we define the right @xmath1142-approximation of @xmath15 . given a triangle @xmath1143 $ ] in @xmath62 with @xmath348 for a fixed @xmath970
, we define @xmath393= ( -w[-1],0 ) : y[-1]{\rightarrow}x\oplus m$ ] , @xmath1144 , where @xmath350 stands for the transpose of the matrix @xmath351 , and let @xmath1 be an admissible subset of @xmath1147 , and let @xmath9 be a triangulated @xmath10-category , and let @xmath7 be an object in @xmath9 .
assume that there are two invertible triangle functors @xmath63 and
@xmath78 from @xmath9 to itself such that @xmath113 is naturally isomorphic to @xmath112 by @xmath1148 , suppose that @xmath761 $ ] is a triangle in @xmath62 such that @xmath12 is a left @xmath1149-approximation of @xmath15 and @xmath13 is a right @xmath1150-approximation of @xmath16 . if @xmath1151 for @xmath1152 , then @xmath1153 and @xmath1154 are derived equivalent .
[ with2functors ] | in this article , a new construction of derived equivalences is given .
it relates different endomorphism rings and more generally cohomological endomorphism rings - including higher extensions - of objects in triangulated categories .
these objects need to be connected by certain universal maps that are cohomological approximations and that exist in very general circumstances .
the construction turns out to be applicable in a wide variety of situations , covering finite dimensional algebras as well as certain infinite dimensional algebras , frobenius categories and @xmath0-calabi - yau categories . * * derived equivalences from cohomological approximations , + and mutations of @xmath1-yoneda algebras * wei hu * , * steffen koenig * and * changchang xi@xmath2 * corresponding author .
email : [email protected] ; fax : 0086 10 58802136 ; tel . : 0086 10 58808877 . ] |
the observed baryon asymmetry of the universe is one of the most interesting problems of particle physics and cosmology .
as was suggested by sakharov , the cosmological baryon excess can be generated through the baryon number violating decays of heavy particles @xcite . any microscopic theory able to reproduce the observed asymmetry
must satisfy sakharov s three conditions : * baryon ( or baryon minus lepton ) number non - conservation ; * c and cp violation ; * deviation from thermal equilibrium .
the condition of baryon number violation makes a strong appeal to physics beyond the standard model ( sm ) .
it has been proven to be difficult to generate an excess of baryons through direct violation of the baryon number
. a very attractive scenario of generating lepton number asymmetry through lepton number violating decays of heavy right handed neutrinos , which is later converted into the baryon number by sphalerons @xcite , has been suggested by fukugita and yanagida @xcite .
this scenario requires any viable extension of the sm to contain heavy right handed neutrinos , which , furthermore , are the reason for small masses of conventional neutrinos .
augmented with three right handed neutrinos , the sm itself fails to reproduce the observed baryon asymmetry of the universe @xmath3 due to the relatively small cp violation and too weak phase transition .
electroweak baryogenesis in the minimal supersymmetric standard model ( mssm ) is not yet ruled out , but is close to being so @xcite . since a significant amount of asymmetry is generated during the electroweak ( ew ) phase transition , from the practical point of view it is of interest to consider simple ` low energy ' extensions of the sm ( i.e. , models with additional @xmath4 or @xmath5 gauge symmetry and extended particle content ) able to reproduce the observed @xmath3 @xcite . from the theoretical point of view , it is interesting to investigate lepto- and baryogenesis in grand unification theories ( gut ) . in this paper ,
an attempt to derive ` low energy ' extensions of the sm , compatible with the fukugita yanagida scenario , from a supersymmetric gut model is made .
the choice of a supersymmetric model is motivated by two reasons .
first , scalar superpartners of right handed neutrinos can be used to break down @xmath1 symmetry spontaneously by the higgs mechanism .
second , supersymmetry ( if it is broken only softly , as is usually assumed ) provides cancellation of large quantum corrections associated with the presence of high scales in the theory .
a minimal gauge group with right handed neutrino contained in the same representation as the 15 known fermions is @xmath6 .
an additional representation is required to fit higgs fields responsible for breaking of the electroweak symmetry . following the ideas of grand unification
, it seems more natural to choose a gauge group which at the unification scale treats all states on an equal footing , i.e. , which contains the known states along with right handed neutrino and higgs doublets in the fundamental representation .
this is the case for the exceptional group @xmath0 , one of the most prominent candidates for a unified theory .
its characteristic features are as follows .
* it naturally follows from breaking of superstring @xmath7 with @xmath8 chiral superfields and * 78 * vector superfields as zero mode spectra . *
it allows chiral representations and its fundamental representation fits the fifteen known fermions along with right handed neutrino and two higgs doublets . *
gauge anomalies are automatically canceled . in , the particle content of the fundamental representation and possible charge assignments are discussed .
there are six charge assignments , compatible with the sm .
charge conservation in the processes involving states of different generations requires that the same charge assignment be used for all generations .
some of the fields in the adjoint representation may lead to a rapid proton decay .
constraints implied by a long - lived proton are considered in section .
in addition to the intermediate gauge groups listed in @xcite the @xmath9 is allowed for two charge assignments .
nevertheless , yukawa interactions implied by the residual @xmath10 symmetry make the rapid proton decay mediated by new down - type quarks unavoidable , unless those are very heavy
. discusses @xmath1 symmetry breaking and masses of the right handed neutrinos .
an interesting feature of the model under consideration is the presence of additional @xmath11 generations , scalar right handed neutrinos of which may be used to break @xmath1 symmetry spontaneously .
introduced simple discrete symmetry ensures that @xmath1 is broken at the scale which is sufficiently high for generating large masses of the right handed neutrinos and that right handed scalar neutrinos of the three known generations do not acquire a vacuum expectation value ( vev ) .
it also forbids yukawa couplings which , if present , would induce large masses of the conventional neutrinos .
the supersymmetric structure of the theory ensures that large quantum corrections to masses of scalars , associated with the presence of heavy gauge fields , cancel out . after the @xmath1 breaking ,
the ` low energy ' gauge group is @xmath12 . finally , is devoted to some concluding remarks .
the ten - dimensional @xmath13 heterotic superstring theory compactifies to the @xmath14 and yields a low energy theory with @xmath15 supersymmetry .
@xmath16 is the calabi yau manifold with @xmath17 holonomy .
if @xmath16 is simply connected , then @xmath13 gauge group breaks down to the @xmath18 subgroup . for a multiply connected manifold the initial gauge group breaks down to @xmath19 where g is a subgroup @xmath0 @xcite . in this scheme ,
chiral superfields @xmath20 and * 78 * vector superfields of @xmath0 emerge as the zero mode spectra . for a wide class of models
states in @xmath22 are denoted by @xmath23 and @xmath24 respectively , whereas states in @xmath25 are denoted by @xmath26 .
the fundamental representation of @xmath0 fits the 15 known states of the sm along with two higgs doublets ( @xmath27 and @xmath28 ) , a pair of down - type quarks ( @xmath29 and @xmath30 ) and two sm singlets @xmath31 and @xmath32 .
llllllllllll & + & + @xmath1 & @xmath33 & @xmath34 & @xmath35 & @xmath36 & & @xmath1 & @xmath33 & @xmath34 & @xmath35 & @xmath36 & + 1/3 & 1/3 & 1/2 & 2/3 & @xmath37 & & 0 & 1 & 1/2 & 1 & @xmath38 & + 1/3 & 1/3 & -1/2 & -1/3 & @xmath39 & & 0 & 1 & -1/2 & 0 & @xmath40 & + -1 & -1 & 1/2 & 0 & @xmath41 & & 0 & -1 & 1/2 & 0 & @xmath42 & + -1 & -1 & -1/2 & -1 & @xmath43 & & 0 & -1 & -1/2 & -1 & @xmath44 & + -1/3 & -4/3 & 0 & -2/3 & @xmath45 & & -2/3 & -2/3 & 0 & -1/3 & @xmath29 & + -1/3 & 2/3 & 0 & 1/3 & @xmath46 & & 2/3 & 2/3 & 0 & 1/3 & @xmath30 & + 1 & 2 & 0 & 1 & @xmath47 & & 1 & 0 & 0 & 0 & @xmath31 & + & 0 & 0 & 0 & 0 & @xmath32 & + a systematic study of quantum numbers of states in the fundamental and adjoint representations is conveniently performed using the cartan
weyl method @xcite . in the framework of this method , each state
is represented by its weight .
weights of * 27 * and * 78 * of @xmath0 are presented in tables [ weights_of_27 ] and [ weights_of_78 ] , respectively .
the hypercharge of any state is given by a scalar product of the hypercharge operator @xmath33 and the weight of the state .
one can easily check that the hypercharge assignment is not unique there are three assignments which reproduce the sm : @xmath48 there are also three @xmath1 operators , i.e. three assignments , which reproduce @xmath1 quantum numbers of the sm states : @xmath49 different assignments correspond to different embeddings of states into subgroups of @xmath0 . out of nine
@xmath50 combinations six are compatible with the sm : @xmath51 since @xmath1 is gauged , the majorana mass of the right handed neutrino , which is an essential ingredient of leptogenesis , is forbidden unless @xmath1 symmetry is broken down .
present data on neutrino masses as well as theoretical estimates of leptogenesis in other gut models favour the @xmath52 gev @xmath31 mass range @xcite . using weights , one can derive the form of trilinear interactions following the algorithm developed in @xcite . in the flavor basis
@xmath53+[\lambda^{ijk}_9 d_i(q_j q_k)+ \lambda^{ijk}_{10 } d_i^{\rm c } u_j^{\rm c } d_k^{\rm c}]\\ \fl\qquad+\lambda^{ijk}_{11}\nu_i^{\rm c}(l_j h_k^{\rm u})\nonumber\end{aligned}\ ] ] using , one can check that the corresponding weights add up to zero at each vertex .
the form of the superpotential is independent of the charge assignment in the sense that the change of assignment will only result in ` permutation ' of vertices .
see assignments i , ii and iii , for example : llllllll & [ @xmath54 & [ @xmath55 & [ @xmath56 & & [ @xmath57 & [ @xmath58 & [ @xmath59 + i & @xmath31 & @xmath41 & @xmath40 & & @xmath32 & @xmath42 & @xmath40 + ii & @xmath32 & @xmath42 & @xmath40 & & @xmath31 & @xmath41 & @xmath40 + iii & @xmath47 & @xmath41 & @xmath44 & & @xmath32 & @xmath38 & @xmath44 + along with vertices with @xmath60 ( i.e. states of the same generation ) superpotential ( [ superpotential ] ) necessarily contains terms with the states of different generations in one vertex .
if the charge assignment for one of the generations differs from that for other generations , then this results in nonzero sum of charges in a vertex .
for instance , if in the example above assignment ii is used for the first weight and assignment i is used for the second and the third weights then @xmath61 if instead of assignment ii , assignment iii is used for the first weight , then it is a sum of electric charges , which is nonzero .
this suggests that the same charge assignment should be used for all generations .
since @xmath10 is a subgroup of @xmath0 , the gauge sector of the model contains @xmath62 and @xmath33 bosons , which are known to mediate proton decay .
there are also new gauge fields leading to a rapid proton decay ( ) .
lllllllllll @xmath1 & @xmath33 & @xmath34 & @xmath35 & @xmath36 & @xmath1 & @xmath33 & @xmath34 & @xmath35 & @xmath36 + & + 2/3 & 5/3 & 1/2 & 4/3 & @xmath62 & -2/3 & 1/3 & 1/2 & 2/3 & @xmath63 + 2/3 & 5/3 & -1/2 & 1/3 & @xmath33 & -2/3 & 1/3 & -1/2 & -1/3 & @xmath64 + to assure that the proton is long lived , these fields must be very heavy of order of @xmath65 gev or more .
if the only source of masses of those particles is the vev of neutral scalars , then the masses of @xmath66 gauge bosons are determined by @xmath67 , @xmath33 boson mass is of the order of ew symmetry breaking scale and the @xmath62 bosons remain massless even after all neutral scalars develop a nonzero vev .
consequently , the gauge superfields which mediate proton decay have to become massive at the first stage of symmetry breaking .
if the manifold @xmath16 is multiply connected , then the effective higgs mechanism @xcite breaks symmetry at the compactification scale and induces large masses of the gauge fields . from the discussion above it follows that after @xmath13 breaking the gauge group of the model is not @xmath0 itself , but a subgroup @xmath68 of @xmath0 .
a very elaborate analysis of many possible breaking chains has been performed in @xcite .
it was argued there that a gauge field gets mass of the order of @xmath69 gev if @xmath70 . here
@xmath71 is the zero root breaking @xmath72 preserving direction , and @xmath73 is the weight of the gauge field .
scalar products @xmath74 for fundamental and adjoint representations of @xmath0 are listed , respectively , in tabels [ weights_of_27 ] and [ weights_of_78 ] .
lllllllllll su(5 ) & @xmath75 @xmath76 & i & ii & iii & iv & v & vi & ( z,@xmath73 ) + * 24 * & ( * 3*,*2*)@xmath77 & @xmath62 & @xmath62 & @xmath64 & @xmath39 & @xmath64 & @xmath39 & @xmath78 + * 24 * & ( * 3*,*2*)@xmath77 & @xmath33 & @xmath33 & @xmath63 & @xmath37 & @xmath63 & @xmath37 & @xmath78 + the requirement of @xmath70 for the gauge fields which mediate proton decay does not allow @xmath79 .
( * 3*,*2*)@xmath80 states ( and their charge conjugate ) are the only states in the adjoint of @xmath10 which are not automatically massless .
as seen from , the requirement that these fields be massless ( i.e. that @xmath10 is unbroken ) does not lead to a rapid gauge mediated proton decay for the charge assignments iv and vi . in other words , in addition to the intermediate gauge groups listed in @xcite as allowed , @xmath81 is allowed for the charge assignments iv and vi .
nevertheless , in this case the residual @xmath10 symmetry implies that couplings in ( [ superpotential ] ) are related in the following way ; @xmath82 and the rapid proton decay mediated by heavy down - type quarks @xmath29 and @xmath83 is unavoidable .
apart from scenarios with an extended color group there are only two options left : @xmath84 if the intermediate scale symmetry group is given by ( [ u1cube ] ) , then the gauge sector of the sm is supplemented by two neutral states ( the rest are superheavy ) , denoted by @xmath85 and @xmath86 ( see ) . if the intermediate scale symmetry group is given by ( [ su2lsu2r ] ) , then the gauge sector of the sm is supplemented by one @xmath87 singlet @xmath86 and one @xmath87 triplet @xmath88 .
there are three possible @xmath89 projections which result in one left
right symmetric ( @xmath90 ) and two skew left
right symmetric ( @xmath91 and @xmath92 ) models .
@xmath87 counterparts of @xmath85 for different choices of the projection and the charge assignment are given in .
lllllll & + & i & ii & iii & iv & v & vi + @xmath90 & @xmath93,@xmath94 & @xmath47,@xmath95 & @xmath93,@xmath94 & @xmath47,@xmath95 & @xmath31,@xmath96 & @xmath31,@xmath96 + @xmath91 & @xmath47,@xmath95 & @xmath93,@xmath94 & @xmath31,@xmath96 & @xmath31,@xmath96 & @xmath93,@xmath94 & @xmath47,@xmath95 + @xmath92 & @xmath31,@xmath96 & @xmath31,@xmath96 & @xmath47,@xmath95 & @xmath93,@xmath94 & @xmath47,@xmath95 & @xmath93,@xmath94 + while @xmath97 in ( [ su2lsu2r ] ) coincide with that of the standard model , @xmath98 is not the sm @xmath99 .
hypercharge @xmath33 , as well as @xmath1 , is a linear combination of @xmath100 , @xmath101 and @xmath102 .
an explicit form of gauge interactions @xmath103 $ ] for the first charge assignment and left
right symmetric model can be read off from : @xmath104\\ + \frac{g_{y_r}}2\ , y_r \left[-({\overline{d^c}d^c}+{\overline{u^c}u^c})+2{\overline{d^c}d^c}+({\overline{h^u}h^u}+ { \overline{h^d}h^d})\right.\nonumber\\ \left.-2{\overline{l}l}+({\overline{e^c}e^c}+{\overline{\nu^c}\nu^c})-2{\overline{s}s}\right ] + \frac{g_{w_r}}2\ , w^i_r \left[(d^c , u^c)^+\tau_i(d^c , u^c)\right.\nonumber\\ \left.+(h^u , h^d)^+\tau_i ( h^u , h^d)+(e^c,\nu^c)^+\tau_i ( e^c,\nu^c)\right]+\dots \nonumber\end{aligned}\ ] ] at the unification scale the relation among the gauge couplings is as follows : @xmath105 instead of @xmath106 , @xmath107 and @xmath108 one can use their linear combination .
for instance , in the limit ( [ unification ] ) linear combination @xmath109 is a gauge field of @xmath110 . for the later purposes it is useful to choose the linear combinations @xmath111 , @xmath112 , @xmath113 in such a way , that only one of the new fields interacts with right handed neutrinos and one of the fields interacts with neither of the sm singlets .
after the standard model singlets develop vevs , @xmath111 and @xmath112 acquire masses , while @xmath113 , which corresponds to @xmath114 , remains massless .
rewritten in terms of these fields , ( [ gaugeint ] ) takes the form @xmath115\\ + g_{\phi^{'}_0}\ , \phi^{'}_0\ , \left[-{\overline{q}q}-{\overline{u^c}u^c}-2{\overline{d^c}d^c}+3{\overline{d^c}d^c}+ 2{\overline{d}d}+2{\overline{h^u}h^u}+3{\overline{h^d}h^d}\right.\nonumber\\ \left.-2{\overline{l}l}-{\overline{e^c}e^c}-5{\overline{s}s}\right ] + g_{\omega^{'}_0}\ , \omega^{'}_0\ , \left[{\overline{q}q}+{\overline{u^c}u^c}-2{\overline{d^c}d^c}+{\overline{d^c}d^c}- 2{\overline{d}d}\right.\nonumber\\ \left.-2{\overline{h^u}h^u}+{\overline{h^d}h^d}-2{\overline{l}l}+{\overline{e^c}e^c}+4{\overline{\nu^c}\nu^c}+{\overline{s}s}\right ] \nonumber+\dots\end{aligned}\ ] ] at the unification scale gauge couplings @xmath116 , @xmath117 , @xmath118 are given by @xmath119 reexpessed in terms of @xmath111 , @xmath112 , @xmath113 gauge interactions of neutral fields are obviously given by ( [ physgaugeint ] ) irrespectively of the intermediate scale symmetry group @xmath68 , charge assignment or particular choice of @xmath120 projection .
there are two sm singlets whose scalar superpartners may be used to break symmetry down to the sm : @xmath32 and the right handed neutrino . the former has zero @xmath1 charge , while the later has @xmath121 .
therefore , it is the vev of the scalar superpartner of the right handed neutrino that breaks @xmath1 symmetry .
@xmath32 couples to higgs doublets and its vev @xmath122 is the origin of the @xmath123term : @xmath124 .
if the right handed sneutrino which develops the vev couples to states of the three known generations , it induces huge dirac masses for components of @xmath125 and @xmath27 doublets via the last term in ( [ superpotential ] ) . neglecting the possibility that one of the @xmath31 superfields is decoupled from the other states of the three known generations ,
one comes to the conclusion that all @xmath126 , and the @xmath1 symmetry is broken spontaneously by nonzero @xmath127 and @xmath128 and , consequently , that @xmath129 and @xmath130 are zero modes . according to @xcite chiral superfields in @xmath131
can be massive through the yukawa coupling @xmath132 . if @xmath70 for a component of * 27 * or @xmath133 with weight @xmath73 , the corresponding chiral superfield gets compactification scale mass , while @xmath134 chiral superfields remain massless .
shows , that for both discussed intermediate gauge groups @xmath135 and @xmath136 and any charge assignment , it is possible to have massless @xmath129 and @xmath130 . in the case of @xmath135
right handed neutrinos @xmath129 and @xmath130 are the only massless states in @xmath137 . in the case of @xmath136
the number of zero modes in @xmath138 depends on the charge assignment and the particular choice of @xmath139 projection ( ) .
lllllll & + & i & ii & iii & iv & v & vi + @xmath90 & @xmath140 & @xmath31 , @xmath141 & @xmath140 & @xmath31 , @xmath141 & @xmath142 & @xmath142 + @xmath91 & @xmath31 , @xmath141 & @xmath140 & @xmath142 & @xmath142 & @xmath140 & @xmath31 , @xmath141 + @xmath92 & @xmath142 & @xmath142 & @xmath31 , @xmath141 & @xmath140 & @xmath31 , @xmath141 & @xmath140 + if supersymmetry is exact , there are no negative mass squared terms needed to break @xmath110 down spontaneously by higgs mechanism . @xmath143 these terms are assumed to come from the @xmath144 sector , where supersymmetry is considered to break down spontaneously . in the gravity - mediated supersymmetry
breaking scenario the magnitude of the soft terms in the visible sector should be roughly of the order of @xmath145 .
for the commonly accepted value @xmath146 gev the scale of supersymmetry breaking in the hidden sector @xmath147 is about @xmath148 gev .
it is interesting to note that this value is of the same order as the desired mass scale of the right handed neutrinos .
one can consider this as a hint that at the stage when right handed neutrinos acquire masses , the temperature is still high enough to produce them thermally .
[ [ bf - su_c3times - su_l2times - u31-model . ] ] @xmath149 model .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + as is well known , the scalar potential consists of an @xmath150-term and @xmath29-term coming from the chiral superfield trilinear couplings and the gauge interactions respectively .
the renormalizable superpotential ( [ superpotential ] ) does not contain terms relevant for the symmetry breaking .
the scalar potential coming from gauge interactions ( [ physgaugeint ] ) and soft supersymmetry breaking is of the form @xmath151 ^ 2 - \left[m^2_{\chi}|\tilde{\chi}_{\nu^c}|^2 + m^2_{\overline{\chi}}|\tilde{\overline{\chi}}_{\nu^c}|^2\right ] + m^2_{\nu^c\,ij}\tilde{\nu}^{c*}_i\tilde{\nu}^c_j \label{dtermpot}\ ] ] on the one hand the vevs of right handed sneutrinos of @xmath138 are expected to be at least of the same order as masses of @xmath31 , i.e. @xmath152 gev ; on the other hand , such a huge vev should not generate large masses of scalar superpartners via the first term in ( [ dtermpot ] ) .
consequently , symmetry breaking should occur in the @xmath29flat direction @xmath153 .
combined with the requirement that all @xmath154 be zero , this means that contribution of the first term in ( [ dtermpot ] ) vanishes . to have symmetry breaking by the higgs mechanism in this direction , the sum of the mass parameters in the second term should be positive : @xmath155 non - renormalizable terms arise due to the interactions with exchange of superheavy fields , which correspond to excitations of internal degrees of freedom @xcite .
the general form of the non renormalizable superpotential is @xmath156 \nonumber\\ & + m^{-3}_c\left[b_1\ , \chi_{\nu^c}\chi_{\nu^c}\chi_{\nu^c}\overline{\chi}_{\nu^c } \overline{\chi}_{\nu^c}\overline{\chi}_{\nu^c}+b^i_2\ , \nu_i^c\chi_{\nu^c}\chi_{\nu^c}\overline{\chi}_{\nu^c } \overline{\chi}_{\nu^c}\overline{\chi}_{\nu^c}\right.\\ & \left.+b^{ij}_3\ , \nu_i^c\nu_j^c\chi_{\nu^c}\overline{\chi}_{\nu^c}\overline{\chi}_{\nu^c}\overline{\chi}_{\nu^c}+b^{ijk}_4\ , \nu_i^c\nu_j^c\nu_k^c\overline{\chi}_{\nu^c}\overline{\chi}_{\nu^c}\overline{\chi}_{\nu^c}\ , \right]+\dots\nonumber\end{aligned}\ ] ] given that all @xmath157 are zero , only @xmath158 terms in ( [ nrsuppot ] ) are relevant for the analysis of symmetry breaking .
these terms are invariant with respect to @xmath159 transformation , whereas soft supersymmetry breaking terms in ( [ dtermpot ] ) are not , unless @xmath160 .
if this condition is not satisfied , then @xmath161 can not be equal to @xmath162 .
nevertheless , since gauge coupling @xmath163 is of the order of unity , while @xmath164 is of many orders of magnitude smaller than unity , the deviation from the @xmath29flat direction @xmath165 is very small .
considering @xmath166 as a small perturbation in ( [ classicalpot ] ) one finds @xmath167 this effect can be entirely neglected in the analysis of non renormalizable superpotential ( [ nrsuppot ] ) . as for gauge interactions , such a deviation from the @xmath29flat direction will result in the generation of masses of scalars @xmath168 via the first term in ( [ dtermpot ] ) which add to mass terms coming from the soft supersymmetry breaking . to simplify the analysis @xmath169 and @xmath170
are taken to be equal in what follows .
if the first non - vanishing terms in ( [ nrsuppot ] ) are @xmath171 then generated vev and masses of right handed neutrinos @xmath31 are @xmath172 so that @xmath173 is of the order of @xmath174 gev for @xmath175 and is of the order of @xmath148 gev for @xmath176 .
large masses of right handed neutrinos suggest that @xmath176 and , consequently , @xmath177 in ( [ nrsuppot ] ) .
moreover , @xmath178 and @xmath179 , as well as coefficient @xmath180 of the ( not indicated ) next similar term in ( [ nrsuppot ] ) , are zero to avoid nonzero @xmath154 and dirac type mass for @xmath31 .
the discrete symmetry of the compactified manifold possibly accomplishes these conditions @xcite as well as the absence of bare mass terms @xmath181 and @xmath182 in the superpotential ( [ superpotential ] ) .
there is no reason to expect scale m to be below compactification scale @xmath183 so that the presence of the first term would make spontaneous breaking of @xmath1 by @xmath129 and @xmath130 impossible , while the presence of the second makes @xmath31 a component of the super heavy dirac neutrino . just as there are no @xmath181 and @xmath182 terms
, there are no soft supersymmetry breaking terms @xmath184 and @xmath185 in ( [ dtermpot ] ) . if present , the contributions of @xmath186 and @xmath187 terms are small and neglected in the following discussion .
then the classical potential is of the form @xmath188 + 9 b_1 b_1^{*}m_c^{-6}u^4 v^4\left[u^2+v^2\right]\\ & + 6 m_c^{-4}\left[b_1^{*}a_{ij}+b_1a^{*}_{ij}\right]\varrho_i\varrho_j v^3 u^3 + \frac{g^2 q_{\nu^c}^2}2 \left[\varrho_i\varrho_i+v^2-u^2\right]^2\nonumber\\ & -\left[m^2_{\chi}v^2+m^2_{\overline{\chi}}u^2\right]+ m^2_{\nu^c\,ij}\varrho_i\varrho_j\nonumber\end{aligned}\ ] ] with @xmath189 , @xmath190 , @xmath191 . for zero @xmath192 the d
flat direction is defined by @xmath193 and the minimum of the potential ( [ classicalpot ] ) corresponds to @xmath194{m_c^6(m^2_{\chi}+m^2_{\overline{\chi}})/(90|b_1|^2)},\qquad m_{\nu_c}^{ij}=2a_1^{ij}v^2_0 m_c^{-1}\ ] ] for nonzero @xmath192 d flat direction is defined by @xmath195 .
the set of products @xmath196 being considered as parameters , the minimizing of ( [ classicalpot ] ) with respect to @xmath197 gives @xmath197 as function of @xmath196 .
true vacuum corresponds to the set @xmath196 which minimizes @xmath198 .
expanding in powers of @xmath196 in the vicinity of @xmath199 and having in mind that partial derivative with respect to @xmath197 is zero : @xmath200\varrho_i\varrho_j\nonumber\end{aligned}\ ] ] since @xmath201 ratio is small while @xmath202 is large , the derivative ( [ expansion ] ) is dominated by the first term in square brackets which is a positive definite .
therefore , @xmath203 is at least a local minimum of the potential ( [ classicalpot ] ) for a wide range of parameters . for large @xmath196
it is sufficient to keep only the first term in ( [ classicalpot ] ) , and an explicit calculation shows that @xmath198 grows with growing @xmath196 , i.e. @xmath199 is a global minimum of the classical potential : @xmath204 a discrete symmetry which allows nonzero @xmath205 , @xmath206 and forbids nonzero @xmath207 , @xmath208 , @xmath209 couplings is essential for having large majorana masses for right handed neutrinos after symmetry breaking .
suppose that right
handed neutrinos @xmath31 , @xmath129 , @xmath130 acquire additional phases under transformations of the discrete symmetry : @xmath210 then from the requirement that nonzero @xmath205 and @xmath206 are allowed while @xmath207 , @xmath208 , @xmath179 , @xmath180 are set to zero by the symmetry : @xmath211 conditions ( [ allowed&forbidden ] ) imply that @xmath212 is not integer and , consequently , that @xmath213 : @xmath214 in other words , @xmath31 and @xmath129 have different transformation properties under the discrete symmetry . if the last term in ( [ superpotential ] ) which is responsible for both small neutrino masses via the see - saw mechanism and leptogenesis is allowed by this symmetry , then the term @xmath215 is necessarily forbidden just as was assumed .
bare mass term @xmath216 is not invariant under transformations of the discrete symmetry and therefore is not allowed . from equations ( [ allowed&forbidden ] ) it also follows that @xmath217 with @xmath218 odd , so that bare mass term @xmath219 is forbidden as well .
finally , coefficients @xmath186 and @xmath187 of the last two terms in ( [ nrsuppot ] ) vanish for the same reasons .
after @xmath220 and @xmath221 develop nonzero vevs , the @xmath4 symmetry , as well as the discrete symmetry ( [ transformations ] ) , is broken down .
components of chiral ( super)fields @xmath129 , @xmath130 and gauge ( super)field become massive .
as @xmath160 is assumed , vevs of @xmath220 and @xmath221 are equal and it is natural to introduce new fields @xmath222 and @xmath223 . the imaginary component of @xmath224 is ` eaten up ' by the vector gauge field @xmath225 , which acquires large mass @xmath226 gev .
the real component @xmath227 of @xmath224 acquire the same mass @xmath228 . from the analysis of gauge interactions alone
, it follows that two - component spinors @xmath229 ( superpartner of @xmath225 ) and @xmath230 ( superpartner of @xmath224 ) form a four component dirac spinor @xmath231 with mass @xmath232 .
renormalizable interactions induce majorana type mass term @xmath233 .
there is also majorana type mass term @xmath234 coming from the soft supersymmetry breaking , so that two linear combinations of @xmath230 and @xmath229 are majorana fermions with large ( @xmath235 ) and close masses .
renormalizable interactions ( [ nrsuppot ] ) also induce masses @xmath236 for the real component of @xmath237 and its fermionic superpartner @xmath238 .
the gauge boson @xmath225 , the fermion @xmath229 and the scalar @xmath227 interact with other states in the fundamental representation of @xmath0 ( in particular with higgses ) so that the self energy of scalars receives large contributions from diagrams with exchange of these heavy fields . for instance , see the case of one loop in . cccccc & & & & & + ( 60,60)(0,-10 ) ( 8,23)[bc]@xmath239 ( 52,23)[bc]@xmath239 ( 32,50)[bc]@xmath225 ( 30,20)2 ( 5,20)(30,20)2(30,20)(55,20)2(30,33)(13,0,360)214 & ( 60,60)(0,-10 ) ( 3,23)[bc]@xmath239 ( 57,23)[bc]@xmath239 ( 32,0)[tc]@xmath239 ( 32,40)[bc]@xmath225 ( 0,20)(15,20)2 ( 45,20)(60,20)2 ( 15,20)2 ( 45,20)2 ( 30,20)(15,0,180)27 ( 30,20)(15,180,0)4 & ( 60,60)(0,-10 ) ( 3,23)[bc]@xmath239 ( 57,23)[bc]@xmath239 ( 32,0)[tc]@xmath240 ( 32,40)[bc]@xmath229 ( 0,20)(15,20)2 ( 45,20)(60,20)2 ( 15,20)2 ( 45,20)2 ( 30,20)(15,0,180 ) ( 30,20)(15,0,180)27 ( 30,20)(15,180,0 ) & ( 60,60)(0,-10 ) ( 8,23)[bc]@xmath239 ( 52,23)[bc]@xmath239 ( 32,50)[bc]@xmath239 ( 30,20)2 ( 5,20)(30,20)2(30,20)(55,20)2(30,33)(13,0,360)4 & ( 60,60)(0,-10 ) ( 3,23)[bc]@xmath239 ( 57,23)[bc]@xmath239 ( 32,0)[t]@xmath239 ( 32,40)[b]@xmath227 ( 15,20)2 ( 45,20)2 ( 0,20)(15,20)2 ( 45,20)(60,20)2 ( 30,20)(15,0,180)4 ( 30,20)(15,180,0)4 & ( 60,60)(0,-10 ) ( 3,23)[bc]@xmath239 ( 57,23)[bc]@xmath239 ( 32,0)[t]@xmath239 ( 30,45)[bc]@xmath239 ( 35,25)[b]@xmath227 ( 30,20)2 ( 30,35)2 ( 0,20)(60,20)2 ( 30,20)(30,35)2 ( 30,50)(15,0,180)4 ( 30,50)(15,180,0)4 + ( a ) & ( b ) & ( c ) & ( d ) & ( e ) & ( f ) + as has already been mentioned in the introduction , the supersymmetric structure of the lagrangian ensures that all large corrections , associated with the exchange of heavy fields , cancel out and only terms proportional to the soft supersymmetry breaking parameters remain : @xmath241[(k - p)^2-y]}\ ] ] [ [ bf - su_c3times - su_l2times - su_r2times - u21-models . ] ] @xmath242 models .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + tables [ masslessadj ] and [ masslessfund ] show , that it is sufficient to consider only the case of the first charge assignment and three @xmath243 projections , the other cases being completely analogous . in the case of the left
right symmetric ( @xmath90 ) model , the zero modes in @xmath244 are @xmath87 doublets @xmath245,@xmath129 ) and @xmath246,@xmath247 ) . the scalar potential coming from gauge interactions
( [ gaugeint ] ) and soft supersymmetry breaking is of the form @xmath248 ^ 2 + \frac{g^2_{w_r}}2\left[\left(\tilde{\chi}_r^{*}\frac{\tau_i}2\tilde{\chi}_r\right)+ \left(\tilde{\overline{\chi}}_r^{*}\frac{\tau_i}2\tilde{\overline{\chi}}_r\right)\right]^2\\ -\left[m^2_{\chi}\left(\tilde{\chi}_r^{*}\tilde{\chi}_r\right ) + m^2_{\overline{\chi}}\left(\tilde{\overline{\chi}}_r^ { * } \tilde{\overline{\chi}}_r\right)\right]\nonumber\end{aligned}\ ] ] let @xmath249 , @xmath250 and @xmath251 , @xmath252 .
the corresponding classical potential is @xmath253\\ + \frac12\ , g^2_{w_r}|\left[v_2^{*}v_1+u_1^{*}u_2\right]|^2 + \frac18\ , g^2_{w_r}\left[\left(|v_2|^2-|v_1|^2\right)+ \left(|u_1|^2-|u_2|^2\right)\right]\nonumber\\- \left[m^2_{\chi}\left(|v_1|^2+|v_2|^2\right)+ m^2_{\overline{\chi}}\left(|u_1|^2+|u_2|^2\right)\right]\nonumber\end{aligned}\ ] ] there are symmetry breaking directions , which are @xmath29 flat .
vanishing of the first and the third term in ( [ lrclasspot ] ) requires that @xmath254 and @xmath255 . with these conditions
satisfied , the second term in ( [ lrclasspot ] ) vanishes if @xmath256 or @xmath257 or @xmath258 .
the non - renormalizable superpotential , which bounds the classical potential from below , is similar to ( [ nrsuppot ] ) with @xmath31 , @xmath129 and @xmath130 replaced with @xmath87 doublets @xmath259 , @xmath260 and @xmath261 .
its explicit form is @xmath262\\ + m_c^{-3}\left[(\chi_r\overline{\chi}_r)(\chi_r\overline{\chi}_r ) ( \chi_r\overline{\chi}_r)+b^i ( \psi_r^i\overline{\chi}_r ) ( \chi_r\overline{\chi}_r)(\chi_r\overline{\chi}_r)\right.\nonumber\\ \left
. + b_3^{ij}(\psi_r^i\overline{\chi}_r)(h^j\overline{\chi}_r ) ( \chi_r\overline{\chi}_r)+b_4^{ijk}(\psi_r^i\overline{\chi}_r ) ( \psi_r^j\overline{\chi}_r)(\psi_r^k\overline{\chi}_r)\right]+\dots \nonumber\end{aligned}\ ] ] the freedom of @xmath87 gauge transformations allows to rotate away a possible vev for one of the isospin components of one of the scalar fields , so one can take @xmath263 at the minimum of the potential . since the classical potential under consideration reaches its minimum in one of the @xmath29 flat directions , @xmath264 is equal to zero as well
. then the following analysis is the same as in the case of the @xmath265 model and furnishes the same result .
the symmetry is broken down to @xmath266 .
the second skew left
right symmetric ( @xmath92 ) model differs from the one above in @xmath32 instead of @xmath47 being an @xmath267 counterpart of the right handed neutrino : @xmath268 ) and @xmath269 ) .
the symmetry is broken down to @xmath12 as well . in the case of the first skew left
right symmetric ( @xmath91 ) model , the right handed neutrino is an @xmath270 singlet .
massless states in @xmath271 are @xmath129 , @xmath272,@xmath273 ) and @xmath130 , @xmath274,@xmath275 ) .
the classical potential coming from renormalizable and nonrenormalizable interactions is similar to ( [ classicalpot ] ) and yields the same results .
the symmetry is broken down to @xmath276 . since the higgs doublets @xmath277,@xmath273 ) and
@xmath278,@xmath275 ) are contained in @xmath279 as zero modes , there are directions in which the @xmath29-term potential vanishes for whatever large vevs of these fields , i.e. , there is a risk of breaking electroweak symmetry at a very high scale . to avoid it ,
the part of the classical potential which comes from the soft supersymmetry breaking should be positive in those directions , as is the case in the mssm .
the coexistence of all the terms in the second row of ( [ superpotential ] ) leads to the rapid proton decay , mediated by new @xmath29 and @xmath30 quarks , unless those are very heavy .
vev of @xmath32 which gives masses @xmath280 to @xmath29 and @xmath30 is also the source of @xmath123term @xmath281 .
although neither for @xmath135 nor for @xmath136 the couplings @xmath282 and @xmath282 are related by symmetry , it is not natural to expect @xmath29 and @xmath30 to be much heavier than 1 tev which is insufficient to suppress the proton decay . a solution to this problem may be provided by an appropriate discrete symmetry , which forbids some of the couplings in ( [ superpotential ] ) .
lllllllllllllllllllll model & & & + & & & & & & & & & + & @xmath46 & @xmath45 & @xmath27 & @xmath28 & @xmath47 & @xmath31 & @xmath30 & @xmath45 & @xmath27 & @xmath125 & @xmath47 & @xmath32 & @xmath30 & @xmath46 & @xmath28 & @xmath125 & @xmath31 & @xmath32 + 1 & + & + & + & + & - & - & - & + & + & - & - & + & - & + & + & - & - & + + 2 & + & + & + & + & - & - & + & + & + & - & - & + & + & + & + & - & - & + + if this is a @xmath283 symmetry , then there are only two models @xcite compatible with leptogenesis and nonzero neutrino masses . in the first model @xmath125 and @xmath47 , @xmath31 , @xmath29 ,
@xmath30 are odd , while the rest of the states are even , so that @xmath284 .
the second model differs from the first in @xmath29 and @xmath30 being even , so that @xmath285 .
transformations of the discrete symmetry should commute with transformations of the gauge symmetry .
shows , that this condition is satisfied only in the case of the left
right symmetric ( @xmath90 ) model , while in both skew left
right symmetric models components of @xmath87 doublets transform differently .
therefore , if proton stability is assured by such a discrete symmetry , the only allowed gauge group after breaking of @xmath68 is @xmath286 .
in this paper , an attempt to derive ` low energy ' extensions of the sm , compatible with the fukugita yanagida scenario , from a superstring inspired @xmath0 model has been made .
this model allows for six charge assignments compatible with the standard model .
charge conservation in the processes involving states of different generations requires that the same charge assignment be used for all generations .
as is discussed above , the gauge symmetry is broken in a sequence of stages .
the first stage is due to calabi yau compactification and the effective higgs mechanism .
the condition that the proton is long - lived requires that the symmetry is broken either to @xmath135 or @xmath136 .
as the temperature goes down , supersymmetry breaks down spontaneously in the hidden sector .
supersymmetry breaking is mediated to the visible sector by gravity and manifests itself in the soft terms . at the next stage right
handed scalar neutrinos and right handed neutrinos of two additional generations @xmath287 develop nonzero vev , breaking the @xmath1 symmetry .
the introduced simple discrete symmetry ensures that @xmath1 is broken at the scale which is sufficiently high for generating large masses of the right handed neutrinos and that right handed scalar neutrinos of the three known generations do not acquire vev .
it also forbids yukawa couplings which , if present , would induce large masses of the conventional neutrinos .
the supersymmetric structure of the theory ensures that large quantum corrections to masses of scalars , associated with the presence of heavy gauge fields , cancel out .
provided that a rapid proton decay mediated by the new quarks is forbidden by a @xmath283 symmetry , the sought ` low energy ' gauge group is given by @xmath288 for both allowed intermediate scale gauge groups @xmath135 and @xmath136 .
apart from the additional ( compared to the sm ) @xmath4 symmetry , the characteristic feature of this ` low energy ' model is the extended particle content .
in addition to the known particles and right handed neutrinos it contains the sm singlets s , new heavy quarks and three generations of higgses , as well as their superpartners . since both new quarks and higgses couple to right handed neutrinos , there are more @xmath1 violating decay channels than in the sm or its supersymmetric extension . at the same time
there are more processes which wash the generated @xmath1 asymmetry out .
the interplay of these effects requires further investigation .
i would like to thank professor e a paschos for inspiration and useful discussions .
financial support from the graduiertenkolleg 841 physik der elementarteilchen an beschleunigern und i m universum at dortmund university is gratefully acknowledged .
20 g. t hooft , phys . rev . lett . * 37 * , 8 ( 1976 ) .
v. a. kuzmin , v. a. rubakov and m. e. shaposhnikov , phys . lett . *
b155 * , 36 ( 1985 ) .
m. fukugita and t. yanagida , phys . lett . *
b174 * , 45 ( 1986 ) .
j. orloff , hep - ph/0307351 .
a. s. joshipura and e. a. paschos , hep - ph/9906498 .
a. s. joshipura , e. a. paschos and w. rodejohann , jhep * 08 * , 029 ( 2001 ) , [ hep - ph/0105175 ] .
a. s. joshipura , e. a. paschos and w. rodejohann , nucl . phys . *
b611 * , 227 ( 2001 ) , [ hep - ph/0104228 ] .
t. matsuoka and d. suematsu , prog .
. phys . * 76 * , 886 ( 1986 ) .
e. witten , nucl .
b258 * , 75 ( 1985 ) .
e. dynkin , am .
, ser . 2 * 6 * , 245 ( 1957 ) .
r. slansky , phys . rept . * 79 * , 1 ( 1981 ) .
m. plumacher , hep - ph/9807557 .
g. w. anderson and t. blazek , j. math .
phys . * 41 * , 4808 ( 2000 ) , [ hep - ph/9912365 ] . y. hosotani , phys . lett . * b126 * , 309 ( 1983 ) . t. matsuoka and d. suematsu , nucl . phys .
* b274 * , 106 ( 1986 ) .
t. hambye , e. ma , m. raidal and u. sarkar , phys . lett . *
b512 * , 373 ( 2001 ) , [ hep - ph/0011197 ] .
m. dine , v. kaplunovsky , m. l. mangano , c. nappi and n. seiberg , nucl .
b259 * , 549 ( 1985 ) . c. a. lutken and g. g. ross , phys
b214 * , 357 ( 1988 ) .
e. ma , phys .
lett . * 60 * , 1363 ( 1988 ) .
lllllllllllllll @xmath0 & & & & @xmath74 + & & & & + & @xmath289 & @xmath125 & @xmath90 & @xmath125 & @xmath90 & @xmath91 & @xmath92 & i & ii & iii & iv & v & vi & + & * 3 * & * 3 * & * 1 * & * 2 * & & & + [ @xmath290 & [ @xmath291 & [ @xmath291 & [ @xmath292 & @xmath293,_{1}$ ] & @xmath294,_{0}$ ] & @xmath294,_{0}$ ] & @xmath294,_{0}$ ] & u & u & u & u & u & u & @xmath295 + [ @xmath296 & [ @xmath297 & [ @xmath291 & [ @xmath292 & @xmath293,_{1}$ ] & @xmath294,_{0}$ ] & @xmath294,_{0}$ ] & @xmath294,_{0}$ ] & u & u & u & u & u & u & @xmath295 + [ @xmath298 & [ @xmath299 & [ @xmath291 & [ @xmath292 & @xmath293,_{1}$ ] & @xmath294,_{0}$ ] & @xmath294,_{0}$ ] & @xmath294,_{0}$ ] & u & u & u & u & u & u & @xmath295 + [ @xmath300 & [ @xmath291 & [ @xmath297 & [ @xmath292 & @xmath301,_{1}$ ] & @xmath294,_{0}$ ] & @xmath294,_{0}$ ] & @xmath294,_{0}$ ] & d & d & d & d & d & d & @xmath295 + [ @xmath302 & [ @xmath297 & [ @xmath297 & [ @xmath292 & @xmath301,_{1}$ ] & @xmath294,_{0}$ ] & @xmath294,_{0}$ ] & @xmath294,_{0}$ ] & d & d & d & d & d & d & @xmath295 + [ @xmath303 & [ @xmath299 & [ @xmath297 & [ @xmath292 & @xmath301,_{1}$ ] & @xmath294,_{0}$ ] & @xmath294,_{0}$ ] & @xmath294,_{0}$ ] & d & d & d & d & d & d & @xmath295 + & * 3 * & * 3 * & * 1 * & * 1 * & & & + & [ @xmath291 & [ @xmath299 & [ @xmath292 & @xmath294,_{\bar{2}}$ ] & @xmath294,_{0}$ ] & @xmath294,_{0}$ ] & @xmath294,_{0}$ ] & d & d & d & d & d & d & @xmath304 + [ @xmath305 & [ @xmath297 & [ @xmath299 & [ @xmath292 & @xmath294,_{\bar{2}}$ ] & @xmath294,_{0}$ ] & @xmath294,_{0}$ ] & @xmath294,_{0}$ ] & d & d & d & d & d & d & @xmath304 + [ @xmath306 & [ @xmath299 & [ @xmath299 & [ @xmath292 & @xmath294,_{\bar{2}}$ ] & @xmath294,_{0}$ ] & @xmath294,_{0}$ ] & @xmath294,_{0}$ ] & d & d & d & d & d & d & @xmath304 + & @xmath307 & * 1 * & @xmath307 & * 1 * & & & & & + & [ @xmath308 & [ @xmath292 & [ @xmath309 & @xmath294,_{0}$ ] & @xmath301,_{\bar{1}}$ ] & @xmath301,_{\bar{1}}$ ] & @xmath294,_{2}$ ] & @xmath45 & @xmath45 & @xmath46 & @xmath30 & @xmath46 & @xmath30 & @xmath310 + [ @xmath311 & [ @xmath312 & [ @xmath292 & [ @xmath309 & @xmath294,_{0}$ ] & @xmath301,_{\bar{1}}$ ] & @xmath301,_{\bar{1}}$ ] & @xmath294,_{2}$ ] & @xmath45 & @xmath45 & @xmath46 & @xmath30 & @xmath46 & @xmath30 & @xmath310 + [ @xmath313 & [ @xmath309 & [ @xmath292 & [ @xmath309 & @xmath294,_{0}$ ] & @xmath301,_{\bar{1}}$ ] & @xmath301,_{\bar{1}}$ ] & @xmath294,_{2}$ ] & @xmath45 & @xmath45 & @xmath46 & @xmath30 & @xmath46 & @xmath30 & @xmath310 + [ @xmath314 & [ @xmath308 & [ @xmath292 & [ @xmath308 & @xmath294,_{0}$ ] & @xmath294,_{2}$ ] & @xmath293,_{\bar{1}}$ ] & @xmath293,_{\bar{1}}$ ] & @xmath30 & @xmath46 & @xmath30 & @xmath46 & @xmath45 & @xmath45 & @xmath315 + [ @xmath316 & [ @xmath312 & [ @xmath292 & [ @xmath308 & @xmath294,_{0}$ ] & @xmath294,_{2}$ ] & @xmath293,_{\bar{1}}$ ] & @xmath293,_{\bar{1}}$ ] & @xmath30 & @xmath46 & @xmath30 & @xmath46 & @xmath45 & @xmath45 & @xmath315 + [ @xmath317 & [ @xmath309 & [ @xmath292 & [ @xmath308 & @xmath294,_{0}$ ] & @xmath294,_{2}$ ] & @xmath293,_{\bar{1}}$ ] & @xmath293,_{\bar{1}}$ ] & @xmath30 & @xmath46 & @xmath30 & @xmath46 & @xmath45 & @xmath45 & @xmath315 + [ @xmath318 & [ @xmath308 & [ @xmath292 & [ @xmath312 & @xmath294,_{0}$ ] & @xmath293,_{\bar{1}}$ ] & @xmath294,_{2}$ ] & @xmath301,_{\bar{1}}$ ] & @xmath46 & @xmath30 & @xmath45 & @xmath45 & @xmath30 & @xmath46 & @xmath319 + [ @xmath320 & [ @xmath312 & [ @xmath292 & [ @xmath312 & @xmath294,_{0}$ ] & @xmath293,_{\bar{1}}$ ] & @xmath294,_{2}$ ] & @xmath301,_{\bar{1}}$ ] & @xmath46 & @xmath30 & @xmath45 & @xmath45 & @xmath30 & @xmath46 & @xmath319 + [ @xmath321 & [ @xmath309 & [ @xmath292 & [ @xmath312 & @xmath294,_{0}$ ] & @xmath293,_{\bar{1}}$ ] & @xmath294,_{2}$ ] & @xmath301,_{\bar{1}}$ ] & @xmath46 & @xmath30 & @xmath45 & @xmath45 & @xmath30 & @xmath46 & @xmath319 + & * 1 * & @xmath307 & * 3 * & * 2 * & & & & & + & [ @xmath292 & [ @xmath312 & [ @xmath291 & @xmath293,_{\bar{1}}$ ] & @xmath293,_{1}$ ] & @xmath293,_{1}$ ] & @xmath294,_{\bar{2}}$ ] & @xmath38 & @xmath38 & n & @xmath41 & n & @xmath41 & @xmath322 + [ @xmath56 & [ @xmath292 & [ @xmath309 & [ @xmath291 & @xmath301,_{\bar{1}}$ ] & @xmath293,_{1}$ ] & @xmath293,_{1}$ ] & @xmath294,_{\bar{2}}$ ] & @xmath40 & @xmath40 & e & e & e & e & @xmath322 + [ @xmath58 & [ @xmath292 & [ @xmath312 & [ @xmath297 & @xmath293,_{\bar{1}}$ ] & @xmath301,_{1}$ ] & @xmath294,_{\bar{2}}$ ] & @xmath293,_{1}$ ] & n & @xmath41 & @xmath38 & @xmath38 & @xmath41 & n & @xmath323 + [ @xmath324 & [ @xmath292 & [ @xmath309 & [ @xmath297 & @xmath301,_{\bar{1}}$ ] & @xmath301,_{1}$ ] & @xmath294,_{\bar{2}}$ ] & @xmath293,_{1}$ ] & e & e & @xmath40 & @xmath40 & e & e & @xmath323 + [ @xmath55 & [ @xmath292 & [ @xmath312 & [ @xmath299 & @xmath293,_{\bar{1}}$ ] & @xmath294,_{\bar{2}}$ ] & @xmath301,_{1}$ ] & @xmath301,_{1}$ ] & @xmath41 & n & @xmath41 & n & @xmath38 & @xmath38 & @xmath325 + [ @xmath326 & [ @xmath292 & [ @xmath309 & [ @xmath299 & @xmath301,_{\bar{1}}$ ] & @xmath294,_{\bar{2}}$ ] & @xmath301,_{1}$ ] & @xmath301,_{1}$ ] & e & e & e & e & @xmath40 & @xmath40 & @xmath325 + & * 1 * & @xmath307 & * 3 * & * 1 * & & & & & + & [ @xmath292 & [ @xmath308 & [ @xmath291 & @xmath294,_{2}$ ] & @xmath293,_{1}$ ] & @xmath293,_{1}$ ] & @xmath294,_{\bar{2}}$ ] & @xmath47 & @xmath47 & @xmath31 & s & @xmath31 & s & @xmath327 + [ @xmath54 & [ @xmath292 & [ @xmath308 & [ @xmath297 & @xmath294,_{2}$ ] & @xmath301,_{1}$ ] & @xmath294,_{\bar{2}}$ ] & @xmath293,_{1}$ ] & @xmath31 & s & @xmath47 & @xmath47 & s & @xmath31 & @xmath328 + [ @xmath57 & [ @xmath292 & [ @xmath308 & [ @xmath299 & @xmath294,_{2}$ ] & @xmath294,_{\bar{2}}$ ] & @xmath301,_{1}$ ] & @xmath301,_{1}$ ] & s & @xmath31 & s & @xmath31 & @xmath47 & @xmath47 & @xmath329 + lllllllllllllll @xmath0 & & & & @xmath74 + & & & + & @xmath289 & @xmath125 & @xmath90 & @xmath125 & @xmath90 & @xmath91 & @xmath92 & i & ii & iii & iv & v & vi & + & * 1 * & * 8 * & * 1 * + [ @xmath330 & [ @xmath292 & [ @xmath331 & [ @xmath292 & [ @xmath332,@xmath333 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & @xmath336 & @xmath336 & @xmath336 & @xmath336 & @xmath336 & @xmath336 & @xmath337 + [ @xmath338 & [ @xmath292 & [ @xmath339 & [ @xmath292 & [ @xmath340,@xmath333 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & @xmath341 & @xmath341 & @xmath341 & @xmath341 & @xmath341 & @xmath341 & @xmath337 + [ @xmath342 & [ @xmath292 & [ @xmath343 & [ @xmath292 & [ @xmath344,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & & & & & & & @xmath345 + [ @xmath346 & [ @xmath292 & [ @xmath292 & [ @xmath292 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & & & & & & & @xmath345 + [ @xmath347 & [ @xmath292 & [ @xmath348 & [ @xmath292 & [ @xmath349,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & & & & & & & @xmath345 + [ @xmath350 & [ @xmath292 & [ @xmath351 & [ @xmath292 & [ @xmath332,@xmath352 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & @xmath41 & @xmath41 & @xmath41 & @xmath41 & @xmath41 & @xmath41 & @xmath353 + [ @xmath354 & [ @xmath292 & [ @xmath355 & [ @xmath292 & [ @xmath340,@xmath352 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & @xmath43 & @xmath43 & @xmath43 & @xmath43 & @xmath43 & @xmath43 & @xmath353 + [ @xmath346 & [ @xmath292 & [ @xmath292 & [ @xmath292 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & @xmath356 & @xmath356 & @xmath356 & @xmath356 & @xmath356 & @xmath356 & @xmath345 + & * 1 * & * 1 * & * 8 * + & [ @xmath292 & [ @xmath292 & [ @xmath331 & [ @xmath334,@xmath335 & [ @xmath332,@xmath333 & [ @xmath344,@xmath335 & [ @xmath332,@xmath352 & @xmath47 & @xmath93 & @xmath31 & @xmath357 & @xmath358 & @xmath359 & @xmath360 + [ @xmath361 & [ @xmath292 & [ @xmath292 & [ @xmath339 & [ @xmath334,@xmath335 & [ @xmath340,@xmath333 & [ @xmath332,@xmath352 & [ @xmath344,@xmath335 & @xmath31 & @xmath357 & @xmath47 & @xmath93 & @xmath359 & @xmath358 & @xmath362 + [ @xmath363 & [ @xmath292 & [ @xmath292 & [ @xmath343 & [ @xmath334,@xmath335 & [ @xmath344,@xmath335 & [ @xmath332,@xmath333 & [ @xmath340,@xmath352 & @xmath93 & @xmath47 & @xmath358 & @xmath359 & @xmath31 & @xmath357 & @xmath364 + [ @xmath346 & [ @xmath292 & [ @xmath292 & [ @xmath292 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & @xmath85 & @xmath85 & @xmath85 & @xmath85 & @xmath85 & @xmath85 & @xmath345 + [ @xmath365 & [ @xmath292 & [ @xmath292 & [ @xmath348 & [ @xmath334,@xmath335 & [ @xmath349,@xmath335 & [ @xmath340,@xmath352 & [ @xmath332,@xmath333 & @xmath358 & @xmath359 & @xmath93 & @xmath47 & @xmath357 & @xmath31 & @xmath366 + [ @xmath367 & [ @xmath292 & [ @xmath292 & [ @xmath351 & [ @xmath334,@xmath335 & [ @xmath332,@xmath352 & [ @xmath340,@xmath333 & [ @xmath349,@xmath335 & @xmath357 & @xmath31 & @xmath359 & @xmath358 & @xmath47 & @xmath93 & @xmath368 + [ @xmath369 & [ @xmath292 & [ @xmath292 & [ @xmath355 & [ @xmath334,@xmath335 & [ @xmath340,@xmath352 & [ @xmath349,@xmath335 & [ @xmath340,@xmath333 & @xmath359 & @xmath358 & @xmath357 & @xmath31 & @xmath93 & @xmath47 & @xmath370 + [ @xmath346 & [ @xmath292 & [ @xmath292 & [ @xmath292 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & @xmath86 & @xmath86 & @xmath86 & @xmath86 & @xmath86 & @xmath86 & @xmath345 + & * 8 * & * 1 * & * 1 * + [ @xmath371 & [ @xmath331 & [ @xmath292 & [ @xmath292 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath345 + [ @xmath372 & [ @xmath339 & [ @xmath292 & [ @xmath292 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath345 + [ @xmath373 & [ @xmath343 & [ @xmath292 & [ @xmath292 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath345 + [ @xmath346 & [ @xmath292 & [ @xmath292 & [ @xmath292 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath345 + [ @xmath346 & [ @xmath292 & [ @xmath292 & [ @xmath292 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath345 + [ @xmath374 & [ @xmath351 & [ @xmath292 & [ @xmath292 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath345 + [ @xmath375 & [ @xmath348 & [ @xmath292 & [ @xmath292 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath345 + [ @xmath376 & [ @xmath355 & [ @xmath292 & [ @xmath292 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & [ @xmath334,@xmath335 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath163 & @xmath345 + & @xmath377 & * 3 * & * 3 * + [ @xmath378 & [ @xmath308 & [ @xmath291 & [ @xmath291 & [ @xmath332,@xmath379 & [ @xmath332,@xmath379 & [ @xmath332,@xmath379 & [ @xmath334,@xmath380 & @xmath381 & @xmath381 & @xmath382 & @xmath383 & @xmath382 & @xmath383 & @xmath384 + [ @xmath385 & [ @xmath312 & [ @xmath291 & [ @xmath291 & [ @xmath332,@xmath379 & [ @xmath332,@xmath379 & [ @xmath332,@xmath379 & [ @xmath334,@xmath380 & @xmath381 & @xmath381 & @xmath382 & @xmath383 & @xmath382 & @xmath383 & @xmath384 + [ @xmath386 & [ @xmath309 & [ @xmath291 & [ @xmath291 & [ @xmath332,@xmath379 & [ @xmath332,@xmath379 & [ @xmath332,@xmath379 & [ @xmath334,@xmath380 & @xmath381 & @xmath381 & @xmath382 & @xmath383 & @xmath382 & @xmath383 & @xmath384 + [ @xmath387 & [ @xmath308 & [ @xmath291 & [ @xmath297 & [ @xmath332,@xmath379 & [ @xmath340,@xmath379 & [ @xmath334,@xmath380 & [ @xmath332,@xmath379 & @xmath382 & @xmath383 & @xmath381 & @xmath381 & @xmath383 & @xmath382 & @xmath388 + [ @xmath389 & [ @xmath312 & [ @xmath291 & [ @xmath297 & [ @xmath332,@xmath379 & [ @xmath340,@xmath379 & [ @xmath334,@xmath380 & [ @xmath332,@xmath379 & @xmath382 & @xmath383 & @xmath381 & @xmath381 & @xmath383 & @xmath382 & @xmath388 + [ @xmath390 & [ @xmath309 & [ @xmath291 & [ @xmath297 & [ @xmath332,@xmath379 & [ @xmath340,@xmath379 & [ @xmath334,@xmath380 & [ @xmath332,@xmath379 & @xmath382 & @xmath383 & @xmath381 & @xmath381 & @xmath383 & @xmath382 & @xmath388 + [ @xmath391 & [ @xmath308 & [ @xmath291 & [ @xmath299 & [ @xmath332,@xmath379 & [ @xmath334,@xmath380 & [ @xmath340,@xmath379 & [ @xmath340,@xmath379 & @xmath383 & @xmath382 & @xmath383 & @xmath382 & @xmath381 & @xmath381 & @xmath392 + [ @xmath393 & [ @xmath312 & [ @xmath291 & [ @xmath299 & [ @xmath332,@xmath379 & [ @xmath334,@xmath380 & [ @xmath340,@xmath379 & [ @xmath340,@xmath379 & @xmath383 & @xmath382 & @xmath383 & @xmath382 & @xmath381 & @xmath381 & @xmath392 + [ @xmath394 & [ @xmath309 & [ @xmath291 & [ @xmath299 & [ @xmath332,@xmath379 & [ @xmath334,@xmath380 & [ @xmath340,@xmath379 & [ @xmath340,@xmath379 & @xmath383 & @xmath382 & @xmath383 & @xmath382 & @xmath381 & @xmath381 & @xmath392 + [ @xmath395 & [ @xmath308 & [ @xmath297 & [ @xmath291 & [ @xmath340,@xmath379 & [ @xmath332,@xmath379 & [ @xmath332,@xmath379 & [ @xmath334,@xmath380 & @xmath396 & @xmath396 & @xmath397 & @xmath398 & @xmath397 & @xmath398 & @xmath384 + lllllllllllllll & @xmath377 & * 3 * & * 3 * + [ @xmath399 & [ @xmath312 & [ @xmath297 & [ @xmath291 & [ @xmath340,@xmath379 & [ @xmath332,@xmath379 & [ @xmath332,@xmath379 & [ @xmath334,@xmath380 & @xmath396 & @xmath396 & @xmath397 & @xmath398 & @xmath397 & @xmath398 & @xmath384 + [ @xmath400 & [ @xmath309 & [ @xmath297 & [ @xmath291 & [ @xmath340,@xmath379 & [ @xmath332,@xmath379 & [ @xmath332,@xmath379 & [ @xmath334,@xmath380 & @xmath396 & @xmath396 & @xmath397 & @xmath398 & @xmath397 & @xmath398 & @xmath384 + [ @xmath401 & [ @xmath308 & [ @xmath297 & [ @xmath297 & [ @xmath340,@xmath379 & [ @xmath340,@xmath379 & [ @xmath334,@xmath380 & [ @xmath332,@xmath379 & @xmath397 & @xmath398 & @xmath396 & @xmath396 & @xmath398 & @xmath397 & @xmath388 + [ @xmath402 & [ @xmath312 & [ @xmath297 & [ @xmath297 & [ @xmath340,@xmath379 & [ @xmath340,@xmath379 & [ @xmath334,@xmath380 & [ @xmath332,@xmath379 & @xmath397 & @xmath398 & @xmath396 & @xmath396 & @xmath398 & @xmath397 & @xmath388 + [ @xmath403 & [ @xmath309 & [ @xmath297 & [ @xmath297 & [ @xmath340,@xmath379 & [ @xmath340,@xmath379 & [ @xmath334,@xmath380 & [ @xmath332,@xmath379 & @xmath397 & @xmath398 & @xmath396 & @xmath396 & @xmath398 & @xmath397 & @xmath388 + [ @xmath404 & [ @xmath308 & [ @xmath297 & [ @xmath299 & [ @xmath340,@xmath379 & [ @xmath334,@xmath380 & [ @xmath340,@xmath379 & [ @xmath340,@xmath379 & @xmath398 & @xmath397 & @xmath398 & @xmath397 & @xmath396 & @xmath396 & @xmath392 + [ @xmath405 & [ @xmath312 & [ @xmath297 & [ @xmath299 & [ @xmath340,@xmath379 & [ @xmath334,@xmath380 & [ @xmath340,@xmath379 & [ @xmath340,@xmath379 & @xmath398 & @xmath397 & @xmath398 & @xmath397 & @xmath396 & @xmath396 & @xmath392 + [ @xmath406 & [ @xmath309 & [ @xmath297 & [ @xmath299 & [ @xmath340,@xmath379 & [ @xmath334,@xmath380 & [ @xmath340,@xmath379 & [ @xmath340,@xmath379 & @xmath398 & @xmath397 & @xmath398 & @xmath397 & @xmath396 & @xmath396 & @xmath392 + [ @xmath407 & [ @xmath308 & [ @xmath299 & [ @xmath291 & [ @xmath334,@xmath380 & [ @xmath332,@xmath379 & [ @xmath332,@xmath379 & [ @xmath334,@xmath380 & @xmath46 & @xmath46 & @xmath45 & @xmath408 & @xmath45 & @xmath408 & @xmath409 + [ @xmath410 & [ @xmath312 & [ @xmath299 & [ @xmath291 & [ @xmath334,@xmath380 & [ @xmath332,@xmath379 & [ @xmath332,@xmath379 & [ @xmath334,@xmath380 & @xmath46 & @xmath46 & @xmath45 & @xmath408 & @xmath45 & @xmath408 & @xmath409 + [ @xmath411 & [ @xmath309 & [ @xmath299 & [ @xmath291 & [ @xmath334,@xmath380 & [ @xmath332,@xmath379 & [ @xmath332,@xmath379 & [ @xmath334,@xmath380 & @xmath46 & @xmath46 & @xmath45 & @xmath408 & @xmath45 & @xmath408 & @xmath409 + [ @xmath412 & [ @xmath308 & [ @xmath299 & [ @xmath297 & [ @xmath334,@xmath380 & [ @xmath340,@xmath379 & [ @xmath334,@xmath380 & [ @xmath332,@xmath379 & @xmath45 & @xmath408 & @xmath46 & @xmath46 & @xmath408 & @xmath45 & @xmath413 + [ @xmath414 & [ @xmath312 & [ @xmath299 & [ @xmath297 & [ @xmath334,@xmath380 & [ @xmath340,@xmath379 & [ @xmath334,@xmath380 & [ @xmath332,@xmath379 & @xmath45 & @xmath408 & @xmath46 & @xmath46 & @xmath408 & @xmath45 & @xmath413 + [ @xmath415 & [ @xmath309 & [ @xmath299 & [ @xmath297 & [ @xmath334,@xmath380 & [ @xmath340,@xmath379 & [ @xmath334,@xmath380 & [ @xmath332,@xmath379 & @xmath45 & @xmath408 & @xmath46 & @xmath46 & @xmath408 & @xmath45 & @xmath413 + [ @xmath416 & [ @xmath308 & [ @xmath299 & [ @xmath299 & [ @xmath334,@xmath380 & [ @xmath334,@xmath380 & [ @xmath340,@xmath379 & [ @xmath340,@xmath379 & @xmath408 & @xmath45 & @xmath408 & @xmath45 & @xmath46 & @xmath46 & @xmath417 + [ @xmath418 & [ @xmath312 & [ @xmath299 & [ @xmath299 & [ @xmath334,@xmath380 & [ @xmath334,@xmath380 & [ @xmath340,@xmath379 & [ @xmath340,@xmath379 & @xmath408 & @xmath45 & @xmath408 & @xmath45 & @xmath46 & @xmath46 & @xmath417 + [ @xmath419 & [ @xmath309 & [ @xmath299 & [ @xmath299 & [ @xmath334,@xmath380 & [ @xmath334,@xmath380 & [ @xmath340,@xmath379 & [ @xmath340,@xmath379 & @xmath408 & @xmath45 & @xmath408 & @xmath45 & @xmath46 & @xmath46 & @xmath417 + & * 3 * & @xmath377 & @xmath377 + [ @xmath420 & [ @xmath291 & [ @xmath308 & [ @xmath308 & [ @xmath334,@xmath421 & [ @xmath334,@xmath421 & [ @xmath332,@xmath422 & [ @xmath332,@xmath422 & @xmath423 & @xmath424 & @xmath423 & @xmath424 & @xmath425 & @xmath425 & @xmath426 + [ @xmath427 & [ @xmath297 & [ @xmath308 & [ @xmath308 & [ @xmath334,@xmath421 & [ @xmath334,@xmath421 & [ @xmath332,@xmath422 & [ @xmath332,@xmath422 & @xmath423 & @xmath424 & @xmath423 & @xmath424 & @xmath425 & @xmath425 & @xmath426 + [ @xmath428 & [ @xmath299 & [ @xmath308 & [ @xmath308 & [ @xmath334,@xmath421 & [ @xmath334,@xmath421 & [ @xmath332,@xmath422 & [ @xmath332,@xmath422 & @xmath423 & @xmath424 & @xmath423 & @xmath424 & @xmath425 & @xmath425 & @xmath426 + [ @xmath429 & [ @xmath291 & [ @xmath308 & [ @xmath312 & [ @xmath334,@xmath421 & [ @xmath332,@xmath422 & [ @xmath334,@xmath421 & [ @xmath340,@xmath422 & @xmath424 & @xmath423 & @xmath425 & @xmath425 & @xmath423 & @xmath424 & @xmath430 + [ @xmath431 & [ @xmath297 & [ @xmath308 & [ @xmath312 & [ @xmath334,@xmath421 & [ @xmath332,@xmath422 & [ @xmath334,@xmath421 & [ @xmath340,@xmath422 & @xmath424 & @xmath423 & @xmath425 & @xmath425 & @xmath423 & @xmath424 & @xmath430 + [ @xmath432 & [ @xmath299 & [ @xmath308 & [ @xmath312 & [ @xmath334,@xmath421 & [ @xmath332,@xmath422 & [ @xmath334,@xmath421 & [ @xmath340,@xmath422 & @xmath424 & @xmath423 & @xmath425 & @xmath425 & @xmath423 & @xmath424 & @xmath430 + [ @xmath433 & [ @xmath291 & [ @xmath308 & [ @xmath309 & [ @xmath334,@xmath421 & [ @xmath340,@xmath422 & [ @xmath340,@xmath422 & [ @xmath334,@xmath421 & @xmath425 & @xmath425 & @xmath424 & @xmath423 & @xmath424 & @xmath423 & @xmath434 + [ @xmath435 & [ @xmath297 & [ @xmath308 & [ @xmath309 & [ @xmath334,@xmath421 & [ @xmath340,@xmath422 & [ @xmath340,@xmath422 & [ @xmath334,@xmath421 & @xmath425 & @xmath425 & @xmath424 & @xmath423 & @xmath424 & @xmath423 & @xmath434 + [ @xmath436 & [ @xmath299 & [ @xmath308 & [ @xmath309 & [ @xmath334,@xmath421 & [ @xmath340,@xmath422 & [ @xmath340,@xmath422 & [ @xmath334,@xmath421 & @xmath425 & @xmath425 & @xmath424 & @xmath423 & @xmath424 & @xmath423 & @xmath434 + [ @xmath437 & [ @xmath291 & [ @xmath312 & [ @xmath308 & [ @xmath332,@xmath422 & [ @xmath334,@xmath421 & [ @xmath332,@xmath422 & [ @xmath332,@xmath422 & @xmath37 & @xmath63 & @xmath37 & @xmath63 & @xmath33 & @xmath33 & @xmath438 + [ @xmath439 & [ @xmath297 & [ @xmath312 & [ @xmath308 & [ @xmath332,@xmath422 & [ @xmath334,@xmath421 & [ @xmath332,@xmath422 & [ @xmath332,@xmath422 & @xmath37 & @xmath63 & @xmath37 & @xmath63 & @xmath33 & @xmath33 & @xmath438 + [ @xmath440 & [ @xmath299 & [ @xmath312 & [ @xmath308 & [ @xmath332,@xmath422 & [ @xmath334,@xmath421 & [ @xmath332,@xmath422 & [ @xmath332,@xmath422 & @xmath37 & @xmath63 & @xmath37 & @xmath63 & @xmath33 & @xmath33 & @xmath438 + [ @xmath441 & [ @xmath291 & [ @xmath312 & [ @xmath312 & [ @xmath332,@xmath422 & [ @xmath332,@xmath422 & [ @xmath334,@xmath421 & [ @xmath340,@xmath422 & @xmath63 & @xmath37 & @xmath33 & @xmath33 & @xmath37 & @xmath63 & @xmath442 + [ @xmath443 & [ @xmath297 & [ @xmath312 & [ @xmath312 & [ @xmath332,@xmath422 & [ @xmath332,@xmath422 & [ @xmath334,@xmath421 & [ @xmath340,@xmath422 & @xmath63 & @xmath37 & @xmath33 & @xmath33 & @xmath37 & @xmath63 & @xmath442 + [ @xmath444 & [ @xmath299 & [ @xmath312 & [ @xmath312 & [ @xmath332,@xmath422 & [ @xmath332,@xmath422 & [ @xmath334,@xmath421 & [ @xmath340,@xmath422 & @xmath63 & @xmath37 & @xmath33 & @xmath33 & @xmath37 & @xmath63 & @xmath442 + [ @xmath445 & [ @xmath291 & [ @xmath312 & [ @xmath309 & [ @xmath332,@xmath422 & [ @xmath340,@xmath422 & [ @xmath340,@xmath422 & [ @xmath334,@xmath421 & @xmath33 & @xmath33 & @xmath63 & @xmath37 & @xmath63 & @xmath37 & @xmath78 + [ @xmath446 & [ @xmath297 & [ @xmath312 & [ @xmath309 & [ @xmath332,@xmath422 & [ @xmath340,@xmath422 & [ @xmath340,@xmath422 & [ @xmath334,@xmath421 & @xmath33 & @xmath33 & @xmath63 & @xmath37 & @xmath63 & @xmath37 & @xmath78 + [ @xmath447 & [ @xmath299 & [ @xmath312 & [ @xmath309 & [ @xmath332,@xmath422 & [ @xmath340,@xmath422 & [ @xmath340,@xmath422 & [ @xmath334,@xmath421 & @xmath33 & @xmath33 & @xmath63 & @xmath37 & @xmath63 & @xmath37 & @xmath78 + [ @xmath448 & [ @xmath291 & [ @xmath309 & [ @xmath308 & [ @xmath340,@xmath422 & [ @xmath334,@xmath421 & [ @xmath332,@xmath422 & [ @xmath332,@xmath422 & @xmath39 & @xmath64 & @xmath39 & @xmath64 & @xmath62 & @xmath62 & @xmath438 + [ @xmath449 & [ @xmath297 & [ @xmath309 & [ @xmath308 & [ @xmath340,@xmath422 & [ @xmath334,@xmath421 & [ @xmath332,@xmath422 & [ @xmath332,@xmath422 & @xmath39 & @xmath64 & @xmath39 & @xmath64 & @xmath62 & @xmath62 & @xmath438 + [ @xmath450 & [ @xmath299 & [ @xmath309 & [ @xmath308 & [ @xmath340,@xmath422 & [ @xmath334,@xmath421 & [ @xmath332,@xmath422 & [ @xmath332,@xmath422 & @xmath39 & @xmath64 & @xmath39 & @xmath64 & @xmath62 & @xmath62 & @xmath438 + [ @xmath451 & [ @xmath291 & [ @xmath309 & [ @xmath312 & [ @xmath340,@xmath422 & [ @xmath332,@xmath422 & [ @xmath334,@xmath421 & [ @xmath340,@xmath422 & @xmath64 & @xmath39 & @xmath62 & @xmath62 & @xmath39 & @xmath64 & @xmath442 + [ @xmath452 & [ @xmath297 & [ @xmath309 & [ @xmath312 & [ @xmath340,@xmath422 & [ @xmath332,@xmath422 & [ @xmath334,@xmath421 & [ @xmath340,@xmath422 & @xmath64 & @xmath39 & @xmath62 & @xmath62 & @xmath39 & @xmath64 & @xmath442 + [ @xmath453 & [ @xmath299 & [ @xmath309 & [ @xmath312 & [ @xmath340,@xmath422 & [ @xmath332,@xmath422 & [ @xmath334,@xmath421 & [ @xmath340,@xmath422 & @xmath64 & @xmath39 & @xmath62 & @xmath62 & @xmath39 & @xmath64 & @xmath442 + [ @xmath454 & [ @xmath291 & [ @xmath309 & [ @xmath309 & [ @xmath340,@xmath422 & [ @xmath340,@xmath422 & [ @xmath340,@xmath422 & [ @xmath334,@xmath421 & @xmath62 & @xmath62 & @xmath64 & @xmath39 & @xmath64 & @xmath39 & @xmath78 + [ @xmath455 & [ @xmath297 & [ @xmath309 & [ @xmath309 & [ @xmath340,@xmath422 & [ @xmath340,@xmath422 & [ @xmath340,@xmath422 & [ @xmath334,@xmath421 & @xmath62 & @xmath62 & @xmath64 & @xmath39 & @xmath64 & @xmath39 & @xmath78 + [ @xmath456 & [ @xmath299 & [ @xmath309 & [ @xmath309 & [ @xmath340,@xmath422 & [ @xmath340,@xmath422 & [ @xmath340,@xmath422 & [ @xmath334,@xmath421 & @xmath62 & @xmath62 & @xmath64 & @xmath39 & @xmath64 & @xmath39 & @xmath78 + | in the framework of the superstring inspired @xmath0 model , ` low energy ' extensions of the standard model compatible with leptogenesis are considered and masses of right handed neutrinos in two scenarios allowed by long lived protons are discussed .
the presence of two additional generations allows breaking of @xmath1 without generating nonzero vacuum expectation values of right handed sneutrinos of the three known generations . after the symmetry breaking , right handed neutrinos acquire majorana masses @xmath2 gev . within the framework of a simple discrete symmetry ,
assumptions made to provide a large mass of right handed neutrinos are shown to be self consistent .
supersymmetric structure of the theory ensures that large corrections , associated with the presence of a ( super)heavy gauge field , cancel out . |
ideal magnetohydrodynamics ( mhd ) , that reliable workhorse of plasma physics , has long been cast into noncanonical hamiltonian form @xcite .
so has the theory from which it is usually derived , the two - fluid model @xcite .
there are many advantages to a hamiltonian form : the discovery and classification of invariants ; the development of numerical algorithms that automatically preserve such invariants ; easily finding the equations of motion in curved coordinates ; conducting equilibrium and stability analysis
. however , there are many theories intermediate in complexity between two - fluid theory and ideal mhd ; kimura and morrison @xcite describe eleven of them .
two are particularly important : hall mhd , which accounts for the difference between the motion of the two species in a typical plasma , and lst s extended mhd @xcite , which includes all terms of first order in the ratio @xmath0 of species masses in the derivation from two - fluid electron - ion theory .
recently , yoshida and hameiri @xcite formulated a noncanonical poisson bracket for hall mhd , and shortly later abdelhamid , kawazura and yoshida did so for extended mhd @xcite ; however , as often happens when working with hamiltonian systems , they had to simply posit a bracket and prove it satisfied all the desired attributes , such as antisymmetry and the jacobi identity .
we will show how to derive these brackets , starting from action principles for each theory .
these action principles have a long and distinguished history in fluids , originating with the work of lagrange in the 18th century @xcite .
the action principle formulation has also been employed in plasma physics since the second half of the 20th century , as evident from the works of @xcite . for ideal magnetohydrodynamics ( mhd ) ,
the first action principle formulation was provided by newcomb in @xcite , shortly followed by other works in the same area @xcite . for extended mhd ,
an eulerian action principle was proposed by @xcite which was subsequently generalized to a eulerian - lagrangian action by @xcite . for recent overviews of action principle formulations of plasma models ,
we refer the reader to @xcite .
the noncanonical hamiltonian formulations for these models can be found in the works of @xcite . in principle
there is an easy process to construct a noncanonical poisson bracket , which goes as follows .
first , construct an action whose variations give the correct equations of motion in some coordinate system . from this tangent - space action principle ,
derive a hamiltonian function via a legendre transform , and produce the corresponding phase - space action principle using the canonical momenta of the original action .
the poisson bracket accompanying the phase - space action will be canonical .
then simply change coordinates in order to produce the desired noncanonical bracket .
this procedure is , indeed , what we use , but there are many complications along the way .
to begin with , the canonical bracket for fluid theories requires lagrangian coordinates : those in which every fluid element is given a distinct label , and the equations of motion are expressed for a given labelled element , despite the fact that the element is changing position .
however , fully lagrangian actions for hall and extended mhd have not yet been given .
the closest are the mixed lagrangian - eulerian actions of ref .
@xcite , `` eulerian '' coordinates being ones that observe fluid quantities at a fixed point rather than following a given element . in this paper
we present fully lagrangian actions .
another complication arises because the legendre transform fails to be invertible for either theory , and an expression for one of the velocities in terms of the phase - space variables must be inserted by hand .
finally , the euler - lagrange map producing the noncanonical brackets requires prior knowledge of the generalized vorticities advected by the theories , so we must devote some time to their discovery and elucidation .
these generalized vorticities turn out to be crucial to the structure of every hamiltonian mhd model .
there are @xmath1 advected vorticities for a theory with @xmath1 distinct charged species , @xmath1 being two in our case .
`` advected '' in this case means that the flux elements defined by the vorticities are carried along with the fluid , their corresponding two - forms obeying a lie - dragging equation . for ideal mhd
, both generalized vorticities collapse down to the same quantity , the magnetic field , which is advected by the fluid velocity . for hall mhd ,
one generalized vorticity is the magnetic field , whose fluxes are carried along with the electron velocity , and the other is the magnetic field plus kinetic vorticity @xmath2 , advected by the ion velocity @xcite . for
extended mhd they turn out to be almost the same , but differing from the hall mhd ones by terms of order @xmath0 in the curl of the current .
both our actions and our derivations of noncanonical brackets would be impossible , but for the fact that we can eliminate the eulerian magnetic field terms in both the action and the euler - lagrange map in favor of fully lagrangian terms , an elimination wholly dependent on the existence of these advected fluxes . before moving on , we note a few ways in which our present work can be readily extended along the lines of past works that utilized these methods .
one can incorporate finite larmor radius ( flr ) effects , such as the braginskii gyroviscosity @xcite .
flr effects for reduced mhd @xcite and generalized fluid models @xcite were implemented via an action principle formulation , and evidently a similar treatment can be undertaken via our extended mhd action principles .
further extensions include stability analyses @xcite , the systematic derivation of reduced extended mhd models ( with potential applications in collisionless reconnection ) , linear and nonlinear waves via a lagrangian approach @xcite , and mhd - like models ( ideal , hall , or extended ) for quasineutral plasmas with more than two charged species .
the paper is organized as follows .
section [ subsec : hamsys ] reviews the basic framework of hamiltonian systems , while section [ subsec : hammhd ] presents a specific example of that framework for ideal mhd , allowing comparisons with the related , but more complex constructions for hall and extended mhd .
we begin our new material by focusing on the simpler theory , hall mhd , in section [ sec : hallmhd ] .
section [ subsec : hallflux ] lays out the needed terminology and facts about hall mhd , which are then used in section [ subsec : hallactions ] to construct both tangent - space and phase - space actions .
section [ subsec : hallliegauge ] is an interesting digression in which we lay out a useful gauge , producing not only advected fluxes but corresponding advected one - forms .
finally , we reach the goal which motivates our entire paper , the derivation in section [ subsec : hallelmap ] of the noncanonical bracket . this derivation is carried out in more algebraic detail than might be necessary , in light of its unfamiliarity to many readers .
we then pivot to extended mhd in section [ sec : exmhd ] , starting with a derivation of its fluxes in section [ subsec : exmhdflux ] .
we give its actions in section [ subsec : exmhdaction ] , and derive its noncanonical bracket in section [ subsec : exmhdelmap ] .
this derivation takes more work than that in [ subsec : hallelmap ] , but the procedure is identical , so this time we omit the details .
we conclude in section [ sec : conclusion ] .
as mentioned , all four models have been put into hamiltonian form . by
that we mean that there is a functional @xmath3 and a bracket @xmath4 so that the time derivative of an arbitrary functional is given by @xmath5 the functionals are integral expressions of the field variables ; for instance , the hamiltonian functional is @xmath6 , where @xmath7 is an energy density . to separate out the time evolution of the field variables ( like a momentum @xmath8 )
, one can use a test functional such as @xmath9 the bracket can be expressed as @xmath10 where the @xmath11 denotes functional differentiation and the components of @xmath12 represent the field variables .
the differential operators @xmath13 must be chosen so that the bracket satisfies its usual properties ( here @xmath14 , @xmath15 , and @xmath16 are functionals , and @xmath17 and @xmath18 are real numbers ) : @xmath19 of these , only the last , the jacobi identity , proves difficult to confirm .
thankfully , the method of this paper provides a relatively easy way to confirm it . in lagrangian coordinates , where every fluid element is given a distinct label @xmath20 and the equations of motion are evaluated at fixed label ( i.e. for a given fluid element ) , the bracket will be canonical : @xmath21 where @xmath22 is the coordinate at fixed label @xmath20 , and @xmath23 is its conjugate momentum , which can be obtained via an action .
this paper has such actions for each of its models .
the jacobi identity is fairly easy to prove for the canonical bracket , relying only on the commutation of functional derivatives , @xmath24 .
however , the map converting lagrangian to eulerian coordinates , in which equations are expressed at fixed spatial coordinates , produces a noncanonical bracket ; for example , even the straightforward definition @xmath25 gives an eulerian momentum dependent on both lagrangian position and momentum .
the jacobi identity can be directly proven for the eulerian bracket , as was done in @xcite for relativistic mhd and @xcite for hall mhd . however , when such a bracket is produced from the canonical lagrangian one , the jacobi identity is assured , as it is invariant under coordinate changes and reductions . in this section
we exhibit the already - discovered hamiltonian forms for ordinary , hall and extended mhd .
it will be our goal in secs .
[ sec : hallmhd ] and [ sec : exmhd ] to derive the latter two .
these theories are all expressed here in terms of eulerian variables , and lagrangian equivalents will be postponed until our later discussion of their actions , where they occur more naturally . in mhd
the eulerian field variables are density @xmath26 , specific entropy @xmath27 , fluid velocity @xmath28 and magnetic field @xmath29 . in the barotropic case
, one can express @xmath27 as a function of @xmath26 and thereby eliminate it , but we consider the more general case .
one can also use as a supplementary variable the current density @xmath30 , using ampre s law in the absence of displacement current .
particle number and entropy are conserved and advected , respectively : @xmath31 the fluid velocity obeys the following momentum equation : @xmath32 while the magnetic field s evolution is determined by ohm s law for a perfect conductor : @xmath33 as can be seen by taking its curl and applying faraday s law : @xmath34 in hamiltonian mhd , while one can express the hamiltonian and bracket in terms of @xmath28 and @xmath27 ( along with @xmath29 and @xmath26 ) , their derivation turns out to be simpler when using the momentum density @xmath35 and the entropy density @xmath36 . in terms of these variables
the hamiltonian is the total energy @xmath37 and the bracket is @xmath38 this bracket also constitutes the bulk of the hall and extended mhd brackets .
it was first given in ref .
@xcite , and the sign convention used is from that paper .
hall mhd differs from ordinary mhd in that the difference between ion and electron velocities is no longer neglected in ohm s law .
the derivation then modifies that law to @xmath39 where @xmath40 is the number density , @xmath41 is the particle mass ( here equal to the ion mass ) and @xmath42 is the electron pressure . the entropy , continuity , and momentum equations are unchanged , as is the hamiltonian .
the bracket , in turn , now has the additional term : @xmath43 this bracket was first described by @xcite , and it requires the assumption of a barotropic electron pressure .
later we will see that this assumption gives us an advected magnetic flux , while barotropic ion pressure gives a second advected quantity .
these will be a necessary part of our construction of the bracket . in extended mhd ,
one additionally retains terms of first order in @xmath44 during the derivation , producing a new momentum equation : @xmath45 and a new version of ohm s law : @xmath46 here the term @xmath47 refers to the symmetric tensorial outer product . the bracket and hamiltonian , however , are more compactly expressed in terms of @xmath48 .
the hamiltonian , which now includes a term for the electron kinetic energy , is @xmath49 in light of the mhd ampre s law @xmath50 .
its bracket , in turn , is @xmath51 this bracket was first given in ref .
it also requires barotropic electron pressure @xmath52 . the structural similarity between the hall and extended mhd poisson brackets
was investigated in ref .
@xcite , and will be elucidated further below .
the essential difference between the various mhd models lies in their flux conservation laws , each one having a different version .
the archetypal flux conservation law is that of ordinary mhd , @xmath53 @xcite . here
the @xmath20 variables are coordinates in a label space @xmath54 , whose continuous values identify fluid elements at @xmath55 ( this condition can be relaxed , as in @xcite ) .
meanwhile , the coordinates @xmath56 describe the point to which a specific element flows ; thus , @xmath57 .
in addition , @xmath58 while @xmath59 .
more explicitly , we write the flux conservation law as @xmath60 this expression can be manipulated into a transformation rule for the magnetic field : @xmath61 where @xmath62 is the jacobian determinant of the invertible transformation from @xmath20 to @xmath63 . , while the magnetic flux is dragged to @xmath64 .
their difference is @xmath65 . ]
there are two distinct ways one can modify the flux conservation law .
first , one can advect a flux different from that of @xmath29 ; with an appropriate choice of this flux , one then gets 2d inertial mhd @xcite .
second , the same flux can be advected , but along a path distinct from that of the fluid .
this second approach gives hall mhd .
specifically , while the fluid itself flows from @xmath20 to a point @xmath56 , the flux element moves from @xmath20 to a distinct point @xmath64 , as illustrated by figure [ labelspacefigure1 ] .
flux conservation is now @xmath66 which gives rise to the transformation rule @xmath67 the flux jacobian @xmath68 is also invertible , and can be written @xmath69 from which one can derive the expression @xmath70 .
taking a full time derivative of @xmath71 in equation gives @xmath72 this equation shows that @xmath29 is advected along @xmath73 as the vector dual to a 2-form , as desired . since @xmath29 is divergenceless , we can add a term proportional to @xmath74 and put the equation in the more familiar faraday form @xmath75 so far , so good .
however , complications arise when you look for the other equations of motion .
some fluid attributes ( density , specific entropy ) are transported along the flow lines @xmath63 , not @xmath73 : mass conservation is described by @xmath76 , and entropy conservation by @xmath77 , recalling that no dissipative terms have been added . as a result
, the label corresponding to the magnetic field will differ from the label on the other quantities .
this situation is shown in figure [ labelspacefigure2 ] . , where a flux has also flowed from a different initial point @xmath78 ; thus @xmath79 . ] in this figure , the fluid element labelled by @xmath20 flows to @xmath56 , while a different label @xmath78 shows the origin of the flux element that has been advected to @xmath80 . for future use
we will need two additional quantities : the point @xmath81 , to which the @xmath78 element flows , and the difference @xmath82 between @xmath83 and @xmath81 .
all these quantities are related via @xmath84 more relations are available , for example @xmath85 . in principle
we could eliminate all but two of the quantities , but it is simpler to keep the extras around .
we also note that , for hall mhd , @xmath63 corresponds to ion flow and @xmath73 to electron flow , so they might also have been written @xmath86 and @xmath87 ; however , in extended mhd , we will use similar quantities , but they will now differ slightly from the electron and ion paths .
thus we use a convention that will be appropriate for both models .
we will also adopt the convention of using superscripts to show vectorial indices , and subscripts to show other attributes , like species identity or initial condition .
there are cases where the distinction between vectors and covectors ( and thus raised and lowered indices ) matters , such as when using curved coordinates , but it may easily be reinstated when needed .
we have , however , left the distinction intact in the expressions @xmath88 and @xmath89 , where it improves readability .
every point corresponds to two labels . in hall mhd , using the convention described above , unprimed labels will denote ion quantities : for example , the number density is advected along the fluid lines , which in our approximation are the ion flow lines
. meanwhile , primed labels will denote electron quantities , such as the magnetic flux density advected along electron flow lines . in light of fig . 2 above
, the variable @xmath63 will appear as both @xmath56 and @xmath81 , as will some quantities ( namely the potentials @xmath90 and @xmath91 ) dependent on them . to simplify expressions
we will write @xmath92 , @xmath93 , @xmath94 and @xmath95 , with unprimed expressions such as @xmath63 denoting unprimed quantities like @xmath56 . if we treat primed and unprimed quantities separately , then the full euler - lagrange equations , using lagrangian density @xmath96
, will be @xmath97_{a ' = q_f^{-1}(q(a , t),t ) } & = 0 \ ; .
\nonumber\end{aligned}\ ] ] with a similar expression for @xmath98 .
many of the terms in the euler - lagrange equations are superfluous : only the first four terms will contribute in the @xmath63 variation , and only the second and fourth terms in the @xmath98 one .
these euler - lagrange equations can be obtained via dirac delta function manipulations on a six - dimensional label space : @xmath99 but for the most part , we will omit this consideration and work with .
however , we emphasize one peculiarity of the action : it only works if the delta function integral is performed after the variations .
if one does so before varying , collapsing back down to a single label space , the variational principle no longer gives the correct expressions .
this peculiarity is shared by the mixed lagrangian - eulerian approach of ref .
@xcite .
if it were written in terms of ion and electron velocities @xmath86 and @xmath87 , the lagrangian density would be standard : @xmath100 in hall mhd , we treat electron velocity as being different from ion velocity ( unlike in regular mhd ) , but nonetheless neglect terms of order @xmath101
. the variables used will be center - of - mass velocity @xmath102 , and the drift velocity of electrons relative to ions , @xmath103 . in terms of ion and electron velocities
we have @xmath104 inverting these equations and neglecting terms of the order of the mass ratio , we have @xmath105 thus , rewriting , setting @xmath106 , and noting the distinction between primed and unprimed labels , the lagrangian density becomes @xmath107 \nonumber \\ & - e n_0 \left[\phi(q , t ) - \phi(q ' + q'_d , t ) \right ] - n_{0 } \left [ u_i \left ( \frac{n_{0}}{\mathcal{j } } , s_{(0)i } \right ) + u_e \left(\frac{n_0}{\mathcal{j}_f } , s_{(0)e } \right ) \right ] \ ; . \label{tangentspaceaction}\end{aligned}\ ] ] in the @xmath63 equation of motion , the terms arising from @xmath108 cancel , plus most of the terms coming from @xmath109 , due to the @xmath110 evaluation .
the only surviving term comes from the advective parts of @xmath111 , which are different for the two terms .
an additional term arises from the @xmath112 partial derivative on @xmath113 . setting @xmath114 , @xmath115 , and @xmath116
, we have , for the @xmath63 equation of motion , @xmath117_{a ' = q_f^{-1}(q(a , t),t ) } = 0\ ] ] which can , by multiplying with @xmath118 and using @xmath119 , be simplified to @xmath120 which is the lagrangian equivalent of . in the @xmath98 equation of motion ,
the three final terms come from the full derivative @xmath121 , and the pressure term comes from the @xmath122 dependence of @xmath68 : @xmath123 with the whole thing evaluated at @xmath124 as usual .
reordering and simplifying , one finds @xmath125 which is ohm s law for hall mhd .
finally , the canonical momenta are @xmath126 with the lagrangian function @xmath127 defined by @xmath128 .
the expression for @xmath129 will allow us to convert the thus - far omitted field term @xmath130 into a term expressed by fluid quantities , once we switch to a phase - space action .
it turns out to be easy to translate the mixed eulerian - lagrangian terms of ref . @xcite into the appropriate terms of , when one minds the difference between primed and unprimed labels .
we translate that paper s @xmath131 and @xmath132 into our variables using @xmath133 and @xmath134
. then its mixed ion terms in the lagrangian are @xmath135 and its electron terms are @xmath136 however , the action in ref .
@xcite also contains a fully eulerian term @xmath137 which is used to produce @xmath138 , a missing piece in our fully lagrangian tangent space action .
we also can not perform the usual legendre transform , because we have no expression @xmath139 .
fortunately , we can solve these problems by switching to a phase space action and invoking .
the four variations of this action give all the needed equations .
the needed action density is @xmath140 \ ; . \label{phasespacedensity}\end{aligned}\ ] ] the middle term , note , is simply @xmath130 . we have expanded it using to express the magnetic field in terms of its initial value , and then applying to express this initial value as the curl of that of a canonical momentum .
there are four phase space variations ; as when using , one sets @xmath110 after taking variations .
thus the @xmath141 variation gives @xmath142
the @xmath129 variation involves an integration by parts on the middle term of the density , giving @xmath143 i.e. @xmath30 , the missing piece of our earlier tangent space action .
note that the @xmath144 term requires varying @xmath145 , which is the value of @xmath129 on the boundary at @xmath146 .
this is permitted since the action principle only requires @xmath147 on the boundary in order to perform an integration by parts , while the momenta are free to vary at @xmath55 .
once again , most of the terms vanish in the @xmath63 variation .
the @xmath148 terms in the middle term of give two factors of @xmath149 , and the @xmath68 in the same term gives a factor of @xmath150 .
the remaining terms proceed similarly as in our tangent space calculation .
the overall result is @xmath151 + \mathcal{j } \partial^i p = 0 \ ; , \ ] ] where @xmath152 is the cofactor matrix to @xmath153 .
given @xmath154 and the @xmath155-@xmath155 identity , is the same as .
finally , the @xmath98 variation gives @xmath156 + n_0
e\partial^i \phi + \mathcal{j}_f \partial^i p_e = 0 \ ; .\ ] ] considering that @xmath157 , and @xmath158 will thus have two terms , this equation is identical to .
however , the action is still slightly unsatisfactory , because we use the quantities @xmath91 and @xmath90 , which are not fully determined by : namely , their gauge freedom remains .
we did use the relation ( [ hallcanonicalmomenta ] ) , viz .
@xmath157 , from the tangent space action to construct the phase space action ; however , does not produce this relation , and neither action gives us the evolution of @xmath90 .
[ subsec : hallliegauge ] develops a gauge condition which resolves this problem in an elegant manner .
look at the hall mhd ohm s law in eulerian coordinates : @xmath159 using @xmath160 and reordering , it becomes @xmath161 and , for a barotropic plasma in which @xmath52 , taking the curl renders it into the form @xmath162 with @xmath163 .
this equation is in the form of , showing that the components of @xmath29 are dual to those of a two - form which is lie - dragged by @xmath164
. it would be even more convenient for @xmath91 to be the components of a lie - dragged one - form , with @xmath29 dual to the components of its exterior derivative .
because the last two terms of are curl - free , they can be expressed as a gradient : @xmath165 .
we then use the gauge freedom in @xmath90 to set @xmath166 which we call the lie gauge , due to the fact that it will produce a lie - dragging equation . with this gauge equation
becomes @xmath167 or @xmath168 so that the vector potential is now a lie - dragged one - form , as desired .
in fact , there exists an entire family of gauges that result in a lie - dragged one - form .
suppose @xmath91 is one such member ( like that already provided ) , and @xmath169 is a lie - dragged zero - form , so that @xmath170 and @xmath171 let @xmath172 .
then , starting from @xmath173 we have @xmath174 collecting the @xmath169 terms inside an overall gradient operator and applying eliminates all of them , showing that @xmath175 is also an advected one - form .
lie - dragging of @xmath91 as a one - form implies that @xmath176 , thus @xmath177 where @xmath178 is the cofactor matrix of the coordinate transformation @xmath179 . because of the relation @xmath180 , the canonical momentum also transforms as a one - form : @xmath181 using the lie gauge , one can eliminate @xmath90 from the phase space action , and using one can also eliminate @xmath182 in favor of its initial value @xmath183 at @xmath146 . however , the other appearance of the vector potential is @xmath184 in is written in terms of `` ion quantities '' ( i.e. unprimed variables ) , whereas expresses it using solely the `` electron quantity '' ( i.e. primed variable ) @xmath185 .
thus , we ve only solved half the problem : we ve expressed @xmath186 in terms of @xmath187 and eliminated @xmath188 with a gauge condition , but we re still left with @xmath189 and @xmath190 .
thankfully , there is a general result showing that , in a system of @xmath1 charged fluid species with barotropic equations of state , there are @xmath1 conserved helicities @xcite and @xmath1 lie - dragged two - forms @xcite . in mhd , this duality is of no concern , because the two collapse and give rise to a single magnetic helicity @xmath191 ; however , in more general models they remain distinct .
we can use the other helicity to eliminate the last two extraneous variables .
hall mhd has the following variable as its second lie - dragged two - form : @xmath192 its advection is straightforward to prove , using eulerian variables . taking the time derivative , and remembering our assumption of barotropic pressures , @xmath193 \\ = & \ ; \nabla \times \left [ \mathbf{v } \times \left(\mathbf{b } + \frac{cm}{e } \nabla \times \mathbf{v } \right ) \right ] = \nabla \times \left(\mathbf{v } \times \mathbf{\mathcal{b } } \right ) = -\pounds_{\mathbf{v } } \mathbf{\mathcal{b } } \ ; .\end{aligned}\ ] ] since @xmath194 is divergenceless , it can be expressed as the curl of a vector @xmath195 .
a fully general expression for such a vector is @xmath196 just as @xmath90 was for @xmath91 , @xmath169 can be chosen to make @xmath195 a lie - dragged one - form , expressed as @xmath197 for we have @xmath198 \ ; . \nonumber\end{aligned}\ ] ] following the reasoning that motivated the lie gauge , we note that the expression in square brackets is the equivalent of @xmath199 from before .
we can thus get a lie - dragged one - form @xmath195 by setting @xmath200 + \frac{1}{ne } \nabla p_e - \frac{1}{ne } \nabla p = - \nabla \left(\frac{\mathbf{v } \cdot \mathbf{\mathcal{a}}}{c}\right ) \ ; .\ ] ] using the lie gauge , this equation simplifies to @xmath201 solved by the action - like quantity @xmath202 just as before , you can add a lie - dragged zero - form to @xmath169 and still have @xmath195 lie - dragged .
however , tantalizing though the action - like expression is , we needed an action expressed entirely in terms of ion quantities and electron quantities , while the above one is mixed due to @xmath203 being an electron quantity .
thus we go back to line , which can be simplified slightly to @xmath204 where @xmath205 is the ion pressure .
except for @xmath90 , which is still an electron quantity , and @xmath169 , which is mixed , these are all ion quantities .
thus we can create the following quantity : @xmath206 which is an ion quantity because all the terms on the right hand side are . with the four quantities
@xmath91 , @xmath207 , @xmath90 and @xmath208 , obeying transformation rules like and subject to the lie gauge , the potential terms in the action can be expressed entirely in terms of their initial conditions , solving the problem mentioned at the end of sec .
[ subsec : hallactions ] .
the new ion variables @xmath207 and @xmath208 introduced in this section deserve a bit more attention . by writing the expression @xmath209 we can say , loosely , that @xmath208 is to @xmath195 what @xmath90 is to @xmath91 .
the parallel is further reinforced by the equivalent of faraday s law , @xmath210 and an easily derived expression @xmath211 reminiscent of ohm s law , but with an extra gradient term . were all the time - dependent terms to be removed from @xmath212 , would be a generalization of bernoulli s law .
we conclude this section with two observations .
first , we can combine , , and @xmath213 to show that @xmath214 which is a nicely compact proof of the conservation of magnetic helicity .
the similar expressions derivable for the ion quantities can similarly be combined to read @xmath215 second , everything that has been said in this section applies to ideal mhd as well , which simply requires one to use @xmath28 instead of @xmath164 and to remember that there is only one distinct helicity .
our phase - space action principle may equivalently be expressed as the set of hamilton s equations @xmath216 , for arbitrary functionals @xmath14 of the phase - space variables .
the bracket in this case is the canonical one : @xmath217 in this section we will show how to convert this bracket into the noncanonical bracket .
the eulerian quantities @xmath26 , @xmath218 , and @xmath219 are defined via standard euler - lagrange maps : @xmath220 the variable @xmath218 is superfluous for barotropic hall mhd , but it is included here for the sake of generality . when we induce variations later on , the quantities @xmath26 and @xmath218 will only have @xmath221 variations ( from the delta functions ) , while @xmath222 will have a @xmath221 and @xmath223 variation .
the odd one is the magnetic euler - lagrange map : @xmath224 this will have @xmath63 and @xmath98 dependence via @xmath73 , and @xmath129 dependence via @xmath225 .
we can now show how the eulerian variables change under variations in the lagrangian phase - space ones , using and : @xmath226 note that the addition of @xmath98 and @xmath129 , which do not appear in regular mhd , nonetheless do not require us to add any new eulerian variables .
they do , however , add a new term in the variation @xmath227 that does not appear in ideal mhd , because now @xmath29 has a @xmath129 dependence via @xmath228 .
the variation induced by an arbitrary function @xmath14 , in both lagrangian and eulerian variables , is @xmath229 substituting the various , except for the one term involving @xmath230 ( which will require more careful attention ) , into the left side of gives the expression @xmath231
d^3 x \ : d^3 a \ ; .\end{aligned}\ ] ] in this expression , the disappearance of @xmath78 is rather startling , but it is still there implicitly via the delta functions , for at a fixed @xmath232 they will pick out values of @xmath20 for the magnetic terms distinct from those of the other terms .
meanwhile , the term that we omitted is , by using @xmath233 , given by @xmath234 here the @xmath235 term in the integration by parts vanishes because it is a symmetric object contracted with an antisymmetric one , and the second factor of @xmath236 appears because we want the delta - function derivative to give a derivative with respect to @xmath63 ( and thus @xmath232 ) .
these factors may be eliminated in the following manner : @xmath237 thus , using , that portion of the @xmath238 variation becomes @xmath239 comparison of the expanded eulerian @xmath238 with the right side of then gives expressions for the lagrangian functional derivatives in terms of the eulerian ones : @xmath240 \delta(x - q ) \\ & + \mathcal{j}_f \left [ b^j \frac{\partial}{\partial x^i } \left(\frac{\delta f}{\delta b^j}\right ) - b^j \frac{\partial}{\partial x^j }
\left(\frac{\delta f}{\delta b^i}\right ) \right ] \delta(x - q - q_d ) \ : d^3 x \\
f}{\delta q_d^i } = & - \int \frac{\delta f}{\delta b^j } b^k_0 \frac{\partial q^j_f}{\partial a^k } \delta'_i \left(x - q - q_d \right ) - \frac{\delta f}{\delta b^i } b^k_0 \frac{\partial q^j_f}{\partial a^k } \delta'_j \left(x - q - q_d \right ) d^3 x \\
= & \int \mathcal{j}_f \left [ b^j \frac{\partial}{\partial x^i } \left(\frac{\delta f}{\delta b^j}\right ) - b^j \frac{\partial}{\partial x^j } \left(\frac{\delta f}{\delta b^i}\right ) \right ] \delta(x - q - q_d ) \ : d^3 x \\
f}{\delta \pi^i_d } = & \int \frac{\delta f}{\delta b^j } \frac{c}{2 n_0 e } \mathcal{j}_f \ : \epsilon^{ijk } \delta'_{k } ( x - q - q_d ) \ : d^3 x \\ = & \ ; \frac{c}{2 n e } \int \left(\nabla \times \frac{\delta f}{\delta \mathbf{b } } \right ) _ i \ : \delta ( x - q - q_d ) \ : d^3 x = - \frac{c}{2 n e } \left .
\left(\nabla \times \frac{\delta f}{\delta \mathbf{b } } \right ) _ i \ ; \right|_{x = q(a , t ) + q_d(a , t)}\end{aligned}\ ] ] finally , we can insert these functional derivatives into the canonical lagrangian bracket .
evaluating the delta function introduces a factor of @xmath241 or @xmath242 , eliminates the @xmath243 integral and converts the remaining lagrangian quantities into eulerian ones : @xmath244 d^3 x \nonumber \\ \equiv & \quad \{f , g\}_{mhd } + \{f , g\}_{hall } \nonumber \ ; .\end{aligned}\ ] ] here the @xmath245 terms are those in the square bracket , and the remaining @xmath246 terms are familiar from ordinary mhd
. the hall portion of the bracket can be greatly simplified .
take the two terms involving the curl of @xmath247 .
they become @xmath248 \\ = \ ; & \frac{c}{2 n e } \left[b^j \left(\nabla \times \frac{\delta f}{\delta \mathbf{b}}\right)^i \ ; \epsilon_{mij}\epsilon^{mkl } \frac{\partial}{\partial x^k}\left(\frac{\delta g}{\delta b^l}\right)\right ] \\ = \ ; & - \frac{c}{2 n e } \mathbf{b } \cdot \left[\left(\nabla \times \frac{\delta f}{\delta \mathbf{b}}\right ) \times \left(\nabla \times \frac{\delta g}{\delta \mathbf{b}}\right)\right ] \ ; .\end{aligned}\ ] ] the other two terms give an identical expressions ; together , they eliminate the factor of @xmath249 and reproduce the hall mhd bracket .
before we move on to produce results for extended mhd , we should pause a moment to discuss our peculiar method of introducing a phase - space constraint .
the simplest phase - space action is a finite - dimensional particle one extremizing @xmath250 with fixed endpoints @xmath251 and @xmath252 .
when doing the @xmath63 variations , an integration by parts must be performed , so @xmath253 on the endpoints of the action integral ; however , when varying @xmath254 , no integration by parts is required , so the momenta can vary on the endpoints . in our lagrangian density ,
the @xmath63 , @xmath141 , and @xmath98 variations occur as normal .
however , @xmath129 has been expressed entirely in terms of its initial value @xmath145 .
thus , when doing the @xmath129 variation , one _ only _ varies at the endpoints ( here the initial surface @xmath146 ) , with the variation at @xmath255 determined , ultimately , by via .
the same substitution for @xmath129 in terms of its initial value occurs in the magnetic euler - lagrange map , making it crucial for the derivation of the hall mhd bracket .
we consider the successful derivation of the bracket to be a sign of this constraint s validity .
however , viewing it as a specific instance of a more general method ( hopefully with applications elsewhere in hamiltonian physics ) , it is clear that we have not established the full conditions under which this method may be applied .
we hope to do so in future work .
in the course of writing an action for extended mhd , we will need to write the field portions in terms of an advected two - form , which will be a vorticity - like quantity .
unfortunately , this time around the magnetic field @xmath29 is not such a quantity
. we will thus begin by showing how to derive a pair of vorticity equations in extended mhd . written in a standard fashion ,
the two central equations for extended mhd are the momentum equation and the generalized ohm s law , rounded off with ampere s law and the continuity equation , plus the isentropy equation if needed
. our goal will be to use these equations to derive a pair of vorticity equations @xmath256 previous experience with ideal and hall mhd suggests that there should be such equations , along with the result that a theory of @xmath1 charged fluids will have @xmath1 such vorticities .
moreover , it was shown in ref .
@xcite that they do exist , by exploiting a map between the poisson brackets for hall and extended mhd
. however , our purpose in this paper is to derive those very brackets , so we should not rely on knowledge drawn from those brackets .
thus we will show how to derive the equations directly . in ref .
@xcite , the form of the generalized ohm s law is derived from the equivalent expression @xmath257 by combining the last two terms and adding a term proportional to @xmath258 ( which is zero ) . instead of doing that , we combine the first term , the last term , and the term proportional to @xmath259 into @xmath260 by using the continuity equation
. we can then replace all occurrences of @xmath261 with @xmath262 , which has units of velocity , producing @xmath263 we next apply the @xmath264 identity and switch to the new field variable @xmath265 to produce @xmath266 \\ & \ ; + \frac{\mathbf{u}\times\mathbf{b}_\star}{c } - \frac{\nabla p_e}{n e } \ ; . \label{ohmslawu}\end{aligned}\ ] ] performing similar operations on the momentum equation produces @xmath267 equations and , being slightly more compact than and , suggest that @xmath268 and @xmath269 are indeed more natural variables than @xmath261 and @xmath29 .
the various gradients in and are crying out for us to take a curl , so we will .
using faraday s law and collecting the time derivative terms in one place , becomes @xmath270 - \frac{c}{ne^2 } \left(\nabla p_e \times \nabla n\right)\ ] ] while becomes @xmath271 + \frac1{mn^2 } \left(\nabla p \times \nabla n\right ) \ ; .\ ] ] going forward we assume barotropic equations of state for both the electrons and ions , because then @xmath272 and the pressure terms drop . in and , shorn of their pressure terms , we have all the ingredients we need to make two equations of the form .
we assume that the quantities appearing in the vorticity equations are linear combinations of those that have appeared in deriving and : @xmath273 the coefficients are all unitless .
expanding , we have @xmath274 \ ; . \label{vorticityexpansion}\end{aligned}\ ] ] on the other hand , we can also use linear combinations of and to express @xmath275 . equating the resulting coefficients with what we find in
, we have the system of equations @xmath276 one equation is redundant , which should be no surprise since @xmath277 is only established up to an overall scale .
we use this extra freedom to set @xmath278 .
then the solutions are @xmath279 , and @xmath280 where @xmath281 is the electron - ion mass ratio .
this confirms what was found in ref .
we can also invert to get @xmath282 because the @xmath277 are divergenceless , they can each be written as the curl of a vector potential @xmath283 .
accounting for the gauge freedom , these potentials are @xmath284 thus , taking appropriate linear combinations of and , we find @xmath285 taking a cue from the hall mhd results , we set @xmath286 and @xmath287 .
we also note @xmath288 , @xmath289 , and @xmath290 . together , all these identities allow a considerable simplification : @xmath291 thus , instead of the differing and , or the greatly different and
, we have two highly symmetric versions of ohm s law expressed using our beloved advected quantities . because we derived only by taking linear combinations and applying lots of vector identities ,
no information has been lost , and the pleasant equations are fully equivalent to the messy and .
finally , in the @xmath292 limit , @xmath293 and @xmath294 , which explains why the @xmath295 term appears in the hall mhd momentum expression but not in its standard ohm s law .
we now turn to the problem of writing a fully lagrangian action principle for extended mhd .
as in ref .
@xcite , this is done by retaining terms up to first order in @xmath281 in the action , and taking @xmath296 after variations .
the coordinate change expressed in now becomes @xmath297 expanding the standard lagrangian and keeping terms up to first order in @xmath0 gives a new lagrangian density @xmath298 \label{extendedtangentaction } \\
& + \frac{e n_0}{c}\left [ \mathbf{\dot{q}}\cdot\mathbf{a } - \mu \mathbf{\dot{q}}_d \cdot \mathbf{a } - \mathbf{\dot{q } } ' \cdot \mathbf{a } ' - ( 1-\mu ) \mathbf{\dot{q}}'_d \cdot \mathbf{a } ' \right ] \nonumber \\ & - e n_0 \left[\phi(q , t ) - \phi(q ' + q'_d , t ) \right ] - n_{0 } \left [ u_i \left ( \frac{n_{0}}{\mathcal{j } } , s_{(0)i } \right ) + u_e \left(\frac{n_0}{\mathcal{j}_f } , s_{(0)e } \right ) \right ] \nonumber \ ; .\end{aligned}\ ] ] here @xmath299 and @xmath300 .
while there are many new terms in this lagrangian , remarkably few make it to the actual equations of motion .
this is partly due to the cancellations that occur ( as in hall mhd ) after setting @xmath301 , and partly due to the ordering @xmath296 imposed after variations .
the @xmath63 variation gives @xmath302 and the @xmath98 variation gives @xmath303 these are the correct equations , given that @xmath304 and @xmath305 will be complicated expressions when eulerianized , particularly the latter .
see sec iv.a of ref .
@xcite for a detailed explanation on how to convert them into eulerian form .
the lagrangian density produces canonical momenta @xmath306 thus @xmath307 the @xmath277 are advected by @xmath308 , whose lagrangian equivalents we will call @xmath309 , where the minus sign comes because @xmath268 and @xmath310 differ by a sign ( see ) . thus we have these fully expanded expressions for @xmath277 , in analogy with : @xmath311 where @xmath312 are the integral lines of @xmath313 , and @xmath314 are the determinants of the matrices @xmath315 .
producing a phase - space action will involve bringing in the omitted term @xmath316 ( see the note about this term in the hall mhd section ) , and converting the terms @xmath317 and @xmath318 to field terms by anticipating the relation @xmath319 . by integrating by parts on the @xmath320 term
, one can combine both of them into the single term @xmath321 .
but first we need to show how to actually perform a variation on such a term . in eulerian variables we have
@xmath322 so that @xmath269 is a function of @xmath29 and @xmath1 . thankfully , the situation is simpler in lagrangian variables .
the variable appears inside a @xmath323 integral , contracted with another vector .
one can integrate by parts to remove a curl , swap in @xmath324 ( any term involving @xmath269 will be an electron quantity ) , and integrate by parts again , so that @xmath325 whenever it appears inside a label space integral .
thus we have @xmath326 in lagrangian variables .
we assume the differential equation can be inverted to produce @xmath327 for some tensorial green s function @xmath328 .
therefore , when varying @xmath329 , the result is @xmath330 where we have used the symmetry @xmath331 of green s functions .
the variation did not affect @xmath332 despite the dependence on @xmath63 because the green s function is translation invariant , i.e. @xmath333 .
we are now in a position to write the full phase - space lagrangian .
it is @xmath334 \delta(a ' - q_f^{-1}(q(a , t),t ) ) \nonumber \\ & + \ ; \frac{j_f}{8\pi ( \delta \beta)^2 } \left [ \frac{\beta_-}{\mathcal{j}_+ } \left(\nabla \times \pi_d - \beta_+ \nabla \times \pi \right)^k \left(\frac{\partial q^i}{\partial a^k } - \alpha_+ \frac{\partial q^i_d}{\partial a^k}\right ) \right .
\nonumber \\ & \ ; \left .
\quad - \frac{\beta_+}{\mathcal{j}_- } \left(\nabla \times \pi_d - \beta_- \nabla \times \pi \right)^k \left(\frac{\partial q^i}{\partial a^k } - \alpha_- \frac{\partial q^i_d}{\partial a^k}\right ) \right ] \label{extendedphasespaceaction } \\ & \ ; \times \left [ \frac{\beta_-}{\mathcal{j}_+ } \left(\nabla \times \pi_d - \beta_+ \nabla \times \pi \right)^l \left(\frac{\partial q^j}{\partial a^l } - \alpha_+ \frac{\partial q^j_d}{\partial a^l}\right ) \right .
\nonumber \\ & \ ; \left .
\quad - \frac{\beta_+}{\mathcal{j}_- } \left(\nabla \times \pi_d - \beta_- \nabla \times \pi \right)^l \left(\frac{\partial q^j}{\partial a^l } - \alpha_- \frac{\partial q^j_d}{\partial a^l}\right ) \right ] g^{ij}(q;q ' ) \ ; \mathrm{d}^3 a \ ; \mathrm{d}^3 a ' \ ; , \nonumber\end{aligned}\ ] ] where we have set @xmath335 and expanded @xmath329 using , the first equation of , and .
thankfully , variations are simplified considerably by the result .
it is also interesting that the usual delta function is replaced by a more specialized green s function in the @xmath329 term .
it is worth pointing out that , while we assumed barotropic equations of state in our development , the above action works for the more general equations of state @xmath336 .
the @xmath141 variation gives @xmath337 the @xmath141 variations occurring in the big field term all cancel , which is not surprising since @xmath269 has no @xmath141 dependence .
the @xmath129 variation gives @xmath338 as it should , with the factor of @xmath68 being absorbed back into @xmath339 when invoking .
after some work , the @xmath63 variation gives @xmath340
+ \mathcal{j } \nabla^i p = 0 \ ; .\ ] ] imposing the @xmath296 condition turns this into @xmath341 + \mathcal{j } \nabla^i p = 0 \ ; .\ ] ] finally , the @xmath98 variation gives @xmath342 \ ; .\end{aligned}\ ] ] to simplify this , note that @xmath343 now @xmath344 , so we end up dropping the @xmath345 term .
next @xmath346 , so we get a plain @xmath347 term , which again gets reduced to @xmath348 . in the end ,
@xmath349 + n_0 e\partial^i \phi + \mathcal{j}_f \partial^i p_e = 0\ ] ] remarkably , equations , , , and are identical to their equivalents , , and from hall mhd .
this is likely the source of the maps discovered in ref .
the differences between hall and extended mhd arise when you switch to eulerian variables .
the field variable @xmath350 will be written as a linear combination of the two - forms @xmath351 , each of which is advected along a linear combination of @xmath63 and @xmath98 .
thus will be rewritten as @xmath352 the coefficients @xmath353 are given by , and those in front of the @xmath354 come from inverting , writing @xmath355 .
the minus signs in front of the various @xmath353 arise because @xmath356 and @xmath268 ( via ) differ by a sign .
equation shows that @xmath357 however , despite the appearance of @xmath358 here , the expression @xmath359 is unchanged , because the terms arising from the two parts of @xmath360 cancel each other .
however , @xmath361 is changed , converting @xmath245 into its extended mhd counterpart .
the following changes appear in the previous calculation : ( i ) all functional derivatives with respect to @xmath29 are now done with respect to @xmath269 ; ( ii ) in the magnetic portion of @xmath246 , @xmath29 is replaced by @xmath269 ; ( iii ) in @xmath245 , @xmath29 is replaced by @xmath362 ( \beta_- \alpha_+ \mathbf{b}_+ - \beta_+ \alpha_- \mathbf{b}_-)$ ] .
this quantity works out to be @xmath363 .
thus we derive the following bracket : @xmath364 with @xmath246 given by the first part of and @xmath365 \ : d^3x,\ ] ] matching .
as expected , the limit @xmath366 reduces @xmath269 to @xmath29 , eliminates the @xmath2 term , and thus reduces @xmath367 to @xmath245 .
we have accomplished many things in the course of deriving the noncanonical brackets for hall and extended mhd .
the need for canonical momenta to serve as the backbone of the brackets led to actions for both theories : first the tangent - space ones and , then the phase - space ones and .
essential to the actions were the advected generalized vorticities : the magnetic field and for hall mhd , and the two expressions for extended mhd . to go with these advected two - forms , we found advected one - form potentials and . for hall
mhd , we found that the momentum equation could be restated in the form of an additional equation resembling ohm s law .
similarly , in extended mhd , after recasting the complex original equations and into variables based around the various advected forms , we could produce the equivalent , but much simpler expressions .
knowing these forms , we were also able to define the natural euler - lagrange maps and , and at last derive the noncanonical brackets and .
a number of interesting concepts had to be used in order to produce these results . to begin with
, the actions required a double label space in order to be fully lagrangian , and we may speculate that lagrangian theories incorporating @xmath1-fluid effects will require @xmath1 label spaces . moreover , padhye and morrison @xcite showed that , in ideal mhd , magnetic helicity is the noether invariant corresponding to the symmetry of relabelling . in each of hall and extended mhd , we have two label spaces , and we speculate that helicities corresponding to each theory s two generalized vorticities will arise from distinct relabelling symmetries on the doubled label space .
this is a matter for future research .
next , our expanded inventory of two - forms and one - forms , defined by their advection properties , allowed us to greatly simplify extended mhd .
this shows that a firm understanding of the geometric nature of the objects appearing in a physical system will allow one to cut away much of its seeming complexity , as briefly noted in @xcite .
finally , our implementation of the euler - lagrange map ( and the phase - space action principle ) required an unusual method of implementing a constraint , a method which may turn out to have broader applicability . prior to attempting work such as ours
, one might have objected that hall and extended mhd are theories too specialized and inelegant to be a fruitful topic for mathematical physics .
fortunately , we have found that it is precisely when investigating such specialized problems that one may discover ideas and methods useful in a broader context . from a practical perspective
, we emphasize that our action principle for extended mhd ( and the concomitant noncanonical hamiltonian formulation ) can be applied to many problems of interest and relevance in diverse areas .
we list a few examples in this category : topological invariants @xcite , particle relabelling symmetries @xcite , reconnection based on hamiltonian models @xcite , tearing modes @xcite , hamiltonian closures @xcite , nonlinear waves @xcite , weakly nonlinear dynamics @xcite , the derivation of gyrofluid and hybrid fluid - kinetic models @xcite , the properties of the equatorial electrojet @xcite , and the rapidly burgeoning field of variational integrators @xcite .
thus , our work serves to advance and flesh out the mathematical foundations of extended mhd , whilst also paving the way for the applications of our methodology in fusion , space and astrophysical plasmas .
pjm and ecd were supported by doe office of fusion energy sciences , under de - fg02 - 04er-54742 .
ml was supported by the nsf grant no .
ags-1338944 and the doe grant no .
de - ac02 - 09ch-11466 .
59ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty link:\doibase 10.1103/physrevlett.45.790 [ * * , ( ) ] link:\doibase 10.1103/physreva.25.2437 [ * * , ( ) ] link:\doibase 10.1063/1.4890955 [ * * , ( ) ] link:\doibase 10.1002/prop.19590070902 [ * * , ( ) ] link:\doibase 10.1088/1751 - 8113/46/33/335502 [ * * , ( ) ] link:\doibase 10.1088/1751 - 8113/48/23/235502 [ * * , ( ) ] @noop _ _ ( , ) link:\doibase 10.1098/rspa.1958.0244 [ * * , ( ) ] link:\doibase 10.1016/0003 - 4916(58)90051 - 4 [ * * , ( ) ] link:\doibase 10.1063/1.1706227 [ * * , ( ) ] link:\doibase 10.1063/1.1761820 [ * * , ( ) ] link:\doibase 10.1017/s002237780002568x [ * * , ( ) ] link:\doibase 10.1063/1.1692854 [ * * , ( ) ] link:\doibase 10.1017/s0022377800006590 [ * * , ( ) ] link:\doibase 10.1017/s0022377800024685 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1063/1.1706844 [ * * , ( ) ] link:\doibase 10.1139/p63 - 216 [ * * , ( ) ] link:\doibase 10.1063/1.1692333 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1063/1.4896336 [ * * , ( ) ] link:\doibase 10.1063/1.1882353 [ * * , ( ) ] in link:\doibase 10.1063/1.3266810 [ _ _ ] , vol .
, ( ) pp . link:\doibase 10.1063/1.4891321 [ * * , ( ) ]
link:\doibase 10.1016/0167 - 2789(83)90017 - 9 [ * * , ( ) ]
link:\doibase 10.1016/j.physleta.2014.12.008 [ * * , ( ) ] link:\doibase 10.1063/1.4926821 [ * * , ( ) ] link:\doibase 10.1016/j.physleta.2016.05.024 [ ( ) , 10.1016/j.physleta.2016.05.024 ] link:\doibase 10.1103/physrevlett.81.4863 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1016/j.physleta.2014.10.013 [ * * , ( ) ] in link:\doibase 10.1063/1.33633 [ _ _ ] , vol . , ( ) pp . link:\doibase 10.1016/0370 - 1573(85)90028 - 6 [ * * , ( ) ] @noop _ _ ( , ) link:\doibase 10.1103/physrevd.91.084050 [ * * , ( ) ] link:\doibase 10.1063/1.4819779 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.88.095001 [ * * , ( ) ] link:\doibase 10.1063/1.4931069 [ * * , ( ) ] link:\doibase 10.1016/0375 - 9601(96)00472 - 0 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1103/physreve.92.063106 [ * * , ( ) ] @noop ( ) link:\doibase 10.1088/0029 - 5515/53/6/063024 [ * * , ( ) ] link:\doibase 10.1063/1.4821840 [ * * , ( ) ] link:\doibase 10.1063/1.4921329 [ * * , ( ) ] link:\doibase 10.1088/1751 - 8113/45/36/365502 [ * * , ( ) ] link:\doibase 10.1016/j.aop.2014.05.011 [ * * , ( ) ] link:\doibase 10.1088/1751 - 8113/48/27/275501 [ * * , ( ) ] link:\doibase 10.1016/j.cnsns.2011.04.010 [ * * , ( ) ] link:\doibase 10.1063/1.4941596 [ * * , ( ) ] link:\doibase 10.1017/s0022377810000760 [ * * , ( ) ] link:\doibase 10.1016/j.aop.2016.02.003 [ * * , ( ) ] link:\doibase 10.1088/1751 - 8113/43/37/375501 [ * * , ( ) ] link:\doibase 10.1088/0741 - 3335/56/9/095008 [ * * , ( ) ] link:\doibase 10.1063/1.4936102 [ * * , ( ) ] @noop ( ) link:\doibase 10.1063/1.4765691 [ * * , ( ) ] link:\doibase 10.1016/j.jcp.2013.01.017 [ * * , ( ) ] link:\doibase 10.1063/1.4897372 [ * * , ( ) ] | there are several plasma models intermediate in complexity between ideal magnetohydrodynamics ( mhd ) and two - fluid theory , with hall and extended mhd being two important examples . in this paper
we investigate several aspects of these theories , with the ultimate goal of deriving the noncanonical poisson brackets used in their hamiltonian formulations .
we present fully lagrangian actions for each , as opposed to the fully eulerian , or mixed eulerian - lagrangian , actions that have appeared previously . as an important step in this process we exhibit each theory s
two advected fluxes ( in analogy to ideal mhd s advected magnetic flux ) , discovering also that with the correct choice of gauge they have corresponding lie - dragged potentials resembling the electromagnetic vector potential , and associated conserved helicities .
finally , using the euler - lagrange map , we show how to derive the noncanonical eulerian brackets from canonical lagrangian ones . |
there are several practical scenarios where it is inappropriate to assume that the distribution of the observations does not change .
for example , financial datasets can exhibit alternate behaviours due to crisis periods . in this case
it is sensible to assume changes in the underlying distribution .
the change in the distribution can be either in the value of one or more of the parameters or , more in general , on the family of the distribution . in the latter case , one may deem appropriate to consider a normal density for the stagnation periods , while a student @xmath0 , with relatively heavy tails , may be more suitable to represents observations in the more turbulent data of a crisis .
the task of identifying if , and when , one or more changes have occurred is not trivial and requires appropriate methods to avoid detection of a large number of changes or , at the opposite extreme , seeing no changes at all .
whilst the literature covering change point analysis from a bayesian perspective is vast when prior distributions are elicited , the documentation referring to analysis under minimal prior information ( i.e. objective bayes ) , is very limited .
in fact , to the best of our knowledge , only two papers tackled the problem from an objective point of view : @xcite and @xcite .
the former discusses the single change point problem in a model selection setting , whilst the latter , which is an extension of the former , tackles the multivariate change point problem in the context of linear regression models .
this work aims to contribute to the methodology for change point analysis under the assumption that the information about the number of change points and their location is minimal .
first , we discuss the definition of an objective prior for change point location , both for single and multiple changes , assuming the number of changes is known a priori .
then , we define a prior on the number of change points via a model selection approach . here
, we assume that the change point coincides with one of the observations .
as such , given @xmath1 data points , the change point location is discrete . to the best of our knowledge ,
the sole objective approach to define prior distributions on discrete spaces is the one introduced by @xcite .
to illustrate their idea , consider a probability distribution @xmath2 , where @xmath3 is a discrete parameter .
then , the prior @xmath4 is obtained by objectively measuring what is lost if the value @xmath5 is removed from the parameter space , and it is the true value . according to @xcite , if a model is misspecified , the posterior distribution asymptomatically accumulates on the model which is the most similar to the true one , where the similarity is measured in terms of the kullback
leibler ( kl ) divergence .
therefore , @xmath6 , where @xmath7 is the parameter characterising the nearest model to @xmath2 , represents the utility of keeping @xmath5 .
the objective prior is then obtained by linking the aforementioned utility via the self - information loss : @xmath8 where the kullback
leibler divergence @xcite from the sampling distribution with density @xmath2 to the one with density @xmath9 is defined as : @xmath10{\mathop{}\!\mathrm{d}}x .
\label{kldef}\ ] ] throughout the paper , the objective prior defined in equation will be referenced as the loss - based prior .
this approach is used to define an objective prior distribution when the number of change points is known a priori . to obtain a prior distribution for the number of change points
, we adopt a model selection approach based on the results in @xcite , where a method to define an objective prior on the space of models is proposed . to illustrate ,
let us consider @xmath11 bayesian models : @xmath12 where @xmath13 is the sampling density characterised by @xmath14 and @xmath15 represents the prior on the model parameter .
assuming the prior on the model parameter , @xmath15 , is proper , the model prior probability @xmath16 is proportional to the expected minimum kullback leibler divergence from @xmath17 , where the expectation is considered with respect to @xmath15 .
that is : @xmath18\right\rbrace \qquad j=1 , \ldots , k. \label{villamodelcompactorg}\end{aligned}\ ] ] the model prior probabilities defined in equation can be employed to derive the model posterior probabilities through : @xmath19^{-1 } , \label{modelposterior}\ ] ] where @xmath20 is the bayes factor between model @xmath21 and model @xmath22 , defined as @xmath23 with @xmath24 .
the paper is structured as follows . in section [ sc_locations ]
we establish the way we set objective priors on both single and multiple change point locations .
section [ sc_numbercp ] shows how we define the model prior probabilities for the number of change point locations .
illustrations of the model selection exercise are provided in sections [ sc_simulation ] and [ sc_realdata ] , where we work with simulated and real data , respectively .
section [ conclusion ] is dedicated to final remarks .
this section is devoted to the derivation of the loss - based prior when the number of change points is known a priori .
specifically , let @xmath11 be the number of change points and @xmath25 their locations .
we introduce the idea in the simple case where we assume that there is only one change point in the dataset ( see section [ sub_sc_onechgpoint ] ) .
then , we extend the results to the more general case where multiple change points are assumed ( see section [ sub_sc_multiplechgpoints ] ) . here , we show that the loss - based prior for the single change point case coincides with the discrete uniform over the set @xmath26 .
let @xmath27 denote an _
n_-dimensional vector of random variables , representing the random sample , and @xmath5 be our single change point location , that is @xmath28 , such that @xmath29{\mbox{\normalfont\tiny\sffamily i.i.d.}}}{\sim}}}f_{1}(\cdot|\tilde\theta_{1 } ) \nonumber \\ x_{m+1 } , \ldots , x_{n}| \tilde\theta_{2 } & { \mathrel{\overset{\makebox[0pt]{\mbox{\normalfont\tiny\sffamily i.i.d.}}}{\sim}}}f_{2}(\cdot|\tilde\theta_{2 } ) .
\label{singlechangepointcase}\end{aligned}\ ] ] note that we assume that there is a change point in the series , as such the space of @xmath5 does not include the case @xmath30 .
in addition , we assume that @xmath31 when @xmath32 . the sampling density for the vector of observations
@xmath33 is : @xmath34 let @xmath35 .
then , the kullback leibler divergence between the model parametrised by @xmath5 and the one parametrised by @xmath7 is : @xmath36 without loss of generality , consider @xmath37 . in this case , note that @xmath38 leading to @xmath39 on the right hand side of equation , we can recognise the kullback leibler divergence from density @xmath40 to density @xmath41 , thus getting : @xmath42 in a similar fashion , when @xmath43 , we have that : @xmath44 in this single change point scenario , we can consider @xmath7 as a perturbation of the change point location @xmath5 , that is @xmath45 where @xmath46 , such that @xmath47 .
then , taking into account equations and , the kullback
leibler divergence becomes : @xmath48 and @xmath49=\nonumber \\&=\min_{m^{\prime}\neq m}\lbrace l \cdot d_{kl}(f_{2}(\cdot|\tilde\theta_{2})\|f_{1}(\cdot|\tilde\theta_{1})),l \cdot d_{kl}(f_{1}(\cdot|\tilde\theta_{1})\|f_{2}(\cdot|\tilde\theta_{2}))\rbrace\nonumber \\&= \min_{m^{\prime}\neq m}\lbrace d_{kl}(f_{2}(\cdot|\tilde\theta_{2})\|f_{1}(\cdot|\tilde\theta_{1})),d_{kl}(f_{1}(\cdot|\tilde\theta_{1})\|f_{2}(\cdot|\tilde\theta_{2}))\rbrace \cdot \underbrace{\min_{m^{\prime}\neq m}\{l\}}_{1}. \label{kl_no_m}\end{aligned}\ ] ] we observe that equation is only a function of @xmath50 and @xmath51 and does not depend on @xmath5 .
thus , @xmath52 and , therefore , @xmath53 this prior was for example used in an econometric context by @xcite with the rationale of giving equal weight to every possible change point location . in this section
, we address the change point problem in its generality by assuming that there are @xmath54 change points .
in particular , for the data @xmath33 , we consider the following sampling distribution @xmath55 where @xmath56 , @xmath57 , is the vector of the change point locations and @xmath58 is the vector of the parameters of the underlying probability distributions .
schematically : @xmath59{\mbox{\normalfont\tiny\sffamily i.i.d.}}}{\sim}}}f_{1}(\cdot|\tilde\theta_{1 } ) \\
x_{m_{1}+1 } & , \ldots , & x_{m_2}|\tilde\theta_{2 } & { \mathrel{\overset{\makebox[0pt]{\mbox{\normalfont\tiny\sffamily i.i.d.}}}{\sim}}}f_{2}(\cdot|\tilde\theta_{2})\\ \vdots&,\ldots , & \vdots & \vdots\ldots\vdots\\ x_{m_{k-1}+1 } & , \ldots , & x_{m_k}|\tilde\theta_{k } & { \mathrel{\overset{\makebox[0pt]{\mbox{\normalfont\tiny\sffamily i.i.d.}}}{\sim}}}f_{k}(\cdot|\tilde\theta_{k})\\ x_{m_{k}+1 } & , \ldots , & x_{n}|\tilde\theta_{k+1 } & { \mathrel{\overset{\makebox[0pt]{\mbox{\normalfont\tiny\sffamily i.i.d.}}}{\sim}}}f_{k+1}(\cdot|\tilde\theta_{k+1}).\\ \end{array}\ ] ] if @xmath60 , then it is reasonable to assume @xmath61 . in a similar fashion to the single change point case
, we can not assume @xmath62 since we require exactly @xmath11 change points .
in this case , due to the multivariate nature of the vector @xmath56 , the derivation of the loss - based prior is not as straightforward as in the one dimensional case .
in fact , the derivation of the prior is based on heuristic considerations supported by the below theorem [ lemmaonechgpoint ] ( which proof is in the appendix ) . in particular , we are able to prove an analogous of equations and when only one component is arbitrarily perturbed .
let us define the following functions : @xmath63 where @xmath64 .
the following theorem is useful to understand the behaviour of the loss - based prior in the general case .
+ let @xmath65 be the sampling distribution defined in equation and consider @xmath66 .
let @xmath67 be such that @xmath68 for @xmath69 , and let the component @xmath70 be such that @xmath71 and @xmath72 .
therefore , @xmath73 where @xmath74 .
[ lemmaonechgpoint ] note that , theorem [ lemmaonechgpoint ] states that the minimum kullback
leibler divergence is achieved when @xmath75 or @xmath76 .
this result is not surprising since the kullback leibler divergence measures the degree of similarity between two distributions .
the smaller the perturbation caused by changes in one of the parameters is , the smaller the kullback leibler divergence between the two distributions is . if now we consider the general case of having @xmath11 change points , it is straightforward to see that the kullback
leibler divergence is minimized when only one of the components of the vector @xmath77 is perturbed by ( plus or minus ) one unit . as such
, the loss - based prior depends on the vector of parameters @xmath78 only , as in the one - dimensional case , yielding the uniform prior for @xmath77 .
therefore , the loss - based prior on the multivariate change point location is @xmath79 where @xmath80 .
the denominator in equation has the above form because , for every number of @xmath11 change points , we are interested in the number of @xmath11-subsets from a set of @xmath81 elements , which is @xmath82 . the same prior was also derived in a different way by @xcite .
here , we approach the change point analysis as a model selection problem . in particular , we define an objective prior on the space of models , where each model represents a certain number of change points ( including the case of no change points ) .
the method adopted to define an objective prior on the space of models is the one introduced in @xcite .
we proceed as follows .
assume we have to select from @xmath83 possible models .
let @xmath84 be the model with no change points , @xmath85 the model with one change point and so on .
generalising , model @xmath86 corresponds to the model with @xmath11 change points .
the idea is that the current model encompasses the change point locations of the previous model . as an example , in model @xmath87 the first two change point locations will be the same as in the case of model @xmath88 . to illustrate the way we envision our models , we have provided figure [ modelconstruct ] . keeping in line with the notation used during the introductory section , for the model @xmath89 , with @xmath90 a non - negative integer , the model parameter @xmath91 is equivalent to the vector @xmath92 .
here , @xmath93 represent the parameters of the underlying sampling distributions considered under model @xmath89 , and @xmath94 are the respective @xmath11 change point locations , as in figure [ modelconstruct ] .
based on the way we have specified our models , which are in direct correspondence with the number of change points and their locations , we state theorem [ remark_model ] ( which proof is in the appendix ) .
let @xmath95 for any @xmath96 integers , with @xmath90 , and the convention @xmath97 , we have the following : @xmath98,\end{aligned}\ ] ] and @xmath99.\end{aligned}\ ] ] [ remark_model ] the result in theorem [ remark_model ] is useful when the model selection exercise is implemented .
indeed , the @xcite approach requires the computation of the kullback
leibler divergences in theorem [ remark_model ] . recalling equation , the objective priors on the model prior probabilities are given by : @xmath100\right\rbrace\qquad j=0 , 1 , \ldots , k. \label{modelpriors}\end{aligned}\ ] ] for illustrative purposes , in the appendix we derive the model prior probabilities to perform model selection among @xmath101 , @xmath102 and @xmath103 .
+ * remark . * in the case where the changes in the underlying sampling distribution are limited to the parameter values , the model prior probabilities defined in follow the uniform distribution .
that is , @xmath104 . in the real data example
illustrated in section [ sub_sc_british_coal_mine ] , we indeed consider a problem where the above case occurs .
let us consider the case where we have to estimate whether there is or not a change point in a set of observations .
this implies that we have to choose between model @xmath101 ( i.e. no change point ) and @xmath102 ( i.e. one change point ) .
following our approach , we have : @xmath105\right\ } , \label{prior_m0_one_chg}\end{aligned}\ ] ] and @xmath106\right\ } \label{prior_m1_one_chg}.\end{aligned}\ ] ] now , let us assume independence between the prior on the change point location and the prior on the parameters of the underlying sampling distributions , that is @xmath107 .
let us further recall that , as per equation , @xmath108 . as such , we observe that the model prior probability on @xmath102 becomes : @xmath109\right\rbrace .
\label{prior_m1_trans_one_chg}\end{aligned}\ ] ] we notice that the model prior probability for model @xmath102 is increasing when the sample size increases .
this behaviour occurs whether there is or not a change point in the data .
we propose to address the above problem by using a non - uniform prior for @xmath110 .
a reasonable alternative , which works quite well in practice , would be to use the following shifted binomial as prior : @xmath111 to argument the choice of , we note that as @xmath112 increases , the probability mass will be more and more concentrated towards the upper end of the support .
therefore , from equations and follows : @xmath113\right\rbrace .
\label{prior_m1_pseudo_binom}\end{aligned}\ ] ] for the more general case where one considers more than two models , the problem highlighted in equation vanishes .
in this section , we present the results of several simulation studies based on the methodologies discussed in sections [ sc_locations ] and [ sc_numbercp ] .
we start with a scenario involving discrete distributions in the context of the one change point problem .
we then show the results obtained when we consider continuous distributions for the case of two change points .
the choice of the underlying sampling distribution is in line with @xcite .
[ [ scenario-1 . ] ] scenario 1 .
+ + + + + + + + + + + the first scenario concerns the choice between models @xmath101 and @xmath102 .
specifically , for @xmath101 we have : @xmath114{\mbox{\normalfont\tiny\sffamily i.i.d.}}}{\sim}}}\mbox{geometric}(p)\ ] ] and for @xmath102 we have : @xmath115{\mbox{\normalfont\tiny\sffamily i.i.d.}}}{\sim}}}\mbox{geometric}(p ) \\
x_{m_{1}+1 } , x_{m_{1}+2 } , \ldots , x_{n}&|\lambda { \mathrel{\overset{\makebox[0pt]{\mbox{\normalfont\tiny\sffamily i.i.d.}}}{\sim}}}\mbox{poisson}(\lambda)\end{aligned}\ ] ] let us denote with @xmath116 and @xmath117 , the probability mass functions of the geometric and the poisson distributions , respectively .
the priors for the parameters of @xmath118 and @xmath119 are @xmath120 and @xmath121 . in the first simulation
, we sample @xmath122 observations from model @xmath101 with @xmath123 . to perform the change point analysis ,
we have chosen the following parameters for the priors on @xmath124 and @xmath125 : @xmath126 , @xmath127 , @xmath128 and @xmath129 . applying the approach introduced in section [ sc_numbercp ]
, we obtain @xmath130 and @xmath131 .
these model priors yield the posterior distribution probabilities ( refer to equation ) @xmath132 and @xmath133 .
as expected , the selection process strongly indicates the true model as @xmath101 .
table [ discrete_results_1 ] reports the above probabilities including other information , such as the appropriate bayes factors .
the second simulation looked at the opposite set up , that is we sample @xmath122 observations from @xmath102 , with @xmath123 and @xmath134 .
we have sampled 50 data points from the geometric distribution and the remaining 50 data points from the poisson distribution . in figure [ fig_gp ]
we have plotted the simulated sample , where it is legitimate to assume a change in the underlying distribution .
using the same prior parameters as above , we obtain @xmath135 and @xmath136 .
again , the model selection process is assigning heavy posterior mass to the true model @xmath102 .
these results are further detailed in table [ discrete_results_1 ] . in scenario 1 . ]
.model prior , bayes factor and model posterior probabilities for the change point analysis in scenario 1 .
we considered samples from , respectively , model @xmath101 and model @xmath102 . [
cols="^,^,^ " , ] from table [ finance_tableresults ] we note that the prior on model @xmath102 and @xmath103 assigned by the proposed method , are the same .
this is not surprising as the only difference between the two models is an additional log - normal distribution with different parameter values .
bayesian inference in change point problems under the assumption of minimal prior information has not been deeply explored in the past , as the limited literature on the matter shows .
we contribute to the area by deriving an objective prior distribution to detect change point locations , when the number of change points is known a priori . as a change point location
can be interpreted as a discrete parameter , we apply recent results in the literature @xcite to make inference . the resulting prior distribution , which is uniform , it is not new in the literature @xcite , and therefore can be considered as a validation of the proposed approach .
a second major contribution is in defining an objective prior on the number of change points , which has been approached by considering the problem as a model selection exercise .
the results of the proposed method on both simulated and real data , allow to show the strength of the approach in estimating the number of change points in a series of observations .
a point to note is in the generality of the scenario considered .
indeed , we consider situations where the change is in the value of the parameter(s ) of the underlying sampling distribution , or in the distribution itself
. of particular interest , is the last real data analysis ( s&p 500 index ) , where we consider a scenario where we have both types of changes , that is the distribution for the first change point and on the parameters of the distribution for the second .
xx berk , r. h. 1966 , ` limiting behavior of posterior distributions when the model is incorrect ' , _ the annals of mathematical statistics _ * 37*(1 ) , 5158 .
carlin , b. p. , gelfand , a. e. smith , a. f. 1992 , ` hierarchical bayesian analysis of changepoint problems ' , _ applied statistics _ * 41*(2 ) , 389405 .
chib , s. 1998 , ` estimation and comparison of multiple change - point models ' , _ journal of econometrics _ * 86*(2 ) , 221 241 .
girn , f. j. , moreno , e. casella , g. 2007 , objective bayesian analysis of multiple changepoints for linear models , _ in _
j. bernardo , m. bayarri , j. berger , a. dawid , d. heckerman , a. smith m. west , eds , ` bayesian statistics 8 ' , oxford university press , pp .
227252 .
kass , r. e. raftery , a. e. 1995 , ` bayes factors ' , _ journal of the american statistical association _ * 90*(430 ) , 773795 .
koop , g. potter , s. m. 2009 , ` prior elicitation in multiple change - point models ' , _ international economic review _ * 50*(3 ) , 751772 .
kullback , s. leibler , r. a. 1951 , ` on information and sufficiency ' , _ the annals of mathematical statistics _ * 22*(1 ) , 7986 .
moreno , e. , casella , g. garcia - ferrer , a. 2005 , ` an objective bayesian analysis of the change point problem ' , _ stochastic environmental research and risk assessment _ * 19*(3 ) , 191204 .
villa , c. walker , s. 2015 _ a _ , ` an objective approach to prior mass functions for discrete parameter spaces ' , _ journal of the american statistical association _ * 110*(511 ) , 10721082 .
villa , c. walker , s. 2015 _ b _ , ` an objective bayesian criterion to determine model prior probabilities ' , _ scandinavian journal of statistics _ * 42*(4 ) , 947966 .
yu , j. 2001 , chapter 6 - testing for a finite variance in stock return distributions , _ in _ j. knight , s. satchell , eds , `
return distributions in finance ' , quantitative finance , butterworth - heinemann , oxford , pp . 143 164 .
here we show how model prior probabilities can be derived for the relatively simple case of selecting among scenarios with no change points ( @xmath101 ) , one change point ( @xmath102 ) or two change points ( @xmath103 ) .
first , by applying the result in theorem [ remark_model ] , we derive the kullback
leibler divergences between any two models .
that is : * the prior probability for model @xmath84 depends on the following quantities : @xmath137 * the prior probability for model @xmath85 depends on the following quantities : @xmath138 * the prior probability for model @xmath88 depends on the following quantities : @xmath139 * for model @xmath84 : @xmath140}_{1}\cdot \left[\inf_{\tilde{\theta}_2}d_{kl}(f_{1}(\cdot|\tilde{\theta}_{1})\|f_{2}(\cdot|\tilde{\theta}_{2}))\right]\\= & \inf_{\tilde{\theta}_2}d_{kl}(f_{1}(\cdot|\tilde{\theta}_{1})\|f_{2}(\cdot|\tilde{\theta}_{2}))\\ \inf_{\theta_2}d_{kl}(m_0\|m_2)=&\underbrace{\left[\inf_{m_{1}\neq m_{2}}(m_{2}-m_{1})\right]}_{1}\cdot \left[\inf_{\tilde{\theta}_{2}}d_{kl}(f_{1}(\cdot|\tilde{\theta}_{1})\|f_{2}(\cdot|\tilde{\theta}_{2}))\right]\\ & + \underbrace{\left[\inf_{m_{2 } \neq n}(n - m_{2})\right]}_{1}\cdot \left[\inf_{\tilde{\theta}_3}d_{kl}(f_{1}(\cdot|\tilde{\theta}_{1})\|f_{3}(\cdot|\tilde{\theta}_{3}))\right]\\ = & \inf_{\tilde{\theta}_{2}}d_{kl}(f_{1}(\cdot|\tilde{\theta}_{1})\|f_{2}(\cdot|\tilde{\theta}_{2}))+\inf_{\tilde{\theta}_3}d_{kl}(f_{1}(\cdot|\tilde{\theta}_{1})\|f_{3}(\cdot|\tilde{\theta}_{3}))\end{aligned}\ ] ] * for model @xmath85 : @xmath141}_{1}\cdot \left[\inf_{\tilde{\theta}_3}d_{kl}(f_{2}(\cdot|\tilde{\theta}_{2})\|f_{3}(\cdot|\tilde{\theta}_{3}))\right]\\=&\inf_{\tilde{\theta}_3}d_{kl}(f_{2}(\cdot|\tilde{\theta}_{2})\|f_{3}(\cdot|\tilde{\theta}_{3}))\\ \inf_{\theta_0=\tilde{\theta}_1}d_{kl}(m_1\|m_0)=&(n - m_{1})\cdot \inf_{\tilde{\theta}_1}d_{kl}(f_{2}(\cdot|\tilde{\theta}_{2})\|f_{1}(\cdot|\tilde{\theta}_{1}))\end{aligned}\ ] ] * for model @xmath88 : @xmath142 * the model prior probability @xmath143 is proportional to the exponential of the minimum between : @xmath144,\mathbb{e}_{\pi_{0}}\left[\inf_{\tilde{\theta}_{2}}d_{kl}(f_{1}(\cdot|\tilde{\theta}_{1})\|f_{2}(\cdot|\tilde{\theta}_{2}))\right.\right . \nonumber \\ & \left.\left .
\hspace*{0.3cm}+\inf_{\tilde{\theta}_3}d_{kl}(f_{1}(\cdot|\tilde{\theta}_{1})\|f_{3}(\cdot|\tilde{\theta}_{3}))\right]\right\rbrace\end{aligned}\ ] ] * the model prior probability @xmath145 is proportional to the exponential of the minimum between : @xmath146 , \right . \nonumber \\ & \left .
\hspace*{0.3 cm } \mathbb{e}_{\pi_1}\left[(n - m_{1 } ) \cdot \inf_{\tilde{\theta}_1}d_{kl}(f_{2}(\cdot|\tilde{\theta}_{2})\|f_{1}(\cdot|\tilde{\theta}_{1}))\right]\right\rbrace\end{aligned}\ ] ] * the model prior probability @xmath147 is proportional to the exponential of the minimum between : @xmath148,\right.\nonumber\\&\left.\hspace*{0.3cm}\mathbb{e}_{\pi_{2}}\left[(m_{2}-m_{1})\cdot \inf_{\tilde{\theta}_1}d_{kl}(f_{2}(\cdot|\tilde{\theta}_{2})\|f_{1}(\cdot|\tilde{\theta}_{1}))+(n - m_{2})\right.\right.\nonumber\\&\left.\left.\hspace*{0.3cm}\cdot\inf_{\tilde{\theta}_1}d_{kl}(f_{3}(\cdot|\tilde{\theta}_{3})\|f_{1}(\cdot|\tilde{\theta}_{1}))\right]\right\rbrace\end{aligned}\ ] ] we distinguish two cases : @xmath149 and @xmath150 . when @xmath149 , equivalent to @xmath151 : @xmath152\,\mathrm{d}\mathbf{x}^{(n ) } \nonumber \\ = & \mathlarger{\sum_{i = m_{j}+1}^{m_{j}^{\prime } } } \int f(\mathbf{x}^{(n)}|\bm{m},\bm{\tilde\theta } ) \cdot \left[\ln\left(\dfrac{f_{j+1}(x_{i}|\tilde\theta_{j+1})}{f_{j}(x_{i}|\tilde\theta_{j})}\right)\right]\,\mathrm{d}\mathbf{x}^{(n ) } \nonumber \\ = & \mathlarger{\sum_{i = m_{j}+1}^{m_{j}^{\prime } } } \left\lbrace 1^{n-1 } \cdot \int f_{j+1}(x_{i}|\tilde\theta_{j+1 } ) \cdot \left[\ln\left(\dfrac{f_{j+1}(x_{i}|\tilde\theta_{j+1})}{f_{j}(x_{i}|\tilde\theta_{j})}\right)\right]\,\mathrm{d}x_{i}\right\rbrace \nonumber \\ = & \sum_{i = m_{j}+1}^{m_{j}^{\prime } } d_{kl}(f_{j+1}(x_{i}|\tilde\theta_{j+1})\|f_{j}(x_{i}|\tilde\theta_{j } ) ) \nonumber \\ = & ( m_j^{\prime}-m_{j } ) \cdot d_{kl}(f_{j+1}(\cdot|\tilde\theta_{j+1})\|f_{j}(\cdot|\tilde\theta_{j } ) ) \nonumber \\ = & ( m_j^{\prime}-m_{j } ) \cdot d_j^{+1}(\bm{\tilde\theta } )
. \label{demopositive}\end{aligned}\ ] ] when @xmath150 , equivalent to @xmath153 , in a similar fashion , we get @xmath154 from equations and , we get the result in theorem [ lemmaonechgpoint ] .
we recall that the model parameter @xmath155 is the vector @xmath156 , where @xmath157 . here ,
@xmath158 represent the parameters of the underlying sampling distributions considered under model @xmath159 and @xmath160 are the respective @xmath161 change point locations . in this
setting , @xmath162 we proceed to the computation of @xmath163 , that is the kullback
leibler divergence introduced in section [ sc_numbercp ] .
similarly to the proof of theorem [ lemmaonechgpoint ] , we obtain the following result .
@xmath164 given equation , if we integrate out the variables not involved in the logarithms , we obtain @xmath165 in a similar fashion , it can be shown that @xmath166 | in this paper we present an objective approach to change point analysis . in particular , we look at the problem from two perspectives . the first focuses on the definition of an objective prior when the number of change points is known a priori .
the second contribution aims to estimate the number of change points by using an objective approach , recently introduced in the literature , based on losses .
the latter considers change point estimation as a model selection exercise .
we show the performance of the proposed approach on simulated data and on real data sets .
* keywords : * change point ; discrete parameter space ; loss - based prior ; model selection ; objective bayes . |
at very low temperatures of a few kelvin rare gas atoms may assemble and form clusters due to the van der waals interaction .
it has been found in a variety of numerical studies that these clusters exhibit a large number of interesting effects . in particular , structural changes of the systems for
an increase or decrease of the temperature provide insight into the behavior and binding mechanisms of the quantum mechanical objects at finite temperatures .
thus , rare gas atomic clusters are a topic of ongoing research . among the rich variety of their thermodynamic properties
are a change from one packing of the atoms to another with increasing energy , or phase transitions from a solid - like behavior to a liquid - like arrangement of the atoms . for the weak van der waals interaction between rare gas atoms the thermal energy corresponding to
a few kelvin already suffices to lead to all impacts mentioned above .
thus , accurate quantum mechanical computations are essential to obtain reliable results . in numerical calculations
the boltzmann operator at inverse temperature @xmath1 , @xmath2 , ( due to its form also called imaginary time propagator ) is the most important quantity .
its trace yields the partition function @xmath3 , and the thermal averages of every observable @xmath4 follow from @xmath5
. however , precise calculations for multi - dimensional systems are still a challenge for today s numerical possibilities .
for example , path - integral monte carlo methods have been used to investigate rare gas clusters . at low temperatures
they become already too expensive for a few dozen atoms , and efficient but sufficiently accurate approximations are required .
these approximations have to reproduce quantum effects correctly since they will be of eminent importance at low temperatures .
still many questions are open .
for example @xmath6 and @xmath7 @xcite might exhibit novel low temperature quantum effects , such as liquid - like zero temperature structures of @xmath7 as compared to a solid - like structure predicted from classical mechanics @xcite .
semiclassical methods help even with presently available computation capabilities to overcome the numerical drawbacks of numerically exact quantum mechanical algorithms and are still developed and applied in a wide context concerning thermodynamic properties .
important semiclassical approximations are based on the idea to restrict the quantum mechanical wave functions to a gaussian shape .
the gaussian functions are defined completely by a low number of parameters as the position of the center , a momentum of the wave packet and a width matrix .
the lowest numerical effort is achieved with a frozen gaussian propagator , i.e. a gaussian wave function of which the width matrix in the exponent is fixed for all imaginary times ( or temperatures ) .
small argon clusters are among those rare gas clusters which attracted large interest for a long time and again recently . despite their apparent simplicity numerical calculations turned out to be nontrivial , and for the argon trimer even sophisticated path - integral monte carlo calculations could not distinguish a complete dissociation from structural changes within a bound system because the numerics suffered strongly from noise @xcite .
the frozen gaussian imaginary time propagator has proved to provide high quality results .
it could solve this question and identified an almost classical dissociation effect @xcite , which was confirmed @xcite with a first - order correction to the semiclassical imaginary time propagator @xcite .
for the more complicated @xmath8 cluster the specific heat and mean energy do not provide enough information to understand its whole structure .
it is important to know which alignment of the atoms is present .
a very detailed knowledge of the atomic structure is available via the inter - particle distances . to obtain their values
it is very common to compute the radial pair correlation function between particles @xmath9 and @xmath10 at positions @xmath11 and @xmath12 , respectively , @xmath13 /z(\beta ) , \label{eq : pair_correlation}\end{gathered}\ ] ] where @xmath14 is the thermal average at inverse temperature @xmath1 and @xmath15 represents the imaginary time propagator at the same temperature @xcite .
this quantity provides information about the distribution of the distances occurring at every temperature .
it is the purpose of this paper to show that the frozen gaussian method applied to atomic clusters in ref .
@xcite can be used to determine the distances of the atoms in clusters in a numerically cheap and easy way . to do so , an extension to the semiclassical frozen gaussian method will be developed .
the mean values and the variances of the distances @xmath16 between atoms @xmath9 and @xmath10 can directly be accessed and provide a very clear information of how the atoms are arranged in the cluster . since the distances @xmath16 of the single pairs can be calculated in parallel with the same monte carlo sampling all of these values can be obtained with low numerical extra cost .
this provides for @xmath17 atoms @xmath18 independent values , whereas in the pair correlation function small differences in the mean distances might be hidden below broad distributions .
the example of the @xmath8 cluster investigated in this article demonstrates the importance of the additional structural information even in this relatively simple system , which does not undergo a structural transformation in the bound state .
the mean energy and the specific heat will show that the cluster dissociates with increasing temperature to a system of six free atoms in one step .
however , precursors of the dissociation will only be observable in the distances between the atoms , which will indicate a loss of the ground state configuration at slightly lower temperatures .
the variances of the distances will turn out to be very sensitive to the breakdown of the structure .
the further sections of this paper are organized as follows . in sec.[sec : method ] mean values and variances for inter - atomic distances with the frozen gaussian method are introduced . for comparison
the same values are introduced within the more flexible thawed gaussian variant .
then the method is applied to the @xmath8 cluster in sec .
[ sec : ar6 ] . after a short introduction of the system ( sec .
[ sec : system ] ) , the confining sphere ( sec .
[ sec : sphere ] ) , and the proper choice of the gaussian width matrix ( sec .
[ sec : width_matrix ] ) derivatives of the partition function ( sec .
[ sec : energy ] ) and the structural information ( sec . [ sec : distances ] ) are investigated .
a discussion in sec.[sec : discussion ] concludes the paper .
the method is based on a semiclassical approximation of the thermal operator @xmath19 where @xmath20 is the inverse temperature .
the approximation consists of evaluating @xmath21 by solving the bloch equation @xmath22 approximately for a frozen gaussian coherent state in position space representation , @xmath23^\mathrm{t } \\ \times \bm{\gamma } [ \bm{x } - \bm{q}(\tau ) ] + \frac{i}{\hbar } \bm{p}^\mathrm{t}(\tau ) \cdot [ \bm{x}-\bm{q}(\tau)]\biggr ) , \label{eq : frozen_state}\end{gathered}\ ] ] where the width matrix @xmath24 is a free parameter and has to be adapted to the given problem as will be explained later for the cluster considered in this article .
the propagation in eq . is done in imaginary time @xmath25 up to the value @xmath26 one is interested in . with gaussian averages of the type @xmath27 the symmetrized form of the frozen gaussian propagator
is given by @xmath28^{-1 } \right ) } \\
\times \exp \left ( -\frac{1}{4 } [ \bm{x } ' - \bm{x}]^\mathrm{t } \bm{\gamma } [ \tanh(\hbar^2 \bm{\gamma } \tau/2)]^{-1 } [ \bm{x}'-\bm{x } ] \right ) \\
\times \int_{-\infty}^\infty \frac{d\bm{q}^{3n}}{(2\pi)^{3n } } \exp \biggl ( -2 \int_0^{\tau/2 } d\tau \langle v(\bm{q}(\tau ) ) \rangle \\ - [ \bm{\bar{x}}-\bm{q}(\tau/2)]^\mathrm{t } \bm{\gamma } [ \bm{\bar{x}}-\bm{q}(\tau/2 ) ] \biggr ) , \label{eq : prop_fg}\end{gathered}\ ] ] where @xmath29 .
the partition function is simply given by the trace @xcite @xmath30 = \sqrt{\det(\bm{\gamma } ) } \exp \left ( -\frac{\hbar^2}{4 } \mathrm{tr}(\bm{\gamma } ) \tau \right ) \\
\times \sqrt{\det \left ( \left [ \bm{1 } - \exp ( -\hbar^2 \bm{\gamma } \tau ) \right ] ^{-1 } \right ) } \\
\times \int_{-\infty}^\infty \frac{d\bm{q}^{3n}}{(2\pi)^{n/2 } } \exp \left ( -2 \int_0^{\tau/2 } d\tau \langle v(\bm{q}(\tau ) ) \rangle \right ) .
\label{eq : pf_fg}\end{gathered}\ ] ] the whole dynamical information is contained in the imaginary time propagation of the variable @xmath31 and is governed by the @xmath32 coupled equations of motion @xmath33 which are relatively simple and can be integrated with a standard integrator for ordinary differential equations .
the remaining numerical task is a single position space integration for the initial positions @xmath34 in eqs . or , which is done with a monte - carlo integration .
as shown previously @xcite a reasonable choice of the width matrix is crucial for the quality of the semiclassical method .
however , highly precise values can be obtained already with a very simple structure .
since all particles are identical and thus the pairwise interactions are the same for all combinations only the center of mass motion has to be distinguished . in center of mass coordinates ,
@xmath35 good semiclassical estimates are obtained with the matrix @xmath36 and the @xmath37 diagonal matrices @xmath38 and @xmath39 describing the three spacial directions of the center of mass and the relative coordinates , respectively . since also the spacial directions are equivalent the best choice are multiples of the unity matrix @xmath40 , such that two scalar coefficients @xmath41 and @xmath42 have to be chosen , i.e. @xmath43 the form of the propagator is represented in cartesian coordinates and is the most efficient for the numerical evaluation . consequently the cartesian representation [ eq : fg_general_matrix ] @xmath44 @xmath45 of the matrix is used .
simple thermal averages requiring only the partition function @xmath3 are the mean energy @xmath46 and the specific heat @xmath47 .
however , @xmath21 provides access to the thermal average of any observable @xmath4 via @xmath48 which is exploited in this article to gain access to the structural information .
a well suited property is the distance between two atoms , i.e.@xmath49 with the frozen gaussian propagator this leads to the expression @xmath50^{-1 } \right ) } \\
\times \int_{-\infty}^\infty \frac{\mathrm{d}\bm{q}^{3n}}{(2\pi)^{3n } } \exp \biggl ( -2 \int_0^{\beta/2 } \mathrm{d}\tau \langle v(\bm{q}(\tau ) ) \rangle \biggr ) \int_{-\infty}^\infty \mathrm{d}\bm{x}^{3n } \exp \biggl ( - [ \bm{x}- \bm{q } ( \beta/2)]^\mathrm{t } \bm{\gamma } [ \bm{x}-\bm{q}(\beta/2 ) ] \biggr ) | \bm{x}_i - \bm{x}_j | , \label{eq : distance_full_term } \end{gathered}\ ] ] in which an explicit integration over the @xmath32 position variables @xmath51 remains in addition to the evaluation of the partition function .
as mentioned previously @xcite it is very important to reduce the numerical effort as much as possible for many - particle systems . in particular , the position space integrations require an expensive monte carlo sampling in a high - dimensional configuration space
. for usual applications a numerical evaluation of the @xmath51 integration can be avoided in a reasonable approximation .
this can be seen with the variable @xmath52 , which transforms the @xmath51 integral in to @xmath53 the widths of the atom s wave functions contribute only at low temperatures significantly to the distance . as will be seen , in practical applications a very narrow gaussian centers all values @xmath54 strongly around zero ,
i.e. @xmath55 is almost identical with @xmath56 for a nonvanishing gaussian weight .
thus , the integral is calculated for the case @xmath57 . with the expansion @xmath58 ^ 2}{|\bm{q}_i - \bm{q}_j|^3 } \label{eq : approximation_dist}\end{gathered}\ ] ] the integral evaluates to @xmath59 , \end{gathered}\ ] ] where @xmath60 is the @xmath37 submatrix of @xmath24 at the rows and columns representing particles @xmath9 and @xmath10 . in total @xmath61^{-1 } \right ) } \\
\times \int_{-\infty}^\infty \frac{\mathrm{d}\bm{q}^{3n}}{(2\pi)^{3n/2 } } \exp \biggl ( -2 \int_0^{\beta/2 } \mathrm{d}\tau \langle v(\bm{q}(\tau ) ) \rangle \biggr ) \bigg [ |\bm{q}_i(\beta/2 ) -
\bm{q}_j(\beta/2)| + \frac{\mathrm{tr } \big ( \bm{\gamma}_{ii}^{-1 } + \bm{\gamma}_{jj}^{-1 } - \bm{\gamma}_{ij}^{-1 } - \bm{\gamma}_{ji}^{-1 } \big ) } { 6 |\bm{q}_i(\beta/2 ) - \bm{q}_j(\beta/2)| } \bigg ] \label{eq : distance_full_approximated } \end{gathered}\ ] ] is obtained . the first term in equation , @xmath62 of the gaussian wave packets .
it corresponds to @xmath63 the second term contains a correction due to the finite width of an atom s wave packet .
it is completely sufficient to include this lowest - order term , of which the @xmath51 integration could be done analytically with a simple result , thus reducing the numerical effort drastically . for the frozen gaussian method
any higher terms beyond those included in the approximation for the mean distances are of lower interest . from the physical point of view
it is expected that the width of the atom s wave function only plays a role at very low temperatures at which the structural configuration is unambiguously in a highly symmetric ground state configuration . indeed , as will be seen in the results already
the correction term in the approximation is very small . to estimate the quality and validity of these mean
values additionally the variances of the distance distributions are calculated . with the operator @xmath64 and the integral @xmath65 ^ 2 \right ) = \frac{1}{z_\mathrm{fg}(\beta ) } \sqrt{\det(\bm{\gamma } ) } \exp \left ( -\frac{\hbar^2}{4 } \mathrm{tr}(\bm{\gamma } ) \beta \right ) \sqrt{\det \left ( \left [ \bm{1 } - \exp ( -\hbar^2 \bm{\gamma } \beta ) \right ] ^{-1 } \right ) } \\ \times \int_{-\infty}^\infty \frac{\mathrm{d}\bm{q}^{3n}}{(2\pi)^{3n/2 } } \exp \biggl ( -2 \int_0^{\beta/2 } \mathrm{d}\tau \langle v(\bm{q}(\tau ) ) \rangle \biggr ) \bigg [ ( \bm{q}_i(\beta/2 ) - \bm{q}_j(\beta/2))^2 + \frac{1}{2 } \mathrm{tr } \big ( \bm{\gamma}_{ii}^{-1 } + \bm{\gamma}_{jj}^{-1 } - \bm{\gamma}_{ij}^{-1 } - \bm{\gamma}_{ji}^{-1 } \big ) \bigg ] \label{eq : distvariances } \end{gathered}\ ] ] the standard deviations @xmath66 of the distances @xmath67 are obtained .
for the @xmath8 cluster there are 15 possible combinations @xmath9 and @xmath10 , and thus 15 distances .
the clusters are oriented arbitrarily in the simulation .
the numbers @xmath9 and @xmath10 of the atoms have no meaning for the true configuration , and thus are not appropriate quantities to define the pairwise distances .
the average of all calculations simply results in identical values for all @xmath68 , which correspond to the mean value of all 15 atom - atom distances in a certain configuration . to obtain a meaningful quantity
the distances are sorted according to their size , @xmath69 and the thermal average of these size - ordered distances is determined , i.e.the thermal mean values of the smallest distance , the second smallest , and so forth are obtained .
these values can be compared with the expectations of geometrical configurations . in an experiment
the single distances are accessible @xcite and can in a given sample be sorted the same way .
alternatively , results from this calculation can be used to determine the distance of the atoms with a well - grounded assumption about the configuration @xcite .
the frozen gaussian method has proved to provide good results for thermodynamic quantities . we want to know whether or not this is also true for the widths calculated in this article .
thus , the structural information of the frozen gaussian method is compared with that of a more flexible thawed gaussian ansatz .
it is based on a time - dependent width matrix @xmath70 , which adapts itself to the given temperature . with the restriction to gaussian wave packets
the thawed gaussian variant is usually the most accurate approximation .
the variable width matrix adds an additional freedom in the parameters .
this is reflected in the quality of the results as has clearly been demonstrated for a double well potential @xcite .
for a large number of degrees of freedom it suffers , however , from the higher numerical costs .
the single - particle ansatz of frantsuzov et al .
@xcite avoids these difficulties by reducing the matrix @xmath70 to a block - diagonal structure , where @xmath37 matrices representing the three spacial coordinates of one particle are the only non - vanishing matrix elements . in the case of six atoms
this reduction is not required and there is no need to ignore the inter - particle correlations .
the thawed gaussian propagator used for comparison with the frozen gaussian method is the time evolved gaussian approximation ( tega ) suggested by frantsuzov et al .
@xcite with a full width matrix @xmath71
. it is based on the solution of the bloch equation with the coherent state @xmath72^\mathrm{t } \bm{g}(\tau)^{-1 } [ \bm{x}-\bm{q}(\tau ) ] \right ) .
\label{eq : tega_gaussian}\end{gathered}\ ] ] the resulting symmetrized propagator reads @xmath73 } { \det[\bm{g}(\tau/2 ) ] } \\ \times \exp \left ( -\frac{1}{2 } [ \bm{x}-\bm{q}(\tau/2 ) ] ^\mathrm{t } \bm{g}(\tau/2)^{-1 } [ \bm{x}-\bm{q}(\tau/2 ) ] \right ) \\ \times \exp \left ( -\frac{1}{2 } [ \bm{x}'-\bm{q}(\tau/2)]^\mathrm{t } \bm{g}(\tau/2)^{-1 } [ \bm{x}'-\bm{q}(\tau/2 ) ] \right ) \label{eq : prop_tg}\end{gathered}\ ] ] with the time - dependent width matrix @xmath70 . in imaginary time
@xmath25 the equations of motion for the gaussian parameters @xmath71 , @xmath74 , and @xmath75 are @xmath76 - \langle v(\bm{q}(\tau ) ) \rangle , \label{eq : tg_eqs_motion_3 } \end{aligned}\ ] ] which have to be integrated from @xmath77 to larger times with the initial conditions @xmath78 in all expressions @xmath79 represents gaussian averaged quantities of the form with the wave packet , and @xmath80 is the @xmath81-dimensional identity matrix .
the relevant quantities are the partition function @xmath82}{\sqrt{\det[\bm{g}(\tau/2 ) ] } } \label{eq : pf_tg}\ ] ] and the mean value of the distances in the same approximation as for the frozen gaussian method , @xmath83}{\sqrt{\det[\bm{g}(\beta/2 ) ] } } \\ \times \bigg [ |\bm{q}_i(\beta/2 ) -
\bm{q}_j(\beta/2)| \\+ \frac{\mathrm{tr } \big ( \bm{g}_{ii}(\beta/2 ) + \bm{g}_{jj}(\beta/2 ) - \bm{g}_{ij}(\beta/2 ) - \bm{g}_{ji}(\beta/2 ) \big ) } { 6 |\bm{q}_i(\beta/2 ) - \bm{q}_j(\beta/2)| } \bigg ] .\end{gathered}\ ] ] the thawed gaussian approximation allows for an additional important information .
its temperature - dependent width matrix @xmath70 provides easier access to the width of the wave function , which influences the variances of the distances . the quantum mechanical part of the variances , i.e. that originating from the spread of the wave function , is expected to increase at lower temperatures . for a frozen gaussian
this can be described correctly if the constant matrix @xmath24 is optimized for every single temperature . in the thawed gaussian case
the variances read @xmath84}{\sqrt{\det[\bm{g}(\beta/2 ) ] } } \\ \times \bigg [ ( \bm{q}_i(\beta/2 )
- \bm{q}_j(\beta/2))^2 + \frac{1}{2 } \mathrm{tr } \big ( \bm{g}_{ii}(\beta/2 ) + \bm{g}_{jj}(\beta/2 ) \\ - \bm{g}_{ij}(\beta/2 ) - \bm{g}_{ji}(\beta/2 ) \big ) \bigg ] - \left ( \bar{d}_{ij}^\mathrm{(tg ) }
\right ) ^2 \label{eq : var_fctg}\end{gathered}\ ] ] and follow directly from the imaginary time evolution of @xmath31 _ and _ @xmath70 .
we are mainly interested in the quantum mechanical part of the variances , viz.@xmath85}{\sqrt{\det[\bm{g}(\beta/2 ) ] } } \\ \times \frac{1}{2 } \mathrm{tr } \big ( \bm{g}_{ii}(\beta/2 ) + \bm{g}_{jj}(\beta/2 ) - \bm{g}_{ij}(\beta/2 ) - \bm{g}_{ji}(\beta/2 ) \big ) .
\label{eq : qm_var_fctg}\end{gathered}\ ] ]
the argon cluster consists of 6 atoms , where the hamiltonian in mass scaled coordinates reads @xmath86 with the laplacian @xmath87 of particle @xmath9 .
the two - body potential @xmath88 of argon is still a very challenging task .
one of the best analytic expressions at hand is a fit to experimental results by aziz and slaman @xcite of which an adoption to a morse potential @xcite is used , @xmath89 - 2 \exp \left [ -\alpha ( r_{ij}-r_\mathrm{e } ) \right ] \right ) \label{eq : morse_potential}\ ] ] with the parameters @xmath90 , @xmath91 , and @xmath92 in consistence with previous studies of the argon trimer @xcite .
the numerical efficiency of the frozen gaussian method is increased with an expansion of the potential in terms of gaussians , viz.@xmath93 this procedure was suggested by frantsuzov et al .
@xcite and has successfully been applied @xcite . in the form
gaussian integrals of the potential or its derivatives can be done analytically .
the required parameters for a fit to three gaussians are listed in table [ tab : gaussian_parameters ] ld .. 8d .. 5 @xmath94 & & + 1 & 3.29610 ^ 5 & 0.6551 + 2 & -1.27910 ^ 3 & 0.1616 + 3 & -9.94610 ^ 3 & 6.0600 + and were previously obtained in ref .
@xcite . an additional potential
is usually introduced to converge the numerical @xmath74 integration .
all particles are confined within a sphere around the center of mass @xmath95 by the condition @xmath96 , where @xmath97 is the confining radius .
this can be achieved with the steep potential @xmath98 added to the hamiltonian or , as in our study , by a restriction of the volume for the @xmath74 integration . of course , an additional potential influences the results and can crucially change the behavior of the cluster @xcite .
if only bound configurations are investigated @xmath99 is usually chosen such that the bound configurations are not affected , i.e. @xmath99 is larger than the extension of the bound cluster .
however , we are interested also in the dissociation process for which the choice of @xmath99 is nontrivial @xcite .
a larger radius @xmath99 always allows for a dissociation at lower temperatures . in principle
it has to be adopted to the physical conditions as , e.g. the pressure .
we are interested in the qualitative behavior at the dissociation and it was checked carefully that the choice @xmath100 does not influence the qualitative change of the relevant observables , i.e. the mean energy , the specific heat , the mean values of the inter - atomic distances and their variances . in particular , it was assured that the case of a completely dissociated cluster is present for temperatures above @xmath101 and the form of the dissociation process is not altered .
the value of the confining radius @xmath99 , i.e. the pressure in physical terms , affects the temperature at which the dissociation occurs . while in a thawed gaussian calculation
the initial condition for the width matrix is defined , the constant matrix @xmath24 of its frozen gaussian counterpart has to be chosen carefully .
it is a free parameter of the system .
it is not trivial to find a good choice of @xmath24 .
however , as was mentioned above , the structure with the @xmath37 submatrices is well suited .
thus , only the two parameters @xmath41 and @xmath42 need to be chosen . a detailed investigation of the @xmath102 cluster revealed that there is a reliable and simple method to find the best choice for the inter - particle width parameter @xmath42 @xcite . propagating the partition function to long imaginary times @xmath103 , i.e. @xmath104 , one can extract the thermodynamic mean energy to correspond to the ground state energy @xmath105 .
the parameter @xmath42 providing the lowest value for @xmath105 has shown to lead to the best agreement with numerically exact calculations and the more flexible thawed gaussian approximation .
this result is almost independent of the temperature at which the partition function , mean energy or specific heat of all methods are compared .
furthermore , calculating the first - order correction to the frozen gaussian propagator showed that this choice also requires the smallest correction .
thus , the simple minimization of the ground state energy gives us a reliable way of determining @xmath42 . for the @xmath8 cluster in this article
it was found that @xmath106 is the best choice .
the center of mass is free and it can exactly be described by a gaussian in the limit @xmath107 .
this means the value should be as small as possible . for the numerical evaluation one needs a finite value
it is known that @xmath108 is small enough @xcite .
the results can not be distinguished from those of even lower values for @xmath41 . to compare the @xmath8 cluster with the trimer considered in @xcite the mean
energy and the specific heat are studied first .
they are shown in fig .
[ fig : energy ] ( color online ) ( a ) mean energies of the @xmath8 cluster calculated with the two - parameter frozen gaussian propagator ( 2p - fg ) and its classical counterpart .
for the high - temperature limit both results agree well . at low temperatures
the classical mean energy tends to the potential minimum and the frozen gaussian propagator approximates the quantum mechanical ground state .
the inset shows a comparison of the 2p - fg method with the fully coupled thawed gaussian method ( fc - tg ) .
the energies differ by a few percent .
( b ) specific heat around the dissociation , which seems to happen in one step . ] for the two - parameter ( @xmath41 and @xmath42 ) frozen gaussian propagator and the derivatives of the classical partition function @xmath109 the observations are very similar to those obtained for trimer . at very low temperatures
the classical calculation tends to the potential minimum . at @xmath110
a mean energy of @xmath111 is found .
the frozen gaussian results are at this temperature already in a very flat regime , in which the mean energy is almost independent of the temperature and approximates the quantum mechanical ground state energy .
the method leads to @xmath112 . in the inset of fig .
[ fig : energy ] ( a ) a comparison with the more flexible fully - coupled thawed gaussian propagator is shown .
it leads to a value of @xmath113 , i.e. the difference of the ground state s binding energy is only @xmath114 .
thus , one may conclude that also for the larger @xmath8 cluster the quality of the frozen gaussian propagator is acceptable in comparison with the numerically more expensive thawed gaussian variant even for the low - temperature limit .
the dissociation appears in the mean energy as a step .
the energy raises almost directly to that of six free particles .
this indicates a dissociation of all atoms at once as was observed for the trimer .
the same information can be gained from the specific heat , which is shown around the dissociation in fig .
[ fig : energy ] ( b ) .
one broad peak confirms that the dissociation occurs in one step .
the classical calculation shows a transition at a slightly lower temperature , and the difference between the two maxima in the specific heat is approximately @xmath115 , which is lower than for the trimer , where a difference of @xmath116 was observed .
certainly the difference can again be related to the zero point energy in quantum mechanics .
this is larger for six atoms than for three and one could expect that also the temperature difference is larger .
however , one has to keep in mind that this energy has to be distributed among a larger number of atoms during the dissociation .
since aside from the small shift in the temperature the dissociation process is almost identical in the classical and the frozen gaussian calculation one may conclude that it is a purely classical phenomenon .
in addition to the information of the simple derivatives of the partition function the mean distances are studied . since so far the dissociation seems to be purely classical it is interesting to also compare the structural information with the classical one . to do so , the classical mean distances @xmath117 are added in the calculations below .
the mean distances obtained for temperatures below @xmath118 , i.e.significantly below the dissociation process , are shown in fig.[fig : dist_low ] , ( color online ) ( a ) comparison of the mean values of all 15 distances calculated with the frozen gaussian method ( 2p - fg ) and a classical calculation at temperatures @xmath119 .
the distances appear in groups .
three distances converge for @xmath120 to values above @xmath121 , whereas the remaining 12 are below @xmath122 .
( b ) a comparison of the frozen gaussian and thawed gaussian method
( fc - tg ) shows that the mean distances agree very well . ]
where first the frozen gaussian method ( 2p - fg ) is compared with the classical calculation in fig .
[ fig : dist_low ] ( a ) , and then the fully - coupled thawed gaussian approximation is added in fig . [
fig : dist_low ] ( b ) .
the most striking observation is that the distances appear in the low - temperature limit in two groups .
a group of three `` long '' distances , of which the values are always above @xmath121 for @xmath120 , and a second group of the 12 remaining `` short '' distances , which converge to a value below @xmath122 , exist .
this already gives a clear answer to the question about the ground state configuration of @xmath8 .
it is consistent with the distances in an octahedron , or in other words , the atoms are located at the centers of the surfaces of its dual polyhedron , viz.the cube .
twelve short distances @xmath123 from the atoms on neighboring surfaces and three longer distances @xmath124 between the atoms on opposite surfaces are expected .
the ratio of the distances is supposed to be @xmath125 , which is fulfilled excellently in both the classical and semiclassical calculations . in the classical case
the atoms seek directly the potential minima , whereas in the quantum case always a wave function with a finite width leading automatically to larger mean distances is present . furthermore ,
in the classical case the fixed octahedron configuration is only observable for @xmath120 .
this is a consequence of the fact that classically every nonvanishing energy allows for a thermal excitation .
in contrast to this there should be no excitation possible if @xmath126 is clearly below the energy difference between the ground state and the first excited state in the quantum mechanical case .
this is also reflected in the mean distances .
for temperatures below @xmath127 no differences between the distances in one group are observed , and the distances do not change for even lower temperatures .
this indicates that the cluster is already in the ground state configuration .
a comparison of the frozen gaussian method with the fully - coupled thawed gaussian propagator reveals that the distances agree very well .
this is in particular true for all larger distances .
also the low temperature limit shows an excellent agreement .
the edge length of the cube containing the octahedron is @xmath128 in the frozen gaussian calculation and @xmath129 in the thawed gaussian approximation .
the difference is below @xmath130 , and thus even smaller than that of the mean energy .
obviously the structural information of the frozen gaussian method is less affected by the constant gaussian width approximation . with the data of fig .
[ fig : dist_low ] we are also able to estimate the quality of the approximation , in which the power series expansion of the distances was introduced .
the first - order term retained in the expansion is of the size @xmath131 where for the width matrix @xmath132 is obtained . with the knowledge that the typical distances @xmath133 are even in the bound phase of the order of a few ngstrms this correction can be estimated to be always less than @xmath130 of the leading order .
hence , it has at most the same size as the difference between the two semiclassical propagators .
higher orders in the series expansion would lead to even smaller corrections , which do not need to be taken into account since they are below the error of the semiclassical approximation .
the distances around the dissociation are shown in fig .
[ fig : dist_high ] , ( color online ) ( a ) mean values of all 15 distances calculated with the frozen gaussian method ( 2p - fg ) around the dissociation
. two of the smaller distances join the group of the three larger distances starting at @xmath134 , but then the dissociation occurs at once .
( b ) the same behavior is observed for the classical calculation . ]
where the classical and 2p - fg results are compared . since the frozen gaussian approximation is known to provide good results at these temperatures ( cf.ref .
@xcite ) , a comparison with the thawed gaussian propagator does not give any new information .
figure [ fig : dist_high ] confirms the finding of the consideration of the mean energy and the specific heat in fig .
[ fig : energy ] .
the dissociation effect is classical .
the semiclassical approximation of the quantum mechanical propagator and the purely classical calculation lead to the same behavior . apart from a small shift in temperature
both diagrams agree very well .
the calculation of the distances gives additional insight into the dissociation process . for temperatures @xmath135 the groups of short and long distances are unchanged .
above this temperature two of the twelve distances @xmath123 are separating from the others and join the three longer distances @xmath124 .
two new groups with five and ten distances start to from . for @xmath0 a second isomer in the form of a tri - tetrahedron
is known to contribute at increasing temperatures @xcite .
this would exactly agree with a grouping of five longer and ten shorter distances and is also visible in fig .
[ fig : cmdist ] , in which the distances of all six atoms from the center of mass are shown for a classical calculation .
( color online ) distances of all six atoms from the center of mass in a classical calculation for @xmath136 .
shown is a thermal average of size - ordered distances as explained in sec .
[ sec : sizeorder ] . with increasing temperature one distance becomes larger than the others before all raise drastically . ] at low temperatures all distances have almost the same size , which agrees with a pure octahedron configuration . with increasing temperature one distance grows and indicates a coexistence of octahedron and tri - tetrahedron configurations .
signatures of further arrangements of the atoms are not found .
however , this rearrangement of the atoms remains in its beginnings .
a new structure can not completely arise since the whole process does not finish before all distances raise drastically and indicate with this increase the dissociation of the cluster .
comparisons with calculations , in which due to a smaller value of @xmath97 no dissociation is allowed , show that this effect only appears in connection with the dissociation .
thus , the separation of the two distances is more a precursor of the total destruction of the cluster .
the dissociation happens then at once .
after the dissociation the distances obtain new almost constant values which correspond to the distribution of atoms moving freely within the confining sphere .
to learn more about the actual distribution of the distances their standard deviations are shown in figs .
[ fig : variances ] ( a ) and ( b ) ( color online ) ( a ) standard deviations @xmath137 of the distances in the frozen gaussian approximation ( 2p - fg ) and ( b ) of the classical calculation . they increase drastically around the dissociation .
( c ) quantum mechanical part of the variances for the fully coupled thawed gaussian propagator . for low temperatures it increases due to an increasing width of the gaussians . ] for the classical and the 2p - fg method .
the most significant feature is the drastic increase of the standard deviations around the transition . at this
temperature range parts of the simulated clusters are still in a bound configuration whereas others are already dissociated .
the effect is similar and of the same size for the classical and the semiclassical calculation . for temperatures above the dissociation
the standard deviations are almost the same for all distances , which is also expected for six free atoms . of more interest
is the behavior of the standard deviations below the dissociation .
the longer distances are expected to show more fluctuations .
additionally two of the shorter distances join the group of the three longer distances for increasing temperatures as a consequence of contributions from two isomers .
it can also be expected that these two show higher standard deviations than the short distances since the separation of the two distances does not happen abruptly at one temperature as can be seen in fig.[fig : dist_high ] .
consequently , below the dissociation the standard deviations are found to form two groups .
one group combines the standard deviations of five distances , i.e. the three longer ones and the two joining them .
the other group consists of the standard deviations of the 10 short distances which stay together up to the temperature of the dissociation . in figs .
[ fig : variances ] ( a ) and ( b ) it seems that the standard deviations always decrease for lower temperatures .
this is definitely expected for the classical calculation .
the frozen gaussian approximation can , since the width of the wave function is determined by the constant values of @xmath24 , not reflect the quantum mechanical expectation that the wave function stretches at lower temperatures . to cover also this effect the quantum mechanical part according to equation
is plotted in fig .
[ fig : variances ] ( c ) .
only the variance of one of the distances is shown since this part is almost identical for all of them in the highly symmetric situation of a monoatomic cluster . at @xmath138
the variance has approximately the same size as that following from the frozen gaussian method @xmath139 .
in particular , these quantum mechanical parts of the variances are considerably lower than other contributions in equations and .
figure [ fig : variances ] ( a ) would not change with the thawed gaussian propagator . only for temperatures @xmath140
the extension of the wave function becomes important for a measurement of the distances . for temperatures in the range of the dissociation
the information from the frozen gaussian method is completely sufficient .
all calculations shown in this section could be implemented and performed easily on a nvidia tesla c2070 gpu . on this architecture
a converged result for low temperatures is obtained in less than 20 hours .
the most critical part is that around the dissociation , where a very detailed sampling for bound configurations has to be done alongside an inclusion of large distances allowing for an unbound cluster .
this can require an increase of the sampling points by a factor of 10 .
in this article it was shown that structural information about a cluster of atoms can be obtained with the frozen gaussian semiclassical method in a numerically cheap way .
the evaluation of the corresponding integrals can be done in parallel to that of the partition function . with this method
the full information about all the distances of all combinations of the atoms can be obtained . a comparison with
the more flexible thawed gaussian propagator revealed that the quality of the distances is on the same level as that of the mean energy or the specific heat , or even better . to avoid inefficient numerical computations of a position space integral an approximation for the distances
was introduced .
it was possible , however , to show that this approximation does not reduce the quality of the results below that obtained in the semiclassical approximation of the propagator . on the physical side it was found that with increasing temperatures the @xmath8 cluster undergoes an almost direct transition to six free atoms
however , it shows precursors in the distances . at temperatures slightly below the dissociation
a reordering of the atoms starts , in which contributions from a second isomer , viz .
a tri - tetrahedron @xcite , appear , but then vanish in the increasing distances at the dissociation .
the dissociation is a purely classical effect .
the semiclassical approximation shows exactly the same behavior with just a small shift in the temperature of @xmath115 . around the dissociation
the standard deviations of the distances are almost completely determined by classical contributions . only for lower temperatures
the extension of the wave functions becomes important as was seen in a thawed gaussian approximation . at low temperatures
the cluster assumes the shape of an octahedron , where the longer distance between the atoms is @xmath141 , and the shorter has the value @xmath142 .
classically the fixed configuration is only obtained in the limit @xmath104 whereas in the quantum mechanical case the ground state configuration is present for all temperatures @xmath143 .
the frozen gaussian method has proved to provide reliable results for quantum mechanical calculations .
there is a large number of investigations which can be done with it . in particular , the results for @xmath102 in @xcite and for @xmath8 in this work indicate that the confinement to very small spheres usually applied in the calculation of the partition function and values deduced from it @xcite only sample bound cluster configurations . this is physically realized at high pressures .
if one is interested in lower pressures , at which a dissociation is allowed , this is too restrictive to fully understand the low - temperature behavior of the clusters .
the dissociation can set in before structural changes or a melting can be observed . to take this into account
it is necessary to advance the investigations done here to clusters with higher numbers of atoms . in particular , the cases of @xmath144 , @xmath6 @xcite or @xmath7 @xcite are of special interest since they showed interesting structural transformations in the non - dissociated cases . whereas the stronger quantum effects in the completely bound case are well covered by a variable width matrix the numerically cheaper frozen gaussian method has advantages in the numerically more challenging case of the dissociation requiring a sampling of bound and unbound configurations of the atoms .
the exactness of both methods can then be monitored and ( if necessary ) improved with the the series expansion of the imaginary time propagator @xcite .
the importance of the series expansion is not restricted to the dissociation .
most effects in rare gas clusters such as structural transformations or dissociations appear at such low temperatures that it is necessary to analyze whether the semiclassical approximations used in the calculations correctly reproduce the true quantum mechanical behavior .
an important example will be @xmath7 , for which strong differences are found between the approximate quantum computations and a purely classical theory @xcite .
44ifxundefined [ 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 * * , ( ) link:\doibase 10.1140/epjb / e2006 - 00353 - 0 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( )
@noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) \doibase http://dx.doi.org/10.1063/1.3599052 [ * * , ( ) ] http://stacks.iop.org/1742-6596/438/i=1/a=012012 [ * * , ( ) ] \doibase http://dx.doi.org/10.1063/1.4730033 [ * * , ( ) ] \doibase http://dx.doi.org/10.1063/1.3239476 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/j.cplett.2012.09.013 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.98.023401 [ * * , ( ) ] link:\doibase 10.1021/jp903282b [ * * , ( ) ] \doibase http://dx.doi.org/10.1063/1.464070 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1103/physrevlett.77.1238 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) | a numerically cheap way to obtain structural information about clusters of rare gas atoms at low temperatures is developed .
the semiclassical frozen gaussian imaginary time propagator is extended such that it can account for the mean values of all inter - atomic distances in the cluster and their variances . to reduce the required numerical effort an approximation for the mean values is developed which preserves the quality of the results offered by the semiclassical ansatz .
the method is applied to the @xmath0 cluster .
it is found that the cluster dissociates almost in one step to six free atoms when the temperature is increased .
precursors of the dissociation are only observable in the distances of the atoms via the appearance of a second isomer .
the process is almost classical .
however , the method is able to resolve small differences in the temperatures at which the dissociation takes place and in the mean distances of the bound configuration . |
with the discovery of a new resonance at the large hadron collider ( lhc ) , having a mass @xmath0 gev @xcite and couplings well consistent with the standard model ( sm ) predictions @xcite , the higgs naturalness problem @xcite has become the foremost problem to be tackled .
the resolution , if any , brings its own new physics structure .
the squared - masses of fundamental scalars , contrary to chiral fermions and gauge bosons whose masses are protected by chiral and gauge invariances , receive additive quantum corrections proportional to @xmath1 the uv boundary of the sm . in explicit terms ,
one - loop quantum correction to higgs squared - mass , originally computed by veltman @xcite , reads as [ vc ] ( m_h^2)_=(6_h+g^2+g^2 - 6g_t^2 ) , where @xmath2 and @xmath3 are the @xmath4 and @xmath5 gauge couplings of the sm , respectively , and @xmath6 ( @xmath7 gev is the vev of the higgs field ) is the top quark yukawa coupling .
the top quark , being the most strongly coupled sm particle to the higgs field , induces the biggest contribution and ensures a nonvanishing , unremovable coefficient before @xmath1 .
the higgs boson mass is stabilized to electroweak scale if @xmath8 .
this is the veltman condition ( vc ) .
the parameters in it have all been measured , and it violates the lhc results for @xmath9 gev @xcite .
having no symmetry to prevent the higgs boson mass from sliding to the higher scales via ( [ vc ] ) , frequently a cancellation mechanism is implemented via fine - tuning of counter terms in which low and high energy degrees of freedom are mixed .
this renders the whole procedure unnatural .
it would be more natural , if the cancellation occurs by means of a symmetry principle at higher scales , or if it arises by accidental cancellations of certain terms .
in fact , models of new physics constructed to complete the sm beyond fermi energies have all been motivated by higgs naturalness problem @xcite ( see also @xcite for studies within supersymmetry ) .
so far , however , in the 7 tev and 8 tev lhc searches reaching out beyond the tev domain , no compelling sign of evidence for new physics has been found @xcite . in consequence ,
having no tev scale new physics for achieving naturalness , one is forced to understand the electroweak unnaturalness within the sm plus general relativity , albeit with some imperative extensions required by specifics of the approach taken . in 1995 ,
conformal symmetry @xcite was proposed as a mechanism for solving the higgs mass hierarchy problem ( the latest studies on the conformal symmetry as a solution to the fine - tuning problem may be found in @xcite ) .
recently , the higgs coupling to spacetime curvature has been found to stabilize the electroweak scale by a harmless , soft fine - tuning @xcite .
furthermore , anti - gravity effects have been claimed to improve higgs naturalness @xcite .
alternatively , one may view the parameters chosen by nature as the necessity of existence , and this leads to anthropic considerations @xcite . in variance with all these approaches , a fine - tuning method based on singlet scalars
@xcite has also been employed . in this approach ,
main idea is to cancel the quadratic divergences in higgs boson mass with the loops of the singlet scalars that couple to higgs field @xcite .
this method , though a fine - tuning operation by itself , nullifies the quadratic divergences and accommodates viable dark matter candidates @xcite . nevertheless , for real singlet scalars with vacuum expectation value ( vev ) , it is not possible to kill the quadratic divergences consistently because there is a mixing between the cp - even component of the higgs field and the real singlet scalar , and it does not allow for simultaneous cancellation of the quadratic divergences in higgs boson and singlet scalar masses @xcite .
there are also studies on two - higgs doublet models without flavor changing neutral currents , demonstrating that , although the cancellation in the coefficient of the one - loop quadratically divergent terms is possible , the parameter space is severely constrained @xcite .
an additional complex scalar triplet extension of the sm has also been studied and proven to be a solution to the fine - tuning problem @xcite . in the present work , as a completely new approach never explored before , we study protection of the higgs boson mass by a sm - singlet gauge field ( not a scalar field as in @xcite ) .
in contrast to the attempts based on hidden scalars @xcite , which are now known to be unable to simultaneously protect the masses of the higgs boson and the singlet scalar @xcite , in the present work , we consider a hidden @xmath10 gauge field @xmath11 whose invariance is ensured by a stueckelberg scalar @xmath12 and whose mass is spontaneously induced by the sm higgs field .
we show that @xmath11 and @xmath12 enable cancellation of the quadratic divergence in higgs boson mass with no quadratic divergence arising in their own masses .
it is important that the sm higgs boson is stabilized at one - loop along with already - stable hidden gauge and stueckelberg scalar .
this phenomenological advantage has important implications not only for stabilizing the higgs boson mass but also for correlating the sm higgs field with hidden sectors .
the paper is organized as follows . in section 2 below
, we construct the model starting from the basic stueckelberg setup .
section 3 is devoted to computation of the quadratic divergences and vanishing of the higgs mass divergence by fine - tuning .
we conclude in section 4 .
in this section , we consider a massive abelian gauge field @xmath11 accompanied by a real scalar field @xmath13 , introduced to preserve the gauge invariance of the theory .
originally proposed by stueckelberg @xcite and noted afterwards by pauli @xcite that , @xmath11 satisfies a restricted @xmath10 gauge invariance , with the gauge function @xmath14 obeying a massive klein - gordon equation .
the mechanism provides an alternative to the higgs mechanism , where the vector boson acquires its mass with the breakdown of the gauge invariance of not the lagrangian but of the vacuum .
these features are encoded in the stueckelberg model @xcite [ basic ] = -v_^2+m^2(v^-^s)^2 -(_v^+m s)^2 , where @xmath15 is the common mass for @xmath11 and @xmath12 . despite its massive spectrum
, this model enjoys a @xmath16 invariance [ u(1)m ] & & v_(x)v_(x)=v_(x)+_(x ) , + & & s(x)s(x)=s(x)+m(x ) , provided that @xmath17 .
consequently , in spite of its nonvanishig hard mass , @xmath11 enjoys exact gauge invariance , albeit with a restricted gauge transformation function @xmath14 @xcite . in the massless limit
, @xmath18 , the stueckelberg lagrangian ( [ basic ] ) reduces to @xmath19 , which is obviously @xmath16 invariant in lorentz gauge ( @xmath20 ) with an unrestricted @xmath14 .
interestingly , the stueckelberg scalar @xmath12 , transforming like the gauge field @xmath11 in massive case , turns into a gauge - singlet scalar in massless limit .
inspired from the stueckelberg model ( [ basic ] ) , we propose the higgsed stueckelberg model [ hs ] = & -&v_^2+_1 h^h(v^-^s)^2 + & -&(_v^+a_h s)^2 , where @xmath21 is a positive dimensionless constant and @xmath22 is a mass parameter .
this model is manifestly gauge - invariant under both the hidden @xmath23 invariance with @xmath24 , and the electroweak gauge group @xmath25 .
the higgs potential @xmath26 and hence the total energy is minimized at the higgs field configuration @xmath27 where @xmath28 is the higgs vev in the broken phase ( @xmath29 ) , to which masses of the sm particles are all proportional . in this phase electroweak gauge group @xmath30 is spontaneously broken down to electromagnetism . in unbroken phase ( @xmath31 )
electroweak group stays exact and all the sm particles but higgs boson are massless . from ( [ hs ] )
it is clear that , the two phases of the sm directly leave distinguishable effects on the mass of @xmath11 and kinetic term of @xmath12 . and the stueckelberg structure in ( [ basic ] ) is achieved properly if the mass parameter @xmath22 can keep track of the two electroweak phases .
this feature is implemented into the higgsed stueckelberg model ( [ hs ] ) by setting [ ah ] a_h=()= _ h & m_h^2<0 , + 0 & m_h^2>0 , which obviously dogs the higgs vev in ( [ higgs - vev ] ) .
it turns out that @xmath32 in both broken and exact electroweak phases , and @xmath33 specifically in the broken phase .
this switching ability of @xmath22 ensures that , in the broken phase of electroweak group , there arises , in addition to the massive sm spectrum , a massive vector @xmath11 with mass @xmath34 and a massive scalar @xmath35 . in the unbroken phase , however , the higgs field stands as the only massive field .
the rest , inclusing @xmath11 and @xmath12 , are all massless . in what follows
, we will work in the physical vacuum of the broken electroweak phase and necessarily set @xmath36 everywhere .
it is instructive to study the transcription of the stueckelberg @xmath16 symmetry in ( [ u(1)m ] ) into the higgsed stueckelberg case . to this end
, one notes that the stueckelberg scalar @xmath13 facilitates @xmath23 gauge invariance of the hidden sector , and also , helps keep the hamiltonian positive definite ) can also be written as @xmath37 , where @xmath38 is a real parameter , similar to thooft s parametrization for abelian higgs model .
the choice of @xmath39 corresponds to the stueckelberg - feynman gauge .
when @xmath40 , the restriction on the gauge function changes to @xmath41 .
it is also possible to choose two different parameters @xmath42 and @xmath43 , to check the gauge independence of the parameters .
however , there is the disadvantage that the terms of the form @xmath44 survives for this choice . in the present work
, we will work in stueckelberg - feynman gauge . ]
@xcite . in this formalism , lorentz subsidiary condition does not follow from equation of motion .
imposing an operator equation of the form @xmath45 , where @xmath46 involves the free field annihilation operators , however , gives rise to conflict between the operator equation and the canonical commutation relations .
this puzzle is solved via the introduction of an additional scalar field @xmath13 , replacing the operator equation with @xmath47|{\rm\bf phys\rangle}=0 $ ] , where @xmath48 also involves free field annihilation operators .
the operator equation decreases the number of degrees of freedom of the lagrangian to four .
the required constraint to decrease it to three for a massive vector field comes into play with the gauge transformation [ u(1)mh ] & & v_(x)v_(x)=v_(x)+_(x ) , + & & s(x)s(x)=s(x)+ _ h ( x ) , which closely follows the stueckelberg transformation ( [ u(1)m ] ) .
the @xmath23 invariance is ensured if @xmath49 .
this restricted gauge invariance changes to an unrestricted , standard gauge invariance in the unbroken ( @xmath31 ) electroweak phase in which @xmath11 and @xmath12 are massless and non - interacting .
moreover , @xmath12 is a gauge singlet in this phase .
the @xmath11 and its stueckelberg companion @xmath12 do possess identical masses in broken and unbroken phases of the electroweak symmetry .
in broken phase , stueckelberg - feynman gauge , their propagators read as _
= -,= , where @xmath50 is the common mass for @xmath11 and @xmath12 .
in this section we study quantum corrections to masses of the higgs boson @xmath51 and stueckelberg fields @xmath12 and @xmath11 . the main constraint on the model is that higgs boson must weigh @xmath52 gev @xcite . as follows from ( [ hs ] ) , there are three - point and four - point interactions among the vector boson @xmath11 , the stueckelberg field @xmath12 , and the higgs field @xmath51 .
the vertex factors are summarized in the appendix .
the higgsed stueckelberg hidden sector then modifies the veltman condition ( [ vc ] ) as ( m_h^2)_=(_h+g^2+g^2 - 6g_t^2+_1 ) , wherein @xmath21 shows up as a new degree of freedom . in the philoshopy of the original attempts in @xcite , one can suppress @xmath53 by choosing @xmath21 appropriately .
in particular , @xmath53 vanishes for @xmath54 .
the @xmath11 and @xmath12 are degenerate in mass , and for this specific value of @xmath21 they weigh @xmath55 gev . it is possible to decrease the value of @xmath21 by simply introducing @xmath56 such fields , which in turn lowers the masses of the new fields while increasing their number . in figure [ fig : bc1 ] , a schematic representation of the one - loop quantum corrections to higgs mass is shown in our extended scenario . as it is apparent from this figure , a hidden abelian gauge sector splendidly cancels the quadratically divergent contributions to higgs mass from the sm fields . denotes the higgs boson , @xmath57 the electroweak bosons , @xmath58 the top quark , @xmath59 the hidden gauge boson @xmath11 and the stueckelberg scalar @xmath12 , respectively .
higgs mass is protected from destabilizing quantum effects when the hidden gauge sector is included . ]
it is clear that suppressing @xmath53 requires @xmath21 to be finely tuned .
the fine - tuning here is of the same size as the fine - tunings required for hidden scalar sectors @xcite .
there is one big difference , however .
indeed , these models based on hidden scalars suffer from the fact that masses of the hidden scalars and of the sm higgs boson can not be protected simultaneously @xcite .
the hidden scalar continues to have a mass @xmath60 after suppressing the radiative contribution to the higgs boson mass . in the higgs - stueckelberg model
this impasse is overcome . to see this , one notes that mass of the stueckelberg field does actually receive quadratically divergent radiative corrections from two self energy diagrams ( one with higgs boson in the loop and another with both higgs and the stueckelberg field @xmath12 in the loop ) .
the self energy diagram with a higgs boson and vector boson @xmath11 in the loop diverges logarithmically .
the spruceness of this scenario emerges at this point in that the quadratically - divergent contributions to the mass of the stueckelberg field from the two loop diagrams cancel out to give ( m_s^2)_=0 . in the same manner , the mass of @xmath11 is protected against quadratically - divergent quantum corrections ( m_v^2)_=0 . leaving aside the logarithmic corrections , masses of @xmath61 and @xmath12 are found to be uv - insensitive .
this is actually expected by gauge invariance because there exists an unbroken @xmath23 invariance in both broken and unbroken electroweak phases .
the invariance protects the mass of @xmath11 .
interestingly , it also protects the mass of @xmath12 because @xmath12 by itself acts like a gauge field when @xmath11 is massive and becomes a non - interacting @xmath23 singlet when @xmath11 is massless .
clearly , the radiative stability of the hidden sector can have important implications for modelling ` dark phenomena ' like dark matter , dark energy , dark photon and neutrino masses .
the discovery of a new scalar @xcite at the lhc , consistent with the sm higgs boson , has accelerated studies on the uv - sensitivity of the higgs boson . as opposed to the physical masses of chiral fermions and gauge bosons , which are protected by chiral and gauge symmetries , there is no symmetry principle to protect the higgs boson mass against quadratically divergent quantum corrections . in the very absence of tev - scale new physics ,
one is left with a finely - tuned higgs sector where nature and degree of fine - tuning vary with the modeling details . in the presence of hidden scalars , despite the protection of the higgs boson mass the hidden sector itself is uv - unstable . in case
the hidden sector is formed by the spacetime curvature scalar , the fine - tuning is severe yet harmless because the sm fields and couplings are immune to its presence .
the fine - tuning is as severe as hidden scalars in other field - theoretic approaches . in this letter
we have shown that a hidden sector spanned by an abelian vector field whose mass is induced by electroweak breaking and whose gauge invariance is sustained by a stueckelberg scalar can lead to stabilization of the higgs boson mass by finely tuning its coupling to the sm higgs field . in spite of this unavoidable fine - tuning , the higgsed stueckelberg model possesses the striking property that the hidden sector is insensitive to the uv scale .
this stability , deriving from unbroken hidden gauge invariance , can have important collider , astrophysical and cosmological implications .
indeed , a stable hidden sector can be utilized in constructing viable models of dark matter , dark energy , dark photon and neutrino masses .
the model can be tested at the lhc ( and its successor fcc ) via direct productions of @xmath11 and @xmath12 fields .
here we list the vertex factors : @xmath62 where @xmath63 is the momentum of @xmath12 .
we used @xmath36 in @xmath22 dependent vertices . * acknowledgments * + this work has been supported in part by tbitak , the scientific and technical research council of turkey , through the grant 2232 , project no : 113c002 and by taek , turkish atomic energy authority , project no : cern - a5.h2.p1.01 - 21 .
9 g. aad _ et al . _
[ atlas collaboration ] , phys .
b * 716 * , 1 ( 2012 ) [ arxiv:1207.7214 [ hep - ex ] ] ; s. chatrchyan _ et al .
_ [ cms collaboration ] , phys .
b * 716 * , 30 ( 2012 ) [ arxiv:1207.7235 [ hep - ex ] ] .
j. ellis and t. you , jhep * 1306 * , 103 ( 2013 ) [ arxiv:1303.3879 [ hep - ph ] ] ; a. djouadi and g. g .
moreau , arxiv:1303.6591 [ hep - ph ] .
v. f. weisskopf , phys .
rev . * 56 * , 72 ( 1939 ) ; k. g. wilson , phys .
d * 3 * , 1818 ( 1971 ) ; l. susskind , phys .
d * 20 * , 2619 ( 1979 ) .
m. j. g. veltman , acta phys .
b * 12 * , 437 ( 1981 ) .
j. l. feng , arxiv:1302.6587 [ hep - ph ] ; j. d. wells , arxiv:1305.3434 [ hep - ph ] ; g. f. giudice , arxiv:1307.7879 [ hep - ph ] ; g. altarelli , phys .
scripta t * 158 * , 014011 ( 2013 ) [ arxiv:1308.0545 [ hep - ph ] ] .
m. liu and p. nath , phys .
d * 87 * , no .
9 , 095012 ( 2013 ) [ arxiv:1303.7472 [ hep - ph ] ] ; i. masina and m. quiros , phys .
d * 88 * , 093003 ( 2013 ) [ arxiv:1308.1242 [ hep - ph ] ] ; x. lu , h. murayama , j. t. ruderman and k. tobioka , arxiv:1308.0792 [ hep - ph ] ; a. fowlie , arxiv:1403.3407 [ hep - ph ] .
m. flechl [ cms and atlas collaborations ] , arxiv:1307.4589 [ hep - ex ] ; j. l. feng , j. -f .
grivaz and j. nachtman , rev .
phys . * 82 * , 699 ( 2010 ) [ arxiv:0903.0046 [ hep - ex ] ] .
w. a. bardeen , fermilab - conf-95 - 391-t .
d. a. demir , arxiv:1207.4584 [ hep - ph ] ; m. heikinheimo , a. racioppi , m. raidal , c. spethmann and k. tuominen , arxiv:1304.7006 [ hep - ph ] ; g. m. tavares , m. schmaltz and w. skiba , phys .
d * 89 * , 015009 ( 2014 ) [ arxiv:1308.0025 [ hep - ph ] ] ; y. kawamura , ptep * 2013 * , no .
11 , 113b04 ( 2013 ) [ arxiv:1308.5069 [ hep - ph ] ] ; m. holthausen , j. kubo , k. s. lim and m. lindner , jhep * 1312 * , 076 ( 2013 ) [ arxiv:1310.4423 [ hep - ph ] ] ; o. antipin , m. mojaza and f. sannino , phys .
d * 89 * , 085015 ( 2014 ) [ arxiv:1310.0957 [ hep - ph ] ] ; j. guo and z. kang , arxiv:1401.5609 [ hep - ph ] .
d. a. demir , arxiv:1405.0300 [ hep - ph ] .
a. salvio and a. strumia , jhep * 1406 * ( 2014 ) 080 [ arxiv:1403.4226 [ hep - ph ] ] .
v. agrawal , s. m. barr , j. f. donoghue and d. seckel , phys . rev .
d * 57 * , 5480 ( 1998 ) [ hep - ph/9707380 ] .
r. s. chivukula and m. golden , phys .
b * 267 * , 233 ( 1991 ) ; r. s. chivukula , m. golden and m. v. ramana , phys .
b * 293 * , 400 ( 1992 ) [ hep - ph/9206255 ] ; j. d. bjorken , int .
j. mod .
phys . a * 7 * , 4189 ( 1992 ) ; m. ruiz - altaba , b. gonzlez , m. vargas , cern - th.5558/89 ( 1989 ) ; m. capdequi peyranere , j. c. montero and g. moultaka , phys .
b * 260 * , 138 ( 1991 ) ; a. a. andrianov , r. rodenberg and n. v. romanenko , nuovo cim .
a * 108 * , 577 ( 1995 ) [ hep - ph/9408301 ] ; a. kundu and s. raychaudhuri , phys .
d * 53 * ( 1996 ) 4042 [ hep - ph/9410291 ] ; f. bazzocchi , m. fabbrichesi and p. ullio , phys .
d * 75 * ( 2007 ) 056004 [ hep - ph/0612280 ] .
i. chakraborty and a. kundu , phys .
d * 87 * , no . 5 , 055015 ( 2013 ) [ arxiv:1212.0394 [ hep - ph ] ] .
j. mcdonald , phys .
d * 50 * , 3637 ( 1994 ) [ hep - ph/0702143 [ hep - ph ] ] ; d. a. demir , phys .
b * 450 * , 215 ( 1999 ) [ hep - ph/9810453 ] ; c. p. burgess , m. pospelov and t. ter veldhuis , nucl .
b * 619 * , 709 ( 2001 ) [ hep - ph/0011335 ] ; v. barger , p. langacker , m. mccaskey , m. j. ramsey - musolf and g. shaughnessy , phys .
d * 77 * , 035005 ( 2008 ) [ arxiv:0706.4311 [ hep - ph ] ] ; v. barger , p. langacker , m. mccaskey , m. ramsey - musolf and g. shaughnessy , phys . rev .
d * 79 * , 015018 ( 2009 ) [ arxiv:0811.0393 [ hep - ph ] ] ; w. -l .
guo and y. -l .
wu , jhep * 1010 * , 083 ( 2010 ) [ arxiv:1006.2518 [ hep - ph ] ] ; a. djouadi , o. lebedev , y. mambrini and j. quevillon , phys
. lett .
b * 709 * , 65 ( 2012 ) [ arxiv:1112.3299 [ hep - ph ] ] ; m. gonderinger , h. lim and m. j. ramsey - musolf , phys .
d * 86 * , 043511 ( 2012 ) [ arxiv:1202.1316 [ hep - ph ] ] ; b. batell , d. mckeen and m. pospelov , jhep * 1210 * , 104 ( 2012 ) [ arxiv:1207.6252 [ hep - ph ] ] ; s. baek , p. ko , w. -i .
park and e. senaha , jhep * 1211 * , 116 ( 2012 ) [ arxiv:1209.4163 [ hep - ph ] ] ; a. biswas and d. majumdar , pramana * 80 * , 539 ( 2013 ) [ arxiv:1102.3024 [ hep - ph ] ] ; g. belanger , k. kannike , a. pukhov and m. raidal , jcap * 1301 * , 022 ( 2013 ) [ arxiv:1211.1014 [ hep - ph ] ] ; j. m. cline and k. kainulainen , jcap * 1301 * , 012 ( 2013 ) [ arxiv:1210.4196 [ hep - ph ] ] ; d. a. demir , m. frank and b. korutlu , phys .
b * 728 * , 393 ( 2014 ) [ arxiv:1308.1203 [ hep - ph ] ] ; n. haba , k. kaneta and r. takahashi , jhep * 1404 * , 029 ( 2014 ) [ arxiv:1312.2089 [ hep - ph ] ] . c. n. karahan and b. korutlu , phys .
b * 732 * , 320 ( 2014 ) [ arxiv:1404.0175 [ hep - ph ] ] .
i. chakraborty and a. kundu , arxiv:1404.3038 [ hep - ph ] . i. chakraborty and a. kundu , arxiv:1404.1723 [ hep - ph ] .
e. c. g. stueckelberg , helv .
acta * 11 * , 299 ( 1938 ) ; helv .
acta * 11 * , 225 ( 1938 ) .
w. pauli , rev .
* 13 * , 203 ( 1941 ) .
b. kors and p. nath , phys .
b * 586 * , 366 ( 2004 ) [ hep - ph/0402047 ] ; jhep * 0507 * , 069 ( 2005 ) [ hep - ph/0503208 ] . | here we show that , a hidden vector field whose gauge invariance is ensured by a stueckelberg scalar and whose mass is spontaneously generated by the standard model higgs field contributes to quadratic divergences in the higgs boson mass squared , and even leads to its cancellation at one - loop when higgs coupling to gauge field is fine - tuned .
in contrast to mechanisms based on hidden scalars where a complete cancellation can not be achieved , stabilization here is complete in that the hidden vector and the accompanying stueckelberg scalar are both free from quadratic divergences at one - loop .
this stability , deriving from hidden exact gauge invariance , can have important implications for modelling dark phenomena like dark matter , dark energy , dark photon and neutrino masses .
the hidden fields can be produced at the lhc . |
procyon a is the 14th nearest star and stellar system to the sun , according to henry ( 2011 ) , at a distance of 3.51 parsecs or 11.46 light years
see van leeuwen ( 2007 ) .
its spectral type is f5 iv - v .
bessel ( 1844 ) discovered that the ( large ) proper motion of procyon on the sky was perturbed by what he recognized to be an unseen companion .
procyon b was first detected visually at the end of the 19th century by schaeberle ( 1896 ) . the first recognition that the companion was like those in the sirius and 40 eri systems , later to be called white dwarfs , may be the work of the irish amateur astronomer john ellard gore .
however , he did not properly publish this result in the astronomical literature see the discussions of his work by fitzgerald ( 1966 ) and holberg ( 2009 ) .
the first orbital elements were determined by auwers ( 1862 ) , who showed that the orbital period would be about 40 years .
we now know that the more exact period is 40.838 years ( h. e. bond & g.h .
schaefer , 2012 private communication ) .
solar - like oscillations in procyon a were first reported by brown et al .
( 1991 ) , and several authors have since investigated the structure and evolution of that star through a seismic approach ( see , e.g. , dogan et al . 2010 and references therein ) .
we come back to the asteroseismological properties of procyon a in 5 . like sirius a studied in liebert et al .
( 2005 , hereafter paper i ) , procyon a has an extremely accurate interferometric radius measurement , and an extremely accurate trigonometric parallax ; the luminosity is also very accurate .
sirius a is certainly a main sequence star , while procyon a could be entering or about to enter the subgiant branch .
in contast to the sirius system , however , the astrometric mass solution for the procyon system has been a controversial issue , as we shall discuss in 2 .
in comparison with the white dwarf sirius b , procyon b is very different from other ( mostly da ) white dwarfs previously used for the problem of the star s initial - to - final mass relation ( ifmr ) .
mainly it has a helium - dominated atmosphere polluted by heavy elements ( dqz spectral type ) , a lower mass , and a cooler temperature .
the mass , radius and @xmath5 determined for the b component will be discussed in 3 . in 4 we address the atmospheric abundances of procyon a and the interior abundances to be used for modelling in 5 .
here we employ the tycho stellar evolution code ( young & arnett 2005 ) to fit the position of procyon a in an hr diagram , and determine the age . in
6 eleven calculations of the cooling time of procyon b are presented , for different assumptions about the mass , the helium envelope mass , and its composition .
the cooling time is the length of time since the degenerate entered the white dwarf sequence ; this has been used for the coolest white dwarfs to estimate ages of the galactic disk ( liebert , dahn , & monet 1988 ) and halo ( hansen et al .
2007 using the globular cluster ngc 6397 ) .
the difference between the systemic age and the white dwarf cooling time yields the best theoretical estimate of the initial mass of the white dwarf .
moreover , as discussed in 6 , procyon b presents a complicated case for the calculation of the cooling age .
further remarks pertinent to this more challenging physical situation are discussed in that section .
this paper adds a valuable new object with accurately determined stellar parameters ( masses , radii , luminosities ) to the initial - to - final mass relation ( ifmr ) for progenitors and white dwarfs .
the vast majority of other data points have white dwarfs with hydrogen - rich atmospheres , while procyon b has a complicated helium - rich envelope and atmosphere .
it is possible that the ifmr for helium - rich white dwarfs may differ from the hydrogen - rich distribution .
we show in 7 how procyon b compares with sirius b and white dwarfs in star clusters for which the ifmr of low to intermediate mass stars has been determined .
as noted in the introduction , a precise radius measurement is available for procyon a , from measurements of the angular diameter with the eso very large telescope interferometer .
the radius is @xmath6 = [email protected]@xmath7 or @xmath00.65% ( aufdenberg , ludwig , & kervella 2005 ) .
this radius value is based on the authors `` best estimate '' angular diameter of [email protected] mas at 2.2@xmath8 .
note that the authors constructed model atmospheres for procyon in three ways : ( 1 ) stand - alone one - dimensional ( 1d ) phoenix structures , radiation fields , and spectra ( hauschildt et al . 1999 ) ; ( 2 ) co@xmath9bold 3d structures temporarily and spatially averaged to 1d , then read by phoenix for computation of the corresponding radiation fields ( freytag et al .
2002 ) ; and ( 3 ) stand - alone 1d atlas 12 ( kurucz 1992 ) structures , radiation fields and spectra .
note that the quantitative assessment of the limb darkening , on which the radius value is determined , depends on the model atmospheres analysis .
the authors discuss in detail their assessment of these different models , leading to their `` best estimate '' .
these authors also cite the earlier measurement on mt .
wilson , california using the mark iii interferometer at 500 and 800 nm ( mozurkewich et al 1991 ) , whose limb - darkened value of the angular diameter ( [email protected] mas ) agrees marginally within the errors .
their new measurement compares to an older , slightly less accurate , but consistent value of [email protected]@xmath7 by the same group ( kervella et al .
2004 ) , which corresponds to an adopted angular diameter of [email protected] mas .
casagrande et al .
( 2010 ) got a lower , less accurate , but consistent [email protected] mas . for the radius ( @xmath10 ) of the primary star in this paper ,
we adopt the aufbenberg et al .
( 2005 ) best value .
note that the kervella et al . and aufdenberg et al .
results differ by less than 1@xmath11 .
adopting either one results in an essentially identical fit .
the _ hipparcos _ satellite mission provided the precise measurement of the trigonometric parallax of [email protected] " ( or @xmath00.3% ) . using the bolometric energy distribution and the effective temperature determination ( 6,543@xmath084 k )
, the luminosity of procyon a is determined in aufdenberg et al .
( 2005 ) to be log @xmath12/@xmath13 = + [email protected] .
note that chiavassa et al .
estimate @xmath14 = 6591 k or 6556 k depending on how the bolometric flux is calculated , within the error bars of aufdenberg et al .
the former value differs the most from aufdenberg et al.s preferred 6543 k , resulting in log @xmath12/@xmath13 = + 0.843 ( with one more significant digit than is warranted ) .
the difference between these two values hardly matters in figure 1 .
two decades ago the mass of the `` a '' component was believed to be near 1.75 @xmath3 ( irwin et al .
. however , main sequence evolutionary models at that mass were too luminous by a factor of two compared with the observed luminosity ( guenter & demarque 1993 ) .
long - term imaging by hst , combined with the interferometric radius and trigonometric parallax measurements cited above led to a great improvement in the stellar parameters .
first , girard et al.(2000 ) produced an improved astrometric solution based on analysis of 250 photographic plates spanning 83 years , augmented with early hubble data .
they found m@xmath15 = [email protected] @xmath3 , almost exactly the 1.50 @xmath3 predicted by guenther & demarque ( 1993 ) .
the problem is that hst imaging has only been possible for only about half of the orbital period .
the combining of space - based imaging and the much longer series of ground - based observations has often been fraught with difficulties .
apparent systematic offsets often occur in the respective measurements of the separations and position angles .
girard et al . ( 2000 ) found that , if they restricted the astrometric analysis to the excellent images obtained with the wfpc2 camera
i.e. , with no ground - based data a substantially lower mass ( 1.465 @xmath3 ) for procyon a was the result .
at that time hst imagery covered only about a decade .
allende prieto et al . ( 2002 ) revisited the procyon a solution using the _ hipparcos _ parallax and the stellar angular diameter by mozurkevich et al .
( 1991 ) discussed previously .
they derived the primary s mass to be a much lower [email protected] @xmath3 .
this is consistent with the ground - based gatewood and han ( 2006 ) value of the primary s mass of [email protected] @xmath3 , determined from their multichannel astrometric photometer * map * camera , built at the university of pittsburg ( gatewood 1987 ) .
the * map * parallax is [email protected] mas , which agrees well with the _ hipparcos _ value .
schaefer et al . ( 2006 ) reported a preliminary solution resulting from a space telescope science institute - led program to extend the imaging program on the procyon binary orbit .
their result was in excellent agreement with the girard et al .
( 2000 ) analysis including ground - based data .
several more years of hst data since 2006 now result in the following masses : [email protected] @xmath3 for procyon a and [email protected] @xmath3 for procyon b ( g. h. schaefer and h. e. bond , 2012 private communication ) .
this solution uses available hst data from 1995 through 2012 plus , appropriately weighted , the archival ground - based data .
it is not difficult to see how the total mass of the system is specified from accurate hst astrometry . using newton s gravitational formula for kepler s third law , ( @xmath16 + @xmath17 ) = @xmath18 / @xmath19 @xmath20 , with the semimajor axis ( @xmath21 ) and period ( @xmath22 ) well known ,
the sum of the masses is fixed at 2.052 @xmath3 .
ground - based observations of the wobble of the a component about the center - of - mass on the sky then fix @xmath23 , @xmath23 + @xmath24 = @xmath21 , and @xmath16/@xmath17 = @xmath24/@xmath23 .
thus the solution provides separate values for @xmath16 and @xmath17 .
the mass determinations remain a matter of dispute in the literature .
we begin with 1.499 @xmath3 as the hst - based , preferred mass of the primary for fitting in 5 .
we also consider 1.42 @xmath3 as a lower bound to the mass of the primary .
note that , in their recent analysis , chiavassa et al .
( 2012 ) give arguments in favor of the 1.43 @xmath3 gatewood and han value .
procyon b was first detected visually at the end of the 19th century by schaeberle ( 1896 ) , and was historically one of the first white dwarfs to be discovered .
even early estimates of the mass ( @xmath17 ) of the white dwarf component , such as that by strand ( 1951 ) , showed that it is much closer to the mean value of field white dwarfs than is sirius b. the updated schaefer et al .
hst solution gives [email protected]
@xmath3 for the mass of the secondary , and we consider this the benchmark value for @xmath17 . to be sure this is much smaller than previous estimates for the secondary , such as 0.622 @xmath3 ( irwin et al .
1992 ) and [email protected] @xmath3 ( girard et al . 2000 ) .
values in between these have been published .
the kervella et al .
( 2004 ) lower mass solution of 1.42 @xmath3 for the primary led to @xmath17 = [email protected] @xmath3 for b. the gatewood and han ( 2006 ) primary mass of [email protected] @xmath3 resulted in [email protected] @xmath3 for the secondary . to illustrate the dependence on mass , the cooling time for a mass of 0.602 @xmath3 ( girard et al .
2000 ) is also discussed in 6 .
as mentioned in the introduction , the nature of the `` b '' white dwarf poses additional problems .
the atmosphere is helium - rich , with trace abundances of carbon , calcium , magnesium and iron , but no detection of hydrogen .
this is an unusually diverse abundance pattern for a white dwarf atmosphere , and complicated greatly the determination of the atmospheric parameters .
not knowing of the complication , in the absence of an ultraviolet spectrum , provencal et al .
( 1997 ) used pure hydrogen and helium atmospheres to model the spectral energy distribution . preferring the helium fit due to its optical `` dc '' spectrum
, they found @xmath14 = 8,688 k. this implied a stellar radius of [email protected] @xmath7 .
given the mass , this small radius implied that the interior composition of the white dwarf could not be carbon - oxygen , but rather something very dense like iron ( shipman & provencal 1999 ) .
this conclusion was unsettling to stellar theorists .
once the complicated ultraviolet spectrum was revealed with the space telescope imaging spectrograph on the hubble space telescope , provencal et al .
( 2002 ) got themselves out of this `` iron box '' .
atmospheres with realistic trace adundances , obtained from fitting the spectrum and the 1,80010,000 spectral energy distribution were employed to obtain @xmath14 = 7,740@xmath050 k. they now estimated the radius to be [email protected] @xmath7 , consistent with a carbon - oxygen interior .
note that procyon b is the only white dwarf in a binary system or star cluster which does not have an atmosphere amenable to a mass determination from either the fitting of hydrogen or he i lines .
it is also the coolest such white dwarf .
there is a long literature trail indicating that the procyon a atmospheric abundance can not be distinguished from solar .
we cite five studies : heiter & luck ( 2003 ) , and luck & heiter ( 2005 ) , as part of a study of all bright , nearby stars in the northern hemisphere , analyzed procyon a using high resolution spectra .
they found an iron abundance [ fe / h ] = @[email protected] .
valenti & fischer ( 2005 ) , as part of a systematic study of nearby stars using echelle spectra , obtained [ fe / h ] = @[email protected] . in allende prieto
et al.s ( 2004 ) study of stars more luminous than m@xmath26 = + 6.5 within 14.5 pc of the sun , a value of + 0.08 was determined for procyon a. finally , in the most comprehensive geneva - copenhagen study of solar neighborhood stars ( nordstrm et al .
2004 ) , a value of [ fe / h ] = + 0.05 was reported .
both of the last two author groups discuss the possible systematic error due to the coupling of the [ fe / h ] determination to the error in the log @xmath27 value .
note that procyon a is cool enough that peculiar diffusion / gravitational support mechanisms should not distort the surface abundances from the interior abundance .
there could , nevertheless , be a small diffusion of heavy elements out of the thin convective envelope . however , recently there has been controversy about what the solar abundance actually is .
inevitably , the abundance parameters to be run for the stellar evolution models of procyon a , to be described next in 5 , must therefore be linked to this controversy , so a few comments are in order . the newer solar abundance scale determined from 3d , non - lte calculations of the solar atmosphere ( asplund et al .
2004 , 2005 , 2009 the latter hereafter a09 ) , and now lodders 2010 ) have significantly lower abundances of oxygen , carbon and nitrogen than previous estimates ( grevesse & sauval 1998 ) . the abundance of neon is a particular problem , since it is impossible to measure this element in the sun ( or procyon a ) , but is likely to be comparable numerically to the three elements mentioned previously .
the overall heavy element abundance parameter for the sun decreases to a primordial value of @xmath1=0.0153 .
the asplund a09 scale has been challenged because it breaks the accordance of the `` standard solar '' interior model with helioseismological measurements ( bahcall et al .
a solar model with these abundances incorrectly predicts the depth of the convection zone , the depth profiles of sound speed and density , and the helium abundance ( basu & antia 2004 ) . as a check on the magnitude of likely systematic errors due to uncertainty in the actual solar ( and
procyon a ) abundance , we retain consideration of the older grevesse & sauval ( 1998 ) solar abundance which corresponds to @xmath1=0.019 .
the tycho code http://chandra.as.arizona.edu/@xmath28dave/tycho-intro.html incorporates the most current microphysics for opacities , equation - of - state , and nuclear reactions . in particular , the code has been upgraded with an algorithm based on a physical analysis of 3d hydrodynamic simulations of convection , not an astronomically - calibrated version of the mixing length theory .
it includes non - locality and time dependence of flow , dynamical acceleration , turbulent dissipation , kolmogorov heating , compositional effects and dynamically - defined boundary conditions ( instead of parameterized overshooting schemes ) , all in a single , self - consistent formulation .
the code is regularly tested against observations of double - lined eclipsing binaries and cluster isochrones to ensure consistency and accuracy .
it produces superior fits to observational test cases without adjustment of parameters .
these changes make possible fits to nuclear - burning stars on the hr diagram , especially those on or near the main sequence ( arnett , meakin , & young 2009 , 2010 ) .
robust , consistent ages for ( nondegenerate ) binary stars with well - determined luminosities , masses , radii and temperatures have been determined by young et al .
( 2001 ) , young & arnett ( 2005 ) . in additon to the mass
, we have to input the abundance parameters @xmath29 and @xmath1 , where @xmath1 scales by the solar ratios of individual elements .
since we can not regard the solar atmospheric abundance issue as fully resolved , we therefore have computed sequences using the lodders ( 2010 ) value of @xmath1 = 0.0153 and also evaluated the older solar abundance of grevesse and sauval ( 1988 ) which imply @xmath1 = 0.019 . in figure 1
we show evolutionary tracks for @xmath1 = 0.0153 of ( 1 ) 1.499 @xmath3 ( in black ) , the best value from preliminary hst astrometry ( 2 ) , fitting the correct luminosity and radius at age 1.74 gyr ; ( 2 ) 1.48 @xmath3 ( red ) at 1.87 gyr , the best fit to the luminosity ( to three significant digits ) ; ( 3 ) 1.463 @xmath3 ( green ) at 1.99 gyr ; and ( 4 ) the lower 1.42 @xmath3 ( blue ) at 2.51 gyr favored by largely ground - based studies cited in 2 .
the highest three masses fit within 1@xmath4 late on the main sequence . for 1.42
@xmath3 the fit misses by more than 2@xmath4 late on the main sequence .
note that best - fit ages include the pre - ms evolution , without pre - ms accretion , but only the tracks from the beginning of the main sequence are shown here .
we also tried a fit with the grevesse & sauval ( 1998 ) @xmath1 = 0.019 abundance at 1.48 @xmath3 , but it is 1.45@xmath4 underluminous when it reaches the correct radius on the main sequence at age 1.96 gyr . given the evidence from a09 and lodders ( 2010 ) that this solar metallicity value is too large , this marginal attempt at a fit is not shown in the figure .
we therefore believe that [email protected] gyr from the 1.48 @xmath3 fit is the best estimate of the age of the binary system , with the uncertainty given by the bracketing age values for the 1.499 and 1.463 @xmath3 fits .
note that provencal et al .
( 2002 ) estimated the nuclear lifetime of 1.3 gyr for component b , estimating @xmath282.1 @xmath3 as its progenitor mass .
these authors use the cooling age from models of wood ( 1995 ) for pure carbon cores the only calculations available then .
we now explore a range of alternative compositions for the white dwarf in 6 .
since our best - fit mass using the tycho models lies well within the error bars of the preliminary hst astrometric mass of 1.499 @xmath3 , we adopt 1.48 @xmath3 as the preferred or `` benchmark '' mass of procyon a. note that the fitting of a star with mass near 1.5 @xmath3 at this stage of evolution presents uncertainties due to having both a small convective core and a small convective envelope . due to the former
, the main sequence lifetime is affected by the uncertainty in the amount of hydrogen fuel available for reactions in the core .
note that remaining uncertainties in the treatment of mixing beyond the convective core boundary , in the primordial solar abundance mix , or in any of the other model physics probably add larger systematic errors , by amounts which are difficult to quantify .
the convective envelope leads to implied small differences between the interior and atmospheric @xmath1 abundances , as noted earlier .
for kervella et al.s ( 2004 ) similar analysis , the fit to the hrd parameters also occurs in a similar way .
if 1.50 @xmath3 is assumed , their fit is late on the main sequence at the much younger age of 1.3 gyr .
however , these authors assumed _ no overshooting _ of the convective core , i.e. , less fuel , vs our calculations .
for their preferred mass of 1.42 @xmath3 , the corresponding fit is early in the h - shell burning phase at ages of 2.312.71 gyr .
this age range is close to what we achieve at 1.42@xmath3 for which we determine an age of 2.51 gyr ( stated three paragraphs earlier ) , plotted as the blue track in figure 1 . as indicated in the introduction , stochastically excited @xmath30-modes
have been observed in procyon a since first reported by brown et al .
( 1991 ) , and several authors have attempted to exploit the seismic potential of that star using different data sets and models ( e.g. , guenther & demarque 1993 ; barban et al .
1999 ; chaboyer et al . 1999
; di mauro & christensen - dalsgaard 2001 ; eggenberger et al . 2005 ; provost et al .
2006 ; bonanno et al .
however , their investigations have been limited by the accuracy of the detected pulsation frequencies .
more recently , better data became available ( bedding et al .
2010 ) and have been modelled in a preliminary way by dogan et al .
we note that dogan et al .
( 2010 ) report two possible seismic model for procyon a , one of which characterized by a total mass of 1.50 @xmath3 and an age of 1.83 gyr ( no uncertainties provided ) , which is remarkably close to our solution .
we find this result most encouraging .
the importance of the cooling age of a white dwarf was introduced in the introduction . for procyon
b we first calculate the cooling time for a non - da white dwarf of @xmath17 = [email protected] @xmath3 , the benchmark mass , but also evaluate the earlier alternative of @xmath17 = 0.602 @xmath3 from kervella et al .
the fitted @xmath31 is 7,740@xmath050 k , as discussed in
3 . these error bars are the published , internal errors of the fit to the spectral energy distribution and absorption features , but are uncomfortably less than 1% of the value itself . for this object with a complicated dqz spectrum ,
we feel more comfortable assuming quadruple the formal @xmath14 error or @xmath0200 k. there have unfortunately been a shortage of available evolutionary sequences for white dwarfs lacking thick hydrogen envelopes .
most cluster white dwarf studies in the last few decades have used the evolutionary models of wood ( 1992 , 1995 ) , but always with an outer hydrogen layer .
these calculations do include one sequence with no hydrogen , a `` thin '' helium envelope of 10@xmath32 , atop a pure carbon core but with solar abundances of elements heavier than helium ( wood 2012 , private communication ) .
hansen ( 1999 ) showed that pure helium envelope cooling models for 0.6 @xmath3 by himself ( for c , o and c / o cores ) and by salaris et al .
( 1997 ) cool considerably faster than the wood sequence ( see hansen s fig . 7 ) .
note that if a trace of heavy elements were assumed , the envelope opacity would be considerably increased .
hence , there has been some confusion as to what the cooling rates of white dwarfs lacking hydrogen layers should be .
this issue is addressed in 6.2 .
for a cool , non - da white dwarf with a simple , featureless spectrum , one could argue that the use of a pure - he envelope sequence such as from hansen ( 1999 ) would yield the more robust cooling time .
however , the spectrum of procyon b shows the effects of ( 1 ) dredge - up of carbon from the diffusion tail of the core , with an atmospheric abundance log [ c / he ] = @[email protected] ( provencal et al .
2002 ) ; and ( 2 ) accretion of heavy elements
log mg / he = @xmath2510.4 , log ca / he = @xmath2512 , log fe / he = @xmath2510.7 with log h / he @xmath33 @xmath254 .
provencal et al .
speculate that the secondary could be accreting a stellar wind from the primary at a rate of about 2@xmath34 10@xmath35 yr@xmath36 .
thus , a calculation including additional opacity in the envelope layer may be more appropriate .
moreover , systematic errors need to be considered also due to the uncertain mix of carbon and oxygen in the core , and the thickness of the helium envelope . for this purpose one of us ( gf )
has calculated a new set of cooling tracks to estimate these errors and parameter dependences , and hopefully thus bound the possible cooling time of this white dwarf .
salaris et al .
( 1997 ) explored the dependence of cooling rates on the internal chemical distribution of carbon and oxygen in the core . for their best choices of the combined effect of convective mixing and the @xmath37c(@xmath38,@xmath39)@xmath40o reaction rate , carbon - oxygen profiles showing an enhancement of oxygen in the central regions
are obtained for all white dwarf masses .
mass fractions of @xmath40o as high as 0.8 are found near 0.6 @xmath3 .
this fraction declines with increasing core mass because the above reaction rate is highest at the lower core temperatures characteristic of lower masses .
thus , a core composition dominated by oxygen , or at least a mixture of carbon and oxygen , is more likely than one of pure carbon for procyon b. sequences have been calculated for pure oxygen , pure carbon , and for mixtures of carbon and oxygen according to the mass - dependent and depth - dependent results of salaris et al .
we pick the last of these three types of sequences as the benchmark core composition .
the cooling time also depends on the thickness of the helium layer .
the helium - rich atmospheric composition indicates that some of the helium layer mass could have been lost in a late helium - shell flash that disposed of all of the hydrogen ( iben 1984 ) .
values in the literature that have been considered include masses between 10@xmath41 and 10@xmath42 .
the atmospheric carbon abundance is a strong constraint , since this results from dredgeup of the diffusion tail of carbon from the core .
the study ( fontaine & brassard 2005 ) fitting dredgeup models to match the spectra of procyon b and a sequence of field dq white dwarfs by dufour , bergeron , & fontaine ( 2005 ) shows that the atmospheric carbon abundance of procyon b favors a layer mass thickness of log @[email protected]@xmath00.5 ( see their fig .
we therefore consider this value as the benchmark , with sequences also calculated for a reasonable range of @xmath252.0 and @xmath253.0 .
finally , it is very difficult to estimate the opacities due to elements heavier than helium in the partially - degenerate helium layer .
this will depend on how deeply the heavy elements have diffused into the envelope below the convection zone . in principle
the carbon abundance profile can be calculated for the helium envelope from diffusion theory , with the atmospheric abundance as a surface boundary . to consider this dependence in an approximate way ,
sequences with @xmath1=0.001 and @xmath1=0 have been calculated , with the former as the benchmark while the latter is contradicted by the observation of carbon and heavier elements in the spectrum . in table 1 , the cooling ages for the parameters of procyon b are listed for 8 sequences at the favored mass of 0.553 @xmath3 , two for values at @xmath00.022@xmath3 , and one for the alternative 0.602 @xmath3 ( girard et al .
also listed for each sequence are the core composition ( c , o , or a more realistic salaris c / o mix ) , the helium layer mass ( @[email protected] as discussed earlier ) , and the heavy element abundance @xmath1=0.001 or 0 .
all assume zero hydrogen abundance .
the benchmark cooling age of procyon b is given by the model for 0.553 @xmath3 , with c / o core , log m(he)/m = @xmath252.5 , and @xmath1 = 0.001 in the envelope ( benchmark parameters ) .
the cooling value is 1.187 gyr , with a formal , internal error of @xmath00.104 gyr due to uncertainties in the assumed effective temperature and the assumed mass . according to table 1
, a range of @xmath0200 k about the estimated effective temperature of @xmath14 = 7740 k translates into an age uncertainty of @xmath00.085 gyr , while a range of @xmath00.022 @xmath3 about the benchmark mass of 0.533 @xmath3 translates into an age interval of @xmath00.061 gyr .
added in quadrature , these uncertainties lead to @xmath00.104 gyr .
in addition , two sources of systematic error should be considered here .
the age difference between a mixed ( c / o ) and a pure oxygen core is relatively small at 0.010 gyr .
for its part , uncertainty due to variations in the helium layer mass is @xmath00.043 gyr .
added again in quadrature with the above internal error , these systematic effects lead to a total uncertainty of @xmath00.113 gyr on the age .
so we reach the value of [email protected] gyr for the white dwarf cooling time of procyon b. on the other hand , the @xmath1 = 0 envelope sequence with otherwise benchmark assumptions yields a considerably larger age of 1.391 gyr . since this is inconsistent with ultraviolet observations of carbon and heavy elements in the ultraviolet spectrum , we consider a metal - free envelope an unrealistic assumption .
note that the atmospheric values determined by provencal et al .
( 2002 ) already have a heavy element abundance ( dominated by carbon ) of [email protected] .
this parameter increases greatly below the convection zone throughout the partially - degenerate helium envelope ( following profile of carbon dredge - up ) , until it reaches the c / o core .
likewise , we do not consider the assumption of a pure carbon core realistic .
the combination of a pure carbon core with zero metallicity envelope yields an even larger cooling age of 1.557 gyr .
the cooling age listed for the alternative mass of 0.602 @xmath3 , assuming benchmark values of the other parameters above , is 1.406 gyr . a mass this large for the secondary disagrees by more than 2@xmath4 from that derived from hst astrometry and we no longer consider it in the rest of this paper .
these several possibilities in this paragraph are not factored into the error bars in the preceding paragraph .
the difference between the age of the procyon system and the cooling time of the white dwarf component , [email protected] gyr , corresponds to the sum of the pre - main sequence lifetime and the main sequence lifetime ( plus the much shorter red giant phases ) of the procyon b progenitor . using again our tycho evolutionary code
, we thus find that the initial mass ( @xmath44 ) of the secondary is 2.59@xmath45 @xmath3 , if the envelope z value is fixed at 0.001 .
however , if one allows for the extreme upper limit to the cooling time from the assumption of z(env ) = 0 , the error bars for the progenitor mass increase to @xmath2 @xmath3 .
the quoted errors on @xmath44 are the quadrature sums of @xmath250.21,+0.40 due to the incertainty in the cooling time of the white dwarf , and @xmath250.16,+0.20 due to systematic error in the age of the binary system ( table 2 ) . in figure 2 ,
discussed in detail in the next section , the solid horizontal error bars for procyon b are for the case of z(env ) fixed at 0.001 , the extended light error bars reflect the increase in cooling time for the limiting case z(env ) = 0
. it may be worthwhile at this point to attract the attention of the reader on some subtleties of white dwarf cooling which leave their signature on the data presented in table 1 .
for instance , as is well known ( and revealed again in the table ) , pure c core models systematically cool slower than models having cores made of heavier elements , all other things being the same . that a pure carbon core takes longer to cool
is well understood , because the thermal energy in a given mass is locked up in more nucleons than for pure oxygen or a mixture . however , that the @xmath1 = 0 envelope takes longer to cool than one with heavy elements is not so intuitive , because one might think that the more opaque envelope with heavy elements would let the interior energy out more slowly .
our calculations show that cooling proceeds in two phases . at first
the less opaque model does let the thermal energy escape more quickly , and core temperature decreases more rapidly . despite this
, however , the decrease in luminosity has to be slower for the less opaque model because of that extra energy that is made available to fuel the luminosity .
hence , in a first phase , the less opaque model actually cools _ more slowly _ to a given luminosity or @xmath14 .
however then , after the less opaque star has let more of its thermal energy escape , there is no turning back and the price to pay , in a second phase , is that it will now cool _
faster _ the rest of the way than its more opaque counterpart . in the case of procyon
b at @xmath14 = 7,740 k , it is still warm emough to be in the first phase of that `` relative '' cooling and , hence , this is why the cooling age is larger for the less opaque model . if the star had a @xmath14 of @xmath285,000 k , then indeed the cooling age based on the less opaque envelope sequence would be shorter than than of the more opaque , as the conventional wisdom would expect .
these arguments may be discerned from a careful perusal of section iii(a ) , subsection ( ii ) of tassoul , fontaine and winget ( 1990 ) , although this discussion compares an opaque envelope of hydrogen with less opaque counterparts .
as discussed in the last paragraph of the introduction , the initial - to - final mass relation ( ifmr ) for progenitors and white dwarfs is fundamental in understanding a stellar population .
this study of procyon adds a data point near the low mass end with well determined stellar parameters ; it may be valuable for testing whether differences exist in the ifmr for the distributions of hydrogen and helium - rich atmosphere white dwarfs .
table 2 lists final ( @xmath46 ) and initial ( @xmath44 ) masses in @xmath3 , cooling times ( log in years ) , and estimated errors , first for procyon b and sirius b. the remaining data points come from white dwarfs found in galactic star clusters , in order of increasing age ( see footnotes ) .
the collective data set is shown in figure 2 .
again , the age of the population is obtained generally from fitting the main sequence turnoff of the cluster color - magnitude diagram . as given in ferrario et al .
( 2005 ) , the errors in @xmath44 fall into two categories observational ( obs ) or random , and systematic ( sys ) , as listed in table 2 .
the observational errors devolve from uncertainties in the @xmath46 from fitting the balmer lines ( or , in one case , he i lines ) ; the systematic errors come from uncertainties in the cluster or binary system age .
since the nuclear lifetime of the progenitor is the difference between the systemic age and the white dwarf cooling time , @xmath44 depends on the age as well .
note that the uncertainties in the age determinations for the young clusters lead to large error bars in the estimates of the initial masses @xmath44 . we add gd 50 to the pleiades `` moving group '' , since dobbie et al .
( 2006 ) make a strong case based on astrometric and spectroscopic data that this ultramassive white dwarf is associated with the star formation event that created the pleiades cluster .
they argue that it evolved as a single star from a progenitor of 6.3 @xmath3 , and that this may represent the first observational evidence that single - star evolution can produce white dwarfs with , in this case , a mass of 1.27 @xmath3 ( bergeron , saffer , & liebert 1992 ) or 1.2(+0.07,@xmath250.08 ) @xmath3 ( bergeron et al .
they make a weaker case that the massive white dwarf pg 0136 + 251 , [email protected] @xmath3 ( bergeron , saffer , & liebert 1992 ) , may also be related to the pleiades .
however , information is lacking on the total ( @xmath47 ) space motion of this object , so we do not include it here .
we adopt an age of 125@xmath025 myr for the pleiades , which depends on the treatment of the convective core .
this excludes a systematic uncertainty to the age which will be true for all clusters with main sequence turnoff stars of mass above @xmath28 1.3 @xmath3 .
williams , bolte , & koester ( 2004 , 2009 ) present studies of white dwarfs in the cluster ngc 2168 ( m35 ) ; this work supercedes the data listed in ferrario et al .
( 2005 ) . in the updated listing , 2168 - 22 is excluded since it is likely magnetic , and the weak balmer lines could not be fit by the authors .
the age of the cluster is believed to be 175@xmath050 myr ( sung and bessell 1999 , von hippel 2005 ) . for ngc 3532 ( age 300@xmath025 gyr )
we take the view that the analysis with one of the 8.2 m vlt telescopes by dobbie et al .
( 2009b ) supercedes the earlier eso observations of koester & reimers ( 1993 ) . for ngc 2099 ( age 490@xmath070 myr )
, we note that star 2099 - 14 at [email protected] @xmath3 likely has a helium core , implying formation from binary star evolution .
a new white dwarf in the 500@xmath0100 myr coma berenices open cluster ( melotte 111 ) , called 1216 + 260 , has been studied by dobbie et al .
( 2009a ) . based on assuming a co core and a thick h layer for this da
, they calculate the parameters listed in table 2 . for the praesepe and
hyades clusters , likely members of the hyades `` moving group '' ( eggen 1958 , zuckerman & song 2004 ) , we adopt as the age 625@xmath050 myr .
the white dwarfs from the hyades were studied by claver et al .
( 2001 ) and ferrario et al .
the only db star among the cluster white dwarfs used here is 0437 + 138 in the hyades .
the praesepe cluster stars were studied by claver et al .
( 2001 ) , ferrario et al .
( 2005 ) , dobbie et al .
( 2004 ) , and casewell et al .
( 2009 ) . for objects in common between these last two papers with many common authors ,
we adopt the values of the last paper , since the results were obtained with the uves spectrograph on the vlt ut2 .
we omit 0837 + 218 since casewell et al .
( 2009 ) make a good case that this is not a cluster member .
white dwarfs in older clusters overlapping and exceeding the likely age of procyon have been analyzed and published since the ferrario et al . work . for the older clusters ( as well as procyon b and sirius b ) , the errors for @xmath44 are relatively smaller .
kalirai et al .
( 2008 ) identified four likely members of the open cluster ngc 7789 , with age [email protected] gyr , and found two likely member white dwarfs in ngc 6819 with age [email protected] gyr .
we note that the signal - to - noise ratio of the spectra with fitted balmer lines are low compared to p. bergeron s usual standards ; we quote the listed errors here .
kalirai et al .
( 2007 ) analyzed a number of the brightest white dwarfs in the very old galactic cluster ngc 6791 ( [email protected] gyr ) ; however , only 6791 - 7 was the only likely member not having a very low mass determination and likely helium core .
the authors note that enhanced mass loss in the red giant phase , which is the likely origin of the helium white dwarfs , probably also occurred in this object , despite the fact that it retained a hydrogen atmosphere
. finally , kalirai et al . ( 2009 ) analyze a number of white dwarfs at the tip of the observed sequence in the globular cluster messier 4 .
this is of course a population ii system with [ fe / h ] = @[email protected] ( mucciarelli et al .
2011 ) , and likely age of [email protected] gyr ( hansen et al .
the basic conclusion is that the @[email protected] @xmath3 stars now producing white dwarfs which have evolved to approximately [email protected] @xmath3 ( but we list the individual determination for each of the latter ) .
procyon b has a final mass below nearly all others of similar initial mass .
it is one of only two white dwarfs here with a helium atmosphere .
the other is the db in the hyades mentioned earlier , which has a mass similar to the other da stars in that cluster and the praesepe .
the dashed line plots the simple linear relation from ferrario et al .
( 2005 ) based on a fit to white dwarfs in more massive clusters available at that time .
the procyon b point lies more than 0.1 @xmath3 or several @xmath4 below this line in the sense of having a lower final mass .
we should note that the close , cool double - dc binary g 107 - 70ab ( harrington , christy , & strand 1981 ) appears to be a similar situation at first glance .
it was barely resolved from ground - based imaging at the u.s .
naval observatory , flagstaff station .
it is no surprise that this old result is superceded by 13 years of hst imaging ( 1995.8 - 2008.8 ) , as of the latest fit
schaefer et al . ( 2006 , and private communication ) .
the preliminary period is [email protected] years , @xmath21 = 0.666 `` , and @xmath48 = [email protected] @xmath3 ( using the hipparcos parallax of [email protected] '' ) . since only a small difference in magnitude between the two components
is indicated , this suggests two components each of not quite 0.6 @xmath3 .
using the integrated light from the two stars , bergeron , ruiz , & leggett ( 1997 ) estimate 4900 k , log @xmath27 = 7.35 , and their broad - band @xmath49 energy distribution fits an h - rich atmosphere .
that is , the spectrum would have been type da if the temperatures of the two stars were warmer .
the hst astrometry above suggests two white dwarfs each of mass near the well known peak of the da mass distribution ( koester , schulz , & weidemann 1979 ) .
note that the log @xmath27 value inferred above is incorrect , since the method used by bergeron et al .
( 1997 ) to estimate a surface gravity assumes a single star .
thus g 107 - 70ab are not cooler analogs of procyon b. we have no explanation for why the procyon b remnant appears so undermassive vs the initial mass we have derived .
suppose that virtually all of the helium envelope of the procyon b nuclear progenitor could have been been lost in the event ( late helium - shell flash ? ) that also removed all of the hydrogen .
however , stellar models of the asymptotic giant branch phase show that the maximum helium envelope mass remaining would have been only @xmath280.01@xmath3 if left intact .
thus the loss of the entire helium envelope fails to account for the total mass shortfall by an order of magnitude .
rather , the evolution of the ( presumably ) carbon - oxygen core was evidently truncated before it could build up to a more normal mass for a single star with the indicated initial mass .
the dilemma suggests that close binary evolution may have been involved , but the existing binary consists of two stars in an elliptical orbit of long period .
hence , we are left with an interesting problem in ( binary ) stellar evolution .
note that there are still rather few stars analyzed from population i clusters of age older than or similar to procyon . in the future the numerous white dwarfs in the old disk clusters m 67 and ngc 188 will make useful additions to round out the low mass end of such a diagram .
we gratefully acknowledge gail schaefer of the chara array at georgia state university , and howard e. bond of stsci , for allowing us to quote preliminary hst results which fix the masses of procyon a and b , and for results on g 107 - 70ab .
this work was originally supported by the national science foundation through grant ast-0307321 ( jl and kaw ) .
gf wishes to acknowledge the contribution of the canada research chair program .
cccccccc & & & & & 7.9676 & 0.012777 & 1.1069 + 1 & c / o & @xmath252.5 & 0.001 & 0.553 & 7.9681 & 0.012768 & 1.1870 + & & & & & 7.9687 & 0.012760 & 1.2772 + & & & & & 7.9297 & 0.013077 & 1.0516 + 2 & c / o & @xmath252.5 & 0.001 & 0.531 & 7.9303 & 0.013068 & 1.1274 + & & & & & 7.9309 & 0.013059 & 1.2122 + & & & & & 8.0045 & 0.012485 & 1.1650 + 3 & c / o & @xmath252.5 & 0.001 & 0.575 & 8.0050 & 0.012479 & 1.2501 + & & & & & 8.0055 & 0.012471 & 1.3438 + & & & & & 7.9637 & 0.012834 & 1.2336 + 4 & c & @xmath252.5 & 0.001 & 0.553 & 7.9642 & 0.012826 & 1.3236 + & & & & & 7.9648 & 0.012817 & 1.4251 + & & & & & 7.9709 & 0.012728 & 1.0992 + 5 & o & @xmath252.5 & 0.001 & 0.553 & 7.9714 & 0.012720 & 1.1770 + & & & & & 7.9720 & 0.012712 & 1.2650 + & & & & & 7.9730 & 0.012696 & 1.2962 + 6 & c / o & @xmath252.5 & 0.0 & 0.553 & 7.9736 & 0.012688 & 1.3914 + & & & & & 7.9742 & 0.012680 & 1.4990 + & & & & & 7.9693 & 0.012752 & 1.4500 + 7 & c & @xmath252.5 & 0.0 & 0.553 & 7.9699 & 0.012743 & 1.5573 + & & & & & 7.9705 & 0.012734 & 1.6774 + & & & & & 7.9762 & 0.012651 & 1.2634 + 8 & o & @xmath252.5 & 0.0 & 0.553 & 7.9767 & 0.012643 & 1.3544 + & & & & & 7.9772 & 0.012635 & 1.4565 + & & & & & 7.9661 & 0.012798 & 1.0705 + 9 & c / o & @xmath252.0 & 0.001 & 0.553 & 7.9667 & 0.012790 & 1.1488 + & & & & & 7.9673 & 0.012781 & 1.2374 + & & & & & 7.9687 & 0.012756 & 1.1521 + 10 & c / o & @xmath253.0 & 0.001 & 0.553 & 7.9692 & 0.012747 & 1.2347 + & & & & & 7.9698 & 0.012739 & 1.3266 + & & & & & 8.0485 & 0.012145 & 1.3120 + 11 & c / o & @xmath252.5 & 0.001 & 0.602 & 8.0489 & 0.012138 & 1.4063 + & & & & & 8.0494 & 0.012132 & 1.5112 procyon b & 0.553 & 0.022 & 9.075 & 0.074 & 2.59 & 0.21,0.40 &
0.16,0.20 & 1 + sirius b & 1.000 & 0.020 & 8.091 & 0.020 & 5.056 & 0.171,0.262 & 0.213,0.273 & 2 + lb 1497 & 1.023 & 0.026 & 7.781 & 0.062 & 6.542 & 0.346,0.458 & 0.866,1.614 & 3 + gd 50 & 1.264 & 0.017 & 7.785 & 0.045 & 6.3 & 2.5,0.8 & , & 4,5 + ngc 2516 - 1 & 0.931 & 0.098 & 7.760 & 0.230 & 5.411 & 1.017,0.500 & 0.524,0.386 & 6 + ngc 2516 - 2 & 1.004 & 0.058 & 7.621 & 0.179 & 5.096 & 0.406,0.239 & 0.415,0.311 & 6 + ngc
2516 - 3 & 0.969 & 0.039 & 7.922 & 0.079 & 6.141 & 0.692,0.427 & 0.924,0.587 & 6 + ngc 2516 - 5 & 1.054 & 0.063 & 7.883 & 0.136 & 5.902 & 0.986,0.524 & 0.786,0.514 & 6 + ngc 2168 - 1 & 0.873 & 0.091 & 7.228 & 0.260 & 4.39 & 0.23,0.09 & 0.35,0.27 & 7 + ngc 2168 - 2 & 1.015 & 0.067 & 7.657 & 0.202 & 4.79 & 0.47,0.26 & 0.46,0.36 & 7 + ngc 2168 - 5 & 0.916 & 0.075 & 6.103 & 0.088 & 4.21 & 0.00,0.00 & 0.29,0.23 & 7 + ngc
2168 - 6 & 0.877 & 0.096 & 6.077 & 0.082 & 4.21 & 0.00,0.00 & 0.29,0.23 & 7 + ngc 2168 - 11 & 0.802 & 0.096 & 8.047 & 0.173 & 6.63 & 0.00,1.38 & 1.89,0.96 & 7 + ngc 2168 - 12 & 0.922 & 0.092 & 7.314 & 0.281 & 4.43 & 0.30,0.13 & 0.36,0.27 & 7 + ngc 2168 - 14 & 1.010 & 0.072 & 7.865 & 0.167 & 5.32 & 0.95,0.44 & 0.73,0.48 & 7 + ngc 2168 - 15 & 0.888 & 0.088 & 7.551 & 0.236 & 4.63 & 0.38,0.22 & 0.41,0.32 & 7 + ngc 2168 - 27 & 1.022 & 0.072 & 7.840 & 0.170 & 5.22 & 0.71,0.42 & 0.68,0.45 & 7 + ngc 2168 - 29 & 0.882 & 0.078 & 7.331 & 0.235 & 4.44 & 0.23,0.12 & 0.36,0.28 & 7 + ngc 2168 - 30 & 1.011 & 0.122 & 7.808 & 0.289 & 5.12 & 1.49,0.55 & 0.63,0.43 & 7 + ngc 2287 - 2 & 0.910 & 0.040 & 7.908 & 0.075 & 4.450 & 0.580,0.380 & , & 8 + ngc 2287 - 5 & 0.910 & 0.040 & 7.964 & 0.071 & 4.570 & 0.640,0.430 & , & 8 + ngc 3532 - 1 & 0.86 & 0.04 & 7.778 & 0.095 & 3.830 & 0.18,0.15 & , & 9 + ngc 3532 - 5 & 0.820 & 0.04 & 7.580 & 0.115 & 3.71 & 0.150,0.130 & , & 9 + ngc 3532 - 9 & 0.760 & 0.04 & 7.00 & 0.113 & 3.57 & 0.12,0.11 & , & 9 + ngc 3532 - 10 & 0.96 & 0.04 & 8.173 & 0.060 & 4.58 & 0.47,0.33 & , & 9 + ngc 2099 - 2 & 0.69 & 0.10 & 7.961 & 0.165 & 2.76 & 0.08,0.05 & 0.162,0.201 & 10 + ngc 2099 - 3 & 0.76 & 0.13 & 8.190 & 0.190 & 2.92 & 0.16,0.14 & 0.197,0.258 & 10 + ngc 2099 - 4 & 0.87 & 0.15 & 8.428 & 0.217 & 3.02 & 0.26,0.19 & 0.324,0.495 & 10 + ngc 2099 - 5 & 0.83 & 0.14 & 8.271 & 0.188 & 3.02 & 0.26,0.19 & 0.219,0.305 & 10 + ngc 2099 - 7 & 0.88 & 0.19 & 8.378 & 0.289 & 3.26 & 0.78,0.40 & 0.278,0.405 & 10 + ngc 2099 - 9 & 0.61 & 0.05 & 8.268 & 0.067 & 2.97 & 0.06,0.07 & 0.218,0.302 & 10 + ngc 2099 - 10 & 0.74 & 0.04 & 8.090 & 0.068 & 2.81 & 0.04,0.02 & 0.178,0.225 & 10 + ngc 2099 - 11 & 0.96 & 0.06 & 8.149 & 0.095 & 2.86 & 0.07,0.07 & 0.188,0.242 & 10 + ngc 2099 - 12 & 0.55 & 0.07 & 8.394 & 0.126 & 3.30 & 0.30,0.22 & 0.290,0.429 & 10 + ngc 2099 - 13 & 0.79 & 0.05 & 8.229 & 0.074 & 2.93 & 0.06,0.07 & 0.207,0.278 & 10 + ngc 2099 - 14 & 0.45 & 0.08 & 8.487 & 0.085 & 3.31 & 0.20,0.26 & 0.400,0.674 & 10 + ngc 2099 - 16 & 0.83 & 0.06 & 8.689 & 0.082 & 5.20 & 7.0,1.20 & 3.737, & 10 + 1216 + 260 & 0.90 & 0.04 & 8.560 & 0.052 & 4.77 & 0.97,5.37 & , & 11 + 0352 + 098 & 0.719 & 0.030 & 8.270 & 0.046 & 3.094 & 0.045,0.052 & 0.114,0.134 & 12 + 0406 + 169 & 0.806 & 0.031 & 8.488 & 0.046 & 3.465 & 0.109,0.147 & 0.172,0.225 & 12
+ 0421 + 162 & 0.680 & 0.031 & 7.970 & 0.055 & 2.892 & 0.020,0.023 & 0.090,0.103 & 12 + 0425 + 168 & 0.705 & 0.031 & 7.549 & 0.077 & 2.789 & 0.011,0.010 & 0.079,0.088 & 12 + 0431 + 125 & 0.652 & 0.032 & 7.752 & 0.068 & 2.825 & 0.014,0.017 & 0.083,0.093 & 12 + 0437 + 138 & 0.740 & 0.060 & 8.470 & 0.047 & 3.663 & 0.030,0.030 & , & 13 + 0438 + 108 & 0.684 & 0.031 & 7.203 & 0.060 & 2.757 & 0.003,0.004 & 0.076,0.084 & 12 + 0833 + 194 & 0.721 & 0.043 & 8.431 & 0.049 & 3.32 & 0.33,0.22 & , & 14 + 0836 + 197 & 0.909 & 0.030 & 8.136 & 0.049 & 2.981 & 0.031,0.035 & 0.101,0.115 & 15
+ 0836 + 199 & 0.831 & 0.037 & 8.612 & 0.069 & 3.997 & 0.350,0.652 & 0.297,0.428 & 15 + 0836 + 199 & 0.752 & 0.044 & 8.489 & 0.049 & 3.46 & 0.43,0.27 & , & 14 + 0836 + 201 & 0.620 & 0.031 & 8.154 & 0.050 & 2.993 & 0.028,0.046 & 0.102,0.117 & 15 + 0837 + 185 & 0.804 & 0.044 & 8.504 & 0.050 & 3.50 & 0.48,0.29 & , & 14 + 0837 + 199 & 0.819 & 0.032 & 8.351 & 0.048 & 3.194 & 0.081,0.062 & 0.153,0.129 & 15 + 0837 + 199 & 0.737 & 0.043 & 8.253 & 0.052 & 3.07 & 0.20,0.15 & , & 14 + 0840 + 190 & 0.849 & 0.045 & 8.566 & 0.050 & 3.73 & 0.71,0.39 & , & 14 + 0840 + 200 & 0.761 & 0.033 & 8.522 & 0.045 & 3.572 & 0.186,0.129 & 0.255,0.196 & 15
+ 0840 + 200 & 0.721 & 0.043 & 8.420 & 0.048 & 3.30 & 0.32,0.21 & & 14 + 0843 + 184 & 0.823 & 0.045 & 8.530 & 0.051 & 3.59 & 0.55,0.33 & , & 14 + ngc 7789 - 4 & 0.560 & 0.020 & 8.061 & 0.032 & 2.080 & 0.080,0.080 & , & 16 + ngc 7789 - 5 & 0.600 & 0.030 & 6.903 & 0.055 & 2.020 & 0.070,0.140 & , & 16 + ngc 7789 - 6 & 0.720 & 0.030 & 8.204 & 0.045 & 2.100 & 0.090,0.090 & , & 16 + ngc 7789 - 8 & 0.640 & 0.040 & 7.462 & 0.080 & 2.020 & 0.090,0.110 & , & 16 + ngc 6819 - 6 & 0.530 & 0.020 & 7.591 & 0.035 & 1.600 & 0.060,0.050 & , & 17 + ngc 6819 - 7 & 0.560 & 0.020 & 8.155 & 0.034 & 1.620 & 0.070,0.050 & , & 17 + ngc 6791 - 7 & 0.530 & 0.020 & 8.176 & 0.060 & 1.160 & 0.040,0.030 & , & 18 + m 4 - 0 & 0.520 & 0.040 & 7.602 & 0.070 & 0.800 & 0.050,0.050 & , & 19 + m 4 - 4 & 0.500 & 0.030 & 7.279 & 0.024 & 0.800 & 0.050,0.050 & , & 19 + m 4 - 6 & 0.590 & 0.040 & 7.255 & 0.050 & 0.800 & 0.050,0.050 & , & 19 + m 4 - 15 & 0.550 & 0.040 & 7.322 & 0.043 & 0.800 & 0.050,0.050 & , & 19 + m 4 - 20 & 0.510 & 0.050 & 7.690 & 0.075 & 0.800 & 0.050,0.050 & , & 19 + m 4 - 24 & 0.510 & 0.030 & 7.204 & 0.028 & 0.800 & 0.050,0.050 & , & 19 + asplund , m. , grevesse , n. , & sauval , a.j .
2005 , in _ cosmic abundances as records of stellar evolution and nucleosynthesis in honor of david l. lambert _ , asp .
series , v. 336 , eds .
t. g. barnes iii & f. n. bash ( san francisco : astronomical society of the pacific ) p. 25
lodders , k. 2010 , `` solar system abundances of the elements '' , in _ principles and perspectives in cosmochemistry , astrophysics & space sciences proceedings _ , vol .
isbn 978 - 3 - 642 - 10351 - 3 ( berlin heidelberg : springer - verlag ) , p. 379 fig . 1
tycho evolutionary tracks in a log luminosity vs. log radius diagram beginning including the pre - main sequence at the asplund , grevesse , & sauval ( 2005 , 2009 = a09 ) primordial solar abundance of z=0.0122 for masses of ( in descending order of luminosity ) 1.499 @xmath3 ( black ) , the favored preliminary hst astrometric value , 1.48 @xmath3 ( red ) , the best fit to the observed luminosity , 1.463 @xmath3 ( green ) , and 1.42 @xmath3 ( blue ) .
fig . 2 the initial to final mass relation ( ifmr ) for procyon b , sirius b and the white dwarfs in a number of star clusters discussed in the text .
procyon b and sirius b ( from paper i ) are plotted as black , filled circles .
the total data set for older open clusters from kalirai et al .
( 2007 , 2008 , 2009 ) are also plotted with filled symbols
ngc 7789 ( red ) , ngc 6819 ( green ) , ngc 6791 ( blue ) and messier 4 ( magenta ) . shown with open circles taken from ferrario et al .
( 2005 ) and paper i are the pleiades ( red ) , ngc 2516 ( cyan ) , ngc 2168 ( magenta ) , the hyades ( blue ) and praesepe ( cyan ) , ngc 3532 ( green ) , and ngc 2099 ( yellow ) .
the error bars include the best estimates of the cluster ages .
the magenta line plots the simple linear relation from ferrario et al .
the procyon b remnant mass appears low relative to the others , as discussed in the text . | the procyon ab binary system ( orbital period 40.838 years , a newly - refined determination ) , is near and bright enough that the component radii , effective temperatures and luminosities are very well determined , although more than one possible solution to the masses has limited the claimed accuracy .
preliminary mass determinations for each component are available from hst imaging , supported by ground - based astrometry and an excellent _ hipparcos _ parallax ; we use these for our preferred solution for the binary system .
other values for the masses are also considered .
we have employed the tycho stellar evolution code to match the radius and luminosity of the f5 iv - v primary star to determine the system s most likely age as [email protected] gyr . since prior studies of procyon a found its abundance indistinguishable from solar , the solar composition of asplund , grevesse & sauval ( @xmath1=0.014 ) is assumed for the hr diagram fitting . an unsuccessful attempt to fit using the older solar abundance scale of grevesse & sauval ( @xmath1=0.019 ) is also reported .
for procyon b , eleven new sequences for the cooling of non - da white dwarfs have been calculated , to investigate the dependences of the cooling age on ( 1 ) the mass , ( 2 ) the core composition , ( 3 ) the helium layer mass , and ( 4 ) heavy - element opacities in the helium envelope .
our calculations indicate a cooling age of [email protected] gyr , which implies that the progenitor mass of procyon b was 2.59@xmath2 @xmath3 . in a plot of initial vs final mass of white dwarfs in astrometric binaries or star clusters ( all with age determinations ) , the procyon b final mass lies several @xmath4 below a straight line fit . |
one of the least tested sectors of the standard model of elementary particle physics is the higgs sector generating the masses of all particles via the mechanism of spontaneous symmetry breaking .
this mechanism implies that the self - coupling @xmath6 of the higgs particle is proportional to the square of its mass , @xmath3 .
hence a heavy higgs particle may cause the breakdown of perturbation theory in @xmath4 .
when calculating decay rates or cross sections beyond the tree level term of the perturbation series one must specify the _
renormalization scheme _ to define the coupling constants and the particle masses appearing in the analytic expressions . in the higgs sector one usually adopts the _ on - mass - shell scheme _ ( oms ) : here the squared mass coincides with the physical pole of the propagator , the vacuum expectation value @xmath7 gev of the higgs field is renormalized such as to cancel tadpole contributions , and the coupling is chosen to satisfy @xmath8 to all orders in perturbation theory .
is hence taken to be zero . ] here @xmath9 is the fermi constant . when discussing the perturbation series of physical observables one must first distinguish processes involving two largely separated mass scales from those containing only one scale , @xmath3 .
prototypes of the first species are cross sections at lhc energies .
they contain potentially large logarithmic terms @xmath10 with @xmath11 being the energy of the process .
these terms may spoil the smallness of radiative corrections . yet
they can be summed to all orders in perturbation theory with the help of _ renormalization group _ ( rg ) methods .
the corresponding equation in the oms is the _ callan - symanzik equation _ , which can be solved in the limit of large @xmath11 when all particle masses can be neglected .
the second kind of physical observables are one - scale processes .
examples are two - body decay rates of the higgs particle into ( almost ) massless particles .
we will see in the following that rg methods are also a useful tool to judge the accuracy of perturbative results in these cases .
due to ( [ defloms ] ) the use of the oms appears natural , because @xmath12 is directly related to physical observables .
yet it is also useful to consider mass independent renormalization schemes , the most prominent example being the @xmath0-scheme .
the reasons are the following : * mass independent schemes allow for an exact solution of the rg equations , i.e. mass effects can be systematically included . *
the analysis of scheme dependences provides a test of the reliability of perturbation theory since the results obtained to order @xmath13 in different schemes formally differ by terms of order @xmath14 .
* results obtained in mass independent schemes involve an arbitrary parameter , the _ renormalization scale _ @xmath15 .
for perturbation theory to work it is necessary that the dependence of the result on @xmath15 diminishes order by order in perturbation theory .
we will use this fact extensively in the discussion of the breakdown of perturbation theory .
the paper is organized as follows : in the following section we discuss the running coupling constant in the two - loop approximation . in sect .
[ sect : scheme ] we give the two - loop relation between the quartic higgs coupling in the oms and the @xmath0 scheme .
we calculate the three - loop oms coefficient of the @xmath2 function as well as the coefficients of the leading and next - to - leading logarithms .
subsequently , we investigate the scheme and scale dependence of two - loop physical observables using the three - loop running coupling . special attention is paid to the breakdown of perturbation theory for heavy higgs masses .
we start by examining the bosonic and fermionic higgs decay widths ( sect . [
sect : scale ] ) . finally , we look at scattering processes which involve the longitudinally polarized gauge bosons @xmath16 and @xmath17 .
these scattering processes give the most stringent bounds on a perturbative higgs mass ( sect .
[ sect : cross ] ) .
we first discuss the coupling constant @xmath18 , whose running is encoded in the @xmath2-function . to three loops ,
the beta function of the higgs quartic coupling is defined as @xmath19 we neglect all contributions from gauge and yukawa couplings .
this is an excellent approximation for large values of @xmath4 . to two loops the coefficients are @xcite
@xmath20 the three - loop coefficient @xmath21 is scheme dependent .
we restrict the discussion of this section to the scheme independent two - loop results and return to @xmath21 in sect .
[ sect : scheme ] .
equation ( [ beta ] ) is valid in any mass independent scheme with @xmath15 being the renormalization scale accompanying the coupling constant in the lagrangian . in the oms scheme ,
the callan - symanzik equation describes the response of some green s function to the scaling of its external momenta according to @xmath22 which is related to the corresponding scaling of the coupling given by eq .
( [ beta ] ) .
the values of @xmath23 and @xmath24 are scheme independent , so that we do nt choose a specific renormalization scheme until later . the determination of @xmath25 proceeds in two steps : first at some initial scale @xmath26 the coupling @xmath27 is obtained from ( [ defloms ] ) or an equivalent relation , if the scheme under consideration is not the oms scheme .
scheme , the expression for @xmath27 is derived in section [ sect : scheme ] , see eq .
( [ coup ] ) . ] then ( [ beta ] ) is solved for @xmath28 $ ] .
the question whether perturbation theory works in a specific physical process therefore depends on the two parameters @xmath29 and @xmath30 , which are related to the physical parameters @xmath3 and the energy @xmath11 of the process .
the cause of a possible breakdown of perturbation theory is twofold : first , the larger the higgs mass , @xmath3 , the larger is @xmath29 due to ( [ defloms ] ) .
second , the resummation of possibly large logarithms , @xmath31 , introduces the running coupling @xmath32 with @xmath33 , and @xmath32 increases for increasing @xmath34 . defining @xmath4 at some initial scale @xmath35 , the solution of ( [ beta ] ) yields @xmath1 at any other scale @xmath15 .
the expansion of @xmath36 around @xmath37 shows that @xmath1 resums powers of the logarithm @xmath38 multiplied by powers of @xmath39 .
the coefficients of the leading logarithms ( ll ) , @xmath40 , depend only on @xmath23 , those of the next - to - leading logarithms ( nll ) depend on both @xmath24 and @xmath23 , and so on .
we will give these coefficients explicitly later .
the first point we want to stress is that beyond the one - loop approximation there are several different solutions of eq .
( [ beta ] ) due to the truncation of the perturbation series .
all of them agree to the order at which the perturbative series of the beta function , ( [ beta ] ) , is truncated , but they differ by terms of the neglected order . for a viable perturbative treatment , these higher - order differences should be negligible .
we now look at four different solutions for the running coupling @xmath1 .
setting simply the coefficients of the neglected terms in ( [ beta ] ) equal to zero and integrating the remaining expressing exactly , we obtain an implicit equation for @xmath1 : @xmath41 in the nll approximation , and @xmath42 \label{analint2}\end{aligned}\ ] ] with @xmath43 in the next - to - next - to - leading log ( nnll ) approximation . , eq . ( [ analint2 ] ) reduces to ( [ analint ] ) , of course . ] here ( [ analint ] ) is the two - loop result published previously @xcite . in ( [ analint ] ) and ( [ analint2 ] ) the useful abbrevation @xmath44 has been used . to obtain @xmath45
one has to solve ( [ analint ] ) or ( [ analint2 ] ) by numerical methods .
direct numerical integration of the original equation , ( [ beta ] ) , yields the same result for @xmath1 .
let us call this form the _ naive _ solution
. it would be the exact result if the neglected coefficients of the @xmath2-function were really identical to zero . yet
( [ analint ] ) and ( [ analint2 ] ) contain logarithmic terms belonging to the neglected higher orders of rg improved perturbation theory .
e.g. , the naive nll result for @xmath18 , ( [ analint ] ) , partially resums nnll logarithms @xmath46 . however , there are further nnll terms from irreducible three - loop contributions
. therefore , the naive nll result contains an inconsistent resummation of nnll terms .
similarly , eq . ( [ analint2 ] ) includes an inconsistent resummation of higher - order terms , even though it is correct to the order considered .
this fact is true whenever beta functions are integrated numerically at and beyond two loops . for a _
consistent _ result , one needs to ignore terms of the neglected order in @xmath4 when integrating ( [ beta ] ) .
this yields the consistent nnll result @xmath47{0pt}{40pt } \frac{\beta_1 ^ 2- \beta_0 \beta_2 } { \beta_0 ^ 2 } \hat{\lambda}(\mu_0 ) \left [ { \hat{\lambda } } ( \mu ) - { \hat{\lambda } } ( \mu_0 ) \right ] } \ , .
\label{consist2lp}\end{aligned}\ ] ] the consistent nll result is obtained from ( [ consist2lp ] ) by dropping the underlined terms . the final result for the consistent solution
is obtained by solving eq .
( [ consist2lp ] ) numerically .
third we solve ( [ consist2lp ] ) iteratively by first substituting the one - loop result for @xmath1 into the rhs of the equation above and then repeating this step with the result of the first substitution .
this yields the _ iterative _ answer @xmath48{0pt}{40pt } 1 -\beta_0 { \hat{\lambda } } ( \mu_0 ) \ln \left ( \frac{\mu}{\mu_0 } \right ) \right .
\nonumber \\ & & + \frac{\beta_1}{\beta_0}{\hat{\lambda } } ( \mu_0 ) \;\ln\ ! \left[\ ; 1 -\beta_0 { \hat{\lambda } } ( \mu_0 ) \ln \frac{\mu}{\mu_0 } \underline{\rule[-17pt]{0pt}{40pt } + \frac{\beta_1}{\beta_0 } { \hat{\lambda } } ( \mu_0 ) \ln \left ( 1- \beta_0 { \hat{\lambda } } ( \mu_0 ) \ln \frac{\mu}{\mu_0 } \right ) } \ ; \right ] \nonumber \\ & & \left .
\underline { \rule[-17pt]{0pt}{40pt } + \frac{\beta_1 ^ 2-\beta_0 \beta_2}{\beta_0 } \ : { \hat{\lambda } } ^3 ( \mu_0 ) \ln \frac{\mu } { \mu_0 } \cdot \left ( 1- \beta_0 { \hat{\lambda } } ( \mu_0 ) \:\ln \frac{\mu}{\mu_0 } \right)^{-1 } } \right\ } ^{-1 } .
\label{pert2lp}\end{aligned}\ ] ] we stress that no further expansions in @xmath49 are possible , because each @xmath49 multiplies a large logarithm @xmath50 .
( [ pert2lp ] ) is the nnll result , and the nll expression is again obtained by dropping the underlined terms . the fourth solution of ( [ beta ] ) is constructed in analogy to qcd : the integration constant obtained in the integration of ( [ beta ] ) can be absorbed into a scale parameter @xmath51 : @xmath52 } { \ln(\frac{\lambda_h^2}{\mu^2 } ) } \underline{\rule[-21pt]{0pt}{40pt } + \frac{4 \beta_1 ^ 2}{\beta_0 ^ 4 } \frac{\ln^2 [ \ln \frac{\lambda_h^2}{\mu^2 } ] } { \ln^2(\frac{\lambda_h^2}{\mu^2 } ) } - 4 \frac{\beta_1 ^ 2}{\beta_0 ^ 4 } \frac{\ln [ \ln ( \frac{\lambda_h^2}{\mu^2 } ) ] } { \ln^2 ( \frac{\lambda_h^2}{\mu^2 } ) } }
\right.\nonumber \\ & & \;\ ; \left.\underline{\rule[-21pt]{0pt}{40pt } + 4 \frac { \beta_0 \beta_2 - \beta_1 ^ 2}{\beta_0 ^ 4 } \frac{1}{\ln^2(\frac{\lambda_h^2}{\mu^2 } ) } } + { \rm o}\biggl(\frac{\ln^3[\ln ( \frac{\lambda_h^2}{\mu^2 } ) ] } { \ln^3(\frac{\lambda_h^2}{\mu^2})}\biggr ) \right ] .
\label{lamdef}\end{aligned}\ ] ] we call this result the _ qcd - like _ solution .
the above equation defines the scale parameter @xmath51 , and it is written such that no term of the form @xmath53 appears in the square brackets .
this is identical to the definition of the qcd scale parameter @xmath54 @xcite .
the definition of the nll parameter @xmath55 is obtained by dropping the underlined terms . the above definition of @xmath51 works in any renormalization scheme .
the actual numerical evaluation of @xmath51 depends on the boundary value @xmath29 , which is scheme dependent .
if for example @xmath1 on the lhs of ( [ lamdef ] ) is given in the @xmath56-scheme , @xmath51 on the rhs equals @xmath57 .
there are two important differences between the higgs sector and qcd : first , eq .
( [ lamdef ] ) holds for @xmath58 , while in an asymptotically free theory like qcd the analogue of ( [ lamdef ] ) is valid for @xmath59 .
second , @xmath60 is the only fundamental scale parameter of qcd ( with massless quarks ) , while in the higgs sector @xmath51 is related to the higgs mass ( see table [ lamtable ] ) . yet
if the higgs mass turns out to be large , one will have to parametrize non - perturbative effects in terms of an effective lagrangian , and @xmath51 will be the natural scale entering the effective couplings . in qcd perturbation theory
breaks down for @xmath61 and binding energies equal a few times @xmath60 . in the higgs sector
the breakdown of perturbation theory likewise occurs for @xmath62 . in table
[ lamtable ] we tabulate the relation between @xmath3 and @xmath51 .
when the oms scheme is adopted , @xmath3 in table [ lamtable ] is the physical mass ( see ( [ defloms ] ) ) . in other schemes the tabulated values for @xmath3 correspond to the tree - level relation @xmath63 and
in contrast to the oms scheme radiative corrections to this relation have to be taken into account in order to obtain the physical higgs mass from @xmath64 .
for the @xmath0-scheme this relation is encoded in ( [ coup ] ) of sect .
[ sect : scheme ] .
@xmath65 we now completed the definition of the four different solutions for the running higgs coupling : naive , consistent , iterative , and qcd - like solution .
it is interesting to note that at one loop all four solutions are identical . in the four different solutions . ]
this is clearly not the case anymore at two - loop and higher orders , and we will discuss the differences below .
let us stress again that _ all _ solutions correctly sum the large logarithm @xmath66 within the calculated order .
the difference between these solutions is of the neglected order as can be seen when expanding the solutions in @xmath27 .
if perturbation theory is applicable this difference should be numerically small , giving a very simple criterion to find the values of @xmath3 and @xmath15 beyond which perturbation theory clearly fails . in fig .
[ rgecurves ] , we compare the @xmath15 dependence of the four solutions at one and two loops .
we choose the three mass values @xmath67 , @xmath68 , and @xmath69 gev . to obtain a meaningful comparison between the one- and two - loop results for a given value of @xmath3
, we take for both orders the same expansion parameter @xmath70 , and choose it to be defined by the tree - level relationship @xmath71 .
this has the additional advantage that we do not yet have to specify a renormalization scheme in fig . [ rgecurves ] .
since the first two coefficients of the @xmath2-function are scheme independent , the whole scheme dependence resides in the relation between the tree - level mass ( which labels the different curves in fig . [ rgecurves ] ) and the physical higgs mass . in the oms , the two mass definitions coincide .
= 4.8 in at one loop , the four solutions are identical ( long dashed curve ) .
the coupling @xmath72 approaches infinity as @xmath73 , which is usually referred to as the landau pole @xcite . at two loops the breakdown of perturbation theory is clearly visible in the two - loop curves for @xmath74 gev .
it manifests itself in a very different behaviour of the four solutions , although they all are defined to have the same value for @xmath75 .
the one - loop landau pole has completely vanished from all two - loop solutions .
the naive ( long - short dashes in fig .
[ rgecurves ] ) and consistent ( solid ) solutions always have positive slopes in accordance with the positiveness of the perturbative @xmath2-function , eq . ( [ beta ] ) .
the consistent solution can be looked at as the `` average '' of the one - loop solution and the two - loop naive solution .
the naive solution approaches the fix - point value @xmath76 , the zero of the two - loop beta function .
this simply reflects the fact that one can not use perturbation theory to gain information about large values of the coupling .
this is even more evident from the other two solutions .
the qcd - like result for @xmath77 has a global maximum .
interestingly enough , for values of @xmath78 gev , the value @xmath79 is larger than the maximal possible value of the qcd - like running coupling , so that @xmath80 can no longer be satisfied .
the unphysical maximum of the qcd - like solution appears for @xmath81 , and perturbation breaks down well before @xmath15 is equal to @xmath51 .
the iterative solution features a maximum at @xmath82 which is located at @xmath83 @xmath84 gev for @xmath74 @xmath85 gev .
the two - loop iterative solution approaches zero at the same value of @xmath15 at which the one - loop solution has its landau pole .
the fact that some perturbative solutions of the two - loop running coupling show a maximum gives a first criterion for the breakdown of the two - loop perturbative treatment . from the definition of the @xmath2 function , eq .
( [ beta ] ) , we know that in the perturbative regime the slope of the running coupling has to be positive as the one - loop term dominates over the two - loop term . if the iterative solution for @xmath1 has negative slope , then perturbation theory is not valid anymore .
hence eq .
( [ maxiter ] ) is possibly a measure of the range of @xmath15 for which perturbation theory is meaningful .
the previous discussion of the four solutions is valid for any value of @xmath3 . yet
if we restrict ourselves to values @xmath86 gev , features like poles and maxima are only apparent for @xmath74 gev , and not for @xmath68 and @xmath87 gev ( see fig .
[ rgecurves ] ) .
for the lower mass values one has to go to ( much ) larger values of @xmath15 to observe the breakdown of perturbation theory . in fact , if @xmath67 gev all four solutions of the running couping agree extremely well for @xmath88 gev .
choosing @xmath89 , however , already results in an uncertainty between the different two - loop solutions of about 25% at @xmath90 tev , and the one - loop solution is more than 50% larger than the two - loop solution at @xmath91 tev .
if the standard model higgs sector is to remain perturbative up to much higher energy scales the restrictions on @xmath3 get much more severe @xcite . to avoid the landau pole and keeping the coupling @xmath4 perturbative
, one can define an embedding scale below the landau pole and require the running coupling to be smaller than a certain value .
for example , taking the embedding scale to be @xmath92 gev the authors of @xcite implement the `` automatic fixing procedure '' requiring ( in our notation ) @xmath93 and find an upper bound of @xmath94 gev when neglecting the top - quark yukawa coupling . using eq .
( [ maxiter ] ) and requiring @xmath95 to be equal or larger than @xmath92 gev , we find the upper limit @xmath96 gev .
the analysis of the running coupling alone can clearly only yield a necessary criterion for the validity of perturbation theory .
taking @xmath74 gev , we can be sure that perturbation theory breaks down for any cross section with @xmath97 tev , because this energy is too close @xcite to the scale parameter @xmath51 and already larger than the corresponding maximum of @xmath15 , @xmath98 gev . for @xmath99 gev
, the perturbative rg treatment of the running coupling could be trustworthy up to values of @xmath100 tev , sufficient for studying higgs physics at the lhc .
nevertheless , choosing @xmath89 gev the running coupling @xmath101 is sizable . to make a final judgement on whether perturbation theory works for @xmath89 gev and tev - energies
, one must in addition investigate the perturbation series of the physical process of interest . as we will see later on , the perturbative solution of the running coupling may show a reasonable convergence , but the numerical value of the running coupling is already too large to calculate physical quantities like cross sections and decay widths in perturbation theory .
this will be the subject of sections [ sect : scale ] and [ sect : cross ] .
so far the discussion of the running coupling has been independent of a special renormalization scheme .
we have already listed the reasons for studying different renormalization schemes in the introduction . in the following
we will look at the @xmath0 and oms scheme , and examine the scheme dependence of the coupling and the three - loop coefficient @xmath21 in ( [ beta ] ) . before going into detail we would like to remark two points : first , the oms coupling @xmath103 is to all orders in @xmath4 directly related to measurable quantities ( the muon lifetime and the higgs mass ) via ( [ defloms ] ) .
this is not so for any other renormalization scheme , where the rhs of ( [ defloms ] ) receives radiative corrections , which depend on an additional parameter , the _ renormalization point _
second the scheme dependence of @xmath21 makes the coupling run differently in different schemes .
one can in general adjust the scheme such as to achieve any desired values for @xmath104 with @xmath105 . a criterion for a `` good scheme '' , however
, can not be founded on the smallness of the running coupling alone .
instead one has to consider physical observables , in which the coefficients of the perturbation series depend on the scheme as well .
this will be done in the following sections .
the starting point of the analysis is the two - loop relation between the bare and the renormalized coupling : @xmath106 \nonumber \\ & + & { \hat{\lambda } } ^3 \xi^{2 \epsilon } \left [ \;\frac{\beta_0 ^ 2}{4 } \frac{1}{\epsilon^2 } + \left ( \beta_0 c_1 + \frac{\beta_1}{4 } \right ) \frac{1}{\epsilon } + c_2 + \epsilon c_{21 } + o ( \epsilon^2 ) \right ] \nonumber\\ & + & { \hat{\lambda } } ^4 \xi^{3 \epsilon } \left [ \;\frac{\beta_0 ^ 3}{8 } \frac{1}{\epsilon^3 } + \left(\frac{3}{4 } \beta_0 ^ 2 c_1 + \frac{7}{24}\beta_0\beta_1 \right ) \frac{1}{\epsilon^2 } \right .
\nonumber\\ & & \left .
\phantom{{\hat{\lambda } } ^4 } \;+ \left(-\frac{5}{12 } \beta_0 ^ 2 c_{11 } + \frac{1}{3 } \beta_0 c_{1}^2 + \frac{7}{6 }
\beta_0 c_{2 } + \frac{7}{12}\beta_1 c_1 + \frac{1}{6}\beta_2 \right ) \frac{1}{\epsilon } + c_3 + o ( \epsilon ) \right ] \ , .
\label{z}\end{aligned}\ ] ] in the @xmath0-scheme the various quantities are given as @xmath107 in the oms scheme , they instead read @xcite : @xmath108 and @xmath109 , and @xmath110 are unknown . knowing the bare coupling to two loops in both the oms scheme and the @xmath56 scheme , we can express the @xmath0 coupling in terms of the oms coupling , and we can calculate the difference between @xmath111 and @xmath112 .
since the oms coupling is a function of @xmath3 ( or @xmath114 ) and is fixed to all orders by ( [ defloms ] ) , the definition of @xmath115 in terms of @xmath103 is equivalent to a relation between @xmath115 and @xmath3 . with @xmath116 and @xmath117 refering to the oms quantities in ( [ omscoeff ] ) we find : @xmath118 , \label{coup } \\ = { \hat\lambda}_{\rm oms}\biggl[1\ , \biggr . &
+ & \ , \left ( 12 \ln \frac{\mu_0 ^ 2}{m_h^2 } + 8.676\right ) { \hat\lambda_{\rm oms}}\ , \nonumber \\ & + & \ , \left ( 144\ln^2 \frac{\mu_0 ^ 2}{m_h^2 } + 52.22\ln \frac{\mu_0 ^ 2}{m_h^2 } + 286.8\right){\hat\lambda_{\rm oms}^2 } \nonumber\\ & + & \biggl.\ , \left ( 1728 \ln^3 \frac{\mu_0 ^ 2}{m_h^2 } - 932.0 \ln^2 \frac{\mu_0 ^ 2}{m_h^2 } + 9736.8 \ln \frac{\mu_0 ^ 2}{m_h^2 } + d_{30 } \right ) { \hat \lambda } _ { \rm oms}^3 \nonumber\\ & & + { \rm o}(\epsilon ) \biggr ] , \label{coupnum}\end{aligned}\ ] ] which agrees to one loop with the result of sirlin and zucchini @xcite .
the three - loop constant term @xmath119 depends on the yet unknown oms coefficients @xmath120 , @xmath110 , and @xmath121 . for convenience
we give the inverse formula of ( [ coupnum ] ) as well : @xmath122 \ , .
\label{coupinv}\end{aligned}\ ] ] eq .
( [ coup ] ) defines the expansion parameter @xmath27 of the running coupling when using the @xmath0 scheme .
it shows that @xmath113 is completely determined by specifying @xmath3 ( and thereby @xmath12 ) and a renormalization point @xmath35 at which ( [ coup ] ) is imposed .
the scale @xmath35 is unspecified .
however , to ensure that the logarithms @xmath123 stay small at all orders , it should be chosen of the order of @xmath3 .
let us emphasize that throughout the paper @xmath3 denotes the pole mass , even when discussing @xmath0 renormalization , and the vacuum expectation value @xmath114 is also chosen as defined in the oms as in @xcite .
expressions involving the running mass of the @xmath0-scheme can systematically be expressed in terms of the pole mass @xmath3 . despite the use of the pole mass @xmath3
, the @xmath0 renormalization scheme maintains all the advantages of mass independent schemes mentioned in the introduction .
for example @xmath124 corresponds to a physical higgs mass of @xmath125 gev at one loop , and @xmath126 gev at two loops . in fig .
[ msbar ] we show the @xmath0 coupling as a function of @xmath35 for different values of @xmath3 , varying @xmath35 in a typical range , @xmath127 .
we find that the @xmath0 coupling is larger than the oms coupling for most of the range of @xmath35 examined . if @xmath3 is larger than @xmath128 gev , the value of @xmath35 at which @xmath129 equals @xmath103 changes significantly when going from one loop to two loops . at one loop ,
the two couplings are equal if @xmath130 for any value of @xmath3 , because @xmath131 . at two loops ,
the relation gets a more complicated mass dependence .
we find that for @xmath132 gev the two - loop couplings of the two schemes are equal if @xmath133 .
= 3.8 in at first sight it appears reasonable to choose @xmath35 in such a way that calculated radiative corrections become small or even zero .
this criterion of _ ( fastest ) apparent convergence _
( fac ) has had some supporters in qcd rg analyses over a decade ago ( for a criticism see @xcite ) .
also in @xcite the one - loop fac scale @xmath134 , at which the one - loop correction in ( [ coup ] ) vanishes , has been used to define @xmath113 . yet in the following we will explain why this choice of the scale is not very useful , in particular in conjunction with the analysis of the breakdown of perturbation theory .
we can already anticipate this fact from the above observation that for large values of @xmath3 the two - loop fac scale differs sizably from its one - loop value . for a more detailed investigation
consider the generalization of the square bracket in ( [ coup ] ) to the order @xmath135 : @xmath136 , \label{ordern}\end{aligned}\ ] ] i.e. one has @xmath137 and so on .
application of the renormalization group equation ( [ beta ] ) ( with @xmath138 ) yields recursion relations for the @xmath139 s .
the solution for the coefficients @xmath140 of the leading logarithms is well - known : @xmath141 similarly we derive the general relationship @xmath142 for the next - to - leading logarithms with @xmath143 , \nonumber\end{aligned}\ ] ] where @xmath144 . for @xmath145 the coefficients @xmath140 and @xmath146 have already been given in ( [ coup ] ) , where also the nnll coefficient @xmath147 involving @xmath21 is displayed .
we remark here that the coefficients ( [ llog ] ) and ( [ nllog ] ) have the same form for any expansion of a running coupling @xmath1 in terms of some @xmath15-independent coupling with only the coefficient @xmath148 changed correspondingly .
for example , the expansion of @xmath149 in terms of @xmath150 yields a series of the form ( [ ordern ] ) with @xmath151 and @xmath152 in ( [ nllog ] ) .
now the choice @xmath153 nullifies all logarithms to all orders in ( [ ordern ] ) .
the fac scale @xmath154 instead nullifies the @xmath155-term in ( [ ordern ] ) and ( [ coup ] ) , but the price to be paid is the appearance of logarithms in higher orders .
since we know the nll coefficients @xmath156 and @xmath157 , we can check the effect of the fac scale setting on the @xmath158-term in ( [ coup ] ) : for @xmath75 the coefficient of @xmath158 equals 287 , but for @xmath154 one finds the larger coefficient 324 which increases the impact of higher order terms , especially if the coupling is large .
this explains the above finding that the fac scale changes significantly for large values of @xmath3 when passing from the one - loop to the two - loop order .
hence the fac scale setting pushes large terms from the calculated orders into the uncalculated higher orders of the perturbation series .
yet a clever choice of @xmath35 should yield the opposite and keep these higher order terms small . of course one
is not forced to use @xmath75 exactly .
but what is the allowed range for @xmath35 ?
clearly @xmath159 in ( [ ordern ] ) must be kept small , so that the logarithms do not become dominant in higher orders . a reasonable interval for @xmath35 should therefore obey @xmath160 for @xmath161 .
especially the size of the logarithms should not exceed the constant term @xmath162 . with the two - loop information ( [ coupnum ] ) at hand
we first compare the size of both @xmath163 and @xmath164 with the non - logarithmic term , @xmath165 , and find that for @xmath166 their magnitudes do not exceed @xmath167 .
second , using ( [ llog ] ) and ( [ nllog ] ) we can compare the leading logarithm @xmath168 with the next - to - leading logarithms @xmath169 to all orders in perturbation theory obtaining the smaller range @xmath170 . at three loops we can use the result ( [ b2oms ] ) to repeat the game with the next - to - next - to leading logarithm , which is multiplied by @xmath171 in ( [ coup ] ) : @xmath172 and @xmath173 are smaller than @xmath174 for @xmath175 . for @xmath176 or @xmath177 this term , however , exceeds @xmath178 , which is above the value one expects for the yet unknown constant @xmath179 . since @xmath180 and @xmath181 the choice @xmath182 , which has been so seductive in the order @xmath155 , leads to the fact that the three logarithmic terms in the order @xmath183 add with the same sign to yield @xmath184 .
moreover from ( [ llog ] ) and ( [ nllog ] ) one realizes that for @xmath185 the leading and next - to - leading logarithmic terms have the same sign to all orders @xmath186 except for @xmath187 and @xmath188 !
hence the one - loop fac choice @xmath189 pushes large terms into the higher orders .
the two - loop fac scale is even lower , increasing all higher order terms even more .
we conclude that our lack of knowledge of the higher order terms of the perturbation series forces us to consider _ any _ choice for @xmath35 in the range @xmath170 ( which may possibly be relaxed to @xmath190 ) with equal right .
changing @xmath35 in the @xmath191-th order perturbative expression for some observable changes the result by terms of the neglected order @xmath14 .
when perturbation theory works the dependence on @xmath35 diminishes order - by - order in @xmath4 . we will use a similar criterion to find the breakdown of perturbation theory in sections [ sect : scale ] and [ sect : cross ] .
let us close this section with a final remark on the arbitrariness of @xmath35 : cross sections with cm - energy @xmath11 and expressed in terms of @xmath149 involve the logarithm @xmath192 . using ( [ coup ] ) with @xmath193
, this logarithm would be large . using the running coupling @xmath149 evaluated at a scale @xmath194 ,
the logarithm is summed to all orders in perturbation theory .
the arbitariness in the choice of @xmath195 and @xmath196 reflects the fact that one can sum an arbitrary small constant together with the large logarithm .
this feature is also present in the oms scheme , but less apparent .
for example , the authors of @xcite sum the constant @xmath148 in ( [ coup ] ) together with @xmath192 . neglecting all couplings except for the higgs coupling @xmath4 ,
the higgs sector of the standard model is equivalent to a spontaneously broken @xmath197 theory with @xmath198 real scalar fields .
theories with spontaneous symmetry breaking have the remarkable property that the counterterms needed to make the theory finite are the same for the broken and unbroken symmetry , if a mass independent renormalization scheme is adopted @xcite . since
the beta coefficients are calculated from the counterterms of the coupling , eq .
( [ z ] ) , the beta function is also identical for both the broken and the unbroken theory .
hence one can calculate the @xmath199-loop coefficient @xmath200 of the higgs sector in two steps : first , obtain @xmath201 in an unbroken @xmath197 theory with @xmath198 real scalar fields by calculating the divergent parts of the four - point function to the @xmath199-loop order .
this is easier than working in the broken theory , because only a four - point coupling is involved .
once this is accomplished , @xmath202 of the broken theory is known , too .
second , calculate the scheme dependence @xmath203 .
this only requires the calculation of the @xmath191-loop ( not @xmath204-loop ) finite parts of self - energy diagrams .
the three - loop coefficient @xmath21 of the @xmath197 theory has been calculated in the @xmath0 scheme @xcite : @xmath205 to obtain @xmath111 , we return to eq .
( [ z ] ) to calculate the scheme dependence of the coefficient @xmath21 .
as stated above , unlike @xmath21 itself the scheme dependent difference between @xmath112 and @xmath111 can entirely be obtained from two - loop quantities : @xmath206 where @xmath207 is the value of the all massive master diagram with @xmath208 , which has been evaluated numerically in @xcite .
( [ schemeres ] ) can be obtained from the definition ( [ beta ] ) of the @xmath2-function or by comparing the bare coupling expressed in terms of @xmath113 with its expression in terms of @xmath12 . using the result for @xmath112
we obtain @xmath209 our analytical which is defined by a single feynman diagram ] result agrees with the numerical result obtained in @xcite to better than six digits . in fig .
[ rgecurves3lp ] we show the @xmath15-dependence of the nnll ( three - loop ) running coupling in the oms scheme .
the consistent solution is almost identical to its nll ( two - loop ) result even for @xmath74 gev , whereas the other nnll solutions show a behaviour very different from their nll results for large @xmath3 . for @xmath89 gev and @xmath210
tev all four nnll solutions show a very nice convergence .
for such a value of @xmath3 , however , the ll running coupling is not an adequate approximation for the upper values of @xmath15 considered . in the @xmath211 scheme ,
the convergence of the running coupling is rather poor for @xmath89 gev and above .
this is also true for the consistent solution . for @xmath212 gev
, the consistent nnll solution of the @xmath211 running coupling is not defined anymore if @xmath213 gev .
the poor performance of the nnll @xmath211 coupling is due to the term involving @xmath21 .
although @xmath214 is only three times larger than @xmath111 the coefficient @xmath215
entering the running coupling ( [ consist2lp]-[lamdef ] ) is 44 times larger in the @xmath211 scheme . this is caused by a numerical cancellation in the oms scheme , where @xmath216 .
we remark here that in a scheme with exact geometrical growth of the coefficients , @xmath217 , the consistent solution ( [ consist2lp ] ) equals the consistent nll result to all orders .
the accuracy of perturbation theory and its breakdown as @xmath3 increases can only be investigated in physical observables . here
processes in which all mass parameters are of the order of @xmath3 are of key importance : if the higgs self - interaction is non - perturbative at the scale @xmath3 at which ( [ defloms ] ) is imposed , we do not expect perturbation theory to work in any other observable .
two - body higgs decay rates are examples for such one - scale processes .
they do not contain large logarithms , which need to be summed to all orders .
nevertheless as emphasized at the end of sect .
[ sect : mscoup ] , rg methods can be used to judge the accuracy of perturbation theory : the dependence on the renormalization scale @xmath15 , at which the decay rate is evaluated , is of the neglected order in @xmath4 .
the stability of the perturbative result with respect to variations of @xmath15 must therefore increase order - by - order in @xmath4 .
this criterion to test perturbation theory has been first used in qcd in @xcite and has become a standard method in qcd .
likewise the ratio of two results obtained in different renormalization schemes must approach unity with increasing order in @xmath4 .
this renormalization scheme dependence will be the second tool used in the analysis of the breakdown of perturbation theory .
at first we consider the decay rate of a higgs into two gauge bosons .
the @xmath218 corrections to the decay rate @xmath219 are the same as the ones for the decay into two @xmath220 bosons .
the decay rates of the two channels only differ by an overall factor of proportionality .
we write for the decay rate @xmath221 here @xmath222 is the born approximation with @xmath223 the corrections stemming from the higgs self - interaction have been calculated in the oms scheme to two loops @xcite : @xmath224 in @xcite it has been stressed that the two - loop oms correction exceeds the one - loop oms term for @xmath225 gev , although the one - loop correction is still small compared to the tree - level term .
this indicates that either perturbation theory does not work for @xmath225 gev or the one - loop term is accidentally small in the oms scheme . using ( [ coupinv ] )
we can express @xmath226 in terms of the @xmath0 coupling : @xmath227 here @xmath15 is the renormalization scale at which the decay rate is evaluated .
the renormalization point @xmath35 , at which the coupling is defined by ( [ coup ] ) in terms of @xmath9 and @xmath3 , is chosen as @xmath75 throughout this section . if we also take @xmath228 , we observe that the apparent convergence of @xmath229 is worse than that of @xmath230 since both @xmath231 and the @xmath211 coefficients of the perturbation series are in magnitude larger than in the oms scheme .
the two - loop @xmath232 correction equals the one - loop @xmath211 term for @xmath233 gev .
the corresponding oms result yields @xmath234 gev .
let us now investigate whether one can refine these bounds on a perturbative higgs mass by the examination of the renormalization scheme and scale dependence .
we start our analysis by looking at the scheme dependence of the decay width . in the left plot of figure fig .
[ figscheme ] we show the normalized scheme dependence @xmath235 at one and two loops , using nll and nnll running couplings , respectively .
the plot indicates that for @xmath236 gev the scheme dependence is reduced when going from one - loop to two - loop order . for larger higgs masses , the scheme dependence increases , suggesting that perturbation theory is not working satisfactory or even fails .
the scheme - dependence criterion indicates problems with perturbation theory in at least one of the two schemes considered for @xmath237 gev .
the criterion , however , is sensitive to the possibility of an accidental smallness of the one - loop correction .
we remark here that the two - loop @xmath211 term in ( [ hwwmsbar ] ) is less than four times the square of the one - loop term , while in the oms scheme , ( [ hwwoms ] ) , this ratio almost equals eight . for @xmath238 gev
the result for @xmath239 is larger than the @xmath0 expression by 23% , and for @xmath240 gev the scheme dependence reaches unacceptable 52% .
the simple criterion of comparing the magnitudes of the one - loop and two - loop contributions gives upper bounds on @xmath3 which seem to be too large .
another criterion for the validity of perturbation theory is the order - by - order reduction of the scale dependence .
the explicit @xmath15-dependent logarithms in ( [ hwwmsbar ] ) compensate the effect of the running coupling @xmath1 to the order considered . to use this criterion in the oms scheme as well , we need to introduce the rg logarithms into the oms decay width ( [ hwwoms ] ) . in the oms scheme ,
the callan - symanzik equation describes the response of some green s function to the scaling @xmath241 of its external momenta .
its solution for the decay rate ( [ hwwoms ] ) reads @xmath242 expanding the running coupling @xmath243 in ( [ defww ] ) and ( [ hwwcs ] ) in terms of @xmath244 , ( [ hwwcs ] ) yields ( [ hwwoms ] ) up to the neglected order @xmath245 , of course .
the oms result expressed in terms of the running coupling ( [ hwwcs ] ) can be used to examine the scale dependence of the decay width .
we use the ll , nll and nnll expressions of the decay rates which consist of the born , one- and two - loop result in ( [ hwwmsbar ] ) supplied with the ll , nll or nnll running coupling @xmath149 , respectively .
then the renormalization scale @xmath15 is varied . the right plot in fig .
[ figscheme ] shows that perturbation theory nicely works for the @xmath0 scheme if @xmath246 gev : the scale dependence diminishes order - by - order .
a similar behavior is found for the oms result using the oms running coupling and the same value of @xmath3 .
next we look for an upper bound for a perturbative higgs mass .
we investigate the scale dependence of the decay rate for values of @xmath3 up to 800 gev .
based on the size of the logarithmic terms in the higher orders of the perturbation series for the @xmath113 we have already advocated the range @xmath247 in sect .
[ sect : mscoup ] . we may add a physical argument for this range as well : suppose one decides to include the non - zero width of the higgs into the analysis
. then the decay diagrams must be calculated with an off - shell higgs boson with invariant mass @xmath11 , and the result is convoluted with a breit - wigner function .
the decay rate would differ from ( [ hwwmsbar ] ) by a function of @xmath248 and @xmath3 vanishing for @xmath249 .
the decay rate then involves two logarithms : @xmath250 and @xmath251 .
the choice of @xmath15 could be with equal right @xmath252 , @xmath253 or any scale in between .
this suggests choosing the range for @xmath15 to be of the order of the total width . for the values of higgs masses we are interested in
, the width is between @xmath254 and @xmath255 @xcite , so that the range @xmath247 is appropriate in both oms and @xmath211 scheme . in fig .
[ figscale ] we have plotted the scale dependence of @xmath256 vs. the physical higgs mass in both oms and @xmath211 scheme . the tree - level coupling is @xmath257 .
the scale dependence at a given order is represented by the smallest and the largest value of @xmath258 when @xmath15 is varied in the range @xmath259 . in the @xmath0 scheme
the scale dependence correctly decreases when passing from ll to nll to nnll order if @xmath260 gev . for @xmath261 gev the scale dependence in the ll and nll order become equal and reach 36% . for @xmath262 gev and @xmath263 ,
the consistent solution of the nnll running coupling is no longer defined in the @xmath232 scheme , indicating the ultimate breakdown . in the oms scheme ,
the nll and nnll scale dependences become equal for @xmath264 gev , but are numerically small ( 8% ) .
we conclude that perturbation theory for bosonic higgs decays breaks down for higgs masses of the order of @xmath265 gev .
the scale - dependence criterion yields similar upper bounds on @xmath3 in both schemes , although the absolute scale dependence is much smaller in the oms scheme than in the @xmath0 scheme .
using running coupling solutions other than the consistent one ( [ consist2lp ] ) , we obtain similar bounds .
it should be noted that our scale - dependence criterion is not only sensitive to the coefficients of the different orders in @xmath266 , but also to the coefficients of the @xmath2-functions which also enter the non - logarithmic terms of the uncalculated higher orders via diagrams connected with counterterms in ( [ z ] ) .
= 3.0 in = 3.0 in = 3.0 in = 3.0 in our next example is the fermionic decay width of the higgs particle . at born level
it reads @xmath267 here @xmath268 ( 3 ) for lepton ( quark ) flavors . at tree level , this process is independent of the higgs coupling @xmath4 . including radiative corrections
, the fermionic decay rate receives corrections in powers of @xmath4 . in the oms scheme
they are @xcite @xmath269 since the tree - level result of the fermionic higgs decay , eq .
( [ eqborn ] ) , is independent of the coupling @xmath4 , we only need the one - loop relation between @xmath270 and @xmath115 to calculate the @xmath271 decay width up to two loops .
likewise , our scale variation criterion only involves ll and nll running coupling in connection with the one - loop and two - loop results . as a result
we can only compare the scale dependence of two instead of three orders . since the ll one - loop result is identical in both schemes ,
the scheme dependence can only be compared at the nll two - loop level .
no bounds on @xmath3 can be derived from the scheme - dependence criterion without knowing the three - loop corrections .
this distinguishes the decay @xmath272 from the case @xmath273 discussed in the preceeding section .
combining eq .
( [ coup ] ) with the previous equation , we obtain the correction to the fermionic higgs decay in @xmath0 quantities @xcite : @xmath274 we also give the scale dependence of the oms result : @xmath275 in fig .
[ hffrge ] we show the scale dependence of the decay width expressed in terms of the running coupling in both @xmath276 and oms scheme at one and two loops as a function of @xmath3 . at each order
the three curves refer to @xmath277 , @xmath278 , and @xmath279 .
we find that the @xmath15 dependence of the two - loop result is larger than the @xmath15 dependence of the one - loop result if @xmath3 is larger than @xmath280 gev in the @xmath211 scheme . in the oms scheme
the corresponding bound is @xmath281 gev .
these results , however , are not as valuable as those found in the previous section , since they are only founded on the comparsion of two orders rather than three
. the low value of @xmath282 gev in the @xmath283 scheme may be accidental .
the scale dependence is very weak , less than a few percent for higgs masses up to 750 gev .
this is due to the fact that the tree - level result does not depend on @xmath4 .
the scheme dependence at nll ( two - loop ) is marginal for the same reason .
we presently have no information on the nnll behaviour of the fermionic decay .
looking at fig .
[ hffrge ] we conclude that the upper bound on a perturbative higgs mass is in agreement with our findings in the case of the bosonic higgs decay .
= 3.0 in = 3.0 in
the previous section considered one - scale processes .
the only logarithms appearing in the rg improved results are @xmath284 testing the perturbativity of higgs sector is much more stringent when considering high - energy two - scale processes , because they involve the running coupling at a high scale .
typical two - scale processes which depend on the coupling @xmath4 are @xmath285 scattering processes involving longitudinally polarized gauge bosons and the higgs boson . in the limit @xmath286
the longitudinally polarized gauge bosons can be replaced by massless goldstone bosons . in this limit
all couplings except for the trilinear and quartic higgs coupling are subleading and can be neglected . also assuming @xmath287 , the high - energy amplitudes and cross sections
have been calculated to two loops by @xcite .
the high - energy limit is approached if @xmath288 @xcite .
its error is less than a few percent if @xmath289 .
for most of our analysis we will consider @xmath290 tev .
a typical example is the process @xmath291 .
the high - energy cross section only depends on @xmath11 , @xmath3 , and @xmath4 .
the oms cross section is @xcite @xmath292 ^ 2\ , \biggl[\,1\,+ \left ( 24 \ln \frac{s}{\mu^2 } - \ , 42.65 \right ) \,\frac{\lambda_{\rm oms}(\mu)}{16\pi^2}\ , \nonumber \\ & & \phantom{\frac{1}{8\pi s } [ \lambda_{\rm oms}(\mu)]^2\ , } + \left ( 432 \ln^2 \frac{s}{\mu^2 } - 1823.3\ln \frac{s}{\mu^2 } + 24.0\ln \frac{s}{m_h^2 } + \;2455.1\,\right ) \frac{[\lambda_{\rm oms}(\mu)]^2}{(16\pi^2)^2}\ , \nonumber \\ & & \phantom{\frac{1}{8\pi s } [ \lambda_{\rm oms}(\mu)]^2\ , } + \ ; { \rm o}\left([\lambda_{\rm oms}(\mu)]^3\right)\,\biggr]\ , , \label{wwzzcross}\end{aligned}\ ] ] where the oms renormalization fixes @xmath75 such that the oms running coupling has the boundary condition @xmath293 .
the scale @xmath15 must be chosen of the order @xmath11 to resum the large logarithms to all orders .
for a complete resummation of all logarithms up to three loops , one uses the nnll running coupling and takes @xmath253 .
the @xmath15-independent logarithm is related to the field anomalous dimension and needs no resummation because of the smallness of its coefficient . at one loop ,
the anomalous dimensions of the higgs and gauge bosons are zero , and at two loops they are numerically unimportant @xcite .
the @xmath211 result is @xmath294 ^ 2\ , \biggl[\,1\,+ \left ( 24\ln\frac{s}{\mu^2 } -\ , 60.0 \right)\ , \frac{\lambda_{\overline { \rm ms}}(\mu)}{16\pi^2}\ , \nonumber \\ & & \phantom{\frac{1}{8\pi s } [ \lambda_{\overline { \rm ms}}(\mu)]^2\ , } + \left ( 432 \ln^2 \frac{s}{\mu^2 } - 2472.0\ln \frac{s}{\mu^2 } + 24.0 \ln \frac{s}{m_h^2 } + \;3367.9 \right ) \,\frac{[\lambda_{\overline { \rm ms}}(\mu)]^2}{(16\pi^2)^2}\ , \nonumber \\ & & \phantom{\frac{1}{8\pi s } [ \lambda_{\overline { \rm ms}}(\mu)]^2\ , } + \ ; { \rm o}\left([\lambda_{\overline { \rm ms}}(\mu)]^3\right)\,\biggr]\ , .
\label{wwzzmsbar}\end{aligned}\ ] ] we choose the renormalization point to be @xmath75 such that the the @xmath211 running coupling is fixed at @xmath138 by ( [ coup ] ) .
the rg structure of the cross section is similar to the one of the decay width @xmath295 since the tree - level result also depends on @xmath4 .
there are two important differences : the tree - level cross section is proportional to @xmath4 _ squared _ , and the running coupling resums terms of order @xmath296 which can lead to a significant increase of the running coupling compared to the tree - level coupling .
choosing @xmath297 gev , we show the scheme dependence of the cross section @xmath291 in fig .
[ wwzzfigscheme ] . for higgs masses larger than 400 gev ,
the scheme dependence is larger than 40% and actually increases when going from nll to nnll cross section . choosing @xmath298 gev
, a value more realistic for @xmath299-scattering at future colliders , we find the critical mass value to be @xmath300 gev .
for such values of @xmath3 and @xmath11 the high - energy approximation yields about 70% of the exact @xmath4-dependence of the cross section @xcite . increasing @xmath11 , the breakdown of perturbation theory as seen in the scheme - dependence criterion happens for smaller and smaller values of @xmath3 .
we will come back to this later .
in addition to the scheme dependence , we also evaluate the scale - dependence criterion .
since the renormalization group is used to resum @xmath301 terms , the scale @xmath15 is varied around @xmath11 rather than @xmath3 .
the arguments of sect .
[ sect : scheme ] , which are based on the size of the beta function coefficients , suggest the range @xmath302 . choosing @xmath303 gev
, the result of this variation is shown in fig .
[ wwzzrge ] for both @xmath211 and oms scheme . in the @xmath211 scheme
, we observe a nice order - by - order reduction of the scale dependence if @xmath3 is less than 357 gev . for larger values of @xmath3 ,
the nll scale dependence exceeds the ll one , and for @xmath304 gev we also find the nnll scale dependence to be larger than the nll one . for the oms
result we find only one crossing point .
there the nnll scale dependence dominates over the nll one if @xmath305 gev .
taking also the scheme dependence into account we conclude that perturbative results for longitudinal gauge boson scattering at @xmath303 gev cease to be meaningful if @xmath3 is of the order 400 gev or above . for a value of @xmath306 gev
we find limits of @xmath307 , but here the additional low - energy contributions though not dominant may change the limit .
the upper bounds on a perturbative higgs mass from scattering processes are more stringent than the results found in higgs decays .
the reasons are two - fold .
first , the one - loop and two - loop coefficients of the perturbative corrections to the cross section are an order of magnitude larger than the corresponding coefficients of the decay processes .
second , the running coupling evaluated at a scale @xmath308 is numerically larger than the coupling involved in the decay widths .
it is also interesting to note that the bounds derived here are similar to the results found in @xcite which are based on perturbative violation of unitarity at two loops , a rather different criterion for the breakdown of perturbation theory .
compared to the nonperturbative lattice results of @xmath309 gev @xcite , our perturbative criteria require smaller higgs masses . in fig .
[ wwzzembed ] we show the upper bound for a perturbative higgs mass as a function of some embedding scale @xmath310 . since the standard model higgs sector is depending on the value of @xmath3 not defined above a certain energy scale , it is a common procedure to introduce an embedding scale above which the physical interactions are to be described by a more complete theory . requiring that the sm higgs sector is still perturbative at such an embedding scale
, it is possible to give upper bounds on the higgs mass as a function of the embedding scale .
here we require the process @xmath311 to be perturbative for energies @xmath312 and apply our criteria for scheme and scale dependence to calculate the upper bound on @xmath3 .
the result is shown in fig .
[ wwzzembed ] . at 2 tev ,
the different bounds on @xmath3 correspond to the values derived from figs .
[ wwzzfigscheme ] and [ wwzzrge ] . for increasing embedding scale , the upper bound on @xmath3 decreases , and the three different criteria give converging bounds . at @xmath313 gev ,
the upper bound is @xmath314 gev . for such low higgs masses
, however , one expects the top - quark yukawa coupling to have an influence on the rg evolution of the higgs coupling . as a matter of fact
, the sm beta function of the higgs coupling can become negative if the higgs particle is too light , invalidating our analysis .
taking @xmath315 gev , this is expected to happen for @xmath316 gev @xcite , which is lower than the values considered by us . using the results of lindner @xcite
, we expect the change of our higgs mass bounds due to a top - quark mass of 175 gev to be less than 10% for the whole range of energy - scales considered . for comparison
, we also display the values of @xmath3 which would lead to a one - loop landau singularity at @xmath317 . calculating the landau pole
, we again neglected yukawa and gauge couplings .
it is worth noting that our criteria for an upper higgs mass lead to values of the running coupling at the embedding scale which are not very large . for values of @xmath11 in the tev range ,
the maximal allowed value of the running coupling is less than 2.2 , and at gut energies we find the maximal value to be less than 1.6 .
these small values are surprising : the beta function and the solutions for the running coupling show excellent convergence for such small values . as a matter of fact
, the usual criteria for the breakdown of perturbation theory assume that the running coupling becomes large . yet
this method is unsatisfactory : the consideration of the one - loop landau pole as in @xcite can not be extended to higher orders .
instead the authors of @xcite use the criterion of ( in our notation ) @xmath318 . here
the choice of the numerical bound involves some arbitrariness .
the plots in fig .
[ rgecurves ] and fig .
[ rgecurves3lp ] show that @xmath319 is clearly chosen too large , because the different solutions differ substantially for @xmath320 .
the second arbitrariness of this methods resides in the fact that the breakdown of perturbation theory is judged from a scheme dependent quantity : @xmath321 depends on the renormalization scheme through the radiative corrections to the matching condition ( [ defloms ] ) and , most important , through the coefficients @xmath104 , @xmath322 , of the @xmath2-function . in contrast
physical observables are scheme independent up to the calculated order .
they , however , seem to become non - perturbative for much smaller values of @xmath4 . for comparison , the unitarity arguments used in @xcite yield an upper bound on the running coupling of 2.3 , independent of any choice of @xmath11 .
= 4.0 in = 3.0 in = 3.0 in = 3.0 in
our analysis of various physical processes of the higgs sector indicates that the breakdown of perturbation theory can not be judged purely on grounds of a large higgs ( running ) coupling .
the breakdown of perturbation theory in both higgs decays ( @xmath323 ) and scattering processes ( @xmath324 ) occurs for relatively small values of the higgs coupling .
the usual criterion the breakdown of the perturbative behaviour of the beta function and the running coupling yields upper bounds for a perturbative higgs mass which are too large . applying our criteria of scheme - dependence and scale - dependence to higgs decays , we find a satisfactory perturbative behaviour of the decay widths if @xmath325 o(700 gev ) . in the case of @xmath285 scattering processes , one needs to specify the energy scale at which the process is to take place .
choosing @xmath11 to be a couple of tev , the higgs mass has to be less than o(400 gev ) to guarantee a good perturbative behaviour of the cross section for @xmath291 scattering .
requiring such processes to be perturbative at @xmath326 gev , the higgs mass has to be less than o(150 gev ) .
one of the authors ( k.r . ) gratefully acknowledges the hospitality of the aspen center for physics where parts of this work were done , and thanks m. lindner and s. willenbrock for interesting and informative discussions .
i. jack and h. osborn , j. phys . * a16 * , 1101 ( 1983 ) .
e. lendvai , g. pcsik , and t. torma , mod .
lett . a * 6 * , 1195 ( 1991 ) ; acta phys . pol .
b * 22 * , 607 ( 1991 ) .
w. a. bardeen , a. j. buras , d. w. duke and t. muta , phys .
d*18 * , 3998 ( 1978 ) .
landau , in : niels bohr and the development of physics ( mcgraw - hill , new york , 1955 ) .
l. maiani , g. parisi , and r. petronzio , nucl .
* b136 * , 115 ( 1979 ) ; + n. cabbibo , l. maiani , g. parisi , and r. petronzio , nucl .
* b158 * , 295 ( 1979 ) . b. grzadkowski and m. lindner , phys .
b*178 * , 81 ( 1986 ) . in the context of lattice studies
this has been discussed in : m. lscher and p. weisz , phys .
* b212 * , 472 ( 1988 ) .
maher , l. durand , and k. riesselmann , phys .
d * 48 * , 1061 ( 1993 ) ; ( e ) * 52 * , 553 ( 1995 ) ; + a. ghinculov and j. van der bij , nucl . phys .
* b436 * , 30 ( 1995 ) .
a. ghinculov , phys .
lett . b * 337 * , 137 ( 1994 ) ; ( e ) * 346 * , 426 ( 1994 ) .
a. sirlin and r. zucchini , nucl .
* b266 * , 389 ( 1986 ) .
p. stevenson , phys .
d*23 * , 2916 ( 1981 ) ; nucl .
* b203 * , 472 ( 1982 ) .
collins , _ renormalization _ ( cambridge university press , cambridge , 1984 ) , chap . 9 .
vladimirov , d.i .
kazakov , and o.v .
tarasov , sov .
jetp * 50 * , 521 ( 1979 ) .
m. lscher and p. weisz , nucl .
* b318 * , 705 ( 1989 ) .
a. j. buras , m. jamin and p. h. weisz , nucl .
* b347 * , 491 ( 1990 ) .
a. ghinculov , univ .
freiburg preprint thep 95/11 , hep - ph/9507240 .
l. durand , b.a .
kniehl , and k. riesselmann , phys .
lett . , 2534 ( 1994 ) ; ( e ) * 74 * , 1699 ( 1995 ) ; phys .
d * 51 * , 5007 ( 1995 ) .
a.i . bochkarev and r.s .
willey , phys .
d * 51 * , 2049 ( 1995 ) ; eq .
( 17 ) of this paper uses incorrect numbers .
k. riesselmann , in : proceedings of the ringberg workshop on `` perspectives for electroweak interactions in @xmath327 collisions '' , munich , germany ( feb . 5 - 8 , 1995 ) ,
ed . : b. kniehl , ( world scientific , singapore , 1995 ) , p. 175 .
k. riesselmann , tech .
munich preprint tum - hep-223/95 ( july 1995 ) , hep - ph/9507413 .
l. durand , p. maher , and k. riesselmann , phys .
d*48 * , 1084 ( 1993 ) .
u. heller , m. klomfass , h. neuberger , and p. vranas , nucl .
phys . , 555 ( 1993 ) .
g. altarelli and g. isidori , phys . lett . *
b337 * , 141 ( 1994 ) .
m. lindner , z. phys .
* c31 * , 295 ( 1986 ) . | we discuss different aspects of the higgs self - interaction in the @xmath0 and the on - mass - shell ( oms ) scheme .
the running coupling @xmath1 is investigated in great detail .
the three - loop coefficient of the @xmath2-function in the oms scheme is derived , and the three - loop running coupling is calculated .
the breakdown of perturbation theory for large higgs masses @xmath3 is analyzed in three physical observables for which two - loop results are known . requiring the dependence on the renormalization scale to diminish order - by - order in @xmath4
, we find that perturbation theory breaks down for @xmath5 o(700 gev ) in higgs decays .
similarly , @xmath3 must be smaller than o(400 gev ) for perturbatively calculated cross sections to be trustworthy up to cm energies of o(2 tev ) .
if the higgs sector shall be perturbative up to the gut scale , the higgs must be lighter than o(150 gev ) .
for the two - loop observables examined , the apparent convergence of the perturbation series is better in the oms scheme than in the @xmath0 scheme . |
we model our approach after standard algorithms for configuration model ( cm ) networks @xcite .
cm networks are useful because all edges from a node are independent of one another , in the sense that whether an epidemic results from following one edge is independent of the result along any other edge because short cycles are negligible .
we begin with @xmath2 nodes . to each node
@xmath3 we assign two degrees , an _ independent edge _ degree @xmath4 and a triangle degree @xmath5 .
the joint probability of @xmath4 and @xmath5 is given by @xmath6 .
then @xmath3 will be part of @xmath5 triangles and have @xmath4 other edges
. each triangle and edge from @xmath3 will be independent of other triangles and edges in the same way that edges in cm networks are independent .
we create an independent stub list and a triangle stub list .
we place @xmath3 into the independent stub list @xmath4 times and into the triangle stub list @xmath5 times .
once all nodes are placed into the lists , we randomize them .
we then take the pairs of nodes in positions @xmath7 and @xmath8 of the independent list and join them , and the triples in positions @xmath9 , @xmath10 , and @xmath11 and join them into a triangle . some repeated edges or loops or short cycles other than the triangles we impose
may appear , but their impact is negligible as @xmath12 .
this algorithm inevitably segregates those nodes with a high proportion of triangles from those nodes with a low proportion of triangles .
if the degrees of nodes with many triangles differs from the degrees of nodes with few triangles , then this effect will cause correlation of different degrees . in order to understand the impact of clustering
, we must be able to compare percolation in these clustered networks with percolation in networks whose nodes are segregated in the same way . for comparative purposes we develop a corresponding unclustered network with the same segregation as the clustered networks . given the joint distribution @xmath6 of independent and triangle degrees , we create a new network where nodes are assigned _ blue _ and _ red _ degrees such that @xmath13 and @xmath14 .
the joint distribution is given by @xmath15 .
we proceed as before .
we create a blue and red list , and pair nodes in positions @xmath7 and @xmath8 in the blue list and then repeat with the red list , joining pairs , not triples .
the resulting network has the same segregation as the corresponding clustered network , but short cycles are negligible .
@xmath1 is usually defined as the number of new infections caused by an average infected individual .
occasionally alternate definitions are used , but in some way it represents the number of new infections attributed to an average infected individual .
@xmath16 is the threshold below which epidemics are impossible ( _ i.e. _ , the percolated network has no giant component ) .
if @xmath17 then epidemics are possible , but not guaranteed . to simplify the analysis , first assume that @xmath3 , @xmath18 , and @xmath19 are members of a triangle and @xmath3 becomes infectious first . there are multiple ways that both @xmath18 and @xmath19 can become infected from edges within the triangle , but they all have the same impact on the epidemic .
it is convenient to treat infections of @xmath18 and @xmath19 as if they came from from @xmath3 regardless of the actual path followed . thus if @xmath3 becomes infected , then with probability @xmath20 it is credited with infecting both @xmath18 and @xmath19 , and with probability @xmath21 it is credited with infecting just @xmath22 . with probability @xmath23
it infects neither . in spirit
this approach is similar to that of @xcite . for book - keeping purposes ,
we define the rank @xmath24 of a node as follows : the index case is given rank @xmath25 .
each node @xmath18 is then assigned rank @xmath24 to be the shortest path from the index case to @xmath18 , bearing in mind the rule above for crediting infections .
this allows us to define a @xmath26 next - generation matrix @xcite .
we separate those nodes infected along an independent edge from those nodes infected along a triangle edge .
we define @xmath27 and @xmath28 to be the number of infections that a node infected from an independent edge is expected to cause along independent and triangle edges respectively .
we symmetrically define @xmath29 and @xmath30 . if @xmath31 and @xmath32 are the number of nodes of rank @xmath24 which were infected along independent and triangle edges respectively , then @xmath33 where @xmath34 , @xmath35 , @xmath36 , and @xmath37 .
the dominant eigenvalue of this matrix is @xmath1 .
we generally want to determine @xmath0 such that @xmath38 .
substituting @xmath16 into the characteristic equation gives @xmath39 the value @xmath40 that solves this equation is the threshold transmissibility below which epidemics are impossible .
the original network has a giant component if @xmath17 when @xmath41 .
thus the conditions for a giant component are @xmath42 and/or @xmath43 the only networks for which the first condition applies but not the second are networks with enough independent edges and triangle edges such that a giant component exists soley within the independent edges and a giant component exists soley within the triangle edges .
we define @xmath44 and @xmath45 in the same manner , except that triangles need not be considered .
then @xmath46 where @xmath47 , @xmath48 , @xmath49 , and @xmath50 .
substituting @xmath16 into the characteristic equation finds the epidemic threshold @xmath51 the network has a giant component when @xmath52 and/or @xmath53 the difference between these conditions and those of the corresponding clustered network comes from the fact that @xmath54 . from this
it can be shown that the epidemic threshold occurs at smaller @xmath0 for the unclustered network .
to calculate the fraction of nodes in the giant component , it suffices to calculate the probability that a random node is _ not _ part of the giant component .
these calculations have been done for cm networks by @xcite .
we follow the approach of @xcite .
a related approach is given by @xcite .
we let @xmath55 be the probability a random node @xmath3 is not part of the giant component .
we have @xmath56 where @xmath57 and @xmath58 are the probabilities that an independent edge or a triangle respectively does not connect to the giant component .
to find @xmath57 , we note that there are two ways an edge can fail to connect @xmath3 to the giant component : it may be deleted in the percolation process with probability @xmath59 , or it may be kept , but @xmath18 , the node reached , is not part of the giant component .
we have @xmath60 where @xmath61 is the probability that a node @xmath18 reached along an independent edge is not part of the giant component . to calculate @xmath61
we note that @xmath18 is selected proportional to @xmath4 , but only has @xmath62 susceptible neighbors along independent edges .
we get @xmath63 for @xmath58 we get @xmath64 ^ 2 - 2t^2(1-t)h_\triangle(1 - h_\triangle ) \ , , \ ] ] where @xmath65 is the probability a node reached along a triangle edge does not connect to the giant component through any edge not in the triangle .
we find @xmath66 the resulting system of equations for @xmath57 , @xmath58 , @xmath61 , and @xmath65 can be solved iteratively , and the result gives @xmath55 .
to find @xmath67 , the probability a random node in the unclustered network is not part of the giant component , we proceed similarly .
we find @xmath68 it can be shown that @xmath69 for the equivalent degree distributions .
consequently the size of the giant component is smaller in clustered networks than in unclustered networks of the same degree distribution and degree correlations .
in figure [ fig : comp ] we consider outbreak spread on three networks , all of which have the same degree distribution . we compare simulated epidemic sizes with predictions from the clustered equations , the unclustered , segregated equations , and the equations derived previously for configuration model networks @xcite .
+ + the nodes are equally distributed between degrees @xmath70 , @xmath71 , and @xmath72 . in each network
the clustering is distributed differently . in the first , @xmath73 , @xmath74 , and @xmath75 .
that is those nodes with degree @xmath72 are only in triangles , nodes of degree @xmath71 have half of their edges in triangles and independent edges , and nodes of degree @xmath70 have just independent edges .
high degree nodes tend to be clustered and contact other high degree nodes .
the tendancy to contact other high degree nodes reduces the epidemic threshold , but the clustering raises the threshold . in the second network , we take @xmath76 , @xmath77 , @xmath78 , @xmath79 , and @xmath80 .
this yields identical distribution of neighbor degrees for nodes reached by either a triangle or an independent edge .
the unclustered , segregated equations yield the same result as the configuration model equations .
the clustered calculations have smaller epidemics .
the third network is an inversion of the first .
nodes with high degree have independent edges while nodes with low degree are clustered .
we take @xmath81 , @xmath74 , and @xmath82 . again
the assortativity reduces the epidemic threshold while clustering reduces the epidemic size . in this particular case , it is the preference for high degree nodes ( which are unclustered ) to contact one another that leads to the reduction in epidemic threshold , and so it is clear that the effect is due to assortative mixing , not clustering .
we have introduced a new model of clustered networks on which we study percolation and epidemics .
this model allows us to make a number of analytic prediction because the edges of the network can be partitioned into sets which are independent of one another ( independent edges or triangles ) .
we have shown that these networks can have a lower epidemic threshold than configuration model networks with the same degree distribution .
however , this is not a consequence of clustering , but rather a consequence of assortative mixing .
the clustering of the network can be proven to raise the epidemic threshold and reduce the epidemic size from networks with the same degree correlations , but without clustering .
this work was supported by the rapidd program of the science & technology directorate , department of homeland security and the fogarty international center , national institutes of health .
o. diekmann , j. a. p. heesterbeek , and j. a. j. metz . on the definition and the computation of the basic reproduction ratio @xmath83 in models for infectious diseases in heterogeneous populations .
, 28:365382 , 1990 . | the social networks that infectious diseases spread along are typically clustered . because of the close relation between percolation and epidemic spread , the behavior of percolation in such networks gives insight into infectious disease dynamics .
a number of authors have studied clustered networks , but the networks often contain preferential mixing between high degree nodes .
we introduce a class of random clustered networks and another class of random unclustered networks with the same preferential mixing .
we analytically show that percolation in the clustered networks reduces the component sizes and increases the epidemic threshold compared to the unclustered networks .
classical random networks contain few short cycles , and the proportion of nodes in short cycles goes to zero as the number of nodes increases .
in contrast social networks typically contain many short cycles .
we refer to such networks as _ clustered _ networks .
the impact of clustering on percolation properties is usually difficult to calculate because cycles prevent the use of branching process arguments , but it is widely expected that clustering significantly alters percolation .
typically studies of infectious disease spread assume that outbreaks begin with a single infected node .
the disease travels to each susceptible neighbor independently with probability @xmath0 , the _ transmissibility _ , and the node recovers
. the process repeats .
we focus on diseases for which recovery provides immunity , so recovered nodes are not susceptible .
typically the outbreak dies out stochastically or becomes an epidemic and spreads until the number of susceptible nodes is reduced .
it is well - established that for fixed @xmath0 , the epidemic spread can be mapped to a bond percolation problem wherein each edge is kept with probability @xmath0 @xcite .
if we perform percolation on the network and then choose the initial infection , the disease spreads from that initial infection along edges of the percolated network , and so an epidemic occurs iff the initial node is in the giant component .
the size of the epidemic matches the size of the giant component .
this establishes that the probability and fraction infected in epidemics are equal if @xmath0 is fixed and all edges are independent .
because social networks frequently exhibit clustering , a number of studies have investigated the impact of clustering on epidemic problems @xcite .
some have found that clustering reduces the sizes of epidemics and raises the epidemic threshold .
that is , clustering reduces the size of giant components and raise the percolation threshold .
however , others have shown that clustering appears to reduce the threshold .
consequently epidemics should be possible at lower transmissibility in the presence of clustering .
this discrepancy occurs because there are many ways used to generate clustered networks , and each nework class results in different behaviors .
it is difficult to separate the impact of clustering from other features introduced by the network generation process . in this article
we introduce a new algorithm to generate random clustered networks .
the clustered networks have correlations between degrees in a well - defined manner which can lead to assortativity , the tendancy for nodes to contact nodes of similar degree .
we show how to generate unclustered networks with the same correlations .
we can make analytic comparisons between the two , and so clearly separate the effect of clustering from degree correlations .
we show that although the clustered networks can have a reduced threshold compared to purely random networks of the same degree distribution , that is entirely an artifact of the assortativityx .
compared to an unclustered network of the same degree correlations , the clustered networks result in smaller epidemics and higher epidemic threshold .
this article is organized as follows : we first introduce our clustered and unclustered networks .
we then calculate and compare the epidemiological quantity @xmath1 which measures how many new infections a typical infected node causes .
finally , we calculate the final size / probability of epidemics assuming constant @xmath0 . |
the absorption lines of heavy elements ( e.g. , c , mg , o , fe ) observed in the spectra of distant quasi - stellar objects ( qsos ) provide one of the best ways of observing and understanding the evolving chemical enrichment of the universe .
several lines of evidence suggest that these absorption lines arise in the circumgalactic medium ( cgm ) surrounding galaxies lying close to the lines of sight to qsos ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the source of the heavy elements in the cgm is not clear , but it is thought to result from enrichment of the halo gas by large - scale galactic outflows driven by a combination of stellar winds , supernovae , active galactic nuclei ( agn ) and by the accretion of gas tidally stripped during mergers from star - forming galactic disks ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the ubiquity of galactic winds associated with star - formation in galaxies over a wide range in redshift is now empirically well established ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) by the detection of blue - shifted absorption lines of mgii and feii in galaxies .
these outflows can carry enriched material out of galactic disks to distances of @xmath950 kpc into galaxy halos @xcite and thus enrich the cgm ( e.g. , * ? ? ?
outflows of enriched material from galaxy disks have also been invoked to regulate the mass - metallicity relationship observed in galaxies ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) : star - formation , fuelled by accreted gas , can increase a galaxy s metallicity , but outflows can remove enriched gas from the galaxy and , in lower mass systems , prevent further star - formation and enrichment ( e.g. , * ? ? ?
. in this picture the metal enrichment of the circumgalactic medium is the result of supernovae and a natural consequence is that the ratio of @xmath10 elements to fe should reflect ( to first order ) the relative rates of core collapse supernovae ( ccsn ; which result from the explosion of a massive star ) and type ia supernovae ( snia ; which result from the explosion of an accreting white dwarf ) .
the heavy element yields from both classes of supernovae have been modeled in various studies @xcite , which typically find that the @xmath10-elements are formed predominantly in ccsn , whereas fe production is dominated by snia
. the observed ratio of @xmath10-element to fe equivalent widths in the spectra of absorption line systems over a range of redshifts can potentially provide an important window into the relative variation of sn rates , the efficiency of outflows in enriching the cgm , and the resultant abundance patterns of diffuse cgm and igm material with cosmic time . in this paper
, we use the johns hopkins university - sloan digital sky survey ( jhu - sdss ) metal absorption line catalog ( * ? ? ?
* hereafter zm13 ) to investigate the variation of the feii / mgii ratio as a function of redshift .
the catalog tabulates the equivalent widths for 35,752 absorption line systems measured in the spectra of qsos observed by the sloan digital sky survey dr7 data release @xcite .
we focus on the measurements of the equivalent widths of the [email protected],2803.53 ( @xmath11 ) and [email protected],2600.73 ( @xmath12 ) transitions ( see , e.g. , * ? ? ?
* ; * ? ? ?
we present the sample in 2 and our analyses in 3 and show that there is evidence for a trend of decreasing @xmath13 with redshift . in
4 we discuss the possible origins of this trend .
we consider the caveats and whether the ratio is a viable proxy for understanding the abundance ratio of fe to mg .
we suggest that the observed trend reflects the possible cosmic evolution of the fe / mg ratio and is consistent with the predictions of a simple model where the abundance ratio variations are driven by the ratio of the rates of snia to ccsn . throughout this paper ,
we adopt a cosmology with [ @xmath14 = [ 0.3,0.7,0.8,0.7 ] .
magnitudes are presented in ab units @xcite unless otherwise specified .
in this work , we make use of the jhu - sdss metal absorption line catalog ( * ? ? ?
* ; hereafter zm13 ) based on the sdss dr7 release @xcite .
this catalog of 35,752 mgii absorption line systems is constructed from 84,533 quasar spectra from sdss dr7 and covers the redshift range between about 0.4 and 2.2 ( determined by the wavelength coverage of the sdss spectrographs ; york et al .
the window function , completeness and purity of the catalog were determined and cross - checked using different approaches , including monte carlo simulation for the whole sample and redshift path determination for individual quasars ( see zm13 for details ) .
the statistical incidence rate of the mg ii doublet as a function of redshift and strength was also presented in zm13 and compared with the literature .
the systematic effects in the catalog are also well understood .
the primary systematic effects are : ( 1 ) the imperfect calibration of the relative spectrophotometry , due to imperfect f - star sed models ; ( 2 ) the absorption induced by ism / cgm of the milky way at given observer - frame wavelengths , most importantly at 3934 and 3969 ( due to ca ii ) , 5891 and 5897 ( due to na i ) , and various diffuse interstellar bands ; ( 3 ) poor subtraction of the telluric emission lines , which were , on average , under - subtracted in the reduction pipeline ; and ( 4 ) relatively low completeness at both the short and long wavelength ends due to the reduced throughput at the blue end of the sdss spectrographs and the poor subtraction of the bright oh lines at the red end .
finally , comparisons of the zm13 catalog to an independently constructed dr7 catalog find good agreement in the measured equivalent widths in the two catalogs , with the difference in @xmath5 measurements less than @xmath15 @xcite .
the zm13 catalog therefore represents a well - understood dataset that spans a large range in redshift ; the known systematics have negligible effects on any of the statistics derived here .
the spectrophotometric calibration errors are at @xmath16 level ( about an order of magnitude smaller than the photon poisson noise ) and therefore have no effect on the statistics presented here .
the absorption by galactic ca ii can be strong along some lines of sight and does affect the detection efficiency of extragalactic mg ii absorbers below redshift 0.43 .
we therefore choose not to include the absorbers at @xmath17 in our sample .
in addition , we restrict the fits and statisitical analyses to @xmath18 to ensure all detected feii@xmath02600 lines lie in the region of higher completeness at @xmath19 .
the terrestrial lines are known to be under - subtracted and the noise ( i.e. , flux errors ) at corresponding wavelengths were adjusted in accordance in the sdss reduction pipeline .
the mg ii detection pipeline has been fully - tested on cases with poor sky subtraction and shown to perform at the same efficiency as along other sightlines . these tests are discussed in detail in @xcite and @xcite . at the red end , we restrict our analyses to @xmath20 , which corresponds to mgii detections at roughly @xmath21 .
we extracted all measurements of the mgii @xmath12796.35,2803.53 , mgi@xmath02853 , feii @xmath12586.00,2600.73 absorption lines .
the catalog contains a total of 35,752 absorption line systems , @xmath2297% of which are strong " absorbers with @xmath23 .
since a few of these are only detected marginally , we restricted our analyses to a subset of 35,049 systems with absorption line signal - to - noise ratios of @xmath24 , where @xmath5 is the sum of the equivalent widths of the mgii 2796.35 and 2803.53 resonance lines and @xmath25 is the measurement uncertainty ; 99.5% of these have @xmath23 .
no restriction was placed on the strength of the feii line ( hereafter feii refers to the 2600.73 line ) .
31,727 of the systems lie at @xmath26 and constitute our primary statistical sample . some data in the catalog required editing .
2,554 sources have values of @xmath27 , 816 of which lie at @xmath28 ( and therefore outside our main sample ) and the rest of which are distributed in redshift similarly with the overall sample of absorbers .
7 more sources have @xmath29 ( most with meaningless values of many 1000s ) and 3 have @xmath30 . in all these systems , we considered feii to be undetected and , for the purposes of our analyses , we set @xmath31 , the minimum detectable equivalent width in the catalog , and the uncertainty @xmath32 . in all cases ,
this resulted in @xmath33 . within @xmath26 ,
the redshift distribution of the @xmath34 @xmath35 is very similar to that of the @xmath36 equivalent widths .
for each absorber , we defined the ratio @xmath2 .
the distribution of @xmath8 shows a broad distribtion which is roughly gaussian with a weak tail to high values of @xmath8 . of the 31,727 absorbers , 4,965 ( 15.6% ) have @xmath37 .
the redshift distribution of these sources is similar to the overall distribution of absorbers . given the large scatter in the points and the non - detections of feii , we considered multiple approaches to determining whether there is a trend of @xmath8 with redshift and estimating its significance . a least - absolute deviation fit to the overall point distribution using 1000 bootstrap samples ( with replacement ) results in a median slope of @xmath38 .
the 95% confidence limits on the slope are [ -0.040,-0.052 ] . given the significant fraction of feii non - detections , we also measured the linear regression using the expectation - maximization ( em ) technique and schmitt s binned linear regression method , as implemented in the asurv package @xcite . all points with @xmath37
were treated as upper limits .
the em method resulted in a slope estimate of @xmath39 .
schmitt s method was used with a redshift binning of @xmath40 and @xmath8 binning of @xmath41 and resulted in a slope estimate of @xmath42 .
we investigated different bin sizes , but the results did not change significantly . finally , we used the bayesian approach to linear regression developed by @xcite .
this was very cpu - intensive and slow for our sample size , but resulted in a median slope of @xmath43 with a 95% confidence range of [ @xmath44 .
we next investigate the redshift evolution in the shape of the distribution of @xmath2 . in this approach ,
non - detections were just placed at their `` measured '' or edited value ; given that the non - detections follow the redshift distribution of the detected sample , any widening of the overall distribution is likely to be similar in all redshift bins . following zm13
, we divided the sample into 12 redshift bins defined between the redshifts @xmath3=[0.43,0.55,0.70,0.85,1.00,1.15, ... ,2.05,2.30 ] and measured the distribution of the ratio @xmath2 in each bin .
we modeled the distribution in each redshift bin as a gaussian ( fit using a non - linear least squares method ) and found the best fit value for the gaussian mean .
we also computed the median value of the distribution within each bin .
[ histograms ] shows the distributions and the fits in a few sample redshift bins .
we then performed linear fits ( using least absolute deviation and least squares techniques ) to the median and mean values to measure the variation of @xmath8 with redshift . in order to understand the uncertainty
that may be inherent in this analysis due to the effects of a complex and censored sample , we carried out a bootstrap analysis by selecting samples using replacement from the data in each redshift bin .
for each bootstrap sample , we carried out the same analyses ( i.e. , measuring medians , fitting gaussians , and fitting the median and means as a linear function of redshift ) .
the results for the primary sample ( i.e. , 31,727 @xmath45 absorbers ) are shown in figure [ absratioplot ] .
the distribution of points is shown as a greyscale with the median and quartiles represented as thin black lines .
the best fit peak and median values in each redshift bin are shown as large filled triangles and circles , respectively , along with the uncertainties derived in the bootstrap analyses .
the peak and median values of the @xmath8 ratio distribution trend to lower values with increasing redshift . the same behavior is reflected in the variation of the quartiles with redshift
. a linear fit to the peak values of the gaussian distribution within the redshift range @xmath46 ( i.e. , using redshift bins with @xmath22900 absorbers ) results in @xmath47 with 95% confidence limits of [ @xmath48 ; a fit to the median values results in a similar slope of @xmath49 .
the observed trend is not due to the censorship of the data . for our primary sample
we used the least censored subset .
however , we also investigated subsamples selected with a range of signal - to - noise ratios @xmath50}$ ] and found significant trends ( with @xmath8 decreasing with increasing redshift in all cases ) and with slopes ( derived for fits to the median values ) ranging from @xmath51 to @xmath52 , with a median derived slope value of @xmath53 .
the steepest ( shallowest ) slope values are found for the @xmath54 and @xmath55 ( @xmath56 and @xmath57 ) sample containing 20,813 ( 2,962 ) absorbers ( see table 1 ) . [ cols="^,^,^,^,^,^,^,^ " , ] while higher snr thresholds tend to select higher equivalent width lines , the correspondence between the two quantities is complicated and the equivalent width distribution at a given signal - to - noise threshold varies in a complicated way with wavelength .
hence , in order to further investigate whether the trend is dominated ( or damped ) by stronger absorbers , we investigated subsamples with a range of @xmath5 and @xmath35 equivalent width cutoffs , selecting samples with equivalent width cuts @xmath22 [ 0,0.5,1.0,1.5,2.0 ] in either line .
the derived slopes do not vary in a monotonic way with the equivalent width cuts , and there is no evidence from the current dataset that a subset of stronger or weaker absorbers are dominating the trend signal ( see also fig .
[ saturation ] right panel ) .
the trend of @xmath8 with redshift appears to be primarily driven by an evolution in the feii equivalent width . fitting the medians of the distributions of the mgii and feii equivalent widths as a function of redshift results in measured slopes of @[email protected] for @xmath5 vs. @xmath3 and @[email protected] for @xmath35 vs. @xmath3 .
the non - evolution in the slope of @xmath5 observed in our data is consistent with the results from previous high - spectral - resolution studies @xcite .
we have established that the absorption line systems listed in the jhu - sdss catalog exhibit a trend where the @xmath13 decreases with redshift .
although the catalog is constructed from the spectra of qsos observed uniformly with the sdss spectrographs , it is important to consider whether the observed trend could be the result of observational selection effects .
as an initial test , we investigated whether the line ratios mgii @xmath02796/mgii @xmath02803 and feii @xmath02586/feii @xmath02600 in our sample show significant variation with redshift ; they do not , and fits ( to snr@xmath63 samples ) result in slopes of @[email protected] and @[email protected] respectively ( see also the left two panels of fig .
[ saturation ] ) .
in addition , the mgi@xmath02853/mgii equivalent width ratio does not vary significantly with redshift ( slope of [email protected] ) . changing the subsamples by varying the signal - to - noise ratio cutoff does not change these results significantly , although in the case of the feii @xmath02586/feii @xmath02600 ratio , higher snr cuts result in a slightly more significant detection of a trend ( e.g. , for snr@xmath225 , feii @xmath02586/feii @xmath02600 vs. redshift has a slope of @xmath60 ) .
while the wavelength separation between these various features is smaller than that between the mgii and feii features , there does not appear to be any systematic effect introduced by the fact that the lines are measured in different parts of the spectrum .
the division in spectral coverage between the blue and red spectrographs at @xmath61 does result in a lower detection efficiency of absorption lines at these wavelengths and results in a slightly higher median equivalent width for the @xmath62 bin ; however , there is no obvious feature in the distribution of the line ratios and this bin does not affect the determination of the overall trend . in order to determine whether the complicated selection function for absorption line detection could be responsible for the trend , we used the measured selection function from zm13 to create 1000 simulated datasets as follows .
we simulated a large number of mgii absorption lines ( distributed randomly in redshift within each redshift bin ) using the best - fit intrinsic exponential distribution of rest equivalent width computed by zm13 ( see equation 3 and table 1 in zm13 ) .
for each mgii absorber , the @xmath63 doublet ratio was drawn from a lognormal model based on a fit to the observed distribution of the ratio in the jhu - sdss dataset .
then , for each mgii absorber , an feii absorption line was drawn randomly from a normal distribution of @xmath8 with a mean of 0.233 and @xmath64=0.143 . in order to test whether the selection would generate a systematic bias , the @xmath8 distribution was assumed to be constant with redshift .
equivalent width uncertainties for the mgii and feii lines were assigned randomly , drawn from model fits to the actual fractional error distributions in these two lines . in each case
, we constructed the two - dimensional distribution of @xmath65 versus @xmath66 and modeled the surface as a lognormal in @xmath65 and a polynomial in @xmath66 .
we then observed " this set of intrinsic mgii+feii absorbers using the equivalent width selection function for the jhu - sdss catalog ( zm13 ) .
we selected absorbers and fit each of the 1000 such datasets in the same manner in which we fit the observed catalog and derived slope measurements for any gradient in @xmath13 ratio with redshift .
since there was no evolution assumed in the data , any resulting slope is purely due to selection effects ( i.e. , completeness as a function of equivalent width and observed wavelength ) and the variation in the intrinsic distribution of @xmath5 with redshift .
we do not detect the observed slope in the simulated @xmath8 distribution with a high degree of reliability .
the 1000 trials resulted in a normal distribution of slopes with a small negative bias ( median recovered slope = @xmath67 ; standard deviation of 0.0033 ) ; the 95% ( 99% ) confidence limits on the slope are [ @xmath68 ( [ @xmath69 ) and none of the 1000 trials resulted in slope determinations @xmath70 .
we therefore conclude that the observed trend is not the result of selection effects .
our results show that the ratio of the equivalent widths of feii to mgii absorption lines along the lines of sight to qsos decreases with redshift .
the trend survives a range of signal - to - noise ratio or equivalent width cuts and does not appear to be due to selection effects within the jhu - sdss zm13 dataset . in this section ,
we discuss possible physical origins for the equivalent width ratio evolution .
the strength of an isolated absorption line as measured by its equivalent width is related to its abundance in the gas phase and the velocity spread of the gas @xcite .
it is tempting to interpret the @xmath13 ratio in terms of the fe / mg abundance ratio .
however , in order to interpret the observed equivalent widths in terms of physical quantities associated with the absorbing clouds , we need to consider the effects of ( i ) saturation , ( ii ) kinematics , ( iii ) ionization , and ( iv ) depletion .
we consider each of these in turn . _
( i ) line saturation : _ the @xmath5 distribution for the snr@xmath63 sample is a lognormal which peaks at @xmath71 , implying that the individual doublet lines have typical equivalent widths @xmath221 and lie in the logarithmic ( i.e. , saturated " ) part of the curve of growth .
it is therefore not obvious that we can interpret any observed trend with redshift as resulting from a trend in the abundance ratio of mg to fe without first understanding the effects of line saturation . in order to investigate the effects of saturation on the observed ratio
, we computed the curves of growth for mgii and feii for two different values of the doppler parameter , @xmath72=10 km s@xmath73 and 50 km s@xmath73 . for each case , we computed the expected @xmath8 values that result from a wide range in mgii column density @xmath74 and [ mg / fe ] varying between 1 and 10 times the solar value . for a given @xmath72 ,
even strongly saturated lines result in the equivalent width ratio @xmath8 depending on the [ mg / fe ] abundance ratio .
hence , even if the observed absorption lines are single and saturated , they only dampen , but not erase , an intrinsic variation of the abundance ratio .
most mgii absorption systems with @xmath75 are multiple systems ( e.g. , * ? ? ?
* ) , and this complicates the simple picture described here . fully modeling
the result of saturation and multiplicity is beyond the scope of this paper , but instead we look to the data for clues to whether saturation effects are responsible for the observed trends .
[ saturation ] shows the effect of saturation on three pairs of equivalent width ratios in our observational dataset .
the @xmath63 and @xmath76 ratios ( shown in the first two panels ) asymptote toward the saturation value with increasing @xmath5 .
however , neither of these ratios shows any evolution with redshift .
in contrast , the @xmath13 ratio clearly evolves , with the median line ratio decreasing with increasing redshift in _ all _ @xmath5 bins .
[ saturation ] also shows that the higher equivalent width bins also exhibit the evolution with redshift , albeit at lower strength than the lower equivalent width ( and less saturated ) bins .
this is consistent with our expectation that saturation effects will tend to weaken the relationship between @xmath8 and the intrinsic feii / mgii abundance ratio of the absorbers , and then that these effects are likely only to dampen ( rather than induce ) a potential trend in @xmath8 with redshift . _
( ii ) multiplicity & kinematics : _ if the observed absorption lines are instead the result of blends of multiple unresolved ( lower equivalent width ) lines , then the individual lines might arise in lower column density clouds and are therefore presumably less saturated .
the real situation is likely to be more complicated , with lines forming in clouds with varying kinematic properties .
high - resolution studies have demonstrated that mgii lines are often made up of a number of components , typically between 2 and 10 , with the number of components increasing with @xmath5 @xcite .
however , they also find that the mgii and feii absorption lines show similar doppler parameters and correlated column densities and the distribution of @xmath77 column ratios does not depend strongly on velocity spread @xcite , nor does the kinematic spread show any evidence for significant evolution with redshift @xcite .
hence , assuming the feii and mgii absorption lines are equally sensitive to the variations in the clouds as they evolve , the evolution of the ratio @xmath13 for individual absorption systems should change monotonically with the abundance ratio @xmath78 , although the magnitude of the change may be dampened by saturation effects .
if , for some reason , the velocity structure or doppler parameter for these two species differs , due to e.g. , differences in ionization structure or incomplete mixing of metal species , this monotonic relation can break down . for example , incomplete metal mixing could arise in a delayed fe - enrichment scenario : if the kinematic spread of fe - rich material is narrower than that of @xmath10-enriched gas , this could give rise to a suppressed @xmath13 ratio without it reflecting a true dearth of fe atoms .
however , even if this somewhat contrived scenario were realized , it would still point to an evolution in the sources of cgm enrichment . while it may be possible to create a trend in @xmath8 with redshift as a result of an evolving @xmath72
, there is no obvious reason why the doppler parameter would evolve with redshift for just one species .
better spectroscopic data ( higher spectral resolution , and on more species and ionization states ) on much larger samples than currently exist can help explore this issue further . _
( iii ) ionization correction : _ the first two ionization potentials for mg and fe are very similar : for mgi@xmath79mgii and fei@xmath79feii , they are 7.646 ev and 7.902ev respectively ; the second ionization potentials ( i.e. , mgii@xmath79mgiii and feii@xmath79feiii ) are 15.0357 ev and 16.1878 ev respectively @xcite .
the ratio of the ionic abundances of the two elements , @xmath80 , is determined by the shape ( at these energies ) of the spectrum of the ionizing metagalactic radiation , which does not vary significantly over the redshift range considered here . using the @xcite model for the uv metagalactic flux as a function of redshift
, we find that the ratio of the emissivity at the mg and fe 1st ionization potentials varies by less than 4% for @xmath81 ; the emissivity ratio at the wavelengths of the 2nd ionization potential varies by less than 2% .
the bulk of the absorbers ( @xmath2296% ) in the jhu - sdss metal absorption line catalog have @xmath82 .
these absorbers therefore correspond to regions with hydrogen column densities of @xmath83 ( if the doppler parameter is 10 km s@xmath73 ; * ? ? ?
@xcite calculate the variation of the n(mgii)/n(feii ) ratio with ionization parameter for the @xcite metagalactic radiation field .
they find that for hydrogen column densities @xmath84 , the n(mgii)/n(feii ) ratio ( and hence its inverse ) is essentially constant over a wide range of ionization parameter ( see also * ? ? ?
it is therefore unlikely that differences in ionization could drive the observed trend . _
( iv ) depletion : _
mgii absorption line systems are known to contain dust , exhibiting dust - to - gas ratios comparable to that observed for normal galaxies @xcite . since both mg and fe can be depleted onto dust grains ,
it is necessary to understand whether differing depletion rates for the two elements as a function of redshift can result in the trend observed in their equivalent width ratios . based on studies of the interstellar medium in our galaxy , the depletion rate for fe is approximately 2@xmath587 times larger than that for mg ( e.g. , * ? ? ?
* ) , i.e. , fe is depleted from the gas phase at a higher rate than is mg . using the tabulated model of @xcite for the dependence of the mg and fe gas phase abundance on the line - of - sight depletion strength @xmath85 , we find that the column density ratio @xmath86 varies as @xmath87 , i.e. , the gas - phase column density ratio increases with increasing depletion and dustier absorbing clouds ( which would presumably be more depleted in mg and fe ) would exhibit higher values of @xmath86 .
hence , the only way to reproduce a trend where the @xmath13 ratio decreases with redshift through purely dust depletion effects is to have higher redshift absorbers be dustier ( and more depleted ) than lower redshift absorbers , an unlikely scenario .
in contrast , we expect fe and dust content to increase as the universe ages and that fe is likely to be fractionally more depleted onto dust at lower redshift .
hence , any detection of an increase ( to lower redshift ) in the observed gas - phase fe / mg abundance ratio is likely to reflect a stronger intrinsic increase in the abundance ratio . in other words
, a measurement of the gas - phase fe / mg ratio would underestimate the true fe / mg , and the factor by which it is underestimated would be larger at low redshift than at high redshift .
this goes in the sense , therefore , of making the observed relation imply a stronger underlying variation .
it is therefore unlikely that depletion effects could explain away the observed trend .
in contrast to various possible ( and somewhat contrived ) explanations discussed above , the expected variation of the rates of snia and ccsn should naturally produce the observed variation in @xmath8 and we therefore consider this next .
we model the average universal fe / mg abundance ratio as resulting from the supernovae formed at each epoch as a result of the cosmic star formation rate history .
we assume that the rate of core collapse supernovae follows the cosmic star formation rate density ,
i.e. , @xmath88 ( e.g. , * ? ? ?
* ) . since ccsn progenitors are massive ( @xmath89 ) stars , the rate of ccsn is closely tied to the cosmic star - formation rate density in the universe , i.e. , @xmath90 where @xmath91 is the cosmic star formation rate density ( in units of @xmath92 ; * ? ? ?
@xmath93 assumes a salpeter initial mass function with @xmath94 @xmath95@xcite . in contrast , snia rates exhibit a `` delay time distribution '' ( dtd ) relative to the time since star formation ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
since there is much debate about the exact form of this distribution , here we assume the simple power - law form suggested by @xcite : @xmath96 where @xmath97 .
we compute the evolving mass density of each element as the sum of the sn yields : @xmath98 where @xmath99 is the average mass yield for metal @xmath100 ( in @xmath95 ) for a sn of type @xmath101 .
the number density of element @xmath100 is @xmath102 , where @xmath103 is the relative atomic mass of element @xmath100 .
the sn yields are very uncertain : based on a review of the literature , @xcite conclude that calculated nucleosynthetic yields from sne vary by factors of @xmath1042 . here
, we assume that the typical snia produces @xmath105@xmath95 of mg and 0.74@xmath95 of fe , and that the typical ccsn produces 0.12 @xmath95 of mg and 0.09 @xmath95 of fe @xcite . choosing a single number for the average yields is a simplification , but our aim here is to explore whether the overall variation observed in the @xmath13 ratio
can be explained by just the simple evolution of the sn rates .
the results are shown in fig .
[ models ] .
the solar abundance ratio of @xmath106 is also shown ( based on the mg and fe photospheric abundances of [email protected] and [email protected] respectively ; * ? ? ?
this simple model over - predicts the observed snia rates by a factor of 2 and over - predicts the solar photospheric abundance ratio of fe / mg at redshift @xmath107 by @xmath108 .
this is not surprising given the large uncertainties in the yields and our other assumptions ( e.g. , ignoring the contribution from agb stars , uncertainties / variations in the imf , dtd , sfrd , and the large scatter in the data ) .
rate by a factor of 0.6 can result in both a better fit to the observed rates and the solar abundance ratio . ]
recently , @xcite modeled the metal enrichment in a galaxy resulting from agb stars and sne , investigating different assumptions for sn yields and snia dtds .
they find that the observed local mass - metallicity relationships and the [ o / fe ] abundance in the milky way can be reproduced by models in which the snia dtd is a power - law and the light @xmath10 elements are expelled in galactic winds .
the redshift variation of the fe / mg ratio computed by @xcite for a model milky way - type galaxy using a power - law dtd is also shown in fig . [ models ] ; rather remarkably , its slope matches the evolution in @xmath8 observed in the jhu - sdss dataset .
this lends further , though still circumstantial , support that the observed evolution can be explained as the result of the increasing dominance of snia yields changing the mean composition of the outflowing gas responsible for the absorption line systems over time .
the comparisons of the models to the slopes measured in the data are only meant to be illustrative and not discriminatory .
the scatter in the slope measurement is large and the models are only notional at this stage .
nevertheless it is instructive that the measured variation shows slopes comparable to those predicted by models where all the evolution in the abundance ratios is the result of star formation in galaxies and enriched material ejected into their gaseos halos .
the large scatter in the data may reflect the diversity of star formation histories , impact parameters of qso lines of sight through halos , and inhomogeneities in the halo gas .
individual absorbers at any given redshift show significant scatter and the trend ( of @xmath8 with @xmath3 ) reported here is subtle ; it was only possible to measure the trend using the very large catalog of absorbers resulting from the sdss qso spectra .
it is therefore not surprising that previous high - resolution studies ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) did not see the trend of @xmath8 with redshift in the small samples ( @xmath109 absorbers ) studied ( however , please see @xcite s analysis of the @xcite dataset , where a possible trend of decreasing @xmath110 is reported ) . in a
vlt / uves high - resolution study of 100 weak ( i.e. , @xmath111 ) mgii absorbers , @xcite found that clouds with high @xmath13 disappear from their sample at high redshift ( @xmath112 ) . in a complementary study of 81 strong absorbers identified in the same dataset
, @xcite found that the @xmath13 ratio for @xmath113 absorbers is typically high and exhibits some evolution with redshift ; weaker absorbers ( 0.3@xmath114 ) instead showed a wider range and scatter in the @xmath13 ratio with no obvious trend .
the authors interpret this as primarily due to kinematics : high - redshift absorbers with @xmath113 tend to be formed in regions with more kinematic spread ( and the feii and mgii lines are less saturated ) than their low - redshift counterparts .
they also consider the effects of ionization and @xmath10-enhancement , but conclude that they can not discriminate between these effects .
the sdss spectra have insufficient resolution to explore this issue adequately @xcite .
however , splitting the sample into two equivalent width bins 0.3@xmath115 and @xmath116 following @xcite we find that both subsets show evolution ( with slopes and 95% confidence ranges of @xmath117 , [ @xmath118 and @xmath119 , [ @xmath120 respectively ) .
there is also evidence that the @xmath8 values show a much broader distribution for the 0.3@xmath115 subset than for the @xmath121 subset .
it is not obvious from the large sdss dataset , therefore , whether splitting the data by equivalent width is justified or if the absorber population should be treated as a continuum , reflecting a range of gas physical conditions in the halos of galaxies .
given the various caveats already described in using low - resolution data , it will be necessary to obtain high resolution spectroscopic observations of a sufficiently large sample ( @xmath1221000s ) of absorbers to properly disentangle the effects of saturation , kinematics and abundance .
a more detailed investigation of the scatter in the equivalent width and ratio distributions compared to the results from numerical simulations will help us further understand the range of physical conditions in galaxy halos .
we have investigated the equivalent width ratio @xmath123 for a large sample of mgii absorption line systems from the jhu - sdss metal absorption line catalog of @xcite and discovered that the ratio shows a trend with redshift .
this trend does not appear to be the result of data artifacts or selection effects in the catalog .
several reasons may exist for the observed evolution .
however , a simple model shows that the trend is naturally explained by the evolving fe / mg abundance ratio that results from the changing relative rates of type ia and core - collapse sne .
star - formation ( and hence supernovae production ) is primarily located in galactic disks , whereas the absorption lines under investigation here form in circumgalactic clouds in galaxy halos at large galactocentric distances .
hence , if the equivalent width ratio reflects abundance variations in the cgm clouds , and if the enrichment by supernovae is indeed the primary cause for the cgm metallicity variations , then outflows must be efficient in transporting metals out of star - forming disks to large galactocentric radii .
future spectroscopic surveys with the next generation of highly multiplexed multi - object spectrographs ( e.g. , desi on the mayall telescope , pfs on subaru , and 4most on vista ) will result in larger samples of absorption line systems and the ability to explore the chemical evolution of the universe in greater detail .
higher signal - to - noise ratio and higher resolution spectroscopy of large samples spanning a large redshift range can confirm whether the trend observed in the fe / mg ratio is observed in other @xmath10-element lines and explore whether the changes are due to the chemical enrichment history of the universe ( as suggested here ) or if evolution in the cloud properties ( e.g. , ionization , kinematics , doppler parameter , substructure , etc . ) is instead responsible for the trend .
detailed comparisons between the data and state - of - the - art cosmological simulations ( e.g. , * ? ? ?
* ) will be critical to this effort .
this research made use of the jhu - sdss metal absorption line catalog @xcite , which is based on the sdss dr7 release @xcite , and the nist atomic spectra database @xcite .
we thank dr .
brice mnard for useful advice and for putting together such a scientifically useful resource .
ad thanks joan najita for assistance with the asurv package and tom matheson for useful comments on the manuscript .
we are grateful to the referee for a constructive report that resulted in improving this paper .
funding for the sdss and sdss - ii was provided by the alfred p. sloan foundation , the participating institutions , the national science foundation , the u.s .
department of energy , the national aeronautics and space administration , the japanese monbukagakusho , the max planck society , and the higher education funding council for england .
the sdss web site is http://www.sdss.org/. gbz acknowledges partial support provided by nasa through hubble fellowship grant # hst - hf2 - 51351 awarded by the space telescope science institute , which is operated by the association of universities for research in astronomy , inc .
, under contract nas 5 - 26555 .
ad s research activities are supported by the national optical astronomy observatory ( noao ) .
ad thanks the radcliffe institute for advanced study and the institute for theory and computation at harvard university for their generous support during the period when this paper was written .
noao is operated by the association of universities for research in astronomy ( aura ) under cooperative agreement with the national science foundation .
abazajian , k. n. , adelman - mccarthy , j. k. , ageros , m. a. , et al .
2009 , apjs , 182 , 543 aller , l. h. 1942 , apj , 96 , 321 asplund , m. , grevesse , n. , sauval , a. j. , & scott , p. 2009
, ara&a , 47 , 481 bahcall , j. n. , & spitzer , l. , jr .
1969 , apj , 156 , l63 baker , j. g. 1936 , apj , 84 , 474 bergeron , j. , & boiss , p. 1991 , a&a , 243 , 344 bordoloi , r. , lilly , s. j. , knobel , c. , et al .
2011 , apj , 743 , 10 churchill , c. w. , mellon , r. r. , charlton , j. c. , et al .
2000 , apj , 543 , 577 churchill , c. w. , vogt , s. s. , & charlton , j. c. 2003 , aj , 125 , 98 claeys , j. s. w. , pols , o. r. , izzard , r. g. , vink , j. , & verbunt , f. w. m. 2014 , a&a , 563 , a83 crighton , n. h. m. , hennawi , j. f. , simcoe , r. a. , et al . 2015 , mnras , 446 , 18 dalla vecchia , c. , & schaye , j. 2008 , mnras , 387 , 1431 dav , r. , finlator , k. , & oppenheimer , b. d. 2011 , mnras , 416 , 1354 erb , d. k. , quider , a. m. , henry , a. l. , & martin , c. l. 2012 , apj , 759 , 26 finlator , k. , & dav , r. 2008 , mnras , 385 , 2181 gauthier , j .-
r . , chen , h .- w .
, cooksey , k. l. , et al .
2014 , mnras , 439 , 342 giavalisco , m. , vanzella , e. , salimbeni , s. , et al .
2011 , apj , 743 , 95 haardt , f. , & madau , p. 1996
, apj , 461 , 20 haardt , f. , & madau , p. 2012 , apj , 746 , 125 hartigan , p. , morse , j. a. , tumlinson , j. , raymond , j. , & heathcote , s. 1999 , apj , 512 , 901 isobe , t. , feigelson , e. d. , & nelson , p. i. 1986 , apj , 306 , 490 jenkins , e. b. 2009 , apj , 700 , 1299 kelly , b. c. 2007 , apj , 665 , 1489 kornei , k. a. , shapley , a. e. , martin , c. l. , et al .
2012 , apj , 758 , 135 kramida , a. , ralchenko , yu . ,
reader , j. , and nist asd team ( 2013 ) .
nist atomic spectra database ( ver .
5.1 ) : http://physics.nist.gov/asd [ 2014 , june 4 ]
. national institute of standards and technology , gaithersburg , md .
lan , t .- w . , mnard , b. , & zhu , g. 2014 , apj , 795 , 31 lavalley , m. , isobe , t. , & feigelson , e. 1992 , astronomical data analysis software and systems i , 25 , 245 madau , p. , & dickinson , m. 2014 , araa , 52 , in press ; arxiv:1403.0007 mannucci , f. , cresci , g. , maiolino , r. , marconi , a. , & gnerucci , a. 2010 , mnras , 408 , 2115 maoz , d. , & mannucci , f. 2012 , pasa , 29 , 447 maoz , d. , mannucci , f. , & brandt , t. d. 2012 , mnras , 426 , 3282 martin , c. l. , shapley , a. e. , coil , a. l. , et al .
2012 , apj , 760 , 127 mnard , b. , & chelouche , d. 2009 , mnras , 393 , 808 mnard , b. , & fukugita , m. 2012 , apj , 754 , 116 menzel , d. h. 1936 , apj , 84 , 462 morton , d. c. 2003 , apjs , 149 , 205 mshar , a. c. , charlton , j. c. , lynch , r. s. , churchill , c. , & kim , t .- s .
2007 , apj , 669 , 135 narayanan , a. , charlton , j. c. , misawa , t. , green , r. e. , & kim , t .- s .
2008 , apj , 689 , 782 nomoto , k. , iwamoto , k. , nakasato , n. , et al . 1997 , nuclear physics a , 621 , 467 oke , j. b. & gunn , j. e. 1983 , apj , 266 , 713 oppenheimer , b. d. , & dav , r. 2006 , mnras , 373 , 1265 prochaska , j. x. , kasen , d. , & rubin , k. 2011 , apj , 734 , 24 rodrguez hidalgo , p. , wessels , k. , charlton , j. c. , et al . 2012 , mnras , 427 , 1801 rubin , k. h. r. , prochaska , j. x. , koo , d. c. , et al .
2014 , apj , 794 , 156 savage , b. d. , & sembach , k. r. 1996 , ara&a , 34 , 279 seyffert , e. n. , cooksey , k. l. , simcoe , r. a. , et al .
2013 , apj , 779 , 161 srianand , r. 1996 , apj , 462 , 643 steidel , c. c. 1995 , qso absorption lines , 139 steidel , c. c. , & sargent , w. l. w. 1992 , apjs , 80 , 1 suresh , j. , bird , s. , vogelsberger , m. , et al .
2015 , mnras , 448 , 895 thielemann , f .- k . , argast , d. , brachwitz , f. , et al . 2003 , nuclear physics a , 718 , 139 torrey , p. , vogelsberger , m. , genel , s. , et al . 2014 , mnras , 438 , 1985 totani , t. , morokuma , t. , oda , t. , doi , m. , & yasuda , n. 2008 , pasj , 60 , 1327 tremonti , c. a. , heckman , t. m. , kauffmann , g. , et al .
2004 , apj , 613 , 898 vogelsberger , m. , genel , s. , springel , v. , et al .
2014 , nature , 509 , 177 weiner , b. j. , coil , a. l. , prochaska , j. x. , et al .
2009 , apj , 692 , 187 wiersma , r. p. c. , schaye , j. , theuns , t. , dalla vecchia , c. , & tornatore , l. 2009 , mnras , 399 , 574 yanny , b. , york , d. g. , & williams , t. b. 1990 , apj , 351 , 377 yates , r. m. , henriques , b. , thomas , p. a. , et al .
2013 , mnras , 435 , 3500 york , d. g. , adelman , j. , anderson , j. e. , jr . , et al .
2000 , aj , 120 , 1579 zhu , g. , & mnard , b. 2013 , apj , 770 , 130 zhu , g. , mnard , b. , bizyaev , d. , et al .
2014 , mnras , 439 , 3139 | we investigate the variation of the ratio of the equivalent widths of the feii@xmath02600 line to the mgii@xmath12796,2803 doublet as a function of redshift in a large sample of absorption lines drawn from the johns hopkins university - sloan digital sky survey absorption line catalog .
we find that despite large scatter , the observed ratio shows a trend where the equivalent width ratio @xmath2 decreases monotonically with increasing redshift @xmath3 over the range @xmath4 .
selecting the subset of absorbers where the signal - to - noise ratio of the mgii equivalent width @xmath5 is @xmath63 and modeling the equivalent width ratio distribution as a gaussian , we find that the mean of the gaussian distribution varies as @xmath7 .
we discuss various possible reasons for the trend . a monotonic trend in the fe / mg abundance ratio
is predicted by a simple model where the abundances of mg and fe in the absorbing clouds are assumed to be the result of supernova ejecta and where the cosmic evolution in the snia and core - collapse supernova rates is related to the cosmic star - formation rate .
if the trend in @xmath8 reflects the evolution in the abundances , then it is consistent with the predictions of the simple model .
intergalactic medium ; quasars : absorption lines ; galaxies : halos , abundances , evolution |
the cosmological matter - antimatter asymmetry is one of the most striking mysteries in the universe .
the five - year observations of the wmap colloboration precisely measured the ratio of baryon number to photon number densities as @xcite @xmath2 according to sakharov s suggestion @xcite , to dynamically generate the matter - antimatter asymmetry of the universe , three conditions must be satisfied , which are the violation of baryon number conservation , the violation of c and cp , and the deviation from thermal equilibrium . in the standard model ( sm ) , c parity is maximally violated and
the phase of the cabibbo - kobayashi - maskawa ( ckm ) matrix is the sole source of the cp violation .
however , this cp violation effect , which has been precisely measured in hadronic physics , is too small to be used to accommodate the baryon asymmetry of the universe .
such that , investigating the new origin of cp violation is a subject with great significance .
more recently , the d0 collaboration @xcite , with @xmath3 of data , has reported a measurement of the like - sign dimuon charge asymmetry in semi - leptonic decay of @xmath4 hadrons , @xmath5 where @xmath6 is the number of events with @xmath7 hadrons decaying semileptonically into @xmath8 .
this result , with the first error being statistical and the second systematic , is @xmath0 deviation from the sm prediction of @xmath9 @xcite .
@xmath10 was also measured by the cdf collaboration @xcite , which has @xmath11 , using @xmath12 of data .
this result is positive but still compatible with the d0 measurement at @xmath13 level because its uncertainties are 4 times larger than those of d0 . combining in quadrature these two results , one has @xmath14 @xcite which is still @xmath15 away from the sm value .
if confirmed , it will be an evidence of new physics beyond the sm .
there have already been some models @xcite attempting to explain the data .
the d0 collaboration s result is the first laboratory evidence for significant matter - antimatter asymmetry .
we suspect there should be some connection between the baryon asymmetry of the universe and this anomaly . in this paper
, we consider the possible extension of the sm with one color - triplet charged scalar and three majorana fermion singlets . in this model , there are tree level processes leading to dimension six effective operators , like @xmath16 , which may contribute to @xmath17 mixing and thus the like - sign dimuon charge asymmetry can be explained . besides
, the out of equilibrium decay of the neutral fermion may produce @xmath18 asymmetry .
this asymmetry may be converted to the baryon asymmetry of the universe via sphaleron process @xcite , which violates @xmath19 but keeps @xmath18 conservation .
the paper is organized as follows : in section ii , we present our model .
section iii is devoted to investigating the like - sign dimuon charge asymmetry .
we study baryogenesis in section iv .
conclusions and remarks are presented in section v.
to reproduce the large anomalous like - sign dimuon charge asymmetry observed by d0 and to explain the matter - antimatter asymmetry of the universe , new physics beyond the sm is needed . in this section ,
we consider the possible extension of the sm with one charged scalar @xmath20 and three fermion singlets @xmath21 , whose representation in @xmath22 gauge group are @xmath23 @xmath24 and @xmath25 @xmath26 , respectively . for simplification , @xmath27 discrete flavor symmetry
is introduced and the presentations of fields on @xmath27 are listed in table i. the newly introduced lagrangian can be written as @xmath28 where @xmath29 and @xmath30 are yukawa coupling matrices , @xmath31 is the sm higgs boson , @xmath32 and @xmath33 represent the second or third generation right - handed up- and down - type quarks .
assuming @xmath20 has baryon number @xmath34 , the interaction , @xmath35 , violates baryon number by @xmath36 unit .
.representations of fields on @xmath27 flavor symmetry . [ cols="^,^",options="header " , ] note that @xmath27 symmetry is explicitly broken down by the yukawa interactions containing the firt generation right - handed quarks and the mass term of @xmath21
. we may introduce another higgs doublet , whose representation under @xmath27 is 1 , to genreate masses for the first generation up- or down - type quarks , and itroduce one higgs singlet , whose representation unde @xmath27 is @xmath37 , to generate masses for @xmath21 .
such that @xmath27 symmetry can be recovered . due to the existence of @xmath27 symmetry , @xmath20 and @xmath21 do not couple to the first generation of quarks . as a result , we need not to warry about constraints from proton stability , @xmath38 mixing or @xmath39 mixing .
notice that @xmath20 is similar to the super - partner of @xmath40 and @xmath41 is similar to @xmath42-parity violating interactions in supersymmetry .
the experimental lower bound for the mass of @xmath20 should be consistent with that of squarks . integrating out heavy degrees of freedom , we can derive an effective operator that contributes to @xmath43 .
the inclusive decay width for this process can be written as @xmath44 where @xmath45 is the contribution of sm penguin diagram and @xmath46 \ ; , \end{aligned}\ ] ] comes from the new interaction in our model , with @xmath47 .
assuming @xmath48 and @xmath49 , we get the predication for the branching ratio of @xmath50 : @xmath51 , which is consistent with its present experimental value , @xmath52 , predicted by the heavy flavor average group ( hfag ) @xcite .
notice that , @xmath10 , appearing in eq .
( 2 ) , is blind as to which flavors of @xmath54 meson produced the two muon , it places a constraint on the semi - leptonic cp asymmetries of both @xmath55 and @xmath56 mesons , which we will call @xmath57 and @xmath58 , respectively .
the relation between @xmath10 and @xmath59 is given by @xmath60 taking into account the current experimental value of @xmath61 @xcite , one obtains @xmath62 , which is in agreement with d0 s direct measurement of @xmath63 @xcite , albeit with large uncertainties .
we may combine all these results to obtain an average value @xmath64 .
theoretically , there are two amplitudes characterizing mixing in @xmath56 system : the off - diagonal element of the mass matrix @xmath65 and the off - diagonal element of the decay matrix @xmath66 .
the violation of cp is caused by the non - zero value of the phase : @xmath67 . in terms of these parameters , to a very good approximation , @xmath68 is given by @xcite @xmath69 where @xmath70 and @xmath71 agrees with data very well , while both @xmath72 and @xmath73 differ from the experimental data by about @xmath74 and @xmath75 , respectively .
it is natural to attribute d0 s result to big corrections to @xmath76 and @xmath73 from new physics . in our model
, there are tree level contributions to the dimension six effective operator @xmath77 , which contributes to @xmath78 for @xmath53 mixing . integrating out @xmath20 at the tree level , we have @xmath79 where @xmath80 are generation indices and @xmath81 are color indices . starting with these four quark interactions , one can derive the expression of @xmath82 : @xmath83 where @xmath84 .
@xmath85 is the bag parameter and we take it to be equal to one .
there are also contributions to the off - diagonal elements , @xmath65 , of the mass matrix of the @xmath86 system due to the box diagram .
using vacuum insertion approximation , we have @xmath87 { \cal f}^ * ( x_\alpha^ { } , x_\beta^{})\ ; , \end{aligned}\ ] ] with @xcite @xmath88 where @xmath89 ( @xmath90 ) . in deriving eq .
( 10 ) , we have ignored the contributions from the @xmath91 term by assuming @xmath92 . as function of @xmath93 with @xmath94 being random parameter in the range
@xmath95 $ ] ( gev ) .
the horizontal line represents the central value of @xmath58 measured by @xmath96.,width=396,height=321 ] taking into account contributions from new physics , the final expressions for @xmath65 and @xmath66 can be written as @xmath97 where the formulae for @xmath98 and @xmath99 are given by @xcite @xmath100 \ ; .\end{aligned}\ ] ] since the couplings are in general complex , we can derive arbitrary phase @xmath73 in general . before giving some numerical analysis ,
let s assume that the yukawa coupling constant @xmath30 is real , and the operator @xmath101 dominates contributions to @xmath102 and @xmath103 . in this case , @xmath73 as well as @xmath58 will be functions of @xmath104 and @xmath94 .
we plot in fig . 1 , @xmath58 as function of @xmath104 with @xmath94 being a random parameter in the range @xmath95~ ( { \rm gev})$ ] .
the horizontal line represents the central value of @xmath58 measured by @xmath96 .
it is clear that , to fit the data , @xmath105 should be larger than @xmath106 .
of course , @xmath107 can be much smaller in the case @xmath30 being complex .
as noticed in the last section , the d0 s like - sign dimuon charge asymmetry can be explained by introducing a color - triplet scalar @xmath20 . as a byproduct , this may induce baryon number violating operators like @xmath108 , which is just what we need to reproduce the baryon asymmetry of the universe .
we assume @xmath21 is thermally produced in the early universe .
its out of equilibrium decay , as shown in fig 2 , may violate the @xmath18 symmetry .
the expression for the cp asymmetry @xmath109 in the decay of @xmath21 , which is similar to that in the decay of heavy majorana neutrinos in the type - i seesaw mechanism @xcite , can be written as @xmath110 \over [ \zeta \zeta^\dagger]_{11}^ { } } f{\left ( m_{\chi_i}^2 \over m_{\chi_1}^2 \right ) } \ ; , \end{aligned}\ ] ] where @xmath111 is given by @xmath112 \ ; .\end{aligned}\ ] ] after the evolution of boltzmann equations governing the baryon number density , we derive the @xmath18 asymmetry stored in right - handed quarks .
because all the quarks are in thermal equilibrium , this asymmetry may be converted to left - handed quarks through the left - right equilibrium @xcite , which can be understood as follows .
let us define the chemical potential associated with the @xmath113 field as @xmath114 , where @xmath115 is the chemical potential contributing to @xmath18 asymmetry and @xmath116 is independent of @xmath18 .
hence at equilibrium we have the chemical potential associated with @xmath117 given by @xmath118 .
thus we see that the same chemical potential is associated with @xmath117 as that of @xmath113 .
finally , the @xmath18 asymmetry is converted to the baryon asymmetry of the universe through the sphaleron process , which violates @xmath19 symmetry but keeps @xmath18 conservation .
then the final baryon asymmetry can be given as @xcite @xmath119 with @xmath120 which measures the effectiveness of decays at the crucial epoch ( @xmath121 ) when @xmath21 must decrease in number if they are to stay in equilibrium .
here , @xmath122 @xmath123 is the planck mass and @xmath124 is effective degrees of freedom , at the temperature where @xmath21 decouples . in deriving eq .
( 16 ) , we have assumed that the factor @xmath125 is greater than 1 , but not too large . if @xmath126 , @xmath127 .
for instance , inputting @xmath128 , k=50 , @xmath129 and maximal cp - phase , we derive the sample predication : @xmath130 . in consequence ,
we deduce @xmath131 , as desired .
we have extended the sm by introducing one color - triplet charged scalar , @xmath20 , and three majorana fermions , @xmath21 .
we have shown that the dimuon charge asymmetry , reported by the d0 collaboration , can be explained by the yukawa interaction , like @xmath132 . besides
, the matter - antimatter asymmetry of the universe can be realized by the out of equilibrium decays of @xmath21 .
notice that , the mass of @xmath20 can be several hundred @xmath133 .
it can be produced and detected at the large hadron collider .
a. dighe , a. kundu and s. nandi , arxiv:1005.4051 [ hep - ph ] ; b. a. dobrescu , p. j. fox and a. martin , arxiv:1005.4238 [ hep - ph ] ; c. h. chen and g. faisel , arxiv:1005.4582 [ hep - ph ] ; a. buras , s. gori , m. carlucci and g. isidori , arxiv:1005.5310 [ hep - ph ] ; z. ligeti , m. papucci , g. perez and j. zupan , arxiv:1006.0432 [ hep - ph ] ; m. jung , a. pich and p. tuzon arxiv:1006.0470 [ hep - ph ] ; k. s. babu and j. julio , arxiv:1006.1092 [ hep - ph ] ; u. nierste , arxiv:1006.2078 [ hep - ph ] ; b. batell and m. pospelov arxiv:1006.2127 [ hep - ph ] ; d. choudhury and d. k. ghosh , arxiv:1006.2171 [ hep - ph ] ; c. h. chen , c. q. geng and w. wang , arxiv:1006.5216 [ hep - ph ] ; j. k. parry , arxiv:1006.5331 [ hep - ph ] ; s. f. king , arxiv:1006.5895 [ hep - ph ] ; c. delaunary , o. gedalia , s. j. lee and f. perez , arxiv:1007.0243 [ hep - ph ] ; y. bai and a. e. nelson , arxiv:1007.0596 [ hep - ph ] ; j. kubbo and a. lenz , arxiv:1007.0680 [ hep - ph ] ; k. blum , y. hochberg and y. nir , arxiv:1007.1872 ; r. m. wang , y. g. xu , m. l. liu and b. z. li , arxiv:1007.2994[hep - ph ] ; c. berger and l. m. sehgal , arxiv:1007.2996[hep - ph ] ; b. dutta , y. mimura and y. santoso , arxiv:1007.3696[hep - ph ] ; m. gronau and j. l. rosner , arxiv:1007.4728[hep - ph ] ; t. gershon , arxiv:1007.5135[hep - ph ] ; a. j. buras , g. isidori and p. paradisi , arxiv:1007.5291[hep - ph ] ; a. kostelecky and r. v. kooten , arxiv:1007.5312[hep - ph]p ; s. oh and j. tandean , arxiv:1008.2153[hep - ph ] . l. d. faddeev , in _ proceeding of the ivth international conference on nonlocal field theories _ , dubna ussr , 1976 ( joint institute for nuclear research , dubna ) ; r. jackiw and c. rebbi , phys . rev . lett . *
37 * , 172 ( 1976 ) . | the d0 collaboration has reported a @xmath0 deviation from the standard model ( sm ) prediction in the like - sign dimuon charge asymmetry . new physics beyond the sm in @xmath1 mixing
is needed to explain the data . in this paper
, we investigate the possible extension of the sm with one generation color - triplet charged scalar as well as three generation majorana fermions .
we study the implications of the model on the d0 s dimuon charge asymmetry as well as matter anti - matter asymmetry of the universe . |
what do we mean by nuclear matter at subsaturation density ?
this is an important conceptual question .
the binding energy of uniform nuclear matter has a minimum at saturation density @xmath0 .
below this density , the system can minimize its energy by forming clusters .
therefore the study of low - density nuclear matter is a study of clusters .
what are the properties of these clusters ?
what role do they play in astrophysics ?
what is the relationship between clusters formed in the thermodynamic limit for an infinite system and those formed in laboratory heavy - ion collisions ? in this paper , we answer these questions using the virial expansion for low - density nuclear matter .
simulations of core - collapse supernovae @xcite , giant stellar explosions , depend on properties of low - density nuclear matter near the neutrinosphere .
if one views a star in visible light , one sees the photosphere .
this is the surface of last scattering for photons .
supernovae emit @xmath1 of their energy in neutrinos , and if one views a supernova in neutrinos , one sees the neutrinosphere .
this is the surface of last scattering for neutrinos and occurs at a density where the neutrino mean - free path is comparable to the size of the system .
direct evidence for the temperature of the neutrinosphere yields roughly @xmath2 , from the small number of neutrinos detected in sn1987a @xcite .
the neutrinosphere density , @xmath3 , then follows from known cross sections for neutrinos with these energies .
_ in this paper , we use the virial expansion to make model - independent predictions for the equation of state and other properties of low - density nuclear matter near the neutrinosphere . _
we will use these results in later works to calculate neutrino interactions .
nuclear statistical equilibrium ( nse ) models are commonly used in nuclear astrophysics @xcite .
these models describe low - density nuclear matter as a system of non- or minimally - interacting nuclei with a distribution of neutron number @xmath4 and charge @xmath5 determined from nuclear masses through statistical equilibrium .
however , nse models fail in an uncontrollable manner as the density increases and strong interactions between nuclei become important .
one would like a model - independent way to include strong interactions and a well - defined criterion to determine the range of validity of nse models . in this paper
we take a step towards this goal by considering a simple nse model with neutrons ( @xmath6 ) , protons ( @xmath7 ) and alpha ( @xmath8 ) particles .
the virial expansion is then used to systematically incorporate strong interactions using nucleon - nucleon ( nn ) , n@xmath8 and @xmath9 elastic scattering phase shifts . pure neutron matter at low density is a very interesting nearly - universal fermi system . in the limit that the neutron - neutron scattering length is large compared to the interparticle spacing and
the range of the interaction is small , the properties of dilute fermi gases are expected to be universal , independent of the details of the interaction . in a separate paper @xcite , we have calculated the equation of state of low - density neutron matter using the virial expansion , in order to assess quantitatively how close the system is to universal behavior at finite temperature .
the equation of state of resonant and dilute fermi gases has also been studied in laboratory experiments with trapped atoms @xcite , and ho _ et al . _
have used the virial expansion to describe fermi gases at high temperatures in the vicinity of feshbach resonances @xcite .
there are many theoretical approaches to low - density nuclear matter .
microscopic calculations start from nn and three - nucleon interactions that reproduce nn scattering and selected few - nucleon data .
in addition to the conventional variational @xcite and brueckner @xcite calculations , renormalization - group methods coupled with effective field theory provide new insights to nuclear forces and the possibility of a perturbative and thus systematic approach to nuclear matter with theoretical error estimates @xcite .
however , all wave functions that are typically used only include two - nucleon correlations and omit four - nucleon correlations that could be important to describe alpha particle formation at low densities .
we note that the equation of state calculation of buchler and coon @xcite is close in strategy to the virial equation of state , but it takes into account pauli blocking on the phase shifts and thus uses a model nn interaction .
moreover , there are a variety of skyrme - type @xcite and relativistic mean - field @xcite parametrizations of the nuclear energy functional , which are used to calculate ground - state energies and densities of intermediate - mass and heavy nuclei .
however , it is not clear what form to take for the energy functional for a gas of nucleons at low densities , where the system forms clusters and the binding energy per nucleon is density - independent with @xmath10 .
there are also some phenomenological models that attempt to provide a reasonable behavior for the equation of state over a wide range of densities , temperatures and compositions .
the lattimer - swesty equation of state @xcite is based on an extended liquid - drop model .
this equation of state is almost universally used in modern supernova simulations @xcite .
in addition , the shen - toki - oyamatsu - sumiyoshi equation of state @xcite is based on an approximate thomas - fermi calculation using a relativistic mean - field interaction . however , it is not clear how reliable these models are , and there does not appear to be any way to systematically improve them . recently , there are lattice simulations for neutron matter using effective field theory @xcite , where the couplings are fitted to reproduce nn scattering regularized on the same lattice .
these calculations are promising , but are presently limited to low orders in the effective field theory expansion and to small lattices .
it is very useful to compare these lattice results to our virial equation of state at low densities . in the virial approach it is simple to assess how errors in the nn phase shifts , because of a low - order truncation in the effective field theory , impact the equation of state .
the virial equation of state thus provides a valuable check for the lattice results .
the virial expansion is a very general and old method for determining the equation of state of a dilute gas .
there are two major assumptions .
first , that the system is in a gas phase and has undergone no phase transition with decreasing temperature or increasing density .
second , that the fugacity @xmath11 is small , so the partition function can be expanded in powers of @xmath12 . here
@xmath13 denotes the chemical potential and @xmath14 is the temperature .
the second virial coefficient @xmath15 describes the @xmath16 term in the fugacity expansion of the partition function , while the third virial coefficient describes the @xmath17 term . a general formula relating @xmath15 to the two - body elastic scattering phase shifts has been known for some time @xcite .
little is known about the third virial coefficient @xcite .
the first application of the virial expansion to hot matter formed in heavy - ion experiments was by pratt _
et al . _
in addition , venugopalan and prakash @xcite , for example , use the virial expansion to study matter formed in relativistic heavy - ion collisions . in general , however , the latter work is interested in much higher temperatures , where thermal mesons are important .
a great advantage of the virial expansion is that it correctly includes both bound states and scattering resonances on an equal footing .
in contrast , nse models only include bound state contributions . since the thermodynamics of the system is continuous when a bound state moves into the resonant continuum , both weakly bound states and low - energy scattering resonances are equally important .
this was explicitly demonstrated for fermi gases in the crossover region of a feshbach resonance by ho and mueller @xcite .
the virial expansion in this paper includes contributions from the two - nucleon ( deuteron ) and four - nucleon ( alpha particle ) bound states and the relevant low - energy scattering resonances : the dominant ones being the @xmath18s@xmath19 two - nucleon resonance ( @xmath20he or @xmath20n ) , the p@xmath21 n@xmath8 resonance and the @xmath9 @xmath22 resonance that is the ground state of @xmath23be .
this paper is organized as follows . in section [ formalism ] , we present our virial formalism .
results are presented in section [ results ] for the virial coefficients and equilibrium properties , such as the composition , pressure , entropy , energy and symmetry energy .
we conclude and give an outlook in section [ conclusions ] .
in this section , we introduce the virial expansion for a system of neutrons , protons , and alpha particles . we first discuss our choice of @xmath6 , @xmath7 and @xmath8 particle degrees of freedom
. then we present the virial equation of state expanded to second order in the fugacities and calculate the entropy and the energy .
finally , we relate the second virial coefficients to the relevant scattering phase shifts . there is a close relation between the virial expansion and nse models .
the latter include a variety of bound states , while the virial expansion includes both bound states and scattering resonances .
the virial equation of state can therefore also be regarded as a systematic extension of nse models to take into account strong interactions in the resonant continuum .
the deuteron is automatically included in the virial expansion as a bound - state contribution to the second virial coefficient @xmath24 with experimental deuteron binding energy @xmath25 .
however , the @xmath8 particle has a large binding energy , @xmath26 , and this will only be included in the fourth virial coefficient describing four - nucleon interactions . in general ,
fourth - order contributions are expected to be small at low densities and high temperatures .
however , for lower temperatures @xmath27 is very large .
this will lead to an unnaturally large fourth virial coefficient that will greatly reduce the radius of convergence of the nucleonic virial expansion .
our solution to this problem is to rearrange the original nucleonic virial expansion and include @xmath8 particles explicitly .
this allows us to retain the large @xmath27 terms at lower order in the virial expansion .
this procedure can be extended to explicitly include a variety of additional heavy nuclei that may be important at even higher densities .
we discuss the effects of heavier nuclei and of three - nucleon contributions in more detail below . in this first paper , for simplicity , we limit ourselves to @xmath6 , @xmath7 and @xmath8 particle degrees of freedom .
we note that three - nucleon bound state effects are expected to be small , because their binding energies are small compared to @xmath28 , and the system will thus favor the formation of alpha particles . therefore , we consider a low - density gas of @xmath6 , @xmath7 and @xmath8 particles . in chemical equilibrium ,
the neutron @xmath29 , proton @xmath30 and alpha @xmath31 chemical potentials satisfy @xmath32 so that the neutron @xmath33 , proton @xmath34 and alpha @xmath35 fugacities are related by @xmath36 as discussed above , we will consider @xmath37 to be the same order as @xmath38 or @xmath39 and expand the partition function through second order in the fugacities @xmath38 , @xmath39 and @xmath37 .
this is equivalent to the virial expansion with only nucleon densities at high temperatures and reduces to the virial expansion of a pure @xmath8 gas at low temperatures ( assuming equal numbers of @xmath6 and @xmath7 , @xmath40 ) .
moreover , it is equivalent to including only the bound - state contribution in the fourth nucleon virial coefficient .
we expand the grand - canonical partition function @xmath41 for a system of @xmath6 , @xmath7 and @xmath8 particles confined to a volume @xmath42 as @xmath43 & + z_n^2 q_{nn } + z_p^2 q_{pp } + z_\alpha^2 q_{\alpha\alpha } \nonumber \\[1 mm ] & + 2z_pz_n q_{np}+ 2z_nz_\alpha q_{\alpha n } \nonumber \\[1 mm ] & + 2z_pz_\alpha q_{\alpha p } + \mathcal{o}(z_i^3 ) \,.\end{aligned}\ ] ] here , @xmath44 and @xmath45 , @xmath46 and @xmath47 are the partition functions for single particle @xmath6 , @xmath7 or @xmath8 systems @xmath48 where @xmath49 ( @xmath50 ) denotes the nucleon ( @xmath8 particle ) thermal wavelength , @xmath51 we use @xmath52 for the nucleon mass @xmath53 and @xmath54 for the alpha particle mass , thus @xmath55 .
below , we will relate the two - particle partition functions @xmath56 with @xmath57 to the second virial coefficients .
next , we expand @xmath58 to second order in the fugacities @xmath59 & + z_p^2 \ , \frac{q_{pp}-\frac{1}{2}q_p^2}{v } + z_\alpha^2 \ , \frac{q_{\alpha\alpha}-\frac{1}{2}q_\alpha^2}{v } \nonumber \\[1 mm ] & + 2 z_p z_n \frac{q_{np}-\frac{1}{2}q_nq_p}{v } + 2 z_p z_\alpha \frac{q_{\alpha p}-\frac{1}{2}q_\alpha q_p}{v } \nonumber \\[1 mm ] & + 2 z_n z_\alpha \frac{q_{\alpha n}-\frac{1}{2}q_\alpha q_n}{v } + \mathcal{o}(z_i^3 ) \,.\end{aligned}\ ] ] we define the second neutron virial coefficient @xmath60 as @xmath61 where the second approximation is due to neglecting the coulomb interaction between protons and assumes charge - independent nuclear interactions .
likewise , the second alpha virial coefficient @xmath62 is given by @xmath63 and the virial coefficients describing strong interactions between @xmath64 and n@xmath8 particles are @xmath65 b_{\alpha n}&=\frac{\lambda_\alpha^3}{v}(q_{\alpha n}-\frac{1}{2}q_\alpha q_n ) \approx
\frac{\lambda_\alpha^3}{v}(q_{\alpha p}-\frac{1}{2}q_\alpha q_p ) \,,\end{aligned}\ ] ] where again the second approximation neglects the coulomb interaction and strong - interaction charge dependences .
coulomb effects typically depend on the ratio of the coulomb energy to the thermal energy @xmath66 @xcite , where @xmath67 is the ion charge , @xmath8 the fine structure constant , and @xmath68 with @xmath69 the ion density .
for example , for symmetric nuclear matter ( assumed composed only of nucleons ) at a density @xmath70 and @xmath71 one has @xmath72 , and thus we expect coulomb effects to be small .
however , this should be checked explicitly in future work and coulomb interactions may be more important when we extend the virial approach to densities where heavy nuclei are present .
note that the choice of @xmath73 in the last equation ( instead of @xmath74 ) is our convention .
below , we will calculate these second virial coefficients from microscopic nn , n@xmath8 and @xmath9 elastic scattering phase shifts . the pressure @xmath75 is given by the logarithm of the partition function , which we truncate after second order in the fugacities . in terms of the second virial coefficients
, the pressure can be written as @xmath76 & + \frac{1}{\lambda_\alpha^3}(z_\alpha + z_\alpha^2 b_\alpha + 2 z_\alpha ( z_n+z_p ) b_{\alpha n } ) \ , .
\label{p}\end{aligned}\ ] ] the @xmath6 , @xmath7 and @xmath8 densities follow from derivatives of @xmath77 with respect to the chemical potential or the fugacity , @xmath78 the resulting @xmath6 , @xmath7 and @xmath8 particle densities are given by @xmath79 n_p & = \frac{2}{\lambda^3}(z_p+2z_p^2b_n+2z_pz_nb_{pn}+8z_\alpha z_p b_{\alpha n } ) \ , , \label{np } \\[1 mm ] n_\alpha & = \frac{1}{\lambda_\alpha^3}(z_\alpha+2z_\alpha^2b_\alpha+
2z_\alpha(z_n+z_p)b_{\alpha n } ) \ , , \label{nalpha}\end{aligned}\ ] ] where we have used @xmath80 .
these equations provide a parametric form for the virial equation of state that is thermodynamically consistent . for values of @xmath38 and @xmath39 ,
the alpha fugacity @xmath37 follows from chemical equilibrium @xmath81 and one can then calculate the pressure @xmath75 and composition , @xmath82 , @xmath83 and @xmath84 , from eqs .
( [ p],[nn]-[nalpha ] ) .
the total baryon density @xmath85 is given by @xmath86 and the proton fraction @xmath87 , the number of protons per baryon , by @xmath88 the dependence of the baryon density and the proton fraction on @xmath38 and @xmath39 can be inverted to yield the virial equation of state in terms of @xmath89 note that these dependences are nonlinear and strongly temperature dependent , mainly due to the @xmath27 term in the chemical equilibrium condition for @xmath37 . in practice , for a given proton fraction , we use eq .
( [ yp ] ) to determine the proton fugacity as a function of the neutron one @xmath90 , and generate the virial equation of state in tabular form for a range of @xmath38 .
this maintains the thermodynamic consistency of the virial equation of state .
we relate the second virial coefficients to the relevant scattering phase shifts .
this extends the standard results for spin - zero particles @xcite to include spin and isospin .
the virial coefficient is related to the partition function of the two - particles system @xmath91 , where the sum is over all two - particle states of energy @xmath92 .
this sum can be converted to an integral over relative momentum @xmath93 weighted by the density of states .
the difference between the density of states of an interacting and a free two - particle system can be expressed in terms of the derivative of the two - body phase shift @xmath94 @xcite .
finally , if one integrates by parts , the virial coefficients can be calculated from an integral over the scattering phase shifts summed over all partial waves with two - body spin @xmath95 , isospin @xmath14 , orbital angular momentum @xmath96 and total angular momentum @xmath97 allowed by spin statistics .
the virial coefficient @xmath60 describes the interactions in pure neutron matter , and we have @xmath98 here , @xmath99 is the free fermi gas contribution due to the pauli principle and @xmath100 is the sum of the @xmath101 elastic scattering phase shifts at laboratory energy @xmath102 .
this sum includes a degeneracy factor @xmath103 , @xmath104 & = \delta_{^1\text{s}_0}+\delta_{^3\text{p}_0}+3\,\delta_{^3\text{p}_1 } + 5\,\delta_{^3\text{p}_2}+5\,\delta_{^1\text{d}_2}+ \ldots \label{deltantot}\end{aligned}\ ] ] here and in the following we have neglected the effects of the mixing parameters due to the tensor force .
we expect that their contributions to the second virial coefficients describing spin - averaged observables vanish .
the @xmath64 virial coefficient can be decomposed as @xmath105 where @xmath106 is the second virial coefficient for symmetric nuclear matter . with @xmath107 and
@xmath60 one can describe asymmetric matter with arbitrary proton fraction . for @xmath106 ,
the contributions from the deuteron bound state and the scattering continuum are given by @xmath108 & + \frac{1}{2^{3/2}\,\pi t } \int_0^\infty de \
: e^{-e/2 t } \ , \delta^{\text{tot}}_{\text{nuc}}(e ) \
, , \label{bnuc}\end{aligned}\ ] ] where the term @xmath109 in the deuteron contribution comes from the partial integration ( the phase shift at zero energy being @xmath110 times the number of bound states ) and the factor @xmath111 counts the spin - isospin degeneracy of the deuteron .
the total phase shift @xmath112 for nuclear matter also receives contributions from @xmath113 states , @xmath114 & = 3\,\delta_{^1\text{s}_0}+3\,\delta_{^3\text{s}_1 } + 3\,\delta_{^1\text{p}_1 } \nonumber \\[1 mm ] & + 3\,\delta_{^3\text{p}_0 } + 9\,\delta_{^3\text{p}_1}+15\,\delta_{^3\text{p}_2 } + 15\,\delta_{^1\text{d}_2 } \nonumber \\[1 mm ] & + 3\,\delta_{^3\text{d}_1}+5\,\delta_{^3\text{d}_2 } + 7\,\delta_{^3\text{d}_3}+ \ldots \label{deltanuctot}\end{aligned}\ ] ] for spin - zero alpha particles we can directly use the results of @xcite , which give for the virial coefficient @xmath62 @xmath115 where @xmath116 describes the contribution for a free bose gas and the total phase shift @xmath117 for elastic @xmath9 scattering is given by @xmath118 & = \delta_{\text{s}}+5\,\delta_{\text{d}}+9\,\delta_{\text{g } } + 13\,\delta_{\text{i } } + \ldots \label{deltaaatot}\end{aligned}\ ] ] for the @xmath9 virial coefficient and for the following n@xmath8 virial coefficient , we have not taken into account the effects of inelasticities . for low temperatures , we expect their contributions to be small due to the tight binding of the alpha particle .
finally , the n@xmath8 virial coefficient is given by @xmath119 with total phase shift @xmath120 for n@xmath8 scattering at nucleon laboratory energy @xmath102 , @xmath121 & = 2\,\delta_{\text{s}_{1/2}}+2\,\delta_{\text{p}_{1/2 } } + 4\,\delta_{\text{p}_{3/2}}+4\,\delta_{\text{d}_{3/2 } } + 6\,\delta_{\text{d}_{5/2 } } \nonumber \\[1 mm ] & + 6\,\delta_{\text{f}_{5/2}}+8\,\delta_{\text{f}_{7/2}}+ \ldots \label{deltaantot}\end{aligned}\ ] ] once the four virial coefficients @xmath122 , @xmath123 , @xmath124 and @xmath125 have been calculated , the pressure is determined from eq .
( [ p ] ) using fugacities @xmath38 and @xmath39 that reproduce the desired baryon density @xmath85 and proton fraction @xmath87 using eqs .
( [ zalpha],[nn]-[yp ] ) . the entropy @xmath95 and
the energy @xmath102 can be calculated from the virial equation of state using thermodynamics @xcite .
the entropy density @xmath126 follows from @xmath127 here the temperature derivative is at constant @xmath30 and @xmath29 , and thus constant @xmath31 due to chemical equilibrium .
this leads to @xmath128 & + \frac{2t}{\lambda^3 } \bigl((z_n^2+z_p^2 ) b_n^\prime + 2 z_p z_n b_{pn}^\prime \bigr ) \nonumber \\[1 mm ] & + \frac{t}{\lambda_\alpha^3 } \bigl(z_\alpha^2 b_\alpha^\prime + 2 z_\alpha ( z_n+z_p ) b_{\alpha n}^\prime \bigr ) \ , , \label{s}\end{aligned}\ ] ] where @xmath129 denotes the temperature derivative of the second virial coefficients @xmath130 .
the energy density @xmath131 can be calculated from the entropy density by @xmath132 & = \frac{3}{2}p - n_\alpha e_\alpha + \frac{2t^2}{\lambda^3 } \bigl((z_n^2+z_p^2 ) b_n^\prime + 2 z_p z_n b_{pn}^\prime \bigr ) \nonumber \\[1 mm ] & + \frac{t^2}{\lambda_\alpha^3}\bigl(z_\alpha^2 b_\alpha^\prime + 2 z_\alpha ( z_n+z_p ) b_{\alpha n}^\prime \bigr ) \ , .
\label{epsilon}\end{aligned}\ ] ] we emphasize that the @xmath8 particle binding - energy contribution @xmath133 is very important for the energy .
finally , the entropy per baryon @xmath134 , energy per baryon @xmath135 and the free energy per baryon @xmath136 are given by @xmath137 with the free energy density @xmath138 .
in this section , we present results for the second virial coefficients , the equation of state and the composition , as well as the entropy , energy and symmetry energy . [
cols="^,^,^,^,^,^,^,^,^,^ " , ] we now discuss the @xmath8 particle concentration .
the @xmath8 mass fraction is given by @xmath139 . in fig .
[ fig4 ] , we show @xmath140 for a temperature @xmath141 and proton fraction @xmath142 with various values of the virial coefficients . the largest @xmath8 particle concentration results for a free @xmath6 , @xmath7 , @xmath8 gas , where all @xmath143 .
including nucleonic interactions @xmath144 , but still keeping @xmath145 reduces @xmath140 significantly .
this is because the attractive nuclear interactions reduce the nucleon chemical potential , and this leads to a reduction of @xmath140 .
including also @xmath146 increases the @xmath8 mass fraction , but @xmath140 is still smaller than its free value .
the latter effect is dominated by the low - energy n@xmath8 resonance . finally , including @xmath147 only increases @xmath140 by a very small amount .
therefore , in general , nn virial coefficients are expected to be most important , while the n@xmath8 virial coefficient leads to small changes , and @xmath62 is least important .
this hierarchy is because typically @xmath148 .
the ls model excludes nucleons from the volume occupied by alpha particles .
although this is a prescription , it leads to an @xmath8 mass fraction that vanishes in the high - density limit @xcite .
the effects of the excluded volume are small at low densities .
however , the excluded volume corresponds to a repulsive n@xmath8 interaction that reduces the @xmath8 particle concentration .
in contrast , the neutron - alpha resonance and the attractive total phase shift leads to an attractive n@xmath8 intraction in the virial expansion . as a result , the @xmath8 mass fraction increases , as shown in fig .
therefore , the excluded volume alone gives the wrong sign for n@xmath8 interactions in the low - density limit , but we caution that the effects of the excluded volume and the n@xmath8 virial coefficient may be small at very low densities .
the third virial coefficient @xmath149 can be used to make a simple error estimate of neglected terms in the virial expansion . from the above conclusion
, we expect @xmath149 to be the most important of the third virial coefficients , and we therefore take into account only @xmath149 in this estimate . for simplicity , we consider symmetric nuclear matter with @xmath150 , for which the virial equation of state , eqs .
( [ p],[nn]-[nalpha ] ) , including @xmath149 reads @xmath151 & + \frac{1}{\lambda_\alpha^3}(z_\alpha + z_\alpha^2 b_\alpha
+ 4z_\alpha z\,b_{\alpha n } ) \ , , \\[1 mm ] n & = n_p+n_n = \frac{4}{\lambda^3}(z+2z^2 b_{\text{nuc } } + 3z^3b_{\text{nuc}}^{(3)}+8z_\alpha z\,b_{\alpha n } ) \ , , \\[1 mm ] n_\alpha & = \frac{1}{\lambda_\alpha^3}(z_\alpha + 2z_\alpha^2 b_\alpha + 4 z_\alpha z\,b_{\alpha n } ) \,.\end{aligned}\ ] ] in fig .
[ fig5 ] , we again plot the @xmath8 mass fraction .
the solid line is our previous result with @xmath152 .
we also give approximate error bands by considering @xmath153 ( which gives a lower @xmath140 ) and @xmath154 .
this estimates the error only from the effect of a typical neglected term in the virial expansion .
our estimate @xmath155 is somewhat arbitrary , as it is based on the observation that all second virial coefficients are comfortably less then 10 for @xmath71 .
we see from fig .
[ fig5 ] that the resulting error band is small .
_ therefore , our virial expansion makes a model - independent prediction for the @xmath8 particle concentration .
_ we also contrast our results in fig .
[ fig5 ] to the @xmath140 predicted by two phenomenological equations of state .
the lattimer - swesty ( ls ) equation of state @xcite is based on an extended liquid drop model and is almost universally used in modern supernova simulations .
the ls model predicts too few alpha particles over the densities where the virial equation of state is applicable .
alternatively , the shen - toki - oyamatsu - sumiyoshi ( shen ) equation of state @xcite is based on an approximate thomas fermi calculation using a relativistic mean - field interaction .
the shen equation of state predicts slightly too high values for @xmath140 at this temperature .
[ fig6 ] it is a simple matter to apply the virial equation of state to different proton fractions and temperatures . in fig .
[ fig6 ] , we show how the @xmath8 particle concentration decreases with decreasing proton fraction @xmath87 for @xmath71 . in addition , we give @xmath140 for @xmath142 and a range of temperatures @xmath156 , @xmath157 and @xmath158 in fig .
[ fig7 ] . at the lowest temperature @xmath159
, @xmath140 rises rapidly at low densities where the virial coefficients make only small contributions .
therefore the error band is very small .
the error band becomes more important for @xmath160 .
the ls equation of state is seen to significantly underestimate @xmath140 at all three temperatures .
this may be due to a simple oversight in the published version @xcite of the ls equation of state .
lattimer and swesty include the neutron - proton mass difference in the proton chemical potential , but do not appear to include twice this difference in the alpha chemical potential .
in contrast , the shen equation of state is close to our virial results at low densities .
however , both ls and shen equations of state predict too few @xmath8 particles at high temperatures @xmath161 .
this can be clearly seen from the @xmath8 mass fraction shown in fig .
[ fig8 ] for @xmath162 .
the differences of the ls and shen equations of state at the higher densities shown for @xmath162 in fig .
[ fig8 ] are due to the differences in the symmetry energy @xcite .
[ fig7 ] [ fig8 ] both phenomenological equations of state shown in fig .
[ fig7 ] give @xmath8 particle concentrations that first increase with density and then rapidly decrease .
this decrease is due to the formation of heavy nuclei .
heavy nuclei are not included in our first studies .
we will extend the virial formalism to include heavy nuclei in future work .
therefore , our virial results are currently limited to lower densities , where one does not expect large contributions from heavy nuclei . in fig .
[ fig9 ] , we plot the density above which the shen equation of state has a heavy nuclei mass fraction larger than 10% .
this provides an estimate for the range of validity of the virial @xmath6 , @xmath7 , @xmath8 equation of state . for this estimate
we have considered the shen instead of the ls model due to the ls error in @xmath140 .
next , we present our results for the pressure . in fig .
[ fig10 ] , we plot the pressure of symmetric nuclear matter for @xmath163 .
the formation of @xmath8 particles reduces the total number of particles in the system , and this significantly lowers the pressure . for comparison
, we also show in fig .
[ fig10 ] the microscopic fermi hyper - netted - chain ( fhnc ) equation of state of friedman and pandharipande @xcite .
the latter calculation is based on a variational wave function with only two - nucleon correlations .
however , the model - independent virial equation of state shows that there are large @xmath8 concentrations in this density regime .
for example , at the threshold density of @xmath164 ( from fig .
[ fig9 ] ) we have @xmath165 .
microscopic calculations may require wave functions with four - nucleon correlations to accurately describe these @xmath8 particle contributions .
_ this suggests that all present microscopic variational or brueckner equations of state are incomplete in their description of low - density nuclear matter . _ in fig .
[ fig10 ] , we also give the pressure for densities beyond the threshold density of fig . [ fig9 ] in order to compare virial and fhnc results , which both omit the formation of heavy nuclei . the error band in fig .
[ fig10 ] is again estimated from a neglected third virial coefficient @xmath166 .
for the pressure , this involves a cancellation .
increasing @xmath149 decreases the nucleonic contribution to the pressure due to the additional attractive interaction between nucleons .
however , a larger @xmath149 also reduces the formation of @xmath8 particles ( due to a lower alpha chemical potential ) , and this acts to increase the total pressure .
therefore , a positive @xmath149 reduces the pressure at very low densities , but then leads to a larger pressure with increasing density . in fig .
[ fig11 ] , we show our results for the pressure of symmetric nuclear matter for @xmath71 in comparison to the phenomenological ls and shen equations of state .
the pressure given in fig .
[ fig11 ] also includes the contribution of an electron and photon gas as in appendix c of @xcite .
we find the pressures given by the virial equation of state and the phenomenological models agree well . in a separate paper @xcite
, we have demonstrated that the equation of state of low - density neutron matter scales to a very good approximation .
table [ table1 ] shows that the neutron virial coefficient @xmath122 is practically independent of the temperature .
if all virial coefficients are temperature independent , and provided the condition for chemical equilibrium does not introduce a strong temperature dependence through @xmath167 , then the power series in the fugacity for the pressure and the baryon density will have no explicit temperature dependence . in this case , the pressure will scale as @xmath168 where @xmath169 is some function of @xmath170 .
if this scaling relation holds , one can predict the pressure @xmath171 at a new temperature @xmath172 from the original @xmath173 through @xmath174 with @xmath175 .
we study this scaling symmetry in fig .
[ fig12 ] , where we have used eq .
( [ pscale ] ) to predict the pressure of neutron matter for @xmath176 from virial pressures calculated for @xmath101 and @xmath158 .
the agreement with the unscaled @xmath71 virial pressure is excellent .
this demonstrates that low - density neutron matter scales , even when interactions are included .
[ fig12 ] also shows scaled pressures of symmetric nuclear matter for @xmath176 using @xmath101 and @xmath158 results as input .
these scaled pressures do not agree with the @xmath71 virial pressure .
therefore , nuclear matter does not scale , which is due to clustering and mainly comes as a result of the strong temperature dependence of @xmath37 and @xmath140 .
the entropy per baryon @xmath134 of symmetric nuclear matter is shown for @xmath71 in fig .
[ fig13 ] .
here , our virial calculation again includes the entropy of an electron and photon gas as in appendix c of @xcite . the entropy is clearly seen to reflect the composition . at low densities , the ls entropy is above the other calculations because of the low @xmath8 mass fraction in the ls model . at high densities ,
the entropy in the phenomenological models is below our virial result due to the formation of heavy nuclei .
this deviation coincides with the threshold density of fig .
[ fig9 ] . in fig .
[ fig13 ] , we also show the entropy per baryon for the wolff equation of state @xcite , to give a measure for the model dependence of the onset of heavy nuclei . the entropy per baryon for extremely neutron - rich matter @xmath177 is shown in fig .
[ fig14 ] .
now there is very little change in composition , and thus good agreement between the virial and phenomenological results .
in addition to the entropy , we show our results for the energy per baryon @xmath135 in fig .
[ fig15 ] . in the limit of very low density
, the energy reduces to the kinetic energy of an ideal classical gas @xmath178 .
therefore , we have subtracted @xmath178 from @xmath135 in fig .
[ fig15 ] , in order to clearly show the interaction effects .
we find that the energy per baryon for pure neutron matter is somewhat below @xmath178 due to the attractive nuclear interactions .
our neutron matter results are in excellent agreement with the scaling law for the energy density @xmath179 , which follows from eq .
( [ epsilon ] ) when the virial coefficients are temperature independent , as is the case for neutron matter with @xmath180 .
in contrast , @xmath135 for symmetric nuclear matter drops very rapidly with increasing density .
this reflects the large binding energy as @xmath8 particles form . as can be seen from fig .
[ fig15 ] , this drop becomes more rapid for lower temperatures .
we therefore expect that the inclusion of heavy nuclei , or many - nucleon correlations , are necessary to obtain a constant energy per particle @xmath181 for @xmath113 nuclear matter at subsaturation densities .
the symmetry energy @xmath182 characterizes how the energy rises as one moves away from equal numbers of neutrons and protons , @xmath183 we evaluate the second derivative of the energy per baryon with respect to the proton fraction numerically . fig .
[ fig16 ] shows the symmetry energy for @xmath71 . at very low density , @xmath182 rises slowly with density .
to lowest order in the density and neglecting @xmath8 particles , the symmetry energy is given by @xmath184 then , as @xmath8 particles form , @xmath182 rises much faster with density . as a result of clustering ,
the symmetry energy is large even at a very small fraction of saturation density .
this is analogous to the observed behavior for the energy .
we also show in fig .
[ fig16 ] the symmetry free energy @xmath185 defined as @xmath186 to lowest order in the density and neglecting @xmath8 particles , the symmetry free energy is given by @xmath187 where the @xmath188 term is from the entropy of mixing in the free energy density .
again , as @xmath8 particles form , we find in fig .
[ fig16 ] that @xmath185 increases rapidly with density .
finally , the temperature dependence of the symmetry energy is shown in fig .
[ fig17 ] for @xmath156 , @xmath157 and @xmath158 .
the lower the temperature , the more rapidly @xmath182 rises with density .
we conclude that , in the thermodynamic limit , the symmetry energy is large at low densities due to cluster formation .
isospin observables in asymmetric heavy - ion collisions , such as the @xmath189 ration of emitted fragments , are often analyzed with semiclassical simulations , where nucleons move in isospin - dependent mean fields and undergo two - nucleon collisions .
the isospin - dependent mean - field implies a certain symmetry energy .
the density dependence of this symmetry energy can be adjusted to improve agreement with the data , e.g. , see @xcite .
this phenomenological symmetry energy at low densities is in general much smaller than our virial @xmath182 .
this is because the symmetry energy in the virial expansion includes the contributions of clusters with high central densities even at low baryon density .
the relation between @xmath182 and the symmetry energy of mean - field models for single nuclei should therefore be investigated in future work .
we have presented the equation of state of low - density nuclear matter using the model - independent virial expansion .
the virial equation of state is thermodynamically consistent and was calculated directly from binding energies and scattering phase shifts , without the need of a nuclear interaction model or nuclear wave functions .
therefore , our results provide a benchmark for all microscopic or phenomenological equations of state at low densities .
the virial equation of state can also be used to constrain models that extend to higher densities . in some cases ,
our virial results may be used to constrain model parameters or functional forms for model interactions . in other cases
, more extensive modifications of the phenomenological equations of state may be necessary in order to reproduce the virial low - density limit .
tightly bound nuclei can lead to unnaturally large bound state contributions to the higher - order virial coefficients .
therefore , we have considered a gas of nucleons and alpha particles to explicitly take into account the large @xmath8 particle binding energy .
this leads to a larger radius of convergence compared to a nucleonic virial expansion .
we have calculated the second virial coefficients for nn , n@xmath8 and @xmath9 interactions directly from the relevant binding energies and scattering phase shifts . for the temperatures of interest ,
the virial coefficients are dominated by the low - energy bound states and resonances , but higher partial wave contributions are significant .
we have found that the second virial coefficient for neutron matter @xmath60 is approximately temperature - independent , which leads to an approximate scaling symmetry of pure neutron matter .
the second virial coefficients for nuclear matter @xmath107 and @xmath106 decrease with temperature , and as a result nuclear matter does not scale . the second virial coefficient for n@xmath8 interactions @xmath190
is well approximated by the resonant p@xmath21 contribution .
this holds over a wide range of energies due to a cancellation among the additional partial waves . the second virial coefficient for @xmath9 interactions @xmath62 increases with temperature , as the @xmath9 phase shifts become large at higher energies .
the inclusion of the coulomb interaction for the virial coefficients , and the proper treatment of inelasticities and mixing parameters ( due to the tensor force ) are important topics for future improvements .
we have used the virial coefficients to make model - independent predictions for a variety of properties of nuclear matter over a range of densities , temperatures and composition .
the resulting @xmath8 particle concentration disagrees with all equations of state currently used in supernova simulations .
the contributions from low - energy resonances show most prominently in the composition of low - density nuclear matter .
the effects of the second virial coefficients follows a natural hierarchy , with the nn virial coefficients being more important than the n@xmath8 virial coefficient , and the @xmath9 virial coefficient leading to very small changes .
we have used this hierarchy to make simple error estimates by studying the effects of a neglected third virial coefficient @xmath191 .
the resulting error bands are small .
detailed investigations of the effects of omitted higher virial coefficients , as well as the effects of heavy nuclei will be the topic of separate studies .
for a better error estimate , it is important to also have a reliable calculation of the third virial coefficients , which could , e.g. , come from a calculation in the effective field theory for halo nuclei @xcite .
furthermore , these error estimates can also help clarify the domain of validity of the virial expansion .
the physics of nuclear matter is very different from neutron matter due to clustering . as the density increases , @xmath8 particles form and this leads to a significant reduction of the pressure of low - density nuclear matter .
similarly , clustering increases the binding energy and it reduces the entropy , which reflects the number of particles in the system . as the density increases further , heavier nuclei and larger clusters form , and @xmath8 particles become less important . in fact , the breakdown of the virial expansion for nuclear matter is due to the formation of heavy nuclei , which is reached before the nucleon fugacities become large . in nuclear matter at @xmath113 , we expect that one has to explicitly include heavy nuclei , or many - nucleon correlations , to obtain a constant energy per baryon @xmath10 and a constant symmetry energy at low densities .
these virial results greatly improve our conceptual understanding of nuclear matter at subsaturation densities .
the virial equation of state can be regarded as a systematic extension of nse models for strong interactions in the resonant continuum .
such a comparison will also provide valuable insights how to systematically organize the virial expansion to include heavy nuclei .
in addition , topics of future studies include model - independent predictions for the neutrino response of low - density nuclear matter and a detailed comparison with nuclear lattice calculations @xcite .
we thank s. karataglidis for the calculation of the nucleon - alpha phase shifts and a. marek for providing us with tables of the ls , shen and wolff equations of state .
this work is supported in part by the us doe under grant no .
de fg0287er40365 and the nsf under grant no .
phy0244822 .
99 m. liebendrfer , m. rampp , h .- th .
janka and a. mezzacappa , astrophys .
j. * 620 * ( 2005 ) 840 .
r. walder , a. burrows , c.d .
ott , e. livne and m. jarrah , astrophys .
j. * 626 * ( 2005 ) 317 .
m. costantinit , a. ianni and f. visanni , phys . rev .
* d70 * ( 2004 ) 043006 .
c. lunardini and a.y .
smirnov , astropart .
* 21 * ( 2004 ) 703 .
meyer , ann .
. astrophys .
* 32 * ( 1994 ) 153 .
c.j . horowitz and a. schwenk , nucl - th/0507064 .
ohara , s.l .
hemmer , m.e .
gehm , s.r .
granade and j.e .
thomas , science * 298 * ( 2002 ) 2179 .
gehm , s.l .
hemmer , s.r .
granade , k.m . ohara and j.e .
thomas , phys . rev .
* a68 * ( 2003 ) 011401(r ) ; j. kinast , a. turlapov , j.e .
thomas , q. chen , j. stajic and k. levin , science * 307 * ( 2005 ) 1296 .
t. bourdel , j. cubizolles , l. khaykovich , k.m.f .
magalhaes , s. kokkelmans , g.v . shlyapnikov and c. salomon , phys .
* 91 * ( 2003 ) 020402 .
m. bartenstein , a. altmeyer , s. riedl , s. jochim , c. chin , j.h .
denschlag and r. grimm , phys .
* 92 * ( 2004 ) 120401 and cond - mat/0412712 .
l . ho and e.j .
mueller , phys .
* 92 * ( 2004 ) 160404 .
l . ho and n. zahariev , cond - mat/0408469 .
b. friedman and v.r .
pandharipande , nucl .
a361 * ( 1981 ) 502 .
a. akmal , v.r .
pandharipande and d.g .
ravenhall , phys . rev .
* c58 * ( 1998 ) 1804 .
m. hjorth - jensen , t.t.s .
kuo and e. osnes , phys .
* 261 * ( 1995 ) 125 ; l. engvik , g. bao , m. hjorth - jensen , e. osnes and e. stgaard , astrophys .
j. * 469 * ( 1996 ) 794 .
m. baldo , in nuclear methods and the nuclear equation of state , international review of nuclear physics , vol .
8 , world scientific , 1999 .
bogner , t.t.s .
kuo and a. schwenk , phys .
* 386 * ( 2003 ) 1 .
bogner , a. schwenk , r.j .
furnstahl and a. nogga , nucl . phys .
* a763 * ( 2005 ) 59 .
buchler and s.a .
coon , astrophys .
j. * 212 * ( 1977 ) 807 .
s. goriely , m. samyn , p .- h .
heenen , j. m. pearson and f. tondeur , phys . rev .
* c66 * ( 2002 ) 024326 .
m. bender , p .- h . heenen and p
reinhard , rev .
mod . phys .
* 75 * ( 2003 ) 121 .
stoitsov , j. dobaczewski , w. nazarewicz , s. pittel and d.j .
dean , phys .
* c68 * ( 2003 ) 054312 .
p. ring , prog .
* 37 * ( 1996 ) 193 .
serot and j.d .
walecka , int .
* e6 * ( 1997 ) 515 .
furnstahl and b.d .
serot , nucl . phys . *
a671 * ( 2000 ) 447 .
lattimer and f.d .
swesty , nucl . phys .
* a535 * ( 2001 ) 331 .
h. shen , h. toki , k. oyamatsu and k. sumiyoshi , nucl .
a637 * ( 1998 ) 435 ; prog . of theo
* 100 * ( 1998 ) 1013 .
d. lee and t. schfer , phys .
* c72 * ( 2005 ) 024006 .
e. beth and g.e .
uhlenbeck , physica * 4 * ( 1937 ) 915 .
a. pais and g.e .
uhlenbeck , phys . rev . * 116 * ( 1959 ) 250 .
bedaque and g. rupak , phys .
b67 * ( 2003 ) 174513 .
s. ichimaru , rev .
* 54 * ( 1982 ) 1017 .
s. pratt , p. siemens and q.n .
usmani , phys .
b189 * ( 1987 ) 1 .
venugopalan and prakash , nucl . phys .
* a546 * ( 1992 ) 718 .
k. huang , statistical mechanics , wiley n.y .
, second edition ( 1987 ) .
for the relation of the second virial coefficient @xmath15 to the two - body elastic scattering phase shifts , see p. 224 - 227 .
stoks , r.a.m .
klomp , m.c.m .
rentmeester and j.j .
de swart , phys . rev . *
c48 * ( 1993 ) 792 ; see also nn - online , http://nn-online.org .
r.a . arndt and l.d .
roper , phys . rev .
* c1 * ( 1970 ) 903 .
k. amos , p.j .
dortmans , h.v .
von geramb , s. karataglidis and j. raynal , adv .
. phys . * 25 * ( 2000 ) 275 ; k. amos and s. karataglidis , private communication .
arndt , d.l .
long and l.d .
roper , nucl . phys . * a209 * ( 1973 ) 429 .
afzal , a.a.z . ahmad and s. ali , rev .
* 41 * ( 1969 ) 247 .
the @xmath9 phase shifts are taken from table ii .
bacher , f.g .
resmini , h.e .
conzett , r. de swiniarski , h. meiner and j. ernst , phys .
* 29 * ( 1972 ) 1331 .
the @xmath9 phase shifts are taken from fig . 1 .
lattimer , c.j .
pethick , d.g .
ravenhall and d.q .
lamb , nucl .
* a432 * ( 1985 ) 646 .
lattimer , private communication .
w. hildebrandt , k. nomoto and r.g .
wolff , astron .
* 133 * ( 1984 ) 175 .
tan , s.r .
souza , r.j .
charity , r. donangelo , w.g .
lynch and m.b .
tsang , nucl - th/0311001 .
bertulani , h.w . hammer and u. van kolck , nucl
a712 * ( 2002 ) 37 .
horowitz , d. lee and a. schwenk , in prep . | we present the virial equation of state of low - density nuclear matter composed of neutrons , protons and alpha particles .
the virial equation of state is model - independent , and therefore sets a benchmark for all nuclear equations of state at low densities .
we calculate the second virial coefficients for nucleon - nucleon , nucleon - alpha and alpha - alpha interactions directly from the relevant binding energies and scattering phase shifts .
the virial approach systematically takes into account contributions from bound nuclei and the resonant continuum , and consequently provides a framework to include strong - interaction corrections to nuclear statistical equilibrium models .
the virial coefficients are used to make model - independent predictions for a variety of properties of nuclear matter over a range of densities , temperatures and compositions .
our results provide constraints on the physics of the neutrinosphere in supernovae .
the resulting alpha particle concentration differs from all equations of state currently used in supernova simulations .
finally , the virial equation of state greatly improves our conceptual understanding of low - density nuclear matter . |
a fraction of t tauri stars , including tw hya @xcite , gm aur @xcite , coku tau/4 @xcite and dm tau @xcite , present spectral energy distributions ( seds ) that combine little excess flux ( above the photospheric level ) at near - ir wavelengths with robust excesses at @xmath0 .
a similar effect is seen in at least one herbig ae / be star , hd100546 @xcite . since circumstellar discs are generally most optically thick in the inner , hottest regions , this observation is noteworthy and indicative of some process that can create an optically thin inner cavity within the disc .
one exciting possibility is that the cavity arises from the tidal truncation of the inner disc by a massive planet @xcite , in which case the sed observations provide unique constraints on the time scale and frequency of giant planet formation . however , non - planetary explanations are also possible .
dust can grow rapidly to form solid bodies that contribute negligible opacity in any observed waveband , while larger bodies may themselves suffer destructive collisions or rapid orbital decay due to aerodynamic drag against the gaseous disc @xcite .
photoevaporation of the outer disc by ionizing stellar radiation can also starve the inner disc of gas , yielding a short inner hole phase even in the absence of a planet @xcite .
protoplanetary disc models that include some or all of these effects @xcite admit the formation of dusty rings and inner holes , even in the absence of planets and gas . as improved seds and spectra from _
spitzer _ become available , more detailed theoretical study of both classes of model ( pure disc and disc plus embedded planet ) is warranted in order to determine whether these models can be distinguished without direct imaging data . a sufficiently massive planet ( or brown dwarf ) can probably create an inner cavity in which both the gas and dust surface densities are extremely low . however , of the four aforementioned t tauri stars around which inner disc ` holes ' are inferred from the lack of near - ir emission
only coku tau/4 appears to be entirely devoid of circumstellar material within 1 au of the star . in the case of tw hya
there is both ongoing low - level gas accretion @xcite , and evidence ( from the existence of a silicate emission feature at 10@xmath1 m ) for a sub - lunar mass of micron sized dust at small radii @xcite .
both gm aur and dm tau are classical t tauri stars , with estimated accretion rates on to the stellar photosphere variously estimated at @xmath2 to @xmath3 @xcite .
unfortunately @xcite failed to detect co emission from an inner disc around gm aur , so there are no observational clues as to the _ gas _ surface density that must accompany this accretion .
however , for these accretion rates protoplanetary disc models predict gas surface densities at @xmath4 of 200 - 2000 g@xmath5 @xcite .
although disc models are subject to large uncertainties arising from the unknown efficiency of angular momentum transport within protoplanetary discs , the robust accretion signatures seen for these stars imply that an inner gas disc must be present . within a planetary interpretation of the observed seds ,
the presence of inner gas discs and ongoing stellar accretion is not in itself surprising .
numerical simulations @xcite show that the tidal barrier erected by a massive planet is quite leaky . for jupiter mass planets gas can readily overflow the gap , and
while much of this gas is accreted by the planet in excess of 10% of the accretion rate at large radius can replenish the inner disc and ultimately accrete on to the star @xcite .
this is consistent with gm aur being a classical t tauri star , but the lack of dust opacity then requires that the planet needs to be relatively more effective at holding back dust
i.e. it needs to filter the gas as it passes the tidal barrier . in this paper
, we show that the required filtration may occur as a consequence of the sharp pressure gradient that exists near the outer edge of the gap .
the pressure gradient leads to super - keplerian rotation of the gas , which in turn allows standard aerodynamic coupling to solid bodies @xcite to inject angular momentum and hold back the inward flow of all but the smallest solid particles .
we estimate the efficiency of this process , and argue that it allows relatively low mass planets which are observed @xcite to be much more common than those planets with @xmath6 that would start to cut off accretion entirely to nevertheless create dust - poor inner cavities . within this scenario ,
the variation in the observed properties of t tauri stars with weak near - ir emission could reflect different mass embedded planets relatively low in the case of gm
aur and dm tau , and substantially higher in the case of coku tau/4 .
the disc gas and the embedded dust particles interact via a drag force @xcite that is a consequence of the different gas and dust orbital velocities . the keplerian velocity , @xmath7 , at a radius , r , around a star of mass @xmath8 is @xmath9 if @xmath10 is the gas pressure and if @xmath11 is the gas density , the azimuthal component of the gas velocity , @xmath12 , in centrifugal equilibrium differs from the keplerian velocity due to the influence of the gas pressure gradient and is given by @xmath13 the dust particles , on the other hand , are not influenced by pressure forces and , in centrifugal equilibrium , orbit at the keplerian velocity , @xmath7 .
the effect of the drag force depends on the size of the dust particles relative to the mean free path of the gas molecules .
if we assume that the gas is primarily molcecular hydrogen , the mean free path , @xmath14 , is @xmath15 where @xmath16 is the mass of the hydrogen molecule and @xmath17 is its cross - section ( @xmath18 ) .
the drag force , @xmath19 is then @xmath20 where @xmath21 is the relative velocity between the gas and dust particles , @xmath22 , @xmath23 , @xmath24 is the mean radius of the dust grains and @xmath25 is the drag coefficient , given by @xmath26 in equation ( [ cd ] ) , @xmath27 is the reynolds number which can be shown to be given by ( see for example @xcite ) @xmath28 in general , the gas pressure in protoplanetary discs decreases with increasing radius .
the pressure gradient is therefore negative and the gas velocity is sub - keplerian .
the drag force then acts to slow the dust particles down causing them to spiral in towards the central star .
one can then determine the radial drift velocity of dust particles by self - consistently solving the radial and azimuthal components of the momentum equation as outlined in @xcite .
figure [ radin ] shows an example of the inward radial velocity ( -dr / dt ) against particle radius at @xmath29 au in disc with a density of @xmath30 g @xmath31 and a temperature of @xmath32 k. the radial velocity is clearly strongly dependent on the particle size , and can reach values of @xmath33 cm s@xmath34 . the inward migration of dust particles , however , only occurs if the the pressure in the disc decreases monotonically with increasing radius . if there are any regions of the disc where the gas pressure increases with increasing radius , the gas velocity will become super - keplerian , and the drag force will cause dust particles to move outwards towards the pressure maxima .
this could occur in the presence of self - gravitating spiral structures @xcite , in the presence of vortices @xcite , or if an embedded planet has opened a gap in the disc @xcite . at the outer edge of a gap opened by a planet ,
the drag force will cause dust particles to migrate up the gap edge , away from the star . if , however , there is still gas accretion through the gap , which is expected for jupiter mass companions around solar - like stars , then dust particles moving too slowly up the gap edge will be dragged by the gas into the gap .
the gap edge will therefore act like a filter , allowing only particles of a certain size through into the inner disc .
even if the inner gas disc has not drained onto the central star , there could still be a substantial difference between the dust populations in the inner and outer disc and this could have a noticeable effect on the disc sed . to investigate this further , we need an approximation for the pressure gradient at the outer edge of a gap opened by an embedded planet . we therefore perform two - dimensional simulations of gaseous discs with embedded planets to determine the approximate disc structure .
we simulate a planet embedded in a gaseous disc using the zeus code @xcite .
we use two - dimensional coordinates ( @xmath35,@xmath36 ) and a resolution of @xmath37 , @xmath38 .
the computational domain extends , in code units , from @xmath39 to @xmath40 and we impose ouflow boundary conditions at both @xmath41 and @xmath42 .
we assume that the disc is isothermal with a radially dependent sound speed given by @xmath43 .
we adopt units in which @xmath44 and the sum of the mass of the central star and embedded planet is @xmath45 .
since the disc thickness @xmath46 is related to the sound speed through @xmath47 , we have a disc thickness of @xmath48 at @xmath49 .
we model angular momentum transport using a kinematic viscosity , @xmath50 , that operates only on the azimuthal component of the momentum equation @xcite . the initial disc surface density , @xmath51 ,
is taken to have a radial dependence of @xmath52 . since @xmath53 and
since we would expect @xmath54 to be constant we assume that , @xmath55 . we normalise the viscosity using the standard alpha formalism , @xmath56 @xcite , and assume that at @xmath49 , @xmath57 .
the planet is located at @xmath49 and we consider planet masses of @xmath58 , @xmath59 , and @xmath60 .
the simulations are evolved for @xmath61 planetary orbital periods .
figure [ plangap ] shows the surface density structure of the simulation with @xmath59 and illustrates that the inner regions of the disc have been significantly depleted .
figure [ surfdens ] shows the surface density profiles for the three different planet masses .
also shown ( dashed line ) is an emperical fit to the gap edge .
the surface density is normalised such that the total disc mass within @xmath62 would be @xmath63 in code units .
it appears that the chosen function is a reasonable fit to the gap edge , especially where the surface density is steepest , which is the region that would have the largest pressure gradient and hence would be the region where the drag force would have the largest influence on the embedded dust particles . from figure
[ surfdens ] a general form for the function representing the gap edge would be @xmath64 \right\}^3 , \label{gaussian}\ ] ] where @xmath65 is the inner radius of the gap edge , and @xmath66 is the gap scalelength . from the fits in figure
[ surfdens ] , the gap scalelength is somewhat greater for @xmath60 than for the lower planet masses , but we assume in general that @xmath67 .
the gap depth itself depends on the value of the coefficient @xmath68 .
a large @xmath68 value results in a deeper gap with a wider edge , and would correspond to a more massive planet .
if we use figure [ surfdens ] as a guide , @xmath69 would correspond to a planet mass of @xmath70 , while @xmath71 would correspond to a planet mass of @xmath60 .
these simulations have all been performed in scale free code units . to convert to real units , we simply need to introduce length and mass scales . for the problem we are considering here
we will generally assume a mass scale of @xmath72 giving a total disc mass of @xmath73 and planet masses of about @xmath74 , @xmath75 , and @xmath29 jupiter masses .
the length scale we introduce depends on the assumed semimajor axis of the planet . in the rest of this paper
we will assume a length scale of @xmath76 au , resulting in a semimajor axis of @xmath76 au and , from figure [ surfdens ] , an outer gap edge at @xmath77 au .
to determine how gas drag influences dust particles at the edge of a gap opened by an embedded planet , we need to know the gas pressure gradient . equation ( [ gaussian ] ) gives an approximate form for the surface density structure at the edge of a gap .
the volume density is then , @xmath78 where @xmath79 . the sound speed , @xmath80 where @xmath81 is the mean molecular weight , @xmath82 is boltzmann s constant and the temperature , @xmath83 @xcite with @xmath84 .
the pressure is then @xmath85 and the pressure gradient at the gap edge can be determined exactly since we have a fully analytic form for the pressure as a function of radius .
we should , however , note that since the discs we consider here have optically thin inner regions , the outer edge of the gap will intercept more stellar flux and will be hotter than if the inner disc were optically thick @xcite .
a higher temperature will produce a larger pressure gradient .
this temperature increase will , however , decay very quickly within the gap edge , and so may not be significant at the location where the pressure gradient is largest .
even if the temperature is slightly enhanced , a steeper temperature gradient will act to reduce the effect of the increased temperature .
overall we would not expect the effect of the increased temperature at the edge of the gap to be particularly significant .
once the pressure gradient is calculated , the technique described in [ gasdust ] can then be used to determine the radial velocity of solid particles at any point on the edge of the gap .
the radial velocity of the solid particles will vary considerably along the edge of the gap .
this is illustrated in figure [ radvelf8 ] which shows the radial velocity against particle size at the edge of a gap in which , @xmath86 au , @xmath87 au , and @xmath88 . in this case
the radial velocity is positive , as opposed to figure [ radin ] in which the radial velocity is negative .
the surface density in figure [ radvelf8 ] is also normalised such that in the absence of a gap the total disc mass within @xmath89 au would be @xmath90 .
the different curves in figure [ radvelf8 ] correspond to different locations ( i.e. , differents values of @xmath35 ) on the gap edge .
all the curves in figure [ radvelf8 ] have the same form as in figure [ radin ] . at a given location ,
the radial velocity intially increases as the particle size increases until it reaches a peak radial velocity , after which the radial velocity decreases with increasing particle size .
as @xmath35 increases , the curves initially move upwards and the radial velocity increases for all particle sizes ( solid line , small dash line , and dash - dot line ) reaching a maximum when @xmath91 au .
as @xmath35 increases beyond this point ( dash - dot - dot - dot line , and long dash line ) , the curves move to the right ( i.e. , the particle size at which the radial velocity is maximal increases ) , and the maximum radial velocity decreases . the radial velocity for all particles smaller than the particle with the largest radial velocity therefore decreases as @xmath35 increases . beyond the peak of the gap edge ,
the gradient changes sign , the radial velocity changes sign , and solid particles will drift inwards , rather than outwards .
what we wish to determine is the maximum outward radial velocity for particles that would contribute to the sed ( eg . , @xmath92 cm ) .
this can be determined by simply varying @xmath35 until these particles achieve their maximum velocity . in figure [ radvelf8 ] this would correspond to the dash - dot line ( @xmath91 au ) .
the two panels in figure [ radveldrho ] show the maximum outward radial velocity as a function of particle size for two different surface densities and for three different @xmath68 values . in the upper panel
the total disc mass within @xmath89 au would be @xmath93 , while in the lower panel the total mass would be @xmath94 . in both figures
@xmath86 au and @xmath87 au .
the three curves in each figure correspond to @xmath95 , @xmath88 , and @xmath96 .
based on the simulation results shown in figure [ surfdens ] , these could correspond to planet masses of @xmath74 m@xmath97 , @xmath75 m@xmath97 , and @xmath29 m@xmath97 .
the curves in figure [ radveldrho ] show that as @xmath68 , or planet mass , increases , the peak radial velocity moves to smaller particle sizes .
all particles smaller than the peak particle size then have increased radial velocities .
the radial velocities in figure [ radveldrho ] are , however , measured in the frame in which the gas radial velocity is zero .
if there is no accretion ( i.e. the radial gas velocity is zero relative to the central star ) , all particles would be drifting up the gap edge , away from the central star .
if there was some gas accretion , then particles with radial velocities smaller than the radial velocity of the gas will be dragged into the gap .
it is somewhat hard to quantify the radial gas velocity .
the presence of a planet can significantly influence the velocity structure of the gas in the vicinity of the planet @xcite .
we are , however , interested in the radial velocity of the gas as it accretes through the outer edge of the gap .
simulations @xcite show that in general the planet does not influence this material significantly , at least for planets with masses below a few jupiter masses @xcite .
the material that is most influenced by the planet is material that has already accreted into the gap .
most of this material also ends up within the planet s hill sphere , ultimately accreting onto the planet .
the material that we are considering here is material that accretes through the gap into the inner disc , and it appears that this is largely unaffected by the planet , at least in the vicinity of the gap edge .
the gas radial velocity through the edge of the gap will therefore be determined by the disc viscosity , and is related to the mass accretion rate through @xmath98 .
for an accretion rate of @xmath99 the radial gas velocity in the absence of a gap would be @xmath100 cm s@xmath34 for the higher of the surface densities ( upper panel ) in figure [ radveldrho ] and @xmath29 cm s@xmath34 for the lower of the surface densities ( lower panel ) in figure [ radveldrho ] .
if the mass transfer through the gap edge is unaffected by the planet , the radial gas velocity should then increase as the surface density decreases .
when @xmath95 , the surface density when the radial velocity of the particles maximises is @xmath101 times lower than at the peak of the gap edge .
the radial gas velocity should then be @xmath101 times higher , which for the upper panel of figure [ radveldrho ] , and for the mass transfer rate assumed above , would give @xmath102 cm s@xmath34 .
this means that only particles smaller than @xmath103 microns will be swept into the gap by the radial motion of the gas since these particles have outward radial velocities less than @xmath101 cm s@xmath34 . for @xmath96 , the surface density changes by a factor of @xmath104 and , if the same applies , then the radial gas velocity would then be @xmath105 cm s@xmath34 .
the maximum particle size that can populate the inner disc is then @xmath106 microns , comparable to that obtained when @xmath95 .
an equivalent result is obtained if we consider the lower panel in figure [ radveldrho ] .
the surface density being 5 times smaller than in the upper panel results in the particles radial velocities also increasing by a factor of @xmath29 . for a given mass accretion rate ,
however , the radial gas velocity also increases by a factor of @xmath29 and hence the size of particles that would be swept into the gap is roughly indepedent of the assumed surface density . on the other hand ,
numerical simulations by @xcite suggest that the mass accretion rate through a gap actually decreases with increasing planet mass , at least for planets with masses in excess of @xmath75 jupiter mass . in this case
the gravitational torques from the planet act to inhibit flow through the gap @xcite .
the rate of decrease with mass appears to be exponential , with the mass accretion rate decreasing by almost an order of magnitude as the planet mass increases from @xmath75 to @xmath107 jupiter masses .
if this is indeed the case , then the size of particles that will be swept into the gap by gas accretion should decrease with increasing planet mass . again considering the upper panel of figure [ radveldrho ] ,
if the three curves correspond to planet masses of @xmath108 , @xmath109 , and @xmath110 then we may expect the gas radial velocity to decrease by an order of magnitude as @xmath68 varies from @xmath111 to @xmath112 .
the maximum particle size swept into the gap would then decrease by an order of magnitude . for the gas radial velocities considered above
, we might expect the maximum particle size to decrease from @xmath103 microns to a few tenths of a micron as @xmath68 increases from @xmath111 to @xmath112 corresponding to the planet mass increasing from @xmath74 m@xmath97 to @xmath29 m@xmath97 .
since the above process results in only the smallest dust grains being swept into the gap , we would expect the dust to gas mass ratio to be different in the inner disc , compared to the outer disc .
it is , however , difficult to make an independent measurement of the gas and dust masses , and hence determining the different dust to gas mass ratios is probably not possible .
the presence of a silicate feature in protostellar disc spectral energy distributions ( seds ) does , however , depend on the grain size distribution .
the presence of large grains tends to wash out the silicate feature .
as we discuss in the next section , sed modelling of those systems with near infrared deficits suggest that the grain size distribution in the inner disc differs from that in the outer disc in a way that is consistent with dust filtration at a gap edge .
although the suggestion that the observed near - ir deficits in some ttauri seds could result from an embedded jupiter - like planet clearing the inner disc is exciting , there is , as yet , no clear evidence that this is indeed the case .
other non - planetary possibilities , for example rapid grain growth , could also result in optically thin regions that may produce seds largely consistent with that observed .
a sufficiently massive planet does , however , have a real effect on the structure of the gaseous disc which may influence the distribution of solid particles .
this effect could , in turn , influence the nature of the resulting sed .
a @xmath113 jupiter mass planet should open a gap in the disc , the inner region of which could clear to form an inner hole . the outer edge of this gap ( or hole ) will have a pressure that increases with increasing radius , increasing the net outward force on the gas and resulting in super - keplerian gas velocities . the resulting drag force on the solid particles
then results in these particles drifting up the gap edge , away from the central star .
if , however , there is still some gas accretion into the gap - which is expected for @xmath113 jupiter - mass planets @xcite - solid particles with radial velocities smaller than the radial gas velocity will populate the inner region of the disc . for disc properties appropriate for ttauri stars , the radial velocity increases with particle size for particles smaller than @xmath100 cm
this means that only the smallest particles will have outward radial velocities small enough to be dragged into the gap by the accreting gas .
gas accretion through the gap edge therefore acts to filter solid particles , with only the smallest particles making it into the gap .
the exact size range that is influenced by the gas depends on the properties of the gas disc and on the radial gas velocity . assuming a functional form for the gap edge ( motivated by two - dimensional , hydrodynamic simulations of embedded jupiter - like planets )
we show that if the inward gas velocity is @xmath114 cm s@xmath34 , only @xmath113 micron and smaller grains will be able to populate the inner disc . if the gas accretion rate is modified by sufficiently massive planets , as suggested by simulations @xcite , the size of particles that may populate the inner disc should decrease with increasing planet mass .
systems , such as coku tau/4 , that appear to have almost no material in the inner disc may indicate the presence of a massive companion while those , such as gm aur and tw hya , that appear to still have some material in the inner disc , may have a companion with a mass that allows some material to continually replenish the inner disc .
what is attractive about this possibility is that sed modelling suggests that system with near - ir deficits do indeed have two populations of dust particles .
this spatial segregation of dust within discs has been inferred in the tw hya disc @xcite , and radiation transfer models of new spitzer space telescope irs spectra of the coku tau 4 , gm aur and dm tau discs also require inner holes with small grains in the inner disc , and a larger grain population in the outer disc @xcite .
the silicate feature present in the irs spectra provide the clues that the dust grains in the inner disc are smaller than in the outer disc . as more data comes in from spitzer , the number of disc systems showing evidence for inner holes is increasing @xcite . to date ,
most candidates have irac and mips photometry , so only the presence of inner disc clearing can be inferred .
follow up observations with irs will enable us to determine whether the spatial segregation of dust sizes is a common feature and thereby further test our dynamical models for disc - planet interactions and dust filtration . the sed modelling of systems like gm aur and dm tau not only require a population of small grains in the inner disc , but also require that the mass , or surface density , is also low enough for the inner disc to be optically thin .
accretion onto a planet can reduce the amount of mass reaching the inner disc by up to @xmath115 @xcite , but this is probably insufficient to produce an optically thin inner disc .
typically the optical depth will need to be reduced by a factor of @xmath33 , a factor of @xmath116 more than can be produced by the planet alone . if solid particles are filtered at the gap edge , the amount of mass reaching the inner disc will depend on the size distribution of the solid particles .
interstellar medium ( ism ) grains are generally assumed to have sizes between @xmath117 and @xmath118 and a size distribution of @xmath119 @xcite .
grain growth can however change this distribution quite significantly .
if there is constant replenishment of the ism grains , this can steepen to @xmath120 @xcite , in which case most of the mass will be in small grains . if , however , there is no source of ism grains , the final size distribution can be much flatter , approaching @xmath121 . a relatively shallow size distribution ( e.g. , @xmath122 )
will already have a reduced opacity , independent of the presence of a planet , but , as shown below ( see also figure [ sedgaps ] ) , a further reduction is needed to match the observed seds .
such a reduction can be obtained through the filtration process described here .
the amount of material that can reach the inner disc will depend on the maximum size that can accrete through the gap edge , and on the size distribution of the particles in the outer disc .
figure [ massratfilt ] shows the ratio of the mass of particles smaller than @xmath123 ( solid line ) and @xmath124 ( dashed line ) to the total mass of solid particles , plotted against the exponent of the power - law size distribution . in this figure
we assume a minimum size of @xmath117 and a maximum size of @xmath75 mm . for steep size distributions
most of the mass is in small grains , and even if the maximum size of filtered particles is @xmath123 , the inner disc would remain optically thick .
it is also likely that in this case , grain coagulation would soon produce large grains in the inner disc @xcite . from sed modelling , however , we know that systems like gm aur , tw hya and dm tau must have a size distribution that extend to larger sizes in the outer disc and , since these systems are reasonably evolved , there is unlikely to be a significant new source of ism grains .
we might , therefore , expect the size distribution in the outer disc to be flatter than the ism grain size distribution @xcite .
it has already been shown that in some systems , flatter grain size distribution ( e.g. , @xmath125 ) provide better fits to the seds than ism size distributions @xcite .
figure [ massratfilt ] shows that a size distribution of @xmath126 and a maximum filtered size of @xmath123 , would reduce the mass of solid material accreting through the gap edge by a factor of @xmath127 .
if the planet accretes @xmath115 of this material , the amount of solid material reaching the inner disc would be reduced by a factor of @xmath128 , sufficient to produce an optically thin inner disc .
this process would also produce a very different dust - to - gas ratio in the inner disc , compared with the outer disc .
that systems like gm aur and dm tau have optically thin inner discs but are still accreting at @xmath129 suggests that there is still a substantial amount of gas in the inner disc .
the optically thin inner disc requires a reduction in solid particles of order @xmath128 , while the ongoing accretion requires a far smaller reduction in the amount of gas in the inner disc .
this is at least qualitatively consistent with the suggestion that the solid particles are filtered at the outer edge of the gap which , together with the subsequent accretion onto the planet reduces the mass of solid particles by a factor of @xmath33 , while at least @xmath130 of the gas is allowed through into the inner disc @xcite .
such a reduction in the solid particle surface density in the inner disc also means that subsequent grain coagulation is probably not very important .
@xcite , using a relatively simple one - particle model @xcite , show that at @xmath100 au grain coagulation can produce @xmath131 particles within a few hundred years for reasonable surface densities . if the mass that can be swept up by a growing particle is then reduced by a factor of @xmath128 , one might assume that the maximum size that this particle could achieve is then reduced by a factor of @xmath132 ( i.e. , a maximum size of @xmath133 microns ) .
this , however , is a significant overestimate since the amount of mass swept up by a growing particle depends strongly on the size of the particle , or its collision cross - section . in the simple one - particle model
, there is almost no grain growth if the surface density is reduced by a factor of @xmath128 from an initial density that would be appropriate for the inner regions of a ttauri disc .
although this is a fairly simple model , it at least suggests that if the amount of mass reaching the inner disc is reduced significantly , coagulation should not play an important role in the subsequent evolution of these grains .
figure[sedgaps ] illustrates how the filtration process discussed above may influence the disc seds .
the sed models in figure [ sedgaps ] are computed assuming a 300 au flared disc of total mass @xmath94 that is passively heated by a star with @xmath134 k , and @xmath135 .
the disc surface density and scaleheight have radial profiles of @xmath136 and @xmath137 , and the scale height is normalised such that @xmath138 au at @xmath139 au .
the temperature is determined self - consistently by the monte - carlo radiation transfer calculation .
the solid curve is the sed resulting from a dust model with a grain size distribution of @xmath140 with minimum and maximum grain sizes of @xmath117 and @xmath75 mm respectively ( see @xcite , table 1 and figure 3 ) .
the other two models simulate the dust filtration discussed above .
these models have the large grain model in the outer disc ( beyond 5 au ) and ism - like grains ( maximum size @xmath123 ) inside 5 au .
the dashed curve has a depletion factor of @xmath128 for the dust density , while the dotted curve has a depletion factor of @xmath141 , still within the possibilties discussed above .
notice the near - ir deficit and prominent silicate features at @xmath124 and @xmath142 for the sed models with ism - like grains and inner disc depletions .
the redistribution of ir excess from short to long wavelengths for discs with optically thin interiors is also apparent as discussed in @xcite .
these seds are not models for any particular system , but demonstrate the observable signatures of the filtration process we have described . in these models , the dust density in the inner disc is significantly reduced by filtration at the gap edge and by accretion onto the planet , while the gas is only influenced by accretion onto the planet . the gas density in the inner disc
is therefore reduced by a much smaller amount than the dust density .
this extremely small dust - to - gas ratio in the inner disc produces an opacity gap " , while still allowing gas accretion onto the star , in line with observations of many t tauri stars that exhibit near - ir deficits in their seds .
it has , however , also been suggested @xcite that opacity gaps " and the presence of small grains in the inner disc could actually be evidence for grain growth and subsequent rapid inward migration , rather than inferring the presence of a gap opening planet .
this conclusion was partly because jupiter - like planets in relatively massive discs should migrate into the central star within the disc lifetime @xcite .
although this may indeed be the case , the viscous timescale at @xmath29 au can be @xmath143 years for reasonable viscosity values .
the observation of such a system is therefore still possible , even if the planet ultimately does nt survive the migration process .
if near - ir deficits are due to grain growth , there must also be some means of sharply truncating the grain size distribution at @xmath144 in the inner disc .
it is possible that we are seeing a population of planets that form early , and potentially do not actually survive the migration process .
although we can not definitively conclude that near - ir deficits in the seds of some ttauri stars are a consequence of an embedded jupiter - like planet , that the filtering effect of such a planet - producing a small grain population in the inner disc that differs from the population in the outer disc - is consistent with sed modelling of these systems is suggestive .
the sed modelling also requires that the inner disc be optically thin , while ongoing gas accretion requires that the inner disc still have a substantial amount of gas .
if the grain size distribution is somewhat flatter than the standard ism size distribution @xcite , filtration at the outer gap edge can significantly reduce the amount of solid material reaching the inner disc , while still allowing sufficient gas through the gap .
although we do nt actually know the grain size distribution in the outer discs of those systems that show inner holes , they do show evidence for grain growth . in some cases
flatter size distributions do provide better fits to ttauri seds @xcite , although this has yet to be tested on the systems we consider here . that filtration of solid particles at the outer egde of a gap opened by an embedded jupiter - like planet can produce not only a population of small grains in the inner disc , but
can potentially also make this region optically thin , makes this an attractive possibility .
the filtration process should also lead to an enhancement of larger particles at the peak of the gap edge .
particles exterior to the gap will continue to drift inwards until they reach the gap edge , at which point the change in the gas pressure gradient will prevent the larger particles from drifting any further .
if the surface density of such particles becomes sufficiently high , further planet formation could occur at this location .
since ` hot ' jupiters are thought to form at modest radii and then migrate inwards via type ii migration to their present locations , one may expect in some cases to find additional lower mass planets in 2:1 resonances with these ` hot ' jupiters .
the presence of even terrestrial mass planets in resonant orbits with ` hot ' jupiters could be detected using transit timing @xcite .
although such a detection would be consistent with the analysis presented here , it has been suggested that this could also occur if low - mass , fast migrating planets ( @xmath145 m@xmath146 ) were resonantly trapped by gap opening planets @xcite . in the latter case , however , one may expect the trapped planets to subsequently grow into gas giants @xcite , while in the former case the planetesimal trapping could occur later when there may only be sufficient time for the newly forming planet to reach a terrestrial - like mass .
the authors would like to acknowledge interesting discussions with richard alexander , and would like to thank the referee , c.p .
dullemond , for some extremely useful and insightful suggestions .
this work was supported by nasa under grants nag5 - 13207 and nng04gl01 g from the origins of solar systems and astrophysics theory programs , and by the nsf under grant ast 0407040 . | the spectral energy distributions ( seds ) of some t tauri stars display a deficit of near - ir flux that could be a consequence of an embedded jupiter - mass planet partially clearing an inner hole in the circumstellar disc . here
, we use two - dimensional numerical simulations of the planet - disc interaction , in concert with simple models for the dust dynamics , to quantify how a planet influences the dust at different radii within the disc .
we show that pressure gradients at the outer edge of the gap cleared by the planet act as a filter - letting particles smaller than a critical size through to the inner disc while holding back larger particles in the outer disc .
the critical particle size depends upon the disc properties , but is typically of the order of 10 microns .
this filtration process will lead to discontinuous grain populations across the planet s orbital radius , with small grains in the inner disc and an outer population of larger grains .
we show that this type of dust population is qualitatively consistent with sed modelling of systems that have optically thin inner holes in their circumstellar discs .
this process can also produce a very large gas - to - dust ratio in the inner disc , potentially explaining those systems with optically thin inner cavities that still have relatively high accretion rates .
solar system : formation planets and satellites : formation planetary systems : formation |
current networks aim to support high data rates to end users by increasing the spectral efficiency in bits - per - hertz of the network , at the expense of the energy efficiency of the network .
indeed , an important part of the energy consumption of mobile networks is proportional to the radiated energy , which relies on the frequency bandwidth and the transmission power .
any energy efficient transmission scheme should exploits the whole system bandwidth by allocating the entire available spectrum to each base station .
such an approach , however , leads to significant interference increase and performance degradation for mobiles located at the cell edges .
the key challenge is to balance interference avoidance and spectrum use to reach an optimal spectral efficiency energy efficiency ( ee - se ) trade - off @xcite .
this challenge has been addressed in the past , for instance using frequency / code planning in 2g/3 g networks or with cooperative multiple point antennas in 4 g @xcite .
the aim of our work is to demonstrate the concept of a recently driven interference management scheme that not only leverage the current technology but also achieves greater overall energy efficiency @xcite .
it exploits interference alignment ( ia ) concept for downlink to reduce inter - cell interference without complex coordination .
the theoretical achievements of ia has been largely discussed e.g. @xcite .
it has been concluded that one of the key results is that , under specific conditions , dense and high - power wireless networks are not fundamentally interference limited . as an example , under idealized assumptions , using ia in the setting of an interference channel formed by @xmath0 transmitter - receiver pairs interfering between one another allows an achievable data rate per pair equal to half of his interference - free channel , regardless of @xmath0 @xcite .
strong efforts in the research community have been done to extend ia far beyond the initial @xmath0 -user interference channel .
the recent review @xcite highlights the different technical challenges to be solved before envisioning a practical application among which implementing accurate feedback loops is probably the most important challenge .
but even beyond the practical implementation of ia solutions in a network , the actual model of the network tends to be complex and involve a large number of hypothesis .
these assumptions , or lack thereof , are needed and will play a significant role in the design of ia schemes .
these ia schemes are in return heavily tuned to the specific hypotheses made and may not adapt to all cellular configurations , thereby justifying the need to develop an experimental evaluation of these techniques .
a downlink cellular network is basically an interfering broadcast channel , where base stations ( bs ) transmit towards a number of users and interfere with each other .
however , several tries to extend this approach in the context of cellular networks revealed some limiting improvement @xcite for the following reasons : i ) the direct extension of the ic model to cellular networks relies on defining first the association of each mobile to a given subset of resources .
thus in each cell , the bs decides without coordination which set of resources is given to each mobile . to have a significant gain ,
ia should be performed for users mutually suffering from interference .
ii ) many works rely on a clustering approach @xcite , however , users located at cluster edges can not benefit from any improvement and are still subject to strong interference .
iii ) signaling requirements reduce the theoretical gain of the system . in @xcite , suh
_ proposed an ia scheme for downlink channels by considering a scenario where each bs uses a reduced space for its own transmission , preserving a given free subspace for other cells .
their solution extended in @xcite , allows users in a cell to cancel their dominant interferer as well as the intra - cell interference for users in the same sub - band , achieving one degree of freedom ( dof ) , without any communication between the bss . in other words ,
each bs create an interference - free hole in which the other bss can serve their mobile users .
each mobile measures and feeds back its own free subspace to its main bs . after receiving this information from all associated ue the bs jointly computes a set of precoders and schedules the ue to maximize the overall capacity , under some fairness / priority constraints .
basically , a certain cooperation between cells is not explicit but exists through the feedback channels .
the performance of such scheme relies on many parameters such as synchronization , feedback capabilities , precoders choice ... in this paper , we review the aforementioned non - classic ia scheme for downlink proposed in @xcite and @xcite .
then , we show the implementation scenario in the experimentation shielded room cortexlab located in lyon , france , and we discuss some experimental issues . finally , we present the results of our demo and highlight the capacity gain of the implemented ia scheme over the classical ofdma scheme .
notations : boldface upper case letters and boldface lower case letters denote matrices and vectors , respectively . the superscripts @xmath1 and @xmath2 stand for the transpose and transpose conjugate matrices , respectively .
the following notation @xmath3 denotes the last @xmath4 vectors of an @xmath5-dimension space matrix @xmath6 defined as the left - singular matrix of a matrix @xmath7 , where svd(@xmath7 ) stands for the singular value decomposition ( svd ) .
for our demo , we consider a cellular network with two bs and three user equipment ( ue ) , all equipped with a single antenna . the transmission scheme is based on ofdm with @xmath0 available sub - carriers .
the transmit signal of the interfering bs is written as @xmath8 where @xmath9 is the data vector of length @xmath10 before hadamard precoding . in the implemented scheme ,
each bs preserves @xmath11 dof for the other bss , thereby creating a hole free of interference at all cell - edges .
the maximum number of transmitted streams is then @xmath12 .
@xmath13 is a unitary trunk hadamard matrix of dimension @xmath14 that allows the use of the @xmath15 reduced space .
we denote @xmath16 the fading channel matrix between the @xmath17 bs and the @xmath18 ue .
let the reduced channel denotes @xmath19 , where @xmath20 and @xmath21 refer to the desired and interfering channel , respectively .
the received signal at the @xmath22 ue is written as @xmath23 where @xmath24 is the noise vector .
in addition to the hadamard trunking matrix , each original stream @xmath25 with @xmath26 is carried over the precoding vector @xmath27 ; the @xmath28 vector of a precoding matrix @xmath29 .
the precoded stream vector at the @xmath17 bs is defined as @xmath30 and includes all streams associated to the @xmath17 bs . in order to accomplish a reliable interference management scheme , we review the non - classic ia technique extended in @xcite that calculates an optimal precoding set lying over the free - interference subspace and then selects the best subset that maximizes the total capacity of the downlink interference channel .
the fading channels connecting the ues to their main bs and to the strongest interferer bs are firstly estimated at the all ues .
this is possible through training sequences provided by both bss .
then , each ue determines the interference null space thanks to the svd of the reduced channel @xmath31 .
it is calculated as @xmath32 by applying a linear projection on the interference null space @xmath33 , the received signal is rewritten as @xmath34 in other words , the signal @xmath35 is being received in the inter - cell interference - free hole created by the strongest interferer .
this means that the best for a ue is to have its desired streams aligned with the equivalent channel @xmath36 which guarantees a maximum received power of the desired signal .
it is worth noting that in practice , the ues at cell edges can be interfered from more than one bs .
dealing with such as a case requires either considering the interference as noise or estimating all interfering channels while the transmission is focused in the joint hole created by the union of the interference . for the sake of simplicity , we only assume one interfering bs for our demo . each ue calculates its optimal @xmath11 precoding vectors and feeds them back to the main bs , which in turn collects @xmath37 precoding vector where @xmath38 is the total number of user fixed to three in our demo .
however , the maximum number of streams allowed for a bs is @xmath15 . when @xmath39 , all vectors are used for transmission .
otherwise , i.e. @xmath40 , an @xmath15 precoding vectors must be selected among the total set .
that is , we have to run a scheduling to select which users and streams are best to serve . among the different existing criteria , we choose to maximize the total channel capacity given by @xmath41 where the term @xmath42 stands for the signal to interference and noise ratio at the @xmath43 selected stream after scheduling .
when a scheduling is required , the precoding vectors set can be presented as an underdetermined matrix of dimension @xmath44 .
the problem is seen as selecting the best determined matrix that maximizes the total rate .
the optimal solution is through exhaustive search . for this to happen
, we need to build all possible subset combination and to calculate the achievable rate of each candidate .
the data symbols are carried over the vectors yielding a maximum achievable rate .
the precoding vector of the @xmath43 stream of the @xmath22 user is calculated as @xmath45 where @xmath46 is the @xmath43 row of the reduced interference - free channel .
in other words , the precoding vector aims at maximizing the power of the desired part of the received signal .
hence , each user relies on the fact that its interference - free directions are independent from the interference - free directions of the other ues associated to the same bs , which means that channel diversity between ues has a major impact for achieving a desired performance of the described interference management scheme .
that is , when the channels are almost orthogonal there is no intra - cell interference and the achievable rate is maximum .
in contrast , when the channels between ues are completely correlated , the interference reaches its maximum and the rate is highly degraded . going back to the scheduling problem , it can be expressed as @xmath47 where @xmath48 and @xmath49 .
such a problem is known as a discrete optimization that requires in most cases an exhaustive search to find out the global optimum .
however , this becomes so costly and expensive with the dimension increases .
other sub - optimal algorithms based on heuristic methods or local minima search can be applied with no guarantee of any improved performance . since we do not assume high dimensions systems and we apply non - classic ia technique over only four sub - carriers , the optimal exhaustive search is feasible with low computational cost .
once the best subset is selected , we proceed to calculate the precoding matrix of all streams that cancels the intra - cell interference .
it is given by @xmath50 where the selected precoding vectors are the columns of the matrix @xmath51 and @xmath52 is the identity matrix .
the achievable rate can be then written as @xmath53 where @xmath54 and @xmath55 is the @xmath56 column of @xmath57 .
it can be noticed that @xmath58 and is equal to one when the selected vectors are all orthogonal .
the performance of the addressed non - classic ia approach has been evaluated through exhaustive simulations in different networks and load scenarios .
it has been found that in comparison to a reference scenario without ia , the cell capacity can on average be increased by a factor @xmath59 and that the spectral efficiency of cell edge users can be increased up to a factor @xmath60 @xcite .
our work herein gives a proof of concept and focuses on the main challenges related to the non - classic ia approach for downlink , namely the knowledge of the interference footprint and the scheduling algorithms to make use of the interference information to maximize the spectral efficiency .
we implement an experimental scenario with two transmitting bss , i.e. the main bs and an interfering one and three ues associated to the main bs .
we show that the ia is feasible , in the sense that ues are able to measure a channel , feed it back to the bs which in turn applies a scheduling to select the best set and calculates the theoretical spectral efficiency ( se ) gain over a classical dumb ofdma allocation .
a wireless network is emulated on cortexlab ( http://www.cortexlab.fr ) @xcite , a controlled hardware facility located in lyon , france with remotely programmable radios and multi - node processing capabilities . during the live demo , a control laptop
is remotely connected to the facility , deploying software on the radios and launching an ia scenario and collecting real - time performance feedback .
the efficiency gain of ia is then shown for various experimental conditions that can be tuned from the control laptop .
represented by the national instruments usrp 2932 , the general - purpose software defined radio ( sdr ) nodes use the gnu radio toolkit for rapid prototyping of transmission techniques mostly reliant on the general purpose processor ( gpp ) of the host pc .
the usrp 2932 is a high end radio platform , counting with a 400 mhz 4.4 ghz rf board , data rates of up to 40 mhz ( with reduced dynamic range , nominal band of 20 mhz ) , a precise ocxo clock source and a @xmath61 gigabit ethernet ( gige ) connection to the host pc .
the host pc is based on a linux environment and allow users to test phy and mac layer techniques .
it is preferable to first use both linux and gnu radio for the development and test at the user s own computer and then to bring the experiment over to cortexlab . among the forty available sdr nodes in cortexlab ,
five are selected where two serve as main and interfering bs and the remainders serve as mobile ues .
all transmitters ( tx ) and receivers ( rx ) are equipped with single antenna .
therefore , the ia scheme lies on the frequency dimensions provided by the ofdm transmission scheme implemented at all txs and rxs . in the first transmission phase ,
the ues need to estimate the main and interfering channels .
this can be done by assigning two orthogonal time slots to the bss through which they transmit the training sequences .
however , the synchronization of the two bs nodes remains a challenge since they are distantly separated .
herein , we propose an over - air synchronization as follows .
holding a unique i d , the interfering bs starts the transmission to let the user nodes measure the interference channel coefficients while the main bs tries to decode the i d of the transmitter .
if the decoded i d corresponds to the interferer , the main bs node transmit ofdm symbols ( one or more ) for channel estimation . by collecting both estimated channels ,
each receiver is able to calculate the @xmath4 optimal precoding vectors in the free dimensions such that the desired signal power is maximized . in order to avoid imperfections in the ia scheme , we use a wire connection provided in the experimentation room between all nodes to perform a perfect feedback to the tx ( main bs ) .
this latter gathers the precoding vectors from all ues and run the scheduling algorithm to seek the best @xmath62 precoding vectors that minimizes the intra - cell interference , and hence maximizes the achievable data rate of the cell . the maximized rate
is then compared to the classical ofdma and the theoretical gain is calculated .
the ofdma scheduler select at each users the @xmath63 streams that maximizes the signal - to - interference and noise ratio ( sinr ) given by @xmath64 where @xmath65 stands for the sinr , @xmath66 is the channel coefficient between the main bs and the @xmath67 user at the @xmath43 stream , @xmath68 is the interference channel coefficient at the @xmath43 stream and @xmath69 is the noise variance .
the classical ofdma achievable rate is then defined as @xmath70 in the live demo , we show a spectral efficiency gain that can largely vary between @xmath61 and @xmath71 .
this variation is due to the influence of many factors summarized by the following parameters : the noise power , the interference power , the channel diversity , the distance between the different nodes , the tx gain ... for instance , in perfect conditions the ratio gain @xmath72 given in ( [ eq9 ] ) and ( [ eq11 ] ) , tends to its maximum when the inter - cell interference power is of the same order or higher than the desired signal power ; i.e. sinr is low , this is the case of cell edge mobile users .
however , as long as the users get closer to the main bs , this ratio decreases .
another critical parameter that highly impacts the performance gain is the channel diversity and correlation , the less the channel is correlated the higher the efficiency gain is .
this is because the applied ia scheme requires a completely decorrelated channel coefficients for the scheduling , otherwise the intra - cell interference dominates the desired signal . in order to get a better decorrelation in the shielded room ,
we emulate a virtual channel on all txs and rxs , however , this still induces some correlation at the different mobile ues .
a better way is to use a multi - antennas node at the tx to generate multi - paths through the different antennas .
each path is randomly attenuated , phase shifted and delayed .
such a multi - paths generation creates a perfect decorrelation since the paths generated by the different nodes are totally independent .
an illustration of our demo is given in figure [ fig2 ] . in our demo , we assume two bss and @xmath73 users . the ia is applied over four sub - carriers which gives a total dimension of @xmath60 , one is free of interference and three other used by each bs .
we focus on the average theoretical capacity gain offered by the non - classical ia over the classical ofdma scheme with respect to the number of transmission as shown in figure [ fig3 ] .
we also plot the channel spectrum at the all sub - carriers and for the different transmission to show how decorrelated the channel coefficients are .
for each channel realization , we display the snr of the interference - free streams and the sinr of all streams when the classical ofdma is applied to see their impact over the efficiency gain .
in this demo , we have implemented the first phase of a non - classic ia approach for interference management .
it consists in measuring the channel coefficients , calculating the interference - free sub - spaces , feeding them back to the main bs , and applying a scheduling to decide which stream to transmit .
we have shown in experiment that a significant theoretical capacity gain over the classical ofdma scheme can be achieved depending on the channel conditions and the interference dominance .
the remaining work is to start the ia transmission and to apply linear decoding criteria at the receiver such as zero - forcing ( zf ) or minimum mean squared error ( mmse ) to recover the original data .
however , the challenge here is to synchronize the transmission between both txs for the ia transmission and to compare the theoretical gain to the practical one .
it is also worth trying to face more practical issues related to de - synchronization , which means to study the impact of the delay and phase shift between the different txs on the ia scheme .
m. a. maddah - ali , a. s. motahari , and a. k. khandani , `` communication over mimo x channels : interference alignment , decomposition , and performance analysis '' , _ information theory , ieee transactions on _ , 2008 , vol .
54 , no 8 , 3457 - 3470 .
a. gshasemi , a. s. motahari , and a. k. khandani , `` interference alignment for the k user mimo interference channel '' , _ in information theory proceedings ( isit ) , 2010 ieee international symposium on _ , pp .
360 - 364 , ieee , june 2010 . o. el ayach , s. w. peters , and r. w. heath jr , `` the feasibility of interference alignment over measured mimo - ofdm channels '' , _ vehicular technology , ieee transactions on _ , 2010 , vol .
59 , no 9 , 4309 - 4321 .
d. aziz , f. boccardi , and a. weber , `` system - level performance study of interference alignment in cellular systems with base - station coordination '' , _ in personal indoor and mobile radio communications ( pimrc ) , 2012 ieee 23rd international symposium on _ , pp .
1155 - 1160 , ieee , september 2012 .
r. tresch , m. guillaud , and e. riegler , `` on the achievability of interference alignment in the k - user constant mimo interference channel '' , _ in statistical signal processing , 2009 .
ssp09 , ieee / sp 15th workshop on _ , pp .
277 - 280 , ieee , august 2009 .
l. s. cardoso , a. massouri , b. guillon , f. hutu , g. villemaud , t. risset , and j. m. gorce , `` cortexlab : a cognitive radio testbed for reproducible experiments '' , _ in proc .
wireless@ virginia tech symposium
_ , may 2014 a. massouri , l. cardoso , b. guillon , f. doru - hutu , g. villemaud , t. risset , and j .-
gorce , `` cortexlab : an open fpga - based facility for testing sdr and cognitive radio networks in a reproducible environment , '' _ in ieee infocom _ , ieee , april 2014 . | our demo aims at proving the concept of a recent proposed interference management scheme that reduces the inter - cell interference in downlink without complex coordination , known as non -classic interference alignment ( ia ) scheme .
we assume a case where one main base station ( bs ) needs to serve three users equipments ( ue ) while another bs is causing interference .
the primary goal is to construct the alignment scheme ; i.e. each ue estimates the main and interfered channel coefficients , calculates the optimal interference free directions dropped by the interfering bs and feeds them back to the main bs which in turn applies a scheduling to select the best free inter - cell interference directions .
once the scheme is build , we are able to measure the total capacity of the downlink interference channel .
we run the scheme in cortexlab ; a controlled hardware facility located in lyon , france with remotely programmable radios and multi - node processing capabilities , and we illustrate the achievable capacity gain for different channel realizations . |
the motif , defined as a small connected subgraph that recurs in a graph , is the basic building block , or functional unit , of complex networks @xcite . in real - world networks
( e.g. , gene regulatory networks ) , motifs represent the elementary interaction patterns between small groups of nodes , and the relative frequencies with which motifs appear represent different functions of the network @xcite .
although it has been found that there is a topological relationship between the large - scale attributes ( scale - free and hierarchical ) and local interaction patterns ( subgraph based ) @xcite , it remains unclear whether there is a relationship between small functional units and other structure properties such as rich - club connections of complex networks . in our previous study
we find that rich - club connections can dominate some global properties ( e.g. , assortativity and transitivity ) of a network @xcite , which implies the possible relation between the rich - club property and the network s subgraph organization .
the rich - club property refers to the organization pattern of rich nodes @xcite , especially whether rich nodes tend to connect to one another , or with the remaining nodes @xcite .
because rich nodes often play a central role in the static property of , and dynamic processes on , complex networks @xcite , significant attention has been paid to the prominent effects of the richest elements @xcite and the organization among them @xcite .
a systematic framework is needed to clearly understand the roles of rich nodes in different real - world networks with distinct degree distributions . in this study , we find the influences of rich nodes and their organization pattern depend largely on the degree distributions of complex networks .
rich nodes are important in scale - free networks @xcite , because a power - law degree distribution indicates that the majority of nodes participate in at most one or two motifs , while a few rich nodes take part in a very large number of small subgraphs . manipulating a very small number of rich - club connections
therefore can strongly affect the frequencies of the basic functional blocks ( motifs ) for a heterogeneous network . in comparison , for the network with a homogeneous degree distribution ( e.g. , the network of us power grid ) ,
the links among rich nodes show a tiny effect on the whole network .
the main reason behind this is that all nodes ( including rich nodes ) in a homogeneous network are engaged in only a few interactions , and there are no hubs linking to a significantly larger number of other nodes .
these results are helpful in understanding the origin of motifs and motif clusters in real - world complex networks , and the mechanisms by which how small subgraphs aggregate into larger superstructures .
our finding has an important potential application : we can build a framework to optimize and control the functional behaviors of complex networks . in most cases we can not regenerate or redesign a real - world network , but manipulating a small number of rich - club connections
gives us a chance to optimize the structure of the network and control the relative frequencies of small functional units in a predictable manner .
furthermore , although pioneer studies have developed a series of methods to judge whether a network has rich - club properties @xcite , these approaches are based on how many links there are among rich nodes instead of how these links affect the whole network . based on subgraph ratio profile , the topological structure among rich nodes can be uncovered from the inspection of the basic functional units . in this study
we develop a novel method to judge whether a network has a rich - club or not .
the new method does not calculate how many links connect to rich nodes compared with its randomized version while it depends on how the organization pattern of rich nodes affects the appearance of different motifs .
taken together , these findings indicate the strong ties between the local subgraphs and rich - club properties of complex networks , which complements our understanding of a network s topological and functional organization . because each network can be characterized by a set of distinct types of subgraphs and rich - club connections are a significant property , our findings are expected to provide new insights in understanding the evolution of dynamical networks and design new models for characterizing real - world networks .
our work is a step in an ongoing effort to bridge the local topology of a network and its global statistical features .
here we select the top @xmath0 of the nodes with the highest degree as rich nodes in a network and manipulate the connections among them .
we use link rewiring algorithms to generate the network with rich - club and the network without rich - club , respectively .
the basic idea is very similar to the random rewiring method @xcite , while the main difference is that our new method only switches the links among rich nodes and a small number of low - degree nodes .
first we make rich nodes fully connected to one another , so they form a completely connected rich - club .
secondly , we completely eradicate the edges among rich nodes , so that the network has no rich - club .
now we specify the rewiring algorithms .
first we make all rich nodes fully connected to generate a network with a significant rich - club .
if there is a link between two rich nodes , their structure remains unchanged [ fig .
[ fig1](a ) ] .
if there is no link between two rich nodes , we perform the operation from fig .
[ fig1](b ) to [ fig1](a ) .
that is , we select another two low - degree nodes that respectively connect to the two rich nodes while do not connect to each other [ fig . [ fig1](b ) ] .
then we cut the two links between the rich nodes and their low - degree neighbors , and connect the two rich nodes as well as the two low - degree nodes , respectively [ fig .
[ fig1](a ) ] .
after repeating this process until all rich nodes form a completely connected rich - club , we can obtain the network with a full - connected rich - club [ fig . [
fig1](c ) ] .
secondly , we completely eradicate the edges among rich nodes , so that the network has no rich - club property .
if there is no link between two rich nodes , we will do nothing [ fig .
[ fig1](b ) ] .
if there is a link between two rich nodes , we do the operation from fig . [ fig1](a ) to [ fig1](b ) .
we randomly select another pair of low - degree nodes which connect to each other while do not connect to either of the two rich nodes [ fig .
[ fig1](a ) ] .
then we cut off the links both between the two low - degree nodes and between two rich nodes respectively , and let each rich node connect to one low - degree node [ fig .
[ fig1](b ) ] . repeating the above process until the links among the whole rich nodes are completely eradicated
, we will get a network without rich - club property [ fig . [
fig1](d ) ] .
because we use the rewiring method , the degree of every node in the original network exactly remains unchanged .
for the topological structure of the original network , there is only small variation induced by manipulating rich - club connections , so we can monitor how the subgraph frequencies are affected by the rich - club property .
furthermore , we can compare the results of the subgraph ratio profile for the original network , the network with rich - club , and the network without rich - club , to make more reliable inference of whether the original network has a rich - club property or not .
each network will be scanned for all possible @xmath1-node subgraphs ( we choose @xmath2 ) . in a network with a skewed degree distribution ,
rich nodes have much higher degrees than the overwhelming majority , so whether they connect to each other to form a rich - club will strongly affect the frequencies of subgraphs .
actually , rich nodes can absorb a very large number of subgraphs and form a motif cluster .
for example , the triangles may not distribute uniformly within a scale - free network but tend to aggregate around the hubs , because a node with @xmath3 links can carry up to @xmath4 triangles @xcite .
the aggregation of motifs into motif clusters is important , because it implies that the potential functional properties of the large number of subgraphs also need to be evaluated at the level of subgraph clusters instead of being evaluated only at the level of a single subgraph .
exploring rich - club connections provides a new way to evaluate the functional properties of abundant subgraphs at the level of subgraph clusters .
a few rich nodes usually take part in a very large number of small subgraphs and they can form motif clusters in real - world complex networks . actually , the organization of rich nodes can dominate the appearance of particular motifs prominently . in the non - rich - club network [ fig .
[ fig2](b ) ] , rich nodes do not tend to connect to each other , so the non - rich - club subgraphs [ fig . [
fig2](d ) ] will be more common .
on the contrary , in the rich - club network [ fig . [ fig2](c ) ] , rich nodes trend to connect to each other , so the network will demonstrate a larger number of the rich - club motifs [ fig . [
fig2](e ) ] .
in the original network [ fig . [ fig2](a ) ] , rich nodes may or may not connect to each other . comparing the appearing frequencies of motifs in the above three networks , we can conclude whether the original network has a rich - club property .
it is obvious that by considering the subnetworks of rich nodes , the frequencies of the non - rich - club motifs and/or rich - club motifs are remarkably more than those of the randomized versions of the subnetworks .
the inherent existence of two distinct classes of subgraphs ( non - rich - club motifs and rich - club motifs ) in a heterogeneous network demonstrates that , in contrast to the homogeneous network , the highly abundant motifs can not exist in isolation but must naturally aggregate into subgraph clusters .
specifically , in the network with a rich - club , the neighbors of a highly connected node are linked to each other , therefore the chance that low - degree nodes participate in highly connected subgraphs is slim . in a homogeneous network , however , all nodes are engaged in only a few interactions and the appearance of motifs is the statistical average of the whole network , for there are no hubs linking to a significantly higher number of other nodes to form motif clusters .
table [ table1 ] lists the results of six undirected networks ( including three real - world networks and three model networks ) arranged with @xmath5 increasing .
the value of the structural cutoff degree @xmath6 can be regarded as the first approximation of the maximum degree within a scale - free network @xcite . here
@xmath5 is a convenient index that can be used in complex networks with any degree distribution to show the proportion of links ( or degrees ) rich nodes possess in comparison with the remaining nodes in a network @xcite .
.statistics of six undirected networks : number of nodes @xmath1 , average degree @xmath7 , the exponent of degree distribution if the distribution follows a power law : @xmath8 ( or `` '' if not ) , structural cutoff degree @xmath9 @xcite , maximal degree @xmath10 .
sw is the network generated by the small - world model @xcite , pg is the network of us power grid @xcite , ba is the network generated by the scale - free model @xcite , epa is the network from the pages linking to www.epa.gov @xcite , pfp is the network generated by the model for the internet topology @xcite and as is the network of the internet topology at the level of autonomous systems @xcite . [ cols="^,^,^,^,^,^,^",options="header " , ] [ table2 ] the above results indicate that rich nodes in homogeneous networks ( e.g. , sw and pg ) only have a very limited effect on the whole network , for all nodes ( including rich nodes ) in such type of networks are engaged in only a few interactions .
obviously , rich - club connections are more involved in the heterogeneous - motif in a heterogeneous network , for a node with higher degree has a more chance to participate in this structure . in a heterogeneous network
( e.g. , ba , epa , pfp and as ) , a few rich nodes with much more links than the overwhelming majority can absorb a very large number of subgraphs and form motif clusters , which makes rich - club connections more influential to the whole network . the percentage of the sum of heterogeneous - motif and homogeneous - motif for all the networks is very high ( up to @xmath11 ) , which means other types of motifs are relative sparse compared with the above two types of motifs .
therefore , to form a specific functional block , the absolute frequencies of a particular subgraph are not necessary very large .
actually , it is enough for the relative frequencies of the motif for the original network are statistically higher than those for its randomized version @xcite .
moreover , in view of the difficulty in forming the specific functional blocks in a randomized network , the sparse distribution of other motifs gives us a chance to control the appearance of small functional subgraphs in real - world networks by manipulating rich - club connections . because undirected networks have only two types of triads ( unclosed triple and triangle )
, we only analyze the profile of the six types of undirected connected tetrads ( @xmath12-node motifs ) .
the normalized @xmath13 scores of tetrads show a significant dependence on the network size , so we use the abundance of each subgraph @xmath14 relative to random networks @xcite : @xmath15 where @xmath16 ensures that @xmath17 is not misleadingly large when the subgraph appears very few times in both the real and random networks . the subgraph ratio profile ( srp )
is the vector of @xmath17 normalized to length 1 : @xmath18 network motifs , which are patterns of interconnections occurring in complex networks are significantly higher than those for randomized networks @xcite .
the motif pattern reflects the local structural properties of complex networks and thus can be used to classify networks .
if different types of networks share the similar result of srp , these networks can be classified into the same `` superfamily '' @xcite .
the networks in the same triad superfamily share not only some particular types of motifs , but also very similar proportions of all types of subgraphs . here
we show the srp results for the original network , the network with rich - club , and the network without rich - club in fig .
if the above three networks belong to the same superfamily , it means that the rich - club property has weak effect on the original network , and this result shows the network is a homogeneous network .
if the three networks belong to different superfamilies , it means that rich - club connections can strongly affect the structure and function of the original network , and this result indicates that the network is heterogeneous .
furthermore , according to the fact that the original network belongs to the same superfamily as the network with rich - club or the network without rich - club , we can judge whether the original network has a rich - club property .
.,scaledwidth=48.0% ] the networks of sw and pg have a homogeneous degree distribution , so rich nodes in the two networks are not significantly higher than others .
therefore , as has been shown in figs .
[ fig3](a ) and [ fig3](b ) , whether the two networks have rich - club properties does not have any influence on srp .
moreover , the original network , and the networks with and without rich - club all belong to the same superfamily .
the above results indicate whether a homogeneous network has a rich - club property is not very important , and rich - club connections can not control the functions of such type of networks . because the networks of ba and epa have a heterogeneous degree distribution , rich nodes possess much more links than the overwhelming majority .
therefore , whether the two networks have rich - club properties can greatly affect the result of srp . as is shown in figs .
[ fig3](c ) and [ fig3](d ) , the original network and the network with rich - club do not belong to the same superfamily .
conversely , the original network and the network without rich - club belong to the same superfamily , so ba and epa both have no rich - club property . for the networks of pfp and as
, they not only have a heterogeneous degree distribution but also have a few superrich nodes , so whether the two networks have a rich - club can affect the result of srp most significantly .
the original network and the network without rich - club do not belong to the same superfamily . for the original pfp and the network with rich - club belong to the same superfamily , pfp has the property of rich - club as is shown in fig .
[ fig3](e ) .
basically we can say that as has a rich - club property , for the original as has the very similar srp to the network with rich - club , except for the motif @xmath19 ( @xmath12-node clique ) in fig .
[ fig3](f ) . yet
the non - consistency of motif @xmath19 for the original network and the network with rich - club may be the origin of arguments on whether the internet topology has a rich - club property @xcite .
in conclusion , we find that the influences of rich - club connections strongly depend on the degree distributions of complex networks .
our findings show that in a homogeneous network , whether the network has a rich - club or not is not very important for its structure and function . while rich - club connections in a heterogeneous network have a crucial implication , for they can partially optimize and control the function of the whole network . our new framework for measuring the subgraph ratio profile can provide a more impartial judgement on whether a network has a rich - club .
previous studies put more attention on finding whether the links among rich nodes appear more frequently in the original network compared with its randomized counterparts @xcite . while the actual influence of rich - clubs in different degree distribution networks has not been studied .
our approach which is based on the effect of the rich - club on the network structure and function , is therefore more advanced .
we demonstrate that strong ties between the rich - club property and local ( subgraph - based ) structure underscore the importance to understand the properties of complex networks as fully integrated systems . indeed , the abundance of some kinds of local interaction patterns reflects the rich - club property of a network , raising intriguing questions about the role of local events in shaping a network s overall behavior @xcite .
these results indicate that the analysis described here may have an impact on our understanding for other types of subgraphs ( e.g. , cliques @xcite and cycles @xcite ) in complex networks .
our results show the significance of the rich - club property and motif distributions in modeling and designing real - world networks @xcite .
an appropriate model should have similar structure and function to the real - world network . to meet this demand
the model can be designed from the basic motifs or the subgraph ratio profile , which can be easily controlled by the rich - club property .
our findings also deepen our understanding of the evolution of dynamical networks .
the existence of the dense rich - club motifs and/or non - rich - club motifs in real - world networks may be a unifying property of evolved systems , so it is interesting to understand the rich - club concept from the perspective of network evolution .
we conjecture that the common origin of the local functional blocks and the rich - club property is primarily the same , because neither the density and topology of subgraphs nor the rich - club property can be dissociated from the evolution of the overall network . following the framework in this work
, we will contrive to bridge the gap between local topologies of a network and its global statistical features in future . | traditionally , there is no evidence suggesting that there are strong ties between the rich - club property and the function of complex networks . in this study
, we find that whether a very small portion of rich nodes connected to each other or not can strongly affect the frequency of occurrence of basic building blocks ( motif ) within networks , and therefore the function , of a heterogeneous network .
conversely whether a homogeneous network has a rich - club property or not generally has no significant effect on its structure and function .
these findings open the possibility to optimize and control the function of complex networks by manipulating rich - club connections .
furthermore , based on the subgraph ratio profile , we develop a more rigorous approach to judge whether a network has a rich - club or not .
the new method does not calculate how many links there are among rich nodes but depends on how the links among rich nodes can affect the overall structure as well as function of a given network .
these results can also help us to understand the evolution of dynamical networks and design new models for characterizing real - world networks .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the function properties of complex networks properties such as synchronizabilty , and efficiency of information transport depend sensitively on the detailed topological structure of the particular network .
we show that , despite this , the functional behavior of a complex network can be largely controlled by rewiring a very small fraction of nodes within the network . for networks with scale - free degree distribution ( heterogeneous networks )
the connectivity pattern among the highest degree nodes determines the functional behavior of the entire network .
that is , whether a network exhibits a dominant rich - club ( whether high degree nodes are mutually connected or not ) will determine the functional behavior of the entire network .
for random graphs and other networks with homogeneous degree distribution this is not the case .
our results provide a mechanism by which the behavior of real - world networks can be effectively controlled by rewiring only a very small portion of links .
this suggests the likely mechanism by which real - world networks such as the internet and gene regulatory networks in various organisms evolve .
the connectivity among a small portion of hub nodes will control the functional behavior of the entire network , and thus a real network can be made more robust or provide more efficient information processing by rewiring those links .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ |
observations with the _ hubble space telescope _ ( _ hst _ ) have revealed that intrinsic uv absorption is a common phenomenon in seyfert 1 galaxies , appearing in over half of the objects with available spectra @xcite .
the uv absorption resonance lines are typically blueshifted in the rest frames of the host galaxies , indicating radial outflow .
the absorption is often highly variable and , in some cases , only partially covers the continuum and emission - line regions of the active nuclei , which implies the absorbers are intrinsic to the agn environments .
the observed variations may be due to changes in ionization in the absorbers @xcite or changes in the total absorbing column , i.e. , as a result of motion into and out of our line - of - sight @xcite .
ngc 3783 is a bright seyfert 1 galaxy that exhibits strong uv absorption features and x - ray warm absorption " .
several observations with _
hst _ over the past decade have revealed dramatic variability in the uv absorption .
ngc 3783 also has a highly variable uv continuum source .
an _ international ultraviolet explorer _
( _ iue _ ) monitoring campaign revealed a factor of @xmath162 flux variations over timescales of 2040 days @xcite .
intrinsic absorption in ly@xmath3 and @xmath171548,1551 was first detected in ngc 3783 with faint object spectrograph ( fos ) observations by @xcite .
three subsequent spectra obtained over a period of about two years with the goddard high resolution spectrograph ( ghrs ) revealed highly variable absorption @xcite .
there was no detectable absorption in the 1993 february ghrs spectrum , however , by 1995 april , two kinematic components appeared , at radial velocities of @xmath11365 km s@xmath2 ( referred to as component 1 ) and @xmath1548 km s@xmath2 ( component 2 ) relative to the systemic redshift ( we adopt @xmath18 0.009760 @xmath19 0.000093 throughout this paper ; * ? ? ? * ) .
additionally , a ghrs spectrum of the @xmath171239,1243 spectral region revealed absorption coincident in velocity with component 2 just 16 days after the 1993 february observation of that showed no absorption , suggesting rapid variability @xcite .
a space telescope imaging spectrograph ( stis ) medium resolution echelle spectrum of ngc 3783 was obtained on 2000 february 27 revealing a third kinematic component in ( @xmath20724 km s@xmath2 , component 3 ) , in addition to the components seen in the final ghrs spectrum @xcite . ly@xmath3 and @xmath171239,1242 also appeared in these three kinematic components in the stis spectrum .
@xmath171394,1403 absorption was only found in component 1 and no lines from lower ionization species , i.e. , , , mg ii , were detected . using the doublet lines
, @xcite found that all three absorption systems have a non - unity effective covering factor .
no correlation was found between the strength of the uv absorption features and continuum flux in the ghrs and stis spectra by @xcite .
they concluded the observed absorption variations were due largely to a change in total column .
however , it was not possible to constrain tightly the variation timescales due to the sampling of these observations
. additionally , these observations did not sample the rapid changes in the continuum flux that were observed in the study by @xcite , thus , it remains unclear what affect variable ionization has on the absorption .
a large ( @xmath21 10@xmath22 @xmath7 ) and variable column of ionized gas was measured in the x - ray spectrum of ngc 3783 with the _ advanced satellite for cosmology and astrophysics _ ( _ asca _ ) ( e.g. , * ? ? ?
subsequent observations with the high energy transmission grating spectrometer ( hetgs ) aboard the _ chandra x - ray observatory _
( _ cxo _ ) showed numerous absorption lines with a mean radial velocity of @xmath23610 @xmath19 130 km s@xmath2 , consistent with components 2 and 3 identified in the uv @xcite .
the coincidence in velocities suggests a link between the uv and x - ray absorbers .
photoionization modeling of the stis spectrum revealed that , although the uv absorbers can produce some of the observed x - ray columns , the ionization is too low and the total column too small to account for all of the features in the x - ray @xcite . hence , the exact relationship of the uv and x - ray absorption remained uncertain .
additionally , @xcite found that two zones are required to explain the uv absorption columns measured in component 1 .
the strength of the absorption is inconsistent with the large / column density ratio measured in the stis spectrum .
they proposed a model where a relatively high density absorber is co - located with more tenuous gas that is more highly ionized .
various studies of intrinsic absorption in agns have demonstrated the importance of determining the absorption covering factor in the line - of - sight to the nucleus .
in addition to affecting column density measurements , the covering factor constrains the absorption and emission geometry , providing tests for physical models .
non - unity effective covering factors , @xmath24 , can arise from scattering of light into our line - of - sight by an extended scatterer ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) or partial occultation of the emission sources @xcite . in the most general case ,
the different covering factors of the continuum and emission - line region need to be determined @xcite , and @xmath24 can vary across the profile of an individual absorption component ( e.g. , * ? ? ?
additionally , instrumental scattering can affect the derived @xmath24 and column densities @xcite .
high - resolution spectra with high signal - to - noise ( s / n ) are required to separate these effects .
we have undertaken an intensive multiwavelength monitoring campaign to probe the intrinsic absorption in ngc 3783 using observations in the uv , far - uv , and x - ray . in this paper
, we present an analysis of the averaged spectrum in the uv and far - uv from stis and the _ far ultraviolet spectroscopic explorer _ ( _ fuse _ ) .
we have co - added all observations to produce a high s / n spectrum that samples numerous lines from a range of ionization states .
analysis of the mean x - ray spectrum is presented in @xcite , hereafter paper i. details of the variable nature of the absorption will be presented in future papers by gabel et al .
( in preparation ) and george et al . ( in preparation )
. this paper is organized as follows : in 2 , we present details of the stis and _ fuse _ observations and data reduction ; in 3 , measurements of the covering factors and column densities for the uv absorbers are given ; in 4 we derive constraints on the physical conditions and geometry of the absorbers ; we summarize our results in 5 .
the nucleus of ngc 3783 was observed with _ hst_/stis at 18 epochs between 2000 february 27 to 2002 january 6 , and with _ fuse _ at 5 epochs between 2001 february 28 to 2001 june 30 .
each stis medium - resolution echelle spectrum was obtained in two orbits through the 0@xmath252 @xmath13 0@xmath252 aperture centered on the nucleus using the e140 m grating , which covers 1150 - 1730 .
exposure times and observation dates are given in table 1 .
all spectra were reduced using the idl software developed at nasa goddard space flight center for the instrument definition team .
we incorporated a procedure to remove the background light from each order using a scattered light model devised by @xcite .
each echelle spectrum was generated by averaging together the individual orders in regions of overlap .
the extracted stis spectra are sampled in @xmath165 km s@xmath2 bins .
we measured residual fluxes in the cores of saturated galactic lines in the stis spectra ( ly@xmath3 , @xmath261302 , @xmath261335 , @xmath261527 ) to test the accuracy of the removal of scattered light .
we find in all cases mean flux levels measured over the cores are consistent with zero within the noise levels .
typical values for the mean fluxes in the cores are @xmath19 03 % of the local continuum flux , compared to uncertainties due to noise ( i.e. , the standard deviations measured over these intervals ) of @xmath16512 % of the local continuum flux . the _ fuse _ spectra , which cover 905 - 1187 , were obtained in time - tag mode with the 30@xmath27 @xmath13 30@xmath27 low - resolution aperture ( lwrs ) .
each observation consists of multiple exposures obtained in consecutive orbits on four detector segments , with two spectra ( from the sic and lif channels ) on each segment .
details of the observations are given in table 2 .
using idl software obtained from the _ fuse _ data analysis web page ( http://fuse.pha.jhu.edu/analysis/ analysis.html ) , we examined each raw two - dimensional image for instrumental effects that may contaminate the data .
event bursts " were found ( see discussion in * ? ? ? * ) and an instrumental artifact affects the long wavelength region of all lif 1b spectra .
this artifact , known as the worm " , is due to shadowing by one of the grid wires above the detector @xcite .
the raw data were screened to remove these effects and processed using the standard _ fuse _ calibration pipeline , calfuse v1.8.7 .
calfuse extracts eight one - dimensional spectra for each exposure , corresponding to the two channels and four detector segments .
since the extracted _ fuse _ spectra are oversampled , we rebinned to 8 km s@xmath2 bin@xmath2 to achieve higher s /
n while preserving the full instrumental resolution ( one resolution element corresponds to two data bins ) , which is nominally @xmath1620 km s@xmath2 @xcite . for each channel / segment , we compared the wavelength scales for all exposures using cross - correlation and corrected for any offset .
all exposures were then co - added , weighted by exposure time . to achieve the highest possible s
/ n , the resulting spectra for the eight channel / segments were also co - added . to correct for the non - linear shifts in wavelength scale in the detector segments , we cross - correlated and aligned the spectra over small bandpasses .
the spectra were then co - added by weighting each channel / segment with its effective area , using the effective area versus wavelength functions given in @xcite .
the absolute wavelength scale was determined from the intrinsic wavelengths of the galactic lines .
we measured flux levels in the cores of saturated galactic h@xmath28 lines , and find they are consistent with zero within the noise of the data .
the individual stis and _ fuse _ observations were co - added to produce an averaged spectrum for each instrument .
we have tested the individual observations extensively and find that our results in this paper are not strongly sensitive to variability in the spectra . specifically , we performed all measurements described in the subsequent analysis on each stis and _ fuse _ observation , as well as on various coadded combinations of the individual observations .
we find that the measurements of the covering factors in 3.1 and 3.2 below are not affected by variability , within the uncertainties of those measurements .
there are variations in column densities for some of the ions , and these will be given a full treatment in an upcoming paper ( gabel et al .
[ in preparation ] ) . thus , the column density measurements presented in 3.2 are average values . in figure 1 , the spectra for the + ly@xmath4 , ly@xmath3 + , , and resonance lines are plotted .
wavelengths corresponding to four previously identified intrinsic absorption components are denoted with tick marks for each line .
components 13 are those identified in the initial stis spectrum by @xcite .
component 4 , at v@xmath291027 km s@xmath2 in the rest frame of the host galaxy , was first identified in and ly@xmath4 in an earlier _ fuse _ spectrum @xcite .
we note that absorption in this component is weak or absent in all other lines and its detection is very sensitive to the fit to the intrinsic emission .
additionally , galactic @xmath261032 absorption may contribute to the observed ly@xmath4 feature ( this will be addressed further in a study of the galactic absorption by mathur et al . [ in preparation ] ) .
given these uncertainties , we consider component 4 to be a tentative detection of intrinsic absorption .
interstellar absorption lines and detector artifacts are also identified on the top of each spectrum in figure 1 .
the local interstellar absorption includes three kinematic components , at 0 , + 60 , and + 240 km s@xmath2 with respect to the galactic rest frame @xcite .
we modeled the intrinsic continuum and emission line fluxes at each absorption feature in the following manner ( " intrinsic flux here refers to the agn emission before passing through the absorbers ) .
first , we modeled the intrinsic continuum _ plus _ emission - line flux by fitting a cubic spline to unabsorbed regions of the spectrum on either side of the line .
next , to determine the relative contribution of continuum and line emission at each absorption feature , we estimated the _ continuum _ flux levels at each absorption feature by measuring the flux in regions of the spectrum that are not affected by line emission and linearly interpolating between these regions . due to heavy contamination of the ly@xmath3 and spectral regions by galactic absorption , we used the emission - line profile as a template in fitting the intrinsic emission for these lines .
specifically , we subtracted our fit to the continuum flux from our fit to the total emission spectrum to derive its emission - line profile .
we then scaled this profile in flux , preserving its velocity , and added it to the continuum flux for ly@xmath3 and to derive the fit to these lines .
we find that the line template does not fit the emission profile fully ; an additional , narrow emission component is required .
we modeled this with a fwhm @xmath30 350 km s@xmath2 gaussian profile centered at the systemic velocity of ngc 3783 .
since this narrow line region component does not appear in the emission - line profiles in the stis spectrum , it likely arises in an extended region that is not fully covered by the 0@xmath252 @xmath13 0@xmath252 stis aperture .
the broad line region ( blr ) emission from ly@xmath4 and higher order lyman lines is negligible in the _ fuse _ spectrum .
we have generated comparison models using the photoionization code cloudy @xcite for typical conditions in blr gas and find this is consistent with the predicted spectrum .
multiple absorptions of the higher order lyman line photons followed by cascade produce an spectrum dominated by ly@xmath3 .
thus , the line emission that is absorbed by ly@xmath4 is high velocity blr emission .
our fits to the total ( dashed line ) and continuum ( dot - dashed line ) fluxes are plotted on the spectra in figure 1 .
we have modeled the galactic h@xmath28 and absorption , using the parameters derived in @xcite for the high velocity cloud component .
the fits to , ly@xmath4 and ly@xmath3 shown in figure 1 include absorption by this model . in figure 2
, we show the spectra of ly@xmath4 and illustrating the kinematic link betweeen the uv and x - ray absorption .
the profile is a combination of the four strongest lines from that ion in the mean _ chandra _ spectrum ( see figure 10 in paper i ) .
the velocities of the four intrinsic uv absorption components are marked with tick marks above the spectra .
structure in the x - ray absorption profiles is revealed in the high s / n mean _ chandra _ spectrum . in figure 2 , at least two kinematic components are apparent in the profile , coinciding with uv components 1 and 2 + 3 ( see discussion in 3.5 of paper i ) .
we can not rule out the presence of absorption in component 4 as well .
we normalized all absorption features by dividing the averaged spectrum by our fits to the intrinsic emission ( see 2.1 ) .
the normalized profiles are plotted as a function of radial velocity in figure 3 .
the centroid velocities of the four intrinsic absorption components are marked with dashed vertical lines on each spectrum .
these were derived from the absorption for components 13 and ly@xmath4 for component 4 .
we have marked interstellar lines and known detector artifacts ( labeled with `` art '' ) with short tick marks in figure 3 .
qualitatively , the absorption in the merged stis data is similar to the initial spectrum presented in @xcite .
ly@xmath3 , , and are strong in components 13 , while is only detectable in component 1 .
the @xmath261551 line in component 3 is blended with @xmath261548 in component 1 .
the blue wing of @xmath261242 component 1 is blended with the red wing of @xmath261238 component 2 . in the _ fuse _ spectrum ,
the @xmath261032 line is heavily contaminated with galactic features , but it appears in all components in @xmath261038 . the deep feature in the red wing of component 1 in @xmath261038 is a detector artifact .
we detect absorption in the lyman lines up to ly@xmath0 in components 13 .
higher order lyman lines may be present , but suffer from severe blending with galactic absorption . additionally , strong galactic absorption contaminates component 1 in ly@xmath31 and component 2 in ly@xmath0 . in the top window of figure 3a , ly@xmath4 is seen to be deeper than ly@xmath3 in components 1 , 3 , and 4 , despite having an optical depth that is six times smaller .
additionally , the depth of ly@xmath0 in components 1 and 3 indicates ly@xmath4 is saturated in these components since @xmath32/@xmath33 11 , but the residual flux levels in the cores of the ly@xmath4 features are non - zero .
these are signatures of a complex absorption geometry in ngc 3783 , which is addressed further in 3.1 .
both @xmath26989 and @xmath34991 , which arises from the fine - structure level , are strong in component 1 .
the @xmath26989 component 1 profile shown in figure 3c exhibits a sharp edge in the blue wing and spiked core blueward of the centroid .
these features are not seen in the profiles of any other lines in this component , most notably the @xmath34991 line .
furthermore , there is a strong galactic h@xmath28 line at this wavelength contaminating the absorption .
thus , this spike is likely due to noise in the deep trough of the h@xmath28 + absorption .
weak absorption is also detected in components 2 and 3 in the @xmath34991 line .
@xmath26977 , also predicted to be strong in component 1 @xcite , is heavily blended with galactic h@xmath28 and can not be measured .
however , the @xmath35117576 multiplet , which arises from metastable levels 6.5 ev above ground , is detected in the stis spectrum , indicating high density in this component @xcite . to our knowledge , this is the first detection of @xmath35117576 absorption in a seyfert galaxy other than ngc 4151 .
weak @xmath26977 absorption also appears in components 2 and 3 .
there is a weak feature corresponding to the @xmath261118 line at the velocity of component 1 .
however , we can not confirm the presence of this component in the other line of the doublet , @xmath261128 , due to contamination with galactic absorption .
the @xmath17933,944 resonance doublet is contaminated with galactic interstellar features for all components and can not be measured .
lines of lower ionization , such as and , are not detectable in the spectrum .
the observed absorption depths are determined by a combination of the effective covering factor and the optical depths of the lines .
the absorption must be properly corrected for @xmath24 before an accurate column density can be measured . in general , the different covering factors of the line and continuum sources , @xmath36 and @xmath37 , must be taken into account . following the method of @xcite , we assume the same column of gas obscures the two emission sources , i.e. , @xmath38 . the observed residual flux in any absorption line is then given by @xmath39 where @xmath40 and @xmath41 are the intrinsic line and continuum fluxes , respectively .
dividing equation 1 by @xmath42 gives the expression for normalized flux , @xmath43 where @xmath44 and @xmath45 , which represent the relative contributions of line and continuum emission to the total intrinsic flux .
the lyman lines provide a unique study of the covering factor .
they give a set of lines that ( a ) span a large range in optical depth and ( b ) have substantially different relative contributions from continuum and line emission , @xmath46 and @xmath47 . since they arise from the same atomic energy level , their optical depths scale simply by their @xmath48 ratios , where @xmath49 is the oscillator strength and @xmath26 the laboratory wavelength of the lines . in figure
3a , the strong absorption in the cores of components 1 and 3 in ly@xmath0 , and component 2 in ly@xmath31 ( ly@xmath0 is blended with galactic in this component ) indicates both ly@xmath3 and ly@xmath4 are saturated in these components and , therefore , _ their core absorption depths are determined entirely by the covering factor_. since ly@xmath3 and ly@xmath4 are saturated , @xmath50 0 for these two lines and their normalized flux given by equation 2 reduces to a simple function of @xmath36 and @xmath37 .
we have combined these simplified equations for ly@xmath3 and ly@xmath4 and solved for @xmath36 and @xmath37 in the cores of the absorption , thereby separating the individual covering factors of the line and continuum sources . in these measurements , we averaged over 50 km s@xmath2 intervals centered on the centroid velocities of the components .
we note that this analysis assumes the @xmath36 of the emission ( which affects the ly@xmath4 absorption profile ) and the ly@xmath3 emission are identical . in table 3
, we list the normalized residual flux , @xmath51 , and the fractional contribution of continuum emission to the total intrinsic flux , @xmath46 , measured in the core of each available lyman line , along with our derived values for @xmath36 and @xmath37 .
the 1 @xmath52 uncertainties quoted in table 3 were computed from repeated measurements using different placements of intrinsic continuum and blr flux levels , and include the uncertainty in the residual flux due to noise .
these results show the discrepancies in the apparent absorption strengths seen in figure 3a are due to different covering factors of the line and continuum sources .
for example , in component 1 where the ly@xmath4 absorption is deeper than ly@xmath3 , @xmath53 0.57 and 0.15 for ly@xmath4 and ly@xmath3 , respectively , indicating @xmath54 in this component
. a closer look at the lyman profiles in figure 3a reveals the absorption in the higher order lines is resolved in the individual kinematic components , and is deep over several resolution elements in the _ fuse _ data .
thus , we extended the above analysis to derive the line and continuum covering factors as a function of radial velocity .
an exact solution to @xmath36 and @xmath37 is obtained at all radial velocities where ly@xmath4 is saturated .
since the optical depths of the lyman lines scale according to their @xmath48 ratios , this can be determined from the strengths of the higher order lyman lines .
in general , @xmath55 , where the @xmath21 in the relation applies if the covering factor is less than unity .
thus , the velocities where @xmath56@xmath57 equals zero within uncertainties gives a conservative estimate of the regions of saturated ly@xmath4 ( x represents @xmath0 for components 1 and 3 and @xmath31 for component 2 ) . for velocities where ly@xmath4 is not saturated , such as in component 4 and in the outer wings of the absorption in components 13 ,
the solution gives only lower limits to @xmath36 and @xmath37 . in figure 4 ,
the derived covering factor profiles are plotted .
velocities with only lower limits on the covering factors are plotted with a dotted line .
figure 4a shows that the continuum covering factor decreases smoothly in the blue wings of components 1 and 3 , at velocities where the covering factor is well determined ( solid line ) .
thus , the decreased absorption observed in these wings in ly@xmath4 ( figure 3a ) is a result of decreasing coverage of the continuum source , and the profile shape traces out the unocculted flux .
the other absorption wings may also exhibit decreased covering factor , however , only lower limits are available at these velocities . as discussed in @xcite , instrumental smearing can introduce an error into computations of the covering factor profiles , particularly in the wings . to test this effect
, we convolved the observed absorption profiles with the line spread function of _ fuse_. we find that the changes in the profiles are indetectable within the noise of the data , and conclude that instrumental smearing has a negligible effect on our results .
the variable covering factor detected in this analysis indicates there is substructure in the individual absorbers and may provide a constraint on physical models of the absorption gas .
this is addressed further in 4 .
one possible source of error associated with our calculations that is not included in the uncertainties quoted in table 3 is if the covering factor of the and ly@xmath3 blrs are not the same .
the intensive _ iue _ monitoring campaign of ngc 3783 showed the ly@xmath3 blr flux variations lagged the continuum variations by 3.6 days , while the blr flux lag was only 1.3 days @xcite , indicating the higher ionization line emission is located closer to the ionizing source .
if the blr is also more compact than ly@xmath3 , then @xmath36 for ly@xmath4 may be greater than ly@xmath3 , leading to an overestimate of the continuum covering factor .
this may explain the discrepancy in figure 4a , where the continuum covering factor derived at velocities just blueward of the centroid in component 1 and in the core of component 4 is greater than unity .
another possible source of error involves the contribution of light from an extended scattering region to the residuals in the absorption features .
if the scattered light profile underneath the absorption has structure , this will affect the covering factor calculations ( e.g. , see * ? ? ?
this has consequences in using limits on the size of the emission sources to constrain the size of the absorbers for cases of partial coverage .
however , it seems unlikely that scattered light can explain all of the non - unity covering factors derived from the absorption .
for example , to account for all of the residual flux in the cores of ly@xmath3 in components 1 and 3 would require that 60% and 50% , respectively , of the intrinsic continuum plus blr flux is scattered light .
we conclude that , at least in these cases , the low derived covering factors represent partial line - of - sight coverage of the emission - line source .
thus , large variations occur in @xmath36 between the cores of the different components , with values decreasing for higher blueshifted radial velocities in components 13 .
we note that unobscured nlr emission can not account for the low line covering factors , since the low @xmath36 values occur at high outflow velocities . at
most , the nlr only affects the red - wing of the lowest velocity component .
furthermore , it is difficult to see how the scattered light profile could explain the dramatic variation in the continuum covering factor derived in the blue wings of components 1 and 3 . in the outer velocities of the wings
, the scattered contribution would have to account for 6070% of the total intrinsic flux .
thus , although we can not rule out that the continuum source is fully occulted by the absorber in the cores of all the components , the decreasing covering factor derived in the _ wings _ of components 1 and 3 implies velocity - dependent , partial coverage of the continuum source . for doublet line absorption , the effective covering factor , which is a combination of the individual line and continuum covering factors , can be derived directly from the residual fluxes in the two lines using their intrinsic optical depth ratio @xcite . however , many of the doublets in ngc 3783 are too heavily contaminated with other absorption to apply this technique . in these cases and for lines not arising from doublets , another measure of @xmath24 is required that accounts for the individual line and continuum covering factors , as highlighted in our analysis of the lyman lines .
this is evident in component 1 , in which @xmath58 0.63 is measured in the core of using the doublet method , but and have only one measurable line .
figure 3 shows the , , and ly@xmath3 absorption depths are all relatively shallow compared to in this component .
our fits to the continuum and line emission reveal they also have a stronger relative contribution of blr emission at the wavelengths of component 1 absorption ( @xmath59 0.74 , 0.82 , and 0.85 for , , and ly@xmath3 , respectively , compared to @xmath59 0.58 for ) . in 3.1 , we found the ly@xmath3 absorption is saturated but it appears weak because of a low covering factor of the blr .
similarly , the shallow absorption observed in and may also be the result of relatively low effective covering factors rather than low column densities .
this was suggested by the photoionization modeling results of @xcite , which showed that substantial columns of and/or should always accompany a large column .
motivated by these results , we used the covering factors for to derive effective covering factors and column densities for all intrinsic absorption lines . following @xcite ,
the effective covering factor is defined in terms of the individual line and continuum covering factors , @xmath60 for each line , we computed @xmath24 as a function of radial velocity from the @xmath37 and @xmath36 profiles derived for and our fits to the intrinsic continuum and line fluxes .
the effective covering factors , averaged over 50 km s@xmath2 intervals in the cores of the absorption components , are listed in table 4 .
uncertainties were derived by propagating the errors associated with @xmath37 and @xmath36 . for the doublet absorption in which the cores of both lines are uncontaminated with other absorption ,
i.e. , component 1 and components 13 , we also computed the effective covering factors in the cores of the absorption directly using the doublet method as described in @xcite .
these values , and associated uncertainties , are listed in parentheses in table 4 .
these results show the effective covering factor derived for in component 1 is equivalent to that determined for , but is larger than the covering factor by more than a factor of two .
_ this provides the only evidence for a covering factor within an individual kinematic component that is a function of ionization .
_ we note this may indeed be the case since component 1 likely consists of at least two distinct subcomponents , one of which is lower ionization and responsible for all of the absorption ( see * ? ? ?
comparison of the line widths in this component reveals further evidence of substructure . in table 5
, we give the fwhm and centroid velocities measured directly from the normalized absorption profiles for each line in component 1 .
the and @xmath15 lines are significantly narrower than the higher ionization lines , and .
substituting equation 3 into equation 2 gives the familiar expression for normalized flux as a function of the effective covering factor @xcite , @xmath61 the factor @xmath62 in equation 4 is the normalized unocculted flux .
this represents the fraction of the total intrinsic flux that is unobscured by the absorber .
the normalized absorption profiles for , , , and are plotted together with the normalized unocculted flux levels derived from equation 4 ( dashed lines ) for component 1 in figure 5a and components 2 + 3 in figure 5b . in these figures
we have plotted only the least contaminated line for the doublets .
figure 5 illustrates how the apparent absorption is a convolution of covering factor and column density . for lines with large column densities , the normalized flux will approach the unocculted flux level , but
never lie below it ( see equation 4 ) .
other than the detector artifact in @xmath261038 component 1 ( figure 5a ) , we find no cases where residual fluxes lie below the unocculted flux levels outside the 1 @xmath52 uncertainties .
this supports the validity of using the covering factors . in figure 5 ,
the absorption depths in several lines closely match the derived unocculted flux levels , indicating possible saturation in these lines . for example
, the residual flux in the @xmath261038 feature in components 2 and 3 traces the unocculted flux levels over nearly the entire profile .
the @xmath261032 absorption gives a consistent result in the narrow windows of component 2 that are uncontaminated with galactic absorption ( @xmath63650 and @xmath1450 km s@xmath2 in figure 3b ) .
additionally , the absorption in the blue wings of @xmath261238 , @xmath261548 , and @xmath261038 in component 1 appears saturated .
however , the absorption is seen to diverge from the unocculted flux in the red wing of and ( is contaminated with a detector artifact at these velocities ) .
thus , if the covering factor profiles derived for are valid for these lines , the column densities are not uniform across the profile in component 1 , providing further evidence for substructure in this component . given the uncertainties in the covering factor profiles due to the uncertain scattered light profile and the possibility that the covering factors of the and ly@xmath3 blrs are not the same , as described in 3.1 , we can not rule out that this apparent substructure is due to our assumptions about the covering factor profiles .
equation 4 can be rearranged to give the optical depth @xcite , @xmath64 where the subscript indicates the calculation is made at each radial velocity , thereby giving the optical depth profile . the column density at each radial velocity
is then obtained from the optical depth , @xmath65 @xcite .
we measured the total column density for each line by integrating equation 6 over the optical depth profiles calculated from equation 5 ; results are given in table 6 .
for the doublets where a direct measurement of the effective covering factor is available from the residuals in the two lines , we used the core value listed in parenthesis in table 4 to compute the column density .
components 2 and 3 were deblended following the method outlined in @xcite . in saturated regions , we measured lower limits for the column densities using @xmath66 in equation 5 , where @xmath67 is the 1 @xmath52 uncertainty in the normalized flux .
upper limits on undetectable lines were determined from uncertainties in the normalized fluxes .
the 1 @xmath52 uncertainties quoted in table 6 were estimated from repeated measurements , and include uncertainties in our fits to the intrinsic emission as well as the photon noise .
we note that the lower limits on the columns from the combined components are consistent with the upper limit of 10@xmath14 @xmath7 derived from the x - ray spectrum in paper i. if the absorption in component 1 is real ( see discussion in 2 ) , then it has a column density of 1.5 @xmath19 0.7 @xmath13 10@xmath68 @xmath7 .
given the important implications of our results involving the covering factors , particularly the lower covering factor derived for , it is worthwhile to address the possible systematic errors that were discussed in 3.1 in more depth .
first , we note that any contribution of scattered light to the residual fluxes in the ly@xmath4 or ly@xmath3 absorption will lead to an _ underestimate _ of the covering factors .
this would lead to an even larger discrepancy between the effective covering factor derived for from the doublet and the prediction from the results .
the second possible systematic error is introduced if the sizes of the and ly@xmath3 blrs are not identical , leading to different emission line covering factors for ly@xmath4 and ly@xmath3 .
since the ly@xmath0 absorption has no underlying blr emission ( see table 3 ) , it can be used to place limits on the covering factors that are unaffected by any discrepancy in the sizes of the blrs of the different lines .
for example , in component 1 , an absolute lower limit on the covering factor in the core follows directly from the depth of the feature , @xmath69 [email protected] .
this is stricly a lower limit , since ( a ) ly@xmath0 is not saturated and ( b ) there may be a contribution from scattered flux . using this result in equation 3
, we derive an absolute lower limit on the effective covering factor predicted for from the covering factors , @xmath71 0.58 . comparing this with the value derived from the doublet , @xmath58 0.30 @xmath190.07
, we conclude that our finding that the covering factor of is less than is robust .
while detailed photoionization modeling of the absorption will be presented in future papers ( kraemer et al . in preparation
; netzer et al . in preparation ) , we can place some immediate constraints on the physical conditions and geometry of the low ionization uv absorber in component 1 .
we refer to the photoionization modeling results in @xcite , which predicted a total hydrogen column density , @xmath72 5 @xmath13 10@xmath73 @xmath7 , and ionization parameter , @xmath74 0.0018 , for this subcomponent .
since the column measured in the averaged spectrum is somewhat higher than in the @xcite study , partly due to the lower covering factor found for this line in the present study , and we have a more rigid upper limit on the column because of the high s / n , we have modified the models slightly . using the ionizing spectral energy distribution and abundances employed in the models of @xcite
, we find that @xmath5 10@xmath6 @xmath7 and @xmath8 0.005 are required to match the column and upper limit simultaneously .
the detection of @xmath34117576 absorption , which arises from metastable levels 6.5 ev above the ground state , implies a high density in the low ionization absorber in component 1 @xcite .
the relative population of the metastable levels is determined by the physical conditions in the absorber , and the ratio of metastable to ground state columns can be used to estimate the electron density , @xmath75 @xcite . due to contamination of the @xmath26977 feature by galactic h@xmath28 , the ground state column density can not be measured directly in ngc 3783 .
instead , we use the photoionization model results , which predict @xmath76 2 @xmath13 10@xmath77 @xmath7 .
the predicted column is not a strong function of the exact @xmath78 , @xmath79 solution needed to match the , since and appear in the same zone due to the similarity in their ionization potentials .
our measured column for the metastable level from @xmath34117576 ( 1.2 @xmath13 10@xmath68 @xmath7 ) then implies the ratio of metastable to ground state columns is @xmath160.006 .
the photoionization models predict @xmath80 20,000 k in the absorber , which is largely insensitive to the exact @xmath79 , @xmath78 solution for our assumed sed . using the calculations in @xcite
, this corresponds to @xmath9 10@xmath10 @xmath11 for the low ionization subcomponent .
this high density implies a very small radial dimension for the absorber , @xmath1610@xmath81 cm for a column of 10@xmath6 @xmath7 , which may indicate the gas is clumpy and has a low volume filling factor ( see discussion below ) . from the density , lower limit on @xmath78 , and estimate of the ionizing continuum luminosity given by @xcite , we derive an upper limit on the location of the absorber with respect to the ionizing continuum source , @xmath12 8 @xmath13 10@xmath14 cm . the partial coverage of the emission sources revealed in the analysis of the lines in
3.1 also places constraints on the absorption geometry . at velocities exhibiting partial coverage ,
the area of the absorbing gas projected on the sky is smaller than the projected area of the emission sources .
the monitoring study of ngc 3783 revealed large amplitude variations of both the continuum and blr emission @xcite .
the most rigid constraint comes from the measured lag of the blr flux variations behind the continuum variations .
for example , @xcite found the ly@xmath3 blr flux varied by up to a factor of 1.6 and lagged the continuum variations by 3.6 days . from light travel time arguments , this constrains the projected size of much of the blr to be @xmath163.6 lt - days ( 10@xmath82 cm ) across .
thus , in components 1 and 3 , with @xmath83 0.31 and 0.55 measured in the cores , the projected area of the absorbers is @xmath16(5.5 @xmath13 10@xmath77)@xmath84 and @xmath16(7.5 @xmath13 10@xmath82)@xmath84 cm@xmath84 , respectively .
these results provide important clues to the physical nature of the absorbers .
the smoothly varying covering factor of the continuum detected in the wings of the absorption in 3.1 indicates there is substructure within the individual kinematic components .
one interpretation of this result is that the absorption components are comprised of many small substructures , such as cloudlets or turbulent cells . in this scenario ,
the decreasing coverage in the wings may be a consequence of the velocity dispersion of these substructures , with a decreasing number of clouds " having velocities that diverge from the centroid velocity .
since , in general , the ionic column densities do not decrease dramatically in the wings of the absorption profiles ( see figure 5 ) , there can not be much overlap of clouds in the cores of the absorbers , indicating they have small volume filling factors .
the extreme compactness of the gas comprising the low ionization subcomponent in component 1 provides additional evidence for this model .
since the radial thickness of this gas is much smaller than the projected dimensions of the emission sources , the most natural explanation is that a large number of very small , dense clouds combine to give the observed absorption .
the alternative interpretation would require that the absorber is an exceedingly thin sheet of material .
previous studies have invoked absorption by a clumpy gas to explain the complex structure seen in other agn absorption systems ( e.g. , * ? ? ?
* ; * ? ? ?
* ) and a highly inhomogeneous absorbing medium is predicted by the multitemperature wind models of @xcite .
ngc 3783 was observed with _ hst_/stis at 18 epochs and with _ fuse _ at 5 epochs as part of an intensive multiwavelength monitoring campaign designed to investigate the intrinsic absorption in this seyfert 1 galaxy .
we have combined the observations to produce a high s / n averaged spectrum in the uv and far - uv .
the major findings of our study follow .
, , , , and the lyman lines up to ly@xmath0 appear in the three kinematic components identified by @xcite in the initial stis spectrum ( components 1 , 2 , and 3 at radial velocities @xmath11320 , @xmath1548 , and @xmath1724 km s@xmath2 , respectively ) .
galactic contamination prevents the detection of higher order lyman lines .
appears in component 1 ; @xmath26977 is heavily contaminated with galactic absorption and can not be measured in this component , but is present in components 2 and 3 .
we detect absorption from the @xmath341175 - 76 multiplet in component 1 , indicating a high density in this absorber .
this is the first detection of metastable in a seyfert galaxy other than ngc 4151 @xcite .
no lower ionization lines appear in absorption , and we place stringent upper limits on their column densities .
a fourth kinematic component of intrinsic absorption is tentatively identified at a radial velocity of @xmath11027 km s@xmath2 .
this component appears strong only in ly@xmath4 , which may be contaminated with galactic absorption , and . in all other lines ,
the detection of this component is very sensitive to the fit to the intrinsic emission .
the lyman lines reveal a complex absorption geometry in ngc 3783 and highlight the importance of determining the individual covering factors of the continuum and emission - line sources .
the strength of the higher order lyman lines indicates ly@xmath3 and ly@xmath4 are saturated in components 13 .
the continuum and emission - line covering factors were separated using the ly@xmath3 and ly@xmath4 absorption in each kinematic component , both in the cores of absorption and as a function of radial velocity .
the covering factor of the blr varies significantly between the cores of the absorption components , with @xmath83 0.33 , 0.84 , 0.55 , and 0.03 derived in components 1 , 2 , 3 , and 4 , respectively .
the large residual fluxes measured in ly@xmath3 in components 1 and 3 imply that , at least in these components , the non - unity emission - line covering factors can not be fully accounted for by light from an extended scatterer , and thus represent partial coverage of the blr .
we also find evidence for variation of the continuum covering factor with velocity in the individual kinematic components .
specifically , @xmath37 decreases by @xmath1660% over several resolution elements in the wings of the absorption .
this smoothly decreasing coverage is a signature of substructure within the absorbers .
the individual continuum and line covering factors derived from were used to derive effective covering factors for all lines .
comparison to the covering factors derived directly by the doublet method reveals the covering factor is smaller than that of and by more than a factor of two in component 1 .
this implies substructure in component 1 and is consistent with the prediction by @xcite that this component is comprised of at least two zones of uv absorption .
furthermore , the fwhm of and are narrower than the higher ionization lines in component 1 , providing additional evidence for substructure . employing the covering factor profiles for the higher ionization lines , we find the relatively weak apparent and absorption compared to in component 1 is due to low effective covering factors rather than small column densities .
if the covering factors derived from are valid for and , then the column densities of these lines are not uniform across the profile of component 1 ; they are saturated in the blue wing , but smaller in the red wing .
additionally , is found to be saturated in components 2 and 3 , as well as in the blue wing of component 1 .
we place lower limits on the total hydrogen column density and ionization parameter in the low ionization subcomponent in component 1 , @xmath5 10@xmath6 @xmath7 and @xmath8 0.005 .
combining model predictions of the total column with our measured @xmath15 column gives @xmath9 10@xmath86 @xmath11 .
these results imply an upper limit on the distance of the absorber from the ionizing continuum source of @xmath87 8 @xmath13 10@xmath14 cm . using upper limits on the size of the blr from the variability study by @xcite , the projected area of the absorbers at velocities exhibiting partial coverage of the emission - line source
is found to be @xmath16(10@xmath82)@xmath84 cm@xmath84 .
the decreasing covering factor exhibited in the wings of the lyman lines and extreme compactness of the @xmath15 absorber are suggestive of a clumpy absorbing medium with a low filling factor .
j. r. g. , d. m. c. , and s. b. k. acknowledge support from nasa grant hst - go-08606.13-a .
w.n.b acknowledges cxc grant go1 - 2103 and nasa ltsa grant nag5 - 8107 .
f.h . acknowledges nsf grant ast 99 - 84040 .
we thank b - g andersson and the _ fuse _ team for assistance with the _ fuse _ spectra and derck massa for providing his code for modeling the h@xmath28 absorption .
we also thank the anonymous referee for very helpful comments .
arav , n. , korista , k. t. , & de kool , m. 2002 , , 566 , 699 barlow , t. a. , & sargent , w. l. w. 1997 , , 113 , 136 barlow , t. a. , hamann , f. , & sargent , w. l. w. 1997 , in asp conf .
128 , mass ejection from active galactic nuclei , ed .
n. arav , i. shlosmann , & r. j. weymann ( san francisco : asp ) , 13 blair , w. 2000 , the _ fuse _ observers guide , version 2.1 bromage , g. e. , et al .
1985 , , 215 , 1 cohen , m. h. , ogle , p. m. , tran , h. d. , vermeulen , r. c. , miller , j. s. , goodrich , r. w. & martel , a. r. 1995 , apj , 448 , l77 crenshaw , d. m. , maran , s. p. , & mushotzky , r. f. 1998 , , 496 , 797 crenshaw , d. m. , kraemer , s. b. , boggess , a. , maran , s. p. , mushotzky , r. f. , & wu , c .- c . 1999 , , 516 , 750 crenshaw , d. m. , & kraemer , s. b. 1999 , , 521 , 572 crenshaw , d. m. , kraemer , s. b. , hutchings , j. b. , danks , a. c. , gull , t. r. kaiser , m. e. , nelson , c. h. , & weistrop , d. 2000 , , 545 , l27 de vaucouleurs , g. , de vaucouleurs , a. , corwin , h. g. , buta , r. j. , paturel , g. , & fouque , p. 1991 third reference catalogue of bright galaxies ( springer - verlag : new york ) ferland , g. j. , korista , k. t. , verner , d. a. , ferguson , j. w. , kingdon , j. b. , & verner , e. m. 1998 , , 110 , 761 gabel , j. r. , crenshaw , d. m. , & kraemer , s. b. 2002 , in asp conf .
255 , mass outflow in active galactic nuclei : new perspectives , ed .
d. m. crenshaw , s. b. kraemer , & i. m. george ( san francisco : asp ) , 81 ganguly , r. , eracleous , m. , charlton , j. c. , & churchill , c. w. 1999 , , 117 , 2594 george , i. m. , turner , t. j. , netzer , h. , nandra , k. , mushotzky , r. f. , & yaqoob , t. 1998 , , 114 , 73 goodrich , r. w. , & miller , j. s. 1995 , , 448 , l73 hamann , f. , barlow , t. a. , junkkarinen , v. , & burbidge , e. m. 1997 , , 478 , 80 hamann , f. w. , barlow , t. a. , chaffee , f. c. , foltz , c. b. , & weymann , r. j. 2001 , , 550 , 142 kaiser , m. e. , & kriss , g. a. 2002 , in asp conf . ser .
255 , mass outflow in active galactic nuclei : new perspectives , ed .
d. m. crenshaw , s. b. kraemer , & i. m. george ( san francisco : asp ) , 75 kaspi , s. , brandt , w. n. , netzer , h. , sambruna , r. , chartas , g. , garmire , g. p. , & nousek , j. a. 2000 , , 535 , l17 kaspi , s. , et al .
2001 , , 554 , 216 kaspi , s. , et al .
2002 , , 574 , 643 ( paper i ) kraemer , s. b. , crenshaw , d. m. , gabel j. r. 2001a , , 557 , 30 kraemer , s. b. , et al .
2001b , , 551 , 671 kriss , g. a. , et al .
1992 , , 392 , 485 kriss , g. a. 2002 , in asp conf .
255 , mass outflow in active galactic nuclei : new perspectives , ed .
d. m. crenshaw , s. b. kraemer , & i. m. george ( san francisco : asp ) , 69 krolik , j. h. , & kriss , g. a. 1997 , in asp conf . ser .
128 , mass ejection from active galactic nuclei , ed .
n. arav , i. shlosmann , & r. j. weymann ( san francisco : asp ) , 132 krolik , j. h. , & kriss , g. a. 2001 , , 561 , 684 lindler , d. 1999 calstis reference guide ( calstis version 5.1 ) lu , l. , savage , b. d. , & sembach , k. r. 1994 , , 426 , 563 maran , s. p. , et al .
1996 , , 465 , 733 oegerle , w. , murphy , e. , & kriss , g. a. 2000 , the _ fuse _ data handbook , version 1.1 onken , c. a. , & peterson , b. m. 2002 , , 572 , 746 reichert , g. a. , et al .
1994 , , 425 , 582 savage , b. d. , & sembach , k. r. 1991 , , 379 , 245 sembach , k. r. , howk , j. c. , savage , b. d. , & shull , j. m. 2001 , , 121 , 992 shields , j. c. , & hamann , f. 1997 , , 481 , 752 wampler , e. j. , bergeron , j. , & petitjean , p. 1993 , a&a , 273 , 15 lrrr o57b01020 & 2000 - 02 - 27 , 09:34:59 & 2700 + o57b01030 & 2000 - 02 - 27 , 11:11:33 & 2700 + o63m01010 & 2000 - 08 - 05 , 00:54:58 & 2200 + o63m01020 & 2000 - 08 - 05 , 02:18:41 & 2700 + o63m02010 & 2000 - 11 - 26 , 01:10:17 & 2200 + o63m02020 & 2000 - 11 - 26 , 02:35:18 & 2700 + o63m03010 & 2001 - 01 - 25 , 20:09:07 & 2200 + o63m03020 & 2001 - 01 - 25 , 21:34:05 & 2700 + o63m04010 & 2001 - 02 - 25 , 04:01:51 & 2200 + o63m04020 & 2001 - 02 - 25 , 05:22:57 & 2700 + o63m05010 & 2001 - 02 - 28 , 01:05:49 & 2200 + o63m05020 & 2001 - 02 - 28 , 02:28:43 & 2700 + o63m06010 & 2001 - 03 - 02 , 17:20:14 & 2200 + o63m06020 & 2001 - 03 - 02 , 18:43:59 & 2700 + o63m08010 & 2001 - 03 - 11 ,
14:54:01 & 2200 + o63m08020 & 2001 - 03 - 11 , 16:20:48 & 2700 + o63m09010 & 2001 - 03 - 15 , 16:51:54 & 2200 + o63m09020 & 2001 - 03 - 15 , 18:19:14 & 2700 + o63m10010 & 2001 - 03 - 19 , 17:12:53 & 2200 + o63m10020 & 2001 - 03 - 19 , 18:40:08 & 2700 + o63m11010 & 2001 - 03 - 23 , 15:59:27 & 2200 + o63m11020 & 2001 - 03 - 23 , 17:24:31 & 2700 + o63m12010 & 2001 - 03 - 27 , 08:19:43 & 2200 + o63m12020 & 2001 - 03 - 27 , 09:43:10 & 2700 + o63m14010 & 2001 - 04 - 04 , 02:35:42 & 2200 + o63m14020 & 2001 - 04 - 04 , 03:57:29 & 2700 + o63m15010 & 2001 - 04 - 08 , 01:22:01 & 2200 + o63m15020 & 2001 - 04 - 08 , 02:41:30 & 2700 + o63m16010 & 2001 - 04 - 11 , 22:33:08 & 2200 + o63m16020 & 2001 - 04 - 11 , 23:49:35 & 2700 + o63m07010 & 2001 - 04 - 16 , 02:09:01 & 2242 + o63m07020 & 2001 - 04 - 16 , 03:28:27 & 2700 + o63m53010 & 2001 - 04 - 23 , 15:37:54 & 2242 + o63m53020 & 2001 - 04 - 23 , 17:02:53 & 2700 + o63m17010 & 2002 - 01 - 06 , 13:39:42 & 2242 + o63m17020 & 2002 - 01 - 06 , 15:15:51 & 2700 + lrrr b1070102000 & 2001 - 02 - 28 , 16:46 & 27830 + b1070106000 & 2001 - 03 - 07 , 08:31 & 26400 + b1070103000 & 2001 - 03 - 11 , 12:56 & 27494 + b1070105000 & 2001 - 03 - 30 , 22:52 & 28099 + b1070105000 & 2001 - 06 - 27 , 05:26 & 27490 + lrrrrrrrrrrr 1 & 0.57 & 0.27 & & 0.23 & & 0.15 & 0.57 & 0.92 & 1.0 & 1.0 ( @xmath190.10 ) & 0.31 ( @xmath190.06 ) + 2 & 0.18 & 0.20 & 0.38 & 0.50 & & 0.10 & 0.52 & 0.89 & 1.0 & 0.74 ( @xmath190.08 ) & 0.84 ( @xmath190.08 ) + 3 & 0.41 & 0.26 & 0.24 & 0.46 & & 0.11 & 0.52 & 0.90 & 1.0 & 0.9 ( @xmath190.10 ) & 0.55 ( @xmath190.08 ) + lllll @xmath261038 & [email protected] & [email protected] & [email protected] & 0.29 + @xmath261238 & [email protected] ( [email protected] ) & [email protected] ( [email protected] ) & [email protected] ( [email protected] ) & 0.41 + * @xmath26991 & [email protected] & [email protected] & [email protected] & + @xmath261548 & [email protected] & [email protected] & [email protected] & 0.19 + @xmath26977 & & [email protected] & [email protected] & 0.84 + * @xmath261775 - 76 & [email protected] & & & + @xmath261393 & [email protected] ( [email protected] ) & & & + lrr @xmath261238 & @xmath11304 ( @xmath1910 ) & 194 ( @xmath197 ) + @xmath261548 & @xmath11322 ( @xmath1910 ) & 180 ( @xmath197 ) + @xmath261393 & @xmath11326 ( @xmath1915 ) & 111 ( @xmath1910 ) + * @xmath26991 & @xmath11331 ( @xmath1920 ) & 124 ( @xmath1915 ) + ly@xmath4 & @xmath11297 ( @xmath1912 ) & 219 ( @xmath1910 ) + ly@xmath0 & @xmath11315 ( @xmath1920 ) & 145 ( @xmath1915 ) + lllll & 140 ( @xmath1920 ) & 37 ( @xmath199 ) & 163 ( @xmath1930 ) & 5.0 ( @xmath192.0 ) + & @xmath2114 & @xmath2112 & @xmath2151 & 6.1 ( @xmath194.0 ) + & @xmath2111 & 2.6 ( @xmath190.5 ) & 11.4 ( @xmath192.0 ) & 0.50 ( @xmath190.4 ) + & 4.4 ( @xmath192.5 ) & & & @xmath881.0 + * & 4.5 ( @xmath191.5 ) & 0.4 ( @xmath190.2 ) & 1.5 ( @xmath190.8 ) & @xmath881.0 + & @xmath214.5 & 0.60 ( @xmath190.13 ) & 2.6 ( @xmath190.5 ) & 0.34 ( @xmath190.25 ) + & & 0.30 ( @xmath190.2 ) & 0.90 ( @xmath190.35 ) & + * & 0.12 ( @xmath190.06 ) & @xmath880.05 & @xmath880.03 & @xmath880.03 + & @xmath880.6 & @xmath880.25 & @xmath880.10 & @xmath880.10 + & 0.89 ( @xmath190.20 ) & @xmath880.1 & @xmath880.03 & @xmath880.03 + & @xmath880.06 & @xmath880.04 & @xmath880.02 & @xmath880.01 + | we present observations of the intrinsic absorption in the seyfert 1 galaxy ngc 3783 obtained with the space telescope imaging spectrograph ( stis ) on the _ hubble space telescope _ ( _ hst _ ) and the _ far ultraviolet spectroscopic explorer _
( _ fuse _ ) .
we have combined 18 stis and 5 _ fuse _ observations to obtain a high signal - to - noise averaged spectrum spanning 9051730 .
the averaged spectrum reveals absorption in , , , , and the lyman lines up to ly@xmath0 in the three blueshifted kinematic components previously detected in the stis spectrum ( at radial velocities of @xmath11320 , @xmath1724 , and @xmath1548 km s@xmath2 ) .
the highest velocity component exhibits absorption in .
we also detect metastable in this component , indicating a high density in this absorber .
no lower ionization lines , i.e. , and , are detected .
a weak , fourth absorption component is tentatively detected in the high ionization lines and ly@xmath3 and ly@xmath4 at a radial velocity of @xmath11027 km s@xmath2 .
the lyman lines reveal a complex absorption geometry .
the strength of the higher order lines indicates ly@xmath3 and ly@xmath4 are saturated over much of the resolved profiles in the three strongest absorption components and , therefore , their observed profiles are determined by the covering factor .
we separate the individual covering factors of the continuum and emission - line sources as a function of velocity in each kinematic component using the ly@xmath3 and ly@xmath4 lines .
the covering factor of the blr is found to vary dramatically between the cores of the individual kinematic components , ranging from 0 to 0.84 . additionally , we find that the continuum covering factor varies with velocity within the individual kinematic components , decreasing smoothly in the wings of the absorption by at least 60% .
comparison of the effective covering factors derived from the results with those determined directly from the doublets reveals the covering factor of is less than half that of and in the high velocity component .
additionally , the fwhm of and are narrower than the higher ionization lines in this component .
these results indicate there is substructure within this absorber .
we also find evidence for structure in the column density profiles of the high ionization lines in this component .
we derive a lower limit on the total column ( @xmath5 10@xmath6 @xmath7 ) and ionization parameter ( @xmath8 0.005 ) in the low ionization subcomponent of this absorber .
the metastable - to - total column density ratio implies @xmath9 10@xmath10 @xmath11 and an upper limit on the distance of the absorber from the ionizing continuum of @xmath12 8 @xmath13 10@xmath14 cm .
the decreasing covering factor found in the wings of the absorption and the extreme compactness of the c iii@xmath15 absorber are suggestive of a clumpy absorption gas with low volume filling factor . |
justifying its use in what follows , one of the great advantages of the woodhouse axiomatics @xcite resides precisely in the absence of the chronometric hypothesis ( ch ) of general relativity as well as in references to somehow unspecified standard or absolute " clocks .
let us recall that the chronometric hypothesis is defined by the equality @xmath0 , where @xmath1 is the proper time of the clock ( also known as the ` _ _ atomic time _ _ ' ) , @xmath2 is the time interval between two events and is defined from a given metric field @xmath3 , and @xmath4 is the so - called ` _ _ gravitational time _ _ ' ( up to the speed of light @xmath5 ) .
@xcite strongly criticized this hypothesis , invoking at least two reasons : 1 ) it can be replaced by the conformal or projective structures without a metric field , and 2 ) the shifts between atomic and gravitational times can be deduced from the kundt - hoffmann protocol @xcite and , as a result , allow the selection of those clocks satisfying the equality defined by the ch .
these axiomatics are one of those appropriate frameworks under which the latter shifts , which are scale factors differing from @xmath5 , enable us to track physical justifications to the ch .
the main tool we use throughout is the geometrical situation encountered in the so - called ` _ _ clock ( pseudo-)paradox _ _ , ' the solutions of which explicitly require the ch .
additionally , this is another great advantage of the woodhouse axiomatics for approaching temporal paradoxes , since it does not refer to any proper time notions . for more on the geometric settings
, we recall that in these axiomatics , the so - called ` _ _ message functions _ _ '
@xmath6 and ` _ _ clock functions _ _ ' @xmath7 ( i.e. the time parameterizations defined by woodhouse ) are only defined for ` _ _ particles _ _ ' @xmath8 ascribed to worldlines of freely falling massive punctual objects , and the metric fields @xmath3 depend only on particles @xmath8 up to conformal factors defined for clock functions @xmath7 .
in addition , upstream of the metric structure given by @xmath3 , there is a unique affine structure given by the totality of the freely falling particles of the woodhouse axiomatics or equivalently , from ehlers _ et al .
_ , a weyl structure , i.e. , in particular , a class of conformally equivalent riemann structures .
weyl structures do not fix rates of clocks a priori independently of their histories .
this is the so - called second clock effect .
the opposite case was dismissed by audretsch @xcite , who concluded that there were several efforts after 1970 to assign a weyl geometry to spacetime rather than a riemannian geometry . in reality , this is close to our viewpoint , i.e. , that rates of clocks , invoked to approach the ch via the clock paradox , are not observable , but fixed only respective to very particular given causal protocols exhibiting the physical and historical relationships among particles themselves , each endowed with particular different sub - structures of a given common conformal one . obviously , whenever the ch is posed and the metric connection is chosen , then the clocks with proper time aging ( or those that are displaying " a proper time ) are automatically inferred from the so - called ` geodesic hypotheses ' as well as singled out , and so , there are clearly no clock paradoxes .
the differential aging is quite simply given from the proper time differences obtained from the integration of the infinitesimal proper times along each worldline .
however , the ch actually raises the point that we must feature a proper time notion in the physical viewpoint .
that is , we must find a proper time definition _ regardless of _ the choice of time parameterizations of the worldlines carried out for each particle of the woodhouse axiomatics . on the contrary ,
the differential aging would only translate or reflect the agreement between the proper time definition , and , in particular , the ch and the selected metric connections .
that is to say , this agreement expresses the metric structure choice from the unique affine structure of a weyl spacetime . additionally , the woodhouse results are situated upstream of the projective , conformal and weyl structures defined in particular in the ehlers _ et al
. _ axiomatics .
terms such as ` _ _ freely falling particles _ _ ' are simply generic , all the more so as their worldlines can not be defined from geodesic equations since metric connections are never utilized in the woodhouse formalism .
the only important restriction in the definition of the particles is mainly , in our opinion , that it must satisfy the following fundamental property : that for each particle @xmath8 passing through each event @xmath9 in spacetime , there exists a neighborhood of @xmath9 into which a four - dimensional ` _
_ @xmath10-congruence _ _ ' can be defined and which can also be an extension of @xmath8 . to some extent , this condition is a topological translation of the possible existence in this neighborhood of a projective connection .
it can finally be substituted for axiom @xmath11 of ehlers _ et al .
_ , specifying the ( projective ) equations assigned to freely falling particles and , in some ways , incidentally and unfortunately , making a contradictory use of proper times .
this indicates another advantage of the woodhouse axiomatics . as a fundamental result ,
the true freely falling particles are those belonging to a .
particles subjected to forces do not belong to it , but even in this case , the message and clock functions are still usable in the woodhouse formalism .
returning to the tool used in the following , a resolution of the clock paradox is possible if the accelerations are involved in the differential aging formulas , since they are the sole data distinguishing the two clocks .
this accounts for the well - known historical attempts made by einstein @xcite to solve it in general relativity .
he used , in his own terms , a ` _ _ pseudo - gravitational _ _ ' field accounting for the accelerating force experienced by the moving clock , which is at the origin of the clock asymmetry .
unfortunately , we run into an important problem with at least two distinct physical situations : the first is met in the case of rectilinear motions and the second is related to circular motions . in the first case , we refer , for example , to the unnikrishnan @xcite results , recalling and clearly showing some possible inconsistencies in the solutions based upon general relativity due to the use of pseudo - gravitational fields .
however , in the second situation for circular motions and among the most precise recent experimental results , differential aging clearly does not depend in any way on a pseudo - gravitational field such as , for instance , the pseudo - field associated with the centrifugal force . in this last case ,
the conclusion is clear : special relativity is more than sufficient to fully account for the experimental results .
therefore , it is necessary to raise the question of why , when motions are rectilinear , we must use general relativity , if that is indeed what must be done , and why , in the case of circular motions , general relativity no longer intervenes .
nevertheless , a certain type of mechanical work might also be used for interpretations and solutions , as we shall see .
it is a fact that in the first , linear case , some forces work , whereas in the case of circular motions , centripetal forces do not .
in addition , at the origin of the conformal proper times , considering an accelerating frame under a kind of primary initial geometry or special relativity , we know , for instance , that there are precession effects influencing space vectors , such as the spin vectors leading to the so - called thomas precession .
the latter has a long - standing history in relation to the clock paradox . at the basis of the precession equations ,
there is , as a particular case , the so - called _ mller term _ @xmath12 , where @xmath13 is a velocity vector , or more generally , if @xmath14 is a velocity field , @xmath15 .
then , if @xmath13 is the restriction of @xmath16 to a worldline and is tangent to the latter , this mller term can exactly account for another metric connection restricted to this worldline . as a result , the precessing vector is the tangent vector of a geodesic .
furthermore , if @xmath16 precesses itself with respect to another precession equation , i.e. , if @xmath16 is _ congruent _ with another vector field , then its dual with respect to the metric defines a _ weyl proper time _
* remark ( d ) p.67 , and p.81 ) .
moreover , the vanishing behavior of this mller term meaning the primary connection @xmath17 is a projective connection with its corresponding geodesics defines the projective geodesics as being those of the primary initial geometry . in other words ,
a non - vanishing mller term points to a change in a non - equivalent projective geometry , i.e. , a change of physical frame of reference , but it is at least compatible with the conformal structure .
consequently , again and in particular , the flat minkowski spacetime should no longer be valid for non - geodesic motions .
we could say that the projective connections and , thus , the weyl structures depend on the non - geodesic motions to which they apply .
this is a way of justifying the use of varying weyl structures ( or projective structures ) attached to each particle .
then , each worldline or particle should correspond to the restriction of a specific weyl structure only , and all these non - projectively equivalent weyl structures should be conformally related , leading to ( relative ) conformal proper times . in the next section , we recall the different definitions given in the woodhouse axiomatics and the malament theorem .
section [ coeur ] is divided into four sub - sections , of which the second is devoted to a difficulty that occurs systematically and is inherited from the simultaneity maps resulting in a non - trivial differential aging . in the third sub - section ,
geometrical arguments that are necessary to define formulas for differential aging and are consistent with the experimental results are presented using conformal proper times . in the fourth sub - section ,
the conformal factors in the previous formulas are physically interpreted , and then in section [ expert ] , these formulas are applied to certain experimental results .
in the woodhouse axiomatics , the spacetime @xmath18 is a set of points , to which a set @xmath19 of subsets of @xmath18 called ` _ _ particles _ _ ' and denoted by @xmath8 is associated , each of them being homeomorphic to @xmath20 .
thus , the set @xmath18 is not a priori a topological manifold , and the particles can not be loops .
it is assumed that at least one particle @xmath8 passes through each point @xmath21 of @xmath18 .
the first axiom of causality , i.e. , axiom 1a , states that , for each particle , an orientation can be selected among the two determined by the homeomorphism with @xmath20 such that , once these orientations are selected and fixed for the whole set @xmath19 of the particles , each event @xmath21 of @xmath18 can not be chronological to itself , i.e. , @xmath22 then @xmath23 .
the ` _ _ chronological relation @xmath24 _ _ ' is then a partial and anti - reflexive order on @xmath18 .
this is a global property of @xmath18 called the ` _ _ chronological principle _ _ ' , i.e. , there are no causal loops , and thus , @xmath18 is also called a ` _ _ chronological _ _ ' spacetime .
axiom 2 states that the intersection of any particle @xmath8 with any open @xmath25 of the alexandrov topology then @xmath26 , and also @xmath27 .
these sets are open for the alexandrov topology .
] is open for the topology on @xmath8 ( issued from the borel topology on @xmath20 ) . from axioms 1a , 1b , 2 and 3 , providing @xmath18 with the alexandrov topology @xmath28 ,
then @xmath29 is hausdorff , the chronological relation @xmath24 is _ _ past- and future - distinguishing _ _ is ` _ _ future distinguishing _ _ ' if @xmath30 , and ` _ _ past distinguishing _ _ ' if @xmath31 . ] and _ full _ , , @xmath32 such that @xmath33 , 2 ) if @xmath34 and @xmath35 then @xmath36 such that @xmath37 , @xmath38 and @xmath39 , and 3 ) the dual relations of 1 ) and 2 ) are satisfied with the dual chronological relation @xmath40 . ] and each event @xmath41 possesses a _ _ past and future reflecting _ _ ) @xmath42 of @xmath21 is an ` _ _ future and past reflecting _ _ ' open if @xmath43 then 1 ) @xmath44 ( _ future reflecting _ ) , and 2 ) @xmath45 ( _ past reflecting _ ) . ]
open ( for @xmath28 ) neighborhood .
then , woodhouse shows that @xmath18 is of the kronheimer - penrose type @xcite , where @xmath46 is the ` _ _ causal relation _ _ ' if @xmath47 and @xmath48 . ] and @xmath49 the ` _ _ horismotic relation _ _ ' if @xmath50 and @xmath51 . ] or ` _ _
horismos__. ' moreover , the chronological relation is transitive , which is not initially required in the kronheimer - penrose causal axiomatics in full generality .
the _ message functions , @xmath52 and @xmath53 . ] _
@xmath54 associated with each particle @xmath8 are defined on the tubular past- and future - reflecting alexandrov open neighborhoods @xmath55 , such that @xmath56 where @xmath57 .
these functions are unique and increase strictly monotonically for the chronological relation , open and continuous on @xmath55 .
the ` _ _ clock functions _ _ ' associated with each particle @xmath8 of @xmath19 are homeomorphisms @xmath58 , which increase monotonically for the chronological relation .
additionally , we define the ` _ _ radar coordinates _ _ '
@xmath59 defined for each particle @xmath8 . from this point on , for each pair of non - coplanar particles @xmath60 and @xmath61 ( if coplanarity has a meaning in @xmath18 ) , axiom 4a is satisfied if @xmath62 , @xmath63 such that @xmath64 , and the four radar coordinates @xmath65 ( @xmath66 ) together define a morphism : @xmath67 which is a one - to - one map .
then , from this axiom , it can be proven that @xmath18 is a topological manifold homeomorphic to @xmath68 .
in addition , from the `
_ _ @xmath10-congruence _ _ ' of particles @xmath69 with the parameterizations @xmath70 , is a _
@xmath10-congruence _ on @xmath71 if : 1 ) @xmath72 , @xmath73 such that @xmath74 and @xmath75 , and 2 ) @xmath76 a neighborhood of @xmath77 , then @xmath78 the ` _ _ evaluation map _ _ ' @xmath79 is a homeomorphism . ] on a set @xmath80 and from the@xmath81-differentiability assumption have no caustics , that is each message function @xmath82 restricted to @xmath8 with @xmath83 are diffeomorphisms . ] of the clock and message functions , woodhouse defined @xmath84-congruences and inductively proved that @xmath18 can be assumed to be _ smooth _ , i.e. , of class @xmath85 .
lastly , we call ` _ _ scalar potentials of metric @xmath86 _ _ ' smooth functions defined on @xmath55 with values in @xmath20 , each one associated with a particle @xmath8 , such as @xmath87 : @xmath88 then , the lorentzian metric @xmath3 on @xmath18 is the hessian of @xmath86 : @xmath89 which no longer depends , nonetheless up to a conformal factor , on @xmath8 and @xmath7 .
in addition , if @xmath90 is in a past and future reflecting open set of @xmath21 and @xmath91 , then there exists ( * ? ? ?
* lemma 4.2 ) a neighborhood @xmath92 of @xmath90 such that @xmath93 ( with @xmath94 ) is a ( unique ) light path and a @xmath85 one - dimensional submanifold .
each light path is a null curve , and they are null conformal geodesics of class @xmath85 .
hence , if @xmath91 then there exists a unique ( no caustics ) light path that is at least piecewise smooth .
one can notice that in the woodhouse axiomatics , we may have @xmath91 without light path joining @xmath21 and @xmath90 , meaning the latter have no common past- and future - reflecting open neighborhoods .
then , defining a new horismos @xmath95 such that @xmath96 if @xmath91 and @xmath51 , this new causal structure on @xmath18 ( or each @xmath55 ) is somehow natural " , i.e. , @xmath18 is a conformal manifold with its field of light cones , such that @xmath95 relates events on a cone or equivalently by a light path . and
then , @xmath49 is identified with @xmath95 .
based on this topology , the projective , conformal and weyl structures can be defined from the choices of metric connections , together with the constraints as indicated in the previous section .
that is to say , the particles must be geodesic for the weyl structure .
moreover , we recall the malament theorem @xcite , which can be used further with the topology provided by the woodhouse causal axiomatics : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ malament theorem _ let @xmath97 and @xmath98 be two past- andfuture - distinguishing spacetimes , each supplied with a metric connection to inherit a riemannian structure , and let @xmath99 be a causal isomorphism ( i.e. , keeping the chronological relation as well as its inverse ) from @xmath18 to @xmath100 .
then , @xmath99 is a smooth conformal diffeomorphism .
furthermore , @xmath99 preserves timelike curve orientations .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ by definition , @xmath99 is a conformal diffeomorphism preserving orientations if @xmath101 , where @xmath102 is a continuous function defined on @xmath18 .
if @xmath103 , then @xmath99 is said to be isometric .
in this preliminary section , we present the general mathematical framework used in the next two sub - sections treating the clock paradox .
let @xmath18 be a woodhouse spacetime supplied with a lorentzian metric @xmath104 ( defined from a scalar potential of metric ) and a metric connection denoted by @xmath105 providing @xmath18 with a riemannian structure .
we assume @xmath104 is at least of class @xmath81 and that it is so defined from clock and message functions at least of class @xmath106 as well as @xmath18 .
then , we consider two intersecting particles ( or _ ` trips ' _ , following the woodhouse definition ) @xmath8 and @xmath107 .
we point out again that the non - intersection is only considered in the definition of the @xmath108-congruences , and so it is associated with the choice for the projective structure defined by @xmath105 but not with the whole set of particles or in particular , @xmath8 and @xmath107 . in other words , there are two other @xmath108-congruences , each containing @xmath8 or @xmath107 and each associated with another possibly non - equivalent projective structure .
we denote by @xmath109 a tubular past- and future - reflecting open neighborhood of @xmath107 , onto which the message functions @xmath110 are defined . the neighborhood @xmath109
should possibly be reduced to avoid including caustics . in addition ,
@xmath8 and @xmath107 intersect at two points @xmath111 and @xmath112 only , the latter being contained in a past- and future - reflecting open sub - neighborhood @xmath113 of @xmath109 .
moreover , we assume that @xmath8 and @xmath107 are contained in @xmath113 between these two points ( chronologically ) .
the same notation @xmath114 is used for the restrictions of @xmath114 to @xmath113 .
we set @xmath115 , where @xmath116 is analogously defined to @xmath113 . in order to apply the malament theorem ,
@xmath117 is assumed to be connected and without boundaries .
moreover , we also consider @xmath117 as a conformal manifold , i.e. , a manifold being endowed with a class @xmath118 $ ] of conformally equivalent lorentzian metrics @xmath3 of which @xmath104 is a representative and being supplied with its natural causal structure , i.e. , we use the natural horismos @xmath95 . in addition , the class @xmath119 $ ] of metric connections @xmath17 considered throughout , each supplying @xmath117 with a riemannian structure compatible with the conformal structure defined by @xmath3 ( or @xmath104 ) , represents those metric connections such that , at least , together @xmath8 _ and _ @xmath107 are projective geodesics between @xmath111 and @xmath112 .
then , only on @xmath8 and @xmath107 are these specific _ restricted _ weyl structures defined .
if @xmath120 and @xmath121 are clock functions , respectively , on @xmath8 and @xmath107 , then @xmath8 can also be parameterized with @xmath122 using @xmath123 .
indeed , we can set @xmath124 : @xmath125 , where @xmath126 .
we choose @xmath122 such that @xmath127 and @xmath128 .
let @xmath129 and @xmath130 be the vector fields tangent , respectively , to @xmath8 and @xmath107 and defined from the parameterizations @xmath131 and @xmath122 .
then , in full generality , we have on @xmath117 : @xmath132 and @xmath133 , where @xmath134 and @xmath135 are real differentiable functions defined , respectively , on @xmath136 and @xmath137 only .
let @xmath138 be a metric field defined on @xmath117 , where @xmath139 is a real differentiable function on @xmath117 such that @xmath140 if @xmath141 and @xmath142 if @xmath143 .
thus , the relations @xmath144 and @xmath145 hold , and @xmath146 $ ] .
thenceforth , we can consider the parallel transport from @xmath141 to @xmath147 along a piecewise @xmath106 null geodesic @xmath148 ( geodesic with respect to both the conformally equivalent metrics @xmath104 and @xmath3 ) in @xmath149 .
we denote by @xmath150 the vector field tangent to @xmath148 , and by @xmath17 a metric connection associated with @xmath3 , as previously indicated .
then , we define the vectors @xmath151 ( @xmath152 $ ] ) parallel transported from @xmath153 along @xmath148 , such that @xmath154 and @xmath155 , where @xmath156 is the covariant derivative in the direction of @xmath150 . the transported vectors are reached using the parallel transport map @xmath157 which is defined such that : @xmath158 where @xmath159 , @xmath160 and @xmath161 , and @xmath162 is continuous and increasing for the horismos . on the other hand , this is a fundamental map accounting for the causal interpretations that each observer makes locally from a signal being propagated along a null geodesic .
obviously , from the vanishing covariant derivative of @xmath163 in the direction of @xmath150 , we have along @xmath148 , @xmath164 .
let @xmath165 and @xmath166 be the dual one - forms defined , respectively , on @xmath8 and @xmath107 such that @xmath167 .
then , we define on @xmath168 the one - form @xmath169 such that : @xmath170 figure 1 we now consider the inverse situation with two other parameterizations for the particles @xmath8 and @xmath107 , namely , @xmath7 and @xmath171 respectively , but using instead the map @xmath172 ( see figure 1 ) . in the same way , we set @xmath173 : @xmath174 where @xmath147 , and @xmath175 and @xmath176
. thus , one obtains an expression analogous to along @xmath177 , where @xmath178 and @xmath179 are defined in a way similar to @xmath130 and @xmath180 : @xmath181 in conclusion , only the parameterizations @xmath7 on @xmath8 and @xmath122 on @xmath107 are relevant , with the others being deduced afterwards from the message functions .
lastly , if @xmath117 is relatively compact , the integrals computed subsequently will always be defined , since the integrands will be lipschitzian . however , in addition , the lipschitz constants will no longer depend on @xmath117 , according to the zeghib theorem @xcite .
indeed , all the maps on @xmath117 are defined from a foliation of codimension one defined from a given @xmath108-congruence . in the present framework
, we shall define a simultaneity map and certain types of proper times from which the first difficulty that arises is the realization of a differential aging formula consistent with the experimental results .
let the set of pairs @xmath182 be such that @xmath183 .
an expression developed for @xmath184 could be : @xmath185 , and thus , @xmath186 a priori . setting this equality is like defining @xmath7 as a function of @xmath122 , or conversely .
then , we impose the condition @xmath187 .
now , let the map @xmath188 be such that @xmath189 when @xmath183 .
thus , the relation @xmath190 necessarily holds , since @xmath187 .
moreover , @xmath191 is bijective and reversible . the map @xmath191 transforms timelike curves into other timelike curves . however , if @xmath192 on @xmath8 , then on @xmath107 , it is equivalent to @xmath193 @xmath194 @xmath195
@xmath194 @xmath196 @xmath194 @xmath197 . thus , the map @xmath191 is a causal isomorphism from @xmath8 to @xmath107 .
however , from the malament theorem , there exists on @xmath117 a ( non - unique ) @xmath106 conformal extension @xmath198 of @xmath191 , which , in addition , preserves the particle orientations while passing from @xmath8 to @xmath107 .
let us note that the map @xmath191 can never be one of the message functions @xmath114 restricted to @xmath8 , because in this case , the relation @xmath199 with @xmath200 should hold with @xmath191 being causal , which is obviously impossible .
one calls @xmath191 the ` _ _ simultaneity map__. ' it is almost fully geometrically unspecified a priori and is non - unique , depending on the parameterizing clock functions .
however , it is well - known that , in contrast , it is causally worked out as being unique , as shown with the malament theorem on simultaneity @xcite . from this point on
, we shall define a variant of the chronometric hypothesis that is nonetheless quite different . instead of assuming that the interval @xmath2 associated with @xmath3 is identified as @xmath201 , where @xmath1 is the proper time of a clock ,
we suppose that @xmath202 on @xmath8 , and @xmath203 on @xmath107 . that is to say , the interval @xmath2 is identified as the sole physically " measurable quantity , which represents the time intervals @xmath204 _ displayed _ by the clocks
. therefore , these are as varying as the clocks used , contrary to the proper times and depending on each particle but not on their parameterizations .
moreover , the constants @xmath205 and @xmath206 are the numerical values for the speed of light when making use of the corresponding parameterizations @xmath7 and @xmath122 . then
, we set : @xmath207 where @xmath208 and @xmath209 are the
usual " proper times associated with ( not defined on ) the particles @xmath8 and @xmath107 , respectively .
the one - form @xmath169 is ascribed to the infinitesimal proper time @xmath210 of @xmath8 , because , first , @xmath180 depends on two points , one on @xmath8 and the other on @xmath107 , and secondly , the term @xmath211 is the well - known @xmath212 factor of special relativity between one observer on @xmath8 and another on @xmath107 .
hence , the proper time of an object on @xmath8 is evaluated on @xmath107 , onto which @xmath169 is defined , hence the notation .
implicitly , this shows the relative " character of the concept of proper time , which will also be truly justified further by using certain other arguments presented below .
then , these two one - forms are defined strictly on the product @xmath213 fibered " by a simultaneity map @xmath191 .
in addition , since @xmath198 puts into correspondence the ( co)tangent vector spaces on @xmath8 and @xmath107 , the relations @xmath214 and @xmath215 hold while preserving the particle orientations , since @xmath216 .
moreover , at least between the points @xmath111 and @xmath112 , we have the important property @xmath217 , but naturally @xmath218 a priori .
we suppose this last equality is satisfied when choosing , among all of the maps @xmath198 , those that satisfy the condition @xmath219 below .
let @xmath198 be a conformal diffeomorphism , such that @xmath217 .
we denote by @xmath219 the condition @xmath220 when @xmath198 is restricted to @xmath8 or @xmath107 only .
therefore , @xmath198 carries out the exchange of the two particles . from all of these relations , we have @xmath221 on @xmath8 and @xmath107 , only if @xmath222 .
this relation would be satisfied if , in particular and due to the malament theorem , the following equalities @xmath223 and @xmath224 held on @xmath8 .
in fact , we would obtain the relation @xmath225 on @xmath8 , because @xmath180 and @xmath179 , as well as @xmath178 and @xmath130 , have the same norms with respect to @xmath3 . in order to satisfy these equalities
, we set below the defining constraint or hypothesis on @xmath198 , namely , the constraint @xmath226 : @xmath198 satisfies the condition @xmath226 if @xmath198 preserves the riemannian structure on @xmath227 .
then , as we will show in the proof below , the relation @xmath228 holds if @xmath219 and @xmath226 are assumed
. then , we have : * lemma * _ if @xmath219 and @xmath226 are satisfied , then @xmath221 on @xmath8 and @xmath107 . _ _ proof _ : indeed , firstly , if @xmath198 preserves the riemannian structure , then for all continuous vector fields @xmath229 and @xmath230 on @xmath117 , there exists a continuous function @xmath231 on @xmath117 such that @xmath232 , where @xmath233 and @xmath234 .
then , considering @xmath150 , a null vector tangent to the piecewise @xmath106 null geodesic @xmath148 from @xmath21 to @xmath235 , and @xmath163 ( @xmath152 $ ] ) , the parallel transported vectors along @xmath148 of @xmath178 at @xmath21 , we find @xmath236 , where @xmath237 and @xmath238 . secondly , since @xmath239 then @xmath240 .
therefore , from lemma 4.2 of woodhouse , the geodesic @xmath148 from @xmath21 to @xmath235 is unique as well as a piecewise @xmath106 null geodesic from @xmath241 to @xmath242 denoted by @xmath243 .
moreover , the latter is such that @xmath244 on @xmath117 .
moreover , since @xmath245 at @xmath241 and @xmath243 is unique , then for all @xmath152 $ ] , we find @xmath246 on @xmath243 .
thus , @xmath247 and @xmath225 .
however , from @xmath219 , we also have @xmath248 when @xmath198 is restricted on @xmath107 and so the expected result .
then , @xmath249 is the pull - back of @xmath169 by @xmath198 .
@xmath250 hence , on the basis of a physical interpretation of the signals received by each particle , i.e. , considering again that the proper time of each particle is actually assessed as such only by another particle running in the time " given by its own clock ( in special relativity , the proper time for a physical frame is always compared , using a factor @xmath212 , to another physical frame ) , then the differential aging @xmath251 between the clocks can be computed _ a priori _ with the formula : we set @xmath251 , the differential aging , with the usual " proper times : @xmath252 then , in making use of @xmath198 and the constraints @xmath219 and @xmath226 , we find : @xmath253 thus , we must clarify the links between the constraints on @xmath198 and some physical properties .
nevertheless , we can not make allowance for concepts inherited from special relativity to define a differential aging formula as above . now , in order to seek and to justify a correct formula for the differential aging , we will also consider a situation with a third particle @xmath254 , which is more related to physical experiments that have actually been performed , such as the hafele and keating experiment @xcite .
let @xmath254 be a third particle not necessarily passing through the points @xmath111 or @xmath112 , with its two message functions @xmath82 .
we assume that @xmath255 and @xmath256 as well as every point @xmath21 of @xmath254 such that @xmath257 are in @xmath117 .
according to @xmath254 and the signals it receives coming from @xmath8 and @xmath107 , i.e. via the parallel transports along light paths , a difference of proper times @xmath258 between the two particles @xmath8 and @xmath107 could be written as follows : we define @xmath259 where @xmath260 and @xmath261 are the parallel transported vectors at @xmath262 from , respectively , the vectors @xmath263 , @xmath264 , and @xmath265 is a vector field tangent to @xmath254 provided with the parameterization @xmath266 ( see figure 2 ) .
figure 2 in addition , we assume that @xmath267 with @xmath268 on @xmath254 , where @xmath269 . moreover , we set @xmath270 and @xmath271 . the expression must be absolutely independent on @xmath254 , and we shall show that a particular condition must be satisfied to obtain such independence .
we denote by @xmath272 the null geodesic from @xmath273 to @xmath274 ( see figure 2 ) , @xmath243 from @xmath275 to @xmath90 , and @xmath148 from @xmath276 to @xmath275 .
in addition , we denote again by @xmath265 the vector field tangent to @xmath254 at @xmath277 and @xmath260 and @xmath180 , the parallel transported vectors , such that : @xmath278 then , we have : * proposition .
* _ there exists a unique differentiable function @xmath279 on @xmath8 , independent on the representative metric connection @xmath17 and @xmath254 , such that @xmath280 and @xmath281 at @xmath275 with @xmath282 and @xmath283 .
we obtain an analogous relation for @xmath284 on @xmath107 . _ _ proof _ : let @xmath285 be the message function @xmath286 restricted to @xmath254 . as for @xmath191
, we extend the latter map to a @xmath106 causal conformal map defined on @xmath117 and denoted by @xmath287 .
the relation @xmath288 holds , and since @xmath265 and @xmath130 have the same norms with respect to @xmath3 , @xmath289 . additionally , we define @xmath290 such that @xmath291
. then , we have the relations : @xmath292 however , @xmath293 acts as an isometry with respect to @xmath3 on the vectors @xmath265 and @xmath260 , as well as @xmath294 on @xmath129 .
thus , we can set the relations : @xmath295 where @xmath296 and @xmath297 are isometries with respect to @xmath3 .
on the particle @xmath8 , the carriage of @xmath178 from @xmath21 to @xmath298 is not a parallel transport with respect to @xmath104 , because @xmath8 is not a priori geodesic under @xmath104 .
accordingly , we can set @xmath299 , although , @xmath300 . setting @xmath301 ,
if @xmath302 , then we necessarily find : @xmath303 on @xmath8 .
then , integrating from @xmath21 to @xmath298 , we deduce the integral definition of @xmath304 : @xmath305 from this definition , we set @xmath306 , where @xmath307 additionally , @xmath308 , since @xmath309 , but @xmath310 at @xmath21 and thus , @xmath311 . in conclusion , there exists an isometric transformation @xmath312 with respect to @xmath104 , such that @xmath313 , and consequently : @xmath314 thus , we deduce with @xmath315 that : @xmath316 however , @xmath180 and @xmath290 are normalized with respect to the metric @xmath3 at @xmath241 , namely , the metric we denote hereafter by @xmath317 . thus , necessarily , the holonomy map along the loop @xmath318 , i.e. , @xmath319 , is a dilatation composed with a lorentz transformation @xmath320 of @xmath317 : @xmath321 such that @xmath322 .
the map @xmath319 can always be composed on the left with another lorentz transformation @xmath323 preserving @xmath290 , but such that @xmath324 . therefore , in full generality , we can always choose a map @xmath319 such that the associated lorentz transformation @xmath320 remains @xmath130 invariant .
thus , we deduce that there exists a map @xmath319 and , moreover , a function @xmath13 on @xmath107 such that : @xmath325 therefore , in relation , the only way to pass from @xmath180 to @xmath290 using signaling , that is to say , passing from @xmath8 to @xmath107 using null geodesics , is to apply @xmath319
. however , on the loop @xmath162 , equivalently , the signals at @xmath90 coming from @xmath21 sometimes via @xmath298 or sometimes via @xmath241 must be the same .
this signal identification , regardless of the light paths followed , means that @xmath319 leaves @xmath317 invariant .
hence , the relation @xmath326 must hold .
additionally , this means that the holonomy group of @xmath17 at each point in @xmath117 is the lorentz group of @xmath3 at this point .
this is the present case , since , from condition @xmath226 , the metric connection @xmath17 provides @xmath117 with a riemannian structure @xcite .
thus , there always exists a map @xmath319 leaving @xmath317 invariant .
then , at the point @xmath241 and with @xmath327 restricted to @xmath107 , we find : @xmath328 with @xmath282 et @xmath283 .
the relation holds .
in addition , we can deduce at @xmath276 , using an analogous transformation @xmath329 and with @xmath327 restricted to @xmath8 : @xmath330 with @xmath331 and @xmath332.@xmath250 then , we obtain in particular : @xmath333 with @xmath282 and @xmath334
. however , @xmath254 may vary and be such that @xmath335 . the relation @xmath336 holds as well . performing a change of variables using the map @xmath198
, we find : @xmath337 however , the two integrands are positive , so we can set new supplementary defining constraints @xmath338 and @xmath339 on @xmath198 , which is not fully defined since on @xmath8 , if @xmath340 , then we find only : @xmath341 such that @xmath342 , then @xmath343\,f^-_{{\tilde p}}(x),f^+_{{\tilde p}}(x'')\,[\,\,\subset{{\tilde p}}$ ] . by definition , the constraints @xmath338 and @xmath339 are : @xmath191 satisfies the condition @xmath338 if @xmath344 and the condition @xmath339 if @xmath345 nevertheless , we must examine whether any of these constraints are redundant . however , setting @xmath346 and @xmath347 , then necessarily , @xmath189 and @xmath348 . hence , a constraint is equivalent to the other if and only if @xmath349 , a condition that is not always satisfied , and therefore , these constraints are not equivalent .
in addition , we point out that these two constraints could be physical " constraints rather than geometrical " constraints if the functions @xmath279 are ascribed to physical quantities , and that is indeed the case , as we shall see .
then , the simultaneity map @xmath191 is related to the physics instead of the causality only .
thus , we can conclude that @xmath350 .
then , although the exponential factors in the integrands may have common constant factors , we set a priori for all @xmath254 the definitions below for the differential aging between @xmath8 and @xmath107 : we set the differentials @xmath351 and @xmath352 such that : @xmath353 and @xmath354 from these new definitions and the previous developments , we have : * theorem . * _ if the conditions @xmath219 to @xmath339 are satisfied on @xmath117 , then @xmath355 regardless of the particle @xmath254 .
_ _ proof _ : obvious from the proposition and the previous definitions.@xmath250 then , @xmath352 is invariant with respect to @xmath254 .
nevertheless , the proper times must be redefined .
the conformal ( and relative ) proper times are defined by the relations : @xmath356 and @xmath357 two other formulas can be given for the expression of the difference of proper times .
the first is obtained with @xmath338 or : @xmath358 and the second on @xmath107 , analogous to the first , with the condition @xmath339 .
we note that they are close to and they indeed reduce to if , for example , @xmath359 in .
the relation allows us to solve the clock paradox , since the unspecified particle @xmath254 can be taken as equal to @xmath8 or @xmath107 .
however , this amounts to modifying the computation viewpoints for the integral of the differential aging formula . by viewpoint " , we mean that the integrating variable specifies it , i.e. , the pairs @xmath360 or @xmath361 in the subindices in the differences @xmath362 . as a result , we see that the computation of the difference of times between the clocks is the same as that evaluated by the two clocks : @xmath363 finally , the chronometric hypothesis must be amended by the conformal factors and the gamma " factors . for this
, we consider the particular viewpoint in the minkowski spacetime with @xmath359 .
it follows that @xmath364 and ( considering that the functions @xmath279 vanish for simplicity ) : @xmath365 and thus finally , if @xmath3 is close to @xmath104 , i.e. , @xmath7 is close to the proper time of the particle @xmath8 , the usual factors @xmath212 obtained from the minkowski metric can be used .
therefore , we find : @xmath366 which is clearly the same formula that is usually used in the minkowski spacetime .
however , we also identify the origin of the ch , by taking for a proper time a very particular clock function @xmath266 , which is substituted , due to the ch , by a known gravitational time .
however , the question then follows : if a clock stops beating its unit of time and remains in a physical steady state " , does its proper time change ?
meaning , is its proper time what the clock displays ?
the reply might be yes ( ! ) , depending on which other times " the clock sequentially displays as its own " displayed time @xmath266 .
the conclusion is essential in itself : the functions @xmath279 make it possible to connect the data of a temporal counter , like a clock with @xmath266 , to a physical duration .
thus , necessarily , the previous functions are physical and associated mainly with a relation between a physical state of an object with its information content and other objects with their own physical states and information content . the physical / geometrical interpretations of the functions @xmath279 , for instance , can not really be performed without the use of some results given by typical physical experiences in general relativity .
nevertheless , they can be related to more general physical aspects .
we propose this as a first step .
first of all , the natural " hypothesis would be that these functions come out of metrics @xmath3 with the general form @xmath367 defined only on @xmath8 ( or @xmath107 ) .
then , the functions @xmath231 can not be scalar fields on @xmath117 and a fortiori on @xmath18 . nevertheless , at a particular @xmath368 on @xmath8 , the value @xmath369 determines a set of continuous scalar fields @xmath102 defined on any tubular neighborhood of @xmath8 contained in @xmath117 such that @xmath370 .
hence , @xmath369 defines what is called a germ at @xmath368 , i.e. , a class of functions @xmath102 satisfying the previous equality .
we denote by @xmath371_z$ ] this class of functions @xmath102 and the full set of such classes on @xmath8 defines a ` sheaf ' of rings of continuous functions onto @xmath8 , denoted by @xmath372_z\,/\,z\in\,p\}$ ] . obviously ,
if @xmath373_z$ ] then we might have @xmath374_{z'}$ ] if @xmath375 , i.e. , @xmath370 and @xmath376
. nevertheless , if we wish to obtain local results at @xmath368 , we can take any continuous function @xmath373_z$ ] to do the computations .
thus , we can set @xmath377 at @xmath368 , and use this expression to compute the metric connection @xmath17 at @xmath368 associated to @xmath3 from the metric connection @xmath105 associated with @xmath104 . as a result , at @xmath368 and if @xmath102 is assumed to be differentiable , the relation : @xmath378 holds , where @xmath178 and @xmath265 are any vector fields defined in an open neighborhood of @xmath368 .
we note that if @xmath102 and @xmath379 both belong to @xmath371_z$ ] , then , in general , @xmath380 at @xmath368 .
thus , a priori , there are as many metric connections @xmath381 at @xmath368 as there are functions in @xmath371_z$ ] whose gradients differ at @xmath368 .
hence , we must define a sub - sheaf of @xmath382 denoted by @xmath383 such that @xmath384_z\,/\,z\in\,p\}\subset s_p(w)$ ] , where @xmath99 is a continuous vector field defined on @xmath8 only , and the germ @xmath385_z$ ] denotes the class of functions @xmath102 such that @xmath386 and @xmath387 .
any function @xmath99 only defined on @xmath8 is suitable ( although continuous on @xmath8 ) , but between @xmath231 and @xmath99 , we impose the relation at @xmath388 : @xmath389 where @xmath390 is the dual one - form of @xmath99 such that @xmath391 on @xmath8 .
this can be presented with a more meaningful formula .
first , let @xmath265 be any vector field on @xmath8 , with its corresponding space vector @xmath392 such that @xmath393 , where @xmath178 is assumed to be the timelike vector field tangent to @xmath8 and normalized with respect to @xmath3 .
of course , @xmath392 satisfies the relations @xmath394 and @xmath395 .
secondly , with @xmath178 , we have the relation : @xmath396 however , since @xmath397 , then @xmath398 , and thus , we deduce the relations : @xmath399 , and @xmath400 therefore , we obtain @xmath401 . the relation is thus written in the simplified form : @xmath402 .
one can rewrite the preceding formula in another form that is more amenable to easy interpretation : @xmath403 these two relations depend on the clock chosen , i.e. , the clock function @xmath7 and , thus , the functions @xmath102 . then
, if there are no applied forces @xmath404 when considering the metric @xmath104 , i.e. , @xmath405 , then naturally it is no longer the case considering instead the metric @xmath3 if @xmath406 and , reciprocally , if @xmath407 and @xmath408 .
thus , this exactly represents the ` _ _ pseudo - fields _ _ ' of the gravity viewpoint , whose explicit terminology can be found in the einstein paper on the twin paradox @xcite as well as in the lift metaphor of the einstein equivalence principle @xcite .
hence , their origins can be ascribed to the choice of a clock _ displaying _ a time differing from the proper time .
then , to the lift metaphor , we could append a sort of _ unregulated clock metaphor _ to special relativity : a massive object is in inertial motion along a straight line at a constant relative speed when compared to a certain clock , and the latter suddenly starts to have a delay that can not be observed in the absence of an absolute time of reference . as a result
, one would observe a sudden increase in the relative speed and , thus , a non - vanishing relative acceleration .
however , we must mention that the accelerations given by accelerometers remain null , so that the motion remains inertial , even though the relative accelerations are modified .
we can ascribe this variation to a geodesic or curvature variation to become non - zero , as well as to a pseudo - field of gravity occurrence .
thus , these fields are pseudo " only relative to the choice of the conformal metric carried out a priori during the variation and , thus , of a clock .
if the variations are not considered as being due to changes in temporal coordinates or clocks , then , quite naturally , two metrics @xmath3 and @xmath104 can be simultaneously associated with each particle @xmath8 . in the geometry defined by @xmath3 , there are no forces applied on @xmath8 , i.e. , @xmath409 , and in the geometry defined by @xmath104 , a force @xmath404 is applied on @xmath8 and is interpreted as fictive on the metric @xmath3 , i.e. , ascribed to a pseudo - field of gravity .
on @xmath8 , one would then have , for example , the forces of gravity or centrifugal forces , which would be fictive forces on @xmath3 but true " forces on @xmath104 .
nevertheless , why would that not remain valid if the variations are considered as changes in clocks or clock functions ?
indeed , how do we note such a change ?
we do not have a local absolute clock against which we can calibrate the others .
this is well highlighted in the _ allan deviation _
@xcite of a pool of identical clocks .
then , deduced from the latter , the precision of an atomic clock is no longer defined from an absolute , standard clock linked to a given temporal orientation , but from a statistical relation between clocks , and moreover , time drift bias can not be defined .
thus in practice , an unobservable change in clock function is also equivalent to the sole observed change in geodesics ( related to curvature ) , and then , the lift metaphor can no longer be distinguished from that of an unregulated clock .
therefore , this at once poses the question of the relation between clock functions and accelerations , forces or geodesics .
thus , if the relation @xmath410 holds , we should interpret @xmath411 as a quantity associated , in some ways , with the mechanical work of a force @xmath404 , fictive on @xmath3 , or with an energy potential . from the previous equality ,
one thus notes that the fact that this scalar product does not change when passing from @xmath412 to @xmath99 means that the work would not be , in any case , fictive .
additionally , we can notice elsewhere that a non - fictive work is at the basis of the schild argument for deducing the gravitational frequency shifts in light ( * ? ? ?
187189 ) .
now , starting from the relation , the function @xmath279 should be a priori the mechanical work of force @xmath99 , along @xmath8 and dependent on the path . as we shall see
the function @xmath279 could be the variation in the total energy of the particle minus the kinetic energy .
lastly , it is significant to note that this interpretation for @xmath279 differs for instance from that given by wheeler @xcite for the weyl potentials .
wheeler identified @xmath279 with the classical action ; it would be very appealing indeed to connect to the physics .
however , the examples presented below prevent us , disappointingly , from making such an identification and lead us to dismiss such a proposal in the present situation .
one considers in this section only verifications up to the order @xmath413 , corresponding to a first approximation of the definitions of the functions @xmath279 , whose refinements in their complete physical interpretations , under study , remain to be made if one wants to move to results up to the order of at least @xmath414 .
this is the simplest situation , occurring when @xmath415 , but with at least two distinct clocks on the same worldline of a freely falling particle , with one of them not running .
thus , in this case , @xmath416 if @xmath241 is on @xmath8 between @xmath111 and @xmath112 , and it starts to run only while arriving at @xmath112 , so that @xmath128
. then , the relation is written in the form : @xmath417 where @xmath418 because the clock on @xmath107 is not running .
moreover , @xmath419 is then linked only to the internal energy of the running clock , since it is in free fall .
let us consider the term of the entropy @xmath420 in classical thermodynamics , such that @xmath421 , where @xmath422 is the infinitesimal internal " work ( work not related to the kinetic energy ) .
the terms @xmath423 and @xmath424 are dimensionless quantities and represent energies per unit of mass or per unit of thermodynamic energy , such as @xmath425 for instance ( @xmath150 being boltzmann s constant ) . setting @xmath426 , where @xmath427 is the entropy , and @xmath428 where @xmath429 is the internal work , and assuming @xmath430 with @xmath431 constant while the clock operates , then : @xmath432 then , we consider that the clock produces a positive variation in entropy during its irreversible counting : @xmath433 , and @xmath434 .
thus , the relation : @xmath435 holds . setting by definition @xmath436 , where @xmath1 is the proper time of this clock , and considering @xmath7 as the time the clock displays as an increasing counter , then at each point
@xmath276 such that @xmath437 , we have the variation : @xmath438 hence , the second law of thermodynamics would be equivalent to the monotonic increase of the proper time @xmath1 with respect to the displayed time @xmath7 , with entropy linking them .
the question remains as to what the verification conditions are for the chronometric hypothesis .
one notes that we must start from the relation and the scalar potentials of metric @xmath439 .
the metric field @xmath3 is independent of @xmath8 and @xmath7 up to a conformal factor .
in fact , the conformal factor depends upon the first - order derivatives of @xmath7 at point @xmath368 only such that @xmath440 . at this point @xmath368
, one can show that the conformal factor is proportional to the square of the differential @xmath204 and , so , of @xmath441 from .
the mere verification of the ch would involve having a constant non - vanishing factor to set down a relation of the form @xmath442 up to a constant factor .
the condition is that for every point @xmath388 between @xmath111 and @xmath112 , @xmath443 .
then , necessarily , we find @xmath444 .
this can not be required for the whole of the interval between @xmath111 and @xmath112 , and then , it remains only to suppose that there exist @xmath445 and @xmath446 such that @xmath447 when @xmath448 .
hence , only in that case , the chronometric hypothesis would be equivalent to einstein s adiabatic hypothesis .
moreover , if @xmath449 , then it requires , first , that @xmath450 is high when @xmath448 , and second , that we have an adiabatic process .
these are conditions that atomic clocks should satisfy with @xmath451 closed " to @xmath111 .
additionally , let us investigate the case with @xmath452 and @xmath431 constant .
for instance , the clock starts to spin or acquire a moment of rotation .
then , to first order , we find the following expression : @xmath453 in other words , this process leads to a slower aging " of the clock compared to the case without rotation .
this phenomenon should be strictly different from the rejuvenation " case found by prigogine _
_ @xcite in the same circumstances of a rotating thermodynamic system . nevertheless
, they did not account for the time of rotation while this rejuvenating effect occurred .
this time could compensate for the rejuvenation , such that the system would only age more slowly , and then , actually , their phenomenon would be very similar to the present case .
this well - known experiment @xcite brings into play three reference frames , each one using a cesium atomic clock : two jet aircrafts , one moving westward @xmath454 and the other eastward @xmath455 , and a ground control base station @xmath8 on the equator at null altitude .
actually , hafele and keating used a fourth reference frame , i.e. , an inertial reference frame @xmath254 ( thus , in particular , without rotation ) associated with a remote star .
following our notations , they computed the integrals @xmath456 and @xmath457 to the first order of the relativistic effects .
the two aircrafts were supposed to fly along the equator at the same velocity @xmath102 with respect to the ground .
the altitudes and the maximal velocities during this experiment were typically about @xmath458 and @xmath459 , respectively .
the earth s equatorial radius is about @xmath460 and the mean angular velocity for the earth rotation about @xmath461 .
thus , the equatorial tangential velocity is about @xmath462 .
the aircraft moving towards the east moves in flight at a speed @xmath463 with respect to the frame @xmath254 , and the aircraft moving towards the west travels at a speed of @xmath464 .
in addition , we denote by @xmath465 the constant of gravity at the equator .
now , considering the first term in the integral in the formula for @xmath457 , the term @xmath466 in is the factor @xmath212 for one of the aircrafts . for the aircraft flying toward the east
, we find : @xmath467 in this expression , @xmath3 is the minkowski metric , since we refer only to the remote inertial frame @xmath254 . then , to the first order , we obtain : @xmath468 the second term in the integral can also be written to the first order as : @xmath469
lastly , according to our interpretations , the coefficient @xmath470 is equal to 1 , because @xmath8 is not subjected to an external working force . on the other hand
, the coefficient @xmath471 will reflect the work @xmath472 of the external force that brings the aircraft @xmath455 to its cruising altitude and speed . under these conditions , and
contrary to the function @xmath279 in the previous example for one clock only , it is necessary to take with @xmath473 the expression @xmath474 , where @xmath472 is the work per energy of mass unit @xmath475 , where @xmath476 is the mass of the aircraft .
thus , with @xmath477 being the work of the forces ( directed to the ground ) given by the accelerometers and considering @xmath465 to be roughly constant , then we find the relation : @xmath478 hence , to the first order , the relation : @xmath479 is exactly that given by hafele and keating .
thus , the following integrands correspond to each integral : @xmath480 the integrals are both computed with respect to the variable @xmath481 , and thus , we can subtract them to find the differential aging @xmath482 between the airplanes .
equivalently , the term : @xmath483 must be integrated .
rigorously , this difference is possible only if the two aircrafts are in flight at the same speed , in opposite directions , and at the same distance from the base @xmath8 .
this hypothesis is also needed to compare with the integrand in the difference @xmath484 , i.e. , to avoid passing by the basis @xmath8 .
then , we easily obtain the corresponding following integrands : @xmath485 subtracting these two terms gives the same result .
one can also consider the use of the formula . as previously indicated , the metric @xmath3 is the minkowsky metric , only far from the earth . in the vicinity of the earth , this is the schwartzschild metric , and from the viewpoint of each aircraft in motion or of the base station @xmath8 , this metric is in a rotating reference frame .
in addition , it must be synchronous , or equivalently , conditions such that @xmath486 must be satisfied .
the schwarzschild metric can be approximated by the expression ( @xmath487 ) : @xmath488 where @xmath489 , and @xmath490 is the gravitational radius of the earth . in a rotating reference frame such that @xmath491 , @xmath492 , and by replacing @xmath493 by @xmath494 with the approximation @xmath487 and taking into account the orders of magnitude for @xmath495 , @xmath496 and @xmath497 , we can write the following approximation of @xmath2 : @xmath498 however , we need a metric in a synchronous frame , so a change of basis is performed by replacing @xmath499 by @xmath500 and also by multiplying all of the terms by a factor @xmath229 ( i.e. , the basis is defined up to a conformal factor and a projective transformation ) : @xmath501 the constants @xmath229 and @xmath502 are chosen in order to normalize the metric ,
i.e. , its determinant is equal to @xmath503 on the basis @xmath504 , and with the constraint : @xmath505 for the value of @xmath506 .
then , to the first order , we find that @xmath507 and @xmath508 , and so : @xmath509 in order to evaluate the factor @xmath212 for this metric , the tangential velocity denoted by @xmath379 of an object with respect to this rotating reference frame is such that : @xmath510 with the upward velocity @xmath511 defined by : @xmath512 setting @xmath513 at the equator , the inverse of the factor @xmath212 for this metric is roughly of the form : @xmath514 \right\}.\ ] ] let us now consider the formula to compute @xmath515 .
we assume the angles are @xmath516 toward the east . with @xmath517 or @xmath518 and assuming @xmath519 , then , for the two cases , we find the relations : @xmath520 \right\},\\ & & \notag\\ \beta^-_e=-\frac{\mathrm{g}\,h}{c^2 } , & & \frac{e^{-\beta^-_e}}{g(\eta,\zeta^+_1)}\simeq 1+\frac{1}{c^2 } \left\ { \mathrm{g}\,h - v\,v_t-\frac{v^2}{2 } -\frac{h}{3r } \left [ 4\,v\,v_t+v^2 \right ] \right\}.\end{aligned}\ ] ] hence , in order to obtain @xmath515 , we need to integrate : @xmath521 with respect to @xmath7 . at the zeroth order in @xmath317 ,
we find again the formula .
we consider now the formulas obtained from the viewpoint of each aircraft , that is to say , @xmath522 or @xmath523 .
first of all , in their proper frames , none of the external forces work , since the origins of the reference frames are the aircrafts themselves . however , the functions @xmath279 are the evaluations of the work of the forces being applied on the aircraft which is not at the origin of the considered reference frame .
formula must be applied : @xmath524 there are the following correspondences : @xmath525 is the work of the forces given by the accelerometers placed on the aircraft moving toward the west and estimated in the reference frame of the same aircraft .
reciprocally , @xmath526 is the work of the forces given by the accelerometers placed on the aircraft moving to the east and estimated by the plane flying toward the west . from the point of view of each airplane
, these two sources of work would be the same if they could simultaneously be estimated with the function @xmath191
. however , in fact , this is not even necessary because the function @xmath527 takes into account the temporal delay via the map @xmath528 .
for the two differences , it is thus necessary to estimate the coefficients @xmath529 ascribed to @xmath455 .
there is the following correspondence : @xmath530 \right\},\ ] ] where @xmath531 , a function of @xmath532 , is the velocity of the airplane travelling towards the east from the point of view of @xmath454 , and @xmath511 is its upward velocity .
the expressions for the functions @xmath279 are : @xmath533 where the altitudes are functions of the values of the clock function @xmath532 , @xmath534 is the altitude of the airplane moving toward the east , and @xmath535 is that for the plane traveling west .
these altitudes are , actually , relative altitudes between the two airplanes , not respective to the ground .
one can thus make the reasonable assumption that @xmath536 . under these conditions , by considering @xmath537 , the integrand of @xmath522 is of the approximate form : @xmath538 \right\}.\ ] ] this expression is far from that obtained in the reference frames of @xmath8 or @xmath254 , but in fact it means that the comparison of the integrated terms is no longer valid .
it is the complete integral that must be computed .
it follows that the doppler terms in @xmath539 must include dilatations and contractions of the durations depending on whether the airplanes are moving away from each other or are approaching one another along the equator .
the contributions of these two modifications in duration will be the same and will be cancelled out at the conclusion of a circumnavigation .
thus , only the effect of the term in @xmath540 will persist , leading to the same final result .
we have seen that the clock paradox is perfectly solved within the relativity framework due to the asymmetry resulting from the factors @xmath541
. however , at the same time , it is necessary to ascribe a new meaning to these conformal factors .
these factors are strongly associated with the concept of total energy minus kinetic energy , which is not necessarily defined by a potential along each timelike worldline ( particles ) .
actually , these functions @xmath279 imply in themselves the data of an additional dimension .
more precisely , we have sheafs of germs of functions @xmath279 that are classes of scalar fields , which mimic " , at each given event , the values of the functions @xmath279 .
in other words , we could have no such scalar fields on @xmath18 , but there is always the need for non - vanishing conformal scale factors , i.e. , values independent on the _ locus _ in the spacetime , which is an extra dimension . in the same vein , we can suggest a reply to a concluding question of ehlers _ et al . _ , asking whether other interpretations of the so - called weyl _ streckenkrmmung _ bivector @xmath542 rather than those ascribing the latter to the electromagnetic field _ might contain some physical truth"_. the fact is that as a field , per se , we may have no such field @xmath542 , but only germs of such fields along particles or beams of particles .
again , this bivector @xmath542 would be related as a germ to derivatives of the functions @xmath279 . however , in any case , such ( bi)vectors would be defined only in jet manifolds of germs of functions and , as such , would be associated with an extra variable .
thus , it seems necessary to consider a spacetime supplied with an extra dimension of energy , i.e. , the conformal scale factors , that translates the various physical transformations without gravitational origins .
therefore , it is basically a spacetime " of five dimensions upon a conformal spacetime of four dimensions .
however , the geometrical structure should also be a product of copies of the spacetime @xmath18 fibered by conformal scale factors ( or equivalently , @xmath191 ) . in conclusion , this extra dimension is absolutely necessary in order to provide an account of the internal evolution of objects in relation to the spacetime structure , and it alone gives a meaning to the concept of the proper time of an object in relation to the chronology given on a spacetime .
ehlers , j. , pirani , f. a. e. , & schild , a. ( 1972 ) .
the geometry of free fall and light propagation . in l.
oraifeartaigh ( ed . ) , _ general relativity , papers in honour of j. l. synge _ ( pp .
oxford : clarendon press .
kundt , w. , & hoffmann , b. ( 1962 ) .
determination of gravitational standard time . in
_ recent developments in general relativity - a book dedicated to leopold infeld s 60th birthday
303306 ) .
new york : pergamon press .
prigogine , i. , & ordonez , g. ( 2006 ) .
acceleration and entropy : a macroscopic analogue of the twin paradox . in f.
f. orsucci and n. sala ( eds . ) , _ new research on chaos and complexity _
( pp . 520 ) .
new york : nova science publishers inc .. | on the basis of the woodhouse causal axiomatics , we show that conformal proper times and an extra variable in addition to those of space and time , precisely and physically identified from experimental examples , together give a physical justification for the ` _ _ chronometric hypothesis _ _ ' of general relativity . indeed
, we show that , with a lack of these latter two ingredients , no clock paradox solution exists in which the clock and message functions are solely at the origin of the asymmetry .
these proper times originate from a given conformal structure of the spacetime when ascribing different compatible projective structures to each woodhouse particle , and then , each defines a specific weylian sheaf structure .
in addition , the proper time parameterizations , as two point functions , can not be defined irrespective of the processes in the relative changes of physical characteristics .
these processes are included via path - dependent conformal scale factors , which act like sockets for any kind of physical interaction and also represent the values of the variable associated with the extra dimension .
as such , the differential aging differs far beyond the first and second clock effects in weyl geometries , with the latter finally appearing to not be suitable . |
a basic example of non discrete dynamics on the euclidean space is given by a linear vector field .
this linear system is infinitesimally hyperbolic if every eigenvalue has nonzero real part , and it has well known properties @xcite .
for instance , when the real part of its eigenvalues are negative ( hurwitz matrix ) , the origin is a global attractor rest point . in the nonlinear case
, there has been a great interest in the local study of vector fields around their rest points @xcite .
however , in order to describe a global phase - portrait , as in @xcite it is absolutely necessary to study its behavior in a neighborhood of infinity @xcite .
the asymptotic stability at infinity has been investigated with a strong influence of @xcite , where olech showed a connection between stability and injectivity ( see also @xcite ) .
this research was also studied in @xcite . in @xcite ,
gutierrez and teixeira study @xmath11vector fields @xmath12 , the linearizations of which satisfy and in an neighborhood of infinity . by using @xcite , they prove that if @xmath13 has a rest point and the index @xmath14 ( resp .
@xmath15 ) , then @xmath13 is topologically equivalent to @xmath16 that is the infinity is a repellor ( resp . to @xmath17 that is the infinity is an attractor ) .
this gutierrez - teixeira s paper was used to obtain the next theorem , where @xmath18 , and @xmath19 is the real part of @xmath20 . [
thm : gs03 ] let @xmath21 be a @xmath11 map , where @xmath22 and @xmath23 the following is satisfied : 1 . if for some @xmath24 @xmath25 then there exists @xmath26 such that the restriction @xmath27 is injective .
2 . if for some @xmath24 the spectrum @xmath28 is disjoint of the union @xmath29\cup\{z\in\mathbb{c}:\re(z)\geq0\}$ ] .
then there exist @xmath30 such that the point @xmath31 of the riemann sphere @xmath32 is either an attractor or a repellor of @xmath33 is given in @xcite , and it has been extended to differentiable maps @xmath34 in @xcite . in both papers the eigenvalues
also avoid a real open neighborhood of zero . in @xcite the author
examine the intrinsic relation between the asymptotic behavior of @xmath35 and the global injectivity of the local diffeomorphism given by @xmath2 .
he uses @xmath36 , where @xmath37 is the linear rotation of angle @xmath38 , and ( motivated by @xcite ) introduces the so called @xmath39condition @xcite , which claims : : : _ for each @xmath38 , there does not exist a sequence @xmath40 with @xmath41 such that @xmath42 and @xmath43 has a real eigenvalue @xmath44 satisfying @xmath45 .
_ by using this , @xcite improves the differentiable version of theorem [ thm : gs03 ] . in the present paper
we prove that the condition @xmath46 is enough in order to obtain theorem [ thm : gs03 ] for differentiable vector fields @xmath34 . throughout this paper , @xmath47 is embedded in the riemann sphere @xmath32 .
thus @xmath48 is the subspace of @xmath32 with the induced topology , and infinity refers to the point @xmath31 of @xmath32 .
moreover given @xmath49 , a closed ( compact , no boundary ) curve ( @xmath50manifold ) , @xmath51 ( respectively @xmath52 ) is the compact disc ( respectively open disc ) bounded by @xmath53 .
thus , the boundaries @xmath54 and @xmath55 are equal to @xmath53 , homeomorphic to @xmath56 .
for every @xmath22 let @xmath57 . outside this compact disk
we consider a differentiable vector field @xmath58 . as usual , a * trajectory of @xmath2 starting at @xmath59 * is defined as the integral curve determined by a maximal solution of the initial value problem @xmath60 .
this is a curve @xmath61 , satisfying : * @xmath62 varies on some open real interval containing the zero , the image of which @xmath63 ; * @xmath64 and there exist the real derivatives @xmath65 , @xmath66 ; * @xmath67 the velocity vector field of @xmath68 at @xmath69 equals @xmath70 and * @xmath71 is the maximal interval of definition .
we identify the trajectory @xmath68 with its image @xmath72 , and we denote by @xmath73 ( resp .
@xmath74 ) the positive ( resp .
negative ) semi - trajectory of @xmath2 , contained in @xmath68 and starting at @xmath75 . in this way
thus each trajectory has its two limit sets , @xmath77 and @xmath78 respectively .
these limit sets are well defined in the sense that they only depend on the respective solution .
a @xmath79vector field @xmath80 ( without rest points ) can be extended to a map @xmath81 ( which takes @xmath31 to @xmath82 ) @xcite . in this manner ,
all questions concerning the local theory of isolated rest points of @xmath2 can be formulated and examined in the case of the vector field @xmath83 .
for instance , if @xmath84 ( resp .
@xmath85 ) is an unbounded semi - trajectory of @xmath34 starting at @xmath86 with empty @xmath87limite ( resp .
@xmath88limit ) set , we say @xmath89 * goes to infinity * ( resp . @xmath85
* comes from infinity * ) , and it is denoted by @xmath90 ( resp .
@xmath91 ) .
therefore , we may also talk about the phase portrait of @xmath2 in a neighborhood of @xmath31 .
as in our paper @xcite , we say that the point at infinity @xmath31 of the riemann sphere @xmath32 is an * attractor * ( resp .
a * repellor * ) for the continuos vector field @xmath34 if : * there is a sequence of closed curves , transversal to @xmath2 and tending to infinity . that is for every @xmath92 there
exist a closed curve @xmath93 such that @xmath94 contains @xmath95 and @xmath93 has transversal contact to each small local integral curve of @xmath2 at any @xmath96 .
* for some @xmath97 with @xmath98 all trajectories @xmath99 starting at a point @xmath100 satisfy @xmath101 that is @xmath84 go to infinity ( resp . @xmath102
that is @xmath85 come from infinity ) .
we also recall that @xmath103 , the * index of @xmath2 at infinity * is the number of the extended line @xmath104 $ ] given by @xmath105 where @xmath106 is a global differentiable vector field such that : * in the complement of some disk @xmath107 with @xmath26 both @xmath2 and @xmath108 coincide . *
the map @xmath109 is lebesgue almost integrable in whole @xmath110 in the sense of @xcite .
this index is a well - defined number in @xmath104 $ ] , and it does not depend on the pair @xmath111 as shown @xcite . the differentiable vector field ( or map ) @xmath112 is called * hurwitz * if every eigenvalue of the jacobian matrices has negative real part .
this means that its spectrum @xmath113 satisfies @xmath3 .
we are now ready to state our result .
[ thm : main-1 ] let @xmath112 be a differentiable vector field ( or map ) , where @xmath22 and @xmath114 .
suppose that @xmath2 is hurwitz : @xmath115
. then 1 .
there are @xmath26 and a globally injective local homeomorphism @xmath116 such that @xmath117 and @xmath2 coincide on @xmath118 . moreover , the restriction @xmath119 is injective , and it admits a global differentiable extension @xmath108 such that the pair @xmath111 satisfies the definition of the index of @xmath2 at infinity , and this index @xmath120 .
2 . for all @xmath121
, there is a unique positive semi - trajectory of @xmath2 starting at @xmath122 .
moreover , for some @xmath123 , the point at infinity @xmath31 of the riemann sphere @xmath32 is an attractor ( respectively a repellor ) for the vector field @xmath124 as long as the well - defined index @xmath125 ( respectively @xmath126 ) .
the map @xmath127 of is not necessarily a homeomorphism .
this @xmath127 is a topological embedding , the image of which may be properly contained in @xmath47 .
furthermore , if @xmath128 is an arbitrary invertible linear map , then applies to the map @xmath129 . improves the main results of @xcite . complements the injectivity work of @xcite ( see also @xcite ) , where the authors consider the assumption @xmath130 ( as in ) . generalizes @xcite , where the authors utilize the second condition of . in our new assumptions
, the negative eigenvalues can tend to zero .
since the local inverse function theorem is true , a map @xmath131 as in is a local diffeomorphism .
thus the level curves @xmath132 make up a @xmath79foliation @xmath133 the leaves of which are differentiable curves , and the restriction of the other submersion @xmath134 to each of these leaves is strictly monotone . in particular , @xmath133 and @xmath135 are ( topologically ) transversal to each other .
we orient @xmath133 in agreement that if @xmath136 is an oriented leaf of @xmath133 thought the point @xmath122 , then the restriction @xmath137 is an increasing function in conformity with the orientation of @xmath136 .
we denote by @xmath138 ( resp .
@xmath139 ) and @xmath140 ( resp .
@xmath141 ) the respective positive and negative half - leaf of @xmath133 ( resp .
@xmath135 ) . thus @xmath142 and @xmath143 . in this context , the nonsingular vector fields @xmath144 given by the partial derivatives are tangent to @xmath136 and @xmath145 , respectively .
this construction has previously been used in @xcite .
we need the following definition @xcite .
let @xmath146 and consider the set @xmath147\times[0,2]:0<x+y\leq 2 \big\}.\ ] ] [ def : hrc ] let @xmath148 be a differentiable local homeomorphism .
given @xmath149 , we say that @xmath150 ( in the domain of @xmath2 ) is a * half - reeb component * for @xmath151 if there is a homeomorphism @xmath152 which is a topological equivalence between @xmath153 and @xmath154 such that : ( ) * the segment @xmath155 is sent by @xmath156 onto a transversal section for the foliation @xmath151 in the complement of the point @xmath157 ; this section is called the compact edge of @xmath150 . * both segments @xmath158 and @xmath159 are sent by @xmath156 onto full half - leaves of @xmath151 .
these half - leaves of @xmath151 are called the non compact edges of @xmath150 .
observe that @xmath150 may not be a closed subset of @xmath47 , and @xmath156 does not need to be extended to infinity .
gives new results on the foliations induced by a local diffeomorphism @xmath131 ( see also proposition [ prop:10 ] ) .
implies that the conditions @xmath160 give the existence of @xmath26 such that @xmath161 can be extended to an injective map @xmath116 .
presents some preliminary results on maps such that @xmath162 .
section [ sec:5 ] concludes with the proof of .
the main step is given in , which implies that hurwitz maps satisfy .
therefore , is obtained by using this and some previous work @xcite .
let @xmath131 be an orientation preserving local diffeomorphism , that is @xmath163 .
next subsection gives preparatory results about @xmath133 in order to obtain that @xmath2 will be injective on topological half planes ( see proposition [ prop:10 ] ) .
subsection [ subsec:2.3 ] presents a condition under which @xmath2 is injective at infinity , that is outside some compact set .
let @xmath164 be a closed curve surrounding the origin .
we say that the vector field @xmath112 has _ contact _
tangency with ; resp .
transversal to ; etc ) with @xmath53 at @xmath165 if for each small local integral curve of @xmath2 at @xmath122 has such property .
[ def : general - position ] a closed curve @xmath164 is in * general position * with @xmath133 if there exists a set @xmath166 , at most finite such that : * @xmath133 is transversal to @xmath167 .
* @xmath133 has a tangency with @xmath53 at every point of @xmath168 .
* a leaf of @xmath133 can meet tangentially @xmath53 at most at one point .
denote by @xmath169 the set of all closed curves @xmath164 in general
position with @xmath133 such that @xmath170 .
if @xmath164 is in general
position , we denote by @xmath171 ( resp .
@xmath172 ) the number of tangent points of @xmath133 with @xmath53 , which are external ( resp .
internal ) . here , external ( resp .
internal ) means the existence of a small open interval @xmath173 such that the intersection set @xmath174 and @xmath175 ( resp .
@xmath176 ) .
[ rem : a , b ] if @xmath164 is in general
position with @xmath177 there exist @xmath178 two different point such that @xmath179.$ ] moreover , @xmath180 and @xmath181 are external tangent points because the map @xmath182 is an orientation preserving local diffeomorphism .
since @xmath53 and @xmath183 are connected and @xmath53 is not contained in any leaf of @xmath133 , we conclude that both external tangencies are different .
corollary [ cor : ext ] gives important properties of the leafs passing trough a point in remark [ rem : a , b ] , if we select @xmath184 with the minimal number of internal tangencies . to this end
, the next lemma will be needed .
[ lem : external - tangency ] let @xmath184 .
suppose that a leaf @xmath185 of @xmath133 meets @xmath53 transversally somewhere and with an external tangency at a point @xmath186 .
then @xmath185 contains a closed subinterval @xmath187_f$ ] which meets @xmath53 exactly at @xmath188 ( doing it transversally at @xmath189 ) and the following is satisfied : * if @xmath187 $ ] is the closed subinterval of @xmath53 such that @xmath190\cup[p , r]_f$ ] bounds a compact disc @xmath191 contained in @xmath192 then points of @xmath193_f$ ] nearby @xmath122 do not belong to @xmath191 .
* let @xmath194 and @xmath195 $ ] be subintervals of @xmath53 satisfying @xmath187\subset(\tilde{p},\tilde{r } ) \subset [ \tilde{p},\tilde{r}]$ ] .
if @xmath196 and @xmath197 are close enough to @xmath122 and @xmath198 respectively ; then we may deform @xmath53 into @xmath199 in such a way that the deformation fixes @xmath200 and takes @xmath195\subset c$ ] to a closed subinterval @xmath195_1 \subset c_1 $ ] which is close to @xmath187_f$ ] .
furthermore , the number of generic tangencies of @xmath133 with @xmath201 is smaller than that of @xmath133 with @xmath53 .
we refer the reader to ( * ? ? ?
* lemma 2 ) .
[ cor : ext ] let @xmath202 suppose that @xmath53 minimizes @xmath203 and @xmath179,$ ] then @xmath204 and @xmath205 we only consider the case of the point @xmath180 .
assume by contradiction that the number of elements in @xmath206 is greater than one i.e @xmath207 .
the last condition of implies that the intersection of @xmath208 with the other point is transversal to @xmath53 .
then there is a disk as in statement ( a ) of . by using the second part of we
can avoid the external tangency @xmath180 and some internal tangency .
this is a contradiction because @xmath203 is minimal .
corollary [ cor : ext ] remains true if we take any external tangency ( not necessarily @xmath180 and @xmath181 ) in a closed curve @xmath209 with minimal @xmath210 .
a oriented leaf of @xmath133 whose distance to @xmath211 is different from zero has unbounded half leaves . given any @xmath136 with unbounded half leaves , we denote by @xmath212 and @xmath213 the two components of @xmath214 in order that @xmath215 be contained in @xmath216 therefore , the image @xmath217 is an open connected subset of the semi
plane @xmath218 [ rem:1 ] if the image @xmath217 is a vertical convex set , all the level curves @xmath219 are connected .
thus the restriction @xmath220 is an homeomorphism , and it sends every leaf of @xmath221 over vertical lines .
therefore , it is a topological equivalence between two foliations .
[ lem : non - convex.hrc ] if @xmath212 is disjoint from @xmath222 and the image @xmath217 is not a vertical convex set , then @xmath212 contains a half - reeb component of @xmath223 by remark [ rem:1 ] , some level set @xmath224 is disconnected . therefore , the result is obtained directly from ( * ? ? ?
* proposition 1 ) .
lemma [ lem : non - convex.hrc ] and remark [ rem:1 ] hold when we consider @xmath225 [ lem : def - eta ] recall that @xmath169 is the set of all closed curves @xmath164 in general
position with @xmath133 such that @xmath226 let @xmath227 be the function given by @xmath228 where @xmath229 minimizes the number of internal tangent points with @xmath223 the following statements hold :
the function @xmath230 is nondecreasing .
2 . if @xmath230 is bounded then , there exist @xmath231 such that @xmath232 for all @xmath233 3
. set @xmath234 $ ] .
suppose that @xmath133 has a half - reeb component @xmath150 whose image @xmath235 is disjoint from @xmath236 then such @xmath231 is not a maximum value of the function @xmath237
as @xmath238 also belongs to @xmath169 we have that @xmath239 .
therefore ( a ) is true . to prove the second part
, we introduce the set @xmath240 , for every @xmath26 . from this definition
it is not difficult to check that : @xmath230 is bounded if and only if @xmath241 . therefore , the first element of @xmath241 is the bound @xmath242 of statement ( b ) .
this proves the second statement .
we shall have established ( c ) if we prove that there is some @xmath243 with @xmath244 such that @xmath245 . to this end , we select @xmath246 large enough for which @xmath97 is enclosing @xmath247 where @xmath248 is the compact edge of @xmath150 .
since , this @xmath97 intersects both leaves @xmath136 and @xmath185 where @xmath122 and @xmath75 are the endpoints of @xmath248 , we obtain that @xmath210 is greater than @xmath249 .
therefore , for some @xmath244 there is @xmath243 such that @xmath245 .
this proves ( c ) .
[ prop:10 ] let @xmath131 be a map with @xmath250 consider @xmath97 and @xmath230 as in lemma [ lem : def - eta ] .
if @xmath251 satisfies that @xmath252 for all @xmath253 and @xmath234.$ ] then , for each @xmath254 at least one of the restrictions of @xmath2 to @xmath255 or @xmath256 is a globally injective map , in agrement that the domain of this restriction is in the complement of @xmath257 .
we only consider the case @xmath258 suppose that @xmath259 is contained in the complement of @xmath260 from remark [ rem:1 ] it is sufficient to prove that @xmath261 is vertical convex .
suppose by contradiction that it is false .
then lemma [
lem : non - convex.hrc ] implies that there is a half - reeb component @xmath262 by using statement ( c ) of lemma [ lem : def - eta ] this @xmath263 is not a maximum value of the function @xmath230 .
this contradiction with our selection of the circle @xmath251 conclude the proof .
the next theorem implies the injectivity at infinity of a map , and it is obtained by using the methods , ideas and arguments of @xcite .
we only give the proof , in the case of continuously differentiable maps .
[ teo:1 ] let @xmath264 be an differentiable local homeomorphism with @xmath265 .
suppose that @xmath162 , and each half reeb component of either @xmath133 or @xmath135 is bounded .
then there exist @xmath26 such that the restriction @xmath266 can be extended to a globally injective local homeomorphism @xmath267 .
we can apply the results of @xcite to the continuous vector field @xmath268 and obtain that for each closed curve @xmath184 the _ index of @xmath268 along @xmath53 , _ denoted by @xmath269 satisfies @xmath270 if @xmath268 is discontinuous , we proceed as in @xcite by using the index of the foliation @xmath133 which also satisfies this formulae . 1 .
we claim that @xmath271 for all @xmath272 suppose that ( a.1 ) is false , so there is @xmath273 whose @xmath274 thus , for some point in @xmath275 the hamiltonian @xmath276 is vertical .
more precisely we can obtain @xmath277 such that @xmath278 and @xmath279 this is a contradiction with the eigenvalue assumptions because @xmath280 .
( if @xmath268 is discontinuous , we refer the reader to @xcite where proves that the index of the foliation @xmath133 is zero )
. therefore , ( a.1 ) holds . 1
. we claim that if @xmath281 _ minimizes _
@xmath282 then every internal tangency in @xmath283 gives a half
reeb component . for every internal tangency @xmath284 we consider the forward poincar map @xmath285 induced by the oriended @xmath133 ( if @xmath286_{\sigma}\subset c_{\sigma}\to c_{\sigma}$ ] the proof is similar ) where @xmath287 is the maximal connected domain of definition of @xmath168 on which this first return map is continuous .
if the open arc @xmath288 intersects @xmath283 we apply lemma [ lem : external - tangency ] , so we can deform @xmath283 in a new circle @xmath289 such that the number of internal tangencies of @xmath275 with @xmath133 is ( strictly ) smaller than that of @xmath290 this is a contradiction .
therefore @xmath288 is disjoint from @xmath290 by using this and our selection of @xmath287 is not difficult to check that there is a half
reeb component of @xmath133 whose compact edge is contained in @xmath291 thus , we obtain ( a.2 ) .
notice that , for every circle as in ( a.2 ) any internal tangency of this @xmath292 gives an unbounded half reeb component , thus by our assumptions and ( a.1 ) we have that @xmath293 therefore , 1 .
if @xmath294 is as in ( a.2 ) then @xmath295 and @xmath296 moreover , @xmath177 restricted to @xmath297 is topologically equivalent to the foliation made up by all the vertical straight lines , on @xmath298 .
since @xmath2 has no unbounded half reeb component , we can use the last section of @xcite ( see proposition 5.1 ) and obtain that the closed curve @xmath283 of ( a.3 ) can be deformed so that , the resulting new circle @xmath53 has an exterior collar neighborhood @xmath299 such that : 1 .
@xmath300 is a non - trivial closed curve , @xmath301 is an exterior collar neighborhood of @xmath300 and the restriction @xmath302 is a homeomorphism . by schoenflies theorem @xcite the map @xmath303
can be extended to a homeomorphism @xmath304 .
we extend @xmath305 to @xmath306 by defining @xmath307 thus @xmath308 is a homeomorphism and @xmath10 ( resp .
@xmath301 ) is a exterior collar neighborhood of @xmath53 ( resp .
@xmath300 ) . consequently
, @xmath117 is a local homeomorphism and @xmath309 is topologically equivalent to the foliation made up by all the vertical straight lines .
the injectivity of @xmath9 follows from the fact that @xmath309 in trivial ( * ? ? ?
* proposition 1.4 ) .
this concludes the proof .
[ cor : ext with no hrc ] suppose that @xmath2 satisfies .
then the respective extension @xmath310 of @xmath2 is a globally injective local homeomorphism the foliations of which , @xmath311 and @xmath312 have no half - reeb components .
we reefer the reader to affirmation ( a.3 ) in the proof of .
suppose that @xmath131 is an orientation preserving local diffeomorphism .
then the foliation @xmath133 ( resp .
@xmath135 ) has at most countably many half
reeb components .
reeb component has a tangency with some @xmath313 of lemma [ lem : def - eta ] with @xmath314 large enough .
we conclude , since the closed curves has at most a finite number of tangent points . by using a smooth embedding @xmath315 the authors of (
* proposition 1 ) prove the existence of foliations @xmath133 which have infinitely many half - reeb components .
in this section we present some properties of a map the spectrum of which is disjoint of @xmath316 .
these results will be used in section [ sec:5 ] to proving the first part of theorem [ thm : main-1 ] . in this context , we consider @xmath133 and their trajectories @xmath317 , @xmath318 and @xmath319 .
+ [ lemma : disk1 ] let @xmath131 be a differentiable local homeomorphism .
suppose that the spectrum @xmath162 and @xmath320 .
then the intersection of @xmath321 with the vertical ray @xmath322 is the one point set @xmath323 .
assume , by contradiction , that @xmath324 intersects @xmath325 .
we take @xmath326 the smallest value such that @xmath327 ( see ) .
we only consider the case in which the compact arc @xmath328_f\subset{l_p^+}$ ] such that @xmath329_f)$ ] equals the interval @xmath330 $ ] with @xmath331 , where @xmath332 ( in the other case , @xmath329_f)=[a_0,a]$ ] , @xmath333 the argument is similar ) .
therefore , if we take the vertical segment @xmath328_a=\{(a , y):c\leq y\leq d\}$ ] joint to the open disk @xmath334 bounded by the closed curve @xmath335_f\cup[p , q]_a$ ] .
we meet two possible cases : _ the first one is that @xmath336 .
_ we select the point @xmath337_f$ ] in order that @xmath338 will be the smallest value of the compact set @xmath339_f$ ] .
let @xmath340 be the closed region bounded by the union of @xmath341 @xmath342 and @xmath343_f\subset[p , q]_f$ ] where @xmath344 .
as @xmath345_f\}$ ] the compact arc @xmath328_f$ ] is tangent to the vertical line @xmath346 at the point @xmath347 .
thus , @xmath348 is vertical , and so @xmath349 .
this implies that @xmath350 . by the assumptions about @xmath351 ,
@xmath352 which in turn implies that the arc @xmath353_f\subset[p , q]_f$ ] must enter into @xmath340 and can not cross the boundary of @xmath340 ( see ) .
this contradicts the fact that @xmath354 . _
the second case happens when @xmath355 .
_ as @xmath328_f$ ] is not contained in @xmath356 either @xmath357 or @xmath358 . if @xmath359 the vertical line @xmath360 meet @xmath328_f$ ] in two different points which define a closed curve as in such that it bounds an open disk disjoint of @xmath222 .
we conclude by using the proof of the first case .
if @xmath358 , we observe the continuous foliation @xmath133 in a neighborhood of @xmath328_f$ ] and meet two points @xmath361 and @xmath362 with @xmath363 , but @xmath364_f)\subset[\sigma,+\infty)$ ] .
it satisfies the conditions of the first case .
therefore the lemma is proved .
[ rem : disk1 ] lemma [ lemma : disk1 ] remains true , if we consider the negative leaf @xmath365 starting at @xmath366 joint to the vertical ray @xmath367 . a positive basis , title="fig : " ] + [ lemma : fluw - positive ] let @xmath131 be a map with @xmath162 .
consider @xmath368 and the projection @xmath332 . if the oriented compact arc @xmath328_f\subset{l_p^+}$ ] and its image @xmath329_f)$ ] is the interval @xmath369\subset(\sigma,+\infty)$ ] with @xmath370 . then @xmath371_f}}\langle x , \nabla_f\rangle dt \geq f(p ) \int_{[p ,
q]_f}f_x ~ dt+ g(p)\big(\pi(q)-\pi(p)\big),\ ] ] where @xmath372 denotes the usual inner product on the plane , and @xmath373 is the first partial derivative . for each @xmath374_f)=[a , b]$ ] ,
the vertical line @xmath375 intersects @xmath328_f$ ] in a non - empty compact set .
so there exist @xmath376_f\}$ ] and @xmath377 .
we also define @xmath378 as the set of critical values of the projection @xmath332 restricted to the differentiable arc @xmath328_f$ ] . by the sard s theorem , presented in ( * ? ? ?
* theorem 3.3 ) this set @xmath379 is closed and has zero lebesgue measure . for @xmath380
the set @xmath375 intersects @xmath328_f$ ] in at most finitely many points and the complement set @xmath381_f\big)\setminus\ { m_{\alpha}\}$ ] is either empty or its cardinality is odd . by lemma [ lemma : disk1 ] , the order of these points in the line @xmath375 oriented oppositely to the @xmath382axis coincides with that on the oriented arc @xmath328_f$ ] ( a behavior as in does not exist ) .
therefore , for every @xmath380 the set @xmath381_f\big)\setminus\ { m_{\alpha}\}$ ] splits into pairs @xmath383 with the following three properties : * ( a.1 ) * @xmath384 , * ( a.2 ) * the compact arc @xmath385_f$ ] lies in the semi - plane @xmath386 and it is oriented from @xmath387 to @xmath388 * ( a.3 ) * @xmath389 .
notice that the tangent vector of @xmath328_f$ ] at @xmath387 has a negative @xmath390component : @xmath391 , and the respective tangent vector at @xmath388 satisfies @xmath392 .
similarly , @xmath393 ( see ) .
assertion : : _ take @xmath394_f\}$ ] and @xmath369_c=\{(\alpha , c):a\leq\alpha \leq b\}$ ] , an horizontal segment . consider the compact set @xmath395 the boundary of which contain @xmath328_f\cup[\pi(p),\pi(q)]_c$ ] ( see ) .
suppose that @xmath53 , the boundary of @xmath51 is negatively oriented ( clock wise )
. then @xmath396_f}\langle f , \nabla_f\rangle dt- { \int_{a}^b g(\alpha , c)d{{\alpha}}},\]]where @xmath397 . _ _ proof of assertion _ we will use the _ green s formulae _ given in ( * ? ? ?
* corollary 5.7 ) ( see also @xcite ) with the differentiable map @xmath398 and the outer normal vector of @xmath53 denoted by @xmath399 ( unitary ) . by using that @xmath400
it follows that @xmath401 where @xmath402 denotes the arc length element .
if @xmath403_c$ ] denote @xmath369_c$ ] oriented from @xmath404 to @xmath405 , then @xmath335_f\cup b\cup[\pi(q),\pi(p)]_c\cup a$ ] with @xmath406 and @xmath407 two oriented vertical segments .
consequently , @xmath408_f}}\langle g,-\eta\rangle ds + { \int_{b}}\langle g,-\eta \rangle ds + { \int_{[\pi(q),\pi(p)]_c}}\langle g,-\eta \rangle ds + { \int_{a}}\langle g,-\eta \rangle ds.\ ] ] in @xmath409 the vector @xmath410 is horizontal i.e @xmath411 then @xmath412 implies that @xmath413 . therefore @xmath414_f}}\langle g,-\eta\rangle ds + { \int_{[\pi(q),\pi(p)]_c}}\langle g,-\eta \rangle ds.\]]in @xmath328_f$ ] , the outer normal vector es parallel to @xmath415 .
then , for all @xmath416_f$ ] we obtain that @xmath417 and @xmath418 because @xmath328_f\subset\{f = f(p)\}$ ] . thus @xmath371_f}}\langle g,-\eta \rangle ds={\int_{[p , q]_f}}\langle f,\nabla_f \rangle dt.\ ] ] similarly , in @xmath403_c$ ] we have @xmath419 thus @xmath420_c}}\langle g,-\eta \rangle ds= { \int_{b}^a g(\alpha , c)d{{\alpha}}}= -{\int_{a}^b g(\alpha , c)d{{\alpha}}}.\ ] ] therefore , and prove the assertion .
@xmath421 in order to conclude the proof of this lemma we consider @xmath51 as in the last assertion joint to the construction of its precedent paragraph .
since the complement of @xmath379 is a total measure set , @xmath422}d\alpha.\ ] ] thus , the formulae in the assertion implies that @xmath423_f}\langle f , \nabla_f\rangle dt= { \int_a^b g(m_{\alpha})d\alpha}+ { \int_{\alpha\not\in{s}}}{\sum_{\pi(p_{\alpha})=\alpha}\big[g(q_{\alpha})-g(p_{\alpha})\big]}d\alpha.\ ] ] but , @xmath424_f}\langle f , \nabla_f\rangle dt=\int_{[p , q]_f}\langle x , \nabla_f\rangle dt - f(p)\int_{[p , q]_f}f_x dt\ ] ] and the property ( a.3 ) of the precedent paragraph to assertion 1 shows that @xmath425 .
therefore , since @xmath426 , implies that @xmath424_f}\langle x , \nabla_f\rangle dt - f(p)\int_{[p , q]_f}f_x dt\geq { \int_a^b g(m_{\alpha})d\alpha}\geq g(p)\big(b - a\big),\ ] ] and concludes this proof . by applying the methods of the last proof
give us the next : [ lem : fluw - negative ] let @xmath131 be a map with @xmath162 .
consider @xmath427 and @xmath332 .
if the oriented compact arc @xmath328_f\subset{l_q^-}$ ] , and @xmath329_f)=[\pi(q),\pi(p)]\subset(\sigma,+\infty)$ ] with @xmath428 .
then @xmath371_f}}\langle x , \nabla_f\rangle dt \geq f(q ) \int_{[p , q]_f}f_x dt+ g(p)\big(\pi(p)-\pi(q)\big).\ ] ] as @xmath429 we again consider the null set @xmath430_f)$ ] given by the critical values of @xmath431 restricted to @xmath328_f$ ] ( see @xcite ) .
similarly , for every @xmath432_f)=[a , b]$ ] we define @xmath376_f\}$ ] and @xmath377 .
therefore , remark [ rem : disk1 ] shows that for each @xmath433 the finite set @xmath381_f\big)\setminus\{\tilde{m}_{\alpha}\}$ ] splits into pairs @xmath383 satisfying : * ( i ) * @xmath384 , * ( ii ) * the oriented arc @xmath385_f\subset\{x\geq\alpha\}$ ] , and * ( iii ) * @xmath389 .
take @xmath51 the boundary of which is the closed curve @xmath434_f\cup\pi^{-1}(a)\cup[\pi(p),\pi(q)]_c$ ] , where @xmath369_c=\{(\alpha , c):a\leq\alpha\leq b\}$ ] for some @xmath435_f\}$ ] . by using the green s formulae with the map @xmath436 and the compact disk @xmath51 we have that @xmath396_f}\langle \tilde{f } , \nabla_f\rangle dt- { \int_{a}^b g(\alpha , c)d{{\alpha}}},\]]where @xmath437 .
since @xmath438}d\alpha\]]and @xmath439_f}\langle \tilde{f } , \nabla_f\rangle dt = { \int_{[p , q]_f}}\langle x , \nabla_f\rangle dt - f(q ) \int_{[p , q]_f}f_x dt,\ ] ] the last property * ( iii ) * implies @xmath371_f}}\langle x , \nabla_f\rangle dt - f(q ) \int_{[p , q]_f}f_x dt \geq{\int_a^b } g(m_{\alpha } ) d\alpha.\ ] ] this concludes the proof because @xmath426 shows @xmath440 .
this section concludes with the proof of the main theorem
. the essential goal of the next proposition is to prove the fact that our eigenvalue assumption ensures the non - existence of unbounded half reeb components .
it is obtained by using the preparatory results of the previous section . with this fact
is just obtained by applying our previous papers @xcite . 1
. any half - reeb component of either @xmath133 or @xmath135 is a bounded subset of @xmath47 .
2 . there are @xmath26 and a globally injective local homeomorphism @xmath267 such that @xmath117 and @xmath2 coincide on @xmath118 .
moreover , @xmath133 and @xmath135 have no half - reeb components . in the proof of ( i ) , we only consider one case .
suppose by contradiction that the foliation , given by the level curves has an unbounded half reeb component . by (
* proposition 1.5 ) , there exists a half
reeb component of @xmath133 the projection of which is an interval of infinite length .
thus , 1 .
there are @xmath441 and @xmath150 , a half
reeb component of @xmath133 such that @xmath442 , and the vertical line @xmath443 intersects ( transversally ) both non compact edges of @xmath150 . here
. a half - reeb component of a fixed foliation contain properly other half - reeb components , and they are topologically equivalent . in this context , the component is stable under perturbations on their compact face as long as the perturbed arc is also a compact face of some component .
therefore , without lost of generality , we may assume that nearby its endpoints , the compact edge of is made up of arcs of @xmath135 .
in this way , there exist @xmath444 and an injective , continuous curve @xmath445 such that 1 .
@xmath446)$ ] is a compact edge of @xmath150 such that both non compact edges @xmath447 and @xmath448 are contained in a half
plane @xmath449 , for some @xmath450 .
the images @xmath451 and @xmath452 are contained in some leaves of @xmath135 such that @xmath453 .
3 . for some @xmath454 there exists a orientation reversing injective function @xmath455\to(1-\tilde{a},1+\tilde{a})$ ] with @xmath456 such that @xmath457 .
furthermore , if @xmath458 $ ] then @xmath459 and there exists an oriented compact arc of trajectory @xmath460_f\subset\mathcal{a}$ ] of @xmath133 , connecting @xmath461 with @xmath462 .
4 . for some @xmath454 ,
small enough if @xmath463 $ ] and @xmath464 then there exists @xmath465 such that @xmath466 belongs to the open arc @xmath467_f$ ] . since @xmath150 is the union of an increasing sequence of compact sets bounded by the compact edge and a compact segment of leaf .
then , from our selection of the compact edge we have that for every @xmath474\}$ ] , large enough there exists @xmath475 $ ] as in ( b.4 ) such that the compact arc @xmath476_f$ ] projects over @xmath477 $ ] , meeting @xmath478 ( ) . more precisely , @xmath479_f\big\}.\ ] ] this defines a closed curve @xmath480 contained in @xmath481_f$ ] .
if @xmath482_{a_{s_{\alpha}}}\subset\pi^{-1}(a_{s_{\alpha}})$ ] is the vertical segment connecting @xmath483 , of ( b.4 ) with @xmath484 , then this clock wise oriented curve satisfies @xmath485_{a_{s_{\alpha}}}\cup[p_{s_{\alpha}},q_{s_{\alpha}}]_{f}\quad \mbox{and } \quad \pi(\gamma^-_{\alpha})=[a_{s_{\alpha}},\alpha],\ ] ] where the oriented compact arc @xmath486_{f}\subset[\gamma(s_{\alpha}),\gamma(\varphi(s_{\alpha}))]_f$ ] .
we select and fix @xmath487 , by using and , respectively we obtain @xmath488_f}}\langle x , \nabla_f\rangle dt \geq f(p_{s_{\alpha } } ) \int_{[p_{s_{\alpha}},m_{\alpha}]_f}f_x dt+ g(p_{s_{\alpha}})(x - a_{s_{\alpha}}),\ ] ] and @xmath489_f}}\langle x , \nabla_f\rangle dt \geq f(q_{s_{\alpha } } ) \int_{[m_{\alpha},q_{s_{\alpha}}]_f}f_x dt+ g(m_{\alpha})(x - a_{s_{\alpha}}).\ ] ] therefore , by adding we conclude @xmath490_f}}\langle x , \nabla_f\rangle dt \geq f(p_{s_{\alpha } } ) \int_{[p_{s_{\alpha}},q_{s_{\alpha}}]_f}f_x dt+ g(p_{s_{\alpha}})(x - a_{s_{\alpha}})\ ] ] because @xmath134 in increasing along @xmath486_f\subset\{f = f(p_{s_{\alpha}})\}$ ] and @xmath491 . 1 .
we claim that , the closed curves @xmath492 , given in define the following functions @xmath493_{a_{s_{\alpha}}}}}f dt \right| \quad \mbox{and } \quad \alpha\mapsto\left|{\int_{[q_{s_{\alpha}},p_{s_{\alpha}}]_f}}f_x dt \right|,\ ] ] they are bounded , when @xmath478 varies in some interval of infinite length contained on @xmath494 . furthermore , @xmath495_f}}f_x = 0.\ ] ] in fact , by a perturbation in the compact face if it is necessary , it is not difficult to prove that there is some half - reeb component @xmath496 of @xmath133 such that @xmath497 , their boundaries satisfy @xmath498 and @xmath499_{s_{\alpha}}$ ] , for all @xmath500 . since the image @xmath501 , the function @xmath502 is bounded in the closure of @xmath496 , which contain the compact set @xmath503_{a_{s_{\alpha}}}}$ ] .
consequently , @xmath493_{a_{s_{\alpha}}}}}f dt \right|\ ] ] is bounded , in some interval of infinite length contained on @xmath494 . in order to prove the second part , we apply the green s formulae to the map @xmath504 in the compact disk @xmath505 .
since the trace is cero , we obtain @xmath506_f}}f_x dt\right| = \mbox{arc length of } [ q_{s_{\alpha}},p_{s_{\alpha}}]_{a_{s_x}}.\ ] ] by compactness we conclude . the last part is directly obtained by using and @xmath484 because the continuity of the foliation @xmath133 and ( d ) imply that @xmath507 therefore ( e.1 ) holds . in the compact arc @xmath486_f\subset\gamma_{\alpha}^-$ ] , the vector @xmath510 is parallel to @xmath511 .
thus and ( e.1 ) imply that @xmath512 where the constant @xmath406 is independent of @xmath478 and @xmath513\}$ ] . since ( d ) and ( b.2 ) imply that @xmath514 , it is not difficult to obtain ( e.2 ) .
by ( e.2 ) we select some @xmath515 such that @xmath516 satisfies @xmath517 . by using the green s formulae with the map @xmath148
, the assumptions over the eigenvalues i.e @xmath518 imply that @xmath519 this contradiction concludes the proof of part ( i ) . in order to obtain ( ii ) , we apply because @xmath2 satisfies its conditions .
this gives the existence of the pair @xmath520 with @xmath26 and @xmath521 a globally injective local homeomorphism such that @xmath117 and @xmath2 coincide on @xmath118 .
furthermore , the last property in ( ii ) is obtained as a direct application of .
therefore , this proposition holds . by
, there exists a globally injective local homeomorphism @xmath116 such that @xmath117 and @xmath2 coincide on some @xmath522 , with @xmath523 .
in particular , the restriction @xmath524 is injective . in order to shown
the existence of the differentiable extension , consider @xmath525 joint to the globally injective map @xmath526 . in this context
, we can apply the arguments of ( * ? ? ?
* theorem 11 ) .
thus there are @xmath527 and a global differentiable vector field @xmath12 such that thus , there exist the index of @xmath2 at infinity , @xmath533 . therefore , @xmath534 is the global differentiable extension of @xmath535 and the pair @xmath536 satisfies the definition of the index of @xmath2 at infinity .
this concludes the proof of ( i ) .
an immediate consequence of ( i ) is that : if @xmath2 is hurwitz , then outside a larger disk both @xmath2 and @xmath13 have no rest points .
in addition , by ( a.1 ) the hurwitz vector field @xmath13 has no periodic trajectory @xmath541 with @xmath542 contained in @xmath543 .
as @xmath544 by green s formulae @xmath13 admits at most one periodic trajectory , say @xmath545 such that @xmath546 .
consequently 1 . for every @xmath549 there exist a closed curve @xmath93 transversal to @xmath13 contained in the regular set @xmath550 .
in particular , @xmath94 contains @xmath95 and @xmath93 has transversal contact to each small local integral curve of @xmath13 at any @xmath96 . 1 .
the point at infinity of the riemann sphere @xmath32 is either an attractor or a repellor of @xmath124 .
more specifically , if @xmath126 ( respectively @xmath125 ) , then @xmath31 is a repellor ( respectively an attractor ) of the vector field @xmath7 .
_ topics on singularities and bifurcations of vector fields .
normal forms , bifurcations and finiteness problems in differential equations _
nato sci . ser .
* 137 * kluwer acad .
dordrecht , 2004 .
_ global injectivity of @xmath551 maps of the real plane , inseparable leaves and the palais smale condition .
math . bull . * 50 * ( 2007 ) 377389 . _ a remark on an eigenvalue condition for the global injectivity of differentiable maps of @xmath47 , _ discrete contin .
* 17 * ( 2007 ) 397402 . | let @xmath0 be a differentiable vector field .
set @xmath1 .
this @xmath2 is called hurwitz if @xmath3 .
suppose that @xmath2 is hurwitz and @xmath4 is the complement of a compact set .
then by adding to @xmath2 a constant @xmath5 one obtains that the infinity is either an attractor or a repellor for @xmath6 that means : exists an unbounded sequence of closed curves , pairwise bounding an annulus the boundary of which is transversal to @xmath7 , and is a neighborhood of infinity with unbounded trajectories , free of singularities and periodic trajectories of @xmath6 this result is obtained after to proving the existence of @xmath8 a topological embedding such that @xmath9 equals @xmath2 in the complement of some compact subset of @xmath10 .
injectivity , reeb component , asymptotic stability primary : 37e35 , 37c10 ; secondary : 26b10 , 58c25 |
study of qed processes in a strong external field is stimulated essentially by existence of very strong magnetic fields in nature .
there is a series of evidences for the existence of neutron stars possessing magnetic fields close the critical field strength @xmath3
g. the rotating magnetic dipole model , in which the pulsar loses rotational energy through the magnetic dipole radiation , was confirmed with the discovery that the spin - down power predicted for many pulsars is in a quite good energetic agreement with the observed radiation , and gives magnetic fields @xmath4 g @xcite .
there are presently around 1600 spin - powered radio pulsars .
another class of neutron stars was discovered at x - ray and @xmath5-ray energies and may possess even stronger surface magnetic fields @xmath6 g. such stars are now referred to as magnetars @xcite , since they most probably derive their power from their magnetic fields rather than from spin - down energy loss .
two different classes of objects are thought to be magnetars : the soft gamma - ray repeaters ( sgr ) , discovered as sources of short bursts of hard x - rays ( @xmath7 kev ) with super - eddington luminosity and the anomalous x - ray pulsars ( aps ) , discovered as persistent , soft ( @xmath8 kev ) x- ray sources with pulsations of several seconds and spinning - down on time scales of @xmath9 years .
an increasing number of common properties , pointing to a close relationship of these two apparently different classes of objects and leading to their interpretation as magnetars , has been found ( see e.g.@xcite ) .
pair creation by a photon in a strong magnetic field is the basic qed reaction which can play the significant role in processes in vicinity of a neutron star .
investigation of pair creation by a photon in a strong magnetic field was started in 1952 independently by klepikov and toll @xcite . in klepikov
s paper @xcite , which was based on the solution of the dirac equation in a constant and homogeneous magnetic field , the probabilities of radiation from an electron and @xmath10 pair creation by a photon were obtained for magnetic field of arbitrary strength on the mass shell and the metric @xmath11 ( @xmath12 is the 4-momentum of photon ) . in 1971
adler @xcite calculated the photon polarization operator in the mentioned magnetic field using the proper - time technique developed by schwinger @xcite and batalin and shabad @xcite calculated the photon polarization operator in a constant and homogeneous electromagnetic field for @xmath13 using the green function in this field found by schwinger @xcite . in 1975
strakhovenko and present authors calculated the contribution of a charged - particles loop with @xmath14 external photon lines having applied the proper - time method in a constant and homogeneous electromagnetic field @xcite .
the explicit expression for the photon polarization operator was found . in a purely magnetic field
( * e*=0 ) the imaginary part of the polarization operator , which determines the probability of the pair creation in general case ( @xmath13 ) , was analyzed in detail near the root singularities originating at each threshold of @xmath10 pair creation in the state with the given landau quantum numbers .
this peculiarity of the pair creation process was observed by klepikov @xcite .
shabad @xcite performed an extensive analysis of the photon polarization operator found in @xcite in a purely magnetic field .
it was shown that its imaginary part in the limit @xmath15 reduces to the result of klepikov s calculation @xcite .
daugherty and harding @xcite analyzed in detail the total probabilities ( the attenuation coefficients ) of the pair creation by a polarized photon in the superstrong field @xmath16 basing on the calculations of toll and klepikov and paying the special attention to the threshold region @xmath17 , where the probabilities exhibit a `` saw - tooth '' pattern due to the mentioned root singularities . for high landau levels
klepikov @xcite ( and toll ) found the quasiclassical representations for the probabilities of radiation and pair creation in the case @xmath1 . in 1968
the present authors developed the operator quasiclassical method for the consideration of electromagnetic processes in an external field @xcite which is valid also in nonuniform and ( or ) depending on time field under conditions : 1 ) @xmath1 , 2 ) the charged particles have relativistic energies . both the radiation and pair creation by a photon were considered .
the method is given in detail in @xcite .
tsai and erber @xcite deduced the quasiclassical probability of pair creation using the photon polarization operator in a magnetic field calculated by adler @xcite . in the sec.2
the exact probability of the pair creation by a photon was obtained for general case @xmath18 starting from the polarization operator in a magnetic field .
the technical details of the derivation are given in appendix a. in sec.3 the standard quasiclassical approximation ( sqa ) is outlined in the limit @xmath19 , where @xmath20 is the photon energy .
the quasiclassical characteristics of the process depend on the parameter @xmath21 .
the correction to sqa is found .
details of calculation of the correction starting from the polarization operator in a magnetic field is given in appendix c. if the photon energy is not high , so that @xmath22 and sqa is not applicable , the new quasiclassical approach ( at low energy near threshold ( th ) ) is developed in appendix b. one can expect that the photon energy distribution in a vicinity of neutron star is wide and smooth . at averaging over this distribution
the `` saw - tooth '' pattern will be washed out .
it is very important that the result the averaging is very close to sqa even at @xmath23 .
the corresponding analysis has been carried out in sec.3 .
the energy spectrum of created pair is discussed in sec.4 . here
sqa is very useful also .
a remark concerning derivation of the energy spectrum in sqa is given in appendix d.
our analysis is based on the expression for the polarization operator obtained in @xcite , see eqs.(3.19 ) , ( 3.33 ) . for pure magnetic field ( * e*=0 ) this polarization operation can be written in the diagonal form @xmath24 where @xmath25 where the axis 3 directed along the magnetic field * b * , @xmath26 , @xmath27 , @xmath20 is the photon energy , @xmath28 is the tensor of electromagnetic field , @xmath29 is the dual tensor of electromagnetic field , @xmath30 here @xmath31 where @xmath32 let us note that the positive @xmath33 axis in the integral eq.([d2 ] ) is turned slightly down , and in the function @xmath34 in the integral over @xmath33 the subtraction at @xmath35 is implied .
the real part of the polarization operator defines the dispersive properties of the space region with magnetic field . at @xmath36
one can turn the integration contour over @xmath33 to the negative imaginary axis ( @xmath37 ) so that @xmath38 the functions @xmath39 in eq.([d5 ] ) are real .
these functions , studied in @xcite at @xmath40 and arbitrary value of @xmath41 , define the index of refraction @xmath42 . on mass shell @xmath43
one has @xmath44 for the mode 2 the photon polarization vector @xmath45 is perpendicular to the plane formed by the vectors @xmath46 and @xmath47 and for mode 3 the photon polarization vector @xmath48 lies in this plane ( @xmath49 and @xmath50 in adler s notation @xcite ) . at
@xmath51 the polarization operator eigenvalue @xmath52 acquires an imaginary part , which determines the probability of @xmath10 pair creation per unit length @xmath53 by a photon with a given polarization @xmath54 details of calculation basing on eqs.([d1]),([d2 ] ) are given in appendix a. using @xmath55 we get the explicit expressions for @xmath56 at @xmath13 : @xmath57 where @xmath58 is the kronecker symbol , @xmath59 is the heaviside function : @xmath60 for @xmath61 , @xmath62 for @xmath63 , @xmath64,\quad d(r)=\frac{2(r-\sqrt{r})}{\mu}. \label{d8}\end{aligned}\ ] ] here @xmath65 $ ] is the integer part of @xmath66 , @xmath67 where @xmath68 is the generalized laguerre polynomial , @xmath69 is defined eq.([d4 ] ) .
obtained probabilities @xmath53 agree with found by shabad @xcite but presented in much more compact form . on the mass shell ( @xmath43 ) in eqs.([d7])-([d9 ] ) one has to put @xmath70 . in this case @xmath71 and only two polarizations @xmath72 and @xmath73 remain .
the probability of pair creation averaged over the photon polarizations acquires especially simple form ( in the frame where @xmath74 ) @xmath75 \label{d10}\end{aligned}\ ] ] the probabilities of pair creation by a photon eqs.([d7 ] ) , ( [ d10 ] ) contain the factor @xmath76 .
the function @xmath77 , if @xmath78 where @xmath79 is the energy of charged particle on the landau energy level .
so the probability diverges when the electron and positron are created on the landau levels with electron and positron momentum @xmath80 , see e.g. @xcite .
the origin the singularity is due the properties of the space volume in the lowest order of perturbation theory ( infinitesimally narrow level ) .
the reason why this singularity is not a pole but a branch point is that motion along a field is not quantized .
in the case @xmath81 the mentioned above divergences are situated very often : @xmath82 . on the photon energy scale
the characteristic distance between the mentioned peaks will be @xmath83 , where @xmath84 is the frequency of motion of created particle in a magnetic field .
this frequency defines the distance between energy levels in the quasiclassical approximation .
so the resonance properties of pair creation probability become apparent if the width of energy levels and the effective spread of photon energy are small comparing with @xmath84 value . in the opposite case
the spectrum peculiarities will be washed out and the process probability will be quite adequate described by approximate expressions ultimately connected with quasiclassical character of created particle motion at large landau quantum numbers .
when the magnetic field is weak comparing with critical one ( @xmath85 ) and for high energy photons ( @xmath86 ) the quasiclassical approximation was created in @xcite , where the process probabilities were derived using the different approaches . in @xcite the asymptotic of eq.([d10 ] ) were obtained .
the approach in @xcite was developed for @xmath87 and the particle motion was considered to be quasiclassical ( and relativistic ) from the very beginning ; it is applicable for nonuniform ( and non - stationary ) fields .
the pair creation probabilities in the form eq.([c6 ] ) ( see appendix c ) were found in @xcite . below we will call it the standard quasiclassical approximation ( sqa ) . in appendix
c the corrections to sqa were calculated . for probability of pair creation by unpolarized photon
the correction is @xmath88\frac{dv}{1-v^2 } , \label{d12}\ ] ] where @xmath89 at @xmath90 the main contribution is given by the term @xmath91 , it is @xmath92 it is seen that at @xmath90 the sqa is applicable if @xmath93 . at @xmath94 sqa
becomes inapplicable . in this region one can use the process probability found in appendix b , which is valid if @xmath95 . using eq.([b5 ] )
we get for the probability of pair creation by unpolarized photon @xmath96 where @xmath97 for large @xmath86 , limited by the condition @xmath98 , we have @xmath99 at @xmath100 eq.([d17 ] ) agrees with eq.([d14 ] ) .
the last expression has essentially wider region of applicability than pure @xmath101 .
the behavior of pair creation probability for unpolarized photon at @xmath102 as a function of @xmath103 is illustrated in fig.1 .
the probability in saw - tooth form is calculated according to eq.([d10 ] ) .
it is seen that the characteristic distance between peaks is of the order @xmath104 .
the result of averaging over this interval @xmath105 is given by the dashed line which is smoothed out all peculiarities of original behavior and which is in a very good agreement with the thick curve calculated according to eq.([d15 ] ) . at @xmath106 eq.([d15 ] ) becomes inapplicable .
however in this case sqa is valid . from eq.([c6 ] ) we have the probability of pair creation by unpolarized photon @xmath107 where @xmath69 is defined in eq.([d13 ] ) .
the correction to this probability is given by eq.([d12 ] ) , its relative value at @xmath106 is @xmath108 .
the ratio @xmath109 and @xmath101 is shown in fig.2 .
the curves 2 and 3 are for polarized photons , the thick curves is for unpolarized photon ( see eq.([d15 ] ) and eq.([d18 ] ) ) .
the point where the ratio attains 1 is the boundary of the applicability region of the corresponding approximation .
the relative contribution of the correction eq.([d12 ] ) to sqa at @xmath110 is shown in fig.3 .
the upper curve is the probability of pair creation by a photon in sqa eq.([d18 ] ) , the lower curve is the sum @xmath111 ( the correction is negative at low @xmath103 ) .
it is seen that the correction is small in the region of sqa applicability @xmath112 .
it is very important that the quasiclassical approximation appears to be applicable also for super - strong magnetic field @xmath23 .
this is true if the characteristic parameter @xmath113 .
it is assumed in this case that parameter @xmath114 is large @xmath115 . in the limit @xmath116 the asymptotic expansion of the probability eq.([d18 ] )
is @xmath117 the asymptotic expansion of correction eq.([d12 ] ) is @xmath118 the relative value of the correction defines the boundary of region where sqa is valid for the super - strong magnetic field @xmath23 @xmath119 it is seen from this expression that at @xmath23 the applicability of quasiclassical approximation is controlled by the parameter @xmath120 .
so , both in the case @xmath87 and in the case @xmath23 the quasiclassical approximation is valid when the particles of pair are created in states with large landau quantum numbers .
it should be noted concerning eq.([d21 ] ) the smallness of numerical factor and relatively large contribution of the second term @xmath121 .
this means that the correction is small even at low @xmath122 , and one have to use the exact expression eq.([d12 ] ) . under these circumstances
one can use eq.([d18 ] ) with the correction eq.([d12 ] ) starting from @xmath123 .
this is illustrated in fig.4 where the probability of pair creation is shown at @xmath110 .
the probability in saw - tooth form is calculated according to eq.([d10 ] ) .
it is seen that the characteristic distance between peaks is of the order @xmath41 ( here @xmath110 ) just as in the case shown in fig.1 where the probability for @xmath104 was presented .
the result of averaging over this interval @xmath124 is given by the dashed line which is smoothed out all peculiarities of original behavior and which is in a very good agreement with the thick curve which is quasiclassical approximation calculated according to eqs.([d18 ] ) and ( [ d12 ] ) .
the energy distribution of created pair is symmetric with respect to exchange of electron and positron , i.e. it is symmetric respect energy @xmath125 and one can consider this distribution in interval @xmath126 . in the frame where @xmath74
the energy of electron and positron is @xmath127 from eq.([s1 ] ) follows that @xmath128 is defined in eq.([d8 ] ) ) . in the plane
@xmath129 the allowed states of electron and positron are bounded by the axes and the curve @xmath130 which is parabola with the nearest point to origin @xmath131 $ ] . when @xmath132 the maximal value @xmath133 is defined in eq.([d8 ] ) ) .
substituting @xmath134 in @xmath135 we get @xmath136 so the distribution over energy of created particle is of the discrete character . the probability @xmath137 is defined by eq.([d10 ] ) , where one has to perform summation at fixed @xmath138 .
the expression for spectrum contains the factor @xmath76 discussed above ( see eq.([d11 ] ) ) .
the energy distribution over electron energy in sqa ( see , e.g. @xcite , eq.(3.50 ) ) is @xmath139 , \label{s3}\ ] ] where @xmath140 at low values of the parameters @xmath41 and @xmath114 the energy distribution of pair creation probability has a sharp peak when @xmath141 ( for details see sec.3.2 in @xcite ) . when @xmath87 and @xmath106 the sqa is not valid on an edge of the energy spectrum where the created particle is no more relativistic ( @xmath142 ) . however at smaller energy ( @xmath143 ) ) the differential probability is suppressed exponentially , so the contribution of the interval @xmath142 , where sqa is not valid , is negligible small . for the super - strong field @xmath23 in the region of applicability of the quasiclassical approximation ( @xmath144 ) the parameter @xmath145 .
in this case the upper limit of electron energy @xmath146 is lower ( or of the same order ) than the electron energy in the maximum of the energy distribution of eq.([s3 ] ) ( @xmath147 ) .
the region of applicability of the quasiclassical approximation near @xmath148 is defined by the quasiclassical character of positron ( electron ) transverse motion ( @xmath149 ) @xmath150 where @xmath151 eq.([d8 ] ) .
in fig.5 the spectrum of created particles is shown for @xmath152 .
the spectrum in saw - tooth form is calculated according to eq.([d10 ] ) .
the dashed curve is the spectrum averaged over the interval @xmath153 .
the thick curve calculated using eq.([s3 ] ) .
it is seen that this curve is in a quite good agreement with the result of averaging . in this case
the argument @xmath154 in eq.([s3 ] ) and one can perform the expansion ( @xmath155 ) @xmath156 , \label{s6}\ ] ] where @xmath157 in fig.6 the energy distribution of the created electron(positron ) at @xmath110 is shown for different r : @xmath158 ( the upper curve ) , @xmath159 ( the middle curve ) , @xmath160 ( the lower curve ) , calculated according with eq.([s3 ] ) ( the solid curve ) and according with eq.([s6 ] ) ( the dashed curve )
. very good agreement of corresponding curves is seen in the region of sqa applicability ( @xmath161 ) .
in this paper we study the process of pair creation by a photon in a strong magnetic field basing on the polarization operator in the field calculated by different methods in a set of papers @xcite . by using the decomposition of functions into series containing the bessel function ( see eq.([a4 ] ) ) we deduced the imaginary part of the polarization operator ( which is the process probability ) into the relatively simple combination of the generalized laguerre polynomial ( see eq.([d10 ] ) ) .
this probability exhibits a `` saw - tooth '' pattern because of the factor @xmath76 .
this pattern will be washed out at averaging over any smooth photon energy distribution .
the range of applicability of the quasiclassical approximation extended in the photon energy interval where the created particles are no more ultrarelativistic ( see appendix b ) . found correction to standard quasiclassical approximation ( sqa ) permitted to extend the range of applicability of the quasiclassical approximation to the region @xmath123 at @xmath23 . from the performed analysis
follows the remarkable conclusion : for both @xmath87 and @xmath162 the results of the quasiclassical calculation are very close to averaged probabilities of exact theory in very wide interval of photon energies . for the total probability of pair creation
the quasiclassical approach is valid if ( see eq.([d8 ] ) ) @xmath163 \gg 1 \label{co1}\ ] ] in relatively weak field ( @xmath87 ) it is valid not far from threshold : @xmath164 . for this energy interval
the relative correction to the probability eq.([d15 ] ) is of the order @xmath165 . in superstrong field @xmath23 , for relativistic motion of the particles of the created pair and if @xmath166 the sqa is valid .
the expansion parameter in this case is @xmath167 ( see eq.([c11 ] ) ) . for such field and energy
it is helpful to compare the probability of pair creation by a photon @xmath168 with the radiation length @xmath169 for photon emission process found in @xcite . conserving the main terms of the decomposition over the landau energy number @xmath170 ( see eq.([d11 ] ) ) in the sqa as well as in the first correction we get ( @xmath155 ) @xmath171,\quad w=\frac{a_{p}}{l}\left[\frac{1}{l_{max}^{1/6}}+\frac{b_{p}}{l_{max}^{5/6 } } \right ] , \label{co2}\ ] ] where @xmath172 where the numbers are taken from eq.(46 ) in @xcite and eqs.([d19 ] ) ) , ( [ d20 ] ) ) above .
it should be noted that the physics characteristics in eq.([co2 ] ) ) do nt depend on the electron mass .
it s important that in the formula not only coefficients @xmath173 and @xmath174 are very close , but the same properties possess the coefficients @xmath175 and @xmath176 reflecting proximity of details of mechanisms of radiation and pair creation by a photon in a magnetic field .
one can expect that this property will be essential for consideration of electron - photon shower in a magnetic field .
set of kinetic equations for such shower with the particle polarization taken into account was investigated in @xcite . *
acknowledgments * the authors are indebted to the russian foundation for basic research supported in part this research by grant 06 - 02 - 16226 .
let us consider the integral in eq.([d2 ] ) @xmath177 using the formula @xmath178 and integrating by parts the integral over @xmath33 in eq.([a1 ] ) we get @xmath179 using eq.7.2.4.(27 ) @xcite we have @xmath180 where @xmath181 is the bessel function .
taking into account that the functions @xmath182 and @xmath183 in the integral over @xmath184 in eq.([d2 ] ) are the even functions of @xmath184 , we can perform following substitutions in sums in eq.([a4 ] ) @xmath185 so the expression for @xmath186 eq.([a1 ] ) can be written as @xmath187 where @xmath188 , \nonumber \\ \hspace{-8 mm } & & f_2^{(n)}=f_1^{(n)}+(-i)^n\exp(iz\cot x ) \frac{2\left(2-\delta_{n0}\right)}{z}\left(b\cot x-\frac{i}{\sin^2x } \right)j_{n}(t ) , \nonumber \\ \hspace{-8 mm } & & f_3^{(n)}=(-i)^n\exp(iz\cot x)(1-v^2)\cot x \left(2-\delta_{n0}\right)j_{n}(t)-f_1^{(n ) } , \nonumber \\
\hspace{-8 mm } & & a_n(v)=nv - b , \label{a7}\end{aligned}\ ] ] and the function @xmath189 is a periodic function of @xmath33 .
let us note that at @xmath190 the asymptotic of @xmath191 is @xmath192 in the case @xmath193 the integration contour over @xmath33 in eq.([a6 ] ) can be unrolled to the lower imaginary semiaxis and the contribution to the imaginary part of @xmath194 vanishes .
one can represent @xmath194 in the form @xmath195 we will use in eq.([a9 ] ) the known equality @xmath196 by virtue of the made above observation @xmath197 substituting eq.([a10 ] ) and eq.([a11 ] ) into eq.([a9 ] ) and taking into account that @xmath198 we have @xmath199 where @xmath200 it is seen from eq.([a13 ] ) that @xmath201 is even . using eq.([a12 ] ) one can represent the imaginary part of @xmath202 in the form @xmath203 along with integers @xmath204 and @xmath14 we will use also @xmath205 and @xmath206 .
we turn now to calculation of integrals in eq.([a15 ] ) .
let us consider the integral @xmath207 changing the variable @xmath208 we get @xmath209 one can close the integration contour in eq.([a17 ] ) by an infinite half - circle in the upper half - plane and than contract the contour to the point @xmath210 .
so , we have to calculate the integral in eq.([a17 ] ) over the contour @xmath211 .
changing successive the variables @xmath212 , @xmath213 we get @xmath214 using eq.10.12.(20 ) @xcite @xmath215 where @xmath216 is the modified bessel function , @xmath68 is the is the generalized laguerre polynomial , we get @xmath217 in the same manner one can calculate the integral @xmath218 somewhat different structure has the integral @xmath219 using eq.10.12.(9 ) @xcite : @xmath220 , \label{a23}\ ] ] one obtains @xmath221 \nonumber \\
\hspace{-10mm}&&=4 \pi e^{-z } z^n \frac{k!}{l!}l\left[l_{k-1}^{n+1}(z)l_{k}^{n-1}(z ) -l_{k}^{n}(z)l_{k-1}^{n}(z)\right ] .
\label{a24}\end{aligned}\ ] ] in derivation of the last expression eqs.10.12.(15),(16),(24 ) @xcite was used . in conclusion of this appendix
let us note , that @xmath202 contains the combination @xmath222 \nonumber \\
\pi i e^{-z }
z^{n-1 } \frac{k!}{l!}\left[\frac{1}{k}(ll_{k-1}^{n}(z)-kl_{k}^{n}(z))^2 + l(l_{k}^{n}(z)-l_{k-1}^{n}(z))^2\right ] \nonumber \\
\pi i e^{-z } z^{n-1 } \frac{k!}{l!}\left[m f(l , k , z ) - 4ll_{k}^{n}(z)l_{k-1}^{n}(z)\right ] .
\label{a25}\end{aligned}\ ] ] it should be noted that the above derivation is the most direct path from the process probability in the form of the integral eqs.([d1]),([d2 ] ) to the process probability in the form eq.([d7 ] ) . in the field which is weak comparing with the critical field @xmath223 and at relatively low photon energy ( @xmath224 )
the created particles occupy mainly states with high quantum numbers if the condition @xmath225 is fulfilled . in this case
the quasiclassical approach is valid , but created particles are no more ultrarelativistic , so the standard quasiclassical formulas @xcite are nonapplicable .
we will develop here another approach , using the method of stationary phase at calculation of the imaginary part of the integral over @xmath33 in eq.([d2 ] ) . granting that the large parameter @xmath226 is the common factor in the phase @xmath227 and does nt contained in the equation @xmath228 which defines the stationary phase point @xmath229 . in this case
the small values of variable @xmath184 contribute to the integral over @xmath184 , so that one can extend the integration limits to the infinity .
so we get @xmath230\right\}dx,\label{b1}\ ] ] where @xmath231 from the equation @xmath232 we find @xmath233 substituting these results in the expressions which defines the integral in eq.([b1 ] ) we have @xmath234 performing the standard procedure of the method of stationary phase one obtains for probability of pair creation by a polarized photon not far from threshold @xmath235 in spite of made above assumption @xmath224 eq.([b5 ] ) is valid also at @xmath86 if the condition @xmath236 is fulfilled .
this can be traced in the derivation of eq.([b5 ] ) .
the first two term of the decomposition of the function @xmath237 over power of @xmath238 are @xmath239 it follows from this formula that that applicability of eq.([b5 ] ) is limited by the condition @xmath240 .
if the second term much smaller than unity the exponent with it can be expanded . as a result
we have from eq.([b5 ] ) at @xmath241 @xmath242 the main term in above expression coincides with the probability of pair creation by a photon in standard quasiclassical theory at @xmath243 . the correction in eq.([b7 ] ) determines the lower boundary over the photon energy of standard approach applicability ( @xmath100 ) .
so exists the overlapping region where both the formulated here and the standard approach for high energy are valid .
the standard quasiclassical approximation valid for ultrarelativistic created particles ( @xmath86 ) can be derived from eqs.([d2 ] ) , ( [ d3 ] ) by expanding the functions @xmath244 over @xmath33 powers .
taking into account the higher powers of @xmath33 one gets @xmath245 \nonumber \\ \hspace{-10 mm } & & f_3(v , x)=-\frac{1-v^2}{6}\left[(3-v^2)x+\frac{1}{60}(15 - 2v^2 + 3v^4)x^3\right ] \nonumber \\ \hspace{-10 mm } & & \psi(v , x)=-\frac{r(1-v^2)^2}{12\mu}\left(x^3+\frac{3-v^2}{30}x^5\right ) -\frac{x}{\mu}. \label{c1}\end{aligned}\ ] ] here the first terms in the brackets give the known probability of the process in the standard quasiclassical approximation , while the second terms are the corrections .
expanding the term with @xmath246 in @xmath247 and making substitution @xmath248 one finds @xmath249dx , \nonumber \\
\hspace{-15 mm } & & g_2(v , t)=\frac{1-v^2}{12}\mu t\left[(3+v^2)-i\frac{9-v^4}{90}\xi\mu^2t^5 + \frac{\mu^2t^2}{15}(15 - 6v^2-v^4)\right ] , \nonumber \\ \hspace{-15 mm } & & g_3(v , t)=-\frac{1-v^2}{6}\mu t\left[(3-v^2)-i\frac{(3-v^2)^2}{90}\xi\mu^2t^5 + \frac{\mu^2t^2}{60}(15 - 2v^2 + 3v^4)\right ] , \label{c2}\end{aligned}\ ] ] where @xmath250 ^ 2}{16},\quad \kappa=2\sqrt{r}\mu=\frac{\omega}{m}\mu . \label{c3}\
] ] we will use the known integrals @xmath251 differentiating the first integral over @xmath252 one gets @xmath253 .
\label{c5}\end{aligned}\ ] ] substituting the integrals eqs.([c3])-([c4 ] ) in eq.([c2 ] ) we obtain the probabilities of pair creation in standard quasiclassical approximation @xmath254 and the corresponding corrections to this approximation @xmath255 where @xmath256 in derivation of eqs.([c7])-([c8 ] ) the integration by parts was used as well as the intermediate equalities in eq.([c5 ] ) .
the obtained representation of probabilities is convenient for calculation asymptotic at @xmath116 which are @xmath257 at @xmath90 the main contribution is given by the terms @xmath91 in @xmath258 eq.([c8 ] ) .
one has @xmath259 the relative magnitude of the corrections at @xmath116 @xmath260 at @xmath90 one has @xmath261 what agrees with eq.([b7 ] ) . substituting in eq.([s3 ] ) @xmath262 one has @xmath263 , \label{ad1}\ ] ] where @xmath69 is defined in eq.([d13 ] ) . if one omits the integration over @xmath184 in eq.([d18 ] ) , the result of this manipulation disagrees with eq.([ad1 ] ) .
so the described manipulation is erroneous . only if one performs integration by parts of the second term in eq.([ad1 ] ) and use the recursion relation @xmath264 one obtains eq.([d18 ] ) for integral probability .
99 m.ruderman , _ the electromagnetic spectrum of neutron stars _ , nato - asi proceedings , 2004 .
r.c.duncan , c.thompson , apj * 392 * , 9 , 1992 .
s.mereghetti , proceedings of 40th rencontres de moriond , _ very high energy phenomena in the universe _ , 2006 , astro - phys/0505491 .
n.p.klepikov , ph.d .
dissertation , moscow state university , 1952 ( unpublished ) .
j.s.toll , ph.d .
dissertation , princeton university , 1952 ( unpublished ) .
n.p.klepikov , zh.eksp.teor.fiz . , * 26 * , 19 ( 1954 ) .
s.l.adler , ann . phys.(n.y . ) , * 67 * , 599 ( 1971 ) .
j.schwinger , phys.rev . ,
* 82 * , 664 ( 1951 ) .
i.a.batalin , a.e.shabad , sov.phys.jetp , * 33 * , 483 ( 1971 ) v.n.baier , v.m.katkov , and v.m.strakhovenko , sov.phys.-jetp , * 41 * , 198 ( 1975 ) .
a.e.shabad , ann . phys.(n.y . ) , * 90 * , 166 ( 1975 ) .
j.k.daugherty and a.k.harding , ap.j . * 273 * , 761 ( 1983 ) .
v.n.baier , v.m.katkov , sov.phys.-jetp , * 26 * , 854 ( 1968 ) .
v.n.baier , v.m.katkov , v.s.fadin , _ `` radiation from relativistic electrons '' _ ( in russian ) , atomizdat , moscow , 1973 . v. n. baier , v. m. katkov and v. m. strakhovenko , _
electromagnetic processes at high energies in oriented single crystals _ , world scientific publishing co , singapore , 1998 .
w.tsai , and t.erber , phys.rev .
, d*10 * , 492 ( 1974 ) .
v.n.baier , a.i.milstein , r.zh.shaisultanov , zh.eksp.teor.fiz . , * 111 * , 52 ( 1997 ) ; journal of experimental and theoretical physics , * 84 * , 29 ( 1997 ) .
v.n.baier , v.m.katkov , and v.m.strakhovenko , sov.phys.-jetp , * 71 * , 657 ( 1990 ) .
v.n.baier , v.m.katkov , and v.m.strakhovenko , phys.lett , * b 70 * , 83 ( 1977 ) .
v.n.baier , v.m.katkov , and v.m.strakhovenko , sov.j . nucl.phys . *
27 * , 389 ( 1978 ) .
h.bateman , a.erdlyi , higher transcendental function , v.ii , mcgraw - hill book co , new york , 1953 .
( 10,120 ) ( 70,160 ) as a function of @xmath103 .
the probability in saw - tooth form is calculated according to eq.([d10 ] ) .
the dashed line is the result of averaging over the interval @xmath105 .
the solid line is calculated according to eq.([d15]).,title="fig:",height=415 ] ( 0,200 ) ( -65,8 ) .
the probability in saw - tooth form is calculated according to eq.([d10 ] ) .
the result of averaging over the interval @xmath105 is given by the dashed line .
the solid curve , which is quasiclassical approximation , calculated according to eqs.([d18 ] ) and ( [ d12]).,title="fig:",height=377 ] ( 0,200 ) ( -65,8 ) vs @xmath14 ( see eq.([s2 ] ) ) . the spectrum in saw - tooth form
is calculated according to eq.([d10 ] ) .
the dashed curve is the spectrum averaged over the interval @xmath265 .
the solid curve calculated according to eq.([s3 ] ) .
, title="fig:",height=377 ] ( 0,200 ) ( -60,8 ) for different r : @xmath158 ( the upper curve ) , @xmath159 ( the middle curve ) , @xmath160 ( the lower curve ) , calculated according to eq.([s3 ] ) ( the solid curve ) and according to eq.([s6 ] ) ( the dashed curve).,title="fig:",height=377 ] | the process of pair creation by a photon in a strong magnetic field is investigated basing on the polarization operator in the field .
the total probability of the process is found in a relatively simple form .
the probability exhibits a `` saw - tooth '' pattern because of divergences arising when the electron and positron are created at threshold of the landau energy levels .
the pattern will be washed out at averaging over any smooth photon energy distribution .
the new results are obtained in the scope of the quasiclassical approach : 1 ) in the case when the magnetic field @xmath0 is the critical field ) the new formulation extends the photon energy interval to the case when the created particles are not ultrarelativistic ; 2 ) the correction to the standard quasiclassical approximation is found showing the range of applicability of the approach at high photon energy as well .
the very important conclusion is that for both cases @xmath1 and @xmath2 the results of the quasiclassical calculation are very close to averaged probabilities of exact theory in a very wide range of photon energies .
the quasiclassical approximation is valid also for the energy distribution if the electron and positron are created on enough high levels . |
extremely metal poor stars provide important clues to the chemical history of our galaxy , the role and type of early sn , the mode of star formation in the proto - milky way , and the formation of the galactic halo .
the number of extremely metal poor ( emp ) stars known is summarized by @xcite .
they compiled a list of the key properties of the 12 stars identified up to that time with [ fe / h ] @xmath0 dex is given by @xmath8 on a scale where @xmath9 h atoms . then [ x / h ] = log@xmath10[n(x)/n(h ) ] @xmath4 log@xmath10[n(x)/n(h ) ] , and similarly for [ x / fe ] .
] , 7 of which are emp giants and subgiants within the range of @xmath2 considered here .
our 0z project has the goal of increasing the sample of such stars through data mining of the hamburg / eso survey ( he s ) @xcite .
this is an objective prism survey from which it is possible to efficiently select qsos @xcite as well as a variety of interesting stellar objects , among them extremely metal poor ( emp ) stars @xcite .
the 0z project has been systematically searching the database of the he s for this purpose over the past five years .
we present in [ section_sample ] a sample of new emp giants with @xmath2 @xmath11 k and [ fe / h ] @xmath0 dex which substantially increases the number of such stars known .
details of the analysis are described in [ section_analysis ] , while the radial velocity data are discussed in [ section_vr ] .
the abundance ratios for @xmath720 elements in each of the sample emp giants are described in [ section_abund ] . in the following section ( [ section_comp ] )
we check the consistency of the analyses and procedures adopted by our 0z project with those of the first stars vlt large project using uves @xcite and offer some comments on the implication of our results on the frequency of carbon - enhanced stars .
the penultimate section presents a discussion of the implications of our results for early sn and for nucleosynthesis in the forming galactic halo .
a summary of the key results follows .
two appendices which discuss details of a comparison of our work with other large spectroscopic and photometric surveys of emp stars complete this paper .
in pursuit of our goal of exploiting the he s to identify new emp stars , we began with a list of candidates selected from the he s database over the 50% of its area on the sky to which we have access .
our 0z project has now taken moderate resolution spectra of more than 600 candidates with the double spectrograph @xcite on the hale 5-m telescope at palomar mountain , and more than 1100 at the 6.5 m clay and baade telescopes at the las campanas observatory .
these spectra have been processed with the algorithm of @xcite , which uses a measure of the strength of h@xmath12 and of the 3933 line of ca ii , to determine rough metallicities , denoted [ fe / h](hes ) .
@xcite provides a brief description of the vetting process , see also @xcite .
those stars of special interest with follow - up spectra from palomar , including all those with [ fe / h](hes ) @xmath13 dex , have been observed with hires @xcite at the keck i telescope over the past four years ; a total of @xmath790 such stars have been observed with hires to date . in this paper
we present analyses of those stars from the palomar sample with @xmath2 @xmath11 k which turned out to be genuine emp stars with [ fe / h ] @xmath0 dex as determined from our high - resolution , high snr spectra , and which have not already been published in our earlier papers
. a future paper will deal with the most extreme metal - poor stars found in the las campanas sample .
the limitation on @xmath2 ensures that internal comparisons of abundance ratios within the sample will be as reliable as possible .
five new emp stars are presented here .
one turned out to be a rediscovery of a hk survey star .
this was not realized for a long time due to the 32 difference in the coordinates of cs22949@xmath4037 from its discovery by the hk survey @xcite , and those of he2323@xmath40256 , found in the he s .
it appears that the hk survey coordinates are sometimes in error by such a large amount , as the updated coordinates for this star given by @xcite are within 1.5 of those from the he s of he2323@xmath40256 .
we also include a new analysis based on better spectra of the only genuine emp giant described previously in our published papers , he01322429 , part of the keck pilot project @xcite .
we do not consider here the three emp dwarfs whose analyses we have published , bs165450089 and he13460427 from @xcite , nor he02182738 , a double lined spectroscopic binary @xcite .
there are many fewer absorption lines detected in these hotter stars and we wish to restrict ourselves to a narrow range in @xmath2 to ensure accurate comparisons within our sample .
we add to our sample the only star from the hamburg / eso r - process enhanced star survey ( heres ) @xcite which they believed to have [ fe / h ] @xmath14 dex , he1300 + 0157 .
we also add the star bs16467062 , which is included in the vlt / uves first stars program @xcite . a detailed analysis for the latter , found in the hk survey ,
was presented by @xcite , which was superseded by that of @xcite . since bs16467062 and he2323@xmath40256
are part of the first stars sample , several analyses have been published for each of these stars . in the present work
they serve as comparison objects to determine the consistency of the absolute iron abundance and relative abundances of various elements as deduced by our group versus those of the first stars vlt large program .
the sample stars are listed in table [ table_sample ] , which gives their j2000 coordinates , new optical photometry , when available , and other relevant data .
we use the procedures described in @xcite and used in all subsequent work by our 0z project published to date .
our @xmath2 determinations are based on broad band colors @xmath15 and @xmath16 .
the ir photometry is taken from 2mass @xcite .
we have obtained new photometry at @xmath17 and @xmath18 for almost all of the stars discussed here .
we use andicam images taken for this purpose over the past two years via a service observing queue on the 1.3 m telescope at ctio operated by the smarts consortium .
andicam is a dual channel camera constructed by the ohio state university instrument group .
our andicam program requires photometric conditions , and additional standard star fields , charged to our andicam allocation through noao , are always taken for us .
appendix a compares our photometry with that of @xcite and with that of the sloan digital sky survey @xcite .
we derive surface gravities through combining these @xmath2 with an appropriate 12 gyr isochrone from the grid of @xcite .
the resulting stellar parameters , which have been derived with no reference to the spectra themselves , are given in table [ table_sample ] .
the random uncertainties in @xmath2 from photometric errors ( see appendix a ) are 100 k. this ignores systematic errors which may be present .
the adopted uncertainties in log(@xmath19 ) are , following the discussion in @xcite , 100 @xmath20[log(@xmath19)]/@xmath20@xmath2 evaluated along the rgb for stars in this extremely low metallicity range , 0.2 dex at @xmath2= 5000 k and 0.15 dex at 5500 k.
the emp stars in our sample were observed with hires at keck during various runs over the past 4 years .
details of the best available spectra for each of the new stars are listed in table [ table_spectra ] .
six of these stars were observed with high snr spectra from hires at the keck observatory after the recent detector upgrade , which provides complete spectral coverage from 3180 to 5990 in a single exposure .
two ( he1356@xmath40622 and he1347@xmath41025 ) have only relatively short exposures with the new detector , used primarily to measure the strength of the nad lines .
their only high snr spectra were taken before the upgrade and hence have spectral coverage restricted to 3840 to 5330 with no gaps between orders for @xmath21 , and only small gaps thereafter .
the slit width ( either 0.86 arcsec , corresponding to a spectral resolution of 46,000 , or 1.1 arcsec , which corresponds to a spectral resolution of 35,000 ) used for each spectrum is indicated in this table as well .
each exposure was broken up into 1200 sec segments to expedite removal of cosmic rays .
the goal was to achieve a snr of 100 per spectral resolution element in the continuum at 4500 ; a few spectra have slightly lower snr .
this snr calculation utilizes only poisson statistics , ignoring issues of cosmic ray removal , night sky subtraction , flattening , etc .
the observations were carried out with the slit length aligned to the parallactic angle .
the processing of the spectra was done with makee and figaro @xcite scripts , and follows closely that described by @xcite .
the equivalent widths were measured as described in @xcite .
table [ table_eqw ] lists the atomic parameters adopted for each line and their equivalent widths measured in the spectra of each of the 8 emp stars .
the analysis is identical to that of @xcite with several important additions .
in particular we use the model stellar atmosphere grid of @xcite and a current version of the lte spectral synthesis program moog @xcite , which treats scattering as lte absorption .
the improved hires spectra now reach into the near - uv , making the nh and redder part of the oh bands accessible .
we use the molecular line list of @xcite , augmented with the strongest atomic features , to analyze the nh band .
we use the line list of @xcite for oh . our nominal solar cno abundances are 8.59 , 7.93 , and 8.83 dex respectively .
these are close to those of @xcite , but somewhat larger than the values obtained using 3d model atmospheres by @xcite and @xcite .
we prefer not to attempt 3d corrections until a full grid of model 3d atmospheres or of corrections to cno abundances derived from the molecular bands from 3d to 1d models becomes available . for ch and for nh
we have adjusted the scale of our @xmath22 values so as to reproduce the solar spectrum , taken from @xcite .
we adopt dissociation potentials of 3.47 and 4.39 ev @xcite for ch and oh respectively .
for nh we adopt 3.40 ev based on the theoretical calculations of @xcite and the laboratory spectroscopy of @xcite .
our analysis assumes classical plane parallel stellar atmospheres and lte , both for atomic and for molecular features . in view of the inclusion of the near - uv in many of these spectra
, a few uv lines of key atomic species have been added to the master line list .
these include , for example , three fe ii lines near 3270 which are stronger than any in the optical band .
this is important as often only the two strongest fe ii lines in the optical are detected in even high quality spectra of such extremely metal poor giants ; the remaining optical fe ii lines are too weak . for the most extreme emp dwarfs ,
none of the optical fe ii lines can be detected .
inclusion of the uv lines strengthens the determination of the ionization equilibrium between fe i , with its multitude of detected lines , and fe ii .
lines of species with no detectable optical features , such as v ii and mn ii , have also been added .
the resonance lines of cu i near 3250 were added as well , as the usual cu i lines seen in stars with [ fe / h ] @xmath23 dex , including those at 5105 and 5782 , become undetectable at the extremely low metallicities of the stars studied here . in such cases ,
the @xmath22 values were adopted from version 3.1.0 of the nist atomic spectra database ( phsics.nist.gov/physrefdata/asd/index.html ) .
the hfs patterns for the uv mn ii lines are given by @xcite , but these lines in the sample stars are mostly so weak that the corrections are negligible . those for the uv cu i lines were downloaded from r. kurucz s web site ; his primary source was @xcite .
the isotope ratio @xmath24cu/@xmath25cu was assumed to be the solar value . where possible
, we have checked the consistency of the scale of the transition probabilities for a given species between the rarely used uv lines and the commonly used optical ones by comparing the derived abundances of lines for the same species as a function of wavelength for a small number of stars with hires spectra taken as part of our 0z project with somewhat higher [ fe / h ] , including as hd 122563 , with weak or no detected ch or nh .
the results of this check were satisfactory for fe ii .
however , even with hfs included , log[@xmath26(mn ) ] deduced from the 4030 triplet of mn i ( the strongest lines in the optical region , two lines of which are sufficiently unblended to use in an abundance analysis ) appear to be @xmath70.3 dex lower than those found from the redder optical lines and from the uv resonance lines of mn ii . in the solar spectrum , as well as for hd 122563 , the nominal solar mn abundance of @xcite is recovered only for lines with @xmath27 , as well as from the mn ii uv lines . a similar problem with the 4030 mn
i triplet was noted by @xcite .
@xcite has carried out such a consistency check for cu i. they compare their [ cu / fe ] determinations based on the near - uv resonance lines with those of @xcite derived from the weak optical lines .
they find good agreement , i.e. a mean difference for 16 stars of @xmath28 dex .
following @xcite , we adopt a non - lte correction for al i , for which we only detect the resonance doublet at 3950 , of + 0.60 dex based on the work of @xcite and @xcite .
we adopt a non - lte correction for na i , for which we can only observe the two d lines , of @xmath29 dex based on the calculations of @xcite .
no other non - lte corrections have been applied .
our abundances for the cno elements are based on molecular bands of ch , nh and oh respectively .
we use 1d model atmospheres to synthesize the molecular features , ignoring any 3d effects , although @xcite suggest that these may be very large .
they claim that cno abundances may be overestimated by @xmath70.8 dex as compared to a 1d analysis when molecular bands are used in emp stars .
table [ table_slopes ] gives the slope of a linear fit to the abundances determined from the set of fe i lines as a function of @xmath30 ( the excitation potential of the lower level ) , @xmath31 , and @xmath32 , which are most sensitive to @xmath2 , @xmath33 , and the wavelength dependence of any missing major source of continuous opacity , respectively .
there are @xmath7 40 to 60 fe i lines detected in each star , with @xmath30 ranging from 0 to 3 ev .
the correlation coefficients @xmath34 of the fits with @xmath32 are , for all except one of the stars , between 0.11 and @xmath40.20 , indicating that these fits are not statistically significant .
the @xmath35 ) for the fit with @xmath31 are within the same range for most of the stars .
the slopes with @xmath30 appear at first sight to be statistically significant with @xmath36 for one of the 8 stars ; the @xmath37 are predominantly negative .
they reach as low as @xmath38 dex / ev for two of the stars .
if these slopes were valid , they would suggest that for these two stars , @xmath2 needs to be decreased by @xmath7300 k to achieve excitation equilibrium for fe i. however , a careful scrutiny of the behavior of the derived fe abundance from individual fe i lines indicates that the problem lies largely in the 0 ev lines ; for those of higher excitation ( the majority of the lines ) , the deduced abundance shows no statistically significant dependence on @xmath30 .
a typical example of this is shown in fig .
[ figure_ep ] .
our plots of abundance versus reduced equivalent width indicate that the fe i overabundances for lines with @xmath39 ev do not appear to arise from an inappropriate choice of microturbulent velocity parameters
. the effect could be due to systematic errors for @xmath22 values of low excitation lines , or may result from resonance scattering ( e.g. see * ? ? ?
* ) . in resonance scattering the source function , s@xmath40 ,
is reduced to below the local planck function , thus leading to a stronger line in non - lte .
resonance scattering is seen in the na d lines of metal - poor stars ( e.g. * ? ? ?
* ) , the oi triplet at 7774 in the sun and may have affected the abundances from the ca i 4226 resonance line of @xcite .
if resonance scattering is the cause of the small abundance enhancement seen in the 0 ev fe i lines , the effect should be more pronounced in the weakest low excitation lines ; thus a plot of @xmath31 versus abundance enhancement should show a positive correlation . proof that resonance scattering is the cause of the apparent overabundances requires a non - lte abundance calculation for fe in our stars , which is beyond the scope of this paper .
the radial velocities were determined using the procedure described in @xcite , updated to improve the long term stability of @xmath41 measurements ; they are given in table [ table_spectra ] .
each individual measured @xmath41 from the spectrum of a star taken with the upgraded hires detector for the snr typical of the present set of spectra taken on a given run now has an internal uncertainty of @xmath42 km s@xmath43 , with possible long term systematic drifts of comparable size ; this was not true of the @xmath41 published in earlier papers of the 0z project . for six of our sample emp stars
we have hires spectra taken between 2 and 5 years apart .
the older spectra have been re - reduced using our improved procedures and codes to measure more accurate @xmath41 .
r. cayrel has generously provided the @xmath41 measured by the first stars project from their uves spectra for the two stars discussed here which are in common .
table [ table_vr ] compiles the available high precision radial velocities for these emp stars from hires and from uves .
two of the stars with @xmath41 measurements from multiple hires spectra show @xmath41 variations , he01322429 at more than 10@xmath44 and he10121540 at the 4@xmath44 level .
in addition , bs16467@xmath4062 may also be a @xmath41 variable , with a difference exceeding 5@xmath44 between the two uves measurements from m. spite & r. cayrel and that from our hires spectrum , but this should be verified .
such @xmath41 variations presumably result from orbital motion in a binary system .
tables [ table_abunda ] and [ table_abundb ] give the derived abundances for each detected species in each of the 8 emp giants .
we divide the stars into two groups : low c stars with the g band of ch barely detectable , if present at all , and three stars which appear to have c enhanced .
the first group includes five of the 8 stars , he0132@xmath42429 , he1347@xmath41025 , he1356@xmath40622 , he1424@xmath40241 and bs16467@xmath4062 .
the second group includes three stars , two of which ( he1300 + 0157 and he2323@xmath40256 ) are in the present analyses hovering just at the border line of being c - rich , defined by @xcite as [ c / fe ] @xmath45 dex .
the third star , he1012@xmath41540 , is a highly c - enhanced emp star .
the fraction of c - rich stars in our sample of emp giants with this definition is a minimum of 1/8 and a maximum of 3/8 , depending on which side of the boundary of [ c / fe ] = + 1.0 dex the two borderline stars fall .
we first consider the 5 stars with low c. these stars span a relatively small total range in @xmath2 of 420 k , from 4950 to 5370 k. their spectra show very weak metallic absorption lines and lack strong molecular features , so measuring @xmath31 for them is straightforward ; the uncertainties in the @xmath31 are thus low , particularly for the three stars with high snr spectra taken with the updated hires detector
. a breakdown of the abundance errors resulting from uncertainties in @xmath2 , log(@xmath19 ) , and @xmath33 can be found in table 6 of @xcite .
given our uncertainties in the determination of the stellar parameters , the dominant contribution to the uncertainties in the abundance ratios [ x / fe ] is that of @xmath2 .
the contribution from errors in @xmath33 , @xmath31 and the assumed metallicity of the model atmosphere are small for most elements due to the use of mostly weak lines and to the stars being so metal poor .
( exceptions are ca i , mg i and ni i , where the expected contributions to the uncertainty in [ x / fe ] from @xmath2 and from log(@xmath19 ) are each @xmath46 dex , and sc ii , la ii , ba ii and eu ii , where they are each @xmath47 dex in absolute value . )
we therefore expect the abundance ratios when compared among this group of stars to have small errors , @xmath48 dex .
systematic errors of comparable size may exist for the cno elements as these are derived from hydride bands and accurate values for many molecular parameters are required for their analysis .
this estimate is too conservative , at least for comparisons internal to the 0z project , for neutral species with at least three detected absorption lines which have a temperature dependence similar to that of fe i , an example of which is cr i , when enough lines are detected in a star . the abundance ratios in these stars among the heavy elements are shown in fig .
[ figure_cnormal_heavy ] for 9 species in the range ca to cu
. the median is indicated , and stars with [ x / fe ] which deviate from the median by more than 0.3 dex are shown individually ; these are always either he0132@xmath42429 or he1424@xmath40241 .
if there is an outlier for a particular species , the individual values of [ x / fe ] for the remaining four stars in the sample are shown as well .
thus this figure demonstrates graphically how deviant the outliers really are .
the medians show ( see fig .
[ figure_cnormal_heavy ] ) a small excesses for ca and for ti with respect to fe .
cr , mn and cu are deficient relative to fe , while co is strongly enhanced relative to fe .
ni appears to be tied to fe so that [ ni / fe ] @xmath49 dex .
these are in agreement with the usual trend seen among extreme emp stars as found in our previous work @xcite and in the first stars project @xcite .
the surprise is the outliers .
five species with outliers are marked in fig .
[ figure_cnormal_heavy ] , all of which arise in only two stars , he0132@xmath42429 and he1424@xmath40241 .
spectral regions illustrating lines of three of these cases , contrasting the outlier star with a star close to the median value of [ x / fe ] , are shown in figs .
[ figure_sc2_4246 ] , [ figure_ti2_4444 ] and [ figure_mn_4030 ] .
the stars displayed in each figure have been chosen to have @xmath2 as close to each other as possible .
these figures demonstrate the reality of the outlier in [ x / fe ] for each of sc ii , ti ii and mn i. the spectra of the each outlier star for each of these five cases have been checked twice .
two independent hires spectra exist for several of these stars .
there is no question that the outliers in each of [ ca , sc , ti , mn , co / fe ] are real .
[ figure_cnormal_light ] shows the behavior of [ x / fe ] for 7 light elements in the range c to ca for the five c - normal stars .
upper limits are ignored , and only one of these stars has a detectable nh band .
again there are outliers
. one might expect outliers among those elements ( c , n , o , na , mg , and al ) where mixing of proton - burning material has already been demonstrated to occur among luminous emp giants by @xcite and @xcite , while @xcite present slight modifications in the details due to non - lte effects without altering the overall picture . large variations in [ n / fe ] are clearly present among this small sample of c - normal emp giants , as illustrated in fig .
[ figure_nh3360 ] , where it is shown that he0132@xmath42429 has a strong enhancement for [ n / fe ] .
the most peculiar star in this sample of emp giants , he1424@xmath40241 , is also the most metal poor , with [ fe / h ] @xmath5 dex .
the very low [ si / fe ] and other abundance anomalies found in a preliminary analysis of this star were briefly reported in @xcite .
he1424@xmath40241 has a ratio of [ si / fe ] which is 1.2 dex below that of any other star in the present sample , a result which is completely unexpected .
[ figure_si3905 ] shows the spectral region of the 3905 si i line to demonstrate the obvious reality of this very discrepant abundance ratio .
this star also has anomalously low [ ca / fe ] and moderately low [ ti / fe ] , accompanied by unusually high [ mn / fe ] ( by 0.6 dex ) and [ co / fe ] .
the g band of ch and the nh band at 3360 are not detected in the hires spectra of this star , implying fairly low upper limits for c and for n. the emp giant he0132@xmath42429 has [ sc / fe ] higher than any other star in the sample of c - normal stars by 0.4 dex , accompanied by very high [ n / fe ] , and [ n / c]@xmath50 .
he1356@xmath40622 shows an apparent small excess for [ na / fe ] and for [ si / fe ] .
the modest anomalies in this star are small enough that their reality is dubious .
the plots comparing spectral regions around key absorption lines presented here reinforce our claim that most if not all of the discrepant points in figs .
[ figure_cnormal_heavy ] and [ figure_cnormal_light ] are unquestionably real , and not the result of observational error nor of uncertainties in the analysis .
they are not consistent with a dependence on condensation temperature nor on first ionization potential .
table [ table_abunda ] presents a summary of the abundance ratios found among the five c - normal stars .
we have presented in tables [ table_abunda ] and [ table_abundb ] the first cu abundances for extreme emp stars , made possible by the high efficiency of hires in the near - uv , so that we can reach cu i resonance lines near 3250 .
[ figure_copper ] displays [ cu / fe ] as a function of [ fe / h ] for our sample of emp giants ( including one whose analysis has not yet been published ) .
earlier results using the weak optical cu i lines by @xcite ( for the giants in their sample ) and by @xcite who compiled the means for galactic globular clusters are shown in this figure . a steady decrease of [ cu / fe ] as the fe - metallicity decreases was found in previous work .
our new results demonstrate that [ cu / fe ] reaches a plateau at low fe - metallicity below [ fe / h ] @xmath51 dex which continues through the extreme emp stars , as might be expected if cu is formed primarily in massive stars .
the behavior of the heavy element ratio [ sr / ba ] is presented in fig .
[ figure_sr_ba ] . here
the values from all the stars in our sample of candidate emp giants from the he s from our published and unpublished work are indicated as well to provide guidance as to the typical behavior .
stars from the present sample often have very weak lines of both of these elements . if no detected line has @xmath31@xmath52 m , we consider the abundance @xmath26(sr ) or @xmath26(ba ) for that star to be an upper limit .
the abundance ratio @xmath26(sr)/@xmath26(ba ) ranges over more than a factor of 100 , in agreement with @xcite and @xcite .
the fractions predicted to arise from pure @xmath53 or @xmath54 process nucleosynthesis for the sun , taken from @xcite , are indicated in fig .
[ figure_sr_ba ] by the dashed and solid horizontal lines .
cs22892@xmath4052 , the prototype for the rare extreme @xmath53-process stars , shows [ sr / ba ] @xmath55 dex ( see , e.g. * ? ? ?
* ) , a value somewhat below the @xmath53-process line indicated in fig . [ figure_sr_ba ] .
the presence of numerous emp stars with [ sr / ba ] larger than that from either the standard @xmath54 or @xmath53-process demonstrates that another process must exist which produces the light neutron capture elements , and in particular sr , in emp stars , without producing those in the second peak ( i.e. ba ) , as was originally suggested by @xcite .
early calculations by @xcite have been updated and augmented by @xcite , who emphasize the many nucleosynthetic processes that can produce sr , y and zr , and who suggest again that a secondary source of sr from an as yet unidentified nucleosynthetic site is required .
supernova calculations by @xcite found production of elements significantly heavier than the iron - peak , up to a@xmath7100 , occurs for high neutron excess material ( greater than @[email protected] ) during the alpha - rich freezeout .
they also suggested that the alpha - rich freezeout might merge , naturally , into an @xmath53-process .
@xcite computed supernova yields for a range of masses and metallicities , including charged particle reactions up to mo . in their models they found that elements heavier zn could only be produced for metallicities greater than z / z@xmath57=10@xmath58 , essentially confirming the neutron - excess sensitivity found by @xcite for the production of elements up to a@xmath7100 .
this metallicity limit excluded the supernovae considered by @xcite as sources for the enhancements of elements up to a@xmath7100 seen in emp stars ( e.g. as found by * ? ? ? * ) .
@xcite explored theoretical supernova and hypernova yields ; while they included the alpha - rich freezeout in their calculations , they only considered species up to a=74 .
their hypernovae were characterized with kinetic energies more than 10 times that of normal core - collapse supernovae .
it was found that in the complete si - burning region of hypernovae elements produced by the alpha - rich freezeout are enhanced .
thus , we speculate that some form of alpha - rich freezeout , perhaps from hypernova explosions , with metallicities lower than the low limit determined by @xcite , may yet provide an explanation for the extra source of a@xmath7100 elements seen in some emp stars ( also known as a `` second r - process '' ) .
most of the stars with [ sr / ba ] significantly less than that of the @xmath53-process ( and of the @xmath54-process as well ) are carbon stars with highly enhanced ba from the @xmath54-process running at low fe - metallicity ( see , e.g. * ? ? ?
some of the emp stars studied here fall somewhat below the @xmath53-process line , but not by more than 2@xmath44 .
all of the emp stars in the present sample , including the three c - rich stars , have [ ba / fe ] @xmath59 dex . with such weak lined stars , and no excess of ba in any of them , no other heavy elements beyond the fe - peak besides ba and sr
could be detected in any of the 8 emp stars studied here .
there are three stars with the g band of ch obviously much stronger than the five c - normal stars discussed above .
two of the then have larger enhancements of c than of n , but one ( he2323@xmath40256 , cs22949@xmath4037 ) ) has [ n / fe ] @xmath60 [ c / fe ] , as shown in fig .
[ figure_crich_light ] . in two of the three stars , including the highly n - enhanced star , na and mg are also enhanced , while the enhancement of al is more modest .
such enhancements have been seen among other very metal poor carbon stars , for example he0336 + 0113 from our 0z survey , with [ mg / fe ] + 1.0 dex @xcite , the emp giant cs29498@xmath4043 , studied by @xcite with highly enhanced na and mg as well , and the very extreme emp dwarf cs22958@xmath4042 ( @xmath2 6250 k , [ fe / h ] @xmath61 dex ) with [ na / fe ] + 2.8 dex analyzed by @xcite . on the other hand , he1300 + 0157 , which has the smallest c+n - enhancement of the three stars , [ c / fe ] @xmath62 dex , has only an upper limit for n ( from nh ) , with highly enhanced o ( from oh ) .
it shows normal abundances for all other detected elements .
each of the three c - rich emp stars shows good agreement for the abundance ratios [ x / fe ] for the elements with detected features from ca through cu with each other and with the median from the five c - normal stars .
this is illustrated in fig .
[ figure_crich_heavy ] . the c - enhancement , even when extreme , does not affect the relative abundances of elements in this range , which includes the fe - peak , as was suggested earlier by , e.g. @xcite .
the 0z project and the first stars project of @xcite using uves at the vlt are two large independent efforts to determine the chemical abundance ratios in emp stars and use those to draw inferences on the properties of the early galaxy , the first supernovae , etc .
our sample largely consists of new emp stars we have found through painstaking , time consuming searches of the he s database coupled with the expenditure of very large and generous allocations of telescope time .
we have observed a few stars from thee sample of the first stars survey at the vlt @xcite and have analyzed them independently .
we compare our results with those of @xcite to determine the consistency of the absolute iron abundance and relative abundances of various elements as deduced by our group versus those of the first stars vlt large program .
we begin by testing the measured equivalent widths for the two stars in common , he2323@xmath40256 ( a.k.a .
cs22949@xmath4037 ) and bs16467@xmath4062 , the latter of which we added to our sample specifically for this purpose .
there are 79 lines in common for bs16467@xmath4062 , with a mean difference in @xmath31 of 1.8 mand @xmath44 of the differences of 4.4 m .
this extremely good agreement is shown in fig .
[ figure_eqw_bs16467 ] .
the agreement in measured @xmath31 for he23230256 ( a.k.a cs22949037 ) is not quite as good ( see fig .
[ figure_eqw_he2323 ] ) ; the dispersion of the differences in measured @xmath31 for the 66 lines in common is 9.5 m , with a mean difference of only 0.3 m .
most of the dispersion arises from 5 discrepant lines , as is shown in fig .
[ figure_eqw_he2323 ] .
m. spite advises ( private communication , june 2007 ) , on behalf of the first stars project , that their published @xmath31 for these 5 lines are not correct , and that the correct values from their uves spectra are much closer to those given here in table [ table_eqw ] .
she further advises that she believes that their problems with @xmath31 are restricted to this particular star .
we next examine the scale of the transition probabilities adopted by each group . for 43 fe i lines in common in the spectrum of bs16467@xmath4062 , the mean difference in log(@xmath22 ) is only 0.004 dex , with @xmath44 for the differences of 0.06 dex .
the scale of the @xmath22 values for all lines of species in common with @xcite have been compared .
the maximum scale difference for the lines of a given species was only 0.04 dex ( occurring for fe ii ) , with the largest dispersion about the mean for the lines in common reaching 0.07 dex ( for ti ii ) .
thus we find that the parameters adopted for atomic lines by the two groups are in very good agreement , and specifically for fe i are identical in the mean to within @xmath63 dex .
we have adopted the @xcite reddening map , which has a small but non - zero reddening at the galactic pole , while the first stars project appears to be using the older @xcite values based on 21 cm hi surveys .
this map has zero reddening at the galactic pole .
hence we have systematically larger reddening values for each star than does the first stars project . a detailed discussion of the differences in the stellar parameters between us and the first stars project and the differences in abundance , both absolute ( i.e. [ fe / h ] ) and ratios with respect to fe ( [ x / fe ] ) , is given in appendix b. overall for a particular star with measured @xmath31 for a set of detected absorption features and values of observed optical and 2mass colors , the [ fe / h ] value derived by the first stars project
as described in @xcite will be systematically @xmath70.3 dex lower than that for the same star as analyzed by the 0z project .
it is interesting to note that the sample of stars analyzed by @xcite also included two stars in common with the first stars project sample . @xcite
derived [ fe / h ] values higher than those of @xcite by @xmath64 dex .
table 1 of @xcite compares the mean for [ x / fe ] for c - normal emp giants between our 0z survey and the first stars survey of @xcite .
a detailed comparison for a small number of individual stars in common is given in appendix b. we find much better agreement of the abundance ratios [ x / fe ] between the two large surveys than for absolute fe - metallicities [ fe / h ] .
this is as expected since many of the error terms in the absolute iron abundance [ fe / h ] largely cancel out in an abundance ratio [ x / fe ] .
if we ignore c deduced from an analysis of the ch band and n inferred from the uv nh band , we find differences in [ x / fe ] for 11 or 12 elements in a star ranging from @xmath65 to + 0.05 dex when we adopt their equivalent widths but use our stellar parameters .
somewhat larger differences , @xmath660.15 dex with @xmath67 dex , occur when we analyze our own hires spectra with our own choice of stellar parameters , in part because of the errors in the @xmath31 of @xcite .
as indicated above , only a maximum of @xmath68 dex of these differences arise from differences in the scale of the transition probabilities adopted by each of these large surveys .
analyses of 253 stars from the hamburg / eso @xmath53-process enhanced star survey ( heres , * ? ? ?
* ) was presented by @xcite .
this survey relied on modest snr high resolution spectra of candidate emp giants from the he s .
we have observed with hires at the keck i telescope the most fe - poor star found in that survey , he1300 + 0157 , as a comparison object .
a very detailed abundance analysis based on a high quality subaru / hds spectrum for this star was recently presented by @xcite . both heres and @xcite utilize the relations between broad band colors and @xmath2 for giants of @xcite evaluated at [ fe / h ] @xmath51 dex , as these relations have not been calibrated adequately at still lower metallicities . as was discussed in @xcite ,
while the marcs and atlas9 @xmath2 color - relations are in very good agreement , they disagree with those of @xcite . in this regime of @xmath2 ,
log(@xmath19 ) and [ fe / h ] , the difference in deduced @xmath2 for a fixed @xmath16 color is @xmath7200 k , with the value derived from the @xcite relations being cooler .
the @xmath2 adopted by @xcite , @xcite , and that we derive for he1300 + 0157 are 5411 , 5450 , and 5630 k respectively .
this difference in @xmath2 corresponds to a difference in [ fe / h ] of @xmath70.35 dex , with he1300 + 0157 having the higher fe - metallicity of @xmath69 dex in our analysis instead of the value they obtained , @xmath70 dex .
we first consider the comparison for the star he1300 + 0157 if we adopt @xmath2 and log(@xmath19 ) from @xcite , but analyze with our own codes and atomic parameters our own set of @xmath31 from our hires / keck spectra .
overall , with this assumption , the agreement between the results presented here based on high snr keck / hires spectra and those of heres is very good given the lower snr of their spectra and the automatic analysis codes employed in the analysis of @xcite . for the 10 elements in common ,
the agreement in log@xmath26(x ) is in all cases within the errors assigned by heres for absolute abundances ( their `` erra '' values , ranging from @xmath70.2 to 0.3 dex ) , and in almost all cases is within the smaller relative errors they assigned for star - to - star comparison within heres .
the techniques employed by heres are certainly more than adequate to find interesting emp stars and determine the general nature of their chemical inventory .
we detected several additional elements beyond those that heres could reach .
we suggest , as did @xcite , that the heres claimed detection of y using their automatic abundance analysis code is almost surely incorrect ; they found [ y / fe ] + 0.56 dex .
they could not detect sr ; we find a marginal detection of a single sr ii line which yields [ sr / fe ] @xmath71 dex . given this very low sr abundance ,
any accessible optical y line would be expected to be undetectable .
we now compare our derived [ fe / h ] and abundance ratios [ x / fe ] for he1300 + 0157 with those of the very detailed and careful analysis by @xcite .
their @xmath2 is 180 k cooler than ours , hence they derive [ fe / h ] 0.34 dex lower than we do .
however , the difference in abundance ratios should be smaller assuming the same stellar parameters are adopted . for the cno elements
, we note that there is agreement to within 0.2 dex for the [ c / fe ] and [ o / fe ] abundance ratios in this star between our analysis and that of @xcite , while both fail to detect the nh band and only have an upper for [ n / fe ] .
our [ c / fe ] is identical to that derived by @xcite , who analyzed the heres spectra for the cno elements .
since their [ fe / h ] for this star was only @xmath72 dex , this implies a difference in @xmath26(c ) between the value they derive and that of either of the two high dispersion analyses of about a factor of 3 ( 0.5 dex ) , with the value of @xcite for he1300 + 0157 being too large .
we have carried out the comparison for this star adopting first the stellar parameters of @xcite , then those we have derived . in both cases we use our own set of @xmath31 measured from our hires spectra .
the results are given in table [ table_he1300 ] .
our measured @xmath31 for fe i lines show no systematic difference with those of @xcite .
similarly @xmath73(fe ) = [ fe i / h ] [ fe ii / h ] has the same sign ( positive ) in both analyses , but our ionization equilibrium is slightly better than theirs ( @xmath73(fe ) = 0.08 vs 0.15 dex ) for the identical stellar parameters .
however , log[@xmath74 is slightly higher in our analysis ( by 0.13 dex ) with same stellar parameters ; the origin of this offset is not clear . adopting their ( cooler ) @xmath2 , we find that of the 17 elements in common ( ignoring upper limits ) , log[@xmath75 ) ] differs by less than 0.10 dex for 7 of them , but disagrees by more than 0.15 dex for 6 species .
adopting our hotter @xmath2 raises [ fe / h ] substantially .
but the values of [ x / h ] are only slightly altered , @xmath76 } } \mid ~ < ~ 0.15 $ ] dex for most elements in our analysis .
the largest change in abundance ratio is seen for c , which is derived from the ch molecular band ; with the higher @xmath2 [ c / fe ] increases by @xmath70.2 dex .
the same holds for o ( from the oh band ) .
our 0z survey and the first stars survey at the vlt ( see , e.g. cohen _ et al . _ 2004 ,
cayrel _ et al . _
2004 ) , following in the footsteps of many earlier studies , including , for example , @xcite and @xcite , have established over the past five years the general behavior of abundances among emp stars , with substantial samples of stars analyzed with [ fe / h ] @xmath77 dex . if one ignores the light elements which might be affected by proton burning , definite trends of [ x / fe ] with [ fe / h ]
have been established beginning with ca and extending through the fe - peak which hold down to [ fe / h ] @xmath5 dex .
the data available to date show that there is a scatter about these trends which for most stars with normal carbon abundances and for most elements is not larger than the observational uncertainties .
a substantial theoretical effort has gone into calculations of nucleosynthesis yields in core collapse sn directed towards understanding the behavior of these `` typical '' vmp and emp stars .
the work of @xcite , @xcite and @xcite are examples of recent computations for grids of metal - poor stars over a range of initial mass and chemical composition .
these models have been tuned to reproduce the previously observed trends of abundance ratios among `` typical '' emp stars . in comparing the properties of our sample of emp giants with the predictions of such calculations
, it behooves us to recall the enormous difficulty of these calculations and the many parameters whose values must be calculated from theory , assumed , or inferred from the data and which substantially affect the resulting predicted nucleosynthesis yields . among the most crucial of these factors
are the explosion energy , the mass cut , the neutron excess , and previous mass loss in earlier evolutionary stages .
there are three stars in the present sample of emp giants which we consider as `` typical '' .
they have normal or low carbon .
their abundance ratios follow the patterns previously delineated by our work and that of the first stars project .
however , there are also two c - normal stars which are definite outliers .
some abundance ratios in these two stars are definitely anomalous .
the small @xmath2 range of our sample discussed here ensures that intercomparisons within the 0z project set of abundance analyses are valid and that differences exceeding 0.3 dex are real .
furthermore , in table 1 of @xcite we compared our mean abundance ratios with those of @xcite for those `` typical '' giants studied by each group . in our case
this included our published and unpublished abundance analyses , and for the first stars project we relied upon the fits tabulated in @xcite .
the agreement was very good , within 0.10 dex , with one exception ( mg ) .
although we emphasize again that we have shown here that our [ fe / h ] scale is systematically 0.3 dex higher than that of @xcite , the consequences for abundance ratios of differences in the details of the analyses between these two surveys are much smaller .
we are therefore confident that any outliers found are real and are not the result of problems or uncertainties in our measurements or analyses .
he1424@xmath40241 , with [ fe / h ] @xmath5 dex , is the most extreme outlier we have found .
its peculiarities were briefly described in @xcite .
this extreme emp giant has a very low abundance of si , and moderately low ca and ti , with respect to fe .
si , ca and ti are produced primarily via explosive @xmath78-burning .
but mg / fe , where mg is produced largely by hydrostatic @xmath78-burning , is normal in this star .
mn and co are enhanced with respect to fe .
only ( low ) upper limits for c and for n could be determined .
older calculations of nucleosynthetic yields by @xcite come close to reproducing at least some of this behavior with ejecta from sn biased towards the lower end of the relevant mass range , but more current grids of snii nucleosynthesis fail to reproduce the very unusual chemical inventory seen in this extreme emp star .
we defer to our theoretical colleagues to try to find an explanation for this very peculiar star .
the anomalies seen in he1424@xmath40241 are unique and , as far as we are aware , are not seen in any other emp giant studied to date .
this is illustrated in fig .
[ figure_si ] and [ figure_ca ] , which show [ si / fe ] and [ ca / fe ] as a function of [ fe / h ] for all the c - normal giants analyzed by our 0z project to date , all those of the first stars project @xcite , and those from several other sources noted in the figure legends . in both cases
, he1424@xmath40241 has the lowest value by far of the relevant abundance ratio .
there is one star , cs22966@xmath4043 , studied by @xcite and again by @xcite which has abundance ratios somewhat similar to those of he1424@xmath40241 .
cs22966@xmath4043 has [ fe / h ] @xmath79 dex , with [ si / fe ] @xmath80 dex @xcite and [ ca / fe ] @xmath81 dex , with [ cr / fe ] and [ mn / fe ] somewhat high and [ sr / fe ] somewhat low for its fe - metallicity .
cs22966@xmath4043 , however , has @xmath2 = 7200
k. it is a sx phe variable and a binary .
it shows rotation , with @xmath82 km s@xmath43 @xcite , not uncommon among stars in the @xcite sample of blue metal poor stars .
it may be a blue straggler , the outcome today of extensive past mass transfer within the binary system .
@xcite attribute its anomalies to local differences in the chemical history within different regions of the galactic halo , presumably arising from accretion of one or more dwarf satellite galaxies .
we assume , perhaps incorrectly , that whatever may be causing its anomalies is not directly relevant to those of the emp giant he1424@xmath40241 .
the second of the two outliers is he0132@xmath42429 , with [ fe / h ] @xmath83 dex .
the spectrum of this star indicates moderately high [ c / fe ] , very high [ n / fe ] , and high [ sc / fe ] .
we suggest that this is the result of a major contribution to its chemical inventory from a snii with a higher than typical mass . @xcite
reproduce the general nature of these anomalies with a massive snii with @xmath84 ( see their fig .
although this calculation was carried out for solar metallicity , we take it as applying at least partially to extremely metal poor snii
. we could not locate any published snii models with nucleosynthetic yields for a large set of isotopes for such massive extremely metal poor stars ; the published grids typically end at 40@xmath85 , with some studies , for example @xcite , then jumping to treat very massive pair instability sn with @xmath86 , omitting the mass range of `` normal '' snii with @xmath87 .
the very recent calculations of @xcite for nucleosynthesis in pop iii snii explosions end at @xmath88 .
the peculiarities in abundance ratios found in he0132@xmath42429 are reminiscent of those found in the most fe - poor star known , he1327@xmath42326 @xcite , which also has n highly enhanced ( with c enhanced as well , but not by as much ) and high sr relative to fe .
the sc ii lines are too weak to be detected in such an extreme star even if the same anomaly were present for this element as well .
he1356@xmath40622 is a modest outlier in [ na / fe ] and in [ si / fe ] , being high in both cases by perhaps @xmath70.5 dex .
however , [ na / fe ] shows a definite range among emp giants @xcite , which @xcite subsequently explained as mixing of material processed through proton - burning along the rgb , thus enhancing na by @xmath70.5 dex ( see also * ? ? ?
* for a discussion of non - lte effects . ) .
since he1356@xmath40622 is one of the coolest stars in our sample of emp giants , it presumably is among those with the highest luminosity and thus has a high probability of being a mixed star .
the apparent anomaly in [ si / fe ] is due primarily to the extremely large deficiency of that ratio in he1424@xmath40241 , which couples with our adopted definition of `` anomalous '' via a median over our small sample of c - normal emp giants .
after careful consideration , we find that he1356@xmath40622 is probably a mixed emp giant and has no statistically significant anomalies in its abundance ratios for the set of elements we have detected .
the three more c - rich stars in the present sample of emp giants all obey patterns previously seen for such stars ( see , e.g. cohen _ et al . _ 2005 or aoki _
et al . _
two of them show strong enhancements of the light elements up to and including al ; one does not .
there are two stars in our sample of emp giants with [ n / c ] @xmath89 , one of which is very highly n - enhanced .
cs 22949 - 037 ( a.k.a he2323@xmath40256 ) , is a n - rich star , with [ n / fe ] + 2.16 dex , while [ c / fe ] is is + 0.97 dex .
he0132@xmath42429 is a milder case of a star with [ n / c ] @xmath89 .
several similar stars are known , although emp c - rich stars with [ n / c ] @xmath90 are much more common than emp giants with [ n / c ] @xmath89 .
if we assume that the c - rich emp giants are the result of mass transfer across a binary system when the former primary was an agb star , then predictions of nucleosynthesis in agb stars ( lattanzio 1992 , herwig 2004 , and references therein ) suggest that hot bottom burning in intermediate mass agb stars ( 3 to 6 @xmath85 ) leads to strong n - enhancements @xcite discuss the predicted frequency of n - enhanced stars with [ n / c ] @xmath89 as a function of mass of the agb star contributing .
they suggest that n - rich stars represent the contribution from the upper mass limit of such stars near @xmath76 m , and note that their observed frequency appears to be quite low compared to that expected for a normal mass distribution of agb stars .
they speculate that factors which decrease the efficiency of mass transfer in binary systems with large mass ratios may be responsible for the apparent lack of n - enhanced stars .
to summarize the situation as we view it , he1424@xmath40241 and he0132@xmath42429 , both analyzed here , are the only emp giants known to us which show peculiar abundance ratios among the fe - peak elements .
one emp dwarf , he2344@xmath42800 with [ fe / h ] @xmath91 dex , first studied in the keck pilot project , was a suspected outlier , with [ cr / fe ] @xmath92 dex higher than typical emp dwarfs , a large sample of which were studied in @xcite .
analysis of a new hires spectrum of this emp dwarf taken in oct .
2004 confirms the excess in [ cr / fe ] , and suggests an excess in [ mn / fe ] of @xmath93 dex as well .
many emp stars show unexpectedly high cno abundances , which are likely due to intrinsic production followed by mixing ( for luminous giants only ) or pollution from a former agb binary companion . a smaller number of stars , including two of the three c - rich stars studied here , show large enhancements of the light elements na , mg and al as well .
all such @xmath78-enhanced stars with the exception of bs16964@xmath4002 , very recently discovered by @xcite , are c - rich ; aoki s new star is , however , highly o - rich .
a few stars , such as cs22952@xmath4015 @xcite and cs22169@xmath4035 @xcite show small ( at least compared to those of the present sample ) deficiencies of the @xmath78-elements , with normal c. with better spectra and analyses , and the larger sample of known emp stars enabled in part by our searching for such in the hamburg / eso survey , we are now able to discern the impact on the chemical inventory of a star from contributions by individual snii among extreme emp stars . at slightly higher fe - metallicity , we see abundance ratios which show slow trends as functions of [ fe / h ] with low dispersions about the mean trends which presumably arise from summing the ejecta of snii over a stellar population with a normal ( i.e. salpeter or similar ) initial mass function .
these trends can often be reproduced in detail by theoretical models of galactic chemical evolution containing the most recent nucleosynthetic yields such as those of @xcite or @xcite .
it is now up to the theorists who model snii explosions to try to develop a set of nucleosynthesis yields which will lead to the variety of chemical inventories we have seen in the emp stellar population in the milky way , particularly among the lowest [ fe / h ] stars known in the galaxy , and especially for the very anomalous extreme emp star he1424@xmath40241 . turning to the elements beyond the fe - peak ,
another by now well established observational fact is the decoupling between the production of the fe - peak elements and the heavy neutron capture elements
. the ratios of [ sr / fe ] and [ ba / fe ] among the emp stars analyzed by our 0z project , including those discussed here , show a very wide range among both the c - normal and c - rich emp stars ( see fig . [ figure_sr_ba ] ) .
it is interesting to note that no star in the present sample , neither c - normal nor c - rich , has [ ba / fe ] @xmath94 dex . among more metal - rich c - rich stars ,
the fraction of stars with enhancements of the @xmath54-process elements is large , exceeding 75% ( see , e.g. * ? ? ?
the fraction of c - enhanced stars among emp stars is a very contentious issue . adopting the definition of c - enhanced stars of @xcite as those with [ c / fe ] @xmath95 dex , recent values for this fraction from several independent survey for stars with [ fe / h ]
@xmath96 dex cover the range from @xmath970.2% @xcite to 9@xmath982% @xcite , with our 0z survey yielding a preliminary value of 14@xmath984% @xcite .
the samples are in each case reasonably large , but there is a fairly large range in the deduced frequency of c - rich emp stars .
this suggests that differences in the analysis are contributing to this problem .
it is not surprising in the context of the previous discussion ( see , for example , that of [ section_heres ] ) that the fraction of c - rich stars calculated from a set of different independent analyses of large samples would result in different estimates from different surveys .
our [ fe / h ] values are systematically higher than those of @xcite by @xmath70.3 dex , and the differences in c and n abundances determined from molecular bands between the two surveys show a larger dispersion than do the abundances based on atomic absorption lines .
stars near the boundary of the c - enhanced class could easily be shifted into or out of the c - rich class as a result of small systematic differences between the various ongoing large projects .
this would affect such frequency calculations , whatever the specific abundance characteristic of interest might be .
since there are many stars near the boundary of the c - enhanced class as defined above , we suggest that a substantial part of the variation in the deduced fraction of c - enhanced emp stars arises from such differences in the details of the analysis .
it is thus extremely important for people engaged in this type of work to publish the full details of their analyses , as we did in @xcite , and to analyze a few stars in common with other major groups working in this area . among the details that must be described , in addition to those discussed above ,
is the issue of the way one handles the recent upheaval in the solar cno abundances through the work of @xcite and @xcite with 3d models .
the modified definition of the cutoff for enhanced [ c / fe ] suggested by @xcite offers the advantage of a cleaner cut with fewer stars near the boundary between c - normal and c - enhanced classes .
it is preferable to the definition we are using , which is that of @xcite . with the latter definition ,
two of the three stars are just at the boundary , while with the newer definition , all three of the c - rich stars discussed here would clearly be considered c - rich .
we have presented detailed abundance analyses of five extremely metal poor giants which are newly discovered from our datamining of the hamburg / eso survey .
one of these turned out to be a rediscovery of a star found in the hk survey with an unusually large error in its published coordinates .
we include here a new analysis based on better spectra of he0132@xmath42429 , part of the keck pilot project ( see cohen et al 2002 and carretta et al 2002 ) .
we also analyze new high resolution and high signal - to - noise ratio spectra of the only emp giant found in the heres project @xcite and of an emp giant from the first stars project to use as a calibration object for comparison of the two projects , for a total sample of 8 emp giants .
the high quality of our hires spectra and our discovery of many more emp stars , including some close to @xmath99 dex , makes it possible to search for , to find , and to confirm outliers which have anomalous abundance ratios [ x / fe ] among the fe - peak elements , where such have not been detected previously .
the lowest metallicity star in our sample , he1424@xmath40241 , with [ fe / h ] @xmath100 dex , has only upper limits for the c and n abundances , based on our non - detection of the g band of ch and of the 3360 band of nh .
this star shows highly anomalous abundance ratios , with extremely low [ si / fe ] ( @xmath101 dex ) and very low [ ca / fe ] ( @xmath102 dex ) and [ ti / fe ] ( @xmath103 dex ) , while [ mg / fe ] is normal .
mg is produced in hydrostatic @xmath78-burning , while the other three elements are made in explosive @xmath78-burning . in essentially all other emp stars , these three abundance ratios are positive .
these deficits in he1424@xmath40241 are accompanied by strong excesses for the odd atomic number elements , so that [ mn / fe ] and [ co / fe ] are significantly larger than is typical of all other emp giants . we speculate that the parcel of gas from which this star formed in the early milky way contained ejecta from only a few snii , and was deficient in ejecta from core collapse sn whose progenitors had masses at the upper end of the relevant range .
the nucleosynthesis was such that the explosive @xmath78-elements ( si , ca and ti ) were not produced by the snii at typical rates , while the hydrostatic @xmath78-element mg , formed during the course of normal stellar evolution even in zero metallicity stars , was produced at normal rates .
a second outlier , he0132@xmath104 , shows enhanced sc relative to fe , with [ n / c ] @xmath105 .
we suggest that this chemical inventory of this star had the opposite bias , namely a larger contribution from snii toward the upper end of the progenitor mass range near 60@xmath85 .
the remaining three c - normal stars have abundance ratios typical of slightly more metal rich emp stars ; they do not show any detectable anomalies in their chemical inventory .
no other emp giant known to us shows peculiar abundance ratios among the fe - peak elements .
one emp dwarf , he2344@xmath42800 , first studied in the keck pilot project , and now revisited with a better hires spectrum , appears to have [ cr / fe ] @xmath92 dex and [ mn / fe ] of @xmath93 dex higher than typical emp dwarfs , a large sample of which were studied in @xcite .
c and n have only upper limits in this hot dwarf , where the molecular bands are , even for normal [ c / fe ] ratios , very weak and difficult to detect .
many emp stars show unexpectedly high cno abundances , which are likely due to intrinsic production followed by mixing ( for luminous giants only ) or pollution from a former agb binary companion . a smaller number of stars , including two of the three c - rich stars studied here , show large enhancements of the light elements na , mg and al as well .
all such @xmath78-enhanced stars with the exception of bs16964@xmath4002 , very recently discovered by @xcite , are c - rich , and this star is highly o - rich .
the behavior of the c - rich stars contains no surprises .
two in our present sample have large enhancements of the light elements through al ; the third shows normal abundance ratios for na , mg and al with respect to fe .
two of the sample stars have [ n / c ] @xmath89 , which is not common among c - rich emp giants .
it is generally believed that c - rich emp stars are the result of mass transfer within a binary system when the former primary was an agb star .
if this is true , nucleosynthesis calculations by @xcite , @xcite and references therein suggest that to produce such a large excess of n the former primary must have a mass towards the upper end of the range for agb stars .
we present the first determination of [ cu / fe ] for emp giants below [ fe / h ] @xmath106 dex based on the rarely measured uv resonance lines of cu i near 3250 .
we find that the plateau level which was suggested for [ cu / fe ] for dwarfs with [ fe / h ] @xmath107 dex by @xcite continues to even lower metallicities .
the heavy neutron capture elements are low in all eight emp stars in our sample , with [ ba / fe ] @xmath59 dex .
this is rare among more fe - rich c - stars , which often have strong @xmath54-process enhancements .
the ratio @xmath26(ba)/ @xmath26(sr ) varies by more than a factor of 100 among the stars studied here , and suggests again that another nucleosynthesis mechanism that preferentially produces the light neutron capture elements such as sr is required .
a careful comparison of the procedures and results of the detailed abundance analyses carried out by our 0z project with those of the first stars project @xcite demonstrates that there is a systematic offset between the deduced [ fe / h ] values ; ours being on average 0.3 dex higher .
thus the most metal poor star we found , he1424@xmath40241 , with [ fe / h ] @xmath100 dex based on our 0z project analysis , would translates roughly into @xmath108 dex if it were analyzed by the first stars project of @xcite . for elements which display measurable atomic absorption features , the differences between these two projects results in systematic changes in abundance ratios with respect to fe which are much smaller .
inter - comparison between surveys of abundances [ x / fe ] derived from molecular bands , typically used to determine cno abundances , is much less common and more difficult to carry out .
systematic differences between surveys for cno abundances from ch , nh , cn , co or oh bands may be common and may be larger than for elements where atomic lines can be utilized .
any such systematic differences between surveys may affect the deduced fraction of c - rich stars found in a survey and may contribute to the wide range in published values for this parameter .
we have observed @xmath7100 emp candidates with andicam in queue mode over the past three years .
we calibrate to the johnson - kron - cousins photometric system using standard star fields from @xcite .
the observer at ctio running the queue only carries out our program if the night is believed to be photometric .
data from nights which at sunset were believed to be photometric , but subsequently the observer changed his opinion of the sky conditions , were discarded .
the zero points of our photometric calibration are based on two sets of images of standard star fields per night in almost all cases , and never more than two .
most stars were observed only on one night ; about 1/3 were observed on two nights , and a few on three nights .
we assess whether our assigned photometric errors for andicam photometry of the he s candidate emp stars are valid by comparison of stars from our sample which are included in much larger , and hopefully better calibrated , photometric surveys .
this is a key issue since we use this photometry , together with @xmath109 from 2mass , to determine stellar parameters .
we compare our values with those from the sdss @xcite dr4 release @xcite , using the transformations of @xcite for johnson @xmath17 and kron - cousins @xmath18 .
there are 26 stars in common .
the mean difference in @xmath17 and in @xmath18 between that we measure using andicam and that from the sdss database appropriately transformed is less than 0.01 mag .
the dispersion about the mean is 0.06 mag for each of @xmath17 and @xmath18 , somewhat larger than one might expect for our nominal errors of @xmath110 mag for each combined with the nominal uncertainties of @xmath111 mag for the sdss .
this suggests that our assessment of the uncertainty in our andicam photometry may be somewhat underestimated .
the sdss is a very large area survey with a very extensive photometric calibration effort , and our andicam measurements agree well in the mean with the sdss values , suitably transformed .
this suggests that the photometric calibration of our andicam data has no systematic errors .
the recent large photometric survey of emp candidate stars of @xcite includes 17 stars from our andicam sample .
@xcite include data taken at many sites with many different instruments on many different runs .
for these 17 stars , our @xmath17 is fainter by 0.05 mag on average , with @xmath44 of the differences being 0.05 mag .
our @xmath112 colors are identical to within @xmath63 mag on average with those of @xcite for the 12 stars with such colors in the survey of @xcite , with @xmath44 of the difference being 0.03 mag .
the ( identical ) systematic difference in @xmath17 and in @xmath18 may arise from calibration difficulties across a survey with such a wide variety of data sources .
he1424@xmath40241 , the most metal - poor star discussed here , has @xmath17(andicam , sdss , beers ) = 15.47 , 15.36 , 15.32 mag respectively , and is among the three stars with the largest deviation in @xmath17 for both of these comparisons .
its @xmath18 mag is the same ( differing by only 0.015 mag ) for the sdss and for the andicam data , while the @xcite photometry for this star has @xmath112 identical to the andicam result , but @xmath17 brighter by 0.15 mag .
if we assume the sdss @xmath113 to be correct , then we have underestimated @xmath2 for this star by @xmath7130 k , the nominal uncertainty we adopt for this key stellar parameter is 100 k.
in this appendix we provide additional details , beyond the discussion given in [ section_comp ] , of the comparison of the analyses we have carried out for emp giants with those of the first stars project as given in @xcite .
we selected a representative sample of stars from @xcite to cover the range of stellar parameters of interest here ( [ fe / h ] below @xmath106 dex and log(@xmath19 ) as expected for rgb stars ) .
three values of @xmath2 determined for each of these stars are shown in table [ table_teff_cayrel ] .
the first is the value we would derive using the codes and procedures of the 0z project , including reddening from the map of @xcite .
the second is that we would derive if we used @xmath114 from @xcite instead .
the third is that adopted by the first stars project , taken from table 4 of @xcite . in all cases ,
the @xmath17 mag given in table 2 of @xcite was used .
@xmath2 for the second and third cases are given in the table as differences from that of our 0z project . at first sight
the differences between the @xmath2 adopted by each project for the representative sample of stars given in the last column of table [ table_teff_cayrel ] are satisfactory considering the quoted accuracies of @xmath2 by us of @xmath115 k and by the first stars project of @xmath6680 k. however , we stress that this is a systematic , not a random , effect . table [ table_teff_cayrel ] shows , by comparing the first and second values of @xmath2 given for each star , that the somewhat larger values of @xmath114 that we adopt result in our @xmath2 being higher than that of the first stars project by up to @xmath7100 k. this difference depends both on the reddening difference for the particular star ( i.e. where the star is in the sky ) and on @xmath2 for the star , as hotter stars have smaller values of @xmath116@xmath2 ) .
in addition , for a specified set of stellar colors and a fixed choice of reddening , comparing the second and third choices for @xmath2 , we find that our @xmath2 are higher by up to 100 k. this must arise from some systematic difference in the transformation within the grid of colors ( specifically @xmath117 and @xmath16 ) and the stellar parameters for emp giants . it should be noted that the @xmath2 given by @xcite are based on the @xcite relations while ours are based on those of the marcs and atlas model atmosphere grids .
this issue was discussed in @xcite , where it was demonstrated that the marcs and atlas9 @xmath2 color - relations are in very good agreement , but that they disagree with those of @xcite .
we also note that the @xcite color relations have not been calibrated adequately at the very low values of [ fe / h ] considered here ; a theoretical calibration is easier to achieve at present than an observational one for emp stars .
the result of adding these two effects is that our @xmath2 would be from 30 to 165 k hotter than those adopted by @xcite for the same input set of observed colors , with the hotter stars in the sample ( @xmath2@xmath118 k ) having larger differences than the cooler ones ; the coolest stars in our sample have @xmath2@xmath119 k. this of course will result in a systematic offset such that the [ fe / h ] value deduced for a star with a particular set of @xmath31 will be higher as determined by the 0z project than as determined by the first stars project @xcite .
this is accompanied by a smaller effect on the deduced abundance ratios , as we will see below . since @xcite use ionization equilibrium to determine log(@xmath19 ) , given a @xmath2 , while we use evolutionary tracks , there is some scatter of the difference of log(@xmath19)between the two projects for a given star . in most cases
the difference is small ; the largest difference among the stars compared in table [ table_teff_cayrel ] is 0.4 dex for the same input parameters .
we next isolate the differences introduced into the abundance ratios by the different abundance analysis codes and model atmosphere grids between our 0z project and the first stars project .
recall that we use model atmospheres from atlas @xcite while the first stars project uses osmarcs models @xcite .
we first note that changing to the latest and currently preferred kurucz models which have a somewhat different treatment of convection with no overshooting described in @xcite and an improved opacity distribution function ( which are labeled as `` odfnew '' models ) leads to very small changes ( not exceeding 0.01 dex ) in the derived log[@xmath26(x ) ] for all species considered here .
the predicted model colors @xmath117 and @xmath16 differ by less than 0.02 mag in this range of [ fe / h ] and @xmath2 , introducing a change of less than 50 k in deduced @xmath2
. since these stars are so metal - poor , the details of the treatment of the line opacity used in computing the model atmosphere do not matter .
for this test , we assume , for a given star , the same stellar parameters as @xcite derived . we then carry out an abundance analysis using our codes and our model atmospheres appropriate to those input stellar parameters , and compare the resulting [ fe / h ] values and abundance ratios [ x / fe ] .
first we compare the deduced values of log[@xmath26(x ) ] for three stars .
the first case we examined is the bright emp giant cd @xmath120 245 .
we use the @xmath31and @xmath22 values from @xcite ; we do not have any hires spectra of this star . for the other two stars , bs16467@xmath4062 and cs22949@xmath4037 ( a.k.a .
he2323@xmath40256 ) we have independent hires spectra ; we use our own measured @xmath31 and our own adopted atomic parameters . the comparison for the three stars is shown in fig .
[ figure_cayrel_diff ] .
this figure shows our absolute abundances being systematically about 0.15 dex larger for all species .
this difference is similar to that found by @xcite ( see their fig . 1 )
, who have carried out a similar test for stellar parameters appropriate to cepheid variables .
this systematic offset would be largely eliminated if we were to use a stellar atmosphere which was @xmath790 k lower than the @xmath2 adopted by the first stars project .
the agreement for the c and n abundances , with differences as large as 0.4 dex , is not as good .
this is not surprising , as we have shown in
[ section_comp ] that the @xmath22 values for atomic absorptoin lines adopted by both projects are in very good agreement .
the c and n abundances , on the other hand , are derived from molecular bands and many more parameters enter , none of which have been compared between the two projects .
the offset from equality shown in fig .
[ figure_cayrel_diff ] does not affect the deduced abundance ratios [ x / fe ] derived from our work and the first stars project as the offset is approximately constant for all species considered .
however the first stars project adopts log@xmath26(fe ) = 7.50 dex for the sun , while we use a value 0.05 dex smaller .
since , as we have shown earlier , the @xmath22 value scales are identical , we will therefore see a systematic difference of 0.05 dex in all values of [ x / fe ] and in [ fe / h ] , with our values being higher ( i.e. more metal - rich ) .
this is in addition to the offset of 0.15 dex due to differences in the model atmosphere grids and analysis codes . for cd@xmath438 245
both analyses give very good ionization equilibrium , so it matters little whether one uses fe i or fe ii as the reference .
ideally , since log@xmath121feii ) is less sensitive to @xmath2 , it would be better to use that in calculating [ fe / h ] and [ x / fe ] , but there are few fe ii lines detected in these emp stars , and many of those detected are very weak with somewhat unreliable @xmath31 .
the differences in [ x / fe ] for the 12 species in common ( ignoring the reference species , fe i ) range from @xmath65 to + 0.05 dex , with a mean of @xmath122 dex and @xmath44 of 0.04 dex , which we consider very good agreement .
the difference in the deduced abundance ratios between our 0z project and the first stars project for he2323@xmath40256 , ignoring n ( deduced from the nh band at 3360 in both cases ) , ranges from @xmath103 to + 0.11 dex for the 11 species in common , with a mean of @xmath65 dex and @xmath44 0.10 dex .
( that for [ n / fe ] is 0.38 dex , with our value being lower . )
part of this may arise in the problem with the @xmath31 for this star ; incorrect values were published by @xcite for at least 7% of the lines in common with our 0z values given here in table [ table_eqw ] ( m. spite , private communication , july 2007 ) . the ionization equilibrium of fe is good in both analyses .
bs 16467062 , the third case we checked , gives similar results . here , with the stellar parameters set by the first stars project , the difference in log@xmath26(fe i ) is 0.16 dex , with our value again being higher .
the ionization equilibrium in our solution for this set of stellar parameters is log(@xmath26)(fe i )
log(@xmath26)(fe ii ) + 0.04 dex ; for the first stars project derived + 0.12 dex . but
the true comparison is what happens when we use the stellar parameters derived with our own codes and procedures and the reddening values we adopt ( i.e. our systematically higher @xmath2 as compared to those of the first star project ) . for bs 16467062 ,
if we compare log@xmath26(fe i ) as published by @xcite versus that given in table [ table_abunda ] and [ table_abundb ] , the difference in [ fe / h ] 0.30 dex , with our value being higher .
the higher @xmath2 we adopt ( 164 k higher than that of the first stars project ) based on our higher reddening and on our @xmath2 scale is somewhat compensated by the difference in adopted log(@xmath19 ) . the difference in adopted @xmath2 for cs22949 - 037 ( a.k.a .
he2323@xmath40256 ) between the two projects is small only because our @xmath17 mag from andicam photometry is 0.05 mag fainter than that adopted by @xcite .
we thus find a difference of + 0.18 dex in the final deduced [ fe / h ] , our value being higher .
this , given the essentially identical @xmath2 , reflects just the difference in analysis details discussed above . in summary
, it appears that the different analysis codes , and stellar atmospheres grids adopted lead to [ fe / h ] values from our 0z project being systematically 0.15@xmath123 dex higher than those of the first stars project .
the difference becomes somewhat larger , @xmath70.25 dex , when the hotter stellar parameters determined from the independent codes , procedures , and adopted reddening map of our 0z project are used instead of those adopted by @xcite .
there is an additional contribution of 0.05 dex to the difference in derived [ fe / h ] from the two projects which arises from the difference in the adopted solar fe abundance .
overall , the [ fe / h ] value derived by the first stars project as described in @xcite will be systematically @xmath70.3 dex lower than that for the same star , the same set of @xmath31 , and the same set of observed stellar photometry as analyzed by the 0z project .
we are grateful to the many people who have worked to make the keck telescope and hires a reality and to operate and maintain the keck observatory .
the authors wish to extend special thanks to those of hawaiian ancestry on whose sacred mountain we are privileged to be guests . without their generous hospitality , none of the observations presented herein would have been possible .
some of the data presented herein were obtained at the palomar observatory .
this publication makes use of data from the two micron all - sky survey , which is a joint project of the university of massachusetts and the infrared processing and analysis center , funded by the national aeronautics and space administration and the national science foundation .
is grateful to nsf grant ast-0507219 for partial support .
n.c . is a research fellow of the royal swedish academy of sciences supported by a grant from the knut and alice wallenberg foundation .
he also acknowledges financial support from deutsche forschungsgemeinschaft through grants ch 214/3 and re 353/44 .
we thank w. huang for help accessing the sdss database .
aoki , w. , norris , j. e. , ryan , s. g. , beers , t. c. , christlieb , n. , tsangarides , s. & ando , h. , 2004 , , 608 , 971 aoki , w. _ et al .
_ , 2005 , , 632 , 611 aoki , w. , beers , t. c. , christlieb , n. , norris , j. e. , ryan , s. g. & tsangarides , 2007 , , in press asplund , m. , grevesse , n. , sauval , a. j. , allende prieto , c. & kisselman , d. , 2004 , , 417 , 751 asplund , m. 2005 , , 43 , 481 gustafsson , b. , edvardsson , b. , eriksson , k. , mizuno - weidner , m. , jrgensen , u. g. & plez , b. , 2002 , in asp conf.ser.288 , _ stellar atmospheres modeling _ , ed . i. hubeny , d. mihalis & k. werner , ( san francisco , asp ) , 331 huber , k. p. & herzberg , g. , 1979 , _ constants of diatomic molecules _ , ( new york , van nostrand ) ivans , i. i. , sneden , c. , renee james , c. , preston , g. w. , fulbright , j. p. , hoflich , p. a. , carney , b. w. & wheeler , j. c. , 2003 , , 592 , 906 mcwilliam , a. , 1998 , , 115 , 1640 mishenina , t. v. , kovtyukh , v. v. , soubiran , c. , travaglio , c. & busso , m. , 2002 , , 396 , 189 nomoto , k. , tominaga , n. , umeda , h. , kobayashi , c. , & maeda , k. 2006 , nuclear physics , a777 , 424 ( see also astro - ph/0605725 ) plez , b. , cohen , j. g. & melendez , j. , 2005 , in iau symposium 228 , _ from lithium to uranium : elemental tracers of early stellar evolution _ , ed . v. hill , p. francois & f. primas , cambridge university press , pg .
267 simmerer , j. , sneden , c. , ivans , i. i. , kraft , r. p. , shetrone , m. a. & smith , v. v. , 2003 , , 125 , 2018 simmerer , j. , sneden , c. , cowan , j. j. , collier , j. , woolf , v. m. & lawler , j. e. , 2004 , , 617 , 1091 skrutskie , m. f. _ et al .
_ , 2006 , , 131 , 1163 sneden , c. , _ et al .
_ , 2003 , , 591 , 936 spite , m. _ et al .
_ , 2005 , , 430 , 655 spite , m. _ et al .
_ , 2006 , , 455 , 291 takeda , y. , zhao , g. , takada - hidai , m. , chen , y. q. , saito , y. & zhang , h. w. , 2003 , chinese jrl astron & astrophys , 3 , 316 wallace , l. , hinkle , k. & livingston , w. c.,1998 , `` an atlas of the spectrum of the solar photosphere from 13,500 to 28,000 cm@xmath43 '' , n.s.o .
technical report 98 - 001 , ftp://nsokp.nso.edu/pub/atlas/visatl .
lclr rrrr he0132@xmath42429 & 01 34 58.8 @xmath125 24 18 & 14.82 & & 0.012 & 5294 & 2.75 & 1.8 + he1012@xmath41540 & 10 14 53.5 @xmath415 55 54 & 14.04 & 13.21 & 0.070 & 5620 & 3.40 & 1.6 + he1347@xmath41025 & 13 50 22.4 @xmath410 40 19 & 15.06 & 14.16 & 0.058 & 5195 & 2.50 & 1.8 + he1356@xmath40622 & 13 59 30.3 @xmath406 36 35 & 14.31 & 13 36 & 0.030 & 4947 & 1.85 & 2.2 + he1424@xmath40241 & 14 26 40.3 @xmath402 54 28 & 15.47 & 14.54 & 0.064 & 5193 & 2.50 & 1.8 + he2323@xmath40256 & 23 26 29.8 @xmath402 39 58 & 14.41 & 13.40 & 0.051 & 4915 & 1.70 & 2.0 + from other surveys : + he1300 + 0157 & 13 02 56.3 + 01 41 51 & 14.11 & 13.39 & 0.022 & 5632 & 3.37 & 1.3 + bs16467@xmath4062 & 13
42 00.6 + 17 48 48 & 14.09 & & 0.018 & 5364 & 2.95 & 1.6 + lllrr r he0132@xmath42429 & 7200 & 289.89 & 95 & 0.86 & + 289.2 + he1012@xmath41540 & 3600 & 489.77 & 109 & 0.86 & + 226.2 + he1300 + 0157 & 3600 & 843.87 & @xmath60110 & 1.1 & + 73.4 + he1347@xmath41025 & 3600 & 149.80 & 80 & 1.1 & + 48.6 + he1356@xmath40622 & 3600 & 149.89 & 105 & 1.1 & + 93.5 + he1424@xmath40241 & 6000 & 844.96 & 90 & 1.1 & + 60.4 + he2323@xmath40256 & 7200 & 312.76 & 100 & 0.86 & @xmath126 + bs16467@xmath4062 & 3600 & 490.02 & 100 & 0.86 & @xmath127 + lllll he0132@xmath42429 & 1811.96 & 3243.00 & 3289.89 + & + 296.0 ( 0.1 ) & + 289.5 ( 0.2 ) & + 289.2 ( 0.1 ) + he1012@xmath41540 & 2396.82 & 3489.77 + & + 226.3 ( 0.4 ) & + 224.8 ( 0.1 ) + he1300 + 0157 & 2830.50 & 3157.74 & 3432.01 & 3843.87 + & + 73.6 ( 2.0 ) & 74.6 ( 0.6 ) & 74.6 ( 0.6 ) & + 73.4 ( 0.1 ) + he1346@xmath41025 & 3149.80 & 4198.92 + & + 48.6 ( 0.5 ) & + 49.4 ( 0.1 ) + he1356@xmath40622 & 3149.89 & 3989.72 + & + 93.5 ( 0.4 ) & + 94.1 ( 0.2 ) + he1424@xmath40241 & 3152.90 & 3844.96 + & + 58.8 ( 0.6 ) & + 60.4 ( 0.1 ) + he2323@xmath40256 & 1764.73 & 2158.73 & 2544.90 & 3312.76 + & @xmath4125.6 ( 0.1 ) & @xmath4125.6 ( 0.1 ) & @xmath126 ( 0.1 ) & @xmath126 ( 0.1 ) + bs16467@xmath4062 & 2064.55 & 2095.45 & 3490.02 + & @xmath490.6 ( 0.1 ) & @xmath490.5 ( 0.1 ) & @xmath127 ( 0.1 ) + l crrr rrrr rrrr 3189.30 & oh & 1.03 & @xmath41.990 & @xmath12821.2 & & & & & & & + 3255.50 & oh & 1.30 & @xmath41.940 & @xmath12814.3 & & & & & & & 40.0 + 5889.95 & na i & 0.00 & 0.110 & 62.8 & 168.8 & 48.2 & 132.8 & 64.9 & 65.7 & 50.1 & 192.7 + 5895.92 & na i & 0.00 & @xmath40.190 & 50.4 & 146.7 & 26.8 & 117.3 & 52.2 & 39.0 & 27.8 & 159.2 + 3829.36 & mg i & 2.71 & @xmath40.210 & 93.3 & 259.0 & 84.2 & & & 79.0 & 81.8 & 139.7 + 4057.52 & mg i & 4.34 & @xmath41.200 & & 22.0 & & 9.0 & & & & 22.6 + 4167.28 & mg i & 4.34 & @xmath41.000 & & & & 10.5 & & & 6.8 & 28.7 + 4703.00 & mg i & 4.34 & @xmath40.670 & 9.0 & 72.2 & & 24.4 & 16.0 & & & 47.5 + 5172.70 & mg i & 2.71 & @xmath40.380 & 101.3 & 301.2 & 92.6 & 128.2 & 110.5 & 89.7 & 92.7 & 160.6 + 5183.62 & mg i & 2.72 & @xmath40.160 & 114.3 & 391.5 & 103.7 & 142.2 & 120.8 & 104.7 & 105.2 & 179.9 + 5528.40 & mg i & 4.34 & @xmath40.480 & 12.1 & 59.3 & & & & & & 48.3 + 3944.01 & al i & 0.00 & @xmath40.640 & 39.1 & & & 61.5 & 58.5 & 23.0 & 34.2 & 123.8 + 3961.52 & al i & 0.00 & @xmath40.340 & 42.7 & 103.0 & 36.4 & 76.8 & 62.5 & 27.1 & 37.1 & 89.7 + 3905.53 & si i & 1.91 & @xmath41.040 & 110.5 & 133.7 & 93.5 & 141.7 & 116.5 & 16.0 & 93.4 & 150.7 + 4102.94 & si i & 1.91 & @xmath43.140 & & & & 41.3 & 8.0 & & & 13.7 + 4226.74 & ca i & 0.00 & 0.240 & 105.7 & & 94.1 & 134.0 & 111.1 & 58.7 & 98.8 & 116.6 + 4289.37 & ca i & 1.88 & @xmath40.300 & & & & 16.9 & 9.8 & & & + 4302.54 & ca i & 1.90 & 0.280 & 22.6 & & & 39.5 & 32.4 & & 12.0 & + 4318.66 & ca i & 1.90 & @xmath40.210 & & & & 16.3 & & & & + 4425.44 & ca i & 1.88 & @xmath40.360 & & & & 14.0 & 9.0 & & & + 4435.69 & ca i & 1.89 & @xmath40.520 & & 7.3 & & 13.0 & & & & + 4454.79 & ca
i & 1.90 & 0.260 & 17.4 & 21.3 & 15.9 & 36.2 & 26.7 & & 17.2 & 82.0 + 3736.90 & ca ii & 3.15 & @xmath40.148 & 75.0 & 50.9 & 69.7 & & & 44.0 & 70.8 & + 4246.82 & sc ii & 0.32 & 0.242 & 79.3 & & 30.9 & 87.6 & 43.3 & 20.5 & 43.3 & 63.8 + 4314.08 & sc ii & 0.62 & @xmath40.100 & 43.1 & & & 49.5 & 20.0 & & 13.4 & + 4320.73 & sc ii & 0.60 & @xmath40.260 & 33.6 & & & 36.8 & & & 11.1 & + 4670.41 & sc ii & 1.36 & @xmath40.580 & & & & 7.5 & & & & + 3958.22 & ti i & 0.05 & @xmath40.160 & 15.3 & & & 18.9 & & & & @xmath12913.5 + 3998.64 & ti i & 0.05 & @xmath40.050 & 13.5 & & & 24.7 & @xmath12925.0 & & 9.6 & + 4533.25 & ti i & 0.85 & 0.480 & & & & 14.0 & @xmath12916.0 & & & + 4534.78 & ti i & 0.84 & 0.280 & & & & 11.0 & & & & + 4981.74 & ti i & 0.85 & 0.500 & 12.0 & & & 18.0 & @xmath12912.0 & & 4.9 & + 4999.51 & ti i & 0.83 & 0.250 & & @xmath1296.0 & & & @xmath12910.0 & & & + 3900.54 & ti ii & 1.13 & @xmath40.450 & 52.2 & 13.6 & 27.9 & 76.5 & 54.1 & 14.0 & 41.6 & 61.1 + 3987.61 & ti ii & 0.61 & @xmath42.730 & & & & 5.0 & & & & + 4012.39 & ti ii & 0.57 & @xmath41.610 & 24.4 & & & 46.8 & 23.7 & 8.5 & 20.3 & 31.4 + 4028.35 & ti ii & 1.89 & @xmath40.870 & 5.3 & & & 14.0 & & & & + 4300.05 & ti ii & 1.18 & @xmath40.490 & 42.6 & & & 67.2 & 46.6 & 7.5 & 30.9 & + 4301.93 & ti ii & 1.16 & @xmath41.200 & 17.9 & & & 37.6 & & & 10.3 & + 4312.86 & ti ii & 1.18 & @xmath41.160 & 23.4 & & & 41.0 & 23.7 & & 8.5 & + 4395.03 & ti ii & 1.08 & @xmath40.510 & 49.5 & & 24.4 & 76.4 & 56.7 & 12.0 & 39.5 & 60.6 + 4399.77 & ti ii & 1.24 & @xmath41.290 & 19.9 & & & 31.6 & 17.5 & & 12.1 & 20.2 + 4417.72 & ti ii & 1.16 & @xmath41.160 & 23.0 & & & 38.6 & 31.3 & & 12.1 & 25.4 + 4443.81 & ti ii & 1.08 & @xmath40.700 & 44.0 & 11.8 & 20.3 & 69.8 & 51.1 & & 39.9 & 58.2 + 4468.51 & ti ii & 1.13 & @xmath40.600 & 46.8 & 19.9 & 22.1 & 74.3 & 50.3 & 10.7 & 31.2 & 53.8 + 4501.28 & ti ii & 1.12 & @xmath40.760 & 41.7 & 11.2 & 20.7 & 66.8 & 49.9 & 12.2 & 31.9 & 51.4 + 4533.97 & ti ii & 1.24 & @xmath40.640 & 37.4 & 14.0 & 15.5 & 63.5 & 42.2 & 8.3 & 24.8 & 49.3 + 4563.77 & ti ii & 1.22 & @xmath40.820 & 30.1 & 10.6 & 13.4 & 56.1 & 32.3 & 5.1 & 24.7 & 39.5 + 4571.98 & ti ii & 1.57 & @xmath40.340 & 33.4 & 13.7 & 17.3 & 58.5 & 28.9 & & 19.7 & 42.8 + 4589.95 & ti ii & 1.24 & @xmath41.650 & & & & 15.5 & & & & + 4111.77 & v i & 0.30 & 0.408 & & & & & @xmath1295.6 & & & + 3545.19 & v ii & 1.10 & @xmath40.390 & & & & & & @xmath12910.9 & 11.6 & 11.8 + 3592.03 & v ii & 1.10 & @xmath40.370 & 12.4 & & & & & & 8.2 & 11.5 + 4254.33 & cr i & 0.00 & @xmath40.110 & 31.5 & & 21.5 & 58.9 & 40.6 & 23.6 & 30.6 & 38.4 + 4274.79 & cr i & 0.00 & @xmath40.230 & 29.2 & & 16.8 & 47.3 & 32.5 & 19.7 & 22.5 & 29.3 + 4289.72 & cr i & 0.00 & @xmath40.361 & 22.8 & & & 42.3 & 26.5 & 13.4 & 20.0 & 31.8 + 5206.04 & cr i & 0.94 & 0.030 & 11.9 & 8.4 & 4.7 & 23.2 & 14.0 & 5.8 & 6.2 & 7.4 + 5208.43 & cr i & 0.94 & 0.158 & 14.3 & 8.5 & 7.6 & & & 7.7 & 12.8 & 18.2 + 4030.75 & mn i & 0.00 & @xmath40.470 & 17.3 & 20.0 & 10.8 & 36.6 & 26.0 & 34.7 & 17.4 & 22.1 + 4033.06 & mn i & 0.00 & @xmath40.620 & 12.3 & 15.0 & 8.5 & 30.0 & 18.0 & 26.3 & 14.4 & 17.9 + 3441.99 & mn ii & 1.78 & @xmath40.273 & 28.0 & 10.0 & 16.2 & & & 48.4 & 32.0 & + 3460.32 & mn ii & 1.81 & @xmath40.540 & 23.5 & & 14.3 & & & 37.9 & 20.4 & 32.3 + 3488.68 & mn ii & 1.85 & @xmath40.860 & 15.9 & & & & & 29.8 & 13.0 & + 3865.52 & fe
i & 1.01 & @xmath40.980 & 63.8 & 46.5 & 48.8 & 88.1 & 74.2 & 50.2 & 66.2 & 72.8 + 3886.29 & fe i & 0.05 & @xmath41.080 & & & & & & 90.5 & & + 3887.06 & fe i & 0.91 & @xmath41.140 & & & & & & 40.5 & & + 3895.67 & fe i & 0.11 & @xmath41.670 & 79.4 & & 69.2 & 101.9 & 83.4 & 69.4 & 78.6 & 96.4 + 3899.72 & fe i & 0.09 & @xmath41.530 & 83.7 & 70.2 & 72.9 & 115.0 & 92.1 & 73.5 & 84.6 & 95.8 + 3902.96 & fe i & 1.56 & @xmath40.470 & 55.9 & 48.0 & 45.9 & 80.6 & 70.8 & 41.4 & 56.1 & 59.7 + 3906.49 & fe i & 0.11 & @xmath42.240 & 60.7 & 36.0 & 41.8 & 92.4 & 73.9 & 47.5 & 60.7 & 68.3 + 3920.27 & fe i & 0.12 & @xmath41.750 & 83.5 & 59.9 & 66.4 & 106.4 & 82.9 & 70.2 & & 95.7 + 3922.92 & fe i & 0.05 & @xmath41.650 & 90.5 & 58.1 & 74.1 & 113.3 & 95.4 & 79.4 & 85.6 & 98.4 + 3930.31 & fe i & 0.09 & @xmath41.590 & & 75.0 & & & & 75.7 & & + 3949.96 & fe i & 2.18 & @xmath41.160 & 8.6 & & & 18.2 & 9.2 & 8.5 & 8.6 & 12.6 + 4005.24 & fe i & 1.56 & @xmath40.610 & 54.9 & 45.8 & 41.3 & 80.3 & 68.1 & 39.5 & 56.7 & 63.0 + 4045.81 & fe i & 1.49 & 0.280 & 90.0 & 94.1 & 79.5 & 121.5 & 106.7 & 78.1 & 89.4 & 99.3 + 4063.59 & fe i & 1.56 & 0.060 & & 69.1 & & 101.3 & 82.9 & & & + 4071.74 & fe
i & 1.61 & @xmath40.020 & & 68.3 & & 99.6 & 83.8 & & & + 4118.55 & fe i & 3.57 & 0.140 & & & & & 18.0 & & & + 4132.06 & fe i & 1.61 & @xmath40.820 & 49.0 & 37.6 & 36.0 & 81.7 & 68.2 & 40.5 & 48.7 & 62.3 + 4143.87 & fe i & 1.56 & @xmath40.620 & 57.8 & 42.6 & 47.7 & 84.9 & 66.8 & 49.6 & 60.2 & 65.9 + 4147.67 & fe i & 1.49 & @xmath42.100 & 13.6 & & & 19.0 & 8.0 & & 6.4 & 8.7 + 4172.76 & fe i & 0.96 & @xmath43.070 & & & & & 7.1 & & & 6.0 + 4181.75 & fe i & 2.83 & @xmath40.370 & 12.2 & & 7.3 & 21.5 & 17.1 & & 14.4 & 13.3 + 4187.05 & fe i & 2.45 & @xmath40.550 & & & & 35.4 & 19.1 & 16.1 & 14.4 & 19.2 + 4187.81 & fe i & 2.43 & @xmath40.550 & & & 12.6 & 38.8 & 27.9 & 13.2 & 20.0 & 25.0 + 4198.33 & fe i & 2.40 & @xmath40.720 & & & & 33.4 & 20.0 & & 12.4 & 20.2 + 4199.10 & fe i & 3.05 & 0.160 & 14.0 & & 14.5 & 31.8 & 22.3 & & 16.3 & 25.3 + 4202.04 & fe i & 1.49 & @xmath40.710 & 59.1 & & 45.9 & 80.6 & 65.0 & 47.2 & 59.5 & 67.3 + 4216.19 & fe i & 0.00 & @xmath43.360 & 18.2 & & & 41.2 & 19.9 & 8.2 & 20.1 & 23.5 + 4222.22 & fe i & 2.45 & @xmath40.970 & & & & & & & 7.8 & + 4227.44 & fe i & 3.33 & 0.270 & 15.1 & & 11.9 & & & & 11.2 & 15.3 + 4233.61 & fe i & 2.48 & @xmath40.600 & 17.7 & & & 32.4 & 22.0 & & 14.3 & 18.7 + 4235.95 & fe i & 2.43 & @xmath40.340 & 26.4 & & & 44.2 & 32.2 & 11.6 & 39.4 & + 4250.13 & fe i & 2.47 & @xmath40.410 & 16.6 & & 15.9 & 44.6 & 25.2 & & 20.6 & 22.1 + 4250.80 & fe i & 1.56 & @xmath40.380 & 54.3 & & 39.5 & 76.0 & 53.3 & 39.4 & 49.4 & 64.0 + 4260.49 & fe i & 2.40 & 0.140 & 41.0 & & 29.0 & 64.9 & 48.3 & 28.2 & 41.6 & 44.4 + 4271.16 & fe i & 2.45 & @xmath40.350 & 30.1 & & 15.0 & 41.5 & 29.3 & 11.8 & 18.1 & 26.5 + 4271.77 & fe i & 1.49 & @xmath40.160 & 74.9 & & 61.5 & 94.3 & 83.4 & 60.6 & 74.4 & 81.9 + 4282.41 & fe i & 2.18 & @xmath40.780 & 19.9 & & & 38.6 & & & 20.0 & 19.5 + 4294.14 & fe i & 1.49 & @xmath40.970 & 51.6 & & & 83.1 & 57.4 & & 46.3 & 66.1 + 4299.25 & fe i & 2.43 & @xmath40.350 & 25.8 & & & 47.8 & 47.3 & 15.0 & 20.4 & + 4307.91 & fe i & 1.56 & @xmath40.070 & 85.0 & & & & 97.6 & 68.9 & 84.2 & + 4325.77 & fe i & 1.61 & 0.010 & 76.0 & & & 102.6 & 80.2 & 60.8 & 72.8 & + 4337.05 & fe i & 1.56 & @xmath41.690 & 21.5 & & & 41.9 & & & 25.5 & 28.3 + 4375.94 & fe i & 0.00 & @xmath106.030 & 32.1 & & 17.4 & 67.7 & 34.6 & 19.7 & 30.8 & 36.9 + 4383.56 & fe i & 1.49 & 0.200 & 96.7 & & 79.2 & 126.2 & 99.0 & 78.9 & 86.1 & 96.9 + 4404.76 & fe i & 1.56 & @xmath40.140 & 75.1 & 64.3 & 66.0 & 100.9 & 86.2 & 62.5 & 72.9 & 83.4 + 4415.13 & fe i & 1.61 & @xmath40.610 & 51.9 & 43.9 & 45.3 & 81.1 & 65.3 & 46.5 & 58.5 & 62.0 + 4427.32 & fe i & 0.05 & @xmath43.040 & 33.6 & & 15.9 & 59.6 & 39.7 & & 30.5 & 40.0 + 4442.35 & fe i & 2.20 & @xmath41.250 & 9.0 & & 7.2 & 24.0 & & 6.6 & 8.5 & 16.2 + 4447.73 & fe i & 2.22 & @xmath41.340 & & & & 13.7 & & & & 9.1 + 4459.14 & fe i & 2.18 & @xmath41.280 & & & & & & & 5.2 & 8.8 + 4461.66 & fe i & 0.09 & @xmath43.210 & 22.1 & 17.5 & 16.4 & 52.1 & 30.6 & 10.2 & 26.6 & 29.0 + 4489.75 & fe i & 0.12 & @xmath43.970 & & & & 12.6 & & & & 5.0 + 4494.57 & fe i & 2.20 & @xmath41.140 & 19.6 & & 9.7 & 23.4 & 16.8 & 8.9 & 10.9 & 17.8 + 4531.16 & fe i & 1.49 & @xmath42.150 & 6.2 & & & 17.5 & & & & 5.0 + 4592.66 & fe i & 1.56 & @xmath42.450 & & & & 9.1 & & & & 5.0 + 4602.95 & fe i & 1.49 & @xmath42.220 & 7.8 & & & 16.2 & & & & 5.0 + 4871.33 & fe i & 2.86 & @xmath40.360 & 11.9 & 10.6 & & 24.9 & 9.8 & & 18.7 & 18.0 + 4872.14 & fe i & 2.88 & @xmath40.570 & 6.7 & 8.0 & & 18.0 & & & 4.6 & 11.2 + 4891.50 & fe i & 2.85 & @xmath40.110 & 19.0 & 21.5 & 17.5 & 35.3 & & & 23.4 & 23.9 + 4919.00 & fe i & 2.86 & @xmath40.340 & & 10.0 & & 22.5 & 14.0 & & 8.5 & 10.1 + 4920.51 & fe i & 2.83 & 0.150 & 17.4 & 16.5 & 13.1 & 36.2 & 26.0 & 9.9 & 19.9 & 22.3 + 4957.61 & fe i & 2.81 & 0.230 & 34.3 & 31.7 & 26.0 & 56.1 & 37.5 & 22.0 & & 37.7 + 5083.34 & fe i & 0.96 & @xmath42.960 & & & & 17.6 & & & & + 5166.28 & fe i & 0.00 & @xmath44.200 & & & & 11.5 & & & & + 5171.61 & fe
i & 1.48 & @xmath41.790 & 19.2 & 17.9 & 13.2 & 38.2 & 27.4 & & 13.0 & + 5192.35 & fe i & 3.00 & @xmath40.420 & 6.1 & & & 14.4 & & & & + 5194.95 & fe i & 1.56 & @xmath42.090 & & & & 18.0 & & & 7.8 & + 5216.28 & fe i & 1.61 & @xmath42.150 & & & & 12.8 & & & & 6.5 + 5227.19 & fe i & 1.56 & @xmath41.350 & 36.4 & 21.8 & 24.8 & 62.8 & 45.9 & & 34.6 & 46.6 + 5232.95 & fe i & 2.94 & @xmath40.100 & 15.5 & 15.0 & & 30.4 & & 5.6 & 16.4 & 17.9 + 5269.55 & fe i & 0.86 & @xmath41.320 & 75.8 & 55.8 & 55.7 & 103.1 & 97.3 & 61.8 & 76.6 & 85.6 + 5324.19 & fe i & 3.21 & @xmath40.100 & & & & 14.6 & & & 6.8 & + 5405.79 & fe i & 0.99 & @xmath41.840 & 39.3 & 25.5 & 25.7 & & & 27.5 & 38.8 & 46.3 + 5434.53 & fe i & 1.01 & @xmath42.130 & 19.7 & 7.0 & & & & & 21.7 & 23.6 + 5506.79 & fe i & 0.99 & @xmath42.790 & & & & & & & 6.8 & + 3255.90 & fe ii & 0.99 & @xmath42.498 & 49.6 & 36.4 & 33.6 & & & 50.4 & 54.2 & 71.2 + 3277.36 & fe ii & 0.99 & @xmath42.191 & 52.3 & 52.1 & 46.6 & & & 53.4 & 58.6 & 74.5 + 3281.30 & fe ii & 1.04 & @xmath42.678 & 41.0 & 28.7 & 24.8 & & & 34.6 & 45.0 & + 4178.86 & fe ii & 2.57 & @xmath42.530 & & & & 16.6 & 10.2 & & 5.1 & + 4233.17 & fe ii & 2.57 & @xmath42.000 & 13.9 & & 9.0 & 40.5 & & 12.0 & 17.8 & 22.7 + 4416.82 & fe ii & 2.77 & @xmath42.430 & & & & 10.8 & & & & 4.1 + 4508.30 & fe ii & 2.84 & @xmath42.280 & 5.2 & & & 10.8 & & & & 6.8 + 4555.89 & fe ii & 2.82 & @xmath42.170 & & & & 14.4 & 16.8 & & & 6.8 + 4583.84 & fe ii & 2.81 & @xmath42.020 & 10.5 & & & 30.5 & 15.0 & & 8.6 & 17.4 + 4923.93 & fe ii & 2.88 & @xmath41.320 & 17.5 & 5.2 & 12.6 & 45.8 & & & 17.7 & 25.2 + 5018.45 & fe ii & 2.89 & @xmath41.220 & 26.6 & 7.8 & 16.5 & 58.6 & 36.6 & 12.6 & 24.6 & 35.1 + 3842.05 & co i & 0.92 & @xmath40.763 & 9.0 & & & & & 18.0 & & + 3845.46 & co i & 0.92 & 0.009 & 19.9 & 10.3 & 16.5 & 36.8 & 33.4 & 28.0 & 25.6 & 26.8 + 3873.11 & co i & 0.43 & @xmath40.666 & 29.7 & & 9.6 & 41.7 & 40.0 & 34.8 & 43.9 & 38.8 + 4121.31 & co i & 0.92 & @xmath40.315 & 19.8 & & 9.9 & 25.0 & 18.2 & 26.0 & 25.3 & 23.1 + 3807.15 & ni i & 0.42 & @xmath41.180 & 41.4 & 16.5 & 29.0 & 58.9 & 45.3 & 42.7 & 49.0 & 41.4 + 3858.30 & ni i & 0.42 & @xmath40.967 & 56.6 & 29.8 & 41.7 & 79.2 & 56.5 & 53.0 & 60.8 & 55.5 + 4401.55 & ni i & 3.19 & 0.084 & & & & 7.0 & & & & + 3247.53 & cu i & 0.00 & @xmath40.060 & 26.0 & & 22.6 & & & & 30.8 & + 3273.95 & cu i & 0.00 & @xmath40.360 & 18.4 & 13.0 & 14.4 & & & 13.9 & 20.0 & 25.5 + 4810.54 & zn i & 4.08 & @xmath40.170 & @xmath1294.5 & & & @xmath1295.0 & @xmath1296.0 & & @xmath1295.0 & + 4077.71 & sr ii & 0.00 & 0.170 & 82.9 & 44.4 & & 27.6 & 37.2 & @xmath1295.0 & & 106.1 + 4215.52 & sr ii & 0.00 & @xmath40.140 & 75.6 & & 4.8 & 13.3 & 34.2 & @xmath1295.0 & 5.2 & 96.1 + 4554.04 & ba ii & 0.00 & 0.170 & 8.4 & 13.7 & @xmath1295.5 & 22.2 & 21.5 & 4.5 & @xmath1294.3 & 19.7 + 4934.16 & ba ii & 0.00 & @xmath40.150 & 5.2 & 8.3 & @xmath1295.5 & 7.0 & 12.6 & & @xmath1295.0 & 21.7 + 5853.70 & ba ii & 0.60 & @xmath41.010 & & & & & & & @xmath1292.0 & + 3774.33 & y ii & 0.13 & 0.220 & & & & & & @xmath12910.0 & & + 3950.36 & y ii & 0.10 & @xmath40.490 & @xmath1296.9 & & & @xmath1299.0 & & @xmath1294.5 & & + 3819.67 & eu ii & 0.00 & 0.510 & @xmath1298.0 & & & @xmath1297.5 & & & & + 3971.96 & eu ii & 0.21 & 0.270 & & & & & & @xmath1294.5 & & + 4129.70 & eu ii & 0.00 & 0.220 & @xmath1298.0 & @xmath1298.0 & & @xmath1299.0 & @xmath12915.0 & & & @xmath1295.0 + 3407.80 & dy ii & 0.00 & 0.180 & & & & & & @xmath12916.1 & & + 3531.71 & dy ii & 0.00 & 0.770 & & & & & & @xmath1298.8 & & + 4057.81 & pb i & 1.32 & @xmath40.220 & & @xmath1298.0 & & & & & & + l rrr c - normal + he0132@xmath42429 & @xmath130 & @xmath131 & + 0.09 + he1347@xmath41025 & @xmath132 & @xmath40.007 & + 0.32 + he1356@xmath40622 & @xmath133 & @xmath134 & + 0.33 + he1424@xmath135 & @xmath136 & @xmath137 & @xmath138 + bs16467@xmath4062 & @xmath139 & + 0.076 & @xmath140 +
c - rich + he1012@xmath41540 & + 0.014 & @xmath139 & + 0.71 + he1300 + 0157 & @xmath141 & + 0.022 & + 0.11 + he2323@xmath40256 & @xmath142 & + 0.005 & @xmath143 +
l | rrrr | rrrr | rrrr | rrrr| rrrr c(ch ) & 0.62 & 5.66 & 1 & & 0.15 & 5.26 & 1 & & @xmath1440.05 & @xmath1285.06 & 1 & & @xmath1280.63 & @xmath1285.26 & 1 & & 0.48 & 5.60 & 1 & + n(nh ) & 1.07 & 5.45 & 1 & & & & & & & & & & @xmath1281.13 & @xmath1285.10 & 1 & & @xmath1280.54 & @xmath1285.00 & 1 & + o(oh ) & @xmath1281.67 & @xmath1286.95 & 2 & 0.20 & & & & & & & & & & & & & @xmath1281.79 & @xmath128 7.15 & 2 & 0.20 + na i & @xmath40.31 & 2.46 & 2 & 0.09 & @xmath40.42 & 2.42 & 2 & 0.08 & 0.31 & 3.15 & 2 & 0.03 & @xmath145 & 2.32 & 2 & 0.07 & @xmath40.60 & 2.25 & 2 & 0.07 + mg i & 0.40 & 4.39 & 5 & 0.15 & 0.49 & 4.55 & 3 & 0.22 & 0.67 & 4.72 & 5 & 0.32 & 0.45 & 4.03 & 3 & 0.12 & 0.31 & 4.37 & 4 & 0.33 + al i & @xmath40.19 & 2.74 & 2 & 0.17 & @xmath40.02 & 2.97 & 2 & 0.17 & @xmath40.13 & 2.86 & 2 & 0.05 & @xmath40.16 & 2.35 & 2 & 0.15 & @xmath40.28 & 2.72 & 2 & 0.18 + si i & 0.57 & 4.57 & 1 & & 0.41 & 4.48 & 2 & 0.23 & 0.81 & 4.88 & 2 & 0.16 & @xmath41.00 & 2.59 & 1 & & 0.27 & 4.35 & 1 & + ca i & 0.27 & 3.08 & 3 & 0.15 & 0.36 & 3.24 & 5 & 0.18 & 0.43 & 3.31 & 7 & 0.17 & @xmath40.56 & 1.84 & 1 & & 0.12 & 3.00 & 3 & 0.12 + ca ii & 0.00 & 2.81 & 1 & & & & & & & & & & @xmath40.30 & 2.10 & 1 & & @xmath122 & 2.87 & 1 & + sc ii & 0.75 & 0.31 & 3 & 0.05 & 0.00 & @xmath40.38 & 2 & 0.12 & 0.32 & @xmath40.06 & 4 & 0.17 & @xmath40.08 & @xmath40.94 & 1 & & 0.16 & @xmath40.22 & 3 & 0.03 + ti i & 0.61 & 2.05 & 3 & 0.10 & @xmath1280.62 & @xmath1282.13 & 4 & 0.09 & 0.40 & 1.91 & 5 & 0.06 & & & & & 0.28 & 1.80 & 2 & 0.09 + ti ii & 0.39 & 1.84 & 15 & 0.10 & 0.26 & 1.77 & 13 & 0.11 & 0.26 & 1.76 & 17 & 0.07 &
@xmath40.17 & 0.85 & 8 & 0.17 & 0.20 & 1.72 & 14 & 0.12 + v i & & & & & @xmath1280.69 & @xmath1281.21 & 1 & & & & & & & & & & & & & + v ii & 0.36 & 0.81 & 1 & & & & & & & & & & @xmath128 0.60 & @xmath1280.64 & 1 & & 0.28 & 0.81 & 2 & 0.14 + cr i & @xmath40.43 & 1.69 & 5 & 0.12 & @xmath40.52 & 1.67 & 4 & 0.11 & @xmath40.52 & 1.66 & 4 & 0.13 & @xmath40.38 & 1.33 & 5 & 0.09 & @xmath40.54 & 1.65 & 5 & 0.10 + mn i & @xmath40.90 & 0.94 & 2 & 0.02 & @xmath40.88 & 1.03 & 2 & 0.04 & @xmath40.98 & 0.92 & 2 & 0.02 & @xmath40.10 & 1.33 & 2 & 0.02 & @xmath40.85 & 1.06 & 2 & 0.04 + mn ii & @xmath40.48 & 1.37 & 3 & 0.16 & & & & & & & & & 0.19 & 1.62 & 3 & 0.13 & @xmath40.48 & 1.44 & 3 & 0.07 + fe i & @xmath83 & 3.90 & 53 & 0.18 & @xmath146 & 3.97 & 50 & 0.22 & @xmath147 & 3.96 & 63 & 0.16 & @xmath148 & 3.49 & 39 & 0.18 & @xmath149 & 3.98 & 57 & 0.19 + fe ii & @xmath40.05 & 3.85 & 8 & 0.18 & 0.20 & 4.17 & 4 & 0.19 & 0.13 & 4.10 & 8 & 0.13 & 0.09 & 3.58 & 5 & 0.19 & 0.04 & 4.01 & 8 & 0.16 + co i & 0.55 & 1.92 & 4 & 0.18 & 0.48 & 1.92 & 3 & 0.19 & 0.24 & 1.67 & 3 & 0.11 & 1.03 & 1.98 & 4 & 0.21 & 0.68 & 2.12 & 3 & 0.28 + ni i & @xmath40.04 & 2.67 & 2 & 0.06 & @xmath40.19 & 2.58 & 1 & & @xmath40.04 & 2.72 & 3 & 0.33 & 0.24 & 2.52 & 2 & 0.01 & 0.15 & 2.93 & 2 & 0.04 + cu i & @xmath40.85 & @xmath40.18 & 2 & 0.06 & & & & & & & & & @xmath40.66 & @xmath40.41 & 1 & & @xmath40.75 & @xmath40.01 & 2 & 0.01 + zn i & @xmath1280.84 & @xmath1281.89 & 1 & & @xmath1280.94 & @xmath1282.06 & 2 & 0.17 & @xmath1280.55 & @xmath1281.66 & 1 & & & & & & @xmath1280.96 & @xmath1282.09 & 2 & 0.12 + sr ii & 0.05 & @xmath40.60 & 2 & 0.06 & @xmath41.13 & @xmath41.71 & 2 & 0.16 & @xmath41.88 & @xmath42.47 & 2 & 0.08 & @xmath1441.69 & @xmath1442.75 & 2 & 0.21 & @xmath41.75 & @xmath42.32 & 1 & + y ii & @xmath1280.33 & @xmath1440.98 & 1 & & & & & & @xmath1440.13 & @xmath1441.38 & 1 & & @xmath1280.26 & @xmath1441.46 & 2 & 0.19 & & & & + ba ii & @xmath40.85 & @xmath42.27 & 2 & 0.04 & @xmath40.62 & @xmath41.96 & 2 & 0.01 & @xmath41.19 & @xmath42.54 & 2 & 0.21 & @xmath40.91 & @xmath42.74 & 1 & & @xmath1440.56 & @xmath1441.91 & 3 & 0.72 + eu ii & @xmath1281.18 & @xmath1441.86 & 2 & 0.18 & @xmath1281.39 & @xmath1441.58 & 1 & & @xmath1280.56 & @xmath1442.41 & 3 & 0.18 & @xmath1281.50 & @xmath1441.95 & 1 & & & & & + dy ii & & & & & & & & & & & & & @xmath1281.56 & @xmath1441.30 & 2 & & & & & + l | rrrr | rrrr | rrrr c(ch ) & 2.22 & 7.38 & 1 & & 1.23 & 6.43 & 1 & & 0.97 & 5.77 & 1 & + n(nh ) & 1.25 & 5.75 & 1 & & @xmath1290.71 & @xmath1295.25 & 1 & & 2.16 & 6.30 & 2 & 0.30 + o(oh ) & 2.25 & 7.65 & 2 & 0.25 & 1.69 & 7.13 & 2 & 0.18 & 1.96 & 7.20 & 1 & + na i & 1.21 & 4.11 & 2 & 0.07 & @xmath40.49 & 2.44 & 2 & 0.07 & 1.45 & 3.98 & 2 & 0.11 + mg i & 1.88 & 5.99 & 6 & 0.44 & 0.32 & 4.47 & 3 & 0.10 & 1.47 & 5.22 & 7 & 0.25 + al i & 0.93 & 3.97 & 1 & & @xmath40.24 & 2.83 & 1 & & 0.48 & 3.16 & 1 & + si i & 1.07 & 5.20 & 1 & & 0.49 & 4.64 & 1 & & 0.56 & 4.32 & 1 & + ca i & 0.57 & 3.50 & 2 & 0.16 & 0.26 & 3.23 & 2 & 0.13 & 0.31 & 2.88 & 2 & 0.11 + ca ii & @xmath40.34 & 2.59 & 1 & & 0.10 & 3.07 & 1 & & 0.08 & 2.65 & 1 & + sc ii & & & & & 0.17 & @xmath40.12 & 1 & & 0.12 & @xmath40.56 & 1 & + ti i & @xmath1290.75 & @xmath1292.31 & 1 & & & & & & @xmath1290.46 & @xmath1291.66 & 1 & + ti ii & @xmath40.03 & 1.53 & 7 & 0.12 & 0.12 & 1.72 & 8 & 0.06 & 0.27 & 1.47 & 11 & 0.06 + v ii & & & & & & & & & 0.09 & 0.30 & 2 & 0.03 + cr i & @xmath40.33 & 1.91 & 2 & 0.09 & @xmath40.53 & 1.74 & 4 & 0.07 & @xmath40.54 & 1.34 & 5 & 0.15 + mn i & @xmath40.55 & 1.41 & 2 & 0.01 & @xmath40.86 & 1.13 & 2 & 0.02 & @xmath40.97 & 0.63 & 2 & 0.03 + mn ii & @xmath41.00 & 0.96 & 1 & & @xmath40.66 & 1.33 & 2 & 0.16 & @xmath40.43 & 1.17 & 2 & 0.01 + fe i & @xmath150 & 4.02 & 28 & 0.17 & @xmath151 & 4.06 & 36 & 0.19 & @xmath152 & 3.66 & 57 & 0.16 + fe ii & @xmath40.28 & 3.75 & 5 & 0.35 & @xmath40.16 & 3.90 & 6 & 0.11 & 0.06 & 3.72 & 9 & 0.18 + co i & 0.19 & 1.68 & 1 & & 0.39 & 1.92 & 3 & 0.08 & 0.42 & 1.55 & 3 & 0.18 + ni i & @xmath40.32 & 2.50 & 2 & 0.09 & @xmath40.02 & 2.83 & 2 & 0.06 & @xmath40.26 & 2.20 & 2 & 0.04 + cu i & @xmath40.63 & 0.15 & 1 & & @xmath40.68 & 0.13 & 2 & 0.02 & @xmath40.78 & @xmath40.36 & 1 & + sr ii & @xmath40.54 & @xmath41.07 & 1 & & @xmath41.55 & @xmath42.05 & 1 & & 0.18 & @xmath40.71 & 2 & 0.02 + ba ii & @xmath40.29 & @xmath41.58 & 2 & 0.02 & @xmath1530.63 & @xmath1531.89 & 2 & 0.20 & @xmath40.66 & @xmath42.32 & 2 & 0.23 + eu ii & @xmath1291.62 & @xmath154 & 1 & & & & & & @xmath1290.73 & @xmath1532.55 & 1 & + pb i & @xmath1292.93 & @xmath1291.46 & 1 & & & & & & & & & + l rrrrr @xmath155}}$ ] & 3 & 0.39 & 0.29 & 0.07 & 0.62 + @xmath156}}$ ] & + @xmath157}}$ ] & + @xmath158}}$ ] & 5 & @xmath159 & 0.36 & @xmath160 & 0.32 + @xmath161}}$ ] & 5 & 0.46 & 0.13 & 0.31 & 0.67 + @xmath162}}$ ] & 5 & @xmath163 & 0.09 & @xmath164 & @xmath122 + @xmath165}}$ ] & 5 & 0.21 & 0.71 & @xmath166 & 0.81 + @xmath167}}$ ] & 5 & 0.12 & 0.40 & @xmath168 & 0.43 + @xmath169}}$ ] & 3 & @xmath170 & 0.17 & @xmath171 & 0.00 + @xmath172}}$ ] & 5 & 0.23 & 0.33 & @xmath173 & 0.75 + @xmath174}}$ ] & 5 & 0.19 & 0.21 & @xmath175 & 0.40 + @xmath176}}$ ] & 2 & 0.32 & 0.05 & 0.28 & 0.36 + @xmath177}}$ ] & 5 & @xmath178 & 0.07 & @xmath179 & @xmath180 + @xmath181}}$ ] & 5 & @xmath182 & 0.36 & @xmath183 & @xmath184 + @xmath185}}$ ] & 3 & @xmath186 & 0.39 & @xmath178 & 0.19 + @xmath187}}$ ] & 5 & 0.08 & 0.10 & @xmath140 & 0.20 + @xmath188}}$ ] & 5 & 0.59 & 0.29 & 0.24 & 1.03 + @xmath189}}$ ] & 5 & 0.02 & 0.17 & @xmath190 & 0.24 + @xmath191}}$ ] & 3 & @xmath182 & 0.09 & @xmath192 & @xmath193 + @xmath194}}$ ] & 4 & @xmath195 & 0.88 & @xmath196 & 0.05 + @xmath197}}$ ] & 4 & @xmath198 & 0.24 & @xmath199 & @xmath200 + l rrrrr @xmath201}}$ ] & 3 & 1.47 & 0.66 & 0.97 & 2.22 + @xmath202}}$ ] & 2 & 1.37 & 0.70 & 0.64 & 2.15 + @xmath203}}$ ] & 3 & 2.03 & 0.30 & 1.69 & 2.25 + @xmath158}}$ ] & 3 & 0.72 & 1.06 & @xmath204 & 1.45 + @xmath161}}$ ] & 3 & 1.22 & 0.81 & 0.32 & 1.88 + @xmath162}}$ ] & 3 & 0.39 & 0.59 & @xmath205 & 0.93 + @xmath165}}$ ] & 3 & 0.71 & 0.32 & 0.49 & 1.08 + @xmath167}}$ ] & 3 & 0.38 & 0.17 & 0.26 & 0.57 + @xmath169}}$ ] & 2 & @xmath206 & 0.32 & @xmath207 & 0.10 + @xmath172}}$ ] & 2 & 0.14 & 0.04 & 0.12 & 0.17 + @xmath174}}$ ] & 3 & 0.12 & 0.15 & @xmath208 & 0.27 + @xmath177}}$ ] & 3 & @xmath209 & 0.12 & @xmath179 & @xmath210 + @xmath181}}$ ] & 3 & @xmath211 & 0.22 & @xmath212 & @xmath213 + @xmath185}}$ ] & 3 & @xmath214 & 0.28 & @xmath166 & @xmath215 + @xmath187}}$ ] & 3 & @xmath206 & 0.17 & @xmath164 & 0.06 + @xmath188}}$ ] & 3 & 0.33 & 0.12 & 0.19 & 0.42 + @xmath189}}$ ] & 3 & @xmath29 & 0.16 & @xmath138 & @xmath122 + @xmath191}}$ ] & 3 & @xmath214 & 0.08 & @xmath216 & @xmath217 + @xmath194}}$ ] & 3 & @xmath218 & 0.87 & @xmath219 & 0.17 + @xmath197}}$ ] & 2 & @xmath220 & 0.21 & @xmath193 & @xmath221 + l rrr [ fe / h ] & @xmath43.73 & @xmath222 & @xmath151 + c(ch ) & 5.89 & 6.02 & 6.43 + n(nh ) & @xmath223 & @xmath224 & @xmath225 + o(oh ) & 6.54 & 6.78 & 7.13 + na i & 2.44 & 2.49 & 2.64 + mg i & 4.10 & 4.30 & 4.47 + al i & 2.64 & 2.66 & 2.83 + si i & 4.50 & 4.44 & 4.65 + ca i & 2.98 & 3.06 & 3.23 + ca ii & 2.75 & 2.96 & 3.07 + sc ii & @xmath226 & @xmath171 & @xmath206 + ti ii & 1.77 & 1.58 & 1.72 + cr i & 1.51 & 1.56 & 1.74 + mn i & 0.65 & 0.91 & 1.13 + mn ii & 0.89 & 1.20 & 1.33 + fe i & 3.72 & 3.85 & 4.06 + fe ii & 3.57 & 3.77 & 3.90 + co i & 1.80 & 1.71 & 1.92 + ni i & 2.61 & 2.61 & 2.83 + cu i & & @xmath184 & 0.13 + sr ii & @xmath227 & @xmath228 & @xmath229 + ba ii & @xmath230 & @xmath231 & @xmath232 + l rrr cd @xmath438 245 & 4830 & @xmath233 & @xmath233 + bs16477@xmath4003 & 4928 & @xmath234 & @xmath235 + he2323@xmath40256 & 4995 & @xmath236 & @xmath237 + cs22948@xmath4066 & 5224 & @xmath238 & @xmath239 + bs16467@xmath4062 & 5364 & @xmath240 & @xmath241 + | we present a detailed abundance analysis based on high resolution and high signal - to - noise spectra of eight extremely metal poor ( emp ) stars with [ fe / h ] @xmath0 dex , four of which are new . only stars with @xmath1 @xmath2 @xmath3 k
are included .
two stars of the eight are outliers in each of several abundance ratios .
the most metal poor star in this sample , he1424@xmath40241 , has [ fe / h ] @xmath5 dex and is thus among the most metal poor stars known in the galaxy .
it has highly anomalous abundance ratios unlike those of any other known emp giant , with very low si , ca and ti relative to fe , and enhanced mn and co , again relative to fe .
only ( low ) upper limits for c and n can be derived from the non - detection of the ch and nh molecular bands .
he0132@xmath42429 , another sample star , has excesses of n and sc with respect to fe .
the strong outliers in abundance ratios among the fe - peak elements in these c - normal stars , not found at somewhat higher metallicities ( [ fe / h ] @xmath6 dex ) , are definitely real .
they suggest that at such low metallicities we are beginning to see the anticipated and long sought stochastic effects of individual supernova events contributing to the fe - peak material within a single star . with spectra reaching well into the near - uv we are able to probe the behvaior of copper abundances in such extreme emp stars .
a detailed comparison of the results of the analysis procedures adopted by our 0z project compared to those of the first stars vlt large project finds a systematic difference for [ fe / h ] of @xmath70.3 dex , our values always being higher . |
hydrocarbons in jupiter are produced in a series of chemical pathways initiated by the photolysis of methane in the upper stratosphere .
vertical transport , mainly turbulent diffusion , redistributes the molecules throughout the stratosphere and down to the troposphere , where they are eventually destroyed . hydrocarbons , and particularly the most stable of them , are therefore good tracers of the upper atmospheric dynamics . in addition , as they strongly contribute to the atmospheric opacity in the uv and ir , hydrocarbons act as major sources and sinks of heat , thereby participating to the stratospheric dynamics .
all these reasons have strongly motivated theoretical studies of the jovian stratospheric photochemistry ( strobel @xcite ; yung & strobel @xcite ; and most recently gladstone et al .
although nowadays very detailed , these models still need to be constrained by observations of minor species .
however , prior to the iso mission , only two molecules , acetylene ( @xmath11 ) and ethane ( @xmath12 ) , had been detected , except in the auroral zones where several other minor species ( @xmath13 , @xmath14 and @xmath15 ) have been observed ( kim et al .
mean stratospheric mole fractions have been inferred for @xmath11 and @xmath12 by various authors , but no precise information on their vertical variations was made available .
in this paper , we analyse the iso - sws spectrum of jupiter between 7 and 17 @xmath0 , in order to determine the vertical distributions of @xmath11 and @xmath12 , and to search for more complex ( c@xmath16 and c@xmath17 ) molecules .
2 presents the observations .
our analysis of the spectrum is presented in sect .
the results are compared with previous observations and theoretical predictions in sect .
descriptions of the infrared space observatory ( iso ) and of the short wavelength spectrometer ( sws ) can be found respectively in kessler et al .
( @xcite ) and de graauw et al . ( @xcite ) .
a preliminary analysis of the jupiter sws spectrum can be found in encrenaz et al .
( @xcite ) .
new iso - sws grating observations of jupiter were obtained on may 23 , 1997 ut using the aot 01 observing mode .
these observations have an average spectral resolution of 1500 , and range from 2.4 to 45 @xmath0 .
however , the useful range is limited to 2.417 @xmath0 , due to partial saturation at longer wavelengths .
the instrument aperture , @xmath18 arcsec@xmath19 at @xmath20 @xmath0 and @xmath21 arcsec@xmath19 at @xmath22 @xmath0 , was centered on the planet with the long axis aligned perpendicular to the ecliptic , thus roughly parallel to the n - s polar axis .
it covered latitudes between 30@xmath23s and 30@xmath23n , and @xmath24 longitude range from the central meridian .
the absolute flux accuracy is @xmath25 .
instrumental fringing generates a spurious signal between 12.5 and 17 @xmath0 , which amounts to @xmath2610% of the continuum level .
this parasitic signal was for the most part removed by fitting the detector relative response function to the observed spectrum and then dividing it out .
residual fringes were further removed by selective frequency filtering .
we analysed the iso - sws spectrum using a standard line - by - line radiative transfer code adapted to jupiter s conditions .
we included the molecular absorptions by @xmath27 , @xmath28 , @xmath11 , @xmath12 , and @xmath29 using the spectroscopic parameters of the geisa97 databank ( jacquinet - husson et al .
we also considered @xmath30 absorption , using a linelist provided by e. ari ( private communication ) and band intensities from koops et al .
( @xcite ) .
spectroscopic parameters for the @xmath31 s(1 ) line were calculated using molecular constants from jennings et al .
( @xcite ) and reuter & sirota ( @xcite ) .
the h@xmath32-he collision - induced continuum was calculated following the work of borysow et al .
( @xcite , @xcite ) .
the @xmath27 vertical distribution was taken from fouchet et al .
( @xcite ) .
we first calculated synthetic spectra in the region of the @xmath28 @xmath33 band . for @xmath28
, we used a deep mixing ratio of @xmath34 ( niemann et al .
@xcite ) and the vertical profile derived by drossart et al .
( @xcite ) from the @xmath28 fluorescence emission at 3.3 @xmath0 .
the @xmath33 band allows one to retrieve 4 independent points on the temperature profile between 35 and 1 mbar .
we also generated synthetic spectra of the @xmath31 s(1 ) rotational line at 17 @xmath0 .
this line probes a broad atmospheric layer at 330 mbar .
the ortho - to - para ratio of @xmath31 was assumed to follow local thermodynamical equilibrium .
the stratospheric temperature profile was adjusted in order to best match the absolute emission in the @xmath28 band , and the line - to - continuum ratio of the @xmath31 s(1 ) line . in practice , starting with the temperature profile measured in - situ by the galileo probe ( seiff et al .
@xcite ) , it was necessary to cool it by 2 k between 30 and 5 mbar . at pressures lower than 5 mbar , the initial profile was warmed by 4 k up to @xmath35 k. the @xmath33 band is also somewhat sensitive to the temperature around the 10-@xmath36bar pressure level .
we found that the temperature remains constant within a few degrees between 1 mbar and 1 @xmath36bar , as already observed by seiff et al .
( @xcite ) . at pressures
lower than 1 @xmath36bar , we adopted the measurements of seiff et al .
( @xcite ) , vertically smoothed in order to remove oscillations due to gravity waves , noting that our measurements are essentially insensitive to this pressure range .
the fit to the @xmath28 and @xmath31 emissions is presented in fig .
[ figtemp ] .
the 20% uncertainty on the absolute flux calibration directly results in an uncertainty of @xmath37k on the temperature profile inferred from the @xmath28 emission .
this uncertainty partly explains the minor disagreement between the modelled and observed @xmath31 line .
indeed , our model , while giving an optimum fit to the @xmath28 emission , slightly ( 510% ) overpredicts the observed line - to - continuum ratio of the s(1 ) line .
we also note that the @xmath31 ortho - to - para ratio could differ from the thermal equilibrium value , as observed in the troposphere by conrath et al .
( @xcite ) .
for example , a synthetic spectrum calculated with a constant para fraction of 0.34 , corresponding to the thermal equilibrium value at 115 k , would fully reconcile the @xmath28 and @xmath31 measurements .
the iso - sws observations in the vicinity of the @xmath11 emission at 13.7 @xmath0 were compared with synthetic spectra obtained with three distinct @xmath11 vertical distributions ( fig . [ figpro ] ) .
all three profiles reproduce the emissions due to the p-,q- , and r - branches of the main @xmath38 @xmath11 band ( fig .
[ figace ] ) .
however , only one profile ( fig . [ figpro ] , solid line ) fits the observed spectrum in the q - branches of the @xmath39 band at 13.68 and 13.96 @xmath0 . indeed , while the @xmath38 band probes atmospheric levels between 2 and 5 mbar , the hot band sounds warmer , higher levels around 0.3 mbar , allowing us to determine the slope of the @xmath11 profile between these two regions .
error bars on the @xmath11 mixing ratio were estimated by taking into account instrumental noise and the uncertainty in the relative strengths of the @xmath11 lines .
the resulting mixing ratios are @xmath2 at 0.3 mbar and @xmath3 at 4 mbar , giving a slope @xmath4 .
the error on the temperature profile ( @xmath37k ) introduces an additional uncertainty on the @xmath11 mixing ratios of about 20% .
this error , however , essentially equally applies to all pressure levels , and thus leaves the retrieved @xmath11 profile slope mostly unaffected .
similarly to @xmath11 , we compare in fig .
[ figeth ] the iso - sws spectrum in the @xmath12 @xmath40 band with synthetic spectra calculated with three distinct @xmath12 vertical profiles ( fig .
[ figpro ] ) .
each of the profiles was designed to reproduce the observed emission in the @xmath41 multiplet at 12.16 @xmath0 , which probes the 1-mbar pressure level .
the rest of the @xmath40 band consists of weaker multiplets , which sound deeper levels extending from 1 to 10 mbar .
this combination of strong and weak multiplets makes this band sensitive to the @xmath12 vertical distribution .
in addition , the pseudo - continuum level in between the @xmath12 emissions is also sensitive to the @xmath12 abundance in the lower stratosphere . our best - fit model ( fig .
[ figpro ] , solid line ) has a slope @xmath42 , but the steep - slope model ( fig .
[ figpro ] , dotted line ) is also marginally compatible with the observations .
this results in a relatively large uncertainty on the slope determination : @xmath5 at 1 mbar , and @xmath6 at 10 mbar , giving @xmath7 .
an additional error of 25% on @xmath1 comes from the uncertainty on the temperature .
again , it should not affect the retrieved slope .
the iso - sws spectrum exhibits a broad emission near 15.8 @xmath0 , which coincides with the @xmath40 band of @xmath29 ( fig .
[ figmet ] ) .
since the @xmath29 lines are optically thin , no information on the vertical profile can be derived .
using a vertical profile similar to that calculated by gladstone et al .
( @xcite ) , we found a column density of @xmath8 molecule@xmath9 .
the synthetic spectrum exhibits small - scale structures which are not seen in the observations .
this mismatch is propably due to an imperfect data reduction .
indeed , the respective frequencies of the fringes and of the @xmath29 features lie close to each other .
it is therefore very difficult to fully remove the former without altering the latter . on the contrary ,
the broad emission is a low frequency signal and is therefore left unaffected by the frequency filtering .
this explanation is admittedly not entirely satisfactory , but given the wavelength match of the emission with the @xmath40 mode of @xmath29 , and in the absence of any other plausible candidates , we regard the detection of methylacetylene as unambiguous . the @xmath43 band of @xmath30 is not detected at 15.92 @xmath0 .
we inferred an upper limit of the @xmath30 column density of @xmath10 molecule@xmath9 , using the gladstone et al .
( @xcite ) profile .
bzard et al . ( @xcite )
first showed from 13.4-@xmath0 high - resolution spectroscopy that the @xmath11 mixing ratio increases with height in the stratosphere , and that most of the acetylene is concentrated above the @xmath260.5-mbar level .
their distribution has a mixing ratio of about @xmath44 at 0.3 mbar , in reasonable agreement with our results .
more recently , btremieux & yelle ( @xcite ) , using uv observations , found an average @xmath11 mixing ratio in the 2060 mbar range of @xmath45 , consistent with our value of @xmath46 , extrapolated to this pressure range . using height - dependent mixing ratio profiles to analyse high - resolution infrared observations , sada et al .
( @xcite ) derived mixing ratios of @xmath47 for @xmath12 at 5 mbar and @xmath48 for @xmath11 at 8 mbar . while the former value exactly agrees with our results , the latter is almost 3 times less than that extrapolated downwards from our @xmath11 profile . also from infrared observations ,
noll et al . ( @xcite ) , using a slope of @xmath49 , found a @xmath12 mixing ratio of @xmath50 at 1 mbar , which compares reasonably well with our measurement at the same pressure level .
we also compared our retrieved mixing ratios with those calculated in the photochemical model of gladstone et al .
( @xcite ) . for @xmath12 ,
our results are in excellent agreement with their model both at 1 and 10 mbar ( fig .
[ figpro ] ) . for @xmath11 , while the agreement is good at 4 mbar , their mixing ratio at 0.3 mbar is higher than ours by a factor of 4 .
our @xmath11 slope is then significantly lower than theirs ( @xmath4 vs. @xmath51 ) .
note that our derived @xmath11 slope is still steeper than that of @xmath12 ( @xmath7 ) , as expected .
indeed , @xmath12 , being less subject to photolysis , has a longer lifetime in the jovian stratosphere than @xmath11 ( gladstone et al .
@xcite , their fig . 6 ) .
hydrocarbons are formed from the photolysis of @xmath28 around the homopause .
small - scale turbulence , parameterized in a photochemical model by the eddy diffusion coefficient ( @xmath52 ) , transports them downwards in the stratosphere .
this process is approximately modelled for long - lived products by the equation @xmath53 , where @xmath54 is the number density at altitude @xmath55 and @xmath56 the vertically integrated net production rate above altitude @xmath55 .
a first hypothesis would attribute the difference between the observed and calculated @xmath11 abundances to an overestimation of the @xmath11 production rate @xmath56 in the gladstone et al . model .
our two abundance measurements at 4 and 0.3 mbar allow us to calculate the mean @xmath57 over this pressure range . comparing with the value of @xmath57 predicted by gladstone et al . at 1 mbar
, we found that @xmath56 should be decreased by a factor of @xmath263 .
a second explanation would be that gladstone et al . underestimated the eddy diffusion coefficient @xmath52 by a factor of @xmath263 around 1 mbar .
this underestimation at 1 mbar should also apply to the level and eddy diffusion coefficient ( @xmath58 ) at the homopause .
however , it is difficult to directly quantify the changes on the homopause parameters , because of the strong coupling between production rates and homopause level .
simply , we note that our analysis could imply an increase in @xmath58 .
it is consistent with drossart et al .
( @xcite ) , who , analysing @xmath28 fluorescence , found @xmath59 @xmath60 , higher than the value of 1.4@xmath61 assumed in the gladstone et al . model . in this case , the agreement on the @xmath12 slope would also imply that the @xmath12 production rate has been underestimated by gladstone et al .
( @xcite ) .
in fact , the most direct conclusion of our measurements is that gladstone et al . have overestimated the @xmath11/@xmath12 production rate ratio .
the iso - sws spectrum enables the first detection of @xmath29 in the equatorial region of jupiter .
we retrieved a column density of @xmath8 molecule@xmath9 .
kim et al . ( @xcite )
had previously detected @xmath29 in the north polar auroral zone of jupiter , and had retrieved a column density of @xmath62 molecule@xmath9 . at least a large part of the difference is explained by different modelling assumptions .
kim et al .
( @xcite ) assumed a uniform vertical distribution throughout the stratosphere and used a temperature profile for the auroral region which is now known to be incorrect ( drossart et al .
@xcite ) . for @xmath30
, we found only an upper limit of @xmath10 molecule@xmath9 , 65 times lower than the stratospheric column density predicted by gladstone et al .
( 4.5@xmath6310@xmath64 molecule@xmath9 ) .
this is consistent with their overestimate of @xmath11 production , since the production of @xmath30 is essentially quadratically dependent on the abundance of @xmath11 . as our @xmath29 column density is 3.5 times larger than calculated by gladstone et al . , we derived a @xmath29/@xmath30 ratio larger than 20 , while they found a ratio of @xmath65 .
this discrepancy is all the more remarkable as , in the case of saturn , where both @xmath29 and @xmath30 are detected ( de graauw et al .
@xcite ) , the photochemical model of moses et al .
( @xcite ) gives , as observed , a @xmath29/@xmath30 ratio of about 10 .
this stresses that the c@xmath16 and c@xmath17 chemistry in jupiter should be reassessed with a complete photochemical model of the jovian stratosphere .
this study is based on observations with iso , an esa project with instruments funded by esa member states ( especially the principal investigators countries : france , germany , the netherlands and the united kingdom ) , and with participation of isas and nasa . | we have analysed the iso - sws spectrum of jupiter in the 1216 @xmath0 range , where several hydrocarbons exhibit rovibrational bands . using temperature information from the methane and hydrogen emissions , we derive the mixing ratios ( @xmath1 ) of acetylene and ethane at two independent pressure levels . for acetylene ,
we find @xmath2 at 0.3 mbar and @xmath3 at 4 mbar , giving a slope @xmath4 , while for ethane @xmath5 at 1 mbar and @xmath6 at 10 mbar , giving @xmath7 .
the ethane slope is consistent with the predictions of gladstone et al .
( @xcite ) , but that predicted for acetylene is larger than we observe .
this disagreement is best explained by an overestimation of the acetylene production rate compared to that of ethane in the gladstone et al .
( @xcite ) model . at 15.8 @xmath0 ,
methylacetylene is detected for the first time at low jovian latitudes , and a stratospheric column density of @xmath8 molecule@xmath9 is inferred .
we also derive an upper limit for the diacetylene column density of @xmath10 molecule@xmath9 . |
let @xmath2 be the hyperbolic plane of curvature @xmath3 ( see j.g .
ratcliffe @xcite for facts and references on hyperbolic geometry , and d.v .
alekseeskij , e.b .
vinberg , a.s .
solodovnikov @xcite for an in depth study on discrete groups of isometries ) .
a compact hyperbolic surface @xmath4 can be obtained as the quotient of @xmath2 by a discrete group @xmath5 of isometries that acts fixed point free on @xmath2 . for a regular polygon @xmath6 in @xmath2
, we say that @xmath4 can be tiled by copies of @xmath6 if there exists an edge to edge tiling of @xmath2 by congruent copies of @xmath6 that is invariant under @xmath5 .
the equivalence classes of the tiles , edges , and vertices of the tiling in @xmath2 with respect to @xmath5 correspond to the tiles , edges , and vertices , respectively , of the tiling on @xmath4 ( as a cw - decomposition of @xmath4 ) . in this case
there exists an integer @xmath7 such that the angles of @xmath6 are @xmath8 .
if @xmath6 is a regular triangle then @xmath9 . in this note
, we provide a simple proof for the following statement . [ main ] for @xmath0 , the minimal area of a compact hyperbolic surface that can be tiled by embedded regular triangles of angle @xmath1 , @xmath0 , is @xmath10 , where @xmath11 is the minimal positive integer such that @xmath12 is divisible by six . if the surface is assumed to be orientable then the minimal area is the same if @xmath13 , and twice the previous value otherwise . *
remark * it follows by a.l .
edmonds , j.h .
ewing , r.s . kulkarni @xcite that there is a compact hyperbolic surface that can be tiled by @xmath14 embedded regular triangles of angle @xmath1 , @xmath0 , if and only if @xmath14 is even , and @xmath15 is divisible by @xmath16 .
+ we note that theorem [ main ] is proved in a more general setup , and using a more involved construction in a.l .
edmonds , j.h .
ewing , r.s .
kulkarni @xcite .
in addition , the case of theorem [ main ] when @xmath16 is divisible by six , is proved by c. bavard @xcite . in other words , for @xmath0 , let @xmath17 be the isometry group of the edge to edge tiling by regular triangles of angle @xmath18 . then theorem [ main ] yields ( using its notation ) that the minimal index of a subgroup of @xmath17 that acts fixed point free on @xmath2 is @xmath19 . our motivation to consider theorem [ main ] comes from the theory of packing and covering by equal circular discs on a compact surface @xmath4 of constant curvature .
let @xmath20 be a regular triangle of angle @xmath21 in the universal covering surface , which is @xmath2 , @xmath22 , and the unit sphere @xmath23 in @xmath24 if the constant curvature is @xmath3 , @xmath25 or @xmath26 , respectively .
in addition , let @xmath27 and @xmath28 be inradius and circumradius , respectively , of @xmath20 , and let @xmath29 denote area . according to the triangle bound due to l. fejes tth @xcite
( see also g. fejes tth , w. kuperberg @xcite for a survey ) , if @xmath4 can be covered by @xmath30 metric balls of radius @xmath28 , then @xmath31 in addition , if @xmath4 contains @xmath30 pairwise disjoint open embedded balls of radius @xmath27 , then @xmath32 if the injectivity radius of @xmath4 is at least @xmath27 , then @xmath33 . in these bounds
, equality holds if and only if @xmath34 for some integer @xmath35 , and @xmath4 can be tiled by copies of @xmath20 using exactly @xmath30 vertices . to characterize the cases when we have equality in the estimate using the injectivity radius ( @xmath36 ) is the subject of c. bavard @xcite .
to charaterize regular triangulations when the curvature of @xmath4 is non - negative , is rather straightforward . if the curvature of @xmath4 is positive , then @xmath37 , as the sum of the angles of @xmath20 is larger , than @xmath38 .
there are two types of possible @xmath4 .
if @xmath4 is a sphere then the possible regular triangulations are determined by the platonic solids : tetrahedron , octahedron and icosahhedron .
if @xmath4 is a projective plane then the possible regular triangulations are obtained by identifying the opposite faces of the octahedron and the icosahhedron .
next , there are exactly two types of surfaces of curvature zero ; namely , tori and klein bottles . in these cases @xmath39 . for any @xmath40 , it is easy to construct a surface of either type with a regular triangulation having @xmath30 vertices .
therefore , it remains to find regularly triangulated hyperbolic surfaces , when @xmath0 .
this is the content of theorem [ main ] and the following remark .
first , we provide some properties of tilings of a compact hyperbolic surface @xmath4 . the euler characteristic of @xmath4 is denoted by @xmath41 , which is negative , and is always even in the case of oriented surfaces .
it follows from the gau - bonnet formula that the area of @xmath4 is @xmath42 , therefore minimizing the area is equivalent to maximizing the euler characteristic .
[ nk6 ] if a compact hyperbolic surface @xmath4 is tiled by embedded regular triangles of angle @xmath1 , @xmath0 , and the tiling has @xmath30 vertices ( as a cw - decomposition ) , then @xmath43 is divisible by six , and @xmath44 _ proof : _ since each vertex of the tiling is incident with exactly @xmath16 edges , the tiling has @xmath45 faces and @xmath46 edges . therefore , we have @xmath47 since @xmath43 is divisible by two and three , it is divisible by six .
after the necessary condition above , we present some ways to construct suitable hyperbolic surfaces . using the oriented double cover of a non - orientable surface
, we have the following . [ nonoriented ] if a compact non - orientable hyperbolic surface @xmath4 is tiled by embedded regular triangles of angle @xmath1 , @xmath0 , using @xmath30 vertices , then there exists a compact oriented hyperbolic surface @xmath48 , which is tiled by embedded regular triangles of angle @xmath1 using @xmath49 vertices , and @xmath50 .
assume that a compact hyperbolic surface @xmath4 is tiled by embedded regular triangles of angle @xmath1 , @xmath0 , and the tiling has @xmath30 vertices .
by duality , it is equivalent that @xmath4 can be tiled by @xmath30 embedded regular @xmath16-gons of angle @xmath51 . to obtain @xmath4
, one may start with @xmath30 regular @xmath16-gons , and identify the suitable pairs out of the total @xmath43 edges of the @xmath30-gons .
we describe a combinatorial way to encode pairings that we need in the sequel .
let @xmath52 be convex polygons with at least seven vertices such that the their total number of vertices ( or edges ) is @xmath53 for some even integer @xmath54 .
a _ proper labeling _ of the vertices of @xmath52 means an assignment of a label from @xmath55 to each vertex with the following properties .
( i ) : : any @xmath56 is the label of exactly three of the vertices .
( ii ) : : for any non - empty proper subset of @xmath52 there exists a label @xmath56 that occurs once or twice as the label of some vertex of a convex polygon in the subset .
( iii ) : : if a @xmath57 has two consecutive vertices with label @xmath56 , then @xmath57 has three consecutive vertices with label @xmath58 .
they determine two consecutive edges of @xmath57 , which are called proper pairs .
( iv ) : : if @xmath59 are the labels of the end points of an edge @xmath60 of some @xmath57 , then there exists exactly one more edge among the all together @xmath53 edges , whose endpoints are labeled @xmath58 and @xmath61 .
this edge is called the proper pair of @xmath60 .
( v ) : : it never occurs that for some @xmath59 , the two neighbours of a vertex with label @xmath58 are both labeled @xmath61 . given a proper labeling
, we give a orientation to each edge .
for an edge @xmath60 of some @xmath57 , whose end points are labeled @xmath58 and @xmath61 , we orient the edge @xmath60 according to the natural ordering of the labels of its vertices if @xmath59 , and according to the positive orientation of @xmath57 if @xmath62 .
we say that the orientation of @xmath60 is positive if it coincides with the orientation induced by the positive orientation of @xmath57 , and the orientation of @xmath60 is negative otherwise .
we say that the proper labeling is oriented , if for any proper pair @xmath60 and @xmath63 of edges , the orientations of @xmath60 and @xmath63 are opposite . for @xmath0 ,
let us assume that @xmath52 are regular @xmath16-gons with angle @xmath1 , and we have have a proper labeling of the vertices . then identifying the proper pairs of edges according their orientations , we obtain a ( connected ) compact hyperbolic surface @xmath4 , which is oriented if the proper labeling is oriented . from a given proper labeling of @xmath30 polygons , we will construct a proper labeling of @xmath30 polygons with a higher number of vertices using the operations ( a ) , ( b ) and ( c ) below .
a polygonal path @xmath64 , @xmath65 , is the union of the segments with end points @xmath66 and @xmath67 , @xmath68 , where @xmath69 are different .
when we replace a polygonal path @xmath64 contained in the boundary of a convex polygon @xmath6 with the polygonal path @xmath70 with @xmath71 and @xmath72 , we always do it in a way to obtain a new convex polygon @xmath73 , where @xmath74 are vertices of @xmath73 .
let us assume that we have a proper labeling of vertices of the convex polygons @xmath52 using labels from @xmath55 .
when replacing a polygonal path with a new one , we only list the corresponding labels .
( a ) at @xmath75 : : assume that for @xmath76 , the union of the boundaries of @xmath52 contains polygonal paths with labels @xmath77 , @xmath78 and @xmath79 . then for @xmath80 and @xmath81 ,
we replace ( see figure 1 ) @xmath82 + ( b ) at @xmath75 : : assume that for @xmath76 , the union of the boundaries of @xmath52 contains polygonal paths with labels @xmath77 , @xmath78 and @xmath79 .
then for @xmath80 , @xmath81 , @xmath83 and @xmath84 , we replace ( see figure 2 ) @xmath85 + ( c ) at @xmath86 : : assume the endpoints of the proper pair of edges @xmath60 and @xmath63 are labeled @xmath87 . for @xmath80 and @xmath81 ,
we replace ( see figure 3 ) @xmath88 + we observe that all operations ( a ) , ( b ) and ( c ) produce a new proper labeling .
in addition , if the proper labeling of the vertices of @xmath52 is oriented , then the new proper labeling constructed in ( b ) is also oriented .
for @xmath0 , let @xmath89 be the smallest positive integer such that @xmath90 is divisible by six .
it follows by lemma [ nk6 ] that if a compact hyperbolic surface @xmath4 is tiled by embedded regular triangles of angle @xmath1 , @xmath0 , then @xmath91 in addition , let @xmath92 ( @xmath93 ) be a hyperbolic surface ( oriented hyperbolic surface ) of smallest area that can be tiled by embedded regular triangles of angle @xmath1 .
theorem [ main ] is equivalent proving that @xmath94 it follows by lemma [ nonoriented ] and ( [ ln ] ) that @xmath95 we construct @xmath92 and @xmath93 , depending on the remainder of @xmath16 modulo @xmath96 , and providing a proper labeling of the vertices of @xmath89 @xmath16-gons .
we only provide the @xmath89 lists of labels according to positive orientation of the @xmath16-gons . + * case 1 * @xmath97 to prove ( [ chixk ] ) for @xmath101 , we apply operation ( b ) at @xmath26 ( to alter @xmath102 , @xmath103 and @xmath104 ) , and then again ( b ) at @xmath105 ( to alter @xmath106 , @xmath107 and @xmath108 ) .
this way we obtain a proper labeling of some six @xmath109-gons .
after this , we apply ( a ) at @xmath26 and @xmath105 to obtain a proper labeling of some six @xmath110-gons . continuing this way ,
applying alternately operation ( b ) at @xmath26 and @xmath105 , and operation ( a ) at @xmath26 and @xmath105 , we obtain a proper labeling of some six @xmath16-gons for any @xmath16 in case 1 . to prove ( [ chixk ] ) for @xmath117 , we apply operation ( b ) at @xmath105 . to prove ( [ chixk ] ) for @xmath118
, we apply operation ( b ) at @xmath105 twice . in particular , using operation ( b ) at @xmath105 , we obtain a proper labeling of some three @xmath16-gons for any @xmath16 in case 2 . to prove ( [ chixk ] ) for @xmath123 , we apply operation ( c ) at @xmath124 . in particular ,
using operation ( b ) at @xmath124 , we obtain a proper labeling of some pair of @xmath16-gons for any @xmath16 in case 3 . to prove ( [ chixk ] ) for @xmath130 , we apply operation ( b ) at @xmath131 twice . to prove ( [ chixk ] ) for @xmath132
, we apply operation ( b ) at @xmath131 . in particular , using operation ( b ) at @xmath131 , we obtain a proper labeling of some three @xmath16-gons for any @xmath16 in case 4 .
to prove ( [ chixk ] ) for @xmath139 , we apply operation ( b ) at @xmath26 three times . in particular , using operation ( b ) at @xmath26 , we obtain an oriented proper labeling of some @xmath16-gon for any @xmath16 in case 5 .
therefore , @xmath140 in this case . +
* case 6 * @xmath141 to prove ( [ chixk ] ) for @xmath144 , we apply operation ( b ) at @xmath26 three times . in particular , using operation ( b ) at @xmath26 , we obtain a proper labeling of some @xmath16-gon for any @xmath16 in case 6 . for these @xmath16
, we have @xmath145 by lemma [ nk6 ] , which is odd as @xmath146 is even . therefore ,
@xmath112 by ( [ chixkori ] ) .
* remark * if @xmath0 is divisible by six , then there is a graph theoretic way to produce a proper labeling of a @xmath16-gon .
let @xmath147 be a connected graph on @xmath55 for @xmath148 such that each vertex is of degree three .
we call a closed path on the edges of @xmath147 a double hamiltonian path if each edge is traveled exactly twice , and consecutive edges in the path are always different .
then a proper labeling of the vertices of the @xmath16-gon where no pair of consecutive vertices have the same label , is equivalent to a double hamiltonian path .
the proper labeling is oriented if and only if each edge is traveled in both directions in the double hamiltonian path .
christophe bavard , [email protected]_ + institut de mathmatiques , universit bordeaux 1 + kroly j. brczky , [email protected]_ + alfrd rnyi institute of mathematics , and + universitat politcnica de catalunya , barcelona tech , and + department of geometry , roland etvs university + borbla farkas , [email protected]_ + department of geometry , technical university of budapest + istvn prok , [email protected]_ + department of geometry , technical university of budapest + lluis vena , [email protected]_ + department of mathematics , university of toronto + gergely wintsche , [email protected]_ + department of didactics , roland etvs university + | for @xmath0 , we determine the minimal area of a compact hyperbolic surface , and an oriented compact hyperbolic surface that can be tiled by embedded regular triangles of angle @xmath1 . based on this , all the cases of equality in lszl fejes tth s triangle bound for hyperbolic surfaces are described . [ section ] [ theo]corollary [ theo]lemma [ theo]remark [ theo]proposition [ theo]conjecture [ theo]example |
gyrokinetics @xcite has proven a very useful tool to study turbulence in the core of fusion devices , making kinetic simulations of turbulent fluctuations possible in reasonable computational times @xcite .
its main advantage is averaging over the gyrofrequency time scale without losing the effect of the finite size of the gyroradius that is of the order of the typical wavelength of the turbulence . to perform this average , it is necessary to assume certain orderings that in the electrostatic limit can be summarized as @xmath2 where @xmath3 is the electrostatic potential , @xmath4 is the magnetic field , @xmath5 is the characteristic frequency of the turbulent fluctuations , @xmath6 is a characteristic macroscopic scale , @xmath7 , @xmath8 and @xmath9 are the thermal speed , the gyroradius and the gyrofrequency of the species of interest , @xmath10 and @xmath11 are the charge and the mass , and @xmath12 and @xmath13 are the magnitude of the electron charge and the speed of light . since this article is about electrostatic gyrokinetics
, we have assumed that the magnetic field is stationary and its characteristic length of variation is of the order of the macroscopic length @xmath6 .
the ordering in ( [ orderings ] ) implies that the electrostatic potential fluctuates with some characteristic frequency @xmath5 and has a strong gradient perpendicular to the magnetic field , on the order of the inverse of the gyroradius , whereas its gradient parallel to the direction of the magnetic field , @xmath14 , is on the order of the inverse of the larger scale @xmath6 .
the frequency @xmath5 of the turbulence is usually much smaller than the gyrofrequency , making the gyrokinetic average over the gyromotion valid .
we have employed the small parameter @xmath15 to make this explicit . in most fusion experiments ,
the ratio of the gyroradius and the macroscopic length is another small parameter that we also order as @xmath0 .
it is easy to see that in drift wave turbulence , for which the characteristic frequency is @xmath16 , the quantities @xmath17 and @xmath18 are indeed of the same order .
more importantly , to obtain the typical gyrokinetic formalism , it is necessary to order the electrostatic potential as small compared with the characteristic energy of the particles .
this assumption is necessary to prove that the gyromotion of the particles is circular to lowest order .
the most common gyrokinetic ordering assumes that the parameter @xmath19 is comparable to @xmath0 , as is done in ( [ orderings ] ) . in this way
, magnetic geometry effects such as the @xmath20 and curvature drifts , of order @xmath18 , are allowed to be comparable to the turbulent @xmath21 drift , of order @xmath22 . ordering these effects so that they are comparable is very important in , for example , the core of tokamaks , where the curvature of the magnetic field lines is believed to be the most important drive for the turbulence @xcite .
the ordering in ( [ orderings ] ) contains the simplest assumptions that are still interesting , but it can be extended to include components of the potential that have perpendicular gradients of the order of the inverse of the macroscopic length @xmath6 @xcite .
the results that we present in this article can be easily extended to some of these more general orderings , but we leave this for future work .
there are different techniques to obtain gyrokinetics ( and for that matter , drift kinetics @xcite , of which gyrokinetics is a natural extension ) . on the one hand
, it is possible to obtain the gyrokinetic equation by working iteratively on the vlasov equation @xcite .
we will call these _ iterative methods_. on the other hand , it is possible to use phase - space lagrangian / hamiltonian methods that solve order by order for the motion of the particle in a given electromagnetic field , uncoupling the gyromotion from the slower time scales @xcite .
once the motion of the particle is known , the vlasov equation is simply obtained by its characteristics .
we will call these _
lagrangian methods_. both procedures are asymptotic expansions in the parameter @xmath0 , and give equivalent equations order by order , but the lagrangian methods have the advantage of giving the equations in a form that exactly conserves some energy - like quantities
. this property may be very important for the global , full @xmath23 simulations that are being developed @xcite . to have an energy - like invariant and at the same time obtain equations of motion and equations for the electromagnetic fields that are the same to first order in @xmath0 as those obtained with the iterative procedure , it is necessary to carry the expansion in @xmath0 to higher order .
for example , in a slab @xcite , it is necessary to obtain the hamiltonian to second order in @xmath0 .
the second order piece of the hamiltonian , quadratic in the electrostatic potential @xmath24 , gives second - order corrections to the equations of motion and hence it is in principle negligible to first order .
however , if the lowest - order quasineutrality equation that contains a linear term in @xmath24 is employed , the second - order correction to the hamiltonian must be kept to obtain an energy - like invariant .
the complete calculation to order @xmath1 has not been done for a general static magnetic field in either formalism so far was given in @xcite . ] .
in the most common lagrangian formulation @xcite , the calculation is done in two steps : first , the turbulent electromagnetic fields are ignored and only the background magnetic field is considered , giving the drift kinetic equation ; in the second step , the turbulent electromagnetic fields are added and the corresponding corrections are calculated . consider the case in which the magnetic field does not vary in time , i.e. electrostatic gyrokinetics . in the first step , the equations are expanded in the small parameter @xmath25 , whereas in the second step , they are expanded in @xmath26 .
the expansion in @xmath0 is only performed to first order because the next order results are very tedious to calculate .
the expansion in @xmath27 is continued to second order because the pieces quadratic in @xmath24 are needed to have an energy - like invariant . in the expansion in @xmath27 ,
the fact that there has been a previous expansion in @xmath0 is ignored , and as a result the terms of order @xmath28 are never calculated .
the missing terms of order @xmath1 and @xmath28 are comparable to the terms of order @xmath29 according to the gyrokinetic ordering in ( 1 ) , making this expansion consistent only when @xmath30 .
in addition , since the cross - terms that contain both the background magnetic field and the turbulent electrostatic potential , of order @xmath28 , are always neglected when the two - step method is presented , it is not obvious how to calculate them following that procedure . in this article
, we present the complete phase - space lagrangian calculation with the standard gyrokinetic ordering ( 1 ) , emphasizing the self - consistent calculation of the terms of order @xmath1 and @xmath28 . in the gyrokinetic equations that result from the new lagrangian , the magnetic geometry effects and the fluctuating potential appear together in the second - order terms , showing that geometry and turbulence can not be separated and dealt with independently .
our main result is the explicit expression for the second - order gyrokinetic hamiltonian given in equations ( [ hamiltonian_o2_2 ] ) , ( [ psi2_phi ] ) , ( [ psi2_phib ] ) and ( [ psi2_b ] ) .
it clearly exhibits the interplay between geometry and turbulence inherent to gyrokinetic theory , possessing terms of three types : terms quadratic in the electrostatic potential , terms that include both the electrostatic potential and the magnetic geometry , and terms that are purely geometrical . at this point
, it is fair to wonder about the motivations beyond formal coherence to carry out the expansion consistently to second order in @xmath0 .
keeping the second order piece of the hamiltonian that is quadratic in the electrostatic potential is necessary for the conservation of an energy - like invariant , as already noted above .
when the other second - order terms computed in this article are included , they have two effects : ( i ) the gyrokinetic poisson s equation is modified by the effect of the non - uniform magnetic field on the gyro - orbits , and ( ii ) the equations of motion are modified to second order .
both of these effects are not conventionally kept in gyrokinetic formulations , but they may be crucial for conservation of momentum .
conservation of momentum in full @xmath23 gyrokinetic formulations has been the center of a recent controversy @xcite . by assuming a gyrobohm level of turbulent transport of momentum at long wavelengths , catto and one of us , fip , have argued that to recover with a full @xmath23 model the correct transport of toroidal angular momentum in a tokamak , it is necessary to have gyrokinetic fokker - planck and poisson s equations correct to third order in @xmath0 in the high flow ordering , for which the average velocity of the ions @xmath31 is of the order of the ion thermal speed @xmath32 , and correct to fourth order in the low flow ordering , for which @xmath33 . in the case of slab gyrokinetics , a consistent calculation of the transport of momentum in the low flow ordering
requires the third order hamiltonian @xcite .
the requirements for a system with general geometry are still to be sorted out , and to do
so it is necessary to study the new terms presented here and terms of even higher order .
in addition to the issues raised for full @xmath23 simulations , the formulation presented here will be very useful for @xmath34 approaches to momentum transport in tokamaks in the low flow ordering .
reference @xcite presents a formulation of this problem in the electrostatic limit that requires the minimum number of modifications to existing @xmath34 simulations .
the most important conclusion in @xcite is that the turbulent pieces of the distribution function and the electrostatic potential have to be calculated to an order higher in @xmath0 than usual because the contribution to momentum transport from the lowest order pieces vanishes due to symmetry arguments @xcite .
these symmetry arguments do not hold if the higher order terms of the gyrokinetic equation are considered . to avoid calculating most of the next order corrections to the gyrokinetic vlasov and poisson s equations , reference @xcite has to resort to a subsidiary expansion based on the fact that in many tokamaks the poloidal component of the magnetic field is much smaller than the toroidal component .
the new contribution to the hamiltonian that we calculate here gives the self - consistent higher order contributions to the gyrokinetic vlasov equation for the first time .
only a higher order gyrokinetic poisson s equation is then lacking to obtain a complete @xmath34 formulation in the low flow ordering that does not require a small poloidal magnetic field ; this will be the subject of a future publication .
the rest of this article is organized as follows . in section [ sect_order ]
we write the non - dimensional phase - space lagrangian of a particle in an electromagnetic field .
the normalization shows explicitly the standard gyrokinetic ordering ( [ orderings ] ) . in the first part of section [ sect_theory ] we review the phase - space lagrangian approach to gyrokinetics to help the understanding of the calculation that follows . in the second half of this section
we proceed to obtain the gyrokinetic lagrangian to second order in our expansion parameter @xmath0 .
as mentioned above , this is our main result . in section [ sec :
gyroeqsmotion ] we obtain the vlasov equation from this lagrangian , and in section [ sec : gyropoissoneq ] we discuss the consequences of this formulation for poisson s equation .
the new vlasov and poisson s equations presented here are correct to second and first order in @xmath0 , respectively .
we should remark that in the limit where the electrostatic potential has a scale of variation much larger than the gyroradius of the species of interest , our gyrokinetic equations provide the highest order guiding - center equations that we are aware of . in section [ sec : fieldtheory ] we borrow tools from classical field theory to obtain poisson s equation in a different way .
we prove that there is an energy - like invariant and we discuss the stringent conditions on the equations to actually conserve it in a simulation .
we finish with a discussion of our results and the future lines of research in section [ sec : conclusions ] .
the appendices contain the most cumbersome parts of the calculation as well as some material included for completeness .
finally , we would like to stress that in this article we have given all our results in an explicit form that can be directly implemented in a computer code .
the phase - space lagrangian for the motion of a particle of mass @xmath11 and charge @xmath10 in an electromagnetic field is given by @xmath35 \cdot \frac{d { \mathbf{r}}}{dt } - h^{\mathbf{x}}({\mathbf{r } } , { \mathbf{v } } , t ) , \label{lagr_orig}\ ] ] with the hamiltonian @xmath36 here @xmath37 is the vector potential that is defined such that @xmath38 .
notice that the phase - space lagrangian depends on the position of the particle @xmath39 , its velocity @xmath40 , the time derivatives of both the position and the velocity , @xmath41 and @xmath42 , and the time @xmath43 . for convenience
, we will sometimes denote the phase - space coordinates @xmath44 as @xmath45 .
we use the superscript @xmath46 in the lagrangian ( [ lagr_orig ] ) because it is a function of the phase - space coordinates @xmath47 .
the equations of motion are obtained by finding the stationary points of the action @xmath48 = \int_{t_0}^{t_1 } dt\ , \mathcal{l}^{\mathbf{x}}({\mathbf{r}}(t ) , { \mathbf{v}}(t ) , \dot{{\mathbf{r}}}(t ) , \dot{{\mathbf{v}}}(t ) , t ) $ ] with respect to variations of the functions @xmath49 and @xmath50 subject to the constraints @xmath51 , @xmath52 , @xmath53 and @xmath54 . from this procedure
we obtain six equations of motion , namely @xmath55 and @xmath56 this differs from the standard lagrangian formalism where the lagrangian function depends only on @xmath39 , @xmath57 and @xmath43 .
actually , the phase - space lagrangian formalism can be viewed as a variational formulation of hamilton equations ( see , for example , @xcite ) . in plasma physics , it was first applied by littlejohn to guiding - center dynamics in @xcite .
the lagrangian ( [ lagr_orig ] ) is non - dimensionalized using the characteristic thermal velocity of the species of interest @xmath7 , the characteristic length @xmath58 and the characteristic time @xmath59 .
we assume that @xmath60 is a small parameter , with @xmath8 and @xmath61 the characteristic gyroradius and the characteristic gyrofrequency of the species of interest , and @xmath62 the characteristic magnitude of the magnetic field .
we assume that the characteristic time and length scales in the electrostatic potential are the sound gyroradius @xmath63 and the sound characteristic time @xmath64 , i.e. , @xmath65 , where @xmath24 has derivatives with respect to its arguments of order unity . here
@xmath66 is the sound speed , @xmath67 and @xmath68 are the mass and the gyrofrequency of the dominant ion species , usually singly charged , and @xmath69 is the characteristic electron temperature .
the assumption on the scales of the electrostatic potential can be easily relaxed to account for other time and spatial scales . since the electrostatic potential @xmath24 is a quantity that enters the equations of the different species , it is normalized using parameters that do not depend on the species , in particular the characteristic electron temperature @xmath69 , the magnitude of the electron charge @xmath12 and the mass of the dominant ion species @xmath67 .
this normalization will be useful in poisson s equation , where several species appear .
the new , non - dimensionalized variables are @xmath70 giving @xmath71 \cdot \frac{d \check{{\mathbf{r}}}}{d\check{t } } - \check{h}^{\check{{\mathbf{x } } } } ( \check{{\mathbf{r } } } , \check{{\mathbf{v } } } , \check{t } ) , \label{lagrange_non}\ ] ] with @xmath72 here , @xmath73 is the characteristic temperature of the species of interest , @xmath74 is the ratio between the sound gyroradius and the characteristic scale length , @xmath75 is the ratio between the sound gyroradius and the gyroradius of the species of interest , @xmath76 is the ratio between the thermal speed of the species of interest and the sound speed , and @xmath77 even though the electrostatic potential is small , its perpendicular gradient is not . this assumption has been formally implemented by writing @xmath78 . here and in what follows
we assume @xmath79 .
this is the maximal ordering that contains in it several interesting regimes as subsidiary expansions in @xmath80 , @xmath81 and @xmath82 .
the form @xmath78 is , however , somewhat deceiving because the gradients along the magnetic field lines must be small , that is , @xmath83 , with @xmath84 the unit vector parallel to the magnetic field .
it is possible to formalize this condition by writing the functions in flux coordinates @xmath85 , @xmath86 and @xmath87 such that @xmath88 and @xmath89 . in these variables , the potential is given by @xmath90 to simplify the notation , we will often use @xmath91 instead of the most complete expression in ( [ phi_formal ] ) . where no confusion is possible
, we will write @xmath92 . in any case
, we always assume @xmath93 and @xmath94 note that @xmath0 is species - dependent whereas @xmath95 is not . in sections
[ sect_theory ] and [ sec : gyroeqsmotion ] where we compute the gyrokinetic phase - space lagrangian and the equations of motion of a single species , @xmath0 is the natural expansion parameter .
however , in sections [ sec : gyropoissoneq ] and [ sec : fieldtheory ] , devoted to the gyrokinetic poisson s equation , we need to consider several different species and @xmath96 is the appropriate , species - independent small parameter . finally , a notational remark is in order . in sections [ sect_theory ] and [ sec : gyroeqsmotion ]
we will be very careful to exhibit the dependence of our results on the mass , charge and temperature of the species through the parameters @xmath80 , @xmath81 and @xmath82 .
although at some places this may seem unnecessary and awkward ( and it would be if our objective were to treat always a single species ) , it is very convenient to write poisson s equation and the gyrokinetic phase - space lagrangian for a mixture of species in sections [ sec : gyropoissoneq ] and [ sec : fieldtheory ] .
from now on we will drop hats @xmath97 in the normalized expressions .
in this section we follow the general strategy of the applications of hamiltonian and phase - space lagrangian techniques to magnetized plasmas @xcite .
we search order by order in the small parameter @xmath0 for a change of phase - space variables such that only one of the variables has fast time dependence .
the gyrophase @xmath98 is the fast variable that evolves in the gyrofrequency time scale .
the rest of the phase - space variables ( gyrocenter position @xmath99 , parallel velocity @xmath100 and magnetic moment @xmath101 ) evolve with the much slower characteristic time scale @xmath59 . to achieve this , their time derivatives @xmath102 , @xmath103 and @xmath104 will be made independent of the gyrophase to the order of interest . had they depended on @xmath98 , they would necessarily show rapid time fluctuations on top of the more physically interesting slow time evolution . to make the time derivatives @xmath102 , @xmath103 and @xmath104 independent of the gyrophase @xmath98
, we search for a lagrangian that does not depend on @xmath98 ( it will still depend on its time derivative @xmath105 ) .
we first review briefly how to perform a change of variables in a phase - space lagrangian in subsection [ sub_change ] .
as we have already announced , the objective is the gyrokinetic lagrangian to order @xmath1 in general magnetic geometry .
since the calculation is quite long and complicated , we have sketched the derivation in subsection [ sub_procedure ] to offer the reader a global perspective of the formalism .
this subsection also contains our own proof that the algorithm to find the gyrokinetic change of variables can be carried out to any order and that there exists an adiabatic invariant @xmath101 to arbitrary order .
finally , in subsections [ sub_drift ] and [ sub_gyro ] we address the calculation of the gyrokinetic lagrangian to second order in detail .
some of the algebra is relegated to [ app_dk ] and [ app_gyro_2 ] .
the results to first order are compared with the iterative method in @xcite in [ app_comp ] . before proceeding
, we must mention that the phase - space lagrangian ( or hamiltonian ) approach to gyrokinetic theory has been geometrized @xcite .
we have chosen not to use the language of differential geometry to make the paper accessible to a broader audience , without losing mathematical rigor .
the reader familiar with the geometrical tools will realize that every step of our presentation can be translated into that language in an obvious way . consider a transformation @xmath106 that can be time dependent to a new set of gyrokinetic phase - space coordinates @xmath107 .
we write stands for the inverse of the transformation that we call @xmath106 .
] @xmath108 .
the phase - space lagrangian ( [ lagrange_non ] ) can be easily written in the new set of variables by using the chain rule , giving @xmath109 where @xmath110 \cdot \frac{\partial { \mathbf{r}}({\mathbf{z}},t)}{\partial z^\alpha}\ ] ] and @xmath111 \cdot \frac{\partial { \mathbf{r}}({\mathbf{z}},t)}{\partial t}.\ ] ] by finding the stationary points of the action @xmath112 = \int_{t_0}^{t_1 } \mathcal{l}^{\mathbf{z } } ( { \mathbf{z}}(t ) , \dot{\mathbf{z}}(t ) , t ) d t$ ] with respect to variations of @xmath113 subject to the conditions @xmath114 and @xmath115 , we obtain the new equations of motion @xmath116 note that the specific form in ( [ lagr_change ] ) implies that the equations of motion can be written as @xmath117 with @xmath118 the @xmath119 antisymmetric matrix @xmath120 although our gyrokinetic change of variables has an explicit time dependence due to the contribution of the electrostatic potential , we will show that it is possible to choose the functions @xmath121 such that @xmath122 . in our derivation
we impose then that @xmath123 .
consequently , we drop the last term in ( [ eqmotionwithuglyterm ] ) and write the equations of motion as @xmath124 from expression ( [ eqmotion ] ) we define the poisson bracket @xmath125 with @xmath126 the inverse and [ app : poissonbrackets ] we show that the matrix @xmath118 of our particular problem is indeed invertible . ] of the antisymmetric matrix defined in ( [ lagr_bra ] ) .
then @xmath127 noting that @xmath118 satisfies ( [ eq : closedform ] ) with @xmath128 , the proof in [ app_poissonbracket ] guarantees that ( [ poi_bra ] ) actually defines a poisson bracket , i.e. , that for any three functions @xmath129 , @xmath130 and @xmath131 , the bracket satisfies skew - symmetry @xmath132 the leibniz rule @xmath133 and the jacobi identity @xmath134 in general , it is impossible to find a change of phase - space variables that makes the lagrangian as written in ( [ lagr_change ] ) independent of gyrophase .
however , the time derivatives @xmath135 may be gyrophase independent even if the phase - space lagrangian is not .
this apparent discrepancy is easily solved considering that the equations of motion remain the same if instead of the lagrangian ( [ lagr_change ] ) we employ @xmath136 where the function @xmath137 depends on the phase - space variables @xmath138 and @xmath43 . indeed , finding the stationary points of the action @xmath139 = \int_{t_0}^{t_1 } \overline{\mathcal{l } } ( { \mathbf{z } } , \dot{\mathbf{z } } , t ) d t = \int_{t_0}^{t_1 } \mathcal{l}^{\mathbf{z } } ( { \mathbf{z } } , \dot{\mathbf{z } } , t ) d t+ s({\mathbf{z}}_{1 } ,
t_1 ) - s({\mathbf{z}}_{0 } , t_0)\end{aligned}\ ] ] with respect to variations of @xmath113 subject to the conditions @xmath114 and @xmath115 gives the same equations of motion as finding the stationary points of the action @xmath112 $ ] because both actions differ only by terms that are held constant .
since the lagrangian is not unique , we are not going to search for new phase - space variables @xmath138 such that the phase - space lagrangian @xmath140 in ( [ lagr_change ] ) is gyrophase independent , but such that there exists a function @xmath141 for which the phase - space lagrangian @xmath142 in ( [ lagr_prime ] ) is gyrophase independent .
this is equivalent to requiring that the time derivatives @xmath135 be gyrophase independent .
thus , we are searching for both the change of variables @xmath138 and the function @xmath141 such that the lagrangian @xmath142 in ( [ lagr_prime ] ) is gyrophase independent .
explicitly , @xmath143 where @xmath144 \cdot \frac{\partial { \mathbf{r}}({\mathbf{z}},t)}{\partial z^\alpha } + \frac{\partial s({\mathbf{z}},t)}{\partial z^\alpha}\ ] ] and @xmath145 \cdot \frac{\partial { \mathbf{r}}({\mathbf{z}},t)}{\partial t } - \frac{\partial s({\mathbf{z}},t)}{\partial t}\ ] ] are gyrophase independent . in what follows , @xmath138 only refers to the gyrokinetic phase - space coordinates .
notice that equations ( [ eqmotionwithuglyterm ] ) , ( [ lagr_bra ] ) , ( [ eqmotion ] ) , ( [ poi_bra ] ) and ( [ ddt_poibra ] ) are valid for the new lagrangian @xmath142 .
simply replace @xmath121 by @xmath146 and @xmath147 by @xmath148 .
as advanced in the introduction , the detailed computation to obtain the final form of the gyrokinetic lagrangian to order @xmath1 is rather involved .
this is why we devote this subsection to schematically show the steps leading to the determination of the change of variables and the function @xmath141 order by order .
we also give an easy proof that the algorithm can be carried out up to arbitrary order ( although in practice the computations would become prohibitively difficult ) . the detailed calculation to second order is done in subsections [ sub_drift ] and [ sub_gyro ] .
our transformation to new phase - space coordinates will be denoted by @xmath149 , explicitly in @xmath150 because the transformation is in general time dependent . ]
@xmath151 where @xmath99 is the gyrocenter position , @xmath100 is the gyrocenter parallel velocity , @xmath101 is the magnetic moment and @xmath98 is the gyrophase . for convenience
, we have written the transformation @xmath149 as the composition of two other transformations that we call _ non - perturbative transformation _ @xmath152 and _ perturbative transformation _ @xmath153 .
first , we perform a non - perturbative change of coordinates @xmath154 where @xmath155 , @xmath156 , @xmath157 and @xmath158 are lowest order approximations to the gyrocenter position , parallel velocity , magnetic moment and gyrophase .
their detailed definitions can be found in subsection [ sub_drift ] .
for simplicity , we will sometimes use the notation @xmath159 .
physically , the non - perturbative change of variables in subsection [ sub_drift ] amounts to saying that to lowest order the gyromotion is circular .
this is a consequence of the smallness of the gyroradius and the fact that the corrections due to the electrostatic potential are of next order . after the change of coordinates we add the total time derivative of a function @xmath160 to the lagrangian .
the details of the calculation are given in subsection [ sub_drift ] .
the final result is @xmath161 \cdot \frac{d{\mathbf{r}}_g}{dt } \nonumber \\ + \left ( - \mu_g + \epsilon \gamma_\theta^{(1 ) } + \epsilon^2 \gamma_\theta^{(2 ) } + \ldots \right ) \frac{d\theta_g}{d(t/\epsilon ) } - h^{(0 ) } ( { \mathbf{r}}_g , v_{||g } , \mu_g ) - \epsilon h^{(1 ) } , \label{lagr_ini}\end{aligned}\ ] ] where @xmath162 and the rest of the terms are defined in subsection [ sub_drift ] .
notice that with the notation in ( [ lagr_ini ] ) we have made explicit the fact that @xmath163 , i.e. , its time variation is of the order of the gyrofrequency time scale and hence much faster than the evolution of the rest of the phase - space variables .
the change of variables is non - perturbative and will give contributions to all orders in @xmath0 .
all the higher order terms @xmath164 , @xmath165 , ... , @xmath166 , @xmath167 , ... and @xmath168 depend on gyrophase . in the lagrangian ( [ lagr_ini ] )
, the hamiltonian has only the first order correction @xmath169 , with the higher order corrections being exactly zero .
this correction @xmath168 is the only term in the lagrangian that has strong perpendicular gradients because it is the only contribution that depends on the electrostatic potential . the gyrophase dependence in the lagrangian ( [ lagr_ini ] )
must be eliminated with the definition of the gyrokinetic variables order by order .
since we only calculate the gyrokinetic variables to some order , we truncate the expansion in @xmath0 to the order of interest . in general
, we need to keep @xmath170 \cdot \frac{d{\mathbf{r}}_g}{dt } \nonumber \\
+ \left [ - \epsilon \mu_g + \sum_{i = 1}^n \epsilon^{i+1 } \gamma_\theta^{(i ) } ( { \mathbf{r}}_g , v_{||g } , \mu_g , \theta_g ) \right ]
\frac{d\theta_g}{dt } \nonumber \\ - h^{(0 ) } ( { \mathbf{r}}_g , v_{||g } , \mu_g ) - \epsilon h^{(1 ) } ( { \mathbf{r}}_{g\bot}/\epsilon , r_{g|| } , \mu_g , \theta_g , t ) + o(\epsilon^{n+1 } , \epsilon^{n+2 } ) . \label{lagr_n}\end{aligned}\ ] ] here , we have not written explicitly @xmath171 . instead , we keep the terms that are multiplying @xmath172 to higher order in @xmath0 . the notation @xmath173 will be extensively used in this paper and indicates that the terms of order @xmath174 that we have neglected are either proportional to @xmath175 or are in the hamiltonian , and that the terms of order @xmath176 that we have neglected are proportional to @xmath177 , @xmath178 and @xmath172 ( the lagrangian ( [ lagr_n ] ) does not contain terms proportional to @xmath177 or @xmath178 ) . when we perform the expansion to obtain the gyrokinetic variables order by order
, it will be apparent that this notation is convenient because we need to keep some terms to @xmath179 and the rest to @xmath180 . employing expression ( [ lagr_n ] ) , we find the gyrokinetic variables by eliminating the gyrophase dependence order by order .
for now , it is enough to simply write the transformation @xmath181 to the order of interest as @xmath182 in subsection [ sub_gyro ] we show the connection of this expansion with lie transforms .
sometimes we will use the abbreviated notation @xmath183 .
notice that the variable @xmath99 is found to an order higher than the rest .
the corrections @xmath184 , @xmath185 , @xmath186 and @xmath187 are obtained by imposing that the lagrangian is gyrophase independent up to terms of order @xmath188 , where the terms of order @xmath189 are in the hamiltonian or are terms proportional to @xmath102 , and the terms of order @xmath174 are terms proportional to @xmath103 , @xmath104 and @xmath105 . to demonstrate the procedure ,
we show schematically how to obtain the first corrections @xmath190 , @xmath191 , @xmath192 and @xmath193 .
we then argue that the same formalism can be extended to arbitrary order .
the proof presented here demonstrates that the gyrokinetic variables can be consistently calculated order by order without running into problems .
recently , this has been a controversial issue @xcite . to calculate @xmath190 , @xmath191 , @xmath192 and @xmath193
, we need to express the lagrangian ( [ lagr_n ] ) in the new gyrokinetic variables to @xmath194 , giving should be included in @xmath140 ( [ lagr_n1 ] ) .
however , adding the time derivative of @xmath195 eliminates any @xmath196 dependence to this order .
the same can be said about ( [ lagr_ngk ] ) and ( [ lagrangian_gyro_o2_1 ] ) , where the terms @xmath197 and @xmath198 could be added .
it is very easy to show that the algorithm gives , of course , exactly the same results with either choice . ]
@xmath199 \cdot \frac{d { \mathbf{r}}}{dt } \nonumber \\ \fl & + \left [ \epsilon { \mathbf{a}}({\mathbf{r } } ) + \epsilon^2 u { \hat{\mathbf{b}}}({\mathbf{r } } ) \right ] \cdot \frac{d\tilde{{\mathbf{r}}}_2}{dt } + \left [ - \epsilon \mu + \epsilon^2 \left ( - \tilde{\mu}_1 + \gamma_\theta^{(1 ) } ( { \mathbf{r } } , u , \mu , \theta ) \right ) \right ] \frac{d \theta}{dt } - \epsilon^2 \mu \frac{d \tilde{\theta}_1}{dt } \nonumber \\ \fl & - h^{(0 ) } ( { \mathbf{r } } , u , \mu ) - \epsilon \left [ u \tilde{u}_1 + \tilde{\mu}_1 b({\mathbf{r } } ) + h^{(1 ) } ( { \mathbf{r}}_\bot/\epsilon , r_{|| } , \mu , \theta , t ) \right ] + o(\epsilon^2 , \epsilon^3 ) . \label{lagr_n1}\end{aligned}\ ] ] notice that in the functions @xmath200 , @xmath201 , @xmath202 and @xmath169 the variables @xmath155 , @xmath156 , @xmath157 and @xmath158 are replaced by the gyrokinetic variables @xmath99 , @xmath100 , @xmath101 and @xmath98 .
for example , from the definition of @xmath203 in ( [ ham_ini ] ) , we find @xmath204 the correction @xmath205 to the hamiltonian contains both the correction @xmath206 and the result of taylor expanding @xmath207 around @xmath99 , @xmath100 and @xmath101 , i.e. , @xmath208 in subsection [ sub_gyro ] we show that the expansion around @xmath99 gives a term that is an order higher and can be ignored to this order . as indicated in subsection [ sub_change ]
, we can always add a time derivative to the lagrangian ( [ lagr_n1 ] ) . by doing so
, we will get a form of the lagrangian in which the derivatives of a function @xmath209 enter . by imposing that the lagrangian is of a specific form , we first obtain the corrections @xmath190 , @xmath191 , @xmath192 and @xmath193 as functions of @xmath210 and we then find @xmath210 by integrating a simple differential equation .
we add to the lagrangian ( [ lagr_n1 ] ) the total time derivative @xmath211 , \label{add_time_derivative}\ ] ] where we have taken into account that @xmath210 depends on @xmath99 in two different ways : a fast dependence due to the potential that has wavelengths on the order of the gyroradius , and a slow dependence due to the background magnetic field .
notice that adding the time derivative ( [ add_time_derivative ] ) to the lagrangian ( [ lagr_n1 ] ) eliminates all the terms proportional to the time derivatives of the corrections @xmath190 , @xmath191 , @xmath192 and @xmath193 , giving @xmath212 \cdot \frac{d{\mathbf{r}}}{dt } \nonumber \\ \fl & + \epsilon^2 \left ( - { \hat{\mathbf{b}}}\cdot \tilde{{\mathbf{r}}}_2 + \frac{\partial s^{(2)}_p}{\partial u } \right ) \frac{du}{dt } + \epsilon^2 \left ( \tilde{\theta}_1 + \frac{\partial s^{(2)}_p}{\partial \mu } \right ) \frac{d\mu}{dt } \nonumber \\ \fl & + \left [ - \epsilon \mu + \epsilon^2 \left ( - \tilde{\mu}_1 + \gamma_\theta^{(1 ) } + \frac{\partial s^{(2)}_p}{\partial \theta } \right ) \right ] \frac{d\theta}{dt } - h^{(0 ) } \nonumber \\ \fl & - \epsilon \left ( u \tilde{u}_1 + \tilde{\mu}_1 b + h^{(1 ) } \right ) + o(\epsilon^2 , \epsilon^3 ) , \label{lagr_n1_2}\end{aligned}\ ] ] where we have used @xmath213 trivially deduced from the identity @xmath214_{ij } = \frac{\partial a_j}{\partial r_i } - \frac{\partial a_i}{\partial r_j } = \sum_{k=1}^3\varepsilon_{ijk}(\nabla_{\mathbf{r}}\times{\mathbf{a}})_{k}.\ ] ] here @xmath215 is the levi - civita symbol and the superscript @xmath216 stands for matrix transposition .
notice that in equation ( [ lagr_n1_2 ] ) we have not made explicit the dependence of the functions on the gyrokinetic variables , but it is assumed that all the terms are functions of @xmath99 , @xmath100 , @xmath101 and @xmath98 . by imposing that the lagrangian ( [ lagr_n1_2 ] ) be equal to @xmath217 where @xmath218 , @xmath219 and @xmath220 are gyrophase independent , we obtain the equations @xmath221 , \label{r2_1}\ ] ] @xmath222 , \label{u1_1}\ ] ] @xmath223 and @xmath224 the corrections @xmath190 , @xmath191 , @xmath192 and @xmath193 can then be found if we obtain @xmath210 .
to do so , we use that @xmath225 first , we take the gyroaverage of this equation to obtain @xmath220 . using the results in equations ( [ u1_1 ] ) and ( [ mu1_1 ] ) ,
the gyroaverage of equation ( [ eq_h1gk ] ) becomes @xmath226 + b({\mathbf{r } } ) \left \langle \gamma_\theta^{(1 ) } ( { \mathbf{r } } , u , \mu , \theta ) \right \rangle \nonumber \\
+ \left \langle h^{(1 ) } ( { \mathbf{r}}_\bot/\epsilon , r_{|| } , \mu , \theta , t ) \right \rangle , \label{eq_hbar1gk}\end{aligned}\ ] ] where @xmath227 is the gyroaverage holding @xmath99 , @xmath100 , @xmath101 and @xmath43 fixed .
notice that we have been able to obtain @xmath220 without explicitly finding @xmath210 . solving for @xmath210 is easy once @xmath220 is known . combining equations ( [ eq_h1gk ] ) and ( [ eq_hbar1gk ] ) ,
we find @xmath228 \nonumber \\ - \left [ \gamma_\theta^{(1 ) } ( { \mathbf{r } } , u , \mu , \theta ) - \left \langle \gamma_\theta^{(1 ) } ( { \mathbf{r } } , u , \mu , \theta ) \right \rangle \right ] \nonumber \\ - \frac{1}{b({\mathbf{r } } ) } \left [ h^{(1 ) } ( { \mathbf{r}}_\bot/\epsilon , r_{|| } , \mu , \theta , t ) - \left \langle h^{(1 ) } ( { \mathbf{r}}_\bot/\epsilon , r_{|| } , \mu , \theta , t ) \right \rangle \right ] . \label{eq_s1gk}\end{aligned}\ ] ] it is then straightforward to obtain @xmath210 by integrating in the gyrophase @xmath98 . the specific calculation of the corrections @xmath190 , @xmath191 , @xmath192 , @xmath193 and @xmath220 is done in subsection [ sub_gyro ] . here
we only want to demonstrate the procedure .
notice that the final solution depends on our choice of @xmath218 and in general on our choice of the final expression for @xmath142 .
however , once the choice is made , @xmath220 is completely determined .
once @xmath190 , @xmath191 , @xmath192 , @xmath193 , @xmath220 and @xmath210 have been determined , we can prove by induction that the procedure gives the corrections to any order .
assume that the corrections @xmath229 , @xmath230 , @xmath231 , @xmath232 and @xmath233 and the functions @xmath234 have been obtained up to @xmath235 in such a way that the lagrangian is of the form @xmath236 \cdot \frac{d{\mathbf{r}}}{dt } - \epsilon \mu \frac{d\theta}{dt } - \overline{h}^{(0 ) } - \sum_{i=1}^{n-1 } \epsilon^i \overline{h}^{(i ) } \nonumber \\
+ o(\epsilon^{n } , \epsilon^{n+1}).\end{aligned}\ ] ] then it is possible to obtain the corrections @xmath184 , @xmath185 , @xmath186 and @xmath187 , the function @xmath237 and the phase - space lagrangian to order @xmath238 . to do so
, the lagrangian is written to @xmath239 .
in general , we can not guess the exact form of the lagrangian to such high order without doing the calculation order by order , but the terms that contain the corrections @xmath184 , @xmath185 , @xmath186 and @xmath187 are very easy to obtain .
we find that @xmath240 \cdot \frac{d{\mathbf{r}}}{dt } \nonumber \\ \fl & + \left [ \epsilon^n { \mathbf{a}}({\mathbf{r } } ) + \epsilon^{n+1 } u { \hat{\mathbf{b}}}({\mathbf{r } } ) \right ] \cdot \frac{d\tilde{{\mathbf{r}}}_{n+1}}{dt } + \epsilon^{n+1 } \tilde{\gamma}_u^{(n ) } \frac{du}{dt}+ \epsilon^{n+1 } \tilde{\gamma}_\mu^{(n ) } \frac{d\mu}{dt } \nonumber \\ \fl & + \left [ - \epsilon \mu + \epsilon^{n+1 } \left ( - \tilde{\mu}_n + \tilde{\gamma}_\theta^{(n ) } \right ) \right ] \frac{d\theta}{dt } - \epsilon^{n+1 } \mu \frac{d\tilde{\theta}_n}{dt } - \overline{h}^{(0 ) } - \sum_{i=1}^{n-1 } \epsilon^i \overline{h}^{(i ) } \nonumber \\ \fl & - \epsilon^n \left [ u \tilde{u}_n + \tilde{\mu}_n b({\mathbf{r } } ) + \tilde{h}^{(n ) } \right ] + o(\epsilon^{n+1 } , \epsilon^{n+2 } ) .
\label{lagr_ngk}\end{aligned}\ ] ] here we have just separated the terms of order @xmath188 into those that depend on the corrections @xmath184 , @xmath185 , @xmath186 and @xmath187 , and the rest that we have lumped into the terms @xmath241 , @xmath242 , @xmath243 , @xmath244 and @xmath245 .
the form of the lagrangian ( [ lagr_ngk ] ) is very similar to the lagrangian ( [ lagr_n1 ] ) .
we can then use the same procedure .
we add the time derivative @xmath246\ ] ] to cancel all the terms that are proportional to the time derivatives of the corrections @xmath184 , @xmath185 , @xmath186 and @xmath187 .
then , by imposing that the lagrangian ( [ lagr_ngk ] ) plus this time derivative be equal to @xmath247 we obtain the equations @xmath248 @xmath249 @xmath250 and @xmath251 the @xmath252-th correction to the hamiltonian becomes @xmath253 and the equation for @xmath237 is @xmath254 in subsections [ sub_drift ] and [ sub_gyro ] we obtain the phase - space lagrangian to @xmath255 .
specifically , in subsection [ sub_drift ] and [ app_dk ] we derive equation ( [ lagr_ini ] ) . in subsection
[ sub_gyro ] and [ app_gyro_2 ] we use the perturbation procedure explained here to go from equation ( [ lagr_ini ] ) to the final result .
we perform a change of variables @xmath256 defined by @xmath257 and @xmath258 with the gyroradius vector defined as @xmath259.\ ] ] the unit vectors @xmath260 and @xmath261 are orthogonal to each other and to @xmath14 , and satisfy @xmath262 at every location @xmath39 . physically , @xmath155 is the guiding center position , @xmath156 the velocity parallel to the magnetic field at the guiding center position , @xmath157 the lowest order magnetic moment , and @xmath158 the lowest order gyrophase . for a homogeneous static magnetic field and in the absence of electric field , the change of coordinates defined by ( [ eq : changer ] ) , ( [ eq : changev ] ) and ( [ eq : defrho ] ) exactly eliminates the gyrophase dependence . note in passing that it is a well defined change of coordinates .
these formulae explicitly give @xmath263 as a function of @xmath264 and it is easy to see that the transformation is invertible for small @xmath0 : it is clearly invertible for @xmath265 and the transformation is continuous in @xmath0 .
we substitute the relations @xmath266 , given in ( [ eq : changer ] ) and ( [ eq : changev ] ) , and @xmath267 into the non - dimensionalized lagrangian ( [ lagrange_non ] ) , @xmath268 .
the resulting lagrangian that we denote as @xmath269 differs from the lagrangian @xmath270 in ( [ lagr_ini ] ) and ( [ lagrange_drift_4 ] ) by the time derivative of a function @xmath271 and even though both lagrangians give the same equations of motion , we have decided to stress the difference . the lagrangian @xmath272 is @xmath273 \cdot \frac{d}{dt } \left ( { \mathbf{r}}_g + \epsilon { \mbox{\boldmath$\rho$}}\right ) \nonumber \\ - h^{(0 ) } - \epsilon h^{(1 ) } , \label{lagrange_drift_1}\end{aligned}\ ] ] with @xmath274 defined in ( [ ham_ini ] ) and @xmath275 where @xmath276 is the heaviside step function , with @xmath277 for @xmath278 and @xmath279 for @xmath280 . any magnetic quantity with subindex @xmath281
is evaluated at @xmath155 , e.g. , @xmath282 .
we write @xmath283 when no confusion is possible .
we have defined a new function @xmath284 ( notice the difference in the font between @xmath285 and @xmath24 ) given by @xmath286 then @xmath287 is the gyroaverage of @xmath285 , @xmath288 and @xmath289 the gyrophase dependent piece , @xmath290 we now prove that the notation in ( [ h1 ] ) , @xmath291 is appropriate .
first , we show that @xmath157 is always divided by @xmath292 and that the sign of @xmath81 determines the phase of @xmath158 , and later we demonstrate that if conditions ( [ eq : orderingvarphi_par ] ) and ( [ eq : orderingvarphi_perp ] ) are satisfied , then @xmath293 and @xmath294 are also satisfied . to show that @xmath157 always appears divided by @xmath292 and that we need to add @xmath295 to @xmath158 when @xmath81 is negative , it is enough to realize that @xmath285 depends on @xmath157 and @xmath158 through the dependence of @xmath24 on @xmath296 and that @xmath297 as defined in ( [ eq : defrho ] ) only depends on @xmath157 through the multiplying term @xmath298 .
it is then obvious that @xmath299 . to prove that ( [ eq : orderingvarphi_par ] ) and ( [ eq : orderingvarphi_perp ] )
imply ( [ eq : orderingphi_par ] ) and ( [ eq : orderingphi_perp ] ) , we employ @xmath300 with @xmath301 .
then , using equations ( [ eq : orderingvarphi_par ] ) and ( [ eq : orderingvarphi_perp ] ) , it is easy to see that equations ( [ eq : orderingphi_par ] ) and ( [ eq : orderingphi_perp ] ) are correct .
note that when we write @xmath168 in ( [ h1 ] ) we are emphasizing the dependence on @xmath0 because the asymptotic procedure is based on expanding in @xmath302 .
the dependence on @xmath80 , @xmath81 and @xmath82 is only written explicitly in the function @xmath285 .
we now show how to simplify ( [ lagrange_drift_1 ] ) . employing @xmath303 @xmath304 and @xmath305 we write the lagrangian in ( [ lagrange_drift_1 ] ) as @xmath306 \cdot \frac{d { \mathbf{r}}_g}{dt } + \frac{1}{2\mu_g } { \mathbf{a}}({\mathbf{r}}_g + \epsilon { \mbox{\boldmath$\rho$ } } ) \cdot { \mbox{\boldmath$\rho$}}\frac{d\mu_g}{dt } \nonumber \\ \fl & + \left [ { \mathbf{a}}({\mathbf{r}}_g + \epsilon { \mbox{\boldmath$\rho$ } } ) \cdot \frac{\partial { \mbox{\boldmath$\rho$}}}{\partial \theta_g } - 2 \epsilon \mu_g \right ] \frac{d\theta_g}{dt } - h^{(0 ) } - \epsilon h^{(1 ) } .
\label{lagrange_drift_2}\end{aligned}\ ] ] to obtain ( [ gradrho ] ) we have used @xmath307 and @xmath308 . to simplify the lagrangian ( [ lagrange_drift_2 ] ) , we add the time derivative of @xmath309 as a result we find @xmath310 \cdot \frac{d { \mathbf{r}}_g}{dt } \nonumber \\ \fl & + \left [ { \mathbf{a}}({\mathbf{r}}_g + \epsilon { \mbox{\boldmath$\rho$ } } ) \cdot \frac{\partial { \mbox{\boldmath$\rho$}}}{\partial \theta_g } - 2 \epsilon \mu_g + \frac{\partial s_{np}}{\partial \theta_g } \right ] \frac{d\theta_g}{dt } - h^{(0 ) } - \epsilon h^{(1)}. \label{lagrange_drift_3}\end{aligned}\ ] ] in [ app_dk ] we prove that @xmath311 + \epsilon [ ( { \mbox{\boldmath$\rho$}}^\prime \times { \hat{\mathbf{b}}}_g ) \cdot { \mathbf{b}}({\mathbf{r}}_g + \epsilon { \mbox{\boldmath$\rho$}}^\prime ) ] \nabla_{{\mathbf{r}}_g } { \hat{\mathbf{b}}}_g \cdot { \mbox{\boldmath$\rho$}}^\prime \nonumber \\ \fl & + \frac{2\epsilon \mu_g^\prime}{b_g } [ { \hat{\mathbf{b}}}_g \cdot { \mathbf{b}}({\mathbf{r}}_g + \epsilon { \mbox{\boldmath$\rho$}}^\prime ) - b_g ] \nabla_{{\mathbf{r}}_g } { \hat{\mathbf{e}}}_{2 g } \cdot { \hat{\mathbf{e}}}_{1 g } \bigg \}\end{aligned}\ ] ] and @xmath312,\ ] ] where we use the abbreviated notation @xmath313 . substituting equations ( [ ds_npdr ] ) and ( [ ds_npdtheta ] ) into the lagrangian ( [ lagrange_drift_3 ] ) finally gives @xmath314 with @xmath315 \nonumber \\ \fl & + [ ( { \mbox{\boldmath$\rho$}}^\prime \times { \hat{\mathbf{b}}}_g ) \cdot { \mathbf{b}}({\mathbf{r}}_g + \epsilon { \mbox{\boldmath$\rho$}}^\prime ) ] \nabla_{{\mathbf{r}}_g } { \hat{\mathbf{b}}}_g \cdot { \mbox{\boldmath$\rho$}}^\prime + \frac{2 \mu_g^\prime}{b_g } [ { \hat{\mathbf{b}}}_g \cdot { \mathbf{b}}({\mathbf{r}}_g + \epsilon { \mbox{\boldmath$\rho$}}^\prime ) - b_g ] \nabla_{{\mathbf{r}}_g } { \hat{\mathbf{e}}}_{2 g } \cdot { \hat{\mathbf{e}}}_{1 g } \bigg \}\end{aligned}\ ] ] and @xmath316.\ ] ] it is easy to write the lagrangian ( [ lagrange_drift_4 ] ) order by order
. we use @xmath317 where our double - dot convention is @xmath318 , to obtain @xmath319 where @xmath320 { \mathbf{b}}_g , \label{gammar_o1}\end{aligned}\ ] ] @xmath321 \nabla_{{\mathbf{r}}_g } { \hat{\mathbf{b}}}_g \cdot { \mbox{\boldmath$\rho$}}\nonumber \\ - \frac{2\mu_g}{3b_g } ( { \mbox{\boldmath$\rho$}}\cdot \nabla_{{\mathbf{r}}_g } b_g ) \nabla_{{\mathbf{r}}_g } { \hat{\mathbf{e}}}_{2 g } \cdot { \hat{\mathbf{e}}}_{1 g } , \label{gammar_o2}\end{aligned}\ ] ] @xmath322 and @xmath323 in this subsection we find a new set of coordinates @xmath324 that makes the lagrangian ( [ lagrangian_drift_5 ] ) gyrophase independent .
we employ the procedure described in subsection [ sub_procedure ] .
the transformation @xmath325 is customarily written in the form of a lie transform @xcite , @xmath326 where @xmath327.\ ] ] instead of this form , we use the form in ( [ newvar ] ) that we find more convenient .
the connection between the two arrangements is trivial .
to first order we find @xmath328 to second order , the relation is @xmath329 @xmath330 @xmath331 and @xmath332 in subsection [ ss_gyro_1 ] , the corrections @xmath333 , @xmath334 , @xmath335 and @xmath336 are calculated following the procedure in subsection [ sub_procedure ] , and the lagrangian is obtained to @xmath194 . in subsection [ ss_gyro_2 ] the lagrangian
is obtained to next order .
it is possible to do so without explicitly obtaining @xmath337 , @xmath338 , @xmath339 and @xmath340 .
we obtain the first - order gyrokinetic correction to the hamiltonian , @xmath220 , by employing equation ( [ eq_hbar1gk ] ) .
we need to know that @xmath341 , @xmath342 and @xmath343 where we have used that @xmath344 with @xmath345 the unit matrix . for the remainder of the section , whenever we do not write explicitly the arguments of the functions , it will be understood that they are evaluated at @xmath346 , i.e. @xmath347 , @xmath348 , and so on . substituting the values of @xmath349 , @xmath350 and @xmath351 into equation ( [ eq_hbar1gk ] ) , we find @xmath352 notice that we have the freedom to choose @xmath218 as we wish
. our choice will affect the corrections @xmath333 and @xmath334 , and the final form of @xmath220 . to coincide with previous derivations in the literature @xcite , we choose @xmath353 giving @xmath354 in equation ( [ gammarbar1_final ] ) ,
we have chosen @xmath355 instead of @xmath356 to manifestly show that the equations of motion are independent of the choice of @xmath357 and @xmath358 @xcite .
the function @xmath210 can be obtained by solving equation ( [ eq_s1gk ] ) .
substituting equations ( [ h1 ] ) , ( [ gammar_o1 ] ) and ( [ gammatheta_o1 ] ) into ( [ eq_s1gk ] ) gives @xmath359 : \nabla_{\mathbf{r}}{\hat{\mathbf{b}}}- \frac{2\mu}{3b } { \mbox{\boldmath$\rho$}}\cdot \nabla_{\mathbf{r}}b - \frac{\lambda { \widetilde{\phi}}}{b } , \label{ds2dtheta}\end{aligned}\ ] ] where we have used that @xmath360.\ ] ] integrating equation ( [ ds2dtheta ] ) in the gyrophase gives @xmath361 : \nabla_{\mathbf{r}}{\hat{\mathbf{b}}}\nonumber \\ - \frac{2\mu}{3b } ( { \mbox{\boldmath$\rho$}}\times { \hat{\mathbf{b } } } ) \cdot \nabla_{\mathbf{r}}b - \frac{\lambda { \widetilde{\phi}}}{b } , \label{s2}\end{aligned}\ ] ] where the function @xmath362 is the integral @xmath363 such that @xmath364 .
here we have used that @xmath365 and @xmath366/\partial \theta$ ] . using @xmath210 in the expressions ( [ r2_1 ] ) , ( [ u1_1 ] ) , ( [ mu1_1 ] ) and ( [ theta1_1 ] ) , the first order corrections to the gyrokinetic variables
become @xmath367 : \nabla_{\mathbf{r}}{\hat{\mathbf{b}}}- \frac{u}{b } { \hat{\mathbf{b}}}\times \nabla_{\mathbf{r}}{\hat{\mathbf{b}}}\cdot { \mbox{\boldmath$\rho$}}\nonumber \\ - \frac{1}{2b } { \mbox{\boldmath$\rho$}}{\mbox{\boldmath$\rho$}}\cdot \nabla_{\mathbf{r}}b - \frac{\lambda}{\lambda b^2 } { \hat{\mathbf{b}}}\times \nabla_{({\mathbf{r}}_\bot/\lambda \epsilon ) } { \widetilde{\phi } } , \label{r2}\end{aligned}\ ] ] @xmath368 : \nabla_{\mathbf{r}}{\hat{\mathbf{b } } } , \label{u1}\ ] ] @xmath369 : \nabla_{\mathbf{r}}{\hat{\mathbf{b}}}- \frac{\lambda { \widetilde{\phi}}}{b } \label{mu1}\ ] ] and @xmath370 : \nabla_{\mathbf{r}}{\hat{\mathbf{b}}}\nonumber \\ + \frac{1}{b } ( { \mbox{\boldmath$\rho$}}\times { \hat{\mathbf{b } } } ) \cdot \nabla_{\mathbf{r}}b + \frac{\lambda}{\lambda^2 b } \frac{\partial { \widetilde{\phi}}}{\partial ( \mu/\lambda^2)}. \label{theta1}\end{aligned}\ ] ] in [ app_comp ] we show that this result is equivalent to the result obtained with the iterative method in @xcite . in this subsection
we apply the change of variables ( [ newvar ] ) to the lagrangian ( [ lagrangian_drift_5 ] ) to @xmath371 .
the idea is to write an expression similar to ( [ lagr_ngk ] ) with @xmath372 so that we can use the technique demonstrated in subsection [ sub_procedure ] . to @xmath255 ,
the lagrangian ( [ lagrangian_drift_5 ] ) becomes @xmath373 \cdot \frac{d{\mathbf{r}}}{dt } \nonumber \\ \fl & + \bigg [ \epsilon { \mathbf{a}}+ \epsilon^2 u { \hat{\mathbf{b}}}+ \epsilon^3 \left ( { \mbox{\boldmath$\gamma$}}_{\mathbf{r}}^{(1 ) } + { \mathbf{r}}_2 \cdot \nabla_{\mathbf{r}}{\mathbf{a}}+ u_1 { \hat{\mathbf{b}}}\right ) \bigg ] \cdot \frac{d{\mathbf{r}}_2}{dt } + \bigg ( \epsilon^2 { \mathbf{a}}+ \epsilon^3 u { \hat{\mathbf{b}}}\bigg ) \cdot \frac{d\tilde{{\mathbf{r}}}_3}{dt } \nonumber \\ \fl & + \bigg [ - \epsilon \mu + \epsilon^2 \left ( - \mu_1 + \gamma_\theta^{(1 ) } \right ) + \epsilon^3 \bigg ( - \tilde{\mu}_2 + \gamma_\theta^{(2 ) } + \mu_1 \frac{\partial \gamma_\theta^{(1)}}{\partial \mu } + \theta_1 \frac{\partial \gamma_\theta^{(1)}}{\partial \theta } \bigg ) \bigg ] \frac{d\theta}{dt } \nonumber \\ \fl & + \bigg [ - \epsilon^2 \mu + \epsilon^3 \left ( - \mu_1 + \gamma_\theta^{(1 ) } \right ) \bigg ] \frac{d\theta_1}{dt } - \epsilon^3 \mu \frac{d\tilde{\theta}_2}{dt } - h^{(0 ) } - \epsilon \left ( u u_1 + \mu_1 b + h^{(1 ) } \right ) \nonumber \\ \fl & - \epsilon^2 \bigg ( u \tilde{u}_2 + \frac{u_1 ^ 2}{2 } + \tilde{\mu}_2 b + \mu { \mathbf{r}}_2 \cdot \nabla_{\mathbf{r}}b + { \mathbf{r}}_2 \cdot \nabla_{({\mathbf{r}}_\bot/\epsilon ) } h^{(1 ) } + \mu_1 \frac{\partial h^{(1)}}{\partial \mu } + \theta_1 \frac{\partial h^{(1)}}{\partial \theta } \bigg ) \nonumber \\ \fl & + o(\epsilon^3 , \epsilon^4 ) , \label{lagrangian_gyro_o2_1}\end{aligned}\ ] ] where we have used that @xmath374 , that @xmath375 and that @xmath376 and @xmath377 only depend slowly on @xmath99 .
we have seen in subsection [ sub_procedure ] that to first order we need to add to the lagrangian ( [ lagrangian_gyro_o2_1 ] ) the time derivative @xmath378 giving as a result the lagrangian @xmath379 \cdot \frac{d{\mathbf{r}}}{dt } + \left ( \epsilon^2 { \mathbf{a}}+ \epsilon^3 u { \hat{\mathbf{b}}}\right ) \cdot \frac{d\tilde{{\mathbf{r}}}_3}{dt } \nonumber \\ \fl & + \epsilon^3 \tilde{\gamma}_u^{(2 ) } \frac{du}{dt } + \epsilon^3 \tilde{\gamma}_\mu^{(2 ) } \frac{d\mu}{dt } + \left [ - \epsilon \mu + \epsilon^3 \left ( - \tilde{\mu}_2 + \tilde{\gamma}_\theta^{(2 ) } \right ) \right ] \frac{d\theta}{dt } - \epsilon^3 \mu \frac{d\tilde{\theta}_2}{dt } - \overline{h}^{(0 ) } - \epsilon \overline{h}^{(1 ) } \nonumber \\ \fl & - \epsilon^2 \left [ u \tilde{u}_2 + \tilde{\mu}_2 b + \tilde{h}^{(2 ) } \right ] + o(\epsilon^3 , \epsilon^4 ) , \label{lagrangian_gyro_o2_2}\end{aligned}\ ] ] with @xmath380 @xmath381 @xmath382 @xmath383 and @xmath384 notice that in equation ( [ gammat_r_o2 ] ) we are using @xmath385 and we have taken into account that @xmath210 depends on @xmath99 in two different ways .
on the one hand , there is the dependence on @xmath386 , that was the only dependence that was taken into account in subsection [ ss_gyro_1 ] .
this dependence gives the strong gradient @xmath387 . on the other hand
there is a slow dependence on @xmath99 that gives the gradient @xmath388 + \frac{\lambda { \widetilde{\phi}}}{b^2 } \nabla_{\mathbf{r}}b - \frac{\lambda}{b } { \hat{\mathbf{b}}}{\hat{\mathbf{b}}}\cdot \nabla_{\mathbf{r}}\tilde\phi . \label{nablas2_slow}\end{aligned}\ ] ] in subsection [ sub_procedure ] we showed that by adding the total time derivative @xmath389 to the lagrangian ( [ lagrangian_gyro_o2_2 ] ) and making the result equal to @xmath390 \cdot \frac{d{\mathbf{r}}}{dt } - \epsilon \mu \frac{d\theta}{dt } - \overline{h}^{(0 ) } - \epsilon \overline{h}^{(1 ) } - \epsilon^2 \overline{h}^{(2 ) } + o(\epsilon^3 , \epsilon^4),\ ] ] where we have explicitly set @xmath391 , we obtain the equations for the corrections @xmath392
@xmath393 @xmath394 and @xmath395 the correction to the hamiltonian is @xmath396 and the equation for @xmath397 is @xmath398 in this article we do not solve for @xmath397 and hence we are not able to write explicitly the corrections @xmath337 , @xmath338 , @xmath339 and @xmath340 .
we only obtain explicitly the correction @xmath399 . in [ app_gyro_2 ]
we evaluate equation ( [ hbar2_1 ] ) in detail .
the final result is @xmath400 with @xmath401 @xmath402 \cdot \nabla_{\mathbf{r}}b \right \rangle - \frac{1}{b } \nabla_{\mathbf{r}}b \cdot \langle { \widetilde{\phi}}\ , { \mbox{\boldmath$\rho$}}\rangle \nonumber \\ \fl & - \frac{u^2}{\lambda^2 b } { \hat{\mathbf{b}}}\cdot \nabla_{\mathbf{r}}{\hat{\mathbf{b}}}\cdot \left \langle \frac{\partial { \widetilde{\phi}}}{\partial ( \mu/\lambda^2)}\ , { \mbox{\boldmath$\rho$}}\right \rangle - \frac{u^2}{2 \mu b } { \hat{\mathbf{b}}}\cdot \nabla_{\mathbf{r}}{\hat{\mathbf{b}}}\cdot \langle { \widetilde{\phi}}\ , { \mbox{\boldmath$\rho$}}\rangle \nonumber\\ \fl & + \frac{u}{4 \lambda^2 } \nabla_{\mathbf{r}}{\hat{\mathbf{b } } } : \left \langle \frac{\partial { \widetilde{\phi}}}{\partial ( \mu/\lambda^2)}\ , \left [ { \mbox{\boldmath$\rho$ } } ( { \mbox{\boldmath$\rho$}}\times { \hat{\mathbf{b } } } ) + ( { \mbox{\boldmath$\rho$}}\times { \hat{\mathbf{b } } } ) { \mbox{\boldmath$\rho$}}\right ] \right \rangle \nonumber \\ \fl & + \frac{u}{4 \mu } \nabla_{\mathbf{r}}{\hat{\mathbf{b } } } : \left \langle { \widetilde{\phi}}\ , \left [ { \mbox{\boldmath$\rho$ } } ( { \mbox{\boldmath$\rho$}}\times { \hat{\mathbf{b } } } ) + ( { \mbox{\boldmath$\rho$}}\times { \hat{\mathbf{b } } } ) { \mbox{\boldmath$\rho$}}\right ] \right \rangle \label{psi2_phib}\end{aligned}\ ] ] and @xmath403 where @xmath404 is the transpose of the matrix @xmath405 . the final phase - space lagrangian is given then by ( [ eq : lagrangiangyroorderepsilon2 ] ) .
we can write it explicitly as @xmath406 \cdot \frac{d{\mathbf{r}}}{dt } - \epsilon \mu \frac{d\theta}{dt } - \overline{h}+ o(\epsilon^3 , \epsilon^4),\ ] ] where @xmath407 and @xmath408 in previous work @xcite , only the contribution @xmath409 was kept because the terms that contained the function @xmath285 were assumed to be larger . with the more natural ordering ( [ orderings ] ) , we find the new contributions @xmath410 and @xmath411 , demonstrating that magnetic geometry and electrostatic potential appear together and can not be separated . in sections [ sec :
gyroeqsmotion ] , [ sec : gyropoissoneq ] and [ sec : fieldtheory ] we show that @xmath410 and @xmath411 modify both the equations of motion and poisson s equation .
we end this section pointing out that we chose the final form of the lagrangian ( [ finall ] ) , with @xmath412 , to have the same poisson brackets as previous authors @xcite .
there are other possible choices , e.g. , making the second order correction of the hamiltonian independent of the parallel velocity , condition that can be achieved by defining the appropriate @xmath413 .
the equations of motion are given by ( [ ddt_poibra ] ) . to obtain them explicitly
we need to find the poisson bracket ( [ poi_bra ] ) that corresponds to the lagrangian ( [ finall ] ) . employing [ app : poissonbrackets ] we find the poisson bracket to be @xmath414 where @xmath415 @xmath416 and @xmath417 employing the poisson bracket in ( [ eq : poissonbracket ] ) and the hamiltonian in ( [ finalh ] ) ,
we find @xmath418 @xmath419 \cdot \bigg ( \frac{\lambda \epsilon}{\lambda } \nabla_{({\mathbf{r}}_\bot/\lambda \epsilon ) } { \langle \phi \rangle}\nonumber \\ \fl & + \frac{\lambda^2 \epsilon^2}{\lambda } \nabla_{({\mathbf{r}}_\bot/\lambda \epsilon ) } \psi^{(2)}_\phi + \frac{\lambda \epsilon^2}{\lambda } \nabla_{({\mathbf{r}}_\bot/\lambda \epsilon ) } \psi^{(2)}_{\phi b } + \lambda^2 \epsilon^3 \nabla_{\mathbf{r}}\psi^{(2)}_\phi + \lambda \epsilon^3 \nabla_{\mathbf{r}}\psi^{(2)}_{\phi b } \nonumber \\ \fl & + \epsilon^3 \nabla_{\mathbf{r}}\psi^{(2)}_b \bigg ) , \end{aligned}\ ] ] @xmath420 and @xmath421 note that we have emphasized the fact that the dependence of the functions @xmath287 , @xmath422 and @xmath410 on @xmath99 and @xmath101 ( recall equation ( [ finalh ] ) ) can be fast or slow .
for this reason we distinguish between derivatives with respect to the argument @xmath423 and derivatives with respect to the argument @xmath99 , and between derivatives with respect to the argument @xmath101 and derivatives with respect to the argument @xmath424 . the new correction to the hamiltonian @xmath410
gives a contribution of order @xmath1 to the perpendicular and parallel motion of the gyrocenter , comparable to the contribution from @xmath409 , the term that is usually kept .
the correction to the hamiltonian @xmath411 gives a negligible contribution to the perpendicular drift , but is needed for the parallel motion .
thus , both corrections must be kept to obtain the equations of motion to order @xmath1 .
it is worth mentioning that equation ( [ drdt ] ) contains the baos drift @xcite in the definition of @xmath100 . to make it clear , instead of choosing the first order lagrangian as in ( [ lagr_n1_2_final ] ) with @xmath425 given in ( [ gammarbar1_final ] ) , we can choose it to have @xmath426 this choice gives a different parallel velocity @xmath427 and a different first order hamiltonian @xmath428 . with this new choice ,
the equation for @xmath429 to first order is @xmath430 note that the baos drift has been made explicit . from here on , we work only with our equations of motion ( [ drdt ] ) , ( [ dudt ] ) , ( [ dmudt ] ) and ( [ dthetadt ] ) that are equivalent to the equations obtained with this alternative choice that makes the baos drift manifest .
the gyrokinetic vlasov equation is readily written for the phase - space distribution @xmath431 in gyrokinetic coordinates , giving @xmath432 or employing the poisson bracket , @xmath433 in the absence of collisions and making use of the fact that @xmath434 , @xmath435 and @xmath436 are independent of gyrophase , the gyrophase independent piece of the distribution function @xmath437 and the gyrophase dependent piece @xmath438 are determined by two decoupled equations @xcite , namely , @xmath439 and @xmath440 since @xmath441 , it is usually assumed that any dependence on the gyrophase disappears in a very short time and @xmath442 can be neglected . as a result
, we assume that @xmath443 . by zeroing the gyrophase - dependent piece of the distribution function we are eliminating the gyrofrequency time scales , a crucial step to obtain the gyrokinetic equation . in the presence of collisions it is possible to estimate the size of @xmath442 and argue that it is small @xcite .
it is possible to write the gyrokinetic equation in conservative form .
indeed , the equations of motion obtained from a phase - space lagrangian conserve phase - space volume .
therefore , the determinant of the jacobian matrix of the gyrokinetic transformation , @xmath444 , found in [ app : jacobians ] to be @xmath445 , satisfies the condition @xmath446 for completeness , we prove this equation in [ app_consps ] . since @xmath436 and @xmath447 do not depend on @xmath98 , this equation reduces to @xmath448 using this expression and equation ( [ vlasov_ave ] ) , and taking into account that our choice of lagrangian ( [ finall ] ) implies that @xmath449 , we find the vlasov equation in conservative form , i.e. , @xmath450
in sections [ sect_theory ] and [ sec : gyroeqsmotion ] we have obtained the gyrokinetic equations of motion and the gyrokinetic vlasov equation keeping the electrostatic potential as an unspecified function .
the system of equations of electrostatic gyrokinetic theory is closed by coupling the gyrokinetic vlasov equation to poisson s equation , which is the subject of study of this section .
first , in subsection [ normalization_poissoneq ] we present the normalization that we employ for poisson s equation . since several species enter in poisson s equation , the normalization given in section [ sect_order ] must be modified . with this new normalization
, we obtain the gyrokinetic poisson s equation in subsection [ sec : gyropoissonchangevar ] by simply changing from the coordinates @xmath39 and @xmath40 to their gyrokinetic counterparts @xmath99 , @xmath100 , @xmath101 and @xmath98 . in section [ sec : fieldtheory ] we provide another way of obtaining poisson s equation that is based on field theory .
poisson s equation in gaussian units reads @xmath451,\end{aligned}\ ] ] where @xmath452 is the particle distribution of the dominant ions , @xmath453 and @xmath454 are the charge and the distribution function of the impurity @xmath455 , and @xmath456 is the distribution function of the electrons . using the species - independent normalization @xmath457 for time , space , vector potential and electrostatic potential , and the species - dependent normalization @xmath458 for the velocities and the distribution functions , we get @xmath459 here
, @xmath460 is a characteristic equilibrium value for the electron density and @xmath461 is the electron debye length .
recall that @xmath74 .
this new normalization is somewhat different from the normalization in section [ sect_order ] .
we have that @xmath462 , @xmath463 , @xmath464 and @xmath465 ( the subindex @xmath466 can take the value @xmath467 , @xmath12 , and any of the values of @xmath455 ) , but the normalization for the time is different , giving @xmath468 . the gyrokinetic variables obtained in the previous sections can be employed here without further changes because both the normalization used in section [ sect_order ] and the normalization in equations ( [ norm_spindep ] ) and ( [ norm_spdep ] ) give the same normalization for @xmath469 , @xmath470 , @xmath471 and @xmath472 .
this normalization is species - dependent and for this reason we use the subscripts @xmath473 to distinguish the gyrokinetic variables for ions , impurities and electrons . in the results of sections [ sect_order ] and [ sect_theory ] ,
the constants @xmath0 , @xmath80 , @xmath81 and @xmath82 depend on the species , and it will be useful to use subscripts for them as well . their values for ions , the impurity species @xmath455 and electrons
are given in table [ table_lambdas ] .
note the sign in the definition of @xmath474 and @xmath475 . in table
[ table_lambdas ] , @xmath476 and @xmath477 are the characteristic temperatures of the ion species and the impurity species @xmath455 , respectively , and @xmath478 is the mass of the impurity species @xmath455 .
the function @xmath285 in ( [ defphi ] ) is now species - dependent and given by @xmath479 later on we will use the the more compact notation @xmath480 [ table_lambdas ] l l l ions & impurity @xmath455 & electrons + @xmath481 & @xmath482 & @xmath483 + @xmath484 & @xmath485 & @xmath486 + @xmath487 & @xmath488 & @xmath489 + @xmath490 & @xmath491 & @xmath492 + to write the total lagrangian of the system , it is necessary to normalize all the particle lagrangians and hamiltonians by the same quantity .
we choose that the normalized lagrangian and hamiltonian be @xmath493 and @xmath494 .
the gyrokinetic lagrangian ( [ finall ] ) , dependent on the species , becomes in the new normalization @xmath495 \cdot \frac{d\underline{{\mathbf{r}}_p}}{d\underline{t } } - \frac{t_{p0 } \epsilon_s}{t_{e0 } \lambda_p \tau_p } \underline{\mu_p } \frac{d\underline{\theta_p}}{d\underline{t } } - \underline{\overline{h}_p}\ ] ] where @xmath496 + \frac{t_{p0 } \epsilon_s^2}{t_{e0 } \lambda_p^2 } \psi^{(2)}_b ( \underline{{\mathbf{r}}_p } , \underline{u_p } , \underline{\mu_p } ) + z_p \epsilon_s \langle \underline{\phi } \rangle ( \underline{{\mathbf{r}}_{p\bot}}/\epsilon_s , \underline{r_{p|| } } , \underline{\mu_p}/\lambda_p^2 , \underline{t } ) \nonumber \\ \fl + \frac{z_p \lambda_p \epsilon_s^2}{\lambda_p } \psi^{(2)}_\phi ( \underline{{\mathbf{r}}_{p\bot}}/\epsilon_s , \underline{{\mathbf{r}}_p } , \underline{\mu_p}/\lambda_p^2 , \underline{t } , \lambda_p ) + \frac{z_p \epsilon_s^2}{\lambda_p } \psi^{(2)}_{\phi b } ( \underline{{\mathbf{r}}_{p\bot}}/\epsilon_s , \underline{{\mathbf{r}}_p } , \underline{u_p } , \underline{\mu_p } , \underline{\mu_p}/\lambda_p^2 , \underline{t } , \lambda_p)\end{aligned}\ ] ] and @xmath497 here , obviously , @xmath498 and @xmath499 .
analogously to ( [ eq : phip ] ) we define @xmath500 and @xmath501 the equations of motion are those obtained in ( [ drdt ] ) , ( [ dudt ] ) , ( [ dmudt ] ) and ( [ dthetadt ] ) multiplied by @xmath502 . as a result ,
the vlasov equation for each species is @xmath503 in what follows we work in non - dimensional variables but do not underline them .
our objective is to write poisson s equation ( [ eq : poissonnondim ] ) in terms of @xmath504 , i.e. the solution of equation ( [ eq : vlasovbar ] ) .
since depends on the species through the values of @xmath505 , @xmath506 and @xmath507 and denote it by @xmath508 . ]
@xmath509 , @xmath510 using the obvious identity @xmath511 and the change of variables formula we can write poisson s equation as @xmath512 where @xmath513 is the jacobian matrix of the transformation @xmath508 , calculated in [ app : jacobians ] , and @xmath514 is the projection onto the spatial part of the coordinates @xmath515 . from an abstract viewpoint
this is , perhaps , the simplest way of writing ( [ eq : poissonnondim ] ) in terms of @xmath516 .
observe that ( [ eq : poissonpullback2 ] ) is an exact relation .
however , in practical terms , we have only computed the explicit expression of @xmath508 up to a certain order in @xmath505 , or equivalently , in the species - independent parameter @xmath96 .
making use of the results of this paper we can give an explicit expression for the change of variables up to order @xmath517 , namely @xmath518 + o(\epsilon_s^3),\end{aligned}\ ] ] which allows us to write the identity @xmath519 where @xmath520 denotes the gradient of the dirac delta function with respect to its natural arguments and the subscript @xmath473 in the corrections @xmath521 , @xmath522 and @xmath523 indicates that these corrections depend on the species through @xmath506 , @xmath507 and @xmath524 in ( [ eq : phip ] ) . substituting ( [ eq : transdelta ] ) into ( [ eq : poissonpullback2 ] ) finally gives @xmath525 + \dots\end{aligned}\ ] ] here we have used [ app : jacobians ] to write @xmath526 .
we employ the subscripts @xmath473 in the determinant of the jacobian @xmath527 because it depends on the species through its dependence on @xmath528 . expression ( [ eq : poissonpullbackfinal ] ) may seem accurate to order @xmath517 , i.e. , that on the right side of ( [ eq : poissonpullbackfinal ] ) we are dropping only terms which are of order @xmath529 or higher .
however , it is easy to see that this is not true . using @xmath530 @xmath531\ ] ] and @xmath532,\ ] ] and integrating by parts so that the delta function does not appear differentiated
, we find that ( [ eq : poissonpullbackfinal ] ) becomes @xmath533 \bigg \ } + o(\epsilon_s^2).\end{aligned}\ ] ] note that terms that seemed to be of second order in @xmath96 are in reality first order contributions .
similarly , terms that seem to be of third order and are neglected in ( [ eq : poissonpullbackfinal ] ) are in reality of second order . to obtain a gyrokinetic poisson s equation
correct to order @xmath517 it is necessary to carry ( [ eq : r_asfunctionofgyrokincoor ] ) to an order higher @xmath534 all the terms entering this equation are computable from the results found in section [ sect_theory ] .
we leave this for future work . by integrating over the delta function in ( [ eq : gkpoisson_v1 ] )
we make the gyrokinetic poisson s equation more explicit , @xmath535 \bigg \}_{{\mathbf{r}}_p = \widehat{{\mathbf{r } } } ( { \mathbf{r } } , \mu_p , \theta_p , \epsilon_s/\lambda_p ) } + o(\epsilon_s^2),\end{aligned}\ ] ] where @xmath536 is the the function defined by solving for @xmath469 the equation @xmath537 .
finally , we would like to point out that usually equation ( [ eq : gkpoisson ] ) can be simplified even more .
the following discussion is not meant to be an exhaustive review of all possible orderings and simplifications , but a brief comment on the most typical approach .
usually @xmath538 has contributions with different characteristic scales of variation , ranging from the shortest scale @xmath539 to the background profile variation scale @xmath6 . at the same time
@xmath540 because the saturation amplitude of the short scale fluctuation @xmath541 is sufficiently small , i.e. , @xmath542 , where @xmath543 is a function of order unity with derivatives with respect to its arguments of order unity . for @xmath540
, we can use the approximate expression @xmath544 because @xmath545 .
in addition , the term containing @xmath546 in ( [ eq : gkpoisson ] ) becomes of next order because @xmath547 . to simplify even more , we use @xmath548 . as a result , @xmath549 \bigg \}_{{\mathbf{r}}_p = { \mathbf{r}}- ( \epsilon_s/\lambda_p ) { \mbox{\scriptsize\boldmath$\rho$}}({\mathbf{r } } , \mu_p , \theta_p ) } + o(\epsilon_s^2).\end{aligned}\ ] ] this is the simplest poisson s equation correct to first order in @xmath96 . as we have already pointed out , it is necessary to include the corrections @xmath550 , @xmath551 and @xmath552 to find the contributions of order @xmath517 .
we finish by checking that the assumption @xmath540 is consistent with the gyrokinetic system of equations .
consider the short scale pieces of the vlasov equation ( [ eq : vlasovbar ] ) and of the gyrokinetic poisson s equation ( [ eq : gkpoissonsimple ] ) .
the result will be a typical @xmath34 formulation @xcite , demonstrating that it is possible to find a closed non - linear system of equations to determine the short scale fluctuations @xmath543 and @xmath24 .
this does not mean that @xmath34 formulations are always valid , but it indicates that the assumption @xmath540 is consistent with the gyrokinetic system of equations .
gyrokinetic field theory is the formulation of gyrokinetics as a classical field theory and is defined by an action functional @xmath553 that depends on the trajectories of the particles and the electromagnetic field .
a closed system of equations coupling the electromagnetic field and the gyrokinetic distribution function is obtained by finding the stationary points of @xmath553 .
the first application of field theory to plasma physics is the work by low in @xcite that was later extended to gyrokinetic theory in the seminal papers by sugama @xcite and brizard @xcite .
the development of gyrokinetic field theory in the last decade has been motivated mostly by the fact that it allows to identify in a systematic way conservation laws from symmetries of the lagrangian . for the sake of completeness we briefly recast now our results in field theory language following the lagrangian formulation of sugama @xcite . in subsection [ sub : variational ]
we show how to obtain the gyrokinetic equations of motion and the gyrokinetic poisson s equation from a variational principle .
the advantage of this procedure is that it ensures that an energy - like invariant is conserved .
we obtain the invariant and prove that it is conserved in subsection [ sub : noethers ] . using the abbreviated notation @xmath554 ,
the action functional is @xmath555 = \frac{\lambda_{de}^2\epsilon_s^2}{2l^2 } \int_{t_0}^{t_1 } dt \int d^3 r \,|\nabla\varphi({\mathbf{r}},t)|^2 \nonumber \\ \fl & + \sum_p \int_{t_0}^{t_1 } dt \int d^6 z_{p0 } \ , b_{|| , p}^ * ( { \mathbf{z}}_{p0 } ) f_{p0}({\mathbf{z}}_{p0 } ) \ , \overline{\mathcal{l}}_p \big({\mathbf{z}}_p({\mathbf{z}}_{p0},t_0;t ) , \dot{\mathbf{z}}_p({\mathbf{z}}_{p0},t_0;t ) , \varphi(\bullet , t)\big),\end{aligned}\ ] ] where @xmath556 is the distribution function of species @xmath466 at time @xmath557 , and @xmath558 is the trajectory in phase - space of a particle of species @xmath466 satisfying the initial condition @xmath559 at @xmath560 .
the first term on the right side of ( [ eq : actionfieldtheory ] ) is the action for the electric field ( recall that we are considering a static magnetic field ) and the second term is the sum of the actions of the particles .
the gyrokinetic lagrangian @xmath561 of the species @xmath466 is defined in ( [ eq : lnewnorm ] ) ( recall that we have dropped the underlining for normalized variables ) .
its third argument stresses that @xmath562 depends as a functional on the electrostatic potential .
the euler - lagrange equations for @xmath553 are obtained by finding its stationary points under infinitesimal variations of the maps @xmath558 and @xmath563 .
the allowed infinitesimal perturbations to @xmath563 vanish at the boundary of the spatial domain of interest , and the perturbations to both @xmath558 and @xmath563 must be zero at @xmath560 and @xmath564 .
the calculation of the variation with respect to @xmath558 gives the gyrokinetic equations of motion and is a repetition of that leading to equations ( [ drdt ] ) , ( [ dudt ] ) , ( [ dmudt ] ) and ( [ dthetadt ] ) ( recall the discussion on the phase - space lagrangian methodology in subsection [ sub_change ] ) .
the distribution function at time @xmath43 is @xmath565 where @xmath566 is the inverse of the map @xmath567 , i.e. , @xmath568 . note that @xmath538 automatically satisfies ( [ eq : vlasovbar ] ) .
observe also that from condition ( [ consps ] ) , which is automatically satisfied by ( [ drdt ] ) , ( [ dudt ] ) , ( [ dmudt ] ) and ( [ dthetadt ] ) , we obtain @xmath569 where @xmath570 is the jacobian matrix of the map @xmath567 , and @xmath571 is its determinant .
this property is proven in [ app : psconserv ] for completeness .
the stationary points of @xmath553 under variations of @xmath563 are given by @xmath572 , with @xmath573 in [ app : poissonvariationalprinciple ] we have evaluated all the terms . the final result is @xmath574 where @xmath575 \bigg \}.\end{aligned}\ ] ] by imposing that @xmath576 for any @xmath577 , we find that @xmath578 .
this is poisson s equation . by integrating by parts to leave the delta function undifferentiated , and then integrating over the delta function , we recover ( [ eq : gkpoisson ] ) .
note that even though the hamiltonian is obtained to second order in @xmath96 , and hence the vlasov equation is also known to second order , the gyrokinetic poisson s equation that we have found using the variational principle is only correct to first order in @xmath96 .
it coincides to first order with the first order equation that we found independently in subsection [ sec : gyropoissonchangevar ] .
observe that equation ( [ eq : gyrokinpoissonfieldtheory ] ) keeps some second order terms that could have been neglected , as done in ( [ eq : gkpoissonsimple ] ) .
these terms are important if we want to take advantage of the field theory formulation of gyrokinetics .
the action ( [ eq : actionfieldtheory ] ) is invariant under time translations , so noether s theorem automatically provides a conserved quantity , interpreted as the total energy of the system , which we denote by @xmath579 because it is indeed the field theory hamiltonian .
however , @xmath580 _ on the equations of motion of _ @xmath553 , that is , on solutions of ( [ drdt ] ) , ( [ dudt ] ) , ( [ dmudt ] ) , ( [ dthetadt ] ) and ( [ eq : gyrokinpoissonfieldtheory ] ) , without neglecting any terms .
we show this in the next subsection . in this subsection we prove that if the equations of motion are obtained via a variational principle of the action ( [ eq : actionfieldtheory ] ) , there is an energy - like invariant @xmath581 .
this is an application of noether s theorem , and it is based on the fact that in the action ( [ eq : actionfieldtheory ] ) the only time dependence is through the functions @xmath582 and @xmath3 . to obtain the energy - like invariant , we use the equations of motion derived from the variational principle in subsection [ sub : variational ] . in order to find the conserved quantity we first perform the change of variable @xmath583 in every integral on the right side of ( [ eq : actionfieldtheory ] ) , which , of course , does not change the value of the action , giving @xmath584 = \int_{t_0 - \delta t}^{t_1 - \delta t } dt ' \bigg [ \frac{\lambda_{de}^2\epsilon_s^2}{2l^2 } \int d^3 r \,|\nabla\varphi({\mathbf{r}},t ' + \delta t)|^2 \nonumber \\
+ \sum_p \int d^6 z_{p0 } b_{|| , p}^ * ( { \mathbf{z}}_{p0 } ) f_{p0}({\mathbf{z}}_{p0 } ) \left .
\overline{\mathcal{l}}_p \right |_{t = t^\prime + \delta t } \bigg ] , \end{aligned}\ ] ] where @xmath585 we expand the right side of the previous equation up to first order in @xmath586 to find @xmath587 = \sigma \big [ { \mathbf{z}}_p(\bullet , \cdot ; \cdot ) , \varphi(\bullet , \cdot ) \big ] \nonumber\\ \fl & + \delta_\varphi \left . \sigma\big [ { \mathbf{z}}_p(\bullet , \cdot;\cdot),\varphi(\bullet,\cdot ) \big ] \right \sum_p \delta_{{\mathbf{z}}_p } \left . \sigma\big [ { \mathbf{z}}_p(\bullet , \cdot;\cdot),\varphi(\bullet,\cdot ) \big ] \right |_{\delta { \mathbf{z}}_p = \delta t \dot { \mathbf{z}}_p } \nonumber\\ \fl & - \delta t\bigg [ \frac{\lambda_{de}^2\epsilon_s^2}{2l^2 } \int d^3 r \,|\nabla\varphi({\mathbf{r}},t')|^2 + \sum_p \int d^6 z_{p0 } \ , b_{|| , p}^ * ( { \mathbf{z}}_{p0 } ) f_{p0}({\mathbf{z}}_{p0 } ) \left .
\overline{\mathcal{l}}_p \right |_{t = t ' } \bigg]_{t'=t_0}^{t'=t_1},\end{aligned}\ ] ] where @xmath588 and latexmath:[$\delta_{{\mathbf{z}}_p } \sigma variations of @xmath553 under perturbations of both @xmath590 and @xmath582 with the specific form @xmath591 and @xmath592 . for the variations with respect to @xmath593
, we obtain @xmath594_{t^\prime = t_0}^{t^\prime = t_1},\end{aligned}\ ] ] where to obtain this last equality we have integrated by parts in @xmath43 and we have used the equations of motion @xmath595 applying @xmath596 we find @xmath597_{t^\prime = t_0}^{t^\prime = t_1}.\ ] ] for the variations with respect to @xmath598 we obtain that @xmath599 where we have employed ( [ eq : gyrokinpoissonfieldtheory1 ] ) and poisson s equation @xmath578 . to obtain ( [ eq : gyrokinpoissonfieldtheory1 ] ) we assumed that @xmath600 vanishes at the boundaries of the domain , and this is not necessarily the case for @xmath601 .
we avoid this problem by assuming either that the domain extends to infinity , where @xmath602 , or that we are in a periodic box and the contribution from one half of the boundary cancels with the contribution from the other half . in either case
, we are assuming that there is no net energy flux through the boundary . by substituting ( [ varzpnoethers ] ) and ( [ varvarphinoethers ] ) into ( [ eq : actiontimetrans2 ] ) and observing that the term linear in @xmath586 in ( [ eq : actiontimetrans2 ] ) has to be identically zero ,
that is , the sum of the second , third and fourth terms in the right side must vanish , we find @xmath603_{t'=t_0}^{t'=t_1 } = 0.\end{aligned}\ ] ] then , @xmath604_{t'=t_0}^{t'=t_1 } = 0.\ ] ] for solutions of the equations of motion , where @xmath605 since this must hold for every interval @xmath606 $ ] , we deduce that @xmath607 on the equations of motion . finally , using equations ( [ fpfp0 ] ) and ( [ eq : psconserv ] ) , we can write the integrals over @xmath608 in ( [ eq : actiontimetrans6 ] ) as integrals over @xmath593 , giving @xmath609 the conservation of this energy - like invariant is only satisfied on the equations of motion , that is , the variational equations of motion and the variational poisson s equation obtained in subsection [ sub : variational ] must be used .
it is important to keep all the given terms , even if they are higher order than first ( the order to which poisson s equation is correct ) .
for example , it is necessary to obtain the _ exact _ function @xmath610 in ( [ eq : gkpoisson ] ) to have exact conservation of @xmath611 .
the gyrokinetic ordering in a static magnetic field consists of the ordering assumptions ( [ orderings ] ) , defined by a single parameter @xmath612 . in this paper
we have strictly implemented the gyrokinetic ordering in the phase - space lagrangian to obtain explicitly the gyrokinetic lagrangian to order @xmath1 for general magnetic geometry .
our approach differs from previous phase - space lagrangian ( or hamiltonian ) derivations of gyrokinetics . in previous work
@xcite the calculation is performed in two steps .
first , with zero fluctuating electrostatic potential , an expansion in powers of @xmath613 is performed and a gyrophase independent guiding - center lagrangian is determined to order @xmath0 .
then , the electrostatic fluctuations , whose size is given by a new expansion parameter @xmath614 , are switched - on , reintroducing a gyrophase dependence that is removed order by order in @xmath27 yielding the final gyrokinetic lagrangian , usually computed up to order @xmath29 .
when the expansion in @xmath27 is performed , the fact that there has been a previous expansion in @xmath0 is ignored and the terms of order @xmath28 are never calculated .
thus , the final lagrangian is missing relevant terms of order @xmath28 and @xmath1
. the novelty of our work can be easily understood by examining the explicit expression of the second - order gyrokinetic hamiltonian , @xmath399 , given in equations ( [ hamiltonian_o2_2 ] ) , ( [ psi2_phi ] ) , ( [ psi2_phib ] ) and ( [ psi2_b ] ) .
it shows in a transparent way that gyrokinetic theory ties together geometry and turbulence , so that no splitting between guiding - center and gyrokinetic dynamics is possible . see , for example , @xmath410 in ( [ psi2_phib ] ) where magnetic geometry and electrostatic potential appear together .
this is the first time that the electrostatic gyrokinetic equations in general geometry are fully computed to order @xmath1 and the calculations are pursued to the point of reaching formulae , like the one for @xmath399 , that can be straightforwardly implemented in a computer code . from the new phase - space lagrangian in ( [ finall ] ) and the new hamiltonian in ( [ finalh ] )
we obtain a new vlasov equation and a new gyrokinetic poisson s equation .
the vlasov equation is correct to second order in the expansion parameter @xmath0 , and it is to our knowledge the highest order full @xmath23 gyrokinetic equation available in the literature for general geometry . in the limit where the electrostatic potential has a scale of variation much larger than the gyroradius of the species of interest , this equation is also the highest order drift kinetic equation that we are aware of .
the gyrokinetic poisson s equation derived from the new lagrangian is , however , only correct to first order in @xmath0 .
we have calculated poisson s equation employing two methods . in section [ sec : gyropoissoneq ] , in the integrals in velocity space that enter in poisson s equation we have simply changed from the coordinates @xmath44 to the gyrokinetic variables . in section [ sec : fieldtheory ] we have used a variational formalism . the equation obtained with the variational formalism
is set by the form of the hamiltonian and it is only correct to first order , although it contains terms that are higher order .
these higher order terms do not add accuracy , but they are necessary to have an exact energy - like invariant . by directly changing from @xmath615 to the gyrokinetic variables in the integrals that enter poisson s equation , it is possible to obtain a higher order poisson s equation if the higher order corrections @xmath337 , @xmath339 and @xmath340 are calculated .
it is not necessary to calculate the hamiltonian to next order .
the calculation of @xmath337 , @xmath339 and @xmath340 is however very tedious and is left for future work .
if this procedure is followed and the third order hamiltonian is not obtained , the conservation of the energy - like invariant will not be exact .
some natural steps following the present work are the extension to the electromagnetic case and the introduction of external flows .
we will also investigate the implications of the new terms in the hamiltonian for the transport of toroidal angular momentum in tokamaks .
the authors are indebted to the programme of visits to tj - ii at laboratorio nacional de fusin of ciemat ( spain ) and to the summer programme gyrokinetics in laboratory and astrophysical plasmas " at the isaac newton institute for mathematical sciences , without which this work would not have been possible .
this research was partially supported by the post - doctoral programme of the engineering and physical sciences research council of the uk , by the junior research fellowship programme of christ church at university of oxford , and by grant ene2009 - 07247 , ministerio de ciencia e innovacin ( spain ) .
let @xmath616 be an invertible , antisymmetric matrix of dimension @xmath617 whose components are functions defined on a region @xmath618 , and such that @xmath619 denote by @xmath620 the inverse of @xmath616 , @xmath621 .
then , @xmath622 defines a poisson bracket by contraction with the differentials of pairs of functions on @xmath623 , i.e. , @xmath624 we must prove that ( [ poi_braappendix ] ) satisfies skew - symmetry , given in ( [ eq : poissonproperty1 ] ) , the leibniz rule , given in ( [ eq : poissonproperty2 ] ) , and the jacobi identity , given in ( [ eq : poissonproperty3 ] ) . skew - symmetry is satisfied because the inverse of an antisymmetric matrix is antisymmetric , i.e. , @xmath625 .
the leibniz rule is trivial to check . as for the jacobi identity ,
it is immediate to see that it is equivalent to @xmath626 showing that ( [ eq : jacobicomponents ] ) is equivalent to ( [ eq : closedform ] ) is a simple exercise of application of the formula of the derivative of the inverse of a matrix , @xmath627
in this appendix we prove equations ( [ ds_npdr ] ) and ( [ ds_npdtheta ] ) . for equation ( [ ds_npdr ] ) , we use that according to ( [ s_np ] ) @xmath628.\end{aligned}\ ] ] recall that the prime @xmath629 here indicates that the function depends on @xmath630 . employing @xmath631 \nonumber \\ = \frac{1}{\epsilon } \frac{\partial}{\partial \mu_g^\prime } [ { \mathbf{a}}({\mathbf{r}}_g + \epsilon { \mbox{\boldmath$\rho$}}^\prime ) ] + \frac{1}{2 \mu_g^\prime } { \mbox{\boldmath$\rho$}}^\prime \times { \mathbf{b}}({\mathbf{r}}_g + \epsilon { \mbox{\boldmath$\rho$}}^\prime ) , \end{aligned}\ ] ] @xmath632 + \frac{\epsilon}{2\mu_g^\prime } \nabla_{{\mathbf{r}}_g } { \mbox{\boldmath$\rho$}}^\prime \cdot [ { \mbox{\boldmath$\rho$}}^\prime \times { \mathbf{b}}({\mathbf{r}}_g + \epsilon { \mbox{\boldmath$\rho$}}^\prime ) ] \nonumber \\
= \nabla_{{\mathbf{r}}_g } { \mbox{\boldmath$\rho$}}^\prime \cdot \frac{\partial}{\partial \mu_g^\prime } [ { \mathbf{a}}({\mathbf{r}}_g + \epsilon { \mbox{\boldmath$\rho$}}^\prime ) ] + \frac{\epsilon}{2\mu_g^\prime } \nabla_{{\mathbf{r}}_g } { \mbox{\boldmath$\rho$}}^\prime \cdot [ { \mbox{\boldmath$\rho$}}^\prime \times { \mathbf{b}}({\mathbf{r}}_g + \epsilon { \mbox{\boldmath$\rho$}}^\prime ) ] \end{aligned}\ ] ] and @xmath633 equation ( [ dsdr_app1 ] ) becomes @xmath634.\end{aligned}\ ] ] to obtain equation ( [ ds_npdr ] ) from this equation we use ( [ drhodmu ] ) to write @xmath635 and we employ ( [ gradrho ] ) to get @xmath636\nabla_{{\mathbf{r}}_g } { \hat{\mathbf{b}}}_g \cdot { \mbox{\boldmath$\rho$}}^\prime \nonumber \\ + \frac{2\mu_g^\prime}{b_g } [ { \hat{\mathbf{b}}}_g \cdot { \mathbf{b}}({\mathbf{r}}_g + \epsilon { \mbox{\boldmath$\rho$}}^\prime ) ] \nabla_{{\mathbf{r}}_g } { \hat{\mathbf{e}}}_{2 g } \cdot { \hat{\mathbf{e}}}_{1g}.\end{aligned}\ ] ] for equation ( [ ds_npdtheta ] ) , we use that according to ( [ s_np ] ) @xmath637.\end{aligned}\ ] ] employing @xmath638 + \frac{\epsilon}{2\mu_g^\prime } \frac{\partial { \mbox{\boldmath$\rho$}}^\prime}{\partial \theta_g } \cdot [ { \mbox{\boldmath$\rho$}}^\prime \times { \mathbf{b}}({\mathbf{r}}_g + \epsilon { \mbox{\boldmath$\rho$}}^\prime ) ] \nonumber
\\ = \frac{\partial { \mbox{\boldmath$\rho$}}^\prime}{\partial \theta_g } \cdot \frac{\partial}{\partial \mu_g^\prime } [ { \mathbf{a}}({\mathbf{r}}_g + \epsilon { \mbox{\boldmath$\rho$}}^\prime ) ] + \frac{\epsilon}{2\mu_g^\prime } \frac{\partial { \mbox{\boldmath$\rho$}}^\prime}{\partial \theta_g } \cdot [ { \mbox{\boldmath$\rho$}}^\prime \times { \mathbf{b}}({\mathbf{r}}_g + \epsilon { \mbox{\boldmath$\rho$}}^\prime ) ] \end{aligned}\ ] ] and @xmath639 equation ( [ dsdtheta_app1 ] ) becomes equation ( [ ds_npdtheta ] ) . to obtain the final form of the equation we have also used ( [ drhodtheta ] ) to write @xmath640 = - \frac{2\mu_g^\prime}{b_g } { \hat{\mathbf{b}}}_g \cdot { \mathbf{b}}({\mathbf{r}}_g + \epsilon { \mbox{\boldmath$\rho$}}^\prime).\end{aligned}\ ] ]
in this appendix we compare the gyrokinetic variable transformation obtained in this article with the variables found in @xcite . to be able to compare with the results in @xcite , given in the form @xmath641 , @xmath642 , @xmath643 and @xmath644 , we will use the transformation @xmath645 to write them as @xmath646 , @xmath647 , @xmath648 and @xmath649 .
after doing so , we will see that the gyrokinetic variables @xmath650 , @xmath651 and @xmath652 in @xcite are gyrophase independent quantities up to the order that they are defined . this is a property that must be satisfied because otherwise the new gyrokinetic variables would have fast time dependence through the gyrophase .
the variable @xmath652 must be a function of @xmath101 only because there is only one adiabatic invariant associated with the gyromotion .
the gyrokinetic variables of @xcite are the gyrocenter position @xmath653 { \raise 0.65ex\hbox{${\scriptstyle { { _ { \displaystyle \cdot}}\atop\times}}$ } } \nabla \left ( \frac { { \hat{\mathbf{b } } } } { b } \right ) \nonumber \\ \fl & + \frac{\epsilon^2 v_{||}}{b^2 } { \mathbf{v}}_\bot \cdot \nabla { \hat{\mathbf{b}}}+ \frac{\epsilon^2 v_{||}}{b^2 } { \hat{\mathbf{b}}}{\hat{\mathbf{b}}}\cdot \nabla { \hat{\mathbf{b}}}\cdot { \mathbf{v}}_\bot + \frac{\epsilon^2}{8b^2 } { \hat{\mathbf{b } } } [ { \mathbf{v}}_\bot { \mathbf{v}}_\bot - ( { \mathbf{v}}\times { \hat{\mathbf{b } } } ) ( { \mathbf{v}}\times { \hat{\mathbf{b } } } ) ] : \nabla { \hat{\mathbf{b}}}\nonumber\\ \fl & - \frac{\lambda \epsilon^2}{\lambda b^2 } \nabla_{({\mathbf{r}}_\bot/\lambda \epsilon ) } { \widetilde{\phi}}\times { \hat{\mathbf{b}}}+ o(\epsilon^3),\end{aligned}\ ] ] where @xmath654 ; the gyrokinetic kinetic energy @xmath655 the magnetic moment @xmath656 : \nabla { \hat{\mathbf{b}}}+ \frac{\lambda \epsilon}{b } { \widetilde{\phi}}+ o(\epsilon^2);\end{aligned}\ ] ] and the gyrophase @xmath657 : \nabla { \hat{\mathbf{b}}}- \frac{\lambda \epsilon}{\lambda^2 b } \frac{\partial { \widetilde{\phi}}}{\partial ( \mu/\lambda^2 ) } + o(\epsilon^2).\end{aligned}\ ] ] to these variables we apply the gyrokinetic transformation .
first we apply the non - perturbative change of variables @xmath658 . for the variable @xmath650
we find @xmath659 { \raise 0.65ex\hbox{${\scriptstyle { { _ { \displaystyle \cdot}}\atop\times}}$ } } \nabla_{{\mathbf{r}}_g } \left ( \frac { { \hat{\mathbf{b}}}_g } { b_g } \right ) \nonumber \\ \fl & + \frac{\epsilon^2 v_{||g}}{b_g } { \hat{\mathbf{b}}}{\hat{\mathbf{b}}}\cdot \nabla_{{\mathbf{r}}_g } { \hat{\mathbf{b}}}\cdot ( { \mbox{\boldmath$\rho$}}\times { \hat{\mathbf{b}}}_g ) - \frac{\epsilon^2}{8 } { \hat{\mathbf{b}}}_g [ { \mbox{\boldmath$\rho$}}{\mbox{\boldmath$\rho$}}- ( { \mbox{\boldmath$\rho$}}\times { \hat{\mathbf{b}}}_g ) ( { \mbox{\boldmath$\rho$}}\times { \hat{\mathbf{b}}}_g ) ] : \nabla_{{\mathbf{r}}_g } { \hat{\mathbf{b}}}_g \nonumber\\ \fl & - \frac{\lambda \epsilon^2}{\lambda b_g^2 } \nabla_{({\mathbf{r}}_\bot/\lambda \epsilon ) } { \widetilde{\phi}}\times { \hat{\mathbf{b}}}_g + o(\epsilon^3),\end{aligned}\ ] ] where we have used @xmath660 , @xmath661 , @xmath662 , @xmath663 , @xmath664 and @xmath665
we then apply the transformation @xmath666 by using @xmath667 , with @xmath333 given in ( [ r2 ] ) , @xmath668 , @xmath669 and @xmath670 . as a result
we find the expression @xmath671 : \nabla_{\mathbf{r}}{\hat{\mathbf{b}}}+ \frac{\epsilon^2}{4b } [ { \mbox{\boldmath$\rho$}}{\mbox{\boldmath$\rho$}}+ ( { \mbox{\boldmath$\rho$}}\times { \hat{\mathbf{b } } } ) ( { \mbox{\boldmath$\rho$}}\times { \hat{\mathbf{b } } } ) ] \cdot \nabla_{\mathbf{r}}b \nonumber \\ \fl & + \frac{\epsilon^2 u}{b } ( { \mbox{\boldmath$\rho$}}\times { \hat{\mathbf{b } } } ) \cdot \nabla_{\mathbf{r}}{\hat{\mathbf{b}}}+ \epsilon^2 u { \hat{\mathbf{b}}}\cdot \nabla_{\mathbf{r}}\left ( \frac{{\hat{\mathbf{b}}}}{b } \right ) \times { \mbox{\boldmath$\rho$}}- \frac{\epsilon^2 u}{b } { \hat{\mathbf{b}}}{\hat{\mathbf{b}}}\cdot \nabla_{\mathbf{r}}{\hat{\mathbf{b}}}\cdot ( { \mbox{\boldmath$\rho$}}\times { \hat{\mathbf{b } } } ) \nonumber \\ \fl & - \frac{\epsilon^2 u}{b } { \hat{\mathbf{b}}}\times \nabla_{\mathbf{r}}{\hat{\mathbf{b}}}\cdot { \mbox{\boldmath$\rho$}}+ o(\epsilon^3),\end{aligned}\ ] ] where @xmath672 and @xmath673 depend on @xmath99 .
to simplify expression ( [ rp_2 ] ) we use @xmath674 and @xmath675 \times { \mbox{\boldmath$\rho$}}= ( \nabla_{\mathbf{r}}\cdot { \hat{\mathbf{b } } } ) ( { \mbox{\boldmath$\rho$}}\times { \hat{\mathbf{b}}}).\ ] ] we finally obtain @xmath676 to find this result we have used @xmath677 .
applying the non - perturbative change of variables to @xmath651 we find @xmath678 to transform to the variables @xmath99 , @xmath100 , @xmath101 and @xmath98 we use @xmath679 , @xmath680 , with @xmath334 and @xmath335 given in ( [ u1 ] ) and ( [ mu1 ] ) , @xmath681 and @xmath682 , giving @xmath683 for @xmath652 , we find that the transformation @xmath684 gives @xmath685 : \nabla_{{\mathbf{r}}_g } { \hat{\mathbf{b}}}_g + \frac{\lambda \epsilon}{b_g } { \widetilde{\phi}}+ o(\epsilon^2),\end{aligned}\ ] ] where we have used @xmath663 , @xmath686 , @xmath687 - \frac{\epsilon \mu_g}{b_g } { \mbox{\boldmath$\rho$}}\cdot \nabla_{{\mathbf{r}}_g } b_g + o(\epsilon^2)\end{aligned}\ ] ] and @xmath688 ^ 2 + o(\epsilon^2 ) \nonumber \\
= 2\mu_g b_g - 2 \epsilon v_{||g } { \mbox{\boldmath$\rho$}}\cdot \nabla_{{\mathbf{r}}_g } { \hat{\mathbf{b}}}_g \cdot ( { \mbox{\boldmath$\rho$}}\times { \mathbf{b}}_g ) + o(\epsilon^2).\end{aligned}\ ] ] then employing @xmath689 , with @xmath335 given in ( [ mu1 ] ) , @xmath681 , @xmath668 and @xmath670 , we find @xmath690 : \nabla_{{\mathbf{r}}_g } { \hat{\mathbf{b}}}_g + o(\epsilon^2 ) \nonumber \\ = \mu + o(\epsilon^2),\end{aligned}\ ] ] where to obtain the final equality we have used ( [ mati ] ) .
finally , for @xmath691 we obtain that the non - perturbative transformation gives @xmath692 : \nabla_{{\mathbf{r}}_g } { \hat{\mathbf{b}}}_g - \frac{\lambda \epsilon}{\lambda^2 b_g } \frac{\partial { \widetilde{\phi}}}{\partial ( \mu/\lambda^2 ) } + o(\epsilon^2),\end{aligned}\ ] ] where we have used @xmath693 \nonumber
\\ = \theta_g + \frac{\epsilon}{2\mu_g b_g } [ ( { \mbox{\boldmath$\rho$}}\times { \mathbf{b}}_g ) \cdot { \hat{\mathbf{e}}}_{1 g } ] { \mbox{\boldmath$\rho$}}\cdot \nabla_{{\mathbf{r}}_g } { \hat{\mathbf{e}}}_{2 g } \cdot ( v_{||g } { \hat{\mathbf{b}}}_g + { \mbox{\boldmath$\rho$}}\times { \mathbf{b}}_g ) \nonumber \\ - \frac{\epsilon}{2\mu_g b_g } [ ( { \mbox{\boldmath$\rho$}}\times { \mathbf{b}}_g ) \cdot { \hat{\mathbf{e}}}_{2 g } ] { \mbox{\boldmath$\rho$}}\cdot \nabla_{{\mathbf{r}}_g } { \hat{\mathbf{e}}}_{1 g } \cdot ( v_{||g } { \hat{\mathbf{b}}}_g + { \mbox{\boldmath$\rho$}}\times { \mathbf{b}}_g ) + o(\epsilon^2 ) \nonumber
\\ = \theta_g - \frac{\epsilon v_{||g}}{2\mu_g } { \mbox{\boldmath$\rho$}}\cdot \nabla_{{\mathbf{r}}_g } { \hat{\mathbf{b}}}_g \cdot { \mbox{\boldmath$\rho$}}+ \epsilon { \mbox{\boldmath$\rho$}}\cdot \nabla_{{\mathbf{r}}_g } { \hat{\mathbf{e}}}_{2 g } \cdot { \hat{\mathbf{e}}}_{1 g } + o(\epsilon^2).\end{aligned}\ ] ] here we have employed @xmath694 and @xmath695 to write the last equality . equation ( [ thetap_1 ] ) is now rewritten by using @xmath696 , with @xmath336 given in ( [ theta1 ] ) , @xmath681 , @xmath668 and @xmath697 , giving @xmath698 : \nabla_{\mathbf{r}}{\hat{\mathbf{b}}}+ o(\epsilon^2 ) = \theta + \frac{\epsilon u}{2b^2 } { \hat{\mathbf{b}}}\cdot \nabla_{\mathbf{r}}b + o(\epsilon^2),\end{aligned}\ ] ] where we have employed ( [ mati ] ) and @xmath699 . from equation ( [ rp_final ] ) , ( [ ep_final ] ) , ( [ mup_final ] ) and ( [ thetap_final ] ) we see that the gyrokinetic variables defined in @xcite are simple functions of the gyrokinetic variables in this article .
notice that @xmath650 and @xmath651 do not depend on the gyrophase and that @xmath652 is equal to @xmath101 to the order of interest , as expected .
in this appendix we show how to obtain ( [ hamiltonian_o2_2 ] ) from equation ( [ hbar2_1 ] ) .
first , by employing equations ( [ gammar_o2 ] ) , ( [ gammatheta_o2 ] ) , ( [ s2 ] ) , ( [ r2 ] ) and ( [ nablas2_slow ] ) , we find that @xmath700 , @xmath701 , @xmath702 , @xmath703 and @xmath704 here we have used equation ( [ rhorhoave ] ) repeatedly .
these results give that equation ( [ hbar2_1 ] ) is equal to @xmath705 where we have used @xmath706 to write @xmath707 = \frac{\mu}{2b^2 } [ ( \nabla_{\mathbf{r}}\times { \hat{\mathbf{b } } } ) \times { \hat{\mathbf{b } } } ] \cdot \nabla_{\mathbf{r}}b = \frac{\mu}{2b^2 } { \hat{\mathbf{b}}}\cdot \nabla_{\mathbf{r}}{\hat{\mathbf{b}}}\cdot \nabla_{\mathbf{r}}b.\end{aligned}\ ] ] the terms @xmath708 , @xmath709 , @xmath710 and @xmath711 in equation ( [ lagrangian_gyro_app1 ] ) are @xmath712 @xmath713 @xmath714 and @xmath715 we proceed to write these terms in more convenient forms .
the term @xmath708 in ( [ t1app2_1 ] ) is rewritten using @xmath716 to obtain @xmath717 employing ( [ r2_1 ] ) , this last result becomes @xmath718 \right \ } \right \rangle.\ ] ] realizing that @xmath719 - \nabla_{({\mathbf{r}}_\bot/\epsilon ) } s_p^{(2 ) } \cdot \left [ { \hat{\mathbf{b}}}\times \frac{\partial { \mbox{\boldmath$\gamma$}}_{\mathbf{r}}^{(1)}}{\partial \theta } \right ] \nonumber
\\ = \frac{\partial}{\partial \theta } \left [ { \mbox{\boldmath$\gamma$}}_{\mathbf{r}}^{(1 ) } \cdot \left ( { \hat{\mathbf{b}}}\times \nabla_{({\mathbf{r}}_\bot/\epsilon ) } s_p^{(2 ) } \right ) \right ] , \end{aligned}\ ] ] we finally obtain @xmath720 \right \rangle .
\label{t1app2_2}\ ] ] these two terms can be expressed in terms of magnetic and electric field gradients . using equation ( [ gammar_o1 ] ) and @xmath721 from ( [ rhorhoave ] )
, we find latexmath:[\ ] ] note that we did not need to keep the gyroaverages in ( [ term2variational ] ) and ( [ term3variational ] ) because there is an overall integral in @xmath472 and neither @xmath828 nor @xmath538 depend on @xmath472 .
we have only left to prove that ( [ poisson_app ] ) is equal to ( [ eq : gyrokinpoissonfieldtheory ] ) . using the result in ( [ r2 ] ) and employing ( [ rhorhoave ] ) and ( [ rhorhowig ] ) to write @xmath829 \cdot \nabla_{{\mathbf{r}}_p } b = { \mbox{\boldmath$\rho$}}{\mbox{\boldmath$\rho$}}\cdot \nabla_{{\mathbf{r}}_p } b,\ ] ] it is clear that @xmath830 is exactly the perpendicular component of @xmath521 . using ( [ mu1 ] ) , we find that @xmath831 .
it only remains to use that @xmath832 , with @xmath833 given in ( [ theta1 ] ) . to obtain this identity we have used @xmath834 , @xmath835/\partial \theta_p$ ] and @xmath836 .
using @xmath837 and integrating by parts in @xmath472 in ( [ poisson_app ] ) gives the final form ( [ eq : gyrokinpoissonfieldtheory ] ) . | gyrokinetic theory is based on an asymptotic expansion in the small parameter @xmath0 , defined as the ratio of the gyroradius and the characteristic length of variation of the magnetic field . in this article , this ordering is strictly implemented to compute the electrostatic gyrokinetic phase - space lagrangian in general magnetic geometry to order @xmath1 . in particular , a new expression for the complete second - order gyrokinetic hamiltonian is provided , showing that in a rigorous treatment of gyrokinetic theory magnetic geometry and turbulence can not be dealt with independently . the new phase - space gyrokinetic lagrangian gives a vlasov equation accurate to order @xmath1 and a poisson equation accurate to order @xmath0 .
the final expressions are explicit and can be implemented into any simulation without further computations . |
this work has been supported by the erc project ultralaser , the ministry of education and science of the russian federation ( agreement no .
14.b25.31.0003 ) , the spanish mineco ( fis2012 - 37655-c02 - 01 ) , eoard ( grant fa9550 - 14 - 1 - 0359 ) , cnpq brazil , russian foundation for basic research ( grant 15 - 02 - 07925 ) , presidential grant for young researchers ( grant 14.120.14.228-mk ) . n.t .
is supported by the russian science foundation ( grant no .
14 - 21 - 00110 ) . | we use advanced statistical tools of time - series analysis to characterize the dynamical complexity of the transition to optical wave turbulence in a fibre laser .
ordinal analysis and the horizontal visibility graph applied to the experimentally measured laser output intensity reveal the presence of temporal correlations during the transition from the laminar to the turbulent lasing regimes .
both methods unveil coherent structures with well defined time - scales and strong correlations both , in the timing of the laser pulses and in their peak intensities .
our approach is generic and may be used in other complex systems that undergo similar transitions involving the generation of extreme fluctuations .
fibre lasers are important practical devices that represent complex nonlinear systems with many degrees of freedom @xcite .
typically , the output of a fibre laser involves nonlinear interactions of millions of longitudinal cavity modes in regimes far from thermal equilibrium @xcite . in general , wave dynamics in fibre lasers is highly complex , as in other optical wave turbulence systems @xcite .
though the underlying physical effect , nonlinear four - wave mixing , is purely deterministic and well understood , it is also well known that a deterministic dynamical description is not fully adequate in this problem , and statistical tools such as entropy and complexity measures should be used to characterize the complex fluctuations in the generated output signals . within this framework of wave turbulence ,
the role of temperature " is played by optical noise that occurs in the gain medium , which in fibre lasers leads to an effective nonlinear noise " due to four - wave - mixing .
recently , the analogy between hydrodynamic transition to turbulence and change of operational regimes in fibre lasers has been studied both experimentally and theoretically @xcite . a transition from highly ordered lasing regime to more irregular lasing , characterized by extreme
, apparently random intensity fluctuations was reported . such transition , being a relevant example of a phase transition in a one dimensional physical system ,
was shown to be accompanied by the occurrence of coherent spatio - temporal structures @xcite . in this work
we address an important question relevant to practical identification of such structures : are there underlying correlations and/or specific time - scales in the easily measurable intensity fluctuations of laser radiation ? in order to investigate this issue we use two nonlinear analysis tools : ordinal analysis @xcite and the horizontal visibility graph @xcite .
we show that both methods allow to clearly identify the presence of long - range temporal correlations in the experimentally measured laser output intensity . in our experiments ,
we measure an output temporal intensity dynamics of a quasi - cw raman fiber laser formed of 1 km of normal dispersion fiber placed between two fiber bragg gratings acting as cavity mirrors @xcite .
state - of - the - art experimental capabilities allowed us to register extremely long time traces with total number of intensity data points of 50 million .
taking into account the discretisation time of 12.5 ps , the intensity dynamics over 625 @xmath0s could be captured . in order to be able to compare among time - series recorded at different pump power ,
each time - series is normalized to have zero - mean and unit variance .
depending on power , the generation regime can be considered as `` laminar '' or `` turbulent '' @xcite .
the transition occurs at pump power 0.9 w ( see @xcite for details ) . despite the radically different coherence properties of radiation in these two regimes , the output intensity ,
@xmath1 , looks similar and irregular at all powers , as seen in fig .
1(a ) , with typical intensity probability distribution function ( pdf ) of intensity values , @xmath2 , shown on fig . 1(b ) . in a time series
are represented with vertical bars , the ordinal patterns formed by @xmath3 are indicated in red , the links in horizontal visibility graph are indicated with arrows , and the numbers indicate the degree ( the number of links ) of each data point ( graph node ) .
( d ) number of intensity peaks vs the pump power for various thresholds ( threshold values are shown by different colours).,title="fig : " ] in a time series are represented with vertical bars , the ordinal patterns formed by @xmath3 are indicated in red , the links in horizontal visibility graph are indicated with arrows , and the numbers indicate the degree ( the number of links ) of each data point ( graph node ) .
( d ) number of intensity peaks vs the pump power for various thresholds ( threshold values are shown by different colours).,title="fig : " ] in a time series are represented with vertical bars , the ordinal patterns formed by @xmath3 are indicated in red , the links in horizontal visibility graph are indicated with arrows , and the numbers indicate the degree ( the number of links ) of each data point ( graph node ) .
( d ) number of intensity peaks vs the pump power for various thresholds ( threshold values are shown by different colours).,title="fig : " ] in a time series are represented with vertical bars , the ordinal patterns formed by @xmath3 are indicated in red , the links in horizontal visibility graph are indicated with arrows , and the numbers indicate the degree ( the number of links ) of each data point ( graph node ) .
( d ) number of intensity peaks vs the pump power for various thresholds ( threshold values are shown by different colours).,title="fig : " ] as a starting step of our analysis , we generate two new data sets from experimentally measured intensity dynamics , @xmath1 . in particular , we filter out from the initial intensity dynamics only local intensity peaks and create a sequence of intensity peak heights , @xmath4 , which are above a certain threshold , also shown in fig . 1(a ) . because each time series is normalized to zero mean and unit variance , the thresholds used are in units of the standard deviation , @xmath5 : with threshold 0 , the peaks considered are only those above the mean value ; with threshold 1 , the peaks are only those above @xmath5 . the number of the peaks found in intensity dynamics measured at different pump power level is shown on fig .
1(d ) depending on the threshold value .
we note that , if the threshold is low , the number of peaks decreases with the pump power , but if the threshold is high , the number of peaks increases with the pump power until 0.9 w , where it is maximum , then , with further increase of the pump power , the number of peaks diminishes .
the transition between two different generation regimes laminar and turbulent regime could be easily detected at 0.9 w. in the following we fix the threshold equal to 2 , i.e. , we analyze only the intensity peaks that are above @xmath6 .
as each local intensity peak has also a time instant @xmath7 at which it occurs , we generate in parallel another data set : a sequence of time intervals between local intensity peaks , @xmath8 . to investigate the complexity of the dynamics and to uncover hidden temporal correlations we use two methods of time - series analysis , which are represented schematically in fig .
1(c ) . the first one , known as _
ordinal analysis _
@xcite , transforms a time series @xmath9 ( where @xmath9 is either the sequence of peak heights , @xmath4 , or the sequence of time intervals between peaks , @xmath8 ) into a sequence of symbols ( referred to as ordinal patterns , ops ) , by considering the order relation among @xmath10 values of the time - series .
for example , there are 2 different ordinary patterns of length 2 : pattern ` 01 ' if @xmath11 and pattern ` 10 ' if @xmath12 .
if @xmath13 , there are 6 possible patterns : @xmath14 gives ` 012 ' , @xmath15 gives ` 210 ' , etc .
the number of possible patterns of the given length @xmath10 is @xmath16 . in this way
, a sequence of patterns could be generated from the sequence of the peak heights or from the sequence of time intervals between the peaks .
after defining the sequence of patterns , one can calculate the probability to find the given pattern in the data set , @xmath17 .
the entropy computed from their probabilities , @xmath17 , of occurrence in the time series , @xmath18 , known as permutation entropy , has been shown to be an appropriated measure of the complexity of a time - series @xcite . when there are no serial correlations in the time - series @xmath9 , then all the patterns are equally probable and @xmath19 . on the contrary , when there are serial correlations , then the ops are not all equally probable , and the permutation entropy will be @xmath20 . in the following we refer to the normalized entropy , @xmath21 , as pe entropy .
thus , with an appropriate choice of pattern length @xmath10 , the op probabilities and the pe entropy will capture the existence of underlying correlations in the time series . to verify independently presence of correlations
, we also apply the so - called _ horizontal visibility graph _ ( hvg ) method @xcite . in this approach a time series , @xmath9 , is converted into a graph by considering each data point , @xmath22 , as a node .
any two nodes are connected by a link ( or edge ) if horizontal visibility exists between them [ see fig .
1(c ) ] : @xmath22 and @xmath23 are connected if it is possible to trace a horizontal line linking @xmath22 and @xmath23 not intersecting intermediate data ; mathematically , this means that @xmath22 and @xmath23 are connected if : @xmath24 for all @xmath25 .
note that this graph representation of the time series @xmath9 takes into account both , the order and the values of the data points .
time series with different dynamics are mapped into graphs that exhibit distinct topological structures @xcite .
the topology of a graph is characterized by the degree distribution , @xmath26 , that is the probability that a node has @xmath27 links .
thus , the entropy of the degree distribution , @xmath28 ( in the following , referred to as hvg entropy ) , is another measure of the complexity of the time series @xmath9 @xcite .
the hvg method also allows to analyze different time - scales by constructing the graph not from all the `` raw '' data points , but from lagged data : @xmath29 .
while both analysis methods share the common feature of transforming the time series @xmath9 into a sequence of integer numbers , @xmath30 ( in the op case , @xmath31 $ ] is the pattern label : if ` 01 ' , @xmath32 ; if ` 10 ' , @xmath33 , etc . ; in the hvg case , @xmath34 $ ] is the degree of @xmath22 , with @xmath35 being the number of data points in the time series ) , they have important differences : while the op method requires the pre - definition of the length of the pattern @xmath10 , and does not take into account the values of the data points , the horizontal visibility graph method does not require to pre - define an analysis length , and considers both , the order relation and the actual values of the data points .
vs the pump power calculated from the sequence of ( a ) intensity peaks and ( b ) time intervals between consecutive peaks .
permutation entropy ( c ) and hvg - entropy ( d ) calculated from the sequence of intensity peaks . ] figures 2(a ) and 2(b ) display the probabilities of the six patterns of length 3 calculated from the sequence of intensity peaks and from the sequence of the time intervals , respectively .
we observe that the variation of the probabilities with the pump power captures the transition between two dynamical regimes : below the transition the ops are equally probable , while during the transition from laminar to turbulent regime their probabilities are different from equiprobability .
it can also be noticed that , during the transition , the patterns 012 and 210 become more probable ( less probable ) when the ops are calculated from the @xmath4 sequence ( from the @xmath8 sequence ) .
we also note that the patterns calculated from the intensity peaks capture more determinism than those computed from the time intervals [ notice the difference in the vertical scales of figs .
2(a ) and 2(b ) ] .
this indicates that the timing of the high intensity peaks is more random than their peak values .
the pe entropy , fig .
2(c ) , quantifies this effect by decreasing sharply at the transition power .
a similar behavior is observed when computing the hvg - entropy , fig .
2(d ) .
we have verified that , during the transition , the op probabilities ( both , for the intensity peaks and for the time intervals ) are not consistent with the null hypothesis of full stochasticity : the probabilities lie outside the region consistent with random ops , @xmath36 , where @xmath37 and @xmath38 with @xmath35 being the number of data points @xcite . in fig .
2(c ) the pe - entropy was computed from @xmath13 ops ; a similar plot was obtained with @xmath39 and @xmath40 ops ( not shown ) .
larger @xmath10 values were not considered due to the finite length of the dataset : while the `` raw '' intensity time series contains 50 million data points , the number of high intensity peaks ( above @xmath6 ) shown in fig .
1(d ) is about @xmath41 , depending on the pump power . .
( b ) pdf - entropy calculated from the distribution of intensity values . ] the probabilities of patterns 012 and 210 provide a measure of the persistence of the time - series , i.e. , the probability that the sign of @xmath42 persists in the next step @xcite .
thus , at the transition , if there are two consecutive peaks with increasing height , the next peak is likely to be larger than the previous one ( and if there are two consecutive peaks with decreasing height , the next one is likely to be smaller than the previous one ) ; on the contrary , in the sequence of time - intervals , two consecutive intervals that are increasingly long ( @xmath43 ) are likely to be followed by shorter interval ( @xmath44 ) , and two consecutive decreasing intervals ( @xmath45 ) are likely to be followed by a longer one ( @xmath46 ) .
let us note that the optimal value of pattern length @xmath10 depends of the length of correlations embedded into the time series @xcite .
to investigate the presence of specific time - scales in the dynamics , we analyze the lagged time series , i.e. the sequence of @xmath47 , where @xmath48 is the lag time .
we begin by considering the case @xmath49 ( i.e. , we analyze all data points ) .
figure 3(a ) displays the pe entropy and the hvg entropy vs the pump power , and same behavior is seen in both entropies : there is a clear transition at pump power 0.9 w , where both entropies smoothly decrease .
it is also observed that for the highest pump power , both entropies increase again .
this reveals that during the transition there is an increase in the `` ordering '' of consecutive intensity values ( that is captured by both entropies , which decrease ) , but for the highest pump power the trend reverses and the disorder increases . in contrast , the entropy computed in the conventional way and referred as pdf - entropy ( i.e. the entropy calculated from the intensity pdfs of the initial intensity dynamics , i(t ) ) does not capture this behavior : as it can be seen in fig .
3 ( b ) , after the transition the pdf - entropy monotonously decreases with the pump power . thus ,
@xmath50 and @xmath51 provide consistent information , which complements that gained from the standard pdf - entropy . the good agreement between the pe and hvg entropies , also seen in figs .
2(c ) and 2(d ) , is remarkable because the two methods transform a time series into a sequence of integer numbers by using very different encoding rules . by varying the value of the lag time @xmath48 , i.e. by taking into account not all points in data sets , but every second ( @xmath52 ) point , every third ( @xmath53 ) point etc .
, we are able to identify a specific oscillation time - scale in the intensity time - series during the transition . the pe entropy vs @xmath48 for pump powers below ( 0.85 w ) , at ( 0.90 w ) and above ( 0.95 w ) the transition is displayed in fig .
4(a ) . here
we can notice that , at the transition , there are specific lags for which the pe entropy decreases sharply .
similar results were obtained with the hvg entropy .
the sharp minima indicate that , for pump power 0.90 w and for specific lags , 6 different patterns of length @xmath13 are not all equally probable , and thus , there are serial correlations in the sequence of lagged intensity values . to explore the length of such correlations , we computed the pe entropy using longer ordinary patterns ( @xmath39 and @xmath54 ) and found that the minima were more pronounced , revealing the existence of long serial correlations .
these correlations are not captured by the classical autocorrelation function shown in fig .
4(b ) for comparison purposes , that displays only a smooth variation with @xmath48 .
vs lag time @xmath48 .
( b)-(d ) spatio - temporal structures identified with the specific lags indicated with arrows in panel ( a ) .
the color scale indicates the value of @xmath55 with @xmath56 and @xmath48=396 , 431 and 496 in units of the sampling time .
the pump power is 0.9w.,title="fig : " ] + vs lag time @xmath48 .
( b)-(d ) spatio - temporal structures identified with the specific lags indicated with arrows in panel ( a ) .
the color scale indicates the value of @xmath55 with @xmath56 and @xmath48=396 , 431 and 496 in units of the sampling time .
the pump power is 0.9w.,title="fig : " ] to investigate the nature of these correlations we plot on fig .
5(a ) how the probabilities of 6 different ordinary patterns of length @xmath13 depend on the lag time .
we note a periodic alternation in which 012 and 210 became the more probable or the less probable patterns .
the probabilities of the other four patterns are similar ( no clear clusters are seen ) .
the lag values for which 012 and 210 are less probable correspond , as expected , to the lag values where the autocorrelation function is minimum ( and negative ) .
however , unexpectedly , the lag values for which 012 and 210 are more probable , do not correspond to the maxima of the autocorrelation ; and moreover , for the lag values where the autocorrelation is maxima , the all six ordinary patterns have similar probabilities .
these observations suggest that ordinal analysis identifies subtle correlations in the _ ordering _ of data points , which are not seen by the standard autocorrelation function , that measures correlations in the _ values _ of the data points .
these `` order correlations '' result into different types of spatio - temporal patterns .
let us recall that the initial intensity spatio - temporal dynamics could be seen as intensity spatio - temporal dynamics @xcite .
similar concept has been recently employed for studying interaction of cavity solitons @xcite and topological solitons as addressable bits @xcite . here
we apply this concept and choose specific lag times defined from fig.5(a ) ( shown by arrows ) to be used as round - trip times in processing of the initial data series ( see @xcite for details of plotting spatio - temporal dynamics ) .
figures 5(b)-(d ) display examples with lags such that patterns 012 and 210 are as probable [ fig .
5(b ) ] , more probable [ fig .
5(c ) ] ) and less probable [ fig .
5(d ) ] than the other four patterns .
we can see that these spatio - temporal dynamics of patterns display clear and different coherent structures .
these observations can be useful for confronting the predictions of state - of - the - art laser models with empirical data , and the theoretical studies could provide insight into the physical mechanisms underlying these correlations . to summarize , by applying two independent tools of nonlinear time - series analysis we have uncovered long - range temporal correlations in the intensity output of a fibre laser during the transition to a wave turbulence regime .
output of laser radiation is easily measurable , making these easily implementable methods useful and valuable techniques for investigating coherent structures in complex laser radiation .
both approaches can be applied to any high - dimensional complex systems that undergo similar transitions accompanied by the generation of extreme fluctuations . |
analysis of complex high dimensional data is an exploding area of research , with applications in diverse fields , such as machine learning , statistical data analysis , bio - informatics , meteorology , chemistry and physics . in the first three application fields ,
the underlying assumption is that the data is sampled from some unknown probability distribution , typically without any notion of time or correlation between consecutive samples .
important tasks are dimensionality reduction , e.g. , the representation of the high dimensional data with only a few coordinates , and the study of the geometry and statistics of the data , its possible decomposition into clusters , etc @xcite .
in addition , there are many problems concerning supervised learning , in which additional information , such as a discrete class @xmath3 or a continuous function value @xmath4 is given to some of the data points . in this paper
we are concerned only with the unsupervised case , although some of the methods and ideas presented can be applied to the supervised or semi - supervised case as well @xcite . in the later three above - mentioned application fields the data is typically sampled from a complex biological , chemical or physical _ dynamical _ system ,
in which there is an inherent notion of time .
many of these systems involve multiple time and length scales , and in many interesting cases there is a separation of time scales , that is , there are only a few `` slow '' time scales at which the system performs conformational changes from one meta - stable state to another , with many additional fast time scales at which the system performs local fluctuations within these meta - stable states . in the case of macromolecules
the slow time scale is that of a conformational change , while the fast time scales are governed by the chaotic rotations and vibrations of the individual chemical bonds between the different atoms of the molecule , as well as the random fluctuations due to the frequent collisions with the surrounding solvent water molecules . in the more general case of interacting particle systems ,
the fast time scales are those of density fluctuations around the mean density profiles , while the slow time scales correspond to the time evolution of these mean density profiles .
although on the fine time and length scales the full description of such systems requires a high dimensional space , e.g. the locations ( and velocities ) of all the different particles , these systems typically have an intrinsic low dimensionality on coarser length and time scales .
thus , the coarse time evolution of the high dimensional system can be described by only a few dynamically relevant variables , typically called reaction coordinates .
important tasks in such systems are the reduction of the dimensionality at these coarser scales ( known as homogenization ) , and the efficient representation of the complicated linear or non - linear operators that govern their ( coarse grained ) time evolution .
additional goals are the identification of the meta - stable states , the characterization of the transitions between them and the efficient computation of mean exit times , potentials of mean force and effective diffusion coefficients @xcite . in this paper , following @xcite , we consider a family of diffusion maps for the analysis of these problems .
given a large dataset , we construct a family of random walk processes based on isotropic and anisotropic diffusion kernels and study their first few eigenvalues and eigenvectors ( principal components ) . the key point in our analysis
is that these eigenvectors and eigenvalues capture important geometrical and statistical information regarding the structure of the underlying datasets .
it is interesting to note that similar approaches have been suggested in various different fields . in graph theory
, the first few eigenvectors of the normalized graph laplacian have been used for spectral clustering @xcite , approximations to the optimal normalized - cut problem @xcite and dimensionality reduction @xcite , to name just a few .
similar constructions have also been used for the clustering and identification of meta - stable states for datasets sampled from dynamical systems @xcite .
however , it seems that the connection of these computed eigenvectors to the underlying geometry and probability density of the dataset has not been fully explored . in this paper
, we consider the connection of these eigenvalues and eigenvectors to the underlying geometry and probability density distribution of the dataset . to this end , we assume that the data is sampled from some ( unknown ) probability distribution , and view the eigenvectors computed on the finite dataset as discrete approximations of corresponding eigenfunctions of suitably defined continuum operators in an infinite population setting . as the number of samples goes to infinity , the discrete random walk on the set converges to a diffusion process defined on the @xmath5-dimensional space but with a non - uniform probability density . by explicitly studying the asymptotic form of the chapman - kolmogorov equations in this setting ( e.g. , the infinitesimal generators ) , we find that for data sampled from a general probability distribution , written in boltzmann form as @xmath6 , the eigenfunctions and eigenvalues of the standard normalized graph laplacian construction correspond to a diffusion process with a potential @xmath7 ( instead of a single @xmath2 ) .
therefore , a subset of the first few eigenfunctions are indeed well suited for spectral clustering of data that contains only a few well separated clusters , corresponding to deep wells in the potential @xmath2 . motivated by the well known connection between diffusion processes and schrdinger operators @xcite , we propose a different novel non - isotropic construction of a random walk on the graph , that in the asymptotic limit of infinite data recovers the eigenvalues and eigenfunctions of a diffusion process with the same potential @xmath2 .
this normalization , therefore , is most suited for the study of the long time behavior of complex dynamical systems that evolve in time according to a stochastic differential equation .
for example , in the case of a dynamical system driven by a bistable potential with two wells , ( e.g. with one slow time scale for the transition between the wells and many fast time scales ) the second eigenfunction can serve as a parametrization of the reaction coordinate between the two states , much in analogy to its use for the construction of an approximation to the optimal normalized cut for graph segmentation .
for the analysis of dynamical systems , we also propose to use a subset of the first few eigenfunctions as reaction coordinates for the design of fast simulations .
the main idea is that once a parametrization of dynamically meaningful reaction coordinates is known , and lifting and projection operators between the original space and the diffusion map are available , detailed simulations can be initialized at different locations on the reaction path and run only for short times , to estimate the transition probabilities to different nearby locations in the reaction coordinate space , thus efficiently constructing a potential of mean field and an efficient diffusion coefficient on the reaction path @xcite .
finally , we describe yet another random walk construction that in the limit of infinite data recovers the laplace - beltrami ( heat ) operator on the manifold on which the data resides , regardless of the possibly non - uniform sampling of points on the manifold .
this normalization is therefore best suited for learning the geometry of the dataset , as it separates the geometry of the manifold from the statistics on it .
our analysis thus reveals the intimate connection between the eigenvalues and eigenfunctions of different random walks on the finite graph to the underlying geometry and probability distribution @xmath8 from which the dataset was sampled .
these findings lead to a better understanding of the advantages and limitations of diffusion maps as a tool to solve different tasks in the analysis of high dimensional data .
consider a finite dataset @xmath9 .
we consider two different possible scenarios for the origin of the data . in the first scenario ,
the data is not necessarily derived from a dynamical system , but rather it is randomly sampled from some arbitrary probability distribution @xmath10 . in this case
we define an associated potential @xmath11 so that @xmath12 . in the second scenario , we assume that the data is sampled from a dynamical system in equilibrium .
we further assume that the dynamical system , defined by its state @xmath13 at time @xmath14 , satisfies the following generic stochastic differential equation ( sde ) @xmath15 where a dot on a variable means differentiation with respect to time , @xmath16 is the free energy at @xmath17 ( which , with some abuse of nomenclature , we will also call the potential at @xmath17 ) , and @xmath18 is an @xmath5-dimensional brownian motion process . in this case
there is an explicit notion of time , and the transition probability density @xmath19 of finding the system at location @xmath17 at time @xmath14 , given an initial location @xmath20 at time @xmath21 ( @xmath22 ) , satisfies the forward fokker - planck equation ( fpe ) @xcite @xmath23 with initial condition @xmath24 similarly , the backward fokker - planck equation for the density @xmath25 , in the backward variables @xmath26 ( @xmath27 ) is @xmath28 where differentiations in ( [ backward_fpe ] ) are with respect to the variable @xmath20 , and the laplacian @xmath29 is a negative operator , defined as @xmath30 . as time
@xmath31 the steady state solution of ( [ fpe ] ) is given by the equilibrium boltzmann probability density , @xmath32 where @xmath33 is a normalization constant ( known as the partition function in statistical physics ) , given by @xmath34 in what follows we assume that the potential @xmath2 is shifted by the suitable constant ( which does not change the sde ( [ sde ] ) ) , so that @xmath35
. also , we use the notation @xmath36 interchangeably to denote the ( invariant ) probability measure on the space . note that in both scenarios , the steady state probability density , given by ( [ mu_x ] ) is identical .
therefore , for the purpose of our initial analysis , which does not directly take into account the possible time dependence of the data , it is only the features of the underlying potential @xmath2 that come into play . the langevin equation ( [ sde ] ) or the corresponding fokker - planck equation ( [ fpe ] ) are commonly used to describe mechanical , physical , chemical , or biological systems driven by noise .
the study of their behavior , and specifically the decay to equilibrium has been the subject of much theoretical research @xcite . in general , the solution of the fokker - planck equation ( [ fpe ] ) can be written in terms of an eigenfunction expansion @xmath37 where @xmath38 are the eigenvalues of the fp operator , with @xmath39 , @xmath40 are their corresponding eigenfunctions , and the coefficients @xmath41 depend on the initial conditions .
obviously , the long term behavior of the system is governed only by the first few eigenfunctions @xmath42 , where @xmath43 is typically small and depends on the time scale of interest . in low dimensions ,
e.g. @xmath44 for example , it is possible to calculate approximations to these eigenfunctions via numerical solutions of the relevant partial differential equations . in high dimensions
, however , this approach is in general infeasible and one typically resorts to simulations of trajectories of the corresponding sde ( [ sde ] ) . in this case , there is a need to employ statistical methods to analyze the simulated trajectories , identify the slow variables , the meta - stable states , the reaction pathways connecting them and the mean transition times between them .
let @xmath45 , denote the @xmath46 samples , either merged from many different simulations of the stochastic equation ( [ sde ] ) , or simply given without an underlying dynamical system . in @xcite , coifman and
lafon suggested the following method , based on the definition of a markov chain on the data , for the analysis of the geometry of general datasets : for a fixed value of @xmath47 ( a metaparameter of the algorithm ) , define an isotropic diffusion kernel , @xmath48 assume that the transition probability between points @xmath49 and @xmath50 is proportional to @xmath51 , and construct an @xmath52 markov matrix , as follows @xmath53 where @xmath54 is the required normalization constant , given by @xmath55 for large enough values of @xmath47 the markov matrix @xmath56 is fully connected ( in the numerical sense ) and therefore has an eigenvalue @xmath57 with multiplicity one and a sequence of additional @xmath58 non - increasing eigenvalues @xmath59 , with corresponding eigenvectors @xmath60 .
the diffusion map at time @xmath61 is defined as the mapping from @xmath17 to the vector @xmath62 for some small value of @xmath43 . in @xcite
, it was demonstrated that this mapping gives a low dimensional parametrization of the geometry and density of the data . in the field of data analysis ,
this construction is known as the _ normalized graph laplacian_. in @xcite , shi and malik suggested using the first non - trivial eigenvector to compute an approximation to the optimal normalized cut of a graph , while the first few eigenvectors were suggested by weiss et al .
@xcite for clustering .
similar constructions , falling under the general term of kernel methods have been used in the machine learning community for classification and regression @xcite . in this paper
we elucidate the connection between this construction and the underlying potential @xmath2 . to analyze the eigenvalues and eigenvectors of the normalized graph laplacian
, we consider them as a finite approximation of a suitably defined diffusion operator acting on the continuum probability space from which the data was sampled .
we thus consider the limit of the above markov chain process as the number of samples approaches infinity . for a finite value of @xmath47
, the markov chain in discrete time and space converges to a markov process in discrete time but continuous space .
then , in the limit @xmath63 , this jump process converges to a diffusion process on @xmath64 , whose local transition probability depends on the non - uniform probability measure @xmath65 .
we first consider the case of a fixed @xmath66 , and take @xmath67 .
using the similarity of ( [ k_epsilon ] ) to the diffusion kernel , we view @xmath47 as a measure of time and consider a discrete jump process at time intervals @xmath68 , with a transition probability between points @xmath20 and @xmath17 proportional to @xmath69 .
however , since the density of points is not uniform but rather given by the measure @xmath70 , we define an associated normalization factor @xmath71 as follows , @xmath72 and a forward transition probability @xmath73 equations ( [ p_ve ] ) and ( [ m_f ] ) are the continuous analogues of the discrete equations ( [ p_ve_discrete ] ) and ( [ m_discrete ] ) . for future use
, we also define a symmetric kernel @xmath74 as follows , @xmath75 note that @xmath76 is an estimate of the local probability density at @xmath17 , computed by averaging the density in a neighborhood of radius @xmath77 around @xmath17 . indeed , as @xmath78
, we have that @xmath79 we now define forward , backward and symmetric chapman - kolmogorov operators on functions defined on this probability space , as follows , @xmath80 ( { { \mbox{\boldmath$x$ } } } ) = \int m_f ( { { \mbox{\boldmath$x$ } } } | { { \mbox{\boldmath$y$ } } } ) \varphi ( { { \mbox{\boldmath$y$ } } } ) d\mu ( { { \mbox{\boldmath$y$ } } } ) \ ] ] @xmath81 ( { { \mbox{\boldmath$x$ } } } ) = \int m_f ( { { \mbox{\boldmath$y$ } } } | { { \mbox{\boldmath$x$ } } } ) \varphi ( { { \mbox{\boldmath$y$ } } } ) d\mu ( { { \mbox{\boldmath$y$ } } } ) \ ] ] and @xmath82 ( { { \mbox{\boldmath$x$ } } } ) = \int m_s ( { { \mbox{\boldmath$x$ } } } , { { \mbox{\boldmath$y$ } } } ) \varphi ( { { \mbox{\boldmath$y$ } } } ) d\mu ( { { \mbox{\boldmath$y$ } } } ) \ ] ] if @xmath83 is the probability of finding the system at location @xmath17 at time @xmath84 , then @xmath85 $ ] is the evolution of this probability to time @xmath86 .
similarly , if @xmath87 is some function on the space , then @xmath88 ( { { \mbox{\boldmath$x$ } } } ) $ ] is the mean ( average ) value of that function at time @xmath47 for a random walk that started at @xmath17 , and so @xmath89 ( { { \mbox{\boldmath$x$ } } } ) $ ] is the average value of the function at time @xmath90 . by definition , the operators @xmath91 and @xmath92 are adjoint under the inner product with weight @xmath93 , while the operator @xmath94 is self adjoint under this inner product , @xmath95 moreover , since @xmath94 is obtained via conjugation of the kernel @xmath96 with @xmath97 all three operators share the same eigenvalues .
the corresponding eigenfunctions can be found via conjugation by @xmath98 .
for example , if @xmath99 , then the corresponding eigenfunctions for @xmath91 and @xmath92 are @xmath100 and @xmath101 , respectively . since @xmath98 is the first eigenfunction with @xmath102 of @xmath94 , the steady state of the forward operator
is simply @xmath76 , while for the backward operator it is the uniform density , @xmath103 .
obviously , the eigenvalues and eigenvectors of the discrete markov chain described in the previous section are discrete approximations to the eigenvalues and eigenfunctions of these continuum operators .
rigorous mathematical proofs of this convergence as @xmath67 under various assumptions have been recently obtained by several authors @xcite .
therefore , for a better understanding of the finite sample case , we are interested in the properties of the eigenvalues and eigenfunctions of either one of the operators @xmath104 or @xmath94 , and how these relate to the measure @xmath70 and to the corresponding potential @xmath2 . to this end
, we look for functions @xmath83 such that @xmath105 where @xmath106 . while in the case of a finite amount of data
, @xmath47 must remain finite so as to have enough neighbors in a ball of radius @xmath77 near each point @xmath17 , as the number of samples goes to infinity , we can take smaller and smaller values of @xmath47 .
therefore , it is instructive to look at the limit @xmath78 . in this case , the transition probability densities of the continuous in space but discrete in time markov chain converge to those of a diffusion process , whose time evolution is described by a differential equation @xmath107 where @xmath108 is the infinitesimal generator or propagator of the forward operator , defined as @xmath109 as shown in the appendix , by computing the asymptotic expansion of the corresponding integrals in the limit @xmath63 , we obtain that @xmath110 similarly , the inifinitesimal operator of the backward operator is given by @xmath111 as expected , @xmath112 is the eigenfunction with @xmath113 of the backward infinitesimal operator , while @xmath114 is the eigenfunction of the forward one .
a few important remarks are due at this point .
first , note that the backward operator ( [ h_b ] ) has the same functional form as the backward fpe ( [ backward_fpe ] ) , but with a potential @xmath1 instead of @xmath2 .
the forward operator ( [ h_f ] ) has a different functional form from the forward fpe ( [ fpe ] ) corresponding to the stochastic differential equation ( [ sde ] ) .
this should come as no surprise , since ( [ h_f ] ) is the differential operator of an isotropic diffusion process on a space with non - uniform probability measure @xmath70 , which is obviously different from the standard anisotropic diffusion in a space with a uniform measure , as described by the sde ( [ sde ] ) @xcite .
interestingly , however , the form of the forward operator is the same as the schrdinger operator of quantum physics @xcite , e.g. @xmath115 where in our case the potential @xmath116 has the following specific form @xmath117 therefore , in the limit @xmath118 , the eigenfunctions of the diffusion map are the same as those of the schrdinger operator ( [ schrodinger ] ) with a potential ( [ vx ] ) .
the properties of the first few of these eigenfunctions have been extensively studied theoretically for simple potentials @xmath116 @xcite . in order to see why the forward operator @xmath108 also corresponds to a potential @xmath119 instead of @xmath2 , we recall that there is a well known correspondence @xcite , between the schrdinger equation with a sypersymmetric potential of the specific form ( [ vx ] ) and a diffusion process described by a fokker - planck equation of the standard form ( [ fpe ] ) .
the transformation @xmath120 applied to the original fpe ( [ fpe ] ) yields the schrdinger equation with imaginary time @xmath121 comparing ( [ eq : schrodinger imaginary time ] ) with ( [ vx ] ) , we conclude that the eigenvalues of the operator ( [ h_f ] ) are the same as those of a fokker - planck equation with a potential @xmath1 .
therefore , in general , for data sampled from the sde ( [ sde ] ) , there is no direct correspondence between the eigenvalues and eigenfunctions of the normalized graph laplacian and those of the corresponding fokker - planck equation ( [ fpe ] ) .
however , when the original potential @xmath2 has two metastable states separated by a large barrier , corresponding to two well separated clusters , so does @xmath119 .
therefore , the first non - trivial eigenvalue is governed by the mean passage time between the two barriers , and the first non - trivial eigenfunction gives a parametrization of the path between them ( see also the analysis of the next section ) .
we note that in @xcite , horn and gottlieb suggested a clustering algorithm based on the schrdinger operator ( [ qm ] ) , where they constructed an approximate eigenfunction @xmath122 as in our eq .
[ p_ve_discrete ] ) , and computed its corresponding potential @xmath116 from eq .
( [ qm ] ) .
the clusters were then defined by the minima of the potential @xmath123 . employing the same asymptotic analysis of this paper
, one can show that in the appropriate limit , the computed potential @xmath123 is given by ( [ vx ] ) . this asymptotic analysis and the connection between the quantum operator and a diffusion process
thus provides a mathematical explanation for the success of their method .
indeed , when @xmath16 has a deep parabolic minima at a point @xmath17 , corresponding to a well defined cluster , so does @xmath123 .
in the previous section we showed that asymptotically , the eigenvalues and eigenfunctions of the normalized graph laplacian operator correspond to the fokker - planck equation with a potential @xmath119 instead of the single @xmath2 . in this section
we present a different normalization that yields infinitesimal generators corresponding to the potential @xmath2 without the additional factor of two .
in fact , following @xcite we consider in more generality a whole family of different normalizations and their corresponding diffusions , and we show that , in addition to containing the graph laplacian normalization used in the previous section , this collection of diffusions includes at least two other laplacians of interest : the laplace - beltrami operator , which captures the riemannian geometry of the data set , and the backward fokker - planck operator of equation ( [ backward_fpe ] ) . instead of applying the graph laplacian normalization to the isotropic kernel @xmath124
, we first appropriately renormalize the kernel into an anisotropic one to obtain a new weighted graph , and then apply the graph laplacian normalization to this graph .
more precisely , we proceed as follows : start with a gaussian kernel @xmath124 and let @xmath125 be a parameter indexing our family of diffusions . define an estimate for the local density as @xmath126 and consider the family of kernels @xmath127 we now apply the graph laplacian normalization by computing the normalization factor @xmath128 and forming a forward transition probability kernel @xmath129 similar to the analysis of section [ sec : continuum diffusion process ] , we can construct the corresponding forward , symmetric and backward diffusion kernels .
it can be shown ( see appendix [ infinitesimal computations ] ) that the forward and backward infinitesimal generators of this diffusion are @xmath130 we mention three interesting cases : * for @xmath131 , this construction yields the classical normalized graph laplacian with the infinitesimal operator given by equation ( [ h_f ] ) @xmath132 * for @xmath133 , the backward generator gives the laplace - beltrami operator : @xmath134 in other words , this diffusion captures only the geometry of the data , in which the density @xmath135 plays absolutely no role . therefore
, this normalization separates the geometry of the underlying manifold from the statistics on it .
* for @xmath136 , the infinitesimal operator of the forward and backward operators coincide and are given by @xmath137 which is exactly the backward fpe ( [ backward_fpe ] ) , with a potential @xmath2 .
therefore , the last case with @xmath138 provides a consistent method to approximate the eigenvalues and eigenfunctions corresponding to the stochastic differential equation ( [ sde ] ) .
this is done by constructing a graph laplacian with an appropriately anisotropic weighted graph . as explained in @xcite ,
the euclidian distance between any two points after the diffusion map embedding into @xmath139 is almost equal to their diffusion distance on the original dataset .
thus , for dynamical systems with only one or two slow time scales , and many fast time scales , only a small number of diffusion map coordinates need be retained for the coarse grained representation of the data at medium to long times , at which the fast coordinates have equilibrated .
therefore , the diffusion map can be considered as an empirical method to perform data - driven or equation - free homogenization .
in particular , since this observation yields a computational method for the approximation of the top eigenfunctions and eigenvalues , this method can be applied towards the design of fast and efficient simulations that can be initialized on arbitrary points on the diffusion map .
this application will be described in a separate publication @xcite .
in this section we present the potential strength of the diffusion map method by analyzing , both analytically and numerically a few toy examples , with simple potentials @xmath2 .
more complicated high dimensional examples of stochastic dynamical systems are analyzed in @xcite , while other applications such as the analysis of images for which we typically have no underlying probability model appear in @xcite .
we start with the simplest case of a parabolic potential in one dimension , which in the context of the sde ( [ sde ] ) , corresponds to the well known ornstein - uhlenbeck process .
we thus consider a potential @xmath140 , with a corresponding normalized density @xmath141 .
the normalization factor @xmath54 can be computed explicitly @xmath142 where , for convenience , we multiplied the kernel @xmath143 by a normalization factor @xmath144 . therefore , the eigenvalue / eigenfunction problem for the symmetric operator @xmath94 with a finite @xmath47 reads @xmath145 the first eigenfunction , with eigenvalue @xmath57 is given by @xmath146 the second eigenfunction , with eigenvalue @xmath147 is , up to normalization @xmath148 in general , the sequence of eigenfunctions and eigenvalues is characterized by the following lemma : * _ lemma : _ * the eigenvalues of the operator @xmath94 are @xmath149 , with the corresponding eigenvectors given by @xmath150 where @xmath151 is a polynomial of degree @xmath43 ( even or odd depending on @xmath43 ) . in the limit @xmath78 , we obtain the eigenfunctions of the corresponding infinitesimal generator .
for the specific potential @xmath152 , the eigenfunction problem for the backward generator reads @xmath153 the solutions of this eigenfunction problem are , up to scaling of @xmath154 , the well known hermite polynomials , which by the correspondence of this operator to the schrdinger eigenvector / eigenvalue problem , are also the eigenfunctions of the quantum harmonic oscillator ( after multiplication by the appropriate gaussian ) @xcite . note that plotting the second vs. the first eigenfunctions ( with the convention that the zeroth eigenfunction is the constant one , which we typically ignore ) , is the same as plotting @xmath155 vs @xmath154 , e.g. a parabola .
therefore , we expect that for a large enough and yet finite data - set sampled from this potential , the plot of the corresponding discrete eigenfunctions should lay on a parabolic curve ( see next section for a numerical example ) .
we now consider a harmonic potential in @xmath5-dimensions , of the form @xmath156 where , in addition , we assume @xmath157 , so that @xmath158 is a slow variable in the context of the sde ( [ sde ] ) . we note that for this specific potential , the probability density has a separable structure , @xmath159 , and so does the kernel @xmath69 , and consequently , also the symmetric kernel @xmath74 .
therefore , there is an outer - product structure to the eigenvalues and eigenfunctions .
for example , in two dimensions the eigenfunctions and eigenvalues are given by @xmath160 where @xmath161 and @xmath162 .
since by assumption @xmath163 , then upon ordering of the eigenfunctions by decreasing eigenvalue , the first non - trivial eigenfunctions are @xmath164 , which depend only on the slow variable @xmath158 . note that indeed , regardless of the value of @xmath47 , as long as @xmath165 , we have that @xmath166 .
therefore , in this example the first few coordinates of the diffusion map give a ( redundant ) parametrization of the slow variable @xmath158 in the system . in figure
[ f : u1 ] we present numerical results corresponding to a 2-dimensional potential with @xmath167 . in the upper left some 3500 points sampled from the distribution @xmath8 are shown . in the lower right and left panels , the first two non - trivial backward eigenfunctions @xmath168 and @xmath169 are plotted vs. the slow variable @xmath158 .
note that except at the edges , where the statistical sampling is poor , the first eigenfunction is linear in @xmath158 , while the second one is quadratic in @xmath158 . in the upper right panel
we plot @xmath169 vs. @xmath168 and note that they indeed lie on a parabolic curve , as predicted by the analysis of the previous section .
+ we now consider a double well potential @xmath170 with two minima , one at @xmath171 and one at @xmath172 .
for simplicity of the analysis , we assume a symmetric potential around @xmath173 , with @xmath174 ( see figure [ f : u2 ] ) . in the context of data clustering , this can be viewed as approximately a mixture of two gaussian clouds , while in the context of stochastic dynamical systems , this potential defines two meta - stable states .
we first consider an approximation to the quantity @xmath175 , given by eq .
( [ p_ve ] ) . for @xmath154 near @xmath171 , @xmath176 , while for @xmath154 near @xmath172 , @xmath177 . therefore , @xmath178 and @xmath179\end{aligned}\ ] ] where @xmath180 and @xmath181 are the first forward eigenfunctions corresponding to a single well potential centered at @xmath171 or at @xmath172 , respectively .
as is well known both in the theory of quantum physics and in the theory of the fokker - planck equation , an approximate expression for the next eigenfunction is @xmath182\ ] ] therefore , the first non - trivial eigenfunction of the backward operator is given by @xmath183 this eigenfunction is roughly @xmath184 in one well and @xmath185 in the other well , with a sharp transition between the two values near the barrier between the two wells .
therefore , this eigenfunction is indeed suited for clustering .
moreover , in the context of a mixture of two gaussian clouds , clustering according to the sign of @xmath186 is asymptotically equivalent to the optimal bayes classifier .
* example : * consider the following potential in two dimensions , @xmath187 in the @xmath154 direction , this potential has a double well shape with two minima , one at @xmath188 and one at @xmath189 , separated by a potential barrier with a maximum at @xmath190 . in figure
[ f : u2 ] we show some numerical results of the diffusion map on some 1200 points sub - sampled from a stochastic simulation with this potential which generated about 40,000 points . on the upper right panel
we see the potential @xmath191 , showing the two wells . in the upper left , a scatter plot of all the points , color coded
according to the value of the local estimated density @xmath54 , ( with @xmath192 ) is shown , where the two clusters are easily observed . in the lower left panel ,
the first non - trivial eigenfunction is plotted vs. the first coordinate @xmath154 .
note that even though there is quite a bit of variation in the @xmath193-variable inside each of the wells , the first eigenfunction @xmath168 is essentially a function of only @xmath154 , regardless of the value of @xmath193 . in the lower right
we plot the first three backward eigenfunctions .
note that they all lie on a curve , indicating that the long time asymptotics are governed by the passage time between the two wells and not by the local fluctuations inside them .
+ we now consider the following two dimensional potential energy with three wells , @xmath194 -5\beta e^{-y^2 } \left[e^{-(x-1)^2 } + e^{-(x+1)^2}\right]\ ] ] where @xmath195 is a thermal factor .
this potential has two deep wells at @xmath196 and at @xmath197 and a shallower well at @xmath198 , which we denote as the points @xmath199 , respectively , the transitions between the wells of this potential have been analyzed in many works @xcite . in figure
[ f : three_wells ] we plotted on the left the results of 1400 points sub - sampled from a total of 80000 points randomly generated from this potential confined to the region @xmath200
^ 2\subset \mathbb{r}^2 $ ] at temperature @xmath201 , color - coded by their local density . on the right we plotted the first two diffusion map coordinates @xmath202 . notice how in the diffusion map space one can clearly see a triangle where each vertex corresponds to one of the points @xmath199 .
this figure shows very clearly that there are two possible pathways to go from @xmath203 to @xmath204
. a direct ( short ) way and an indirect longer way , that passes through the shallow well centered at @xmath205 .
+ we conclude this section with a diffusion map analysis of one of the most popular multivariate datasets in pattern recognition , the iris data set .
this set contains 3 distinct classes of samples in four dimensions , with 50 samples in each class . in figure [ f : iris ] we see on the left the result of the three dimensional diffusion map on this dataset .
this picture clearly shows that all 50 points of class 1 ( blue ) are shrunk into a single point in the diffusion map space and can thus be easily distinguished from classes two and three ( red and green ) . in the right plot
we see the results of re - running the diffusion map on the 100 remaining red and green samples .
the 2-d plot of the first two diffusion maps coordinates shows that there is no perfect separation between these two classes .
however , clustering according to the sign of @xmath206 gives misclassifications rates similar to those of other methods , of the order of 6 - 8 samples depending on the value chosen for the kernel width @xmath47 .
in this paper , we introduced a mathematical framework for the analysis of diffusion maps , via their corresponding infinitesimal generators .
our results show that diffusion maps are a natural method for the analysis of the geometry and probability distribution of empirical data sets .
the identification of the eigenvectors of the markov chain as discrete approximations to the corresponding differential operators provides a mathematical justification for their use as a dimensional reduction tool and gives an alternative explanation for their empirical success in various data analysis applications , such as spectral clustering and approximations of optimal normalized cuts on discrete graphs .
we generalized the standard construction of the normalized graph laplacian to a one - parameter family of graph laplacians that provides a low - dimensional description of the data combining the geometry of the set with the probability distribution of the data points .
the choice of the diffusion map depends on the task at hand .
if , for example , data points are known to approximately lie on a manifold , and one is solely interested in recovering the geometry of this set , then an appropriate normalization of a gaussian kernel allows to approximate the laplace - beltrami operator , regardless of the density of the data points .
this construction achieves a complete separation of the underlying geometry , represented by the knowledge of the laplace operator , from the statistics of the points .
this is important in situations where the density is meaningless , and yet points on the manifold are not sampled uniformly on it . in a different scenario ,
if the data points are known to be sampled from the equilibrium distribution of a fokker - planck equation , the long - time dynamics of the density of points can be recovered from an appropriately normalized random walk process . in this case
, there is a subtle interaction between the distribution of the points and the geometry of the data set , and one must correctly account for the density of the points .
while in this paper we analyzed only gaussian kernels , our asymptotic results are valid for general kernels , with the appropriate modification that take into account the mean and covariance matrix of the kernel .
note , however , that although asymptotically in the limit @xmath67 and @xmath78 , the choice of the isotropic kernel is unimportant , for a finite data set the choice of both @xmath47 and the kernel can be crucial for the success of the method .
finally , in the context of dynamical systems , we showed that diffusion maps with the appropriate normalization constitute a powerful tool for the analysis of systems exhibiting different time scales .
in particular , as shown in the different examples , these time scales can be separated and the long time dynamics can be characterized by the top eigenfunctions of the diffusion operator .
last , our analysis paves the way for fast simulations of physical systems by allowing larger integration steps along slow variable directions .
the exact details required for the design of fast and efficient simulations based on diffusion maps will be described in a separate publication @xcite .
* acknowledgments : * the authors would like to thank the referee for helpful suggestions and for pointing out ref . @xcite .
in this appendix , we present the calculation of the infinitesimal generators for the different diffusion maps characterized by a parameter @xmath207 .
suppose that the data set @xmath208 consists of a riemannian manifold with a density @xmath0 and let @xmath69 be a gaussian kernel .
it was shown in @xcite that if @xmath209 is scaled appropriately , then for any function @xmath210 on @xmath208 , @xmath211 where @xmath212 is a function that depends on the riemannian geometry of the manifold and its embedding in @xmath64 . using the notations introduced in section [ ref : anisotropic diffusion maps ] , it is easy to verify that @xmath213 and consequently , @xmath214 let @xmath215 then , the normalization factor @xmath216 is given by @xmath217\ ] ] therefore , the asymptotic expansion of the backward operator gives @xmath218 and its infinitesimal generator is @xmath219 inserting the expression @xmath8 into the last equation gives @xmath220 similarly , the form of the forward infinitesimal operator is @xmath221 r. r. coifman , s. lafon , a. b. lee , m. maggioni , b. nadler , f. warner and s. zucker , geometric diffusions as a tool for harmonic analysis and structure definition of data , part i : diffusion maps , _ proc .
_ , in press .
m. saerens , f. fouss , l. yen and p. dupont , _ the principal components analysis of a graph and its relationships to spectral clustering _ , proceedings of the 15th european conference on machine learning ( ecml 2004 ) , lecture notes in artificial intelligence , vol . 3201 , springer - verlag , berlin , 2004 , pp 371 - 383 .
kevrekidis , c.w .
gear , j. m. hyman , p.g .
kevrekidis , o. runborg , c. theodoropoulos , `` equation free multiscale computation : enabling microscopic simulators to perform system - level tasks '' , _ comm .
_ , submitted . | a central problem in data analysis is the low dimensional representation of high dimensional data , and the concise description of its underlying geometry and density . in the analysis of large scale simulations of complex dynamical systems , where the notion of time evolution comes into play ,
important problems are the identification of slow variables and dynamically meaningful reaction coordinates that capture the long time evolution of the system . in this paper
we provide a unifying view of these apparently different tasks , by considering a family of _ diffusion maps _ , defined as the embedding of complex ( high dimensional ) data onto a low dimensional euclidian space , via the eigenvectors of suitably defined random walks defined on the given datasets . assuming that the data is randomly sampled from an underlying general probability distribution @xmath0 , we show that as the number of samples goes to infinity , the eigenvectors of each diffusion map converge to the eigenfunctions of a corresponding differential operator defined on the support of the probability distribution .
different normalizations of the markov chain on the graph lead to different limiting differential operators .
for example , the normalized graph laplacian leads to a backward fokker - planck operator with an underlying potential of @xmath1 , best suited for spectral clustering .
a specific anisotropic normalization of the random walk leads to the backward fokker - planck operator with the potential @xmath2 , best suited for the analysis of the long time asymptotics of high dimensional stochastic systems governed by a stochastic differential equation with the same potential @xmath2 .
finally , yet another normalization leads to the eigenfunctions of the laplace - beltrami ( heat ) operator on the manifold in which the data resides , best suited for the analysis of the geometry of the dataset , regardless of its possibly non - uniform density . |
we consider a massless scalar field @xmath0 interacting with a cylindrical @xmath1-function potential background with corrugations , @xmath2 where indices 1 and 2 represent the two concentric cylinders with radii @xmath3 and @xmath4 shown on the left side of fig . [ corru ] .
the functions @xmath5 and @xmath6 describe the corrugations associated with the cylinders and we are going to assume they are sinusoidal . due to the geometry of our set up we have translational invariance on the @xmath7-axes as well as time independence allowing us to make a fourier transform of the green s function on those coordinates , @xmath8 where the reduced green s function satisfies the differential equation , @xmath9 g(r , r^\prime,\theta,\theta^\prime;k,\omega ) = \frac{\delta ( r - r^\prime)}{r}\delta(\theta-\theta^\prime ) .
\label{g - differential - eq}\end{aligned}\ ] ] the central formula in our calculation is the multiple scattering formula for the casimir energy @xcite , @xmath10 where @xmath11 is the green s function associated with the background . in this sense
we are going to define the background as two concentric semitransparent cylinders and the corrugations are considered to be a perturbation over such a background
. therefore @xmath12 satisfies @xmath13 g_m^{(0)}(r , r^\prime;k,\omega ) = \frac{\delta ( r - r^\prime)}{r}. \label{g0-differential - eq}\end{aligned}\ ] ] here @xmath14 is the potential for the background associated with cylinder @xmath15 and we define @xmath16 as the deviation of the total potential @xmath17 from it , @xmath18^n}{n ! } \frac{\partial^n}{\partial r^n } v_i^{(0)}(r)= \sum_{n=1}^\infty v_i^{(n)}(r,\theta ) .
\label{potential - expansion}\ ] ] this allows us to extract the interaction term @xmath19 from eq .
( [ ceoriginal ] ) , which is the only term in @xmath20 that gives rise to the torque .
it is expressed in terms of the perturbation potential and the green s function @xmath21 associated to each cylinder , @xmath22 .
\label{cecorrug}\ ] ] where @xmath21 satisfies @xmath23g_i(r,\theta ) = 1 $ ] , and can be expanded in terms of the green s function associated with the background , @xmath24 for @xmath25 .
once we know the energy associated with this configuration , we can calculate the torque corresponding to a shift of one of the cylinders with respect to the other described by an angular rotation @xmath26 , where @xmath27 , is the lateral shift between the plates . ]
we concentrate now in the case when the corrugations can be treated as small compared to their wavelength . in this case we can make an expansion in the corrugation amplitude in eq .
( [ cecorrug ] ) and keep the leading order to start with .
let s assume sinusoidal corrugations and induce a rotation by an angle @xmath26 on one of the cylinders .
then these corrugations can be expressed as , @xmath29 $ ] and @xmath30 $ ] , where @xmath31 are the amplitudes and @xmath32 is the frequency associated with the corrugations .
the leading order contribution plays a role in the total energy only when both cylinders have the same frequency , @xmath33 , giving rise to a second order term in the amplitude , @xmath34= \frac{i}{2\tau } \,\rm{tr } \mathcal{k}^{(2)}. \label{e12(2)-general}\ ] ] in terms of the reduced green s functions , the @xmath35 is given by : @xmath36g_{m^\prime}^{(0)}(r^\prime , r)\big[\frac{\partial}{\partial r}\delta(r - a_2)\big],\end{aligned}\ ] ] where we have used eq .
( [ potential - expansion ] ) . to evaluate this expression
we integrate by parts in @xmath37 and @xmath38 and fourier transform the functions @xmath39 that describe the corrugations on the cylinders , @xmath40 .
we can now integrate on @xmath41 and switch to imaginary frequencies by an euclidean rotation , @xmath42 , @xmath43\bigg|_{\bar{r}=a_1 , r = a_2}. \label{e12(2)}\end{aligned}\ ] ] here @xmath44 and the @xmath45 s are the solutions of eq .
( [ g0-differential - eq ] ) in terms of the modified bessel functions . the explicit form of the @xmath45 s in the different regions
are sorted out in ref .
and they have been used there to evaluate the integral in the above equation .
we denote this integral by @xmath46 , @xmath47 @xmath48 .
\label{i2cyl}\end{aligned}\ ] ] where @xmath49 and we have used the notation @xmath50 ( same for @xmath51 ) and with index @xmath52 for the tilde s .
on the other hand the fourier transforms , @xmath53 can be written explicitly as @xmath54,\qquad ( \tilde{h}_2)_m = h_2 \frac{\pi}{i}\big[\delta_{m,\nu}-\delta_{m,-\nu } \big],\ ] ] and eq .
( [ e12(2 ) ] ) becomes @xmath55 .
dirichlet limit .
for the case of the dirichlet limit ( @xmath56 ) , eq . takes the form @xmath57}\frac{1}{[\tilde{i}_2 \tilde{k}_1 - \tilde{i}_1 \tilde{k}_2]}.\end{aligned}\ ] ] if we plug this into the above equation for @xmath58 and use eq .
, we find @xmath59 where we have divided by a factor of @xmath60 , which is the mean circumference . weak coupling limit . for the case of weak coupling ( @xmath61 ) the casimir torque per unit area becomes @xmath62,\ ] ] where @xmath63
the calculation of the leading - order contribution in the case of parallel plates with corrugations , is analogous to the case of concentric cylinders and we are not going to give its derivation here . in the case of parallel plates ,
the corrugations are described by the functions @xmath64 $ ] and @xmath65,$ ] , where @xmath66 is the wavenumber corresponding to the corrugation wavelength @xmath67 ( see fig . [ corru ] ) , and @xmath27 is the shift that gives rise to the lateral force . [
cols="^,^ " , ] the next - to - leading - order contribution is calculated from eq .
( [ cecorrug ] ) by keeping terms up to fourth order in the corrugation amplitudes .
the reason is that , as happens for corrugated concentric cylinders , for the case of corrugations with the same wavelength on both plates , the leading order is a second - order contribution while the third - order perturbation terms do not contribute to the energy . therefore , the next - to - leading order is a fourth order term . following the notation introduced in eq .
( [ e12(2)-general ] ) we can write @xmath68 , where in a simplified notation @xmath69 can now be written as , @xmath70\nonumber\\ & + & \big[g^{(0)}v_1^{(1)}g^{(0)}v_2^{(3)}-g^{(0)}v_1^{(1)}g^{(0)}v_1^{(1)}g^{(0)}v_2^{(2)}-g^{(0)}v_1^{(1)}g^{(0)}v_2^{(2)}g^{(0)}v_2^{(1)}\nonumber\\ & + & g^{(0)}v_1^{(1)}g^{(0)}v_2^{(1)}g^{(0)}v_2^{(1)}g^{(0)}v_2^{(1)}\big]\nonumber\\ & + & \big[g^{(0)}v_1^{(3)}g^{(0)}v_2^{(1)}-g^{(0)}v_1^{(2)}g^{(0)}v_2^{(1)}g^{(0)}v_2^{(1)}-g^{(0)}v_1^{(2)}g^{(0)}v_1^{(1)}g^{(0)}v_2^{(1)}\nonumber\\ & + & g^{(0)}v_1^{(1)}g^{(0)}v_1^{(1)}g^{(0)}v_1^{(1)}g^{(0)}v_2^{(1)}\big]=\mathcal{k}^{(2,2)}+\mathcal{k}^{(1,3)}+\mathcal{k}^{(3,1)}. \label{k4th}\end{aligned}\ ] ] the upper - indices in the last line indicate the order of contribution of the potentials @xmath71 and @xmath72 respectively .
we explicitly write one of the above terms to illustrate the procedure .
@xmath73 can be written as @xmath74 where the last term gives the following contribution to the interaction energy , @xmath75 here , similarly to the case of concentric cylinders @xmath76 $ ] and we get @xmath77 from @xmath78 by setting @xmath79 ) and the solutions can be found in appendix a of ref . .
the rest of the terms in eq .
( [ k4th ] ) can be written n the same fashion
. dirichlet limit .
the next - to - leading order contribution to the lateral casimir force in the limit @xmath80 becomes @xmath81\\ & & - 2 \cos ( k_0y_0 ) \,\frac{h_1}{a } \frac{h_2}{a}\left [ \frac{s^2}{\sinh^2 s } \frac{s_-^2}{\sinh^2 s_- } + 2 \frac{s^2}{\tanh^2 s } \frac{s_+}{\sinh s_+ } \frac{s_-}{\sinh s_-}\nonumber \right]\bigg{\ } } , \label{latf-4d}\end{aligned}\ ] ] where @xmath82 , and @xmath83
. weak coupling limit .
if we now take the limit @xmath84 we find that the fourth - order contribution to the lateral casimir force in the weak coupling equals @xmath85 . \label{latf-4w}\end{aligned}\ ] ]
we have used multiple scattering techniques to calculate the leading - order torque between concentric cylinders with corrugations and the next - to - leading - order corrugated parallel plates .
our results for the casimir torque on corrugated cylinders reproduce the results for the lateral force on corrugated parallel plates in the limit of large radii and small corrugation wavelengths , see ref .
and _ lateral casimir forces on parallel plates and concentric cylinders with corrugations _ by the same authors in the procceedings of the seminar `` 60 years of the casimir effect '' .
we thank the us national science foundation ( grant no .
phy-0554926 ) and the us department of energy ( grant no .
de - fg02 - 04er41305 ) for partially funding this research .
icp would like to thank the french national research agency ( anr ) for support through carnot funding .
we especially thank the organizers of the alexander friedmann seminar and of the satellite symposium devoted to 60 years of the casimir effect .
99 o. kenneth and i. klich , _ phys .
lett . _ * 97 * , 160401 ( 2006 ) .
t. emig , n. graham , r. l. jaffe and m. kardar , _ phys .
d _ * 77 * , 025005 ( 2008 ) k. a. milton and j. wagner , _ j. phys . a _ * 41 * , 155402 ( 2008 ) i. cavero - pelez , k. a. milton , p. parashar and k. v. shajesh , _ phys . rev .
d _ * 78 * , 065018 ( 2008 ) t. emig , a. hanke , r. golestanian and m. kardar , _ phys .
a _ * 67 * , 022114 ( 2003 ) .
r. b. rodrigues , p. a. maia neto , a. lambrecht , and s. reynaud , _ phys .
lett . _ * 96 * , 100402 ( 2006 ) . i. cavero - pelez , k. a. milton , p. parashar and k. v. shajesh , _ phys .
d _ * 78 * , 065019 ( 2008 ) | we derive explicit analytic expressions for the lateral force for two different configurations with corrugations , parallel plates and concentric cylinders . by making use of the multiple scattering formalism
, we calculate the force for a scalar field under the influence of a delta - function potential that has sinusoidal dependence in one direction simulating the corrugations . by making a perturbative expansion in the amplitude of the corrugation we find the leading order for the corrugated concentric cylinders and the next - to - leading order for the corrugated parallel plates . |
the @xmath1 detector @xcite at the facility for antiproton and ion research ( fair ) in darmstadt , germany , will provide the ideal experimental setup to study several exclusive final channels in proton - antiproton collisions . also the measurement of the production of heavy hadron pairs emerging from @xmath2-annihilations is planned ,
for which there is thus the need to have theoretical input . in ref .
@xcite we have studied the meson - pair production @xmath3 within a double handbag approach in the same line as the study of the process @xmath4 in ref .
@xcite , where arguments have been given in favor of a generalization of the handbag approach , the validity of which has been demonstrated in the case of deeply virtual compton scattering and meson production .
we argue that , having the heavy @xmath5-quark mass @xmath6 as large intrinsic scale and assuming restricted parton virtualities and intrinsic transverse momenta , the @xmath7 amplitude factorizes into a hard subprocess amplitude and soft hadronic transition matrix elements . neglecting intrinsic proton charm contributions
, the production of the @xmath8-pair can only occur in the partonic subprocess .
the hadronic transition matrix elements are transition distribution amplitudes @xcite , which generalize the concept of generalized parton distributions for @xmath9-quark operators .
we have developed an overlap representation according to ref .
@xcite to model the hadronic transition matrix elements in terms of hadronic light - cone wave functions . in our studies
we only consider the valence fock - state components of the hadrons and we describe the proton as a quark - diquark system , where we only take the scalar diquark into account . having the @xmath7 amplitude at hand we are able to predict differential and integrated @xmath7 cross sections .
scattering amplitude . blobs represent transition distribution amplitudes ( tdas).,scaledwidth=65.0% ] the assignment of particle momenta and helicities can be seen in fig . [
fig : ppbar - to - ddbar ] .
we work in a symmetric center - of - momentum system ( cms ) in which the 3-axis is chosen to be along the 3-vector component @xmath10 of the average momentum @xmath11 . the transverse momentum transfer @xmath12 , where @xmath13 , is symmetrically shared between the incoming and the outgoing hadron momenta . in light - front
coordinates we can write the particle momenta as follows : @xmath14 \,,\quad p^\prime \,=\ , \left [ ( 1-\xi)\bar{p}^+,\ , \frac{m^2+\boldsymbol{\delta}_\perp^2/4}{2(1-\xi)\bar{p}^+},\ , + \frac{{\bm \delta}_\perp}{2 } \right ] \,,\\
q \,=\ , & \left [ \frac{m^2+\boldsymbol{\delta}_\perp^2/4}{2(1+\xi)\bar{p}^+},\ , ( 1+\xi)\bar{p}^+,\ , + \frac{{\bm \delta}_\perp}{2 } \right ] \,,\quad q^\prime \,=\ , \left [ \frac{m^2+\boldsymbol{\delta}_\perp^2/4}{2(1-\xi)\bar{p}^+},\ , ( 1-\xi)\bar{p}^+,\ , -\frac{{\bm \delta}_\perp}{2 } \right ] \ , , \label{eq : particle - momenta } \end{split}\ ] ] where we have introduced the skewness parameter @xmath15 which parameterizes the relative momentum transfer into longitudinal light - cone plus direction .
we consider the process @xmath16 within a perturbative qcd motivated framework where the hadronic amplitude can be split up into a hard partonic subprocess and into soft hadronic matrix elements of partonic field operators .
specifically , we investigate the process @xmath17 in a double handbag mechanism as shown in fig .
[ fig : ppbar - to - ddbar ] .
in such a framework only the minimal number of hadronic constituents which are required to convert the initial @xmath2 into the final @xmath18 pair actively take part in the partonic subprocess . working in a quark - scalar diquark picture for the proton
, we consider the partonic subprocess @xmath19\overline{s[ud ] } \,\to\ , \bar{c}c$ ] .
the remaining partons inside the parent hadrons act as spectators . in order to produce the heavy @xmath8-pair in the partonic subprocess the gluon has to be a highly virtual one .
the @xmath5-quark mass serves as a natural hard scale , allowing us to treat @xmath19\overline{s[ud ] } \,\to\ , \bar{c}c$ ] perturbatively .
thus , the hadronic @xmath17 amplitude can be written as @xmath20}(+z_1/2 ) \mid p : \ , p,\,\mu \rangle \,\tilde{h}(\bar{k}_1,\ , \bar{k}_2)\ , \\ & \times \langle d^0 : \ , q^\prime \mid \mathcal{t } \phi^{s[ud]\,\dagger}(+z_2/2 ) \overline{\psi}^c(-z_2/2 ) \mid \bar{p } : \ , q,\,\nu \rangle \ , , \label{eq : a - priori - dhm - amplitude } \end{split}\ ] ] where we have omitted color and spinor labels for the ease of writing .
@xmath21 denotes the hard scattering kernel of the @xmath19\overline{s[ud ] } \,\to\ , \bar{c}c$ ] subprocess .
the @xmath22 transition is written as @xmath23}(+z_1/2 ) \mid p : \ , p,\,\mu \rangle \ , , \label{eq : p - to - dbar - transition - matrix - element}\ ] ] which is a fourier transform of a hadronic matrix element of a time - ordered , bilocal product of a @xmath5-quark operator and an @xmath19$]-diquark operator ( @xmath24 denotes the time - ordering of the fields ) . in eq .
( [ eq : p - to - dbar - transition - matrix - element ] ) @xmath25}(+z_1/2)$ ] takes out the @xmath19 $ ] diquark that enters the hard subprocess from the proton state @xmath26 at space - time point @xmath27 .
@xmath28 reinserts the produced @xmath29 quark into the remainders of the proton at space - time point @xmath30 , which then gives the final @xmath31 state .
the @xmath32 transition is treated in an analogous way .
given the heavy quark mass as a hard scale and taking into account the physically plausible assumption that the partons are almost on mass - shell and their intrinsic transverse momenta are smaller than a typical hadronic scale of the order of @xmath33 the transverse and minus ( plus ) components of the active ( anti)parton momenta are small as compared to their plus ( minus ) components .
the parton momenta are then approximatly proportional to the hadron momenta and one can perform the integrations over @xmath34 , @xmath35 , @xmath36 and @xmath37 in the convolution integral .
moreover , the field operators in the hadronic matrix elements are forced to have a light - like distance and thus the time ordering of the fields in the hadronic matrix elements can be dropped .
the @xmath38 amplitude then reads @xmath39}(+z_1 ^ -/2 ) \mid p : \ , p,\,\mu \rangle \,\tilde{h}(\bar{k}_1,\ , \bar{k}_2)\ , \\ & \times \langle d^0 : \ , q^\prime \mid \phi^{s[ud]\,\dagger}(+z_2^+/2 ) \overline{\psi}^c(-z_2^+/2 ) \mid \bar{p } : \ , q,\,\nu \rangle \ , , \label{eq : dhm - amplitude } \end{split}\ ] ] with the @xmath22 transition matrix element @xmath40}(+z_1 ^ -/2 ) \mid
p : \ , p,\,\mu \rangle \label{eq : lc - p - to - dbar - transition - matrix - element}\ ] ] and an analogous one for the @xmath41 transition .
using the same projection techniques as in ref .
@xcite we pick out the dominant components of the @xmath5-quark field @xmath42 here @xmath43 contains the dynamically independent components of the @xmath5-quark field operator . with the help of eq .
( [ eq : c - quark - field - operator ] ) and an analogous one for the antiquark the @xmath16 amplitude ( [ eq : dhm - amplitude ] ) becomes @xmath44}(+z_1 ^ -/2 ) \mid p : \
, p,\,\mu \rangle \\ & \times \bar{q}^- \int \frac{d z_2^+}{2\pi } e^{\imath \bar{x}_2 \bar{q}^- z_2^+ } \langle d^0 : \ ,
q^\prime \mid \phi^{s[ud]\,\dagger}(+z_2^+/2 ) \overline{\psi}^c_\text{good}(-z_2^+/2 ) \mid \bar{p } : \ , q,\,\nu \rangle \gamma^- u(k_2^\prime,\,\lambda_2^\prime ) \
, , \label{eq : dhm - amplitude - simplified } \end{split}\ ] ] where we have defined @xmath45 and introduced the average momentum fractions @xmath46 and @xmath47 . in order to make
predictions one has to model the @xmath48 and @xmath41 transition matrix elements .
we will do that by means of an overlap formalism in terms of light - cone wave functions , which has been first developed in ref .
@xcite to represent the generalized parton distributions of the proton .
we consider the proton and the @xmath49 as @xmath50 \
\rangle$ ] and @xmath51 ( bound)states , respectively .
in addition we assume the bound - state wave functions to be pure s - wave , such that the parton helicities have to add up to the total hadron helicity .
thus , the @xmath52 quark inside the proton has to have the same helicity as the proton itself and the @xmath29 and @xmath52 quark inside the @xmath49 have to have opposite helicities .
the corresponding bound - state ( light - cone ) wave functions of the proton and the @xmath49 are denoted by @xmath53 and @xmath54 , respectively .
those wave functions do not depend on the total hadron momentum but only on the relative parton momenta with respect to the parent hadron momentum .
working out the wave - function overlaps as in ref .
@xcite and inserting the resulting @xmath22 and @xmath55 transition matrix elements into eq .
( [ eq : dhm - amplitude - simplified ] ) we obtain @xmath56 quantities with a tilde ( hat ) relate to a frame where the incoming ( outgoing ) hadron has vanishing @xmath57- and @xmath58-momentum component . ) and ( [ eq : dhm - amplitude - peaking - approximation ] ) we correct eqs .
( 56 ) and ( 57 ) in ref .
@xcite by a missing factor 4 .
correspondingly the cross sections given in ref .
@xcite have to be multiplied with a factor 16 . ]
the wave function for the heavy @xmath59-meson is strongly peaked around @xmath60 with respect to its momentum - fraction dependence @xcite .
this behaviour is also reflected in the overlap representation of the hadronic transition matrix elements .
it means that the kinematical regions close to the peak position contribute most to the @xmath61 integrations in eq .
( [ eq : dhm - amplitude - as - overlap ] )
. thus it is justified to replace the momentum fractions appearing in the hard scattering amplitude by the value of the peak position . after doing that it is possible to pull the hard subprocess amplitude out of the convolution integral , which is then rendered solely to an integral over the hadronic transition matrix elements . after applying this peaking approximation
the @xmath16 amplitude simplifies to @xmath62 ^ 2 \ , .
\label{eq : dhm - amplitude - peaking - approximation } \end{split}\ ] ] the hard @xmath19\overline{s[ud ] } \,\to\ , \bar{c}c$ ] amplitudes can now be calculated by applying the usual feynman rules augmented with the feynman rules for diquarks @xcite ; we obtain @xmath63 the form factor @xmath64 accounts for the composite nature of the diquark at the @xmath65-vertex @xcite .
in order to end up with an overlap representation of the @xmath48 transition we have to specify the valence fock state light - cone wave functions for the proton and the @xmath49 . according to refs . @xcite we take @xmath66 as light - cone wave functions for the proton and the @xmath49 , respectively . the light - cone wave function for
the @xmath49 generates the peak around @xmath67 with the help of the mass exponential .
each of the wave functions has two free model parameters , the normalization constant @xmath68 and the transverse size parameter @xmath69 . for the proton we choose @xmath70 and @xmath71 which amounts to @xmath72 and the valence - fock - state probability @xmath73 . for the @xmath49
we take @xmath74 and @xmath75 , which leads to @xmath76 ( cf .
@xcite ) and the valence - fock - state probability @xmath77 . in fig .
[ fig : overlap ] we show results for the overlap integral occurring within the squar- function parametrization introduced above .
the differential @xmath16 cross section reads @xmath78 we show @xmath79 versus @xmath80 ( @xmath81 ) in the left panel of fig .
[ fig : dsdt ] for mandelstam @xmath82 .
the differential cross section is strongly decreasing with increasing @xmath83 .
this behaviour comes from the decrease of the model overlap with increasing cms scattering angle @xmath84 , cf .
[ fig : overlap ] .
its decrease becomes even more pronounced for higher values of mandelstam @xmath85 . in the right panel of fig .
[ fig : dsdt ] we show the integrated cross section @xmath86 versus mandelstam @xmath85 , whose magnitude is in the range of a few @xmath87 .
we have investigated the process @xmath88 within a double handbag approach where the hard scale is given by the heavy @xmath5-quark mass . we have argued that under physically plausible assumptions the @xmath89 amplitude factorizes into a hard subprocess on the partonic level and transition distribution amplitudes . to model the latter we have constructed an overlap representation in terms of hadronic light - cone wave functions .
our predictions for the differential and integrated @xmath90 cross section should now be confronted with future experimental data from the @xmath1 experiment at fair .
b. pire , k. semenov - tian - shansky and l. szymanowski , phys .
d * 82 * , 094030 ( 2010 ) , phys .
d * 84 * , 074014 ( 2011 ) and phys .
b724,99 ( 2013 ) ; j. p. lansberg , b. pire , k. semenov - tian - shansky and l. szymanowski , phys . rev .
d * 85 * , 054021 ( 2012 ) .
m. diehl , t. feldmann , r. jakob and p. kroll , nucl .
b * 596 * ( 2001 ) 33 [ erratum - ibid .
b * 605 * ( 2001 ) 647 ] [ hep - ph/0009255 ] ; s. j. brodsky , m. diehl and d. s. hwang , nucl .
b * 596 * , 99 ( 2001 ) . | we estimate the scattering amplitude of the process @xmath0 within a double - handbag framework where transition distribution amplitudes , calculated through an overlap representation , factorize from a hard subprocess . this process will be measured in the @xmath1 experiment at gsi - fair . |
it is well - known that the einstein - cartan theory represents the simplest extension of general relativity ( gr ) in the presence of the torsion field @xcite .
however , it is a non - renormalizable quantum theory , in spite of describing physics with a very good approximation at the classical level . in view of that
, several investigations followed in order to set up approaches to the gravitational interaction that may prove relevant to our understanding of the quantum gravity problem .
a fruitful approach has been the theory of higher - order curvature invariants , whose actions contain terms quadratic in the curvature , considering also all possible quadratic terms that we can form with torsion , in addition to the einstein - hilbert term
. furthermore , these terms are necessary if we wish to obtain an effective action for quantum gravity at scales close to the planck length @xcite .
this proposal is not yet free from problems .
the main interest in such theories comes , in particular , from cosmology . in this case
, the possibility that the graviton have a small mass to explain the early stages of the universe depends on the particular quantum theory of gravity .
moreover , torsion could , in principle , be associated to any astrophysical scale . in a more contemporary context of non - perturbative approaches to quantum gravity ,
loop quantum gravity arose as another possible quantization method @xcite .
the heartwood of this scheme is the idea that the fundamental variable for quantization is the connection , rather than the metric .
it was proposed by asthtekar @xcite by working only with the self - dual part of the hilbert - palatini action . in this formulation ,
the phase space variable is a complex su(2)-connection .
unfortunately , this approach exhibits a difficult step , namely , to recover the phase space of real general relativity , we must impose reality conditions on the original complex asthtekar formulation of gravity . in order to come over this difficult , barbero and ,
later on immirzi introduced a family of canonical variables , associated to the bi parameter @xcite ( henceforward represented by @xmath0 ) which is actually a canonical theory of gravity in terms of a real su(2)-connection , referred to as asthtekar - barbero connection . in a lagrangian approach ,
the @xmath0- parameter is taken as a new dimensionless parameter introduced with the meaning of a coupling constant of a topological odd - parity term added to the hilbert - palatini action , which is treated in the first - order formalism and yields the so - called holst action @xcite .
this topological term , in gr , does not affect the equations of motion in the absence of torsion since it appears as a factor in a term that vanishes on the mass - shell . although @xmath0 is non - physical at classical level , we find that it becomes an essential parameter at the quantum level , appearing in several contexts , such as , the spectrum of area and volume operators @xcite , in the black hole entropy formula @xcite and in the dynamics provided by the more modern spin foam models @xcite .
it should be noticed that there is another method to quantize gravity known as the quantum regge calculus @xcite , which is @xmath0-independent . for these reasons , over the latest years
, several attempts to fix the @xmath0- parameter have been worked out .
dreyer @xcite , motivated by hod s work @xcite on the asymptotic quasi - normal modes of black holes , proposed a novel way to fix this parameter . in order that the bekenstein - hawking entropy formula and quasi - normal mode agree ,
the bi parameter should be given by @xmath1 .
in a recent paper , considering faddeev s formulation of gravity , khatsymovsky @xcite found out that the elementary area spectrum is proportional to the bi parameter with the value @xmath2 . requiring the validity of the bekenstein - hawking area law relating the two parameters that characterize the quantum states of black hole horizon
, a. majhi et al obtained the following range for @xmath0 : @xmath3 @xcite . in the latest paper that , to our knowledge , addresses this issue ,
the authors have obtained the value @xmath4 , through the calculation of the entropy for an arbitrary non - rotating isolated horizon @xcite .
it is worthy mentioning that all references previously quoted estimate the @xmath0- parameter making use of a non - perturbative approach to qg . in this paper , we follow an alternative route , by studying perturbative qg @xcite in the first - order ( tetrad ) formalism , to understand which is the role of the @xmath0- parameter in the loop corrections to a model discussed in ref .
@xcite , enriched by implementing the holst term , with minimal coupling between fermions and torsion .
this choice is motivated by the fact that , in the einstein - cartan theory , the torsion field becomes relevant only in the presence of fermionic currents and , more recently , coupling fermions to classical gr has attracted a great deal of attention , since that a. perez and c. rovelli @xcite have demonstrated that the minimal coupling to fermions renders gravity sensitive to the @xmath0- parameter @xcite .
we tackle the issue of a 1-loop mass generation mechanism for the fermion as well as the possible influence of the bi parameter on this mechanism .
the outline of this work is the following . in section [ sec:2 ] , we briefly review the main results of the holst action at the lagrangian level and we present a description of the model discussed in ref .
@xcite , as well as our conventions together with the propagators of the model . also ,
since the dirac field couples to the affine connection and the intrinsic spin of fermions acts as a source for torsion , in section [ sec:3 ] , we study an extension of the our model in order to include the minimal coupling between torsion and the dirac spinor .
finally , in section [ sec:4 ] , we cast our discussions and conclusions .
we adopt to work with the first - order formalism , in which configuration space consists of two independent field variables , namely , the space - time @xmath5 gauge spin connection , @xmath6 @xmath7 , and the vielbein , @xmath8 @xmath9 .
this choice is justified for we think that it is a more fundamental approach to gravitation , since it is based on the fundamental ideas of the yang - mills approach .
the gravitational lagrangian ( einstein - cartan term ) together with the holst invariant , where the latter is multiplied by a coupling constant @xmath10 , the bi parameter , can be written as follows : @xmath11 where @xmath12 , @xmath13 denotes the absolute value of the determinant of the co - tetrad , @xmath14 @xmath15 is the field strength associated with the spin connection @xmath6 @xmath16 which is not torsion - free , and @xmath17 denotes the 4-dimensional levi - civita tensor . in pure gravity , @xmath0 does not affect the graviton equation of motion .
this is no longer the case when fermions are present . in this respect ,
the effective dirac action contains an axial current - current interaction whose dependence on @xmath0 becomes non - trivial ; also , this parameter can not take an arbitrary value , so that we have to fix it by calculating some observable effect .
our motivation in this paper is to investigate the possible relevance of the bi parameter at the quantum level , considering the most general parity - preserving lagrangian that describe massive gravitons studied in the ref .
@xcite with an additional holst term , as given below : @xmath18 where @xmath19 , @xmath0 , @xmath20 , @xmath21 , @xmath22 , @xmath23 and @xmath0 are arbitrary dimensionless constants and the others parameters have canonical dimensions given by @xmath24 = \left [ t\right ] = $ ] @xmath25 = 1 $ ] .
the particular choice of the lagrangian above , eq . ( [ 2 ] ) , is justified by the fact that , by suitably choosing special regions in parameter space , it yields a healthy spectrum of excitations , with massive gravitons and no ghosts or tachyons present , despite the presence of higher powers of curvature and torsion terms .
this has been discussed in the works of refs .
other terms with powers of torsion could be considered ; they would however spoil the consistency of the action , in that ghosts or tachyons would be present and would not decouple from the physical sector @xcite .
this is why we restrict ourselves to the terms present in the action ( [ 2 ] ) .
our conventions are : @xmath26@xmath27 @xmath28 and @xmath29 where the greek indices refer to the world manifold and the latin ones stand for the frame indices .
the torsion tensor has 24 independent components in four space - time dimensions and , by considering the irreducible representations of @xmath30 , it may be split into a vector , @xmath31 , an axial vector @xmath32 , and a rank - three tensor @xmath33 , which satisfies the the conditions @xmath34 @xmath35 and @xmath36 . with this in mind , the generic torsion can be expressed as @xcite @xmath37 where @xmath38 @xmath39 @xmath40 is the metric tensor .
the splitting of eq .
( [ 7 ] ) for @xmath41 in @xmath30-irreducible components is simply a group - theoretic decomposition .
so long as a dynamical model is concerned , it is perfectly legitimate to truncate some of the components in rhs of eq .
( [ 7 ] ) if we choose to restrict our considerations to a cartan s restricted geometry , which amounts to taking @xmath42 .
moreover , this irreducible piece and the @xmath31-vector component of the torsion do not couple to the fundamental fermions of the standard model .
this is the reason why we allowed to suppress them and we build up our model in such a way that only the @xmath32-vector acquires dynamical and can be excited by the interaction with the fundamental fermions of the standard model .
next , we perform the split of the vielbein , @xmath43 namely , the minkowski background and a perturbation , @xmath44 @xmath45 , around the background .
in addition , the fluctuation can be decomposed as @xmath46 before we proceed , we recall that the spin connection @xmath47 @xmath48 may be expressed as@xmath49 where @xmath50 @xmath16 is the riemannian part of the spin connection @xmath51 where @xmath52 stands for the rotation coefficients .
@xmath53 are the components of the contorsion tensor , defined by @xmath54 it is worthy noticing that , while the contortion tensor is antisymmetric in the last two indices , the torsion tensor is antisymmetric in the first two indices . using equations ( [ 7 ] ) to ( [ 12 ] ) , the spin connection fluctuation may be put into a more useful form , namely : @xmath55 at this point , it is worthy to stress that in the presence of torsion , the gauss - bonnet term does not correspond any longer to a topological invariant , for it can not be expressed as a total divergence @xcite .
thus , all the quadratic terms in the curvature present in the lagrangian ( [ 2 ] ) must be kept in our considerations .
finally , we can obtain the relevant part of the bilinear expansion of the lagrangian ( [ 2 ] ) . making use of equation ( [ 13 ] ) , after a lengthy but straightforward calculation , we obtain @xmath56 \square s \nonumber \\ & & + v^{\mu}\left [ \frac{7\alpha}{2}-\left ( \frac{7\rho}{6}+\gamma+\xi-3\chi + k+\frac{5\lambda}{12}\right ) \square\right ] \partial_{\mu}s \nonumber \\ & & + v^{\mu}\left [ -\left ( \frac{\rho}{4}+\frac{2\lambda}{9}+\frac{8\xi}% { 9}\right ) \eta_{\mu\nu}\square \right .
\nonumber \\ & & \left . - \frac{4}{3}\left ( \frac{2\rho}{3}% + \gamma-\frac{\xi}{3}+k+3\chi+\frac{\lambda}{3}\right ) \partial_{\mu}\partial_{\nu } \right .
\nonumber \\ & & \left .
+ \left ( \frac{2\alpha}{3}+\frac{2x}{3}-\frac{y}{3}+z\right ) \eta _ { \mu\nu}\right ] v^{\nu } \nonumber \\ & & + w^{\mu}\left [ \left ( \frac{\rho}{2}-\frac{\gamma } { 2}+2\xi-2k\right ) \eta_{\mu\nu}\square^{2 } \right .
\nonumber \\ & & \left .
+ \left ( -\frac{\rho}{2}-\frac{\gamma}{2}-2\xi+2k-\lambda\right ) \square\partial_{\mu}\partial_{\nu } \right .
\nonumber \\ & & \left .
- \frac{5\alpha}{2}\eta_{\mu\nu}% \square+\frac{5\alpha}{2}\partial_{\mu}\partial_{\nu}\right ] \square w^{\nu } \nonumber \\ & & + a^{\mu}\left [ \left ( \rho-\gamma+4\xi-4k-\frac{\lambda}{2}\right ) \square^{2}+2\alpha\eta_{\mu\nu}\square \right .
\nonumber \\ & & \left .
+ 2\alpha\partial_{\mu}\partial_{\nu } + \left ( -\rho+\gamma-4\xi+4k-\frac{5\lambda}{2}\right ) \square \partial_{\mu}\partial\nu\right ] w^{\nu } \nonumber \end{aligned}\ ] ] @xmath57 a^{\nu } \nonumber \\ & & - a^{\mu}\left ( \frac{21\alpha}{8\beta}\right ) \partial_{\mu}s - a^{\mu } \left ( \frac{2\alpha}{\beta}\right ) \eta_{\mu\nu}v^{\nu } .
\label{15}\end{aligned}\ ] ] two important observations are in order .
the @xmath58-field is not sensitive to the bi parameter , so that we decide to truncate it from the decomposition of the metric .
the second observation concerns the fact that , in the next section , we shall consider the minimal fermion coupling in the presence of massive dirac fields .
as well - known , a dirac particle only interacts with the totally antisymmetric part of the torsion , so that we can disregard the vector component of the torsion .
therefore , the lagrangian ( [ 15 ] ) can be written only in terms of the scalar component of the metric and the pseudo - vector component of the torsion .
we understand that only the @xmath59- and @xmath32- fields should contribute effects whenever fermions are introduced .
this is why we truncate the @xmath31- and @xmath58- fields from the lagrangian ( [ 15 ] ) .
thus , @xmath60 the coefficients in the expression ( [ 16 ] ) , except the constant @xmath61 are arbitrary parameters . by setting @xmath62@xmath63 and @xmath64 we get @xmath65 .
so , the lagrangian ( [ 16 ] ) can be written in terms of the new parameters as @xmath66 where @xmath67 @xmath68 if we make use of equation ( [ 18 ] ) , we can immediately obtain the equation of motion for the axial vector field , @xmath32 , which reads @xmath69 with the free parameters of our model chosen according to the relations of eqs .
( [ 17a])-([17c ] ) , we are led to the lagrangian of eq .
( [ 18 ] ) , where it is clear that the vector @xmath32 becomes an auxiliary field : it appears algebraically , so it has no independent dynamics .
therefore , it can be solved by its algebraic equation of motion in terms of the @xmath59 - field , as it will be done below . for the sake of a better comprehension
, we should perhaps justify why the relations of eqs .
( [ 17a])-([17c ] ) are imposed .
the reason behind the choices of eqs .
( [ 17a ] ) and -([17b ] ) is to suppress the unphysical ( ghost - type ) excitation that would show up from @xmath70 and from the symmetric part @xmath71 : the spin - 0 ghost carried by @xmath70 and @xmath72 would spoil the tree - level unitarity of our model . finally , condition ( [ 17c ] ) is assumed in order to exclude from our model a planckian mass that would otherwise be present in the spectrum of excitations of @xmath32 , as the calculation of its propagator reveals .
hence , the lagrangian ( [ 18 ] ) finally takes the form @xmath73 before going on , one remark is in order .
the next step is to read off the feynman propagator for the scalar field @xmath59 . after a direct computation from ( [ 21 ] ) , we attain@xmath74 where @xmath75 since that we are working in scenarios with low scales , up to a few tev , and the mass @xmath76 is at the planck scale , the appelquist - carazzone theorem ensures that the effects of the planckian loops are not detectable in the tev scale .
so , in view of the suppression of the effects of the planckian massive excitation ( @xmath77 ) by powers of @xmath78 , for the sake of our considerations the propagator for the scalar field may be taken as follows : @xmath79 once we have analyzed the spectrum of excitations that arise from the gravity sector , we shall be now concerned with the introduction of fermions and our main goal shall be the computation of the effects of the ( virtual ) gravity models on the fermion self - energy and , in doing so , we shall track the effects the bi parameter contributes to the fermion masses .
over the past years , there has been a great deal of interest in the study of dirac fields in curved space - time .
particularly , it was shown by perez and rovelli [ 26 ] and by freidel , minic and takeuchi [ 1 ] that , whenever minimally coupled , spinor fields present in the einstein - cartan theory with the holst term prevent the latter of being a truly topological action . in presence of fermions
, we have a modification in the structure of the space - time , leading to a non - vanishing torsion field .
therefore , we have a classical - level interpretation of this parameter , which is related with axial - axial current interactions .
now , in order to better understand the effect of the bi parameter in particle physics , let us calculate the corrections to the fermion mass due to this parameter in the model describe in the previous section .
the lagrangian density which corresponds to the dirac equation in curved space - time is @xmath80 the covariant derivative acting on a spinor is@xmath81 \psi , \label{25}%\ ] ] where @xmath82 @xmath83 , @xmath84 , @xmath85 , @xmath86 , @xmath87 , @xmath88 , @xmath89 are the usual flat space - time dirac matrices , @xmath90 is a dimensionless coupling constant and @xmath91 is the mass of the dirac field . according to the discussion in the section [ sec:2 ] , we are working only with the scalar field , @xmath59 , and the pseudo - vector , @xmath32 , so that the spin connection can be written as@xmath92 replacing ( [ 25 ] ) and ( [ 26 ] ) into ( [ 24 ] ) , and making use of the identity @xmath93 we arrive , after some algebra , to the relation @xmath94 using equation ( [ 21 ] ) , together the dirac lagrangian ( [ 21 ] ) , we can obtain the new equation of motion for the auxiliary axial vector field , @xmath32 , which reads@xmath95 here , it is worthwhile to recall that , since @xmath32 appears as an auxiliary field , it is perfectly licit procedure to replace it in the dirac coupled action by its equation of motion ( [ 29a ] ) .
previously , @xmath32 was given by eq .
( [ 20 ] ) . with the coupling to the fermions
, the axial vector condensate @xmath96 contributes and , as it follows below , upon replacement of @xmath32 by eq .
( [ 29a ] ) in the original dirac action , one obtains the following interaction lagrangian : @xmath97 where now a 4-fermion contact appears as a consequence of the non - dynamical character of the @xmath32-axial field . one can observe that the coupling constant of this interaction depends on the bi parameter .
let us consider the feynman rules for the lagrangian ( [ 21 ] ) .
the momenta in this paper are incoming to the vertices .
thus , the fermion - scalar 3-vertex takes the form @xmath98 and the other two vertices that describes the four - fermion interaction are given by@xmath99 and @xmath100 the propagator for the spinor field is given by @xmath101 from these results , we turn into the calculation of the self - energy corrections to the fermion propagator and discuss the mechanism of mass generation for the fermion field , by looking at the pole(s ) of its 1-loop corrected 2-point function for the fermion self - energy graph , we find @xmath102 the result for the divergent part of the 1-loop self - energy of the fermion has the following final form @xmath103 \right .
\nonumber \\ & & \left .
-\frac{3g^{2}m_{0}^{2}}{8}\right\ } i_{4\times 4 } \nonumber \\ & & + \frac{\beta ^{2}}{12\left ( 13\beta ^{2}+49\right ) } \left [ \frac{1}{3}% \left ( 79-\frac{3087g^{2}}{64\beta ^{2}}\right ) p ^{2}\right .
\nonumber \\ & & + \left .
\left ( 1-\frac{441g^{2}}{64\beta ^{2}}\right ) m_{0}^{2}\right ] p_{\mu } \gamma ^{\mu } \nonumber \\ & & + \left . \frac{7
i g\beta } { 48\left ( 13\beta ^{2}+49\right ) } \left ( 19p ^{2}+3m_{0}^{2}\right ) p_{\mu } \gamma ^{\mu } \gamma _ { 5}\right\ } , \label{38}\end{aligned}\ ] ] where @xmath104 is the parameter of dimensional regularization@xmath105 is the cut - off and @xmath106 is an arbitrary mass parameter .
the insertion of this 1-loop graph leads to the following 1-loop corrected lagrangian@xmath107 we are now ready to find the 1-loop corrected fermionic propagator and then the relation between the bare and the physical mass of the fermion in our model .
the practical calculation for that follows the standard scheme and we do not report the details .
the result is the on - shell relation @xmath108 where we have used the simplified notation , @xmath109 , and the fact that @xmath77 .
we omit the full result of eq .
( [ 40 ] ) , which is very lengthy .
we report below only the parts of interest for our actual considerations : @xmath110 + 1\right\ } , \label{41}\ ] ] @xmath111 \right\ } p_{\mu } , \label{42}\ ] ] @xmath112 we could not achieve an analytic expression for the mass , @xmath76 , as function of the @xmath0 parameter from eq .
( [ 40 ] ) .
therefore , the remainder of our calculation will be carried out using maple computer algebra system . in the next section
, we present the results of our numerical analysis of the model we are discussing and compare them with the known results for the bi parameter cited in the introduction .
, with @xmath113 mev , @xmath114 mev , @xmath115 mev .
[ fig:100 ] ] our main interest in this paper has been the study of the effect of the bi parameter on the mass of a fermion in the context of a particular extended theory of gravity with explicit torsion terms . for this purpose ,
we have performed one - loop calculations in this model with gravity and fermions minimally coupled . proceeding this way , and adopting the physical mass of the fermion at the lhc scale
, we expected to obtain an estimate for the bi parameter .
unfortunately , this could not be done , for the equation ( [ 40 ] ) depends on the free parameters , namely , the cut - off @xmath116 , the arbitrary mass parameter @xmath106 and the bare mass of the fermion , @xmath91 . for this reason , we proceed further to adopt an alternative route . as a first step , we fix the parameters @xmath116 , @xmath106 and @xmath91 , choosing an appropriate energy scale , such as the lhc energy range . with these considerations in mind
, we obtain a relation between @xmath76 and @xmath0 .
thus , one can check if the range of @xmath0 , which has currently been reported in the literature , is compatible with the observed mass of the fermion in the energy scale chosen . in the figure [ fig:100 ]
, we cast the results obtained for @xmath117 mev ( corresponding to qcd cut - off ) , @xmath118 mev , @xmath119 , @xmath120 mev and @xmath121 .
as one should expect , the corrections to the fermion mass are extremely small . by inspection of the plot
, we see that the fermion mass increase if the bi parameter lies in the range @xmath1221.185 , reaching its maximum value at @xmath0 = 1.185 . in this interval of values ,
the potential term in dirac interaction lagrangian dominates the kinetic term .
furthermore , when @xmath123 , the fermion mass is finite . in particular , in the range @xmath124 , the physical mass of the fermion is less than his bare mass . on the other hand , for values of @xmath0 greater than 1.185
, the kinetic term dominates the potential term . in conclusion ,
our result is compatible with the known results in the literature for the bi parameter .
the point of view we have tried to convey in this paper is that the present established limits for the bi parameter are compatible with the spectrum of fermions with masses at the lhc scale .
our study shows that the standard - model charged leptons and quarks would not be sensitive to the bi parameter , but we understand that the whole sector of tev - scale massive fermions , such as the charginos and neutralinos , could be a good probe of the bi parameter in the realm of particle physics .
we would like to point out that the model we present here is still too limited .
to our mind , the next immediate step towards a more realistic model , namely , the inclusion of the bi parameter in the framework of supersymmetry ( more specifically , the minimal sugra model ) , would be the right landscape for the investigation we are reporting in this paper .
we are now concentrating efforts in this direction and we expect to report soon our first efforts elsewhere .
d. cocuroci and j. a. helayl - neto express their appreciation to cnpq - brazil and faperj - rio de janeiro for the invaluable financial support .
the authors are also grateful to antnio duarte for his pertinent suggestions on original manuscript of this paper . | in this paper , we analyse a curvature- and torsion - square quantum gravity action with an additional holst term minimally coupled to a massive dirac field in four dimensions . the main purpose here is to try to estimate and compare the value of the barbero - immirzi ( bi ) parameter with its currently known results . to do that , we work out the physical mass of the fermion as a function of this parameter in a perturbative one - loop calculation , assuming the scenario of a physics at the tev - scale .
pacs numbers : : 12.10.-g , 12.38.bx , 04.50.kd . |
several decades after the discovery of quasars , the physical origin of their variability is still not fully understood .
an unambiguous explanation of the intrinsic variations in quasar emission will likely provide the key to understanding the energy supply of quasars .
the most frequently invoked model connects variability of quasars to instabilities arising in the accretion disk around a super - massive black hole ( e.g. @xcite ) . in the starburst model quasar light curves
are a superposition of very frequent supernova explosions in the host galaxy ( e.g. @xcite ) . gravitational microlensing by compact bodies within the host such as normal stars or certain candidates for dark matter also produces observable flux changes ( e.g. ) , but their contribution to the observed level of variability remains uncertain .
the models can be distinguished based on their predictions for the dependence of the slope and the amplitude of the structure function on the time - scale of variability in the quasar rest frame .
recently , individual quasar light curves have been successfully modeled as a damped random walk ( drw ) process ( @xcite , @xcite , @xcite ) .
while the method allows one to study the structure function of individual objects as a function of black hole mass , quasar luminosity , wavelength of observation ( @xcite ) , it also fixes the slope of the power spectrum in the high frequency regime at @xmath5 for every quasar .
moreover , the drw model excludes the possibility that the structure function of positive flux variations differs from the one for negative changes . in this paper
we rely on more traditional statistics using ensemble structure functions
. a recent influx of survey data on variability of tens of thousands of qsos is fueling numerous observational studies and calls for a new look at the proposed scenarios .
the most recent papers favor the disk instability model ( @xcite , @xcite , @xcite ) with the exception of @xcite who find evidence in support of the microlensing model . here , we re - analyze the repeated @xmath0-band imaging of stripe 82 collected by the sdss using difference image photometry and build an ensemble structure function over a wide range of time - scales in order to test the theoretical predictions .
our results show that in their present form none of the considered models explains the behavior of the structure function on all time - scales .
instead , a hybrid model including both a starburst and disk instabilities naturally follows from the data .
the paper is organized as follows . in section [ sec :
data ] we describe the data and a method of extracting light curves .
basic properties of the structure function are given in section [ sec : sf ] . in section [ sec : sf_ensemble ] we build the structure function for the ensemble of quasars .
we discuss the results in section [ sec : discussion ] and a summary is given in section [ sec : summary ] .
for the purpose of this study we reanalyze all available @xmath0-band drift - scan images of the sdss stripe 82 in the final sdss data release , hereafter dr7 @xcite .
quasar light curves were obtained using difference image photometry and a slightly modified version of the photometric software employed by the ogle project @xcite .
the original motivation for difference imaging approach was to facilitate a sensitive search for strongly lensed quasars @xcite .
but the application of the method also returns high quality relative light curves of all sources in a given data set , which we use to study the variability of known quasars .
the full database of stripe 82 contains 303 runs covering nearly 300 deg@xmath6 of sky along the celestial equator ( @xmath7 and @xmath8 ) .
the area was scanned repeatedly during 10 years with some regions observed more than 80 times @xcite .
the sdss imaging system consists of 30 ccd detectors organized in 6 raws ( camcols ) , where a single ccd in each row is taking data in one of the 5 filters @xmath9 .
since there are gaps between camcols , the width of the major stripe is split into north and south strips covered by independent scans .
the sdss database of stripe 82 contains 162 north and 141 south runs .
the spatial and temporal coverage of all runs is shown in figure [ fig : geom ] , where the blue and red lines correspond to the north and south runs .
each run is divided into @xmath10 pix sections called fpc frames with a @xmath11 pix margin on both ends @xcite .
the locations of the beginning and the end of each scan vary from one observation to another .
this arrangement is not compatible with subtracting multiple images of the same area , so the runs must be `` glued '' back together and then subdivided using consistent field boundaries .
the rebinning starts from ra@xmath12 , the first location covered by at least 20 runs ( determined separately for north and south runs ) .
the rearranged frames of @xmath10 pix are written as fits files and the process continues untill we reach ra@xmath13 , beyond which fewer than 20 runs are available .
some of the data in stripe 82 were obtained in non - photometric conditions and in relatively poor seeing ( dr7 ) .
the performance of the difference image analysis method is sensitive to the quality of the images , and therefore our next step is to reject images with poor seeing , high background , and significant background gradients .
a small number of images taken toward the ends of the run have strong background gradients due to the sun or moon rising or setting . occasionally , a problem with the bias level subtraction caused a jump in the count level between the left and right parts of the image corresponding to individual amplifiers on the ccd chip .
such images were removed from the analysis .
a good quality reference image was prepared for each field using 10 images with the best overall seeing , background , and transparency .
this is done with the help of the sextractor program by requiring a median @xmath14 for point sources ( @xmath15 ) and rejecting background maps with obvious problems .
the fields with fewer than 20 remaining images were discarded .
interpolation of images to the same pixel grid was accomplished using approximately 50 stars identified on all images of the field .
the time base line of the stripe 82 data set approaches a decade .
we found that a significant fraction of stars shows a detectable proper motion producing time - variable residuals in the shape of a dipole in the difference images ( cf .
@xcite ) . in order to limit the impact of such problems on the quality of the final photometry
, we selected the reference stars with the proper motion @xmath16 mas yr@xmath17 from the @xcite catalog in the overlap area ( about 50% ) .
the spatial variability of the psf matching kernel was computed using a third - order polynomial fit to model coefficients and a second - order polynomial for the differential background .
again , slow stars from the catalog of @xcite were preferentially used to compute the solution .
the dia light curve is computed by adding difference fluxes measured with the dia package to the flux in the reference image .
object magnitudes were converted to standard sdss @xmath0 magnitudes by adding a median offset between stars identified on the reference image and corresponding objects in the sdss catalog .
the photometric quality is demonstrated in figure [ fig : qa ] showing the magnitude scatter versus baseline magnitude for a typical field . here
, the baseline is the median magnitude and the variability of the source is calculated as the 68-th percentile of absolute deviations from the median .
black points correspond to the sources identified in that field and blue line shows median of the variability of the sources from all fields . in most fields the frame - to - frame accuracy reaches 7 mmag for bright unsaturated sources .
for a significant fraction of stars the photometric scatter is worse than that and limited by the residuals due to proper motions .
the photometric uncertainty is 1% or better for sources with @xmath18 and around 10% for sources with @xmath19 .
there are 52 quasars with @xmath20 .
those were excluded from the structure function analysis section [ sec : sf_ensemble ] ) due to large uncertainties in their light curves . in the next step ,
we match spectroscopically confirmed quasars from dr7 quasar catalog @xcite with our photometric database .
there are 9519 dr7 catalog quasars within the area covered by stripe 82 , and we can build dia light curves for 7614 of them .
there are several reasons why we miss a significant number of quasars .
after exclusion of poor quality images many fields are left with fewer than 20 epochs and are removed from further analysis .
an average 10% of the image area is lost due to masking of very bright stars .
some fields are rejected for lack of stars that can be used to calibrate light curves .
finally , our pipeline occasionally fails on a small fraction of images with problems in fpc frame headers .
an example quasar light curve resulting from this procedure is shown in the figure [ fig : qso_ex ] .
there are 43 observations in the light curve and time counts from the first epoch in stripe 82 ( run 94 , fpc frame 12 ) . the number of epochs per quasar in our sample varies between 20 and 56 with the mean of 36 .
the time sampling of our quasar light curves is relatively sparse ( see figure [ fig : qso_ex ] ) and insufficient for the use of spectral methods such as the frequency - power spectrum analysis . following the common practice in agn variability studies , we focus on the information contained in the structure function .
a number of structure function definitions are in use differing in relatively minor details that are not important for this study .
we adopt a structure function ( hereafter sf ) in the form : @xmath21 where @xmath22 some random quantity at moment @xmath23 , the sum is taken over all measurements for which @xmath24 and @xmath25 is the number of such data points . in short , the structure function is a time dependent standard deviation of some random process .
sf of white noise is flat and equal to @xmath26 .
therefore , the sf of a random process is also flat on time - scales , where the amplitude of the intrinsic variability is comparable with the measurement errors , as well as on the longest timescales where the r.m.s . stops changing as the length of the time interval keeps increasing . in the context of quasar variability
the random quantity is the magnitude ( essentially the relative luminosity ) @xmath27 measured at time @xmath23 and @xmath28 is taken in the quasar rest frame . for a given quasar we can write : @xmath29 where we converted to flux units .
the flux consists of a stationary component @xmath30 and a time dependent variable part @xmath31 , and @xmath32 is the r.m.s . calculated for points separated by time interval @xmath28 .
from here it is obvious that for two quasars at the same redshift and with equal @xmath33 the brightest one will have smaller variability amplitude , @xmath34 , at all time scales . the number of time intervals that can be constructed from an average of 36 epochs for a single quasar is @xmath35 and results in a very noisy structure function .
therefore , we focus our analysis on the ensemble quasar sf , i.e. we assume that the ensemble variability reflects on average the variability of a single quasar .
the final sample contains 7562 light curves .
all possible pairs of measurements within each light curve are collected for a total of 4,796,468 magnitude differences . after reducing the time lag to the quasar rest frame using spectroscopic redshifts from the sdss quasar catalog @xcite
, the data are binned in time .
the bin size was adjusted to reach 20,000 measurements per bin .
the resulting ensemble sf consists of 239 bins and is shown in figure [ fig : sf ] ( blue points ) .
error bars were estimated by dividing each timescale bin into 20 smaller bins , evaluating the mean sf in each , and calculating the r.m.s .
scatter of the resulting values ( see also @xcite ) . in order to verify that we detect quasar variability and guard against the possibility of contamination by unrelated factors , we repeated this calculation for constant stars .
for that purpose we select stars below the magenta line in figure [ fig : qa ] with @xmath36 .
the result is shown in the top panel of figure [ fig : sf ] ( green points ) . as expected
, the sf of non - variable stars is flat over the entire range of time - scales .
we conclude that the observed variability of quasars is intrinsic and not caused by unrecognized measurement errors .
the stellar structure function exceeds the noise level for the quasar sample because the stars selected for this comparison are fainter on average than the quasars .
the quasar sf in the top panel of figure [ fig : sf ] displays several slopes and breaks .
it starts flat at @xmath37 days , then at @xmath38 days gradually transitions to an approximate power - law increase over @xmath39 days , and finally breaks to a shallower slope above @xmath40 .
the flat portion of the sf is clearly due to noise and the transition to the first power - law section is due to sfs of single quasars with different mean brightness @xmath41 reaching the level of noise at different @xmath28 ( see equation [ eq : sf2 ] ) .
the slopes seen at @xmath39 days and @xmath42 have a physical origin and are discussed further below .
the time interval between 40 and 100 days is poorly sampled and difficult to interpret . this is a result of the sdss scanning strategy for stripe 82 that produces relatively few observations separated by 90270 days in the observer frame ( cf .
figure [ fig : qso_ex ] ) combined with the mean redshift of the sample @xmath43 .
the noise contributes to the sf equally at all timescales and must be subtracted in order to expose the intrinsic quasar variability .
the noise - corrected sf is shown in the bottom panel of figure [ fig : sf ] ( blue points ) .
the shape of the quasar sf is dominated by two slopes that we estimate by fitting a broken power law model to the data @xmath44 we find @xmath2 , @xmath3 , @xmath45 days , and @xmath46 ( solid red line ) .
the reduced @xmath47 of the fit is far from unity , i.e. equation [ eq : bpl ] does not provide a complete description of the data . however , the main goal here is to detect and estimate the two slopes in the overall shape of the sf for an ensemble of quasars , and compare those coefficients with theoretical predictions .
we also estimate errors on the derived parameters using jackknife resampling .
namely , we randomly select 1/3 part of the sample , drop it , and fit ensemble sf built from the rest .
we repeat the procedure 100 times and estimate the error on the given parameter as r.m.s . of proper distribution .
such estimation gives very small errors ( 0.01 , 0.02 , 3.9 days , and 0.01 for @xmath48 , @xmath49 , @xmath50 , and @xmath51 correspondingly ) saying that the shape of ensemble sf is quite stable to small variations of the sample . for comparison
we plot the @xmath0-band sf from @xcite ( black points in the bottom panel of figure [ fig : sf ] ) .
the overall agreement in shape between the sfs is very good and the discrepancy in normalization is explained by the different sf type used in @xcite .
there are two types of sf commonly used in literature .
following @xcite we refer to them as sf@xmath52 and sf@xmath53 : @xmath54 the structure function in the form sf@xmath53 was introduced by @xcite .
our noise - corrected sfs and the @xmath0-band sf@xmath53 from @xcite are shown in figure [ fig : sf_a_b ] .
there is a nearly perfect agreement in both shape and normalization between the b - type sfs derived from the two data sets .
the interpretation of results obtained using sf as a statistical tool is not straightforward .
@xcite show that spurious breaks can appear in the sf due to sparse time sampling and windowing effects .
moreover , the slopes derived from sf modeling are typically much more uncertain than suggested by naive error estimates .
does it mean that the presence of two slopes in figure [ fig : sf ] is an artifact ? to answer this question and validate our results we perform a set of monte - carlo simulations .
the goal of this calculation is to verify that for a set of light curves generated from a single power - law spectrum with a given value of @xmath55 and having the time sampling of real quasars , the algorithm of computing sf does not introduce any significant breaks or biases in the estimated slope .
the simulation is based on the algorithm of and reproduces both the correct power spectrum as well as the phase mixture .
the power spectrum in the form @xmath56 corresponds to sf @xmath57 .
we simulate light curves according to a prescription in @xcite .
the simulated time series are then normalized to have the same r.m.s . and mean as the real light curves .
examples are shown in the figure [ fig : sim_lc ] .
we simulate two sets of light curves with the power - law spectral density ( @xmath58 and @xmath59 ) , select points on the simulated light curves corresponding to the time sampling of a real light curve , and apply proper rest frame correction . in figure
[ fig : sim_slopes ] the simulated sfs are compared to the best fit model for real quasars .
the slopes of the simulated sfs do not exactly match the slope of the model from which they were generated , i.e. the time sampling slightly biases the measured shape of the sf .
however , none of the simulated sfs is even close to a broken power law with @xmath60 and @xmath61 .
therefore , the large observed difference in the sf slope between the short and long timescales is not due to the time sampling , but rather of a physical origin .
possible interpretations are discussed in the next section .
the ensemble sf considered here is built from quasars with different intrinsic characteristics such as black hole mass or luminosity .
therefore , it is interesting to check whether two slopes in sf are characteristic for the hole range of , say , black hole masses or have dependence on mass . using black hole mass estimations from @xcite , we select three subsamples of quasars with black hole masses in ranges @xmath62 ( 1953 sources , low mass range ) , @xmath63 ( 2108 sources , mean mass range ) , @xmath64 ( 1276 sources , high mass range ) and build sf for each range of masses .
all three sfs show existence of two slopes ( see figure [ fig : sf_mbh ] , top panel ) .
we fit sfs by broken power low model ( equation [ eq : bpl ] ) .
the best fit parameters and formal estimation of errors by jackknife resampling are given in the table [ tab : fit_par ] . in order to have better understanding of the uncertainties in the parameters we also plot regions showing minimal and maximal values of the fits for all three sfs ( see figure [ fig : sf_mbh ] , bottom panel ) . as it is seen from the figure [ fig : sf_mbh ] ,
the sfs slopes tend to increase with the growth of the mean subsample mass but still have common values .
more subtle study is needed in order to show that the trend in slopes as a function of black hole mass is real . here
we only conclude that the presence of two slopes in the ensemble sf is independent on the considered black hole mass range .
.best fit parameters [ cols="<,^,^,^,^",options="header " , ]
we consider three processes frequently invoked to explain the variability of quasars : instabilities in the accretion flow around a supermassive black hole , supernova explosions related to the starburst phenomenon , and gravitational microlensing by compact bodies in the host galaxy @xcite .
sfs of light curves predicted by each model are characterized by different slopes .
the expected sf slopes are @xmath65 , @xmath66 , @xmath67 respectively for variability generated by miscrolensing , disk instability , and starburst mechanisms @xcite .
the sf slopes found in section [ sec : sf_ensemble ] are 0.79 and 0.33 correspondingly for time lags below and above @xmath68 days . taken at face value , the first slope is consistent with the starburst model , and the second slope falls half way between the predictions of the disk instability model and the microlensing model .
note that sf slopes in the range 0.300.35 have been obtained in other works , but the authors still attributed the variability to disk instabilities . @xcite and present an overview of recent results .
asymmetries in the sf provide an additional test of model predictions .
the sf that only includes those pairs of measurements for which the flux increases with time , @xmath69 , may be different from @xmath70 , the sf characterizing a decrease in flux .
@xcite show that in starburst models @xmath71 .
disk instabilities produce @xmath72 and microlensing is symmetric , i.e. @xmath73 when averaged over sufficiently long time intervals @xcite .
the difference between @xmath70 and @xmath69 depends on the parameters of the model and therefore @xmath74 is still possible in both the starburst and the disk instability model @xcite .
noise - corrected @xmath69 and @xmath70 functions are shown in figure [ fig : sf_plus_minus ] with blue and green lines respectively . for long time lags in the range 3001600 days , where the sf slope is @xmath75
, we have @xmath76 , as predicted by the disk instability model .
the significance of the difference is not very high , but allows us to exclude the microlensing model .
the difference disappears for time - scales below 100 days , where the slope is consistent with the starburst model .
this may suggest that a typical supernova rate in a starbursts in the host galaxies of quasars is as high as @xmath77 yr@xmath17 and effectively washes out any detectable asymmetry ( see fig . 6 in @xcite ) .
however , the rate should be treated with caution since it is not derived from direct measurements .
moreover , the predictions for @xmath69 and @xmath70 were derived from simulations of a single variability mechanism .
the addition of another variability process can wash out the difference between @xmath69 and @xmath70 .
this is an interesting topic for future research .
the difference in @xmath69 and @xmath70 was also investigated by @xcite and @xcite .
the former study found tentative evidence that @xmath78 ( light curves have fast rise and slow decay as in starburst model ) for 400 days @xmath79 1500 days , while in the latter work @xmath80 over the full range of measurements ( @xmath81 days ) .
the explanation of the difference in results is difficult because each paper is based on entirely different data and requires further research .
the @xcite analysis relies on cross survey flux comparisons .
the introduction of the supernova variability mechanism may be redundant as the shape of the sf can be explained by detailed variability mechanisms acting in accretion disks .
for instance , the psd of x - ray light curves of some active galactic nuclei have two ( e.g. @xcite ) or more distinct slopes @xcite . considering that the optical emission tends to follow the x - ray flux with some time delay
one may expect that psds of optical light curves will display two slopes .
moreover , the sf shape found here is in qualitative agreement with the model proposed in @xcite , where large amplitude optical variations on the time scales of hundreds of days are attributed to the fluctuations in accretion rate and short - term ( order of days ) , small amplitude variations arise due to reprocessing of x - rays .
we applied the image differencing technique to the entire @xmath0-band imaging data set of sdss stripe 82 and constructed high quality light curves for 7562 spectroscopically identified quasars with no less then 20 epochs .
the variability analysis was performed using the ensemble structure function .
while the shape and normalization of the quasar structure function derived here are in perfect agreement with the results of @xcite based on an earlier sdss data release , our structure function is estimated from a much larger data set and covers a substantially wider range of time lags from 8 hours to 6.9 years in the quasar rest frame .
we find that the ensemble sf reveals two distinct power - law slopes with the break around @xmath68 days and confirm the presence of these features with monte - carlo simulations .
the presence of two slopes in the ensemble sf is independent on the black hole mass range of quasars used to build it .
the slopes estimated from a broken pawer - law fit to the data are @xmath82 for @xmath83 days and @xmath84 for @xmath85 days . using predictions from the theoretical models ,
the variability of quasars on times scales @xmath83 days can be explained by a starburst model .
the slope of the structure function on time - scales longer than about 3 months is rather close to models where variability is explained by gravitational microlensing .
however , on time - scales @xmath85 days we detect a significant asymmetry in the sf characteristic for disk instabilities that rules out a substantial contribution of microlensing .
it is also possible to explain the shape of sf without invoking the starburst model , but using the model , where short term optical variations are caused by x - rays reprocessing and long therm variations originate from optical emitting regions of accretion disk @xcite .
the existing sdss stripe 82 data offer a great opportunity to better understand the origin of variability in quasars .
the main limiting factor at present is the lack of detailed theoretical predictions that go beyond the basic characteristics such as the sf slope and include the detailed shape and normalization of the sf , as well as asymmetries and color information .
av thanks przemysaw woniak for his invaluable contribution to the preparation of the paper , eugene churazov for fruitful discussions on the statistics of the structure function and acknowledges support from the research grant rffi 10 - 02 - 01442-a .
funding for the sdss and sdss - ii has been provided by the alfred p. sloan foundation , the participating institutions , the national science foundation , the u.s .
department of energy , the national aeronautics and space administration , the japanese monbukagakusho , the max planck society , and the higher education funding council for england .
the sdss web site is http://www.sdss.org/. the sdss is managed by the astrophysical research consortium for the participating institutions .
the participating institutions are the american museum of natural history , astrophysical institute potsdam , university of basel , university of cambridge , case western reserve university , university of chicago , drexel university , fermilab , the institute for advanced study , the japan participation group , johns hopkins university , the joint institute for nuclear astrophysics , the kavli institute for particle astrophysics and cosmology , the korean scientist group , the chinese academy of sciences ( lamost ) , los alamos national laboratory , the max - planck - institute for astronomy ( mpia ) , the max - planck - institute for astrophysics ( mpa ) , new mexico state university , ohio state university , university of pittsburgh , university of portsmouth , princeton university , the united states naval observatory , and the university of washington . | theoretical predictions for the ensemble quasar structure function are tested using multi - epoch observations of stripe 82 collected by the sloan digital sky survey .
we reanalyze the entire available volume of the @xmath0-band imaging data using difference image photometry and build high quality light curves for 7562 spectroscopically confirmed quasars .
our structure function includes @xmath1 pairs of measurements and covers a wide range of time lags between 3 days and 6.9 years in the quasar rest frame
. a broken power - law fit to this data shows the presence of two slopes @xmath2 and @xmath3 with the break at @xmath4 days .
the structure function compiled using only flux increases is slightly lower than that for variations of the opposite sign , revealing a slight asymmetry between the leading and trailing edge of a typical flare .
the reality of these features is confirmed with monte - carlo simulations .
we give simple interpretation of the results in the frames of existing theoretical models . |
statistical physics furnishes diverse tools for the study of human social dynamics . despite the apparent dissimilarities between social and physical systems , many social processes present a phenomenology that resembles that found , for instance , in the physics of frustrated or disordered materials @xcite
this is because , despite the high heterogeneity of the individuals , and their interactions , not all details are relevant for the emergence of collective patterns .
collective behaviors make social systems interesting for the physicist and , reciprocally , the physicist might contribute with a new perspective to the comprehension of social phenomena .
one of the basic ingredients to be taken into account for modeling people s interactions is imitation , or social contagion .
in fact , imitation is observed in diverse social contexts , such as in the dynamics of language learning or decision making .
the recurrent conformity to the attitudes , opinions or decisions of other individuals , or groups of individuals , has led to the formulation of models based on social contagion as the primary rule of opinion dynamics , e.g. , the voter @xcite , sznajd @xcite and majority rule @xcite models , to give just a few examples .
however , aside from imitation , individuals also dissent and resist to be influenced , in several ways . in the present work , we study the effect of nonconformity attitudes through a kinetic exchange model @xcite . within this modeling ,
the influences that an individual exerts over another are modulated by a coupling strength .
the coupling strength between connected individuals typically takes positive values but also , with certain probability , it can adopt negative ones .
negative values represent negative influences that , instead of imitation , promote dissent .
notice that this kind of dissent is not a characteristic of the individual but of the tie ( or link ) between each pair of individuals .
there are also other types of nonconformity , which are not associated to the links but to the individuals ( or nodes in the network of contacts ) , as taken into account in several models @xcite .
one of these types is anticonformity @xcite .
an anticonformist actively dissents from other people s opinions , which is the case contemplated , for example , by galam s contrarians @xcite . actually , although these anticonformists defy other people s opinions or the group norm , they are similar to conformists in the sense that they take into account other s opinions too .
a different kind of nonconformity is independence , such that the individual tends to resist the influences of other agents or groups of agents , ignoring their choices in the adoption process .
it can be thought either as an attribute of the agent , that acts as independent with certain probability , or an attitude that any individual can assume with certain frequency .
it is in the latter sense that we will incorporate independence into the model .
another important ingredient for modeling opinion dynamics is conviction or persuasion ( a kind of stubbornness or resistance to change mind ) @xcite .
we will consider the heterogeneity of the individuals in that respect .
each agent will be characterized by a parameter that measures its level of conviction about the opinion it supports .
this parameter , typically defined positive , will be allowed to take also negative values to represent volatile individuals that change mind easily .
heterogeneities and disorder can act on opinion dynamics as stochastic drivings able to promote a phase transition , playing the role of a source of randomness or noise similar to a social temperature @xcite .
we will focus on the impact of all the abovementioned sources of disorder on the steady state distribution of opinions in the population , and investigate the occurrence of nonequilibrium phase transitions .
our model belongs to the class of kinetic exchange opinions models @xcite .
we consider a fully - connected population of size @xmath1 participating in a public debate .
each agent @xmath2 in this artificial society has an opinion @xmath3 .
most models of opinion dynamics deal with a discrete state space @xcite , which may be enough to tackle certain problems that involve binary or several particular choices .
however , to investigate , for instance , the emergence of extreme opinions , it seems to be more suitable to represent opinions by means of a continuous variable , to reflect the possible shades of peoples attitudes about a given subject @xcite .
then , we will consider a continuous state space , where opinions can take values in the real range @xmath0 $ ] .
positive ( negative ) values indicate that the position is favorable ( unfavorable ) to the topic under discussion .
opinions tending to @xmath4 and @xmath5 indicate extremist individuals .
finally , opinions near @xmath6 mean neutral or undecided agents . at a given step @xmath7 , the following microscopic rules control the opinion dynamics : 1 .
we choose two random agents @xmath2 and @xmath8 .
2 . with probability @xmath9 ,
the agent @xmath2 acts as independent . in this case
, the agent chooses a position about the subject under discussion , i.e. , @xmath10 is randomly selected from a uniform distribution @xmath0 $ ] , independently of the current states @xmath11 and @xmath12 .
that is , as commented above , we consider that independence is not an attribute of the individual but an attitude that any individual can occasionally assume with certain probability . in that opportunity
, individuals choose their own position ( state ) independently of the other individuals @xcite .
3 . otherwise , with probability @xmath13 , the agent @xmath2 acts as a partially conformist individual , and will be influenced by agent @xmath8 by means of a kinetic exchange . in this case , the opinion of agent @xmath2 will be updated according to the rule @xcite @xmath14 where @xmath15 and @xmath16 are real parameters that measure the level of conviction of agent @xmath2 , and the strength of the influence that agent @xmath2 suffers from agent @xmath8 , respectively .
the opinions are restricted to the range @xmath0 $ ] .
therefore , whenever eq .
( [ eq1 ] ) yields @xmath17 ( @xmath18 ) , it will , actually , lead the opinion to the extreme @xmath19 ( @xmath20 ) , as considered in refs . @xcite .
this re - injection of the opinions into the interval @xmath0 $ ] introduces a nonlinearity in the mapping given by eq .
( [ eq1 ] ) , that becomes linear by parts . smoothing this nonlinearity , for instance through
@xmath21 $ ] , is not expected to affect the results significantly , as will be discussed below in connection with the outcomes of simulations .
+ typically , the strength @xmath16 is positive , meaning agreement , but it can also take negative values to represent ties that induce dissent .
the fraction of negative influences is given by a parameter @xmath22 , such that @xmath16 is uniformly distributed either in @xmath23 $ ] or in @xmath24 $ ] , with probabilities @xmath22 and @xmath25 , respectively .
+ parameter @xmath15 is also typically positive , representing the weight or contribution that the current opinion of agent @xmath2 has , with respect to that of other agents , in the opinion formation process .
the larger @xmath15 , the less other individuals contribute to mold the opinion of agent @xmath2 .
sufficiently large values of @xmath15 ( larger than 1 for the current choice of the set @xmath16 ) mimic intransigence or stubbornness , such that the opinion of agent @xmath2 will not evolve .
parameter @xmath15 can also take negative values , to represent volatile individuals that are not persistent in their positions , that is , have a propensity to change mind .
this spontaneous change must be distinguished from the change of opinion in independent behavior described above .
due to the minus sign , an agent with negative conviction will always change side , adopting a new opinion which is the opposite of the previous one .
differently , in the independent case , since the new opinion is chosen at random , the individual opinion will not necessarily change sign but can take any value in the spectrum .
+ the fraction of negative convictions will be controlled by a parameter @xmath26 , such that @xmath27 is uniformly distributed either in @xmath23 $ ] or in @xmath24 $ ] , with probabilities @xmath26 and @xmath28 , respectively .
the unit of time is defined by the application of the above - mentioned steps @xmath1 times . concerning the random nature of the convictions @xmath15 and the couplings @xmath16 , which are assumed to be uncorrelated , we consider quenched ( frozen ) variables , as far as opinion formation supposedly occurs in a time scale much faster than the changes in agents attributes and couplings
. we will consider populations of @xmath1 fully - connected individuals , situation which corresponds to a mean - field limit .
the initial state of the system is assumed to be fully disordered , that is , at the beginning of the dynamics , each individual has an opinion drawn from the uniform distribution in the range @xmath0 $ ] . in the following
, we will describe separately two distinct cases characterized by : ( i ) the existence of competitive positive / negative interactions among pairs of agents , while convictions are homogeneous ( with @xmath29 , @xmath30 ) , and ( ii ) the heterogeneity of agents convictions , which interactions are always positive ( @xmath31 , for all @xmath32 ) . in both cases ,
the noise introduced by independent attitudes is included .
we calculate the parameter @xmath33 given by @xmath34 where @xmath35 denotes average over disorder or configurations , computed at the steady states .
notice that @xmath33 is a kind of order parameter that plays the role of the `` magnetization per spin '' in magnetic systems .
it is sensitive to the unbalance between positive and negative opinions .
a state with a large value of @xmath33 ( @xmath36 ) means that an extremist position reached consensus .
intermediate values indicate : ( i ) the dominance of either one of the extreme opinions , ( ii ) that opinions are moderate but one of the sides wins , or ( iii ) a combination of both .
all these states can be identified with ordered ones in the sense that the debate has a clear result ( favorable or unfavorable ) , be extremist or moderate .
a small value ( @xmath37 ) indicates a symmetric distribution of opinions : ( i ) polarization such that opposite opinions balance , ( ii ) the dominance of very moderate or undecided opinions around the neutral state , or ( iii ) a combination of both . in all symmetric cases ,
the debate will not have a clear winner position . in that sense
the collective state can be identified with a disordered one .
we consider also the fluctuations @xmath38 ( or `` susceptibility '' ) of parameter @xmath33 , @xmath39 and the binder cumulant @xmath40 @xcite , defined as @xmath41 all these quantities will be used to characterize the phase transitions between ordered and disordered phases .
additionally , those phases will be described by means of the pattern that the distribution of opinions presents .
let us focus first on the effect of competitive interactions ( with a fraction @xmath22 of negative couplings ) and independence ( which occurs with probability @xmath42 ) , by studying the case of homogeneous agents with @xmath29 for all @xmath2 .
in figure [ fig1 ] we exhibit @xmath33 as a function of the independence parameter @xmath9 , for different values of @xmath22 .
as observed in the figure , independence makes the system undergo a phase transition , an effect which is typical of social temperatures @xcite . versus the independence probability @xmath9 , for several values of the fraction @xmath22 of negative interaction strengths , with @xmath29 for all @xmath2 .
one can observe transitions at different critical points @xmath43 , but the maximal value of @xmath33 decreases with @xmath22 and the transition is eliminated for sufficiently large values of @xmath22 .
the population size is @xmath44 and data are averaged over @xmath45 simulations.,scaledwidth=30.0% ] the transition occurs even when couplings are all positive ( @xmath46 ) , then the mere presence of independence leads the system to undergo a phase transition .
a similar effect was observed in the discrete version of the present model , considering only opinions @xmath47 , @xmath6 , @xmath4 , although the critical point @xmath48 has a different value ( 1/4 in the discrete case @xcite ) .
when we introduce negative interaction strengths ( @xmath49 ) , states characterized by @xmath50 , meaning consensus of extreme opinions , become unlikely ( see figure [ fig1 ] ) .
moreover , the critical values @xmath51 decrease with @xmath22 and , above sufficiently large @xmath22 , the system becomes disordered for any @xmath9 .
the critical value @xmath52 is in accord with that found in ref .
@xcite . to estimate the critical points @xmath43
, we performed a finite - size scaling ( fss ) analysis .
as a typical example , we exhibit in figure [ fig2 ] the results of fss for @xmath53 .
the critical exponents are @xmath54 , @xmath55 and @xmath56 .
thus , as expected , the universality class known for the mean - field implementation of the model when @xmath57 @xcite is not altered by the presence of the microscopic disorder introduced by independence .
, for @xmath53 and @xmath29 for all @xmath2 .
the best data collapse was found for @xmath58 , @xmath54 , @xmath55 and @xmath56.,title="fig:",scaledwidth=30.0% ] , for @xmath53 and @xmath29 for all @xmath2 .
the best data collapse was found for @xmath58 , @xmath54 , @xmath55 and @xmath56.,title="fig:",scaledwidth=30.0% ] + , for @xmath53 and @xmath29 for all @xmath2 .
the best data collapse was found for @xmath58 , @xmath54 , @xmath55 and @xmath56.,title="fig:",scaledwidth=30.0% ] the fss analysis also provides the critical values @xmath43 , that allow to build a phase diagram in the plane @xmath9 versus @xmath22 , depicted in figure [ fig3 ] .
the boundary separates the ordered and disordered phases . in the ordered phase , for @xmath59 , one of the sides ( positive or negative opinions ) will win the debate , while in the disordered phase , there will be balance of opposite opinions and/or dominance of moderate ones . following the analytical expression found for the discrete kinetic exchange opinion models @xcite , namely a quotient of two first order polynomials in @xmath22 , we propose a qualitative description of the phase boundary through @xmath60 which for fitting parameters @xmath61 gives a heuristic description of the phase boundary , in good agreement with the critical points obtained from fss analysis , as shown in figure [ fig3 ] .
the extreme critical values @xmath62 and @xmath63 are given respectively by @xmath64 and @xmath65 .
the latter value is in agreement with the estimate found in ref .
@xcite , which deals with such limiting case .
( probability of independence ) versus @xmath22 ( probability of negative interactions ) , when @xmath66 .
the symbols are the numerical estimates of the transition points , whereas the dashed line is given by eq .
( [ eq5 ] ) .
the error bars were estimated from the fss analysis . , scaledwidth=30.0% ] ( top , left pannel ) , @xmath67 ( top , right pannel ) and @xmath68 ( bottom ) , for several values of @xmath22 , when @xmath66 .
empty ( full ) symbols represent ordered ( disordered ) states . in the insets
we show a zoom of the main frames , excluding the extremist agents with majority opinions .
the population size is @xmath44 , and data are accumulated over @xmath45 independent simulations , as explained in the text .
, title="fig:",scaledwidth=30.0% ] ( top , left pannel ) , @xmath67 ( top , right pannel ) and @xmath68 ( bottom ) , for several values of @xmath22 , when @xmath66 .
empty ( full ) symbols represent ordered ( disordered ) states . in the insets
we show a zoom of the main frames , excluding the extremist agents with majority opinions .
the population size is @xmath44 , and data are accumulated over @xmath45 independent simulations , as explained in the text .
, title="fig:",scaledwidth=30.0% ] + ( top , left pannel ) , @xmath67 ( top , right pannel ) and @xmath68 ( bottom ) , for several values of @xmath22 , when @xmath66 .
empty ( full ) symbols represent ordered ( disordered ) states . in the insets
we show a zoom of the main frames , excluding the extremist agents with majority opinions .
the population size is @xmath44 , and data are accumulated over @xmath45 independent simulations , as explained in the text .
, title="fig:",scaledwidth=30.0% ] since there is ambiguity in the interpretation of the values of parameter @xmath33 , we complemented the previous analysis , computing the distribution of opinions in the population at the steady states of the model . in figure [ fig4 ]
, we exhibit the outcomes for @xmath57 , @xmath67 and @xmath68 , at several values of @xmath22 .
each normalized histogram is obtained from @xmath45 independent simulations , for population size @xmath44 .
when there is unbalance of positive and negative opinions , we arbitrarily selected simulations with dominance of positive opinions to build the histograms . in this way , each histogram is representative of each single realization but with improved statistics . instead , if we would have chosen the simulations with predominantly negative outcomes , the distribution would be the symmetric counterpart of those shown in figure [ fig4 ] .
first , we notice that there is condensation of opinions at the extreme values @xmath69 .
this is a consequence of the type of nonlinearity introduced in eq .
( [ eq1 ] ) .
if a smoother form were used instead , then there will be a spread of the extreme values in the condensates , which will not affect significantly the number of individuals that can be identified as those who are extremists . for each frame ( fixed value of @xmath9 ) , we observe the following scenario . both the winner and loser sides present a bulk of moderate individuals as well as a condensation at @xmath70 ( extremists ) . when @xmath22 is small enough so that the system becomes ordered ,
unbalance between positive and negative opinions emerges , consistently with a non null order parameter .
when @xmath22 becomes too large , it is not possible the formation of an ordered phase , where one of the sides dominates , then the distribution becomes flat , with the coexistence of all opinions ( except for the symmetrical condensation in the extreme states @xmath69 ) , as can be seen in the insets of figure [ fig4 ] .
moreover , we see that the number of extremists ( opinions @xmath69 ) decreases with increasing values of @xmath22 , as well as with increasing @xmath9
. that is , the inclusion of dissent interactions , as well as the growth of independent attitudes , inhibit extremism growth , leading to a more homogeneous distribution of opinions , which is associated to the disordered phase in the diagram of figure [ fig4 ] .
in this section we consider independence together with the heterogeneity of the convictions @xmath15 , which can take negative values .
recall that we assume that @xmath27 are ( quenched ) random variables that are uniformly distributed in @xmath23 $ ] with probability @xmath26 and uniformly distributed in @xmath24 $ ] with the complementary probability @xmath28 .
hence , the parameter @xmath26 denotes the fraction of negative convictions .
positive convictions mean that the agent is sure about the adopted belief and therefore contributes in eq .
( [ eq1 ] ) to maintain the current opinion .
a conviction close to zero reflects uncertainty about the belief and high susceptibility to other people s opinions ( conformist behavior ) , while a negative conviction contributes to form an indifferent or an opposite position to the current one . when all the couplings are @xmath71 , the curves for the order parameter @xmath33 vs @xmath9 , are qualitatively similar to those shown in figure [ fig1 ] ( after the substitution @xmath72 ) , although the critical values are not the same . moreover the distribution of opinions are also similar .
hence the cases ( i ) @xmath29 for all @xmath2 , and variable @xmath22 , and ( ii ) @xmath71 for all @xmath32 with variable @xmath26 , are in some way equivalent . since case ( ii ) does not present a new phenomenology with respect to case ( i ) , then we will exhibit , instead of @xmath71 , the results for a more realistic case , where all the interactions are positive but random , uniformly distributed in the range @xmath24 $ ] , which corresponds to @xmath46 .
in figure [ fig5 ] we exhibit the behavior of the order parameter as a function of @xmath9 for several values of @xmath26 .
although the curves are different from those in figure [ fig1 ] , the effect of convictions is rather similar to that introduced by negative interactions , in the sense that the increase of @xmath26 , as well as the increase of @xmath22 , both tend to disorder the system .
however , the population does not reach the state @xmath50 anymore , differently to the case @xmath46 of figure [ fig1 ] . versus the independence probability @xmath9 , for several values of the fraction @xmath26 of negative convictions , when @xmath46 .
one can observe order - disorder transitions at different points @xmath73 , but the transition is suppressed for sufficiently large values of @xmath26 .
the population size is @xmath44 and data are averaged over @xmath45 simulations .
, scaledwidth=30.0% ] in addition , one observes that the critical values @xmath73 are much smaller than in sec .
[ case1 ] , where @xmath74 for all @xmath2 but negative pairwise interactions @xmath16 ( i.e. , @xmath49 ) were allowed .
we estimated the values of the critical points @xmath73 by means of the crossing of the binder cumulant curves , as well as the critical exponents , as illustrated in figure [ fig6 ] for the case @xmath75 .
these exponents are the same as the ones of the previous section , namely @xmath54 , @xmath55 and @xmath56 .
thus , the universality class of the model is not changed by the presence of disorder in the convictions , as expected . and @xmath76 .
the best data collapse was found for @xmath77 , @xmath54 , @xmath55 and @xmath56.,title="fig:",scaledwidth=30.0% ] and @xmath76 .
the best data collapse was found for @xmath77 , @xmath54 , @xmath55 and @xmath56.,title="fig:",scaledwidth=30.0% ] + and @xmath76 .
the best data collapse was found for @xmath77 , @xmath54 , @xmath55 and @xmath56.,title="fig:",scaledwidth=30.0% ] figure [ fig7 ] shows the resulting phase diagram in the plane @xmath9 versus @xmath26 , using the values of @xmath73 obtained from fss .
the boundary separates the ordered and disordered phases . following the same analysis of sec .
[ case1 ] , we estimated the qualitative behavior of the order - disorder frontier by means of eq .
( [ eq5 ] ) , with the change @xmath78 .
the result is plotted in figure [ fig7 ] together with the critical points obtained from fss analysis .
one can see a good agreement between the data and the qualitative frontier . in this case ,
@xmath79 and the extreme critical values are @xmath80 and @xmath81 .
( probability of independence ) versus @xmath26 ( probability of negative convictions ) , when @xmath46 .
the symbols are the numerical estimates of the transition points , whereas the dashed line is obtained by substituting @xmath22 by @xmath26 in eq .
( [ eq5 ] ) , as explained in the text .
the error bars were estimated from the fss analysis.,scaledwidth=30.0% ] to understand the nature of the disordered phase , also in this case we obtained the distribution of opinions in the steady state as done in sec .
[ case1 ] . in figure [ fig8 ]
we show the normalized histograms for several values of @xmath9 and @xmath26 .
each histogram is built from @xmath45 independent simulations , for population size @xmath44 , as in sec .
[ case1 ] .
notice that the distributions are completely different from those in figure [ fig4 ] , both when the system is in the ordered and disordered phases . for values of the parameters in the disordered phase ,
i.e. at the right of critical line in figure [ fig7 ] , the plot is not flat in the center , as it was in the case of figure [ fig4 ] , but the distribution becomes symmetrically concentrated around the neutral state @xmath82 where there is a peak .
that is , the population becomes essentially neutral and opposite opinions are balanced .
when we move away to the left of the critical line in figure [ fig8 ] , the distribution of opinions loses symmetry , in such a way that one of the sides of the debate ( the negative opinions in the example of the figure ) tends to disappear , while the opposite ( positive ) side tends to become uniform , except for the condensation peak at @xmath83 , that is the distribution becomes bimodal .
this effect is more pronounced in the absence of negative convictions ( @xmath84 ) , which represents the farther points from the frontier , at fixed @xmath9 . for increasing values of @xmath26 ,
the fraction of extremists decreases ( this effect was quantified in figure [ fig9 ] ) , as well as the fraction of agents with moderate opinions , while negative opinions emerge in the population . from other viewpoint
, the absence of negative convictions ( @xmath85 ) leads to the emergence of extremists in the population , and the introduction of volatile agents with negative convictions makes moderate and indifferent opinions dominant , which seem realistic features of the model .
in addition , if the independent behavior becomes more frequent ( increasing @xmath9 ) , the pdfs become more symmetric and there are less agents sharing extreme opinions
. see figure [ fig9 ] , where we exhibit the fraction of extremists ( @xmath69 ) as a function of @xmath26 for three different values of @xmath9 .
notice that the increase of independent attitudes ( increasing @xmath9 ) tends to reduce the fraction of extremists in the population .
the increase of volatile attitudes ( increasing @xmath26 ) favors the emergence of moderate opinions and also reduces condensation at the extreme positions .
indeed , the increase of volatile individuals reduces , even extinguishes , the fraction of extremists and also prevents the arrival to consensus .
finally , let us comment that when both @xmath16 and @xmath15 are allowed to take negative values i.e. , @xmath86 , the phenomenology is similar to that discussed in this section , although the domain of the ordered phase shrinks ( not shown , as far as there are not new qualitative features ) .
therefore , the opinion patterns observed in this section can be attributed to the joint heterogeneity in the convictions and in the interactions , be them all positive or not .
when individuals have all the same conviction @xmath29 , or the same interaction strength @xmath71 , then , patterns of the type observed in figure [ fig4 ] emerge .
the substitution of @xmath29 by positive heterogeneous convictions , like in the case @xmath85 , is enough to yield patterns of opinions similar to those shown in figure [ fig8 ] , when we vary @xmath22 instead of @xmath26 .
( top , left pannel ) , @xmath67 ( top , right pannel ) and @xmath87 ( bottom ) , for several values of @xmath26 ( and @xmath46 ) .
empty ( full ) symbols represent ordered ( disordered ) states .
the population size is @xmath44 , and data are accumulated over @xmath45 independent simulations , as explained in the text.,title="fig:",scaledwidth=30.0% ] ( top , left pannel ) , @xmath67 ( top , right pannel ) and @xmath87 ( bottom ) , for several values of @xmath26 ( and @xmath46 ) .
empty ( full ) symbols represent ordered ( disordered ) states .
the population size is @xmath44 , and data are accumulated over @xmath45 independent simulations , as explained in the text.,title="fig:",scaledwidth=30.0% ] + ( top , left pannel ) , @xmath67 ( top , right pannel ) and @xmath87 ( bottom ) , for several values of @xmath26 ( and @xmath46 ) .
empty ( full ) symbols represent ordered ( disordered ) states .
the population size is @xmath44 , and data are accumulated over @xmath45 independent simulations , as explained in the text.,title="fig:",scaledwidth=30.0% ] as a function of @xmath26 ( probability of negative convictions ) for typical values of @xmath9 ( probability of independence ) , when @xmath46 .
one can see that the number of extremists in the population can be reduced ( even to zero ) for sufficiently large values of @xmath26 .
population size is @xmath44 and averages were performed over @xmath45 simulations.,scaledwidth=30.0% ]
in this work , we have studied a kinetic model of opinion formation with continuous states .
we considered pairwise interactions among randomly chosen agents , together with the possibility of independent behaviors that occur with probability @xmath9 .
two sources of heterogeneity were included : random competitive couplings ( interaction term ) and random convictions ( self - interaction term ) , that can take negative values with probabilities @xmath22 and @xmath26 , respectively .
we characterized the critical transitions , as well as the emerging phases , in terms of moderate and extremist individuals .
parameters @xmath22 , @xmath26 and @xmath9 rule microscopic disorder in the system , and their increase triggers transitions from collective order to disorder .
these transitions are characterized by mean - field exponents and their critical curves are shown in figures [ fig3 ] and [ fig7 ] . when only one source of heterogeneity is present , like in the case of sec .
[ case1 ] , where the convictions are given by @xmath29 for all @xmath2 , but there is a probability @xmath49 of negative couplings , patterns of opinions like those shown in figure [ fig4 ] occur .
qualitatively similar patterns arise when the pairwise interactions are @xmath71 for all @xmath32 but convictions are random . in the disordered phase , opinions are evenly distributed . in the ordered ones , extremists dominate .
the steady states show the presence of a large number of extremists with opinions either @xmath83 or @xmath47 .
the fraction of these extremists decreases for increasing values of the independence parameter @xmath9 , and the system tends to become disordered , indicating that independent attitudes reduce extreme opinions .
different patterns emerge when there are two sources of heterogeneity , with negative values or not , like in the case of sec .
[ case2 ] , where the couplings and the convictions are random . in the disordered phase , opinions are not evenly distributed , but moderate opinions dominate , with a peak around the neutral state . in the ordered phases , one of the opinion sides disappears .
when a second source of heterogeneity is present , the increase of volatile individuals ( increasing @xmath26 ) or of dissent interactions ( positive @xmath22 ) , reduces , even extinguishes , the fraction of extremists and also prevents the arrival to consensus .
the models considered in this work contribute to the analysis of the role of the diverse features that characterize individuals and their interactions . as can be seen in this paper ,
each modification yields phase diagrams of order parameters that seem similar , however a quite different phenomenology arises in terms of the distribution of opinions .
this hinders unification , justifying a separate analysis of different versions of the model .
these versions shed light on the origin and role of undecided individuals and extremism uprise , and the effect of realistic characteristics like conviction and independence in the dynamics of opinion formation and evolution .
in particular , the models contribute to understand the circumstances which favor the emergence and development of extremism in a fraction of the population , relating that emergence with the existence of individuals with strong positive convictions in the population . on the other hand , the presence of negative convictions , that represent volatile individuals that have a propensity to change mind , as well as nonconformity represented in the model by the independent behavior , lead to the dominance of moderate individuals .
the authors acknowledge financial support from the brazilian funding agencies cnpq and faperj . | we investigate opinion formation in a kinetic exchange opinion model , where opinions are represented by numbers in the real interval @xmath0 $ ] and agents are typified by the individual degree of conviction about the opinion that they support .
opinions evolve through pairwise interactions governed by competitive positive and negative couplings , that promote imitation and dissent , respectively .
the model contemplates also another type of nonconformity such that agents can occasionally choose their opinions independently of the interactions with other agents .
the steady states of the model as a function of the parameters that describe conviction , dissent and independence are analyzed , with particular emphasis on the emergence of extreme opinions .
then , we characterize the possible ordered and disordered phases and the occurrence or suppression of phase transitions that arise spontaneously due to the disorder introduced by the heterogeneity of the agents and/or their interactions .
keywords : dynamics of social systems , collective phenomena , computer simulations , critical phenomena |
sharpless 140 is a relatively diffuse h ii region at the edge of the much denser l1204 molecular cloud that harbors several clusters of young b stars at a distance of about 900 pc @xcite .
the most luminous object in this cloud was first observed in the near- and mid - infrared by @xcite and by @xcite in the far - infrared .
blair et al
. found a very red , compact object with a spectral energy distribution ( sed ) similar to that of the bn object in orion in addition to several other less remarkable infrared sources . using the kuiper airborne observatory , harvey et al .
found strong , nearly spatially unresolved emission in the far - infrared from 35 - 175 .
they fit several crude models to the sed that showed that the dust optical depth to the central source had to approach unity in the 50 - 100 spectral region and that the total luminosity was equivalent to that of an early b star . in the intervening years a host of much higher angular resolution
near - infrared and radio observations have been published .
these have revealed many additional infrared sources likely to be young stars or protostars as well as energetic outflows similar to those found in many other molecular clouds harboring young stars , e.g. @xcite .
it also seems likely that one or more of the strong mid - ir sources contains two or more young b stars @xcite and that the radio jet from irs1 is likely to be due to an ionized equatorial wind @xcite .
very cold , dense condensations only emitting longward of @xmath0 300 are also seen and may be even younger stages of star formation within the molecular cloud @xcite .
most of the luminosity of very young stellar objects and protostars is emitted in the far - infrared , wavelengths that are essentially invisible to ground - based telescopes due to absorption by water vapor . until the recent commissioning of the herschel space observatory ( pilbratt et al .
2010 ) and the stratospheric observatory for infrared astronomy ( sofia ; gehrz et al .
2011 ) , the far - infrared , however , has suffered from a substantial deficit in angular resolution compared to neighboring wavelengths observable with large ground - based telescopes .
high mass young stars typically form in dense clusters , and this puts a special premium on the value of high angular resolution to resolve individual components to understand their evolutionary state .
we present here the first imaging observations of the s140 cluster at wavelengths longward of the ground - based atmospheric window at @xmath0 20 25 with an angular resolution of a few arcseconds . with this resolution
we are able to distinguish the three main luminosity sources seen at shorter wavelengths as well as diffuse emission between the objects .
we describe the observations in [ obs ] and then in [ super ] describe some efforts to improve the angular resolution with various deconvolution techniques . in
[ compare ] we compare our 31 and 37 maps with those at shorter and longer wavelengths and show the spectral energy distributions for the individual members of this cluster , and in [ summ ] we summarize our results .
our observations were made with the forcast camera ( herter et al .
2012 ) on board sofia on the night of ut 1 dec 2010 .
the forcast camera has two separate 256@xmath1256 infrared arrays that cover the wavelength range from @xmath2 5 40 in a variety of bands with [email protected] pixels .
the two arrays can observe simultaneously through a dichroic beamsplitter that divides the wavelength range shortward and longward of 26 , or the long wavelength array can be used by itself without the intervening dichroic .
the observations presented here were made with the 37.1 filter ( hereafter , 37 ) without the beamsplitter and at 11.1 and 31.5 with the beamsplitter ( hereafter `` 11 '' and `` 31 '' ) .
in addition to the observations of s140 , a number of other science investigations were undertaken that are also reported in this issue , and several calibration targets were observed for both flux and point - source - response ( psf ) calibration .
the forcast images were calibrated to a flux density per pixel using a standard instrument response derived from measurements of standard stars and solar system objects over several flights ( herter et al .
the color correction from a flat spectrum source to that of s140 is negligible ( @xmath3 ) .
the estimated @xmath4 uncertainty in the calibration due to variations in flat field , water vapor burden , and altitude is approximately @xmath5 .
the point source response has been derived from averages over several flights as described in detail by buizer et al .
we discuss below the relatively minor uncertainties involved in using this average psf , and tests of its applicability using @xmath6 cep data taken on the same flight as s140 .
the nominal diffraction limit of sofia at 37 , 1.019@xmath7d with its 14% central obscuration , is 3.1 , and our observations show that achieved resolution is quite close to this limit , @xmath2 4 fwhm . at shorter wavelengths
the sofia psf is affected more by telescope pointing jitter and seeing , with seeing being dominant shortward of 5 because of the shear layer just above the telescope @xcite .
our images of both s140 and @xmath6 cep were obtained with standard thermal infrared techniques of chopping and nodding . because the cluster is relatively compact , we were able to chop and nod completely on the @xmath2 200 arcsec square array .
the chop frequency used was 2 hz , and chop amplitude of 90 with a nod amplitude also of 90 perpendicular to the chop direction .
each observation resulted in 120 seconds of on - chip integration as the result of one abba nod pair .
we obtained 10 useful observations of s140 and two of @xmath6 cep at 37 , and two of s140 at 11 and 31 .
the individual observations were co - added by cross - correlating all the subsequent images with the first .
the sky rotation was essentially negligible between the individual s140 observations .
the telescope software to enable accurate absolute pointing was not yet operational for these observations , so we used the highest resolution , nearest wavelength observations ( 24.5 ) from a large ground - based telescope @xcite to establish the absolute coordinate system for our observations .
for the purposes of deconvolution discussed later , the s140 frames were left in the original instrument coordinate system to match those of the psf calibration image .
figure [ decons140 ] shows the results of the processing described so far for s140 as well as related images from the literature discussed below . using aperture photometry we have derived flux densities for the three obvious components of the image as well as for two additional local flux maxima .
these results are listed later in table [ apflux ] for a 9 " diameter aperture .
we discuss these further in
[ compare ] after we describe our deconvolution work and the resulting higher resolution image .
the statistical uncertainties in these fluxes are negligible because of the very high signal - to - noise ratio , well over 100 , for all the sources except for the faintest objects at 11 where s / n @xmath2 10 .
table [ apflux ] also lists the luminosities of these `` objects '' for the assumed distance of 960 pc , using the 50100 flux densities of @xcite .
with nasa s previous airborne observatory , the kao , @xcite for example showed that significant improvements in the apparent angular resolution in far - infrared images were possible through super - resolution techniques such as maximum entropy deconvolution . with the kao s instrumentation these techniques
were limited to 1-d scans rather than 2-d images . in the intervening years
, the improvement in detectors and in analysis techniques suggests the possibility of substantial resolution improvements in 2-d far - infrared images with sofia . to that end
we have experimented with a variety of algorithms and parameters available as part of the `` mra '' package @xcite to improve the resolution in our s140 image .
our best results were obtained using the van - cittert algorithm with multi - resolution support for regularization . with this mode
the final deconvolved images did not depend strongly on the number of iterations , and therefore this was used for all our results below .
for example , the final image shown used 500 iterations , but the image changed very modestly after as few as 20 iterations .
additionally , we did not force positivity in the image .
as mentioned above we have used the psf derived by buizer et al .
( 2012 ) and done several tests to give us confidence in its applicability to the s140 images .
first we deconvolved the observed @xmath6 cep image and compared the result with the much higher angular resolution , but shorter wavelength 24.5 image of @xcite .
we found the spatial extent and spatial asymmetry to be very similar .
we also found the amplitude of the wings to the northeast and southwest to be relatively larger at 37 , which is entirely consistent with the expected dust temperature of order 150k at a radius of a few arcseconds . in a second test of our derived psf
, we produced several similar psf s with small changes in the width in both telescope axes , elevation and cross elevation , and then deconvolved the s140 image with these modified psf s . in all cases the major details of the deconvolved images
were essentially the same as for the derived psf .
so we are confident that none of our conclusions depends on the small uncertainties in our knowledge of the psf .
finally , figure [ decons140 ] shows the observed and deconvolved images of s140 at 37 together with the observed subaru image @xcite .
it is clear that virtually all the structure seen at the 0.6 diffraction limit of subaru at 24.5 is also visible at 37 in the deconvolved image , albeit at somewhat lower resolution . with the native 37 resolution of @xmath2 4 ,
however , much of this structure is smoothed to the point of being barely detectable .
the final derived psf is equivalent to the theoretical diffraction - limited psf convolved with a gaussian of roughly 1.5 fwhm , and the apparent resolution of the deconvolved s140 image is @xmath2 2.4 .
figure [ psffig ] shows the derived psf along with the observed @xmath6 cep image illustrating the larger extent of @xmath6 cep at 37 .
we have already noted above the excellent agreement between our 37 deconvolved image and the much higher resolution 24.5 image of @xcite . in particular
, in addition to the good agreement in position of the three main peaks , the agreement in morphology of the low level , diffuse emission north of irs 1 and also to the southwest suggests that the deconvolution process has not developed spurious features as might have been the case with a poor estimate for the psf or a poor algorithm .
there are probably still some residual positional problems in our 37 image at the level of @xmath2 1 due to the inherent uncertainties in fixing the aboslute positions by registering to the shorter wavelength ground - based images , but the agreement between the two images is remarkably satisfying . as a corollary it is also clear that there are no new sources present in the 37 data , either raw or deconvolved , that are not present in the subaru 24.5 image .
most importantly , we do not see any evidence for emission above the background from either of smm 1 , smm 2 , or smm 3 , the 450 sources discovered by @xcite . at 37 any emission from these sources must be below a level of @xmath2 30 jy to be invisible in our image .
this confirms their likely interpretation as very cold , prestellar condensations .
four very faint sources in the 37 image are indicated in figure [ decons140 ] that have counterparts in the _ spitzer _ irac images in the _ spitzer _ archive , presumably low luminosity young stellar objects .
as noted earlier , table [ apflux ] lists flux densities derived from our 37 image .
we list both the values measured using the raw image , as well as from the deconvolved image . since the deconvolution algorithm conserves flux , the total flux densities for both methods are identical , but we believe the deconvolved image provides a more realistic measurement of the flux densities in small beams required for a reasonable comparison of the relative fluxes and colors of the individual components in this region . table [ apflux ] also lists the observed 11 and 31 flux densities for a 9 aperture .
figure [ sed ] shows the spectral energy distributions for the various local maxima that have been identified previously in this region and which are clearly detected in our deconvolved map .
this shows graphically the fact , as noted by previous investigators , that the coldest mid - infrared condensations in this region are the diffuse local flux maxima called `` nw '' and `` vla 4 '' ( labelled in figure [ recon ] ) , while the hottest sed belongs to irs 1 .
irs 2 and 3 exhibit sed s that are intermediate in color temperature with irs 3 being the cooler source .
quantitatively , irs 1 dominates the shorter wavelength emission with 75% of the total emission at 11 , but at 37 it contributes less than 40% with the rest arising from the cooler sources , irs 2 , 3 , nw , vla4 , and the surrounding diffuse emission . in their original kao study of s140 @xcite showed that the dust optical depth was likely close to unity in the far - infrared based on the resolving power possible with that instrumentation . with the substantially higher spatial resolution available with sofia / forcast we can re - examine this conclusion .
for example , the ratio of 37 to 53 flux densities shown in figure [ sed ] for irs1 implies a color temperature of order 85k .
if we assume that half of the total flux arises in the central 4 core , this implies @xmath8 0.5 at 53 , quite consistent with the original kao study . in order to compare quantitatively the diffuse emission in our 37 map with that seen at much higher angular resolution at 24.5 by @xcite , it is necessary to compare the two images at identical levels of resolution
this technique was described as `` beam matching '' by @xcite .
since the native resolution at 37 is more than a factor of 10 worse than that at 24.5 it is sufficiently accurate to simply convolve the 24.5 image with the 37 psf .
figure [ recon ] shows the original , undeconvolved 37 sofia image together with the 24.5image of @xcite smoothed to what would have been observed with our 37 sofia psf .
this shows that there is substantially less extended diffuse emission at 24.5 .
figure [ recon ] also shows the relative 24.5 , 31 , and 37 fluxes along a one dimensional cut through irs 1 from northeast to southwest along the line shown in the figure .
this comparison shows quantitatively the higher level of extended , diffuse emission at the longer wavelengths , consistent with a radial temperature gradient away from the center of heating .
this effect can also be seen in the more steeply rising 24.537 sed slopes noted earlier for the sources labeled `` nw '' and `` vla4 '' well away from irs1 . in order to fully model the various components of this compact cluster , it will be necessary to obtain spatially resolved images at the peaks of their sed s , i.e @xmath0 70 . for the 3 brightest sources , this will be possible with sofia using the upcoming hawc far - ir camera . to understand the structure of the cores of the sub - millimeter sources would
, however , require the sensitivity and higher angular resolution of _ herschel _ with its pacs and spire cameras .
we have demonstrated the utility of the forcast camera on sofia for mapping compact regions of star formation at wavelengths beyond the familiar ground - based mid - ir atmospheric windows .
this allows us to define the sed s and colors of the major emission sources in s140 with unprecedented angular resolution in the far - ir .
since the data were taken with quite high s / n , we were able to successfully pursue spatial deconvolution of the image and show that the basic structure which appears at @xmath0 25 with sub - arcsecond resolution from large ground - based telescopes can also be identified in our far - ir images .
we find emission from _ all _ the major sources seen in this cluster at shorter wavelengths from ground - based telescopes and confirm that the coolest sources are irs3 and the diffuse knots , vla4 and nw . we also see no evidence for emission from the sub - mm peaks in this region .
this work is based on observations made with the nasa / dlr stratospheric observatory for infrared astronomy ( sofia ) .
sofia science mission operations are conducted jointly by the universities space research association , inc .
( usra ) , under nasa contract nas2 - 97001 , and the deutsches sofia institut ( dsi ) under dlr contract 50 ok 0901 .
financial support for forcast was provided to cornell by nasa through award 8500 - 98 - 014 issued by usra .
support at the university of texas was provided by usra award 08521 - 012 .. we also especially thank w. de wit for providing the digital versions of the 24.5 images for both @xmath6 cep and s140 and acknowledge insightful comments from d. lester and an anonymous referee .. this research used the facilities of the canadian astronomy data centre operated by the national research council of canada with the support of the canadian space agency . observed 37 & 2176 & 485 & 669 & 465 & 381 & 6730 + deconvolved 37 & 2624 & 577 & 710 & 438 & 358 & 6730 + observed 11 & 110 & 4.0 & 9.7 & 3.5 & 2.6 & 145 + observed 31 & 1585 & 368 & 401 & 242 & 184 & 3780 + luminosity ( ) & 10,700 & 3200 & 2900 & 800 & 650 & 19,150 + bally , j. , reipurth , b. , walawender , j. & armund , t. 2002 , apj , 124 , 2152 blair , g.n . , evans , n.j .
ii , vanden bout , p.a . & peters , w.l .
1978 , apj , 219 , 896 de buizer , j.m . ,
morris , m.r . ,
becklin , e.e . ,
zinnecker , h. , herter , t.l . ,
adams , j.d . , shuping , r.y . & vacca , w.d .
2012 , ... crampton , d. & fisher , w.a . 1974 , pub .
, 14 , 283 de wit , w.j .
2008 , , 685 , l75 de wit , w.j .
2009 , a&a , 494 , 157 erickson , e. f. & dunham , e. w. 2000 , proc .
spie , 4014 , 2 evans , n.j . ,
ii , mundy , l.g .
, kutner , m.l . & depoy , d.l .
1989 , , 346 , 212 gehrz , r.d .
2011 , adv .
space res . , 48 , 1004 harvey , p. m. , campbell , m. f. , & hoffmann , f. 1978 , apj , 219 , 891 herter , t. l. , adams , j. d. , de buizer , j. m. , gull , g. e. , schoenwald , j. , henderson , c. p. , keller , l. d. , nikola , t. , stacey , g. , and vacca , w. d. 2012 , apjl , this issue hoare , m. g. 2006 , apj , 649 , 856 ivezi , z. & elitzur , m. 1997 , mnras , 287 , 799 kraemer , k.e .
2001 , apj , 561 , 282 kurtz , s.e . ,
watson , a.m. , hofner , p. & otte , b. 1999 , apj , 514 , 232 | we present 37 imaging of the s140 complex of infrared sources centered on irs1 made with the forcast camera on sofia .
these observations are the longest wavelength imaging to resolve clearly the three main sources seen at shorter wavelengths , irs 1 , 2 and 3 , and are nearly at the diffraction limit of the 2.5-m telescope .
we also obtained a small number of images at 11 and 31 that are useful for flux measurement .
our images cover the area of several strong sub - mm sources seen in the area
smm 1 , 2 , and 3 that are not coincident with any mid - infrared sources and are not visible in our longer wavelength imaging either .
our new observations confirm previous estimates of the relative dust optical depth and source luminosity for the components in this likely cluster of early b stars .
we also investigate the use of super - resolution to go beyond the basic diffraction limit in imaging on sofia and find that the van cittert algorithm , together with the `` multi - resolution '' technique , provides excellent results .
[ 1995/12/01 ] |
a rapidity gap is a region in rapidity space with no hadrons .
it has been studied for a long time in diffractive scattering physics . in 1986 dokshitzer , khoze and troyan
@xcite suggested that due to its peculiar color flux , a w fusion process producing a higgs boson could also lead to a rapidity gap .
as each hadron is a color singlet , when they emit a w boson , which is a color singlet too , they remain in singlet states , although separated in systems of quark and diquark . therefore
, the initial hadrons do not need to exchange color .
the quark which emitted the w may exchange color with the diquark belonging to the same hadron , in order to fragment . according to the lund string model @xcite
, a string will be stretched between the quark and the diquark of each hadron ( see fig.[fig - fusao ] ) . when both strings fragment , almost all the hadrons formed are expected to be comprised , roughly speaking , between the incident beam direction and that given by the jet containing the quark that emitted the massive vector boson .
[ gap ] shows a typical plot in @xmath0 ( rapidity @xmath1 azimuthal angle ) space , for an event of this kind @xcite .
the jets shown in the figure come from the hadronization of the quark which emitted the vector boson ( it is possible to use them as tagging jets ) .
each jet is supposed to occupy about a 0.7 radius circle @xcite ( in @xmath0 space ) .
the idea is to look for a gap in the region between the tangents to those jets . imposing the higgs to decay into a z pair , which in turn is forced to decay into muon pairs , no hadron production occurs besides that from the hadronization of the beam remnants . on the other hand , in gluon fusion
each hadron emits a color octet ( the gluon ) , and therefore turns into a colored object .
hence the initial hadrons remnants must exchange color with each other in order to become color neutral again ( see fig .
[ fig - fusao ] ) . moreover , unlikely vector bosons , gluons themselves emit quarks and other gluons , which will hadronize later , thus filling the central rapidity region .
the purpose of this work is to discuss the possibility of using the existence of the rapidity gap for distinguishing higgs boson production by w fusion from that by gluon fusion .
it is organized in the following way : in section 2 , some problems which appear in a gap analysis are discussed , and the approaches already existing on this subject are presented ; in section 3 the event generation for this work is described ; in section 4 , the generated events are analyzed , the gap survival probability and the number of events having gaps occurring for a fixed integrated luminosity are obtained for various higgs mass , and many situations , where different cuts had been imposed ; it is shown that although for some cases there is a big gap survival probability , for higgs produced by w fusion the number of events having a gap which occur in an accelerator like the lhc is very small for low higgs masses and null for the bigger ones ; when the number of events having gaps is bigger for w fusion , it is of the same order for gluon fusion , thus being impossible to distinguish the two higgs production processes , independently of some earlier conjectured problems , like pile up events . finally , in section 5 , some ideas and conclusions are presented .
in order to use a rapidity gap for identifying the higgs production mechanism , some problems have to be overcome .
for example , although in a w fusion the hadrons are expected to be close to the initial beams directions , it might occur that some of them appear in the central region ( in fact , it will be shown that this is the case in many events ) .
furthermore , each incident proton is composed of several partons , and it is possible that more than one scattering occurs , therefore filling up the region between the tagging jets ( these are called _ multiple interactions _ in this paper ) .
moreover , problems arise from the high luminosity designed for the accelerator ( as is the case for the lhc )
. more than one proton - proton scattering may occur in the same bunch crossing , producing the so - called _ pile up _ events . that will produce even more hadrons , probably filling completely the central region .
it will be shown , however , that at least for lhc pile up events are not the worst problem .
it is hard to distinguish between the two higgs production mechanisms , even when no pile up events are considered .
besides those problems , there are others that will not be discussed here such as gaps produced by statistical fluctuations in the background events , gaps produced by other color singlet scattering ( such as @xmath2 ) and the most common ones , diffractive scatterings .
these problems have already been studied by many authors @xcite and do not address the problem investigated in this paper , which is the use of rapidity gaps to distinguish between higgs production by gluon and w fusion .
due to the difficulties mentioned above , added to the smallness of the involved cross sections it is necessary to quantify the probability that a gap be observed , and furthermore , the possibility of using its existence in experimental analysis .
bjorken @xcite proposed a variable , the _ gap survival probability _ , determined in the following way : if @xmath3 is the probability that two protons pass through each other with impact parameter @xmath4 with no interaction occurring , except the hard one , the gap survival probability is given by : @xmath5 where @xmath6 is a factor associated with the hard collision , being essentially a measure of the overlap of the parton densities in the colliding hadrons . eq . [ probsob ] is evaluated under some conditions ( eikonal approximation , gaussian form for @xmath6 and for the eikonal itself ) ; @xmath7 is obtained . in ref .
@xcite the same calculation is performed , but using several different models for hadron collisions .
they obtain , for lhc energies , a gap survival probability lying between 5.3 % and 22.1 % ( five models are analyzed , and just for one of them , the reggeon model , one gets a probability as high as 22.1 % ; for the other four , it is below 8.2 % ) .
an important observation must be made at this point .
the gap survival probability , as proposed by bjorken , and used in ref .
@xcite , gives the probability that , in a proton
proton collision , only one parton parton scattering occurs .
it does not imply , necessarily , that a rapidity gap exists , because an eventual gap can be filled by the hadronization of the beam remnants , as already mentioned .
this aspect has not been taken into account in the above references .
previous papers @xcite had used monte carlo simulation to analyze some aspects of the problem .
the main conclusions are : a ) for w fusion a rapidity distribution shows two peaks ; the dip between these peaks increases for increasing higgs masses and the distance between the peaks increases with cm energy @xcite ; when multiple interactions are included , a @xmath8 gev cut on charged particles recovers the dip @xcite ; b ) @xmath9 is obtained for w fusion , @xmath10 for the background ( @xmath11 and @xmath12 ) and a null gap survival probability for @xmath13 .
the present work broadens the scope of the former papers , taking into account a larger range of higgs masses , and mainly , using the processes cross - sections and the lhc luminosity in order to obtain the number of events per year presenting a surviving rapidity gap instead of the probability of having such a gap .
the events for this work have been generated with pythia @xcite , using as distribution functions cteq set l2
the top quark is supposed to have mass 174 gev .
a simple calorimeter is simulated with lucell , a jet algorithm included in pythia .
that calorimeter covers the rapidity region from @xmath14 to @xmath15 , with segmentation : @xmath16 .
pythia includes some models for simulating multiple interactions , and we had chosen the default , that is the simplest one . this model is described both in the program manual and in ref . @xcite .
we have considered higgs produced both by gluon fusion and by w fusion , with mass varying from @xmath17 to @xmath18 , supposing a pp collision with 14 tev of cm energy . in both processes the higgs decay into a z bosons pair ,
each of which then decay into a muon pair .
that choice helps preventing the production of hadrons that could fill an eventual gap , has been produced with pyhtia s process number 102 , while w fusion , @xmath19 , has been generated with pythia s process number 124 .
@xcite ] . for both processes , three groups of events from
now on called group i , group ii and group iii have been produced .
there is a set of common cuts , cuts a , applied to the three groups , described below .
the three groups have been submitted to different cuts ( besides cuts a ) in order to determine the fraction of events containing gaps using different selection criteria for events and gap definitions .
all these cuts have been largely discussed in literature , and therefore they will be presented here without further justification . for each group
the number of generated events is such that , after imposing the respective group cuts ( without including cuts a ) , 10,000 events remain .
the groups are defined in the following way : a ) * group i * a tag is applied to two jets , and the gap is looked for between these jets .
it is known that although this double tagging eliminates considerably the background , it reduces the signal too .
nevertheless we adopted such a cut because it could enhance the rapidity gap signature . here
we demand that the event has at least two jets with @xmath20 gev and @xmath21 .
these choices are imposed because the quarks that emit the w boson acquires transversal momentum of order @xmath22 and follows approximately the initial beam direction .
if more than two jets satisfy the above conditions , the two with the largest transversal moments are picked up .
the rapidity gap width is looked for in a region defined as being the distance in rapidity space between the tagging jets , @xmath23 , as seen in fig .
[ gap ] , where @xmath24 and @xmath25 are the jets rapidities .
the value 1.4 is subtracted due to jet width in rapidity space .
cuts a are applied too .
b)*group ii * here only one jet is tagged and the applied cuts are similar to those used in ref .
the event is accepted if there is at least a jet with energy @xmath26 tev and @xmath27 . in this case a gap is looked for inside a fixed interval , symmetric around @xmath28 , and with width @xmath29 ( @xmath30 ) .
cuts a are applied too .
c)*group iii * no cut beyond that from cuts a have been applied to this group .
gem @xcite adopted this kind of analysis .
the gap widths here is defined in the same way as for group ii .
cuts a are applied to all events .
they consist of @xcite : a ) @xmath31 and @xmath32 gev .
most of the papers on this matter demand four leptons obeying these conditions .
but , as this is very restrictive , we have relaxed it , demanding four leptons with @xmath33 gev but just three of them had to have @xmath31 ; b ) for the signal , leptons are produced isolated .
they were accepted if inside a region of radius @xmath34 around each of them no more than 5 gev of transversal energy had been deposited ; c ) it must be possible to produce , using all four accepted leptons , two pairs with invariant mass next to the z mass : @xmath35 gev ; d ) muon identification and track matching ; e ) at least one z for which @xmath36 , where @xmath37 is the z pair invariant mass .
the figures presented in this sections have been obtained in the following way : a ) 10,000 w fusion and 10,000 gluon fusion for each of the five higgs masses considered have been subjected separately to cuts a ; the fraction of events surviving the cuts in each case @xmath38 were then obtained .
this procedure does not affect the sample , because the cuts applied concern the part of the event that will not be used in the final analysis , i.e. , the leptons and z s .
it is not relevant which of the events are thrown away in this case .
@xmath39 is about 60 % for gluon fusion and about 70 % ( @xmath40 gev ) to 80 % ( @xmath41 gev ) for w fusion events ; b ) for each group events , i , ii and iii , a number of events is generated such that , after being applied the specific cuts , 10,000 events remain .
c ) @xmath42 is the number of events , for each group , for each process and for each higgs mass , which survive , that is , which maintain the region where the gap is searched for with no hadrons .
@xmath42 is obtained for three different cases : i ) all charged particles are included , except for the muons selected by cuts a ; ii ) charged particles with @xmath43 gev , except for the muons selected by cuts a ; iii ) charged particles with @xmath44 gev , except for the muons selected by cuts a. c ) @xmath42 is divided by the * total * number of generated events ( before any kind of cut is done ) , producing @xmath45 for the three situations analyzed above , ( i ) , ( ii ) and ( iii ) .
d ) for each case , @xmath45 is multiplied by the corresponding fraction of events surviving to cuts a and by 100 , producing @xmath46 .
next figures represent @xmath46 .
a ) * fig .
[ a71 - 72 ] * shows the gap survival probability for events from group i. in the upper part , multiple interactions have not been added yet . if no @xmath47 cut is applied ( fig .
[ a71 - 72].a ) , @xmath46 is very small for any higgs mass , unlike what could be expected from theoretical approaches .
it occurs probably because hadrons produced by the fragmentation of the beam remnants reach the detector central region in many events .
this fact shows that even if the main interaction could occur separated from the secondary ones , few events would be completely clean .
when @xmath47 cuts are considered , however , the gap survival probability increases , as may be seen in fig [ a71 - 72 ] ( b ) and ( c ) , and if just @xmath48 gev particles are accepted , @xmath46 lies between 5 and 9% for higgs masses between 300 and 700 gev .
@xmath46 grows with higgs mass as later observed @xcite .
things change completely when multiple interactions are included .
good results are obtained just when particles with @xmath49 gev are left behind ( fig .
[ a71 - 72 ] ) .
if no @xmath47 cut is imposed , no gap is found for any higgs mass and any process .
b ) in fig .
[ c74 - 75 ] the same analysis is performed for group ii events . in such case , even without @xmath47 cuts
, @xmath46 lies between 4 and 8% . using @xmath47 cuts , results are even better for w fusions , but some of the gluon fusions will have rapidity gaps too .
it should be noted that unlike for w fusions , for gluon fusion , @xmath46 decreases with the higgs mass .
taking into account multiple interactions , it is again clear that @xmath47 cuts have to be imposed ; if only particles with @xmath43 gev are accepted , @xmath46 lies between 3 and 5% for w fusion .
nevertheless , some gluon fusion events will produce gaps too , mainly for lower higgs masses .
if @xmath44 gev is imposed , a more expressive @xmath46 value is found for w fusion , between 22% and 33% .
but once more , the same cut leads gluon fusion to produce events with gaps .
c ) fig .
[ e78 - 710 ] shows the same analysis for group iii events .
once more the gap survival probability increases when @xmath47 cuts are applied , both for w and gluon fusion .
the results here are slightly smaller than for group ii .
group iii events ( fig .
[ e78 - 710 ] ) show results quite similar to that from group ii .
if no @xmath47 cuts are applied , @xmath50 , and for increasing @xmath47 cuts , @xmath46 grows up both for w and gluon fusion .
nevertheless , @xmath46 behaves oppositely with increasing higgs mass in w fusion and in gluon fusion .
based upon what had been seen until now , one could conclude that in some circumstances the gap presence is very clear .
for example , for a heavy higgs @xmath51 gev , @xmath52 % for group iii events and @xmath53 % for group ii events , in both cases taking into account final charged hadrons with @xmath44 gev , @xmath29 and with multiple interactions included . in both cases , for gluon fusion @xmath54 % .
for a lighter higgs , the results are not so good . for @xmath40 gev , @xmath55 for group
iii events and @xmath56 for group ii events , but the respective @xmath46 values for gluon fusion are 6% and 8% .
as all those events passed by the same cuts , one could have events with a rapidity gap that could had been produced either by gluon or w fusion .
but there is one very important point that should be included in the analysis .
it is the cross section for each of the processes , gluon and w fusion producing higgs . to get the next figures ,
@xmath46 has been multiplied by the respective process cross section , and by the luminosity lhc is supposed to have ( @xmath57 ) . when these factors are taken into account , fig .
[ g1ea ] , [ g2eta4ea ] and [ g3eta4ea ] show the number of events which will keep a gap in an year for @xmath58 in the range 300 - 700 gev .
they are obtained , respectively from group i , group ii with @xmath29 and group iii with @xmath29 events .
each figure presents in the upper part , the results obtained without including multiple interactions and in the lower part , results including multiple interactions . as before , for both situations , three cases have been considered : a ) all charged hadrons have been taken into account ; b ) charged hadrons with @xmath43 gev taken into account ; c ) charged hadrons with @xmath44 gev taken into account .
it is not difficult to see that the results are not very good , either including or not multiple interactions . for group i ( fig .
[ g1ea ] ) , less than two events with a higgs produced by w fusion will produce a gap in each year , when final charged hadrons with @xmath44 gev are counted . with a softer cut , not even one event will be observed in one year . for group
ii , although the number of events with rapidity gap produced by w fusion processes is larger than that for group i , it is still small and , what is worse , has the same magnitude that gluon fusion process has . for group iii ,
the only situation in which event with rapidity gap could be expected is that showed in fig .
[ g3eta4ea](a ) , which is not a realistic one , since no multiple interaction has been included .
some conclusions may be drawn from our investigation : a)for w fusion , @xmath59 increases with @xmath58 , and for gluon fusion this behavior is opposite ; b ) on the other hand , @xmath60 decreases with @xmath58 , both for gluon and w fusion ; c)no gap could be observed without @xmath61 cuts ; d ) after @xmath61 cuts , both w and gluon fusion have gaps , therefore being impossible to use the gap existence in distinguishing them ; e ) the above results do not depend on pile up , which have not been included ; f ) as the integrated luminosity is the same for w and gluon fusion , and for any higgs mass , the cross section is responsible for the @xmath60 behavior with higgs mass .
this work was supported by fapesp ( tll ) and by cnpq ( coe and tll ) . | the possibility of distinguishing between two higgs production mechanisms , w fusion and gluon fusion , due to rapidity gap existence is investigated using the monte carlo event generator pythia .
it is shown that , considering the designed cm energy and luminosity for the lhc , it is not possible to distinguish between the two higgs production processes as , for a given integrated luminosity , they lead to the same number of events containing a rapidity gap . +
pacs numbers:13.85.qk , 12.15.ji , 13.87.fh , 14.80.bn |
near - earth objects ( neos ) are temporary visitors in the region around the terrestrial planets , and significant work has gone into understanding their dynamical behavior and lifetimes .
gladman et al . ( 2000 ) and others found that neos have short lifetimes , only about 10 myr on average .
gravitational perturbations from planets , collisions with the terrestrial planets , or loss into the sun or ejection out of the solar system limit their average lifetimes . contrasted with the age of the solar system
, the much shorter neo lifetimes suggest that today s neos are simply the current incarnation of a constantly re - filled steady - state population . cratering records ( grieve and shoemaker 1994 ; stffler and ryder 2001 ) find that the impact flux on the terrestrial planets and the moon has been relatively constant on gyr timescales , supporting the idea of a steady - state population .
the problem of how to refill neo space from the seemingly static main asteroid belt was answered with the discovery that asteroids are quite mobile ( bottke et al . 2000 , 2006 ) .
thermal forces work via the yarkovsky - effect to change asteroid s semimajor axes over time ( farinella and vokrouhlick 1999 ; bottke et al .
2000 , 2001 ; see also a review by bottke et al .
this discovery provided the key piece of physics needed to understand how main belt asteroids become neos .
it had been known that some of the more powerful resonances located near , or in , the main asteroid belt could rapidly excite asteroid s eccentricities , sending them onto planet - crossing orbits leading to their delivery to neo orbits .
the yarkovsky effect showed how to efficiently get asteroids into these resonances .
the two most important resonances for this delivery process are readily visible in a plot of the asteroid belt .
first , the 3:1 mean motion resonance ( mmr ) with jupiter is located at heliocentric distance @xmath12.5 au , and is responsible for the large gap in the asteroid distribution at this semi - major axis .
the other is the @xmath4 secular resonance , which occurs when the precession frequency of an asteroid s longitude of perihelion is equal to the mean precession frequency of saturn .
this resonance is inclination dependent , and is very efficient at delivering low - inclination bodies to neo orbits .
it is the effective inner edge of the asteroid belt at @xmath12.15 au , and estimated to deliver @xmath137% of all neos with @xmath5 ( bottke et al .
2002 ) .
the inner main belt ( imb ; @xmath6 au ) bound by these two resonances is a predominant source of neos .
dynamical models predict that @xmath161% of the @xmath7 neo population comes from there .
the majority of the detected imb asteroids , @xmath1 4/5 those with @xmath8 , are on low - inclination orbits ( @xmath9 ) .
although a compositional gradient is known to exist in the main belt , with low - albedo , primitive asteroids being predominant in the central ( 2.5 [email protected] au ) and the outer asteroid belt ( @xmath112.8 au ) ( gradie and tedesco 1978 ; moth - diniz et al .
2003 ; masiero et al .
2011 ) , the imb contains numerous primitive asteroids ( campins et al . 2010 ; gayon - markt et al . 2012 ; masiero et al . 2011 ) .
for instance , in a sample of wise - studied asteroids limited to absolute magnitude @xmath12 , about 1/6 of these bodies in the imb with measured albedos and sizes have geometric visible albedos @xmath13@xmath14 , where we use @xmath13@xmath14 as a simple way to separate low - albedo primitive bodies from more processed or igneous bodies typically with higher albedos .
recent studies devoted to finding the origin of the primitive neos 1999 rq@xmath15 , 1999 ju@xmath16 , and 1996 fg@xmath16 ( baseline targets of the sample return space missions , osiris - rex , hayabusa - ii , and marco polo - r ) have found that each of these three bodies are almost certainly ( @xmath1790% ) delivered from the imb following the well - studied dynamical pathway from the main belt to neo - orbits ( campins et al .
2010 , 2012 ; walsh et al . 2012 ; bottke et al .
2000 , 2002 ) .
moreover , the current low inclination ( @xmath18 8@xmath19 ) orbits of these bodies is indicative of origins on similarly low inclination orbits in the main belt .
furthermore , jenniskens et al .
( 2010 ) , and gayon - markt et al .
( 2012 ) have idenified the imb at low inclination ( @xmath18 8@xmath19 ) as the likely source ( @xmath1790% ) of the neo 2008 tc@xmath20 , the asteroid whose impact produced the almahata sitta meteorites .
the work presented here develops the fundamental hypotheses to be tested by planned asteroid sample return missions .
dynamical evolution studies are critical to obtaining the maximum scientific benefit from these missions .
such studies support primary mission objectives to characterize the geologic and dynamic history of the target asteroids and provide critical context for their returned samples .
in addition , cosmochemical analyses of e.g. , cosmogenic isotope ratios , radionuclide abundances , and nuclear track densities will , in return , provide important constraints on the dynamical evolution of the parent asteroid .
this synergy will result in improved understanding of the dynamical pathways that transform imb objects into neos .
furthermore , the osiris - rex mission specifically will provide the first ground - truth assessment of the yarkovsky effect as it relates to the chemical nature and dynamical state of an individual asteroid .
these studies will provide important input to models of the evolution of asteroid families by yarkovsky drift and delivery to orbital resonances in the imb .
the landscape of the inner main belt is dominated by a handful of asteroid families and a diffuse population of `` background '' objects ( see fig .
[ fig : imbalbedo ] ) . among primitive , low - albedo , bodies
the largest known family is the low - albedo component of the nysa - polana complex , followed by the erigone family and the sulamitis family ( nesvorn 2010 ; gayon - markt et al .
2012 ; campins et al . 2012 ) . following an asteroid break - up event ,
fragments are launched onto orbits that are distinct from , but similar to , the parent body .
specifically , their `` proper orbital elements '' , roughly the long - term average of their osculating orbital elements , remain linked over time ( zappal et al . 1990 , 1994 ;
kneevi and milani 2003 ) .
smaller fragments are typically launched with higher velocity , creating a size - dependent spread in orbital elements . over time the thermal yarkovsky effect induces a size - dependent drift in semi - major axis , such that smaller bodies drift faster , and thus further , over time than the larger bodies .
therefore families of asteroids are identifiable in two ways : clustering in proper orbital element space , and correlated shapes in size vs. semimajor axis due to size - dependent yarkovsky drift ( see vokrouhlick et al .
2006a and bottke et al .
2006 ) . in the context of neo - delivery
, the size , age and location of an asteroid family are important properties .
first , the larger the family , the more potential asteroids that can be delivered .
second , the age of a family determines how far it has spread by the yarkovsky - effect , where older families can spread further .
finally , the location of the family determines how far its fragments must drift via the yarkovsky effect to reach a resonance .
as we find with the nysa - polana complex , a location very near a resonance makes locating the center and determining the age of the family much more difficult .
not all asteroids are associated with a specific asteroid family .
there are a similar amount of `` background '' objects in this region as there are members of an identified family ( gayon - markt et al .
this work will , however , focus on asteroid families for a few reasons .
first , asteroid families imply a physical correlation between fragments .
thus , we may learn more about an interesting asteroid by studying a much larger and brighter parent body or a suite of associated family members .
second , families orbital elements evolve in a relatively simple way - objects drift in semimajor axis at a speed relative to the inverse of their size due to the yarkovsky effect .
this means that a family member that has drifted into a resonance can be followed by equal or smaller - sized members of the same family at the same time - delivery of objects can be correlated temporally .
the specific asteroid family targeted in this work is part of the larger nysa - polana complex ( npc ) - which appears to be two overlapping asteroid families of different asteroid taxonomies .
their overlapping nature has made detailed study difficult , but now the large database of albedos measured by the wise mission has made a study of each of the two componenets of the family possible ( masiero et al .
2011 ; mainzer et al .
2011 ) .
section [ s : familyidentification ] describes the analysis of the families in the low - albedo component of the nysa - polana complex - seeking to find the family center , parent and age .
section [ sec : properties ] describes the known physical properties of the family and section [ discussion ] analyzes a second primitive family in the same region .
as suggested in works by zappal et al .
( 1995 ) , cellino et al .
( 2001 , 2002 ) , and nesvorn ( 2010 ) one component of the nysa - polana complex consists of low - albedo objects belonging to c- or b - type taxonomies ( see also campins et al .
this work will re - analyze the low - albedo population of the inner main belt ( at low inclinations and moderate eccentricities ) in order to find the center , the parent and the age of any low - albedo families in this region .
the first target is the low - albedo component of the nysa - polana complex seeking to constrain the center and find the parent of the family , followed by an age estimate .
we have used the calculated albedos from masiero et al .
( 2011 ) to study asteroids observed by the nasa wise mission in the imb with low albedo .
the wise mission measured fluxes in 4 wavelengths 3.4 , 4.6 , 12 and 22 @xmath21 m , and combined with a neatm thermal model calculated visible albedos , ( @xmath13 ) , and diameters ( @xmath22 ) for over 100,000 main belt asteroids .
not all asteroids were observed and not all observed asteroids were detected the same number of times .
therefore errors on albedo and diameter can vary from object to object , and also the sample of objects is not complete at any diameter .
however , this work is not aiming to build a definitive list of family members , rather the aims are to better understand the family structure in the inner main belt .
this database gives accurate measurements for a representative sample of bodies .
within this dataset there are 6702 asteroids in the inner main belt ( @xmath23 au@xmath24 au ) that are measured to have albedo , @xmath13@xmath25 0.1 , by wise and that also have synthetic proper orbital elements in the database of 318,112 objects as of may 2012 from the astdys database http://hamilton.dm.unipi.it/astdys ( see also kneevi and milani 2003 ) . the nysa - polana complex ( npc ) of asteroids
is centered around @xmath26 and @xmath27 .
hence out of the entire imb , we take a subset with orbital eccentricities @xmath28 with @xmath29 .
this area is centered on the npc , but contains a large portion of the entire imb .
the largest @xmath30 value in this dataset is @xmath31 .
this subset of asteroids has a distinct `` v - shape '' , or rather half of a v - shape , in the distribution of the absolute magnitude @xmath30 as a function of @xmath32 ( fig .
[ fig : vshape ] ; see also gayon - markt et al .
this v - shape is typical of asteroid families due to the size dependence of the yarkovsky effect semimajor axis drift of the family members ( see vokrouhlick et al . 2006a or bottke et al .
this is expected since the selection of asteroids surrounding the nysa - polana complex should be dominated by low - albedo family members that are expected to show a correlation in semimajor axis and absolute magnitude
. however , two half v - shapes are visible , indicating the possible presence of two families with asteroids of low albedo in the region .
this section is focused on the higher density v - shape structure , with the other analyzed in section [ discussion ] .
, for all wise - observed asteroids with albedo @xmath13 below 0.1 , eccentricities between 0.1 and 0.2 and inclinations below 10@xmath19 , plotted as a function of their semi - major axes ( au ) .
the points with circles around their dots indicate the objects with published visible and near - ir spectra .
the bottom panel is the same as the top , except lines are drawn to outline the yarkovsky `` v - shape '' in the data .
[ fig : vshape ] ] identification and analysis of the low - albedo families in the inner main asteroid belt pose a few unique problems .
first , half of the typical asteroid family `` v''-shape is visible in fig .
[ fig : vshape ] , which suggests that the parent body is very close to the 3:1 mean motion resonance with jupiter .
so close , in fact , that we can assume that at least half of the family was immediately lost to the powerful resonance shortly after the formation of the family ( though we do search for family members on the other side of the resonance ) .
second , the families are very extended , nearly reaching from the 3:1 to the @xmath33 , suggesting an old age and increasing the possibilities of interlopers and confusion in determining family membership .
lastly , there may be more than one family .
typically , asteroid families are centered on a known parent body as determined by hierarchical cluster method ( hcm ) fitting . with a known center
the next variable to match is the age of the family which is determined by the width of the bounding envelope due to the size - dependent yarkovsky drift of the asteroids .
however , in this case , the shape of the asteroid distribution was not aligned with the previously proposed parent asteroid ( 142 ) polana .
given the uncertainty in both the center and boundaries of the family , we seek to constrain these one at a time .
first , following vokrouhlick et al .
( 2006a ) , the @xmath34 parameter can be used to normalize asteroid distance from a possible family center by their size .
the yarkovsky drift rate is inversely proportional to an asteroid s diameter , so removing this measureable allows a search for clustering of drift distances from possible asteroid family centers .
the @xmath34 parameter is defined @xmath35 where @xmath30 is the absolute magnitude of the asteroid and @xmath36 is the distance an asteroid has drifted from its initial semimajor axis ( @xmath37 ) over a time @xmath38 at a rate @xmath39 .
this formulation of @xmath34 normalizes each asteroid by its size , as the yarkovsky drift rate is inversely proportional to its size @xmath40 such that @xmath41 . for a simplistic asteroid family formation event where all bodies start with the same @xmath37 and similar spin axes
, all family members will have the same @xmath34 , since this value normalizes their drift rate by their size .
this is simplified by ignoring both the spin - axis evolution over time and different @xmath37 for each asteroid due to the dispersion of fragments following the family - forming impact . however , the @xmath34 parameter can help search for the center of the family .
the distribution of @xmath34 for a given @xmath37 would be flat for a random distribution of asteroids .
if a family is present that had correlated @xmath34 values due to their common semimajor axis drift times , then for an @xmath37 near the center of the family there will be a spike in @xmath34 values .
thus , the @xmath34 distribution can be re - calculated at different @xmath37 , seeking the value that provide the tightest clustering of @xmath34 .
the search algorithm simply counted the number of asteroids in the maximum bin of @xmath42 au , and the range of @xmath37 was selected to cover the centers suggested by fig .
[ fig : vshape ] . as @xmath37 was varied from @xmath432.422.51 au ,
a clear trend was observed , with a peak at @xmath44 au ( fig .
[ fig : censearch ] ) .
value as a function of a center semimajor axis value ( @xmath37 ) .
the bottom panel shows a histogram of @xmath34 values for all asteroids in the subset of wise data ( @xmath45 au @xmath46 au ; @xmath47 ; @xmath48 ) , for the best - fit family center location of @xmath44 au .
the bin size is @xmath49 au , and the maximum is found at @xmath50 au .
[ fig : censearch ] ] the previous section found the center of the family is best fit at @xmath51 au , which is very close to the 3:1 mmr with jupiter nominally located at @xmath1 2.5 au . in the selected subset of asteroids
there is only one asteroid with @xmath52 au @xmath53 au with an absolute magnitude less than 12 and only three with @xmath30 less than 13 .
we have also searched known asteroids with proper elements in the same region , finding only a few with measured low albedo from iras or akari ( asteroids 877 and 2159 , tedesco et al .
2002 , usui et al .
2011 ) , but both have inclinations around 3.5@xmath19 , outside the boundaries of the estimated family . the only asteroid with suitable inclination , semimajor axis and absolute magnitude is ( 495 ) eulalia , with @xmath54 an albedo @xmath13=0.051 and a diameter @xmath55 km ( masiero et al . 2011 ) .
the synthetic proper orbital elements are @xmath56 au , @xmath57 ( kneevi and milani 2003 ) .
the synthetic proper eccentricity of @xmath58 is well below the npc family average , which is why it was never previously identified as a candidate parent for any of the nysa - polana complex .
however , the orbit of ( 495 ) eulalia is extremely close the the 3:1 mean motion resonance ( mmr ) with jupiter .
the location of this resonance is nominally at @xmath1 2.5 au where a body has an orbital period 1/3 that of jupiter .
the location of the resonance is also a function of the small body s orbital eccentricity , where larger eccentricities expands the location of the resonance in semimajor axis ( see wisdom 1993 , 1995 ) .
the boundary of the resonance itself is not precise , rather the chaoticity of an orbit increases as it approaches the resonance .
therefore the precise behavior of ( 495 ) eulalia over time can not be estimated trivially ( see also guillens et al .
2002 ) . to estimate the chaotic affects on ( 495 ) eulalia due to its location near the 3:1 mmr with jupiter
, its long - term evolution was modeled .
based on the published uncertainties for its current osculating orbital elements , we integrated 1000 clone particles ( created from the 6d hyper - ellipsoid in the osculating elements space compatible with today s orbit ) .
the clones were integrated using the swift_mvs code , which is part of the swift integration package ( levison and duncan 1994 ) .
all 8 planets were included in the simulation , and each clone was treated as a test particle ( feeling only the gravity of the planets ) .
the simulations spanned 500 myr ( with some tests to 4 gyr ) .
the specific behavior for any single particle at very short timescales ( kyr ) is dominated by the proximity to the 3:1 mean motion resonance ( mmr ) with jupiter via the slow circulation of the principal resonance angle @xmath59 ( where @xmath60 is the orbital frequency of an orbit and @xmath61 is the precession frequency of an orbit ; see fig . [
fig : resangle ] ) .
on longer timescales the interaction with @xmath62 secular resonance results in the corresponding resonance angle @xmath63 librating about @xmath64 with a period of @xmath65 ky ( where @xmath66 , @xmath67 and @xmath68 are the proper frequencies associated with the precession of the asteroid , jupiter and saturn respectively ) .
this interaction , together with other secular terms of the asteroid s orbit , is responsible for osculating eccentricity oscillations between 0.060.19 . to analyze long - term stability and orbital evolution the entire suite of particles
was considered with averaged orbital elements over 10 myr time periods as a proxy for proper elements .
each particle s averaged elements was sampled every 10 myr , creating a distribution mapping the time spent at different orbits - specifically at different eccentricities ( fig .
[ fig : diffuse ] ) .
the semimajor axis of the 1000 clones is quite constant with @xmath69@xmath70 au , while the average eccentricity diffuses between @xmath71=0.110.15 , with most of its time spent at @xmath71=0.12 . from the limited tests run to 4 gyr we find some clones
are lost to the 3:1 mmr ; at 2 gyr only 30% of the clones are lost and at 4 gyr only about 60% of the clones are lost .
so despite being chaotic , the gravitational - only orbit of ( 495 ) eulalia is surprisingly stable for long timescales .
the role of thermal effects on the dynamics of eulalia may be important given the putative old age of the family . at @xmath55
km its maximum drift rate over 1 gyr is around 0.005 au , which is larger than the diffusion found in semimajor axis in the long timescale integrations above .
also , given its precarious location very near the 3:1 mmr , a drift of 0.005 au could change its long - term evolution dramatically , including a slightly different orbit at the time of a possible family - forming impact .
a more realistic estimate of its potential drift rate depends heavily on its obliquity and various thermal properties .
its rotation rate and obliquity have been estimated at [email protected] hr and @xmath188@xmath19 ( binzel 1987 ) . increasing the accuracy of these measurements and estimating possible yarkovsky drift scenarios , along with the possible role of the yorp effect and secular spin - orbit resonances , is beyond the scope of this work .
the conclusion of this test is that when searching for family membership using an hcm model , the proper eccentricity of ( 495 ) eulalia should be tested at values between @xmath71=0.110.15 , as they are all possible given the asteroid s proximity to the 3:1 mmr . .
the top panel shows the slow ( relative to the orbital periods ) circulation of the principal resonance angle @xmath73 .
the bottom panel shows the much longer timescale interaction with the @xmath67 and @xmath68 frequencies with the argument @xmath74 librating about @xmath64 with a period of @xmath65 ky ( where short - period oscillations were eliminated by digital filtering ) . ]
bin is occupied @xmath1 13% of the time . ]
if ( 495 ) eulalia is the parent of a family caused by a collision , the family - forming collision could have happened when the asteroid had any of its allowable orbital eccentricities .
therefore a hierarchical clustering method search for possible family members associated with ( 495 ) eulalia , necessarily needs to be done varying its proper eccentricity .
family association was tested using the hierarchical clustering method ( see bendjoya and zappal 2002 and references therein ) .
hcm searches among asteroid s proper orbital elements , as it is assumed that proper orbital elements are stable on very long timescales , with the exception of the drift in semimajor axis due to the yarkovsky affect .
the algorithm relies on the `` standard metric '' of zappal et al .
( 1990 , 1995 ) to calculate a velocity difference between two asteroid orbits .
it then connects bodies falling within a cut - off velocity ( @xmath75 ) , where @xmath75 is typically @xmath76 m / s . however , the region surrounding ( 495 ) eulalia is particularly low - density , presumably due to its location directly adjacent to the 3:1 mmr .
even small yarkovsky drifts to higher semimajor axis would drive an asteroid directly into the resonance , thus eulalia presumably lost many of its closest neighbors long ago .
the hcm search here relied on the the wise - surveyed low - albedo asteroids , and has only 6702 objects in the database , as compared to over 100,000 imb asteroids with synthetic proper elements .
so , where typically a @xmath77 m / s is at risk of linking to a substantial fraction of the asteroid belt , here even @xmath78 m / s still only links asteroids with similar orbits . with ( 495 )
eulalia as the parent , with its @xmath79 , hcm tests were run varying @xmath75 between @xmath80 m / s .
( m / s ) , using the hcm family search algorithm .
each panel is a result using a different proper eccentricity for ( 495 ) eulalia .
the results for @xmath71=0.15 show a jump to large numbers of family members at the lowest velocity cutoff ( @xmath81 m / s ) , and the slow upward trend after that is typical of hcm analysis of asteroid families . ] the hcm results for @xmath71=0.14 and @xmath71=0.15 show the traits of the typical asteroid family behavior , with a sharp jump in asteroid family membership numbers at a specific @xmath75 , followed by a slow increase thereafter ( fig . [
fig : hcmall ] ) .
this is not surprising , as the low - albedo component of the nysa - polana complex has a mean eccentricity of @xmath71=0.152 .
neither case , however , shows the characteristic jump in family membership number after exceeding the critical velocity cutoff for the family , as seen in many examples of vokrouhlick et al .
( 2006a ) .
it is more similar to the results for the large and old family eos ( vokrouhlick et al .
the @xmath71=0.15 case has the lowest @xmath75 jump in family numbers , at @xmath82 m / s , and thus @xmath71=0.15 is the favored case .
one outcome from the initial hcm results is the inclusion of asteroid ( 142 ) polana in the family at @xmath83 m / s ( figs .
[ fig : euhcm ] and [ fig : euah ] ) .
it seems very clear from the results in fig .
[ fig : vshape ] that multiple families are overlapping and being selected by the hcm routine .
this highlights the problems with differentiating membership between the two families for the bodies within the yarkovsky envelopes of each .
the inclusion of this secondary family will pose problems in later sections when the bounding envelope of the `` v - shape '' distribution is used to calculate the age of the family .
in fact , by estimating the lower bound of the family in @xmath30 vs. @xmath32 space by the yarkovsky drift curves , we can eliminate numerous interlopers that belong to another family , rather than the low - albedo part of the nysa - polana complex .
the slope of the cumulative distribution of absolute magnitude @xmath30 value for the hcm - determined family members is sometimes used to help select viable ranges of @xmath75 from an hcm search .
this calculation was performed on the selected family at each increment in @xmath75 ( see fig . [
fig : sfdvel ] ) .
this calculation follows vokrouhlick et al .
( 2006a ) , by using the power - law of the form @xmath84 where @xmath85 .
the results show no clear trend , and rather a low @xmath86 compared to what is found for some young ( @xmath1100s myr ) families such as erigone @xmath87 ( vokrouhlick et al .
2006a ) and is closer to the entire asteroid belt population with @xmath88 ( ivezi et al .
young families can have a range of size frequency distributions with slopes sometimes dramatically different than the background population . over time
collisional evolution will change the size frequency distribution eventually driving it to similar values as the background ( morbidelli et al .
2002 ; bottke et al .
2005 ) . in this case , where there is no distinction in the size frequency distribution from the background
, this property can not be used to either secure its selection as a family or to make a statement about its age .
we adopt asteroid ( 495 ) eulalia as the parent of the low - albedo component of the nysa - polana complex - now referred to as the `` eulalia family '' .
its proper semimajor axis at @xmath89 au is adopted as the center of the family . in the next section
we can determine the extent and age of the family by using the members found in this hcm clustering search , and we will use the family as defined by the hcm criteria with @xmath90 m / s , which by visual inspection was preferable to the @xmath82 m / s selection due to larger numbers without increasing the size of the family in @xmath32 , @xmath71 , or @xmath91 .
m / s , where eulalia had an orbital eccentricity of @xmath71=0.15 , and the location of ( 495 ) is shown with a red square .
note that other large families , such as erigone have been removed prior to the hcm .
[ fig : euhcm],title="fig : " ] + as a function of their semimajor axis ( au ) .
this plot is done for a v@xmath92=120 m / s , where eulalia had an orbital eccentricity of @xmath71=0.15 and its location is shown with a red square .
the top panel are the members found in the hcm search , and the bottom panel shows the asteroids in the sample not selected as part of the family .
note that other large families , such as erigone have been removed prior to the hcm.,title="fig : " ] + ) using the hcm family search algorithm for the case where the proper eccentricity of eulalia is @xmath93 .
the bottom panel shows the exponent @xmath94 for the power law size frequency distribution for @xmath85 .
there is no discernable change in the size frequency distribution for different velocity cutoffs as is seen for younger families . ] the age of the family can be estimated by the spreading in semimajor axis of family members over time , which is a function of an asteroid s size , thermal properties and spin axis ( bottke et al .
the size or absolute magnitude of many main belt asteroids is well measured through the combination of recent wise results , years of optical surveys such as sloan digital sky survey ( sdss ) and past infrared surveys such as iras and wise ( some potentially important biases in the absolute magnitude values are discussed in section 4 . ) .
the thermal properties have been directly measured for only a few asteroids ( see delb et al .
2007 ) , but have been inferred for populations due to ensemble behavior for asteroid families with a range of ages ( bottke et al .
the evolution of the spin axis of a body is related to its thermal properties and the thermal yorp - effect , and this behavior , though still being investigated , is constrained by asteroid family investigations ( vokrouhlick et al .
2006a ; bottke et al .
similarly , the initial distribution of fragment orbits following an asteroid impact provides an uncertain initial condition preceeding the family spreading .
vokrouhlick et al .
( 2006a ) have previously employed a complex fitting routine that simultaneously solves for family age , yorp reorientation timescales and strength , and a size - dependent initial velocity dispersion . given the large extent of the eulalia family , that at least half of the family is lost into the 3:1 mmr , and its location overlapping another primitive family , we focus here on the family boundaries in order to estimate family age .
a more detailed method of the age determination , solving for yorp reorientation and the initial velocity disperion , will be presented in later works .
the @xmath34 distribution of asteroids was used to locate the center of the family , based on a cluster of similar values .
a bounding @xmath34 value for the family can be used to estimate an age for the family , based on the maximum yarkovsky drift timescales . using the @xmath34 value we find , @xmath95 where @xmath96 , and @xmath97 with @xmath98 being the maximum yarkovsky drift rate for a diameter @xmath99 ignoring any spin axis variation .
it will be useful to use @xmath100 , and combine them so that @xmath101 where @xmath102 is @xmath103 au / myr for a 5 km asteroid in the inner main belt with a density of 2.5 g @xmath104and thermal conductivity @xmath105 w m@xmath106 k@xmath106 ( taken from bottke et al .
the formulation uses a @xmath107 km , so the nominal drift rate ( in au / myr ) is , @xmath108 for an @xmath109 with @xmath107 km and a density of 1 g @xmath104 the rate is @xmath110 au / myr .
there is no published measurement of the density of any family members , so we selected our nominal value of 1 g @xmath104 for simplicity and the few measurements of c - type asteroids clustering around this value ; with ( 379 ) huenna at @xmath111 g @xmath104 ( marchis et al . 2008 ) , and ( 253 ) mathilde with @xmath112 g @xmath104 ( yeomans et al .
1997 ) are two with accurate estimates .
the drift rate is inversely proportional to the density , and thus this estimate can be a systematic source of error in our calculations .
now we can estimate the age based on the estimates for @xmath34 by , @xmath113 with an albedo @xmath1 0.052 for ( 495 ) eulalia . to calculate the age of the family we are interested in the boundary of the @xmath34 distribution , or the envelope of yarkovsky lines for @xmath30 as a function of @xmath32 .
we devised a fitting routine whereby the value of @xmath34 was varied and the ratio of asteroids with @xmath114 au was compared with @xmath115 au .
a strong contrast in numbers indicates the boundary of the family has been reached .
this was done for all the members selected by the hcm routine for @xmath116 m / s in steps of @xmath117 au ( where these bin sizes were selected via trial and error testing ) . due to the increasing number of asteroids at greater @xmath30 ( smaller sizes )
, we measured this ratio for three different size ranges , @xmath118 .
the average of the three ratios finds two peaks , at @xmath119 au and @xmath120 au ( fig .
[ fig : agefit ] ) . the smaller value of @xmath119 au is favored by the smaller asteroids , while @xmath120 au is favored by the larger asteroids .
this could be a real effect caused by thermal effects acting differently at small sizes , or simply confusion caused by interlopers .
the best fit boundaries are plotted in figs .
[ fig : testage10.5 ] and [ fig : testage9.2 ] . alone
, these values of @xmath34 would correspond to ages of @xmath121 and @xmath122 myr .
however , the calculation does not yet account for the initial fragment semimajor axis distribution produced immediately after the family forming event . the size - dependent velocity dispersion of the fragments due to the asteroid impact creates a distribution of @xmath37 preceeding any yarkovsky drift .
again , following vokrouhlick et al .
( 2006a ) , we can estimate size - dependent @xmath37 for collision outcomes with different velocity dispersions , where here we calculate another @xmath123 as the distance from the center of the family , @xmath124 where @xmath125 is the mean motion of the target asteroid and @xmath126 is the transverse velocity of the fragment and @xmath127 simply denotes other eccentricity - dependent terms .
we can use a very idealized equation , @xmath128 vokrouhlick et al .
( 2006a ) uses @xmath129 , where @xmath130 is the standard deviation of a directional component of a velocity field following an asteroid disruption .
the @xmath130 can be equated to @xmath126 in this simple estimation , and @xmath131 can be equated very roughly to the escape speed at the surface of the target body , which scales linearly with diameter .
so that , @xmath132 and using a mean motion for the target body , eulalia in this case , the @xmath123 can be calculated as a function of fragment size .
we have not yet estimated the total mass in the family or estimated the size of the parent of the family .
this calculation is made later in section [ s : agessfd ] , finding a parent body @xmath133 km . that is used as a guide here since the velocity dispersion should be similar to the parent body escape speed that scales linearly with the parent body size .
so we include calculations of the estimated size - dependent @xmath123 due to the initial collision event with @xmath134=50 m / s and @xmath134=150 m / s ( see the red lines in figs .
[ fig : testage10.5 ] and [ fig : testage9.2 ] ) .
since the @xmath123 from the initial fragment displacement follows the same @xmath135 relationship as the yarkovsky drift , we can relate the initial fragment distribution in terms of @xmath34 , and simply subtract off this @xmath34 and re - calculate the age from the total yarkovsky drift .
these distributions correspond to a maximum initial semimajor axis @xmath136 au and @xmath137 au , such that the best - fit bounding value of @xmath119 au , would simply be a @xmath138 au and @xmath139 au , which correspond to ages of @xmath140 and @xmath141 myr .
while for @xmath120 au , would correspond to total ages of @xmath142 and @xmath143 myr .
the relationship between these calculations are shown in fig .
[ fig : testage9.2 ] , where the black line is the @xmath123 due to yarkovsky drift from a center at @xmath89 au with an age of 1160 myr .
the red lines are the @xmath123 due to the fragment velocity dispersion of 50 and 150 m / s . from these calculations
we find the age of the family between 9001500 myr .
one important caveat in these simplistic age calculations is the role of obliquity changes due to the yarkovsky - okeefe - radzievskii - paddack ( yorp ) effect ( rubincam 2000 ; vokrouhlick and apek 2002 ; bottke et al .
the yorp effect acts by the asymmetrical reflection and re - emission of thermal radiation to affect the rotation state of an irregularly shaped body .
when the yorp effect changes the obliquity of an asteroid , the semimajor axis drift rate due to the yarkovsky effect will also change .
the preferred spin states are those that maximize the yarokovsky drift ( 0@xmath19 and 180@xmath19 ) , such that family members with initial obliquities resulting in slow yarkovsky drift are expected to evolve over time to a faster - drifting state .
large uncertainty exists in modeling the evolution of the yorp re - orientation of obliquities due to the yorp evolution of the rotation rate as well and complications at very high and very low spin rates .
an estimate of re - orientation timescales is @xmath1 300600 myr for @xmath1 5 km objects ( vokrouhlick and apek 2002 ; apek and vokrouhlick 2004 ) .
objects well above this size are then likely to have not experienced a full yorp reorientation since the formation of the family .
any dramatic effects due to yorp cycles for this family would therefore not be significant for asteroids larger than @xmath144 to be conservative
. one possible tracer of @xmath134 would be family members reaching the _ other _ side of the 3:1 resonance ( for @xmath145 au ) .
given the center of the family and the yarkovsky drift boundaries from the asteroids with @xmath146 au , we selected the subset of asteroids with @xmath147 au that would host any family members .
there is no clear evidence of a high density of asteroids around the yarkovsky lines ( fig .
[ fig : eul31 ] ) , confirmed by the distribution of @xmath34 values around an @xmath148 au . the lack of existence of , or inability to detect , members on the far side of the 3:1 mmr limits the @xmath134 for the fragment distribution .
very large velocity distributions would place many asteroids on orbits beyond the 3:1 mmr , where they could be observable today . for a conservative boundary of the 3:1 mmr of @xmath149 au at the relevant eccentricities of @xmath28 a @xmath150 m / s would only place some objects of @xmath151 beyond the resonance , while at @xmath152 m / s only some with @xmath153 .
while a substantial mass of the family can be contained in bodies of these sizes and smaller , it also implies a large fraction of the mass never made it past the 3:1 mmr .
thus the @xmath134 estimates used in this work , presented to bracket possible ranges of fragment velocity dispersion , also are consistent with the lack of a clear family signature across the 3:1 mmr . ) au over the number with ( @xmath115 ) au , where a large value represents a strong density contrast at the tested value of @xmath34 .
this test was done for the family center at @xmath154 au , and combined ratio values for three different ranges of absolute magnitude @xmath30 : the dotted line is @xmath155 , the short dashed line is for @xmath156 , and the long dashed line is for @xmath157 .
the dark solid line is the average of the three , with the value of @xmath158 au is slightly preferred over the value of @xmath159 au due to the strong preference at smaller asteroids ( @xmath157 ) for the former fit .
[ fig : agefit ] ] au is the solid black line .
the red lines represents a simple initial dispersion for a velocity dispersion of fragments for @xmath160 m / s and @xmath161 m / s by @xmath162 where @xmath125 is the mean motion .
[ fig : testage10.5],title="fig : " ] + au is the solid black line .
the red lines represents a simple initial dispersion for a velocity dispersion of fragments for @xmath160 m / s and @xmath161 m / s by @xmath162 where @xmath125 is the mean motion .
[ fig : testage9.2],title="fig : " ] + @xmath1630.1 with @xmath164 and inclination @xmath165 with semimajor axis @xmath166 au . the yarkovsky line for the family boundary at @xmath119 au is the solid black line .
the red lines represents a simple initial dispersion for a velocity dispersion of fragments for @xmath160 m / s and @xmath161 m / s .
this selection of asteroids , matching the orbital properties of the detected family but on the other side of the 3:1 mmr show no sign of a family signature .
[ fig : eul31],title="fig : " ] +
the size frequency distribution of the family can be estimated from the objects selected in the hcm routine , using the power - law of the form @xmath167 . in section 2.4 a @xmath94 for only the range of @xmath85
was compared to the background , now we consider the size distribution of all prospective family members , using it to estimate the mass of the entire family
. simply summing the mass of all observed bodies is a quick estimate of the original parent s size .
however , we are only now detecting half of the family , as the other half were launched into , or very quickly entered , the 3:1 mmr and were lost .
there is another factor of two , as only bodies launched to smaller semimajor axis survive the initial displacement from the parent , but then any asteroid with an initial outward yarkovsky drift will very quickly drift into the 3:1 mmr .
similarly , the dataset is missing objects as the wise survey is not complete , even at large sizes .
a comparison of the wise survey and databse of known asteroids , for large asteroids in the inner main belt , found that it detected @xmath180% of all asteroids with @xmath168 , with that fraction dropping to 55% for @xmath12 .
combined with the previous family scaling factor of 4 , scaling by 5 to account for the @xmath180% wise detection efficiency is closer , though this scaling factor should grow at smaller sizes .
finally , the numbers are scaled at each size according to the wise detection efficiency .
it should be noted that collisional evolution could play a major role in altering the family sfd over time , and this should be considered in future models .
the integrated size for the selected family members from the @xmath169 m / s hcm routine is @xmath170 km ( the black line in fig .
[ fig : sfdtest ] ) . note that this includes a limiting @xmath30 vs. @xmath32 criteria to eliminate interlopers .
when a factor of 4 for the loss of 3/4 of the family to the 3:1 mmr is accounted for ( except the largest body , parent eulalia ) the estimate increases to @xmath171 km ( blue line in fig .
[ fig : sfdtest ] ) .
finally , a full factor of 5 , accounting for the survey limitations , equals a body with @xmath172 km ( red line in fig .
[ fig : sfdtest ] ) . the final correction for size - dependent wise detection efficiency increases the size to @xmath173 km ( green line in fig .
[ fig : sfdtest ] ) . since only the largest bodies are detected with very high efficiency , an alternative method to estimate total family mass is needed .
an alternative method relies on the size frequency distributions for asteroid families derived from smoothed particle hydrodynamics ( sph ) simulations of catastrophic asteroid impacts that are coupled to @xmath174-body gravitational simulations of their reaccumulation ( durda et al . 2007 ;
benavidez et al .
this method matches size frequency distributions between simulations and observations at large fragment sizes , and then uses the simulation to estimate how much unobservable material may also belong to the family .
one sph and @xmath174-body asteroid collision simulation was selected from the suite of simulations by benavidez et al .
it used a 100 km basalt rubble pile target with a 3 km / s 34 km impactor hitting at a 15@xmath19 angle .
the largest remnant in the simulation outcome is @xmath175 km , slightly smaller than ( 495 ) eulalia ( @xmath176 km ) . however , the shape / slope of the curve is a reasonable match for the eulalia family .
these simulations are based on a 100 km target , and a match in slope must be coupled with a horizontal shift to scale for the original parent size . shifting the size distribution by a factor of 1.6 is equivalent to increasing the target body size by the same factor , thereby creating a size frequency distribution for a @xmath1160 km target .
given the relative close match in both slope and horizontal scaling , this suggests that the simple integrated size estimate of @xmath177 km is close to the sph progenitor target scaled to @xmath178 km , and the latter serves as a resonable upperbound given the uncertainties with this particular family .
the wise mission facilitated the analysis of this family while also providing some physical characterization of its constituents .
other public databases and catalogs of observations are available to assess the physical properties of the family members . here
we summarize the available data . from the wise mission ,
there are calculated albedos for each object in the dataset used here .
we compare the albedo distribution of the bodies found by the hcm routine for @xmath116 m / s , with the @xmath119 au boundaries .
we find only a slightly tighter cluster of albedo values in this sample compared to the background ( fig .
[ fig : albedo ] ) . given that the entire sample is less than @xmath13@xmath179 , and typical formal errors in albedo measured by wise are @xmath10.008 for @xmath180 and @xmath10.018 for the entire sample , a clear cluster in albedo was unlikely .
the bulk of the asteroids included in the family are small objects , the distribution may be dominated by objects with larger errors in their measurement due to their small sizes . the 128 objects with @xmath180 show no clear distinction from the entire sample of 1563 selected for these boundaries of the family ( fig .
[ fig : albedo ] ) .
pravec et al .
( 2012 ) found that there is divergence between @xmath30 values found in the mpc database and those measured through an observational campaign , where the mpc values are systematically too small , reaching a maximum mean offset of -0.4 to -0.5 around @xmath181 .
this would lead to albedo estimates that are slightly too large and becomes significant and systematic at @xmath30 above 10 .
the wise calculated albedos relied on the @xmath30 values in the mpc database , and this therefore could account for a source of error as well .
@xmath182 0.1 ) , while the black outline histogram shows the distribution for the selected eulalia family members inside the @xmath183 au boundary .
the dotted line is the albedo distribution for only the subset of the family members with @xmath180 , accounting for 128 of the 1563 total members plotted here .
[ fig : albedo ] ] the sloan digital sky survey ( sdss ) observed 471,000 moving objects , for which 220,000 observations were matched to 104,000 different asteroids .
the photometry was done in 5 wavelengths and is a valuable tool for establishing taxonomic relationships for large groups of asteroids .
the five wavelengths can be combined in numerous ways in order to capture the many possible features found in visible spectra of different asteroid types , but a common method is to compare the @xmath184 colors as a function of @xmath185 ( where the @xmath91 filter is centered at 769 nm , the @xmath186 at 925 nm and @xmath185 is the derived first principal component for the distribution of asteroid colors in the sdss @xmath187 vs. @xmath188 color - color diagram ) .
the former measures the depth of the 1 micron silicate absorption bands mainly due to olivines and pyroxenes , the latter is a proxy for the slope of the spectra ( see fig 3 . in parker
et al . 2008 ) .
in this space , primitive asteroids of the c- and b - types are typically found with negative slope ( @xmath189 ) .
here we matched 668 asteroids from selected family bounded by the @xmath120 au that were also measured by sdss , and have plotted them on the same axes and color scheme as parker et al .
as expected , the family is tightly clustered in the region dominated by primitive asteroids ( fig .
[ fig : sdss ] ) . without more significant quantitative analysis ,
it is clear that cluster is actually much tighter than that found for all primitive bodies .
however , it is also clear that the cluster is not tight enough to use sdss colors as a tool to differentiate members from different family associations .
au ) and all the sdss observed asteroids plotted as @xmath184 colors as a function of the slope @xmath190 where the @xmath91 filter is centered at 769 nm , the @xmath186 at 925 nm and @xmath185 is the derived first principal component for the distribution of asteroid colors in the sdss @xmath187 vs. @xmath188 color - color diagram .
the color scale is that used by parker et al .
( 2008 ) that associates colors with locations on this color - color plot .
[ fig : sdss ] ] asteroid ( 495 ) eulalia was observed on 11 june 2007 at the nasa irtf with the spex instrument between 0.72.5 microns ( fieber - beyer et al .
2008 ; fieber - beyer 2011 ) .
there are no clear features for the ( 495 ) eulalia spectral reflectance , with only a minimal upturn in the spectra beyond 2.0 microns ( fig .
[ fig : specdiversity ] ) . as the parent of the family
this spectra is potentially valuable in finding differences between members of it and the neighboring background .
however , flat and featureless spectra are somewhat common for low - albedo objects , which does not aid the determination of family membership for an asteroid with unknown affiliation .
there exists numerous databases to search for spectra of other family members , and two visible spectra ( asteroids 1076 and 2509 ) were found among the small main - belt asteroid spectroscopic survey ( smass ; xu et al .
1995 , 1996 ) and two ( asteroids 495 and 3999 ) in the second phase of that survey ( smassii ; bus and binzel 2002 ) .
an additional two objects found in the family using the @xmath120 au boundaries ( see fig . [ fig : testage10.5 ] ) were observed in smass database ( asteroids 1768 and 2809 ) .
no objects were found in the s3os2 database ( lazzaro et al .
2004 ) or in the pds release of multiple irtf surveys ( bus 2011 ) . the data in longer , near - ir , wavelengths is even more sparse .
as mentioned above ( 495 ) eulalia , was observed with the irtf spex instrument as part of a campaign to study objects near the 3:1 mmr . as part of the wider asteroid survey ,
the mit - uh - irtf joint campaign for neo reconnaissance ( http://smass.mit.edu/ ) , there are near - infrared observations of asteroids ( 1076 ) viola and ( 1768 ) appenzella .
( 1076 ) viola , which belongs to the eulalia family , is plotted with ( 495 ) eulalia ( fig .
[ fig : specdiversity ] ) and shows similar morphology throughout , in particular they have a similar slope beyond 2.0 microns .
asteroid ( 1768 ) appenzella is within the yarkovsky boundaries for both families , and is thus plotted alone .
it shows an upturn starting around 1.0 micron , with a much shallower slope than found for ( 142 ) polana ( discussed below ) .
for comparison ( clark et al .
[ fig : specdiversity ] ]
an important aspect to the analysis of the eulalia family is the possibility of a larger , older family with similar orbits and similar low albedo . here
we investigate and characterize this proposed family with some of the tools used above .
this family is even more extended and diffuse and much of the analysis is beyond the scope of this current work . beyond the clear correlation of the asteroids in the eulalia family as seen in fig .
[ fig : vshape ] and analyzed in the previous sections , there appears to be a possible second family in the same dataset .
the lack of asteroids with @xmath191 2.3 au and @xmath192 14 is clear and defies even the expected dropoff in the number of asteroids with decreasing semimajor axis due to the inclination - dependent effects of the @xmath33 resonance and eccentricity limits of mars crossing orbits .
the high density of objects with @xmath193 and @xmath23 au @xmath194 au has the characteristic `` ear '' of an asteroid family seen at higher semimajor axis for the eulalia family due to the yarkovsky semimajor axis drift .
this possible family suffers from the same , and even amplified , difficulties found in analyzing the eulalia family ; a center near a powerful resonance , overlapping with another family and possibly a very old age . in many tests below we will remove the eulalia family from the data set ( using previously defined boundaries ) in an attempt to measure only the possible background family
. given the very diffuse nature of this possible family with an unknown parent , we reverse the procedure used before , and begin with an hcm test to make sure there is , in fact , an orbital correlation among the high density region of asteroids . given the very low density of large objects , this hcm test was centered on the heart of the visible `` ear '' of the family located at @xmath195 au @xmath196 au and @xmath197 .
we simply selected a few objects in this orbital range and ran the hcm at varying @xmath75 checking to see whether any clump would acquire the characteristic shape in @xmath30 vs. @xmath32 space .
the first test was done with asteroid ( 2446 ) lunacharsky .
this asteroid has proper orbital elements of @xmath198 au , @xmath199 , @xmath200 , and is also one of the few in this region with published visible and near infrared spectra .
the hcm results for this body show that there is a highly diffuse and extended association with few family members found below a velocity cutoff of @xmath201 150 m / s ( fig . [ fig : lunahcm ] , bottom ) .
however , the association linked by hcm at @xmath201 180 m / s and larger shows the distinct signs in the @xmath30 vs. @xmath32 space of a correlated family and the family size increases quickly .
the largest bodies linked to this association are asteroids ( 142 ) polana and ( 112 ) iphigenia with proper elements @xmath202 au , @xmath203 , @xmath204 and @xmath205 au , @xmath206 , @xmath207 respectively ( iphegenia is linked at @xmath201 180 m / s and polana at @xmath201 190 m / s ) .
the orbital element distribution of the selected asteroids is quite large in eccentricity , spanning @xmath208 , but more tightly clustered in inclination between @xmath209 .
each of these large asteroids are in distinct clusters determined by their eccentricity , where ( 142 ) polana is in the midst of the high eccentricity grouping and ( 112 ) iphigenia with the smaller low eccentricity grouping ( fig .
[ fig : lunaaei ] ) .
m / s , which shows the distinct sign of a yarkovsky `` v - shape '' .
there is a clear over density of asteroids in the region of the eulalia family , suggesting that our conservative attempt to remove the eulalia family did not capture all family members .
[ fig : lunahcm ] ] ] we focus on ( 142 ) polana and ( 112 ) iphigenia due to their size and the suggestions from the @xmath30 vs. @xmath32 plot that they are near the center of the family .
differentiating them is difficult due to these similarities , other than ( 142 ) polana being embedded in a larger and denser grouping in eccentricity .
when these groups are separated and plotted as @xmath30 vs. @xmath32 it is clear that ( 142 ) polana is associated with the extended family structure ( fig .
[ fig : polseparate ] ) .
meanwhile ( 112 ) iphigenia is associated with objects with larger @xmath30 and larger @xmath32 , that are near the highest - density region of the eulalia family .
it is possible that if the eulalia family was linked with a larger @xmath75 in the hcm routine that many of these objects would be added to the eulalia family .
thus , the extended family here appears to be associated with asteroid ( 142 ) polana , and we use its proper semimajor axis @xmath202 au as the nominal center of the family . using the same fitting routine as before , we then measure a best - fit at @xmath210 au ( fig .
[ fig : pol16.9 ] ) .
this fit was substantially easier due to the dramatic density fall - off , and the measured best - fit @xmath34 value is nearly twice as large as for the eulalia family .
this @xmath34 directly correlates to @xmath211 myr , not accounting for the initial velocity dispersion of the fragments following the collision .
we can use the same initial velocity dispersions of @xmath160 m / s and @xmath161 m / s , despite not having a solid estimate of parent body size , which puts the age range of the polana family at @xmath120002500 myr .
however , uncertainty about yorp reorientation of spin axes adds age uncertainty which becomes increasingly important at increasingly larger asteroid sizes for older families .
this adds substantial uncertainty to the upper boundary of this age estimate , and the conservative estimate is simply an age greater than 2000 myr . ]
au is the solid black line .
the red lines represents a simple initial dispersion for a velocity dispersion of fragments for @xmath160 m / s and @xmath161 m / s by @xmath162 where @xmath125 is the mean motion .
[ fig : pol16.9 ] ] given the low density of asteroids in the wise dataset at low @xmath32 ( @xmath212 au ) and @xmath30 ( @xmath213 ) , it is fair to question if the characteristic shape of the new polana family is not simply a statistical fluke caused by decreasing asteroid density near the @xmath33 secular resonance .
specificially , if the gaping hole in the sample with @xmath214 au and @xmath215 is due simply to statistical flucuation , then the boundary of the new polana yarkovsky `` v - shape '' may simply be an artifact and not the edge of an asteroid family .
a simple test of the significance of this was to take the entire distribution of asteroids in our dataset and randomly re - distribute them while preserving the orbital and size distributions .
more simply , the values of @xmath32 , @xmath71 , @xmath91 and @xmath30 were each separately randomized 1000 times .
a test of number of asteroids with @xmath214 au and @xmath215 was performed for each .
the wise dataset has only 4 asteroids in this region , while the average number found in this region was 15.7 , and the smallest number was 5 for all 1000 iterations ( fig . [
fig : fooled ] ) .
this test did not specifically test if the v - shape was a fluke , but rather whether the stark absence of asteroids at lower @xmath30 was random . finding that it is clearly significant suggests that the density contrast seen is related to actual correlated family and supports the existence of the new polana family .
one alternative explanation for the lack of large low - albedo asteroids inside of @xmath1 2.3 au is simply that this is the innermost edge to c - type asteroids in the asteroid belt .
it is long known that primitive bodies predominate in the outer asteroid belt ( gradie and tedesco 1982 ) , and some recent work has investigated various means to `` implant '' primitive bodies in the asteroid belt from more distant regions of the solar system ( levison et al .
2009 ; walsh et al . 2011 ) .
levison et al . ( 2009 ) focus on capture of d- and p - type asteroids during the solar system instability associated with the late heavy bombardment , and they find a good match to the distribution of these taxonomic classes , with the inner edge of both populations around @xmath1 2.6 au .
walsh et al .
( 2011 ) focused on the origin of the c - class populations that are possibly scattered into the asteroid belt much earlier in solar system history .
this work shows a steep decline in primitive asteroid numbers at decreasing semimajor axis , and theoretically could be responsible for the absence of large primitive bodies inside of 2.3 au .
au and @xmath215 for the 6702 asteroids ( observed by wise with @xmath13 @xmath216 ) re - distributed with randomized @xmath32 , @xmath71 , @xmath91 , and @xmath30 distributions .
the actual data had only 4 objects in this @xmath32 vs. @xmath30 range ( indicated by the vertical line ) and is outside the bounds of the distribution found in the monte carlo test .
[ fig : fooled ] ] the albedo distribution , sdss colors and visible spectra of objects in the family regions are very similar and not useful to distinguish between families .
therefore any available spectra - especially in the near - infrared - may be able to differentiate family membership . there is publised near - ir spectra for one asteroid , ( 2446 ) lunacharsky ( de len et al .
2012 ) , that could be a member of the proposed polana family , and it is plotted with ( 142 ) polana , the candidate parent ( fig .
[ fig : specdiversity ] , also see fig . 4 in de len et al .
both objects show similar positive slopes beyond 1.01.2 micron , similar negative slopes in the visible and generally very similar morphologies .
asteroid ( 495 ) eulalia shows no clear feature at 1.0 or 1.2 micron in contrast to ( 142 ) polana .
there is also a clear slope difference between the two family parents beyond 1.0 micron , which has recently been studied as a diagnostic tool by de len et al .
( 2010 ) and ziffer et al . ( 2011 ) to differentiate primitive asteroids with otherwise very similar featureless visible spectra .
the hint of distinction between family parents , including additional candidate family members , is tantalizing for the possibility that near - infrared spectra could distinguish membership between the two families for the numerous smaller bodies that fall within the yarkovsky boundaries of both .
similarly , this is enticing as a possible tool to distinguish a possible origin for primitive neos that have likely originated in this region of the asteroid belt , especially future space - mission targets ( for example , see panel * d * of figure [ fig : specdiversity ] that displays the spectrum for neo 1999 rq@xmath15 , the baseline target for nasa s osiris - rex sample return mission ) .
the structure and past evolution of the families in the inner main belt are critically important in order to understand the history and evolution of the asteroid belt as well as the delivery of asteroids to neo orbits .
we have found that the primitive population in the inner main belt is dominated by a few very old and large families .
the low - albedo component of the nysa - polana complex is an asteroid family parented by asteroid ( 495 ) eulalia .
the family s age is between 9001500 myr , and the parent body had a size between 100165 km .
we have also found evidence for the existence of an older and larger family centered on ( 142 ) polana , dubbed here the `` new polana '' family .
this family is parented by ( 142 ) polana and formed more than 2000 myr ago . in light of this work
, it appears that the major contributors of primtive neos could be the eulalia and new polana family .
our dataset is limited to @xmath217 , and therefore we can not directly observe how the smallest members ( @xmath218 ) of the family are behaving .
we can , however , extrapolate based on the yarkovsky curves , and estimate the @xmath30 value where the yarkovsky curve crosses a major resonance . asteroids with a larger @xmath30
can reach that resonance by simple yarkovsky drift , while those with smaller @xmath30 can not .
this method ignores complications due to yorp cycles , or other effects such as collisions , that may have a different size - dependence than the yarkovsky effect
. in particular , we are interested in where the curves cross the @xmath33 resonance , as it is efficient at delivering neos . the new polana family , being older , is delivering asteroids up to @xmath219 , while the eulalia family is only now contributing bodies of @xmath220 .
each family has also contributed a large fraction of their total families into the 3:1 mmr .
most of these asteroids would have been directly placed into the resonance immediately following the family formation event , and therefore would have resulted in more of an rapid influx , rather than the slow and measureable contribution via the @xmath33 resonance .
further dynamical models beyond the scope of this work may also help to better discern the flux from each family to neo orbits over time via both the @xmath33 and 3:1 mmr .
one implication of the location of both families near the 3:1 mmr , is that nearly all surviving family members have drifted inward from the family center over time .
this inward drift requires a retrograde spin sense , which is a property that can be determined with lightcurve observations .
it is possible to change the spin sense of an asteroid , by either collisions or yorp reorientation .
both families show an absence of small objects near the center of the family , suggesting that most have continued their inward drift without a substantial number suffering a spin - axis reorientation .
this supports the implication that most members of both families should be retrograde rotators .
each of the three proposed neo space mission targets ( 1999 rq@xmath15 , 1999 ju@xmath20 and 1996 fg@xmath20 ) are almost certainly ( @xmath221% ) delivered from the inner region of the main asteroid belt , following the well - studied dynamical pathway from the main belt to neo - orbits ( bottke et al . 2000,2002 ; campins et al .
2010 ; walsh et al .
more specifically , they appear to have come from the @xmath33 resonance at 2.15 au , which is the dominant supplier of neos ( bottke et al .
2002 ) , and their current low inclination is indicative of similarly low inclination orbits in the main belt . though entire works are dedicated to understanding the origin of these specific objects ,
the two families studied here are likely the dominant sources of primitive neos .
these conclusions have been reached previously in works by campins et al .
( 2010 ) finding the nysa - polana complex as the likely source of 1999 rq@xmath15 ( now probably from the eulalia family ) , or gayon - markt et al .
( 2012 ) finding that the background ( now largely belonging to the new polana family ) is an equal or greater contributor of primitive neos .
a major part of this work was the description of relatively new techniques to detect and analyse very diffuse and old asteroid families . despite significant previous work on these topics ,
the discovery of both families was greatly aided by the use of the wise albedo data , which allowed specific analysis of low - albedo objects . beyond the advantage of focusing on asteroids with similar physical properties
, it also removed the `` noisy '' background of high - albedo asteroids which otherwise would have convoluted the hcm tests ( resulting in the hcm linking the huge chunks of the asteroid belt ) .
thus the slew of new data allows a pre - selected subset of the asteroid belt from which to detect possible family correlations .
the possible pitfalls of these technique are that the portion of the asteroid belt selected could be biased or incomplete .
asteroids not included in the subset of data , those not observed by wise for example , need to be tracked down in further iterations of the work to build a full list of family members .
though this work attempted to search for all possible parent members using numerous databases , ( 495 ) eulalia required special dynamical care to be recognised as the parent , a task not realistic for a larger survey of very old families . finally , there are potentially vast implications in assigning much of what was considered to the be the background of the asteroid belt to an older family , the new polana family .
while this family was particularly difficult to find , even while scouring a small subset of the entire database of asteroids , it highlights what might still be lurking in the data .
how much of the background is actually part of old and diffuse families ? if we continue with this exercise and continue to discover old and diffuse families will there be any background left , and what does this tell us about the origin of the asteroid belt ? certainly , the answers to these questions are beyond the scope of this project , but hopefully the flood of data from current and future spacecraft ( the esa gaia mission for example ) and larger ground - based surveys might help to answer these .
kjw would like to thank nlsi for funding , and also the gaia - great program of the european science foundation for funding a visit to nice .
md thanks the centre national detudes spatiales ( cnes ) for financial support .
dsl and wfb thank the nasa new frontiers program for supporting this research ( contract nnm10aa11c ) .
dv was supported by the grant agency of the czech republic through grant 205/08/0064 .
we thank alex parker for the sdss color palette .
benavidez , p. g. , durda , d. d. , enke , b. l. , bottke , w. f. , nesvorn , d. , richardson , d. c. , asphaug , e. , merline , w. j. , 2012 . a comparison between rubble - pile and monolithic targets in impact simulations : application to asteroid satellites and family size distributions icarus 219 , 57 - 76 .
binzel , r. p. , harris , a. w. , bus , s. j. , burbine , t. h. , 2001 .
spectral properties of near - earth objects : palomar and irtf results for 48 objects including spacecraft targets ( 9969 ) braille and ( 10302 ) 1989 ml .
icarus 151 , 139 - 149 .
binzel , r. p. , rivkin , a. s. , stuart , j. s. , harris , a. w. , bus , s. j. , burbine , t. h. , 2004 . observed spectral properties of near - earth objects : results for population distribution , source regions , and space weathering processes .
icarus 170 , 259 - 294 .
bottke , w. f. , rubincam , d. p. , burns , j. a. , 2000 . dynamical evolution of main belt meteoroids : numerical simulations incorporating planetary perturbations and yarkovsky thermal forces .
icarus 145 , 301 - 331 .
bottke , w. f. , morbidelli , a. , jedicke , r. , petit , j .-
m . , levison , h. f. , michel , p. , metcalfe , t. s. , 2002 .
debiased orbital and absolute magnitude distributions of the near - earth objects .
icarus 156 , 399 - 433 .
bottke , w. f. , durda , d. d. , nesvorn , d. , jedicke , r. , morbidelli , a. , vokrouhlick , d. , levison , h. f. , 2005 . linking the collisional history of the main asteroid belt to its dynamical excitation and depletion .
icarus 179 , 63 - 94 .
cloutis , e. a. , hudon , p. , romanek , c. s. , bishop , j. l. , reddy , v. , gaffey , m. j. , and hardersen , p. s. , 2010 .
spectral reflectance properties of ureilites .
meteoritics and planetary science 45 , 1668 - 1694 .
de len , j. , moth - diniz , t. , licandro , j. , pinilla - alonso , n. , campins , h. , 2011 .
new observations of asteroid ( 175706 ) 1996 fg3 , primary target of the esa marco polo - r mission .
530 , l12 .
emery , j. p. , fernndez , y. r. , kelley , m. s. , hergenrother , c. , ziffer , j. , lauretta , d. s. , drake , m. j. , campins , h. , 2010.thermophysical characterization of potential spacecraft target ( 101955 ) 1999 rq36 . lunar and planetary institute science conference abstracts , 41 , 2282 .
fieber - beyer , s. k. , gaffey , m. j. , reddy , v. , 2008 .
compositional and dynamical studies of asteroids located in / near the 3:1 kirkwood gap : 495 eulalia a first step .
lunar and planetary institute science conference abstracts 39 , 1149 .
hiroi , t. , zolensky , m. e. , pieters , c. m. , lipschutz , m. e. 1996 , thermal metamorphism of the c , g , b , and f asteroids seen from the 0.7 micron , 3 micron and uv absorption strengths in comparison with carbonaceous chondrites meteoritics and planetary science 31 , 321 - 327 .
hiroi , t. , jenniskens , p. , bishop , j. l. , shatir , t. s. m. , kudoda , a. m. , shaddad , m. h. 2010 , bidirectional visible - nir and biconical ft - ir reflectance spectra of almahata sitta meteorite samples meteoritics and planetary science 45 , 1836 - 1845 .
morbidelli , a. , jedicke , r. , bottke , w. f. , michel , p. , tedesco , e. f. 2002 . from magnitudes to diameters : the albedo distribution of near earth objects and the earth collision hazard .
icarus 158 329 - 342 .
nesvorn , d. , ferraz - mello , s. , holman , m. , morbidelli , a. 2002 .
regular and chaotic dynamics in the mean - motion resonances : implications for the structure and evolution of the asteroid belt .
w. f. bottke jr . , a. cellino , p. paolicchi , and r. p. binzel ( eds ) , asteroid iii , university of arizona press , tucson , pp . 379 - 394 . | the so - called nysa - polana complex of asteroids is a diverse and widespread group studied by cellino et al .
( 2001 , 2002 ) as a dynamically linked asteroid family .
it carries the name of two asteroids because it appears to be two overlapping families of different asteroid taxonomies : ( 44 ) nysa is an e - type asteroid with the lowest number in the midst of a predominantly s - type cluster and ( 142 ) polana is a b - type asteroid near the low - albedo b- and c - type cluster . the latter has been shown to be a very important source of primitive near earth asteroids . using the data from the wide - field infrared survey explorer ( wise ) mission we have re - analyzed the region around the nysa - polana complex in the inner main belt , focusing on the low - albedo population .
( 142 ) polana does not appear to be a member of the family of low - albedo asteroids in the nysa - polana complex .
rather , the largest is asteroid ( 495 ) eulalia .
this asteroid has never before been linked to this complex for an important dynamical reason : it currently has a proper eccentricity slightly below the range of most of the family members .
however , its orbit is very close to the 3:1 mean motion resonance with jupiter and is in a weak secular resonance .
we show that its osculating eccentricity varies widely ( @xmath0 ) on short timescales ( @xmath11 myr ) and the averaged value diffuses ( between @xmath2 ) over long timescales ( @xmath1100 myr ) .
the diffusive orbit , low - albedo , taxonomic similarity and semimajor axis strongly suggests that despite its current proper eccentricity , ( 495 ) eulalia could have recently been at an orbit very central to the family .
hierarchical clustering method tests confirm that at an eccentricity of @xmath30.15 , ( 495 ) eulalia could be the parent of the family .
the `` eulalia family '' was formed between 9001500 myr ago , and likely resulted from the breakup of a 100160 km parent body .
there is also compelling evidence for an older and more widespread primitive family in the same region of the asteroid belt parented by asteroid ( 142 ) polana .
this family , the `` new polana family '' , is more extended in orbital elements , and is older than 2000 myr . |
the gravitino is the most important prediction of supergravity ( sugra ) @xcite .
its mass is expected to be in a wide range , @xmath11 ev@xmath12 tev , depending on the mediation mechanism of supersymmetry ( susy ) breaking effects . in particular , the gravitino with a mass of @xmath13 mev@xmath14 gev is very interesting in the sense that it is most likely the stable and lightest susy particle ( lsp ) and can be a candidate of cold dark matter in the universe . moreover ,
if the gravitino mass lies between a few gev and a few 10 s gev and the next lsp ( nlsp ) is a charged scalar lepton ( @xmath0 ) such as stau ( @xmath15 ) , the sugra can be tested in high - energy collider experiments @xcite .
the key observation is that the gravitino mass @xmath16 , the stau mass @xmath17 and the stau lifetime @xmath18 could be measured in high precision if @xmath16 is of @xmath19 gev ( cf .
then the planck scale @xmath20 is extracted through the relation , @xmath21 .
this independent determination of @xmath20 will provide us with a crucial test of the sugra . for the gravitino mass in such an interesting region
as discussed above , the stau has necessarily a long lifetime .
for example , it is of order one month for @xmath22 gev and @xmath23 gev .
however , such a long - lived particle is potentially dangerous in cosmology .
if it decays after the big - bang nucleosynthesis ( bbn ) , the decay products destroy the light elements and ruin the success of the sbbn ( standard bbn ) @xcite .
more seriously , if the long - lived particle is negatively charged as @xmath1 , it forms bound states together with positively charged nuclei , which leads to an enhancement of some nuclear reaction rates .
that is , @xmath1 plays as a catalyzer of nuclear fusion ( catalyzed bbn or cbbn in short ) and affects cosmological nuclear abundances . produced in accelerators
may be of practical use as a catalyzer for d - d and d - t fusions similar to the muon catalyzed fusion @xcite . ] in a recent article @xcite , pospelov has argued that too much @xmath4 is produced through the cbbn if the lifetime of @xmath1 is long enough , @xmath24 sec , and its abundance is large enough , @xmath25 , where @xmath26 and @xmath27 are the number density of @xmath1 and the entropy density of the universe , respectively .
bound states have been considered in @xcite . ]
the production rate of @xmath4 estimated in ref .
@xcite is based on a naive comparison of the standard process @xmath28 and the new catalyzed process @xmath29 where @xmath30 is the 1s coulomb bound state of @xmath6 and @xmath1 .
it is assumed that the standard process is dominated by the e2 transition induced by the interaction @xmath31 with @xmath32 being the quadrapole operator .
then , by applying the same interaction to the photonless process with @xmath33 replaced by the coulomb field associated with @xmath1 , it was concluded that the astrophysical @xmath8-factor of the new process at zero incident energy is about 0.3 mev barn .
however , the assumptions adopted in the above estimate do not have firm ground : for example , the angular momentum of the initial @xmath6-@xmath7 system in the catalyzed process is dominated by zero , @xmath34 , while the angular momentum of the initial system in the standard process is dominated by @xmath35 @xmath36 for the e2 ( e1 ) transition .
such a kinematical difference invalidates the use of the sbbn process to estimate the cbbn process .
furthermore , the quantum interplay between the @xmath37 nuclear fusion and the tunneling of @xmath7 through the coulomb barrier plays an important role in low - energy catalyzed fusion and can not be treated in a perturbative manner .
the purpose of this paper is to solve the shrdinger equation for the three - body system ( @xmath38 , @xmath7 and @xmath1 ) exactly and derive a reliable @xmath8-factor for the @xmath1 cbbn .
the method we adopt is the state - of - the - art coupled - channel technique developed by two of the present authors ( m.k . and y.k .
) together with e. hiyama @xcite . since we can treat the catalyzed process directly , we do not need to refer to any of the sbbn processes .
also , the method has been already proven to be highly accurate and useful in atomic and nuclear physics ( reviewed in @xcite ) .
we find that the obtained astrophysical @xmath8-factor at @xmath39 kev ( the position of the gamow peak for @xmath9 kev which is relevant to the cbbn ) is 0.038 mev barn .
this is about 10 times smaller than the estimate in ref .
@xcite at the same energy . for long lived stau
, our @xmath8-factor leads to the @xmath40 abundance from cbbn as @xmath41/10^{-16})$ ] .
therefore , an observational upper bound on the @xmath40 abundance @xmath42 ( @xmath43 ) @xcite leads to a bound on the @xmath1 abundance , @xmath44 .
this requires a dilution of the relic stau by some entropy production at late time by a factor of @xmath45 .
such a dilution factor can easily be consistent with the thermal leptogenesis @xcite for a reheating temperature @xmath46 gev .
the organization of this paper is as follows . in sec.ii , we summarize the basic reaction of the @xmath1 catalyzed fusion as well as its atomic analogue , the muon transfer reaction . in sec.iii , we present our method of solving the @xmath6-@xmath7-x@xmath47 three - body problem . in sec.iv , we show our result of astrophysical @xmath8-factor as a function of the incident energy .
sec.v is devoted to conclusion and discussions in which the @xmath4 abundance in cbbn and its implication to particle physics are also mentioned .
as we have mentioned in the introduction , we will treat the following process directly in a fully quantum mechanical manner : @xmath48 as shown later in section v , the relevant temperature @xmath49 for cbbn is about 10 kev .
this corresponds to the gamow peak of the incident channel located at the c.m .
energy @xmath5036.4 kev with the full @xmath51 width of 44 kev .
thus the c.m .
energy of the incident channel will be typically up to @xmath52 kev .
we adopt the @xmath37 cluster model in which the @xmath53li nucleus is a bound state of a @xmath54he nucleus and a deuteron .
this model has extensively been utilized in the studies of the structure and reactions of light nuclei and is very well established ( see e.g. @xcite ) .
such a cluster - model treatment allows us to investigate the reaction ( [ eq : transfer ] ) as a quantum three - body problem , @xmath55 .
the rearrangement of the three ingredients takes place during the reaction process through the long range coulomb interaction and the short range nuclear interaction .
although solving the three - body problem accurately is quite an elaborate task , two of the authors ( m.k . and y.k .
) have an experience @xcite to have solved a similar subject , the muon transfer reaction , @xmath56 at incident energies of @xmath57 ev in the context of the muon catalyzed d - t fusion cycle @xcite . among the calculations of this reaction in the literatures ,
their three - body coupled - channel method was found to provide with the most accurate result ( see @xmath588 of @xcite ) . in the present paper , the same coupled - channel method
is applied to the reaction , eq .
( [ eq : transfer ] ) .
we note here that it is very important to employ an appropriate nuclear interaction between @xmath54he and deuteron in treating the reaction ; we choose the interaction which can reproduce the binding energy and the charge form - factor of @xmath53li as well as the low - energy @xmath37 scattering phase shift .
following ref.@xcite , we briefly explain the three - body coupled - channel method to investigate the reaction ( [ eq : transfer ] ) . as shown in fig.[fig : stau_fig1 ] , we consider all the jacobi coordinates @xmath59 of the three possible sets , @xmath60 and @xmath61 .
the entrance channel @xmath62 and the exit channel @xmath63 in our reaction are best described by the coordinates for @xmath64 and those for @xmath65 , respectively .
since we are interested in the incident energy of the entrance channel below 100 kev , there is no other open channel than the above two .
all the excited states of @xmath30 and @xmath53li as well as all the states of @xmath66 can be excited only virtually in the intermediate stage of the reaction . to describe such an intermediate state where particle - rearrangement takes place , it is convenient to utilize the coordinates for @xmath67 together with those for @xmath60 .
the schrdinger equation for the total wave function @xmath68 of the @xmath55 system having an angular momentum @xmath69 and its @xmath70-component @xmath71 is given by @xmath72 with the hamiltonian , @xmath73 as far as we use the reduced masses ( @xmath74 and @xmath75 ) associated with the coordinates ( @xmath76 and @xmath77 ) , every choice of @xmath78 in the kinetic term is equivalent .
@xmath79 denotes the potential between the particles @xmath80 and @xmath81 @xmath82 and will explicitly be given below .
spin of the deuteron is neglected , and therefore the angular momentum of the ground states of @xmath53li is zero as well as that of @xmath30 .
we denote the ground - state wave function of @xmath30 in @xmath64 by @xmath83 and its eigenenergy by @xmath84 , and similarly , that of @xmath53li in @xmath65 by @xmath85 and its eigenenergy by @xmath86 .
they are obtained by solving @xmath87 \ ,
\phi_{\rm g.s.}^{(c)}({\bf r}_c ) = 0 . \qquad ( c=1,2 ) \label{eq : subsystem}\ ] ] center - of - mass scattering energy of the channel @xmath78 associated with the coordinate @xmath77 , say @xmath88 , is introduced by @xmath89 together with the corresponding wave number @xmath90 by @xmath91 . for a given total energy @xmath92 , the schr odinger equation ( [ eq : schroedinger ] )
should be solved under the scattering boundary condition : @xmath93 y_{jm}(\widehat{\bf r}_c ) , \qquad ( c=1,2 ) \qquad \label{eq : boundary}\end{aligned}\ ] ] where @xmath94 are the asymptotic outgoing and incident coulomb wave functions .
@xmath95 is the s - matrix for the transition from the channel 1 to @xmath78 and @xmath96 is the velocity of the channel @xmath78 . by introducing a simplified notation , @xmath97 and @xmath98 , the cross section of the rearrangement process ( [ eq : transfer ] )
is given by @xmath99 and the astrophysical @xmath8-factor is derived from @xmath100 where @xmath101 is the sommerfeld parameter of the entrance channel .
the three - body wave function which describes the transfer reaction ( [ eq : transfer ] ) and the elastic @xmath102 scattering simultaneously is written as @xmath103 the first and the second terms represent the open channels , @xmath64 for @xmath102 and @xmath65 for @xmath104 .
the factors @xmath105 describe the scattering waves along the coordinates @xmath77 and are to be solved under the boundary condition ( [ eq : boundary ] ) .
the third term , @xmath106 , stands for all the closed ( virtually - excited ) channels in the energy range of this work ; in other words , this term is introduced to represent all the asymptotically - vanishing three - body amplitudes that are not included in the first two scattering terms .
for example , the third term describes such an effect that the incoming deuteron attracts the @xmath54he in the 1s - orbit around @xmath1 and distorts the orbit before picking up the @xmath54he .
since @xmath106 vanishes asymptotically , it is reasonable and useful @xcite to expand it in terms of a complete set of @xmath107-integrable three - body basis functions , @xmath108 , spanned in a finite spatial region ( see sec.3.c ) : @xmath109 equations for @xmath110 , @xmath111 and the coefficient @xmath112 are given by the @xmath113 simultaneous equations @xmath114 and @xmath115 here , @xmath116 denotes the integration over @xmath117 and @xmath118 .
since @xmath119 are constructed so as to diagonalize the three - body hamiltonian as @xmath120 the coefficients @xmath112 can be written , from eqs.([eq : eq27 ] ) , as @xmath121 inserting eqs.([eq : bjnu ] ) into @xmath112 in @xmath68 in eqs.([eq : eq26 ] ) , we reach two coupled integro - differential equations for @xmath110 and @xmath111 .
the integro - differential equations , though not recapitulated here ( c.f . 8 of @xcite for them ) , are solved by using both the direct numerical method ( finite - difference method ) and the kohn - type variational method @xcite , and we have obtained the same result for the incident energies relevant to cbbn , @xmath122 kev .
the coupling between the entrance and exit channels as well as the contribution from the closed channels @xmath123 were found to be significantly large as will be discussed later .
it is essential for the three - body calculation to employ appropriate nuclear interaction between @xmath54he and deuteron which governs the @xmath54he - transfer process . in this subsection
, we define the potentials @xmath124 and @xmath125 in our hamiltonian ( [ eq : hamiltonian ] ) .
first , we assume gaussian - shape charge distributions of @xmath54he and deuteron as @xmath126 and @xmath127 , respectively .
we take @xmath128 fm and @xmath129 fm , which reproduce observed r.m.s .
charge radii , @xmath130 fm of @xmath54he and 2.14 fm of deuteron .
the potential between @xmath54he and @xmath1 is then given by @xmath131 and that between deuteron and @xmath1 is written as @xmath132 where @xmath133 is the error function .
energy of the @xmath134 state is @xmath135 kev and the r.m.s .
radius is @xmath136 fm .
the potential @xmath137 is a sum of the nuclear potential , @xmath138 , and the coulomb potential , @xmath139 .
the latter is given by @xmath140 the nuclear potential is assumed to have a two - range gaussian shape as @xmath141 with @xmath142 fm , @xmath143 mev , @xmath144 fm and @xmath145 mev .
the first term , a repulsive core , is introduced to simulate the pauli exclusion principle that nucleons in the incoming deuteron should not occupy the nucleon s - orbit in the @xmath54he nucleus during the reaction process ( see e.g. @xcite for this role of pauli principle ) .
the above parameters of @xmath138 were so determined that the solution to the schrdinger equation ( [ eq : subsystem ] ) could reproduce observed values of the energy @xmath146 mev and the r.m.s .
charge radius 2.54 @xmath147 of the ground state of @xmath53li .
furthermore , the charge density of @xmath53li reproduces observed charge form factor of the electron scattering from @xmath4 as shown in the left panel of fig.[fig : stau_fig2 ] .
simultaneously , use of the potential @xmath137 explains the low - energy s - wave phase shifts of the @xmath148 scattering as shown in the right panel of fig.[fig : stau_fig2 ] . here
, it should be noted that simple attractive potential without a repulsive core leads to the phase shift which increases with increasing @xmath149 and can not explain the observed data .
we thus have a good @xmath37 potential @xmath137 in order to perform a precise study of the three - body reaction ( [ eq : transfer ] ) .
the potential @xmath137 and the wave function @xmath150 are illustrated in fig.[fig : stau_fig4 ] in which @xmath151 is denoted by @xmath152 for simplicity . the @xmath107-integrable three - body basis functions @xmath108 used in ( [ eq : phi - def ] ) to expand @xmath106 are introduced as follows @xcite : @xmath153 are written as a sum of the component functions in the jacobi - coordinate sets @xmath154 ( fig.[fig : stau_fig1 ] ) , @xmath155 each component is expanded in terms of the gaussian basis functions of the coordinates @xmath76 and @xmath77 : @xmath156_{jm } \qquad ( c=1 - 3)\ ; , \label{eq : bases}\ ] ] where @xmath157 the gaussian ranges are postulated to lie in geometric progression .
basis functions so chosen are suited for describing both short range correlations ( mainly due to the @xmath37 nuclear interaction ) and long range asymptotic behavior simultaneously @xcite , and therefore they are efficient to describe the three - body configurations ( closed - channel contribution ) in the interaction region in the intermediate stage of reactions @xcite .
the coefficients @xmath158 in ( [ eq : bases ] ) are determined by diagonalizing the three - body hamiltonian @xmath159 as eqs.([eq : diag ] ) . in the calculation for @xmath160
, we took @xmath161 and @xmath162 for @xmath154 .
therefore , total number of the three - body gaussian basis functions @xmath163_{jm}$ ] to construct @xmath164 amounts to @xmath165 , which was found to be large enough for the present calculation . as for the gaussian ranges
, we took @xmath166 and @xmath167 to be 0.5 , 15.0 , 1.0 , 40.0 fm , which are sufficiently precise for the present purpose .
the expansion ( [ eq : phi - def ] ) was found to converge quickly with increasing @xmath168 , and @xmath169 mev above the @xmath170 threshold ) is very sufficient .
table [ table : percent ] lists the cross section @xmath171 and the astrophysical @xmath8-factor @xmath172 of the reaction ( [ eq : transfer ] ) obtained by the full coupled - channel calculation at several energies around @xmath173 kev . in fig.[fig : stau_fig5 ] , @xmath172 is shown together with the gamow peak . the @xmath8-factor increases with decreasing energy @xmath149 since there are three - body bound states below the @xmath4-@xmath1 threshold .
note that @xmath174 estimated in @xcite is an order of magnitude larger than the value given by the present exact calculation .
since the @xmath8-factor can be approximated around @xmath175 as @xmath176 with @xmath177 , the reaction rate , using eqs.(4.56 ) and ( 4.74 ) of @xcite , is expressed as @xmath178 where @xmath179 is the avogadro constant and @xmath174 , @xmath180 and @xmath181 are to be given in units of mev@xmath182 barn , barn and @xmath183 k , respectively . taking @xmath184 mev barn and @xmath185 barn , we obtain @xmath186 cccc @xmath149 [ kev ] & @xmath187 @xmath188 [ barn ] & @xmath187 @xmath189 [ mev barn ] & @xmath187 + + 10 & @xmath187 @xmath190 & 0.0426 + 20 & @xmath187 @xmath191 & 0.0410 + 36.4 & @xmath187 @xmath192 & 0.0380 + 50 & @xmath187 @xmath193 & 0.0357 + 100 & @xmath187 @xmath194 & 0.0286 + + + [ table : percent ] to check the necessity of the exact coupled - channel treatment ( [ eq : psi12])@xmath195([eq : bjnu ] ) , we have calculated the cross sections in the following approximations : ( i ) a coupled - channel treatment with only the entrance and exit channels where the closed channel @xmath123 in eq.([eq : psi12 ] ) is neglected , and ( ii ) an one - step approximation where the transition from the entrance channel to the exit channel in the case ( i ) is treated in the leading order perturbation . for the case ( i ) , the cross section @xmath196 become approximately 3 times smaller than those in table [ table : percent ] .
for the case ( ii ) , the cross section becomes an order of magnitude larger than that in the case ( i ) .
therefore , these approximations are not justified .
we have also checked that the cross section @xmath196 is dominated by the total orbital angular momentum @xmath160 , and contribution from the higher angular momenta @xmath197 is three orders of magnitude smaller .
furthermore , any pair among @xmath54he , @xmath7 and @xmath1 has dominantly zero orbital angular momentum , i.e. s - wave .
this is in sharp contrast to the sbbn reaction @xmath198 where the entrance channel is dominated either by d - wave ( the e2 transition ) or by p - wave ( the e1 transition ) from the selection rule .
since we consider the gravitino masses of order 10 gev , the scalar lepton @xmath0 ( such as stau ) is most likely the next lightest supersymmetry particle and could have a long lifetime more than @xmath199 sec .
thus , the negatively charged @xmath1 forms a bound states @xmath2 with positively charged nuclei @xmath3 during the big - bang nucleosynthesis ( bbn ) . then , some nuclear abundances are enhanced through the catalyzed process . in this letter , we have calculated a production cross section of @xmath4 from the catalyzed process @xmath5 by exactly solving the schrdinger equation for three - body system of @xmath6 , @xmath7 and @xmath0 .
we utilize the state - of - the - art coupled - channel method , which is known to be very accurate to describe other three - body systems in nuclear and atomic reactions .
the use of appropriate nuclear potential and the exact treatment of the quantum tunneling in the fusion process turned out to be important .
we have found that the astrophysical @xmath8-factor at the gamow peak corresponding to @xmath9 kev is 0.038 mev barn as shown in table [ table:1 ] , fig .
[ fig : stau_fig5 ] and eq .
( [ eq : sigmav ] ) .
they are the main results of this letter . before closing , let us briefly discuss the particle physics implication of our result . with the cross section @xmath200 in eq .
( [ eq : sigmav ] ) , the @xmath4 production via cbbn is described by @xmath201 where @xmath202 and @xmath203 are the abundances of these elements normalized by the baryon number density .
the ratio of the number density of the bound states @xmath204 to the entropy density is given by is in chemical equilibrium . for more detailed analysis of the evolution of the bound state abundance ,
see ref .
@xcite . ] @xmath205 where @xmath206 and @xmath207 are the number density and the mass of the @xmath208 , and @xmath209 gev is the binding energy of @xmath210 .
the fraction of @xmath1 which form bound states with @xmath208 is shown in fig .
[ fig : stau_fig6](a ) , as a function of temperature @xmath49 for various lifetimes of @xmath0 .
, @xmath211 , @xmath212 , @xmath213 , @xmath214 gev @xcite and @xmath215 gev ( see the previous section ) . ] the bound state abundance is peaked at around @xmath9 kev for @xmath216 sec , sec the cbbn effect is negligible .
see fig .
[ fig : stau_fig6](b ) . ] which justifies the expansion in eq.([eq : expansion ] ) .
thus , the cbbn occurs at @xmath217 sec , when the sbbn processes ( @xmath218 sec ) are already frozen . at such late time , the effect of the dissociation @xmath219 is also negligible @xcite , and hence we have neglected it in eq .
( [ eq : ddtli6 ] ) .
[ fig : stau_fig6](b ) shows the contour plot of @xmath40 abundance produced by the cbbn process , as a function of initial @xmath1 abundance @xmath220 and the lifetime @xmath18 . in the limit of long lifetime , one obtains @xmath221 therefore , the observational upper bound on the @xmath40 abundance @xmath42 ( @xmath43 ) @xcite leads to a bound on the @xmath1 abundance , @xmath44 . on the other hand ,
the thermal relic abundance of stau is given by @xmath222 @xcite .
( conservative cases are shown in fig .
[ fig : stau_fig6](b ) as the dashed lines for @xmath223 gev and 300 gev ) .
therefore , in the limit of long life time , an entropy production with a dilution factor @xmath224 for the primordial stau abundance is necessary to avoid the overproduction of @xmath4 . on the other hand
, this means that the primordial baryon asymmetry is also diluted by @xmath225 .
standard thermal leptogenesis @xcite , an attractive mechanism for baryogenesis , can normally explain the observed baryon asymmetry if the reheating temperature satisfies @xmath226 gev @xcite .
since the maximal baryon asymmetry is proportional to @xmath227 in thermal leptogenesis , this means that , for a late time dilution @xmath224 , the reheating temperature of @xmath228 is necessary .
such a reheating temperature can be obtained in inflationary models .
d. z. freedman , p. van nieuwenhuizen and s. ferrara , phys .
d * 13 * ( 1976 ) 3214 ; + s. deser and b. zumino , phys .
b * 62 * ( 1976 ) 335 .
w. buchmuller , k. hamaguchi , m. ratz and t. yanagida , phys .
b * 588 * , 90 ( 2004 ) [ arxiv : hep - ph/0402179 ] .
w. m. yao _ et al .
_ [ particle data group ] , j. phys .
g * 33 * ( 2006 ) 1 .
t. asaka , k. hamaguchi and k. suzuki , phys .
b * 490 * ( 2000 ) 136 [ arxiv : hep - ph/0005136 ] .
m. fujii , m. ibe and t. yanagida , phys .
b * 579 * ( 2004 ) 6 [ arxiv : hep - ph/0310142 ]
. w. buchmuller , k. hamaguchi , m. ibe and t. t. yanagida , arxiv : hep - ph/0605164 .
w. buchmuller , p. di
bari and m. plumacher , annals phys .
* 315 * ( 2005 ) 305 [ arxiv : hep - ph/0401240 ] ; + see , for a review , w. buchmuller , r. d. peccei and t. yanagida , ann .
nucl . part .
* 55 * ( 2005 ) 311 [ arxiv : hep - ph/0502169 ] . | if the gravitino mass is in the region from a few gev to a few 10 s gev , the scalar lepton @xmath0 such as stau is most likely the next lightest supersymmetry particle .
the negatively charged and long - lived @xmath1 may form a coulomb bound state @xmath2 with a nucleus @xmath3 and may affect the big - bang nucleosynthesis through catalyzed fusion process . we calculate a production cross section of @xmath4 from the catalyzed fusion @xmath5 by solving the schrdinger equation exactly for three - body system of @xmath6 , @xmath7 and @xmath0 .
we utilize the state - of - the - art coupled - channel method , which is known to be very accurate to describe other three - body systems in nuclear and atomic reactions . the importance of the use of appropriate nuclear potential and the exact treatment of the quantum tunneling in the fusion process are emphasized .
we find that the astrophysical @xmath8-factor at the gamow peak corresponding to @xmath9 kev is 0.038 mev barn .
this leads to the @xmath4 abundance from the catalyzed process as @xmath10/10^{-16})$ ] in the limit of long lifetime of @xmath0 .
particle physics implication of this result is also discussed . |
performing stellar archeology on white dwarfs , by far the most common stellar remnant , provides valuable information for not only understanding stellar evolution and mass loss but also galactic evolution .
one of the fundamental relations in the analysis of white dwarfs is the initial - final mass relation ( hereafter ifmr ) , where the masses of white dwarfs are compared directly to the zero - age main sequence mass of their progenitors .
this semi - empirical relation is critical to our understanding of integrated mass - loss over the lifetime of a star and how it changes with stellar mass .
the ifmr has a variety of additional applications including predicting type ia supernovae rates ( pritchet et al .
2008 ; greggio 2010 ) and overall stellar feedback in galaxy models ( agertz & kravtsov 2014 ) , interpreting the white dwarf luminosity function ( cataln et al .
2008 ) , and providing a technique for measuring the age of the galactic halo ( kalirai 2013 ) .
analysis of the ifmr began with weidemann ( 1977 ) , where it was shown that the models of the time greatly underestimated the observed stellar mass loss .
subsequent work on the ifmr by a number of groups ( see weidemann 2000 for review ) resulted in a broad but sparsely populated relation that showed a clear trend with higher - mass main sequence stars producing increasingly more massive white dwarfs . in the past 15 years
the amount of ifmr data has greatly increased ( e.g. , claver et al .
2001 ; dobbie et al .
2004 , 2006a ; williams et al . 2004 ;
kalirai et al .
2005 ; liebert et al . 2005 ; williams & bolte 2007 ; kalirai et al .
2007 ; kalirai et al . 2008
; rubin et al . 2008 ; kalirai et al .
2009 ; williams et al . 2009 ; dobbie & baxter 2010 ; dobbie et al . 2012 ) .
these newer data retain the general trend of the previous ifmr work , but the scatter in the data remains significant .
the source of this scatter may be attributable to several factors including the possible stochastic nature of mass loss , effects from variation in metallicity or environment , or systematic differences between the studies .
one important systematic is the challenge in defining the ages of the clusters these white dwarfs belong to , which creates uncertainty in the derived lifetimes of their progenitor stars .
cummings et al .
( 2015 ; hereafter paper i ) began to analyze the important intermediate - mass ifmr ( progenitor masses of 3 - 4 m@xmath1 ) from the rich ngc 2099 ( m37 ) .
this work strengthened the observational evidence that in this mass range the ifmr is steep , where the final white dwarf mass increases more rapidly with increasing progenitor mass .
comparison to the rich population of comparable mass white dwarfs in both the hyades and praesepe from kalirai et al .
( 2014 ) showed strongly consistent ifmrs .
this consistency also suggests that across this mass range the slightly metal - rich progenitor stars from the hyades and praesepe ( [ fe / h]@xmath00.15 ) have no significant increase in mass - loss rates compared to those in the solar metallicity ngc 2099 . expanding beyond paper i
, we now look at the challenging higher - mass region ( m@xmath4 of 46.5 m@xmath1 ) of the ifmr by focusing on white dwarfs in younger clusters .
while younger clusters do not provide a broad mass range of white dwarfs , they provide several important advantages . first , the highest - mass white dwarfs are the most compact and lowest luminosity , and because they form first they remain bright only in the youngest clusters ( @xmath5 200 myr ) .
second , the cooling rates are far more rapid in young and hot white dwarfs and , as a result , errors in temperature lead to far smaller errors in both cooling age and luminosity . third , high - mass white dwarfs ( @xmath20.9 m@xmath1 ) may be prone to be ejected from their parent population clusters , either due to dynamical interactions or potential velocity kicks due to asymmetric mass loss during their formation ( fellhauer et al . 2003 ;
tremblay et al .
therefore , the probability of finding high - mass white dwarfs still within their cluster population may decrease with age .
these three reasons are why younger clusters provide far more advantages in analyzing high - mass white dwarfs with the best precision . in this paper
we begin our analysis with the rich , young , and nearby cluster ngc 2323 ( m50 ) .
based on population analysis , kalirai et al .
( 2003 ) find that it is approximately three times as rich as the pleiades , making it an excellent environment to search for the rare remnants of higher - mass stars .
additionally , to expand our sample we self - consistently reanalyze publicly available data on all published high - mass white dwarfs from clusters ( pleiades [ m45 ] , ngc 2516 , ngc 2287 , ngc 3532 , and ngc 2168 [ m35 ] ) and sirius b. to further limit systematics , we also self - consistently analyze the cluster parameters for all of the parent cluster populations based on published high - quality ubv photometry .
to look at the broader picture , we connect these high - mass data to the moderate - mass data from paper i and further analyze the broader characteristics of the ifmr .
the structure of this paper is as follows , in section 2 we discuss the spectroscopic white dwarf observations of ngc 2323 , the publicly available data we have used , and our reduction and analysis techniques . in section 3
we discuss the ubv photometry based cluster parameters for our six open clusters being analyzed . in section 4
we discuss the cluster membership of our white dwarf candidates in ngc 2323 . in section 5
we discuss the high - mass ifmr and compare to the intermediate - mass ( 34 m@xmath1 progenitors ) ifmr from paper i. in section 6 we summarize our study .
based on the deep bv photometric observations of ngc 2323 ( kalirai et al .
2003 ) with the canada - france - hawaii telescope and the cfh12k mosaic camera , a sample of white dwarf candidates in ngc 2323 were spectroscopically observed at keck i using the low resolution imaging spectrometer ( lris ; oke et al .
1995 ) . in total , 10 of these candidates had sufficient signal to properly analyze their characteristics .
the 400/3400 grism was used with 1 " slits giving a spectral resolution of @xmath06 , which provides us the wavelength coverage of @xmath03000 to 5750 and a series of 5 balmer lines ( h@xmath6 , h@xmath7 , h@xmath8 , h@xmath9 , and h8 ) .
these observations were performed on 2008 december 23 and 24 , on 2011 december 27 and 28 , and on 2015 february 19 . for spectral flux calibration , flux standards
were observed each night .
three independent lris masks were used in 2008 to observe white dwarfs candidates from ngc 2323 .
mask 1 was observed for 40 minutes , mask 2 was observed for 40 minutes , and mask 3 was observed for 2 hours and 40 minutes . in 2011 , individual longslit observations were performed for five different white dwarf candidates in ngc 2323 ranging from 20 minutes to 1 hour and 50 minutes .
lastly , in 2015 an additional 70 minutes was acquired on wd11 , which is a white dwarf candidate of interest that only had 20 minutes of observation in 2011 .
we have reduced and flux calibrated our lris observations using the idl based xidl pipeline .
of our total observed sample of 10 white dwarfs candidates , eight are da white dwarfs and wd23 and wd38 have no clear spectral features . to provide additional high - mass white dwarfs for comparison we have taken from the vlt archive the observations of seven white dwarf members of ngc 3532 , three white dwarf members of ngc 2287 , and four white dwarf members of ngc 2516 ( based on observations made with eso telescopes under program ids : 079.d-0490(a ) ; 080.d-0654(a ) ; 084.d-1097(a ) ; pi : dobbie ) .
these observations were performed with fors1 and fors2 using the 600b grism ( appenzeller et al .
1998 ) giving comparable spectral resolution to our lris observations of @xmath06 .
analysis of the parameters and membership for the white dwarfs of ngc 2287 , ngc 2516 , and ngc 3532 were originally published in dobbie et al .
( 2009 ; 2012 ) . additionally , from the keck archive we have taken the observations of 11 white dwarf members of ngc 6128 ( program i d : u49l-2002b ; u60l-2004a ; u15l-2004b ; u18l-2005b ; pi : bolte ) . similar to our ngc 2323 observations , these observations were performed using lris with a majority of them using the 400/3400 grism and 1 " slits , giving the same characteristics to our data .
analysis of the parameters and membership for the white dwarfs of ngc 6128 were original published in williams et al .
( 2009 ) .
we have performed our own reductions and analyses of these data from both the vlt and keck , but we do not redetermine their membership status in this paper .
the vlt data were reduced using the standard iraf techniques for reduction of longslit data , while the keck data were reduced using the same xidl pipeline used to analyze our ngc 2323 data . as a test for our vlt data reduction , we were provided with the published spectrum of white dwarf j0646 - 203 ( ngc 2287 - 4 ) ( p.d .
dobbie ; private communication 2014 ) , and we found that there are no meaningful differences in our reduced spectra of the same data .
hence , there are no systematics caused by our spectral reduction techniques , and the systematic differences between our parameters and those presented in dobbie et al .
( 2012 ) are due to differences in our applied white dwarf models and fitting techniques .
l c c c c c c c c c c * - ngc 2323 white dwarf initial and final parameters * + id&t@xmath10&log g&m@xmath11 & m@xmath12&t@xmath13 & y@xmath3 m@xmath14 & parsec m@xmath14 & m@xmath15 & m@xmath16&s / n + & ( k ) & & ( m@xmath1 ) & & ( myr ) & ( m@xmath1)&(m@xmath1)&(m@xmath1)&(m@xmath1 ) & + + ngc 2323-wd10 & 52800@xmath171350 & [email protected] & [email protected] & [email protected] & 1.6@xmath18 & 4.69@xmath19 & 5.07@xmath20 & 4.98 & 4.45 & 85 + ngc 2323-wd11 & 54100@xmath171000 & [email protected] & [email protected] & [email protected] & 1.3@xmath21 & 4.69@xmath19 & 5.07@xmath22 & 4.98 & 4.45 & 130 + + ngc 2323-wd21 & 18200@xmath17850 & [email protected] & [email protected] & [email protected] & 170@xmath23 & & & & & 28 + ngc 2323-wd7 & 16800@xmath17250 & [email protected] & [email protected] & [email protected] & 112@xmath24 & & & & & 122 + ngc 2323-wd17 & 19800@xmath17300 & [email protected] & [email protected] & [email protected] & 96@xmath24 & & & & & 111 + ngc 2323-wd12 & 17100@xmath17400 & [email protected] & [email protected] & [email protected] & 101@xmath25 & & & & & 60 + + ngc 2323-wd22 & 24400@xmath171550 & [email protected] & [email protected] & [email protected] & 33@xmath26 & & & & & 13 + ngc 2323-wd30 & 13400@xmath17900 & [email protected] & [email protected] & [email protected] & 290@xmath27 & & & & & 23 + for the high - mass lb 1497 , from the pleiades , and sirius b we have taken the t@xmath10 and log g parameters from gianninas et al .
we also have taken these parameters from gianninas et al .
( 2011 ) for the supermassive gd50 and pg 0136 + 251 white dwarfs , where dobbie et al .
( 2006b ) used three - dimensional space velocities to argue that gd50 s progenitor was related to and coeval with the pleiades cluster . similarly , but based on only proper motions , dobbie et al .
( 2006b ) argued that pg 0136 + 251 s progenitor is likely consistent with coeval formation with the pleiades .
gianninas et al .
( 2011 ) use white dwarf atmospheric models and fitting techniques equivalent to ours , and we adopt their t@xmath10 and log g in our analysis of these four white dwarfs . for our spectroscopic analysis we adopted the same analysis techniques as those described in paper i. in brief
, we used the recent white dwarf spectroscopic models of tremblay et al .
( 2011 ) with the stark profiles of tremblay & bergeron ( 2009 ) , and the automated fitting techniques described by bergeron et al .
( 1992 ) to fit our balmer line spectra and derive t@xmath10 and log g. however , for our derived parameters ( mass , luminosity , and cooling age ) we expand upon the methods of paper i because our current sample has a far broader mass range . for deriving the parameters for white dwarfs of mass less than 1.10 m@xmath1 we applied our t@xmath10 and log g to the cooling models for a carbon / oxygen ( co ) core composition with a thick hydrogen layer by fontaine et al .
for the highest - mass white dwarfs ( @xmath28 1.1 m@xmath1 ) we derived the parameters based on the oxygen / neon ( one ) core models of althaus et al .
( 2005 ; 2007 ) . of the cooling rates for the several white dwarf models that we have discussed , in addition to the basti co cooling models from salaris et al . ( 2010 ) . in the lower panel ,
the hottest temperatures are shown and the systematics are always @xmath01.5 myr or less in magnitude , which is not concerning . at cooler temperatures , all models other than the carbon core models of wood ( 1995 ) are consistent .
this is consistent with the systematic effects of differing core composition , which grow as white dwarfs cool further . ]
a discussion of our adopted models is warranted .
first , we do not adopt the more recent white dwarf atmospheric models of tremblay et al .
( 2013 ) because those focus only on 3d modeling of convective atmospheres ( @xmath2914,000 k ) , and our current analysis looks hot fully radiative white dwarfs . for our cooling models , while the fontaine et al .
( 2001 ) models are widely used , we should acknowledge two limitations they have .
first , they assume a 50/50 carbon and oxygen core composition , which based on full stellar evolutionary models is not accurate ( e.g. , romero et al .
the effect of this on the calculated cooling ages can be important , but for the relatively young white dwarfs we are analyzing this effect remains small .
second , these cooling models do not begin at the tip of the asymptotic giant branch ( hereafter agb ) and instead begin at a t@xmath10 of @xmath060,000 k. in comparison , the widely used wood ( 1995 ) co cooling models do begin at the tip of the agb and they do not adopt a simple 50/50 carbon and oxygen core composition . unfortunately , their co core models have an upper mass limit of 1.0 m@xmath1 , which limits their application in our analysis but they provide an important test for systematics . in figure 1
we compare the cooling ages from the fontaine et al .
( 2001 ) co core models at 1.0 m@xmath1 to both the co and pure - carbon core models of wood ( 1995 ) and the basti co models from salaris et al .
the basti co cooling models adopted c / o ratio profiles based on the basti scaled solar stellar evolution models . the lower panel of figure 1 shows that at high t@xmath30@xmath2845,000 k all of the differences between these models never result in cooling age differences of more than 1.5 myr .
the upper panel of figure 1 similarly shows that at cooler temperatures 30,000@xmath5t@xmath10@xmath540,000 k the co models , irrespective of their adopted c / o ratios or starting points , all give strong agreement in cooling ages , but the systematic effects introduced from pure carbon core models ( @xmath010 myr ) are now clearly seen . for ultramassive white dwarfs ,
the mass at which they transition to one white dwarfs remains uncertain , and it also likely depends on metallicity ( doherty et al .
here we have adopted a somewhat conservative 1.10 m@xmath1 , but we note that the models of garcia - berro et al .
( 1997 ) argue that it may be as low as 1.05 m@xmath1 .
reassuringly , in this mass range the co and one cooling ages at constant mass are consistent for such young white dwarfs , but for these white dwarfs the dependence of the mass - radius relationship on core - composition is very important .
for example , applying the gravities of white dwarfs in the mass range of 1.05 to 1.10 m@xmath1 ( based on co - core models ) to the one models derives masses @xmath00.05 m@xmath1 lower and places them all below 1.05 m@xmath1 .
therefore , it is appropriate to adopt 1.10 m@xmath1 as the transition mass . in table 1
we present our white dwarf parameters for the eight da white dwarfs from ngc 2323 . for clarity in table 1
we have distinguished between members and nonmembers , where membership is based on our comparisons of model based and observed photometry in addition to comparisons of cooling ages to the cluster age ( see our detailed discussion of membership in section 4 ) .
additionally , we have considered s / n ( per resolution element ) and the resulting errors given in table 1 .
the wd22 and wd30 spectra have low s / n and mass errors greater than 0.1 m@xmath1 .
therefore , their parameters are presented for reference but have been cut from our final analysis .
this is because their membership determinations are unreliable , but their low masses do suggest they are field white dwarfs .
with ifmr analysis , the parameters of the star clusters are as critical as the white dwarfs themselves .
this is particularly true for the highest - mass white dwarfs , where the derived masses of their progenitors change rapidly with evolutionary time .
uniform photometric data sets are not available for these six clusters . but in paper i we showed that for ngc 2099 , the adopted isochrones and fitting techniques had as large , if not larger , of an effect on its cluster parameters as did any systematics between the cluster s different photometry sets .
this is best illustrated by the systematic effects on derived cluster ages , where the yi et al .
( 2001 ; hereafter y@xmath3 ) and the ventura et al .
( 1998 ) isochrones both gave our final adopted age of 520 myr for ngc 2099 .
the parsec version 1.2s isochrones ( bressan et al .
2012 ) derived a comparable age of 540 myr , and lastly the girardi et al .
( 2000 ) and bertelli et al .
( 1994 ) isochrones both gave significantly younger ages of 445 myr .
between these isochrone sets there is nearly a 100 myr range of derived ages when using identical photometry , and further systematics can be introduced based on how the isochrones are fit to the data . in this paper
we have redetermined as uniformly as possible the reddenings , distance moduli , and ages for these six clusters using available high - quality ubv photometry that covers up to the full turnoff .
the y@xmath3 isochrones provide our final adopted cluster ages , but to broaden our results we also determine ages with the parsec isochrones .
these two isochrones give only slightly different cluster ages , but in these younger clusters the masses of the progenitor stars have a far more significant dependence on evolutionary time , and even a 20 myr systematic has a significant effect on our results . in our cluster
photometric analysis we first make use of color - color diagrams ( b - v vs u - b ) , which provide direct photometric information on the cluster reddening . the photometric metallicity can also be derived but it is quite sensitive to systematics in u magnitude , a concern considering the typically more complex standardization process for u magnitudes and the varying sources of our photometry
therefore , in the case of the pleiades , ngc 2168 , and ngc 2516 we consider more detailed spectroscopic metallicities , but for ngc 2287 , ngc 2323 , and ngc 3532 we will simply adopt solar metallicity .
however , we note that these adopted metallicities do show strong consistency with the observed photometry .
our color - color analysis adopts two techniques , the first is semi - empirical and based on the hyades fiducial and the second is based directly on the y@xmath3 isochrones , which reach to higher masses than those available in the hyades fiducial . for both methods
the reddening relation adopted is that of cardelli et al .
( 1989 ) and the metallicity correction is based on that of the y@xmath3 isochrones .
the methods using the hyades fiducial have been developed in deliyannis et al .
( in prep ) , where the fiducial was derived from single - star cluster members ( see perryman et al . 1998 ) .
the hyades ubv photometry of johnson & knuckles ( 1955 ) was adopted with a cluster [ fe / h ] of + 0.15 and e(b - v ) of 0 .
the pleiades provides a good example for our color - color analysis techniques . in the left panel of figure 2
we have plotted the photoelectric ubv photometry from johnson & mitchell ( 1958 ) with several reddening curves based on the hyades fiducial , and we have applied a metallicity of [ fe / h]=0.01 ( z=0.0185 ) to match the reddening insensitive region where all of the reddening curves intersect near b - v of 0.6 .
fitting by eye the blueward color - color trend we find that a reddening curve of e(b - v)[email protected] matches the pleiades ubv photometry the best . both this reddening and metallicity are consistent with the typically derived values and spectroscopic analyses of the pleiades ( e.g. , [ fe / h][email protected] schuler et al .
2010 ; [ fe / h][email protected]@xmath170.05 soderblom et al .
2009 ) we also note that the hyades fiducial ends at b - [email protected] , where the older hyades turnoff occurs .
in the right panel of figure 2 we fit the higher - mass stars bluer than b - v=0.0 with three different 135 myr y@xmath3 isochrones of differing metallicity .
all three metallicities fit a reddening of e(b - v)=0.03 in these bluest stars .
this demonstrates that these higher - mass stars create a nearly linear trend that is insensitive to variations in metallicity .
we note that the position of this blue linear trend is also insensitive to cluster age , where as we look at older clusters the trend only shortens in length and does not shift its position .
therefore , this linear trend s position provides a reliable reddening measurement independent of all other cluster parameters . in the upper left panel of figure 3
we compare directly our hyades fiducial fit and our y@xmath3 fit for the pleiades .
there are systematic differences that are noted in the b - v range of 0.1 - 0.3 , where it appears that the y@xmath3 isochrones are too blue in u - b relative to both the data and the hyades fiducial .
however , when adopting a y@xmath3 isochrone with an age consistent to the hyades ( 650 myr ; not shown ) the isochrone is nearly identical to that of the hyades fiducial , so at these younger ages the isochrones appear to overestimate the u flux at this intermediate color range . at all other colors
this does not seem to be a concern because , reassuringly , both the hyades fiducial and the y@xmath3 isochrones find the same reddening in their respective regions , and they also agree in the region sensitive to metallicity ( b - [email protected] ) and redder . in regard to the parsec isochrones , they are not independently considered in our color - color analysis because their u - b colors do not fit the observations .
a key advantage to color - color analysis is that it can be used to clean young cluster turnoffs .
dating younger clusters is prone to several difficulties , including that turnoffs are relatively sparse even in the richest clusters and that these higher - mass turnoff stars have high binarity fraction ( kouwenhoven et al .
2007 ) . additionally , many higher - mass stars fall into the peculiar groups of blue stragglers or be stars .
for example , mermilliod 1982a , 1982b , and ahumada & lapasset ( 2007 ) found that several of the brightest stars in the clusters being analyzed in this paper are blue stragglers that are far too blue or too bright , if not both , for their age .
lu et al .
( 2011 ) modeled the formation of blue stragglers on the short time scales necessary in these young clusters and found that they can rapidly be created through binary mass transfer . identifying these peculiar stars
can greatly improve the fit of the turnoff ages of these younger clusters .
mermilliod 1982b do find that when plotted in color - color space be stars deviate from the approximately linear trend of the
normal " high - mass stars .
additionally , several of our clusters may have variable reddening , and fitting this high - mass linear trend identifies the richest group of cluster stars with consistent reddening . in the pleiades shown in the upper - left panel of figure 3
, we mark in red the high - mass stars consistent with the trend , and several stars deviate from this trend that are likely peculiar .
however , we do acknowledge that the brightest star in the pleiades , alcyone , does not deviate from this trend .
alcyone is a multiple system and has several peculiar characteristics like spectral emission and rapid rotation ( hoffleit & jaschek 1991 ) and is commonly referred to as both a blue straggler and be dwarf .
this suggests that while this color - color method does identify many problematic stars , it does not remove all peculiar stars . in the upper right panel of figure 3
we have similarly plotted the ngc 2323 ubv photometry from claria et al .
we fit the bluest stars and see that independent of an assumed metallicity , we find a large reddening of e(b - v)[email protected] at ( b - v)@xmath31=0 , where we deem this reddening large enough to account for the color dependence of reddening ( see fernie 1963 and our discussion in paper i ) .
when we have adopted a color dependent reddening we will define the reddening at ( b - v)@xmath31=0 .
again , our selected final turnoff stars are shown in red .
l c c c c c c * - open cluster parameters * + cluster & e(b - v)@xmath32 & ( m - m)@xmath31 & [ fe / h ] & y@xmath3 age ( myr ) & parsec age ( myr ) & photometric sources + pleiades & [email protected] & [email protected] & + 0.01 & 135@xmath1715 & 110@xmath1715 & 1 + ngc 2323 & [email protected] & [email protected] & 0.00 & 140@xmath1720 & 115@xmath1720 & 2,3 + ngc 2516 & [email protected] & [email protected] & + 0.065 & 170@xmath1720 & 150@xmath1720 & 4,5 + ngc 2168 & [email protected] & [email protected] & -0.143 & 190@xmath1720 & 170@xmath1720 & 6 + ngc 2287 & [email protected] & [email protected] & 0.00 & 220@xmath1730 & 205@xmath1730 & 7,8 + ngc 3532 & [email protected] & [email protected] & 0.00 & 320@xmath1720 & 325@xmath1720 & 9 + l c c c c c c c
c * - membership data * + id&@xmath33&@xmath8&m@xmath12 & v & ( b - v)@xmath31 & b - v & t@xmath13 + & ( j2000 ) & ( j2000 ) & ( model)&(obs.)&(model)&(obs.)&(myr ) + + ngc 2323-wd10 & 7:02:41.02 & -8:26:12.8 & [email protected] & [email protected] & [email protected] & [email protected] & 2.9@xmath34 + ngc 2323-wd11 & 7:03:22.14 & -8:15:58.7 & [email protected] & [email protected] & [email protected] & [email protected] & 2.6@xmath35 + + ngc 2323-wd21 & 7:02:08.56 & -8:25:48.0 & [email protected] & [email protected] & [email protected] & [email protected] & 170@xmath23 + ngc 2323-wd7 & 7:02:47.87 & -8:35:56.4 & [email protected] & [email protected] & [email protected] & [email protected] & 112@xmath24 + ngc 2323-wd17 & 7:03:12.62 & -8:30:53.6 & [email protected] & [email protected] & [email protected] & [email protected] & 96@xmath24 + ngc 2323-wd12 & 7:03:17.79 & -8:19:59.7 & [email protected] & [email protected] & [email protected] & [email protected] & 101@xmath25 + + ngc 2323-wd22 & 7:02:33.93 & -8:31:04.3 & [email protected] & [email protected] & [email protected] & [email protected] & 33@xmath26 + ngc 2323-wd30 & 7:03:22.56 & -8:29:25.7 & [email protected] & [email protected] & [email protected] & [email protected] & 290@xmath27 + ngc 2323-wd38 & 7:03:31.87 & -8:28:25.0 & & [email protected] & & [email protected] & + ngc 2323-wd23 & 7:03:39.85 & -8:28:16.7 & & [email protected] & & [email protected] &
+ our four additional clusters are also shown in figure 3 . in ngc 2516 ,
we have used two ubv photometric studies , dachs ( 1970 ) for the brightest stars and sung et al .
( 2002 ) for the fainter stars shown in blue . for the moderately large reddening we fit a e(b - v ) of [email protected] at ( b - v)@xmath31=0 and a [ fe / h]=0.065 , and as with the pleiades this is consistent with the typically adopted parameters and our spectroscopic analysis ( cummings 2011 ) . for ngc 2287
we have used the ubv photometry of ianna et al .
( 1987 ) and fit a e(b - v ) of [email protected] with an assumed solar metallicity . for ngc 3532
we have used the ubv photometry of fernandez & salgado ( 1980 ) and fit a e(b - v ) of [email protected] with an assumed solar metallicity .
lastly , for ngc 2168 we have used the ubv photometry of sung & bessell ( 1999 ) .
we derive a e(b - v ) of [email protected] at ( b - v)@xmath31=0 for this cluster and adopt a metal - poor [ fe / h ] of -0.143 ( steinhauer & deliyannis 2004 ) , which provides a reliable fit to the bluest stars and follows the u - b range well but it appears that there is a systematic shift in b - v where the stars becoming increasingly too blue at redder colors .
otherwise , changes in either adopted metallicity or reddening can not fit the full b - v color range . in figure 4
we display our by eye turnoff age fits with y@xmath3 isochrones in green when adopting from figure 3 the reddenings , metallicities , and the cleaned turnoff stars shown in red . in magenta
we similarly display the parsec isochrone fits , where we have adopted the same reddenings , distance moduli , and metallicities ( z ) . in table 2 the derived cluster parameters and the photometric sources are listed .
we note that in the youngest clusters the parsec isochrones systematically derive ages 25 myr younger , while in the 320 myr ngc 3532 they derive ages 5 myr older , and lastly from paper i the parsec isochrones derive an age 20 myr older in the 520 myr old ngc 2099 . therefore , these two isochrones not only have changing systematics at differing ages , but they are in opposite directions in young versus older clusters .
the possible causes of the systematics between these two isochrones include the differences in their adopted opacities , equations of state , and solar compositions . while the y@xmath3 isochrones do not consider evolution past the tip of the red giant branch ( rgb ) , in these typically young clusters we can not reliably fit the giants because their populations are very sparse or they have no giants at all .
therefore , we have chosen the y@xmath3 isochrones for our final parameters because they more successfully fit both the color - color data and the the main sequence features .
these cluster ages provide several advantages over adopting literature values .
1 ) the ages are based on a uniform system of isochrones , while literature values adopt wide ranging models that have systematic differences that become more pronounced at younger ages .
2 ) the fitting techniques applied are by eye but consistent , while fitting techniques for literature values can greatly vary .
3 ) the difficulty of peculiar turnoff stars are addressed in a systematic way , while their consideration can have important differences in the literature values , if they are considered at all .
we will not comment on the absolute accuracy of the various isochrone model ages , but in this study uniformity and precision is the goal .
we must also reiterate two remaining limitations with our cluster parameters .
first , all of our photometries are primarily from differing groups , which still may leave important systematics remaining in our parameter analyses . additionally , uniformly measured spectroscopic metallicities are also needed to address the metallicity sensitivity in the turnoff isochrone fits .
lastly , for sirius b there is no parent cluster that we can self - consistently analyze for the total age , but the age of the sirius system is well studied and here we adopt solar metallicity and the age of [email protected] myr determined in liebert et al .
, h@xmath7 , h@xmath8 , h@xmath9 , and h8 fits are shown from bottom to top . ] ( solid line ) , 1.0 , 0.8 , and 0.6 m@xmath1 ( dotted lines ) , and 0.4 m@xmath1 ( dashed line ) . the mean distance modulus and reddening derived from the white dwarf members
are applied to these cooling models . ]
cluster membership determination is key in analyzing the formation history of these white dwarfs and applying them to the ifmr . in the case of ngc 2323
, we have the advantage that because it is such a young cluster , any hot and young high - mass white dwarfs observed in its field would already have reliable membership .
for example , in the sloan digital sky survey sample of field white dwarfs , only @xmath02.6% of da white dwarfs have a mass greater than 1 m@xmath1 and less than 10% of these are young with a t@xmath10@xmath2820,000 k ( e.g. , kleinman et al .
2013 ; kepler et al .
2016 ) . to further confirm cluster membership in ngc 2323 , however ,
we have also performed magnitude and color analysis . in table 3
we list the observed photometry for these white dwarfs and also list their cooling age and model based m@xmath12 and ( b - v)@xmath31 . for our membership analysis
we directly compare the model based and observed magnitudes and colors in figure 5 to derive an effective distance modulus and reddening for each white dwarf .
we define our 1@xmath36 color and magnitude errors by adding in quadrature the respective model fitting errors , the observational errors , and in the case of magnitude the distance modulus errors or in the case of color the reddening errors .
we select white dwarfs as consistent with membership if their effective distance modulus and reddening are within 2@xmath36 of the cluster parameters derived in section 3 . of our six
high signal white dwarfs , only wd10 , wd11 , and wd21 have both magnitudes and colors consistent with membership .
however , wd21 has a cooling age longer than the age of the cluster .
therefore , we do not consider it a member but a field white dwarf at comparable distance to ngc 2323 . in figure 6
we show the balmer line fits of wd10 and wd11 , the two cluster members .
they both are high mass at @xmath01.07 m@xmath1 and have very high temperature ( @xmath2850,000 k ) with accordingly extremely short cooling times of @xmath03 myr .
their membership further suggests that all of our other observed white dwarfs , which all have both lower masses and longer cooling times are not consistent with ngc 2323 membership . in figure 7
we display the cmd of all observed candidates relative to several white dwarf cooling models taken from http://www.astro.umontreal.ca/.17exbergeron/coolingmodels/ ( holberg & bergeron 2006 , kowalski & saumon 2006 , tremblay et al .
2011 , and bergeron et al .
l c c c c c c c c * - reanalyzed white dwarf initial and final parameters * + id&t@xmath10&log g&m@xmath11 & m@xmath12 & t@xmath13 & y@xmath3 m@xmath14 & parsec m@xmath14 & s / n + & ( k ) & & ( m@xmath1 ) & & ( myr ) & ( m@xmath1 ) & ( m@xmath1 ) & + ngc 2287 - 2 & 25900@xmath17350 & [email protected] & [email protected] & [email protected] & 76@xmath37 & 4.61@xmath38 & 4.81@xmath39 & 150 + ngc 2287 - 4 & 26500@xmath17350 & [email protected] & [email protected] & [email protected] & 127@xmath40 & 5.50@xmath41 & 5.93@xmath42 & 130 + ngc 2287 - 5 & 25600@xmath17350 & [email protected] & [email protected] & [email protected] & 77@xmath37 & 4.63@xmath38 & 4.83@xmath39 & 170 + ngc 2516 - 1 & 30100@xmath17350 & [email protected] & [email protected] & [email protected] & 42@xmath43 & 4.85@xmath44 & 5.19@xmath45 & 160 + ngc 2516 - 2 & 35500@xmath17550 & [email protected] & [email protected] & [email protected] & 24@xmath46 & 4.61@xmath47 & 4.88@xmath48 & 90 + ngc 2516 - 3 & 29100@xmath17350 & [email protected] & [email protected] & [email protected] & 48@xmath46 & 4.94@xmath48 & 5.31@xmath49 & 170 + ngc 2516 - 5 & 32200@xmath17400 & [email protected] & [email protected] & [email protected] & 38@xmath50 & 4.80@xmath51 & 5.12@xmath38 & 170 + ngc 3532 - 1 & 23100@xmath17300 & [email protected] & [email protected] & [email protected] & 131@xmath52 & 4.17@xmath44 & 4.13@xmath44 & 240 +
ngc 3532 - 5 & 27700@xmath17350 & [email protected] & [email protected] & [email protected] & 31@xmath50 & 3.57@xmath53 & 3.55@xmath53 & 220 + ngc 3532 - 9 & 31900@xmath17400 & [email protected] & [email protected] & [email protected] & 9.3@xmath54 & 3.48@xmath19 & 3.46@xmath19 & 210 + ngc 3532 - 10 & 26300@xmath17350 & [email protected] & [email protected] & [email protected] & 51@xmath55 & 3.66@xmath56 & 3.64@xmath56 & 200 + ngc 3532-j1106 - 590 & 21100@xmath17350 & [email protected] & [email protected] & [email protected] & 163@xmath57 & 4.46@xmath58 & 4.41@xmath58 & 110 + ngc 3532-j1106 - 584 & 20200@xmath17300 & [email protected] & [email protected] & [email protected] & 197@xmath59 & 4.91@xmath60 & 4.83@xmath60 & 120 + ngc 3532-j1107 - 584 & 20700@xmath17300 & [email protected] & [email protected] & [email protected] & 211@xmath61 & 5.16@xmath62 & 5.07@xmath62 & 180 + sirius b & 26000@xmath17400 & [email protected] & [email protected] & [email protected] & 99@xmath63 & 4.69@xmath64 & 4.69@xmath64 & - + pleiades - lb 1497 & 32700@xmath17500 & [email protected] & [email protected] & [email protected] & 54@xmath65 & 5.86@xmath66 & 6.85@xmath67 & 170 + pleiades - gd50 & 42700@xmath17800 & [email protected] & [email protected] & [email protected] & 70@xmath40 & 6.41@xmath68 & 8.05@xmath69 & - + pleiades - pg0136 + 251 & 41400@xmath17800 & [email protected] & [email protected] & [email protected] & 48@xmath24 & 5.64@xmath70 & 6.52@xmath71 & - + ngc 2168-lawds1 & 33500@xmath17450 & [email protected] & [email protected] & [email protected] & 19@xmath50 & 4.27@xmath72 & 4.48@xmath73 & 95 + ngc 2168-lawds2 & 33400@xmath17600 & [email protected] & [email protected] & [email protected] & 25@xmath74 & 4.33@xmath75 & 4.55@xmath76 & 70 + ngc 2168-lawds5 & 52700@xmath17900 & [email protected] & [email protected] & [email protected] & 1.0@xmath77 & 4.10@xmath19 & 4.28@xmath19 & 210 + ngc 2168-lawds6 & 57300@xmath171000 & [email protected] & [email protected] & [email protected] & 0.5@xmath77 & 4.10@xmath19 & 4.28@xmath19 & 260 + ngc 2168-lawds11 & 19900@xmath17350 & [email protected] & [email protected] & [email protected] & 149@xmath57 & 7.93@xmath78 & 11.60@xmath79 & 100 + ngc 2168-lawds12 & 34200@xmath17500 & [email protected] & [email protected] & [email protected] & 36@xmath65 & 4.44@xmath44 & 4.69@xmath38 & 100 + ngc 2168-lawds14 & 30500@xmath17450 & [email protected] & [email protected] & [email protected] & 54@xmath63 & 4.67@xmath80 & 4.98@xmath58 & 100 + ngc 2168-lawds15 & 30100@xmath17400 & [email protected] & [email protected] & [email protected] & 64@xmath81 & 4.80@xmath82 & 5.16@xmath83 & 130 + ngc 2168-lawds22 & 53000@xmath171000 & [email protected] & [email protected] & [email protected] & 1.0@xmath77 & 4.10@xmath19 & 4.28@xmath19 & 250 + ngc 2168-lawds27 & 30700@xmath17400 & [email protected] & [email protected] & [email protected] & 78@xmath24 & 5.06@xmath84 & 5.49@xmath85 & 120 + ngc 2168-lawds29 & 33500@xmath17450 & [email protected] & [email protected] & [email protected] & 34@xmath55 & 4.42@xmath86 & 4.67@xmath87 & 120 + ngc 2168-lawds30 & 29700@xmath17500 & [email protected] & [email protected] & [email protected] & 33@xmath88 & 4.41@xmath48 & 4.65@xmath39 & 60 + with the white dwarf cluster members , a simple comparison of their cooling age to the total cluster age gives the evolutionary time to the tip of the asymptotic giant branch ( agb ) for their progenitor .
application of this time to evolutionary models gives the white dwarf s progenitor mass .
in paper i we adopted the models of hurley et al .
( 2000 ) for the progenitor masses in ngc 2099 ( 3 - 4 m@xmath1 ) . at these masses ,
the difference between predicted progenitor mass by the hurley et al .
( 2000 ) models and the parsec isochrones are less than 1% .
we note that the parsec models do not include the thermally pulsing - agb phase , but this phase is very short and does not meaningfully affect the resulting progenitor masses .
for the y@xmath3 isochrones we can not infer progenitor masses directly because these isochrones do not evolve beyond the rgb .
but we note that while the y@xmath3 isochrones predict slower evolution to the turnoff in these younger clusters , they predict more rapid evolution through the rgb .
this results in the total evolutionary time scales to the tip of the rgb being comparable in both model isochrones for all but the highest masses .
for example , as we reach evolutionary times of 100 myr or shorter ( progenitors of @xmath25.3 m@xmath1 ) , the systematic differences become significant between all three models . for our current analysis , with both our y@xmath3 and parsec ages ,
we use progenitor masses derived from the parsec isochrones because it will provide the strongest consistency with our cluster age fits . for ngc 2323-wd10 and wd11 , with their cooling times and our y@xmath3 isochrones based age for ngc 2323 of 140 myr ,
the corresponding progenitor masses for both white dwarfs are 4.69 m@xmath1 .
( see table 1 , where we also give the progenitor masses based on the cluster age errors [ 140@xmath1720 myr ] and the parsec based age of 115 myr . )
it is quite remarkable that we have two independently formed high - mass white dwarfs from the same cluster that are so consistent in both initial and final mass . across their lifetimes they both lost 77.2% of their total mass .
this argues for the consistency of single - star mass loss at higher masses .
to look at these data in context , in table 4 we present the initial - final mass data for 30 mid to very high - mass ( 0.73 - 1.25 m@xmath1 ) white dwarfs that have been self - consistently analyzed from publicly available data . in the left panel of figure 8 we plot the initial and final masses of the
two analyzed white dwarfs from ngc 2323 and the 30 white dwarfs we have reanalyzed from the literature , adopting the y@xmath3 ages . in the right panel of figure 8
we plot these same data , but with application of the parsec based ages .
we also compare these high - mass white dwarf data to the sample of 31 intermediate - mass white dwarfs taken from ngc 2099 , the hyades , and praesepe ( paper i ) .
for the higher - mass white dwarfs , while there is some dispersion in the trend there is only one clear outlier that is from ngc 2168 .
lawds11 is an extreme outlier with a far longer cooling time than the other members of ngc 2168 , giving it a massive progenitor . for clusters in the rich galactic plane , contamination from common field white dwarfs
is expected and likely explains this white dwarf .
proper motions may be necessary to further constrain its membership , but in our current analysis and ifmr fits we do not consider lawds11 . in our initial analysis from paper i
, we demonstrated that the intermediate - mass white dwarfs ( 0.70.9 m@xmath1 ) create a steep ifmr slope . with the continued adoption of y@xmath3 ages in this paper , which systematically derive older ages in young clusters , we find
there is no meaningful change in slope at higher masses and the data appears to be defined well with a weighted linear relation : @xmath89 the linear nature of the ifmr across such a broad range ( 36.5 m@xmath4 ) is of interest .
for example , this may suggest we can extrapolate to derive the progenitor of a chandrasekhar mass limit white dwarf to be @xmath07.75 m@xmath1 , but the still limited data at the highest masses and the remaining uncertainties in the evolutionary models suggests this is unreliable .
furthermore , in theoretical models a moderate turnover in the slope of the ifmr is predicted near initial mass of 4 m@xmath1 ( e.g. , marigo & girardi 2007 and meng et al .
this predicted turnover is the result of the second dredge - up , which only occurs in stars of @xmath04 m@xmath1 and higher .
this diminishes their core mass and hence their final white dwarf mass .
a comparable turnover could still be lost in our data s remaining scatter , causing further problems with a linear extrapolation . in figure 9
we more closely analyze the residuals of this linear data fit . to illustrate the importance of our adopted isochrones ,
the right panel of figure 8 shows a clear turnover in the ifmr near an initial mass of 4 m@xmath1 when adopting parsec isochrone based ages , consistent with theoretical predictions .
we note that while our comparison ifmr data from paper i adopts y@xmath3 isochrone ages , this is not the cause of the strong overturn .
this is because in these older clusters ages from paper i the systematics shift direction and the parsec isochrones give cluster ages 20 myr older than the y@xmath3 isochrones . for the intermediate - mass comparison ifmr
this would increase its slope by @xmath05% , and this would further increase the magnitude of the turnover .
we fit the data based on the parsec ages for our younger clusters with a two - piece linear function with a kink at initial mass of 4 m@xmath1 : @xmath90 this is a strong turnover , one that is stronger than predicted in marigo & girardi ( 2007 ) or meng et al .
( 2008 ) . taking a moment to focus specifically on ngc 3532 may help to clarify these systematic differences .
this is both because it is the only cluster with progenitors that cover both above and below initial mass of 4 m@xmath1 , and it is the cluster least effected by systematic differences between the y@xmath3 and parsec isochrones .
the ngc 3532 data alone is limited but consistent with an overturn , which suggests the y@xmath3 may be overestimating the ages of the youngest clusters , but the massive progenitors of the pleiades white dwarfs in the right panel of figure 8 conversely suggests that the parsec isochrones may be underestimating the ages of the older clusters .
further work is needed on isochrones and evolutionary times at these higher masses , but for now we continue to adopt our y@xmath3 isochrone results as final . in figure 10
we compare to multiple empirical ifmrs from weidemann ( 2000 ) , dobbie et al .
( 2006a ) , cataln et al .
( 2008 ) , casewell et al .
( 2009 ) , kalirai et al .
( 2009 ) , and salaris et al .
we find that our relation based on y@xmath3 ages gives that the ifmr is steeper at higher masses than all of these relations besides those of dobbie et al .
( 2006a ) and cataln et
al.(2008 ) , which have comparable slopes .
our relation is strongly consistent with that of cataln et al .
( 2008 ) , but the zeropoint of dobbie et al .
( 2006a ) is @xmath00.05 m@xmath1 lower .
therefore , in comparison to most previous relations , ours predicts that white dwarf mass increases more rapidly with increasing progenitor mass , and that overall mass - loss rates are lower .
this finding is primarily the result of our ages being systematically older due in part to our analysis cleaning the turnoffs of peculiar stars .
additionally , the y@xmath3 isochrones give systematically older ages in comparison to the commonly adopted girardi et al .
( 2000 ) isochrones .
for example , a consistent turnoff fit to the pleiades with the isochrones of girardi et al .
( 2000 ) gives a 100 myr age , 35 myr younger than the y@xmath3 isochrones .
the most remarkable difference to previous ifmrs is that the large scatter has significantly decreased by @xmath050% .
this not only illustrates the importance of doing uniform analysis of the white dwarf spectra , but of the consistent examination of the cluster parameters with detailed consideration of peculiar stars in the turnoffs .
our ifmr s relatively smaller scatter also begins to show that while variations in mass - loss rates may still occur during a progenitor s lifetime , in singly evolved stars it is likely quite minor and the previously observed scatter is not indicative of stochastic mass loss .
metallicity dependence of the ifmr is another important parameter to consider , but it will not be quantitatively analysed here with such incomplete metallicity information . for our three clusters pleiades ,
ngc 2168 , and ngc 2516 with spectroscopic metallicity determinations , however , we note that the high - mass white dwarfs from the metal - poor ngc 2168 primarily fall above the mean of the ifmr while the white dwarfs from the slightly metal - rich ngc 2516 fall below the mean .
direct comparison of the ngc 2323 white dwarfs to the rest of our sample shows that they deviate more from the relation than a majority of the others .
individually , their errors in white dwarf mass are large enough to make these deviations not appear significant , but both white dwarfs being so consistent suggests that this deviation may be telling us something . for example , in comparison to the progenitors of the three comparable mass white dwarfs in other clusters ( lb 1497 , ngc 2168-lawds27 , and ngc 2287 - 4 ) , their three progenitors are all higher mass with lb 1497 giving a @xmath01.2 m@xmath1 larger progenitor .
is this indicative of mass - loss variations between these different clusters or remaining systematics ?
for example , a possible explanation for this is that our assumed solar metallicity for ngc 2323 is not appropriate .
a more metal - rich isochrone would give a younger age and help to decrease the observed differences in progenitor masses . or this deviation may simply be the result of potential systematic differences between our adopted photometries , and in general this may be a cause for much of the remaining scatter in our total ifmr . in any case , this new ngc 2323 white dwarf analysis further emphasizes the importance of this nearby , rich , and young open cluster , and of the need for spectroscopic metallicity and precise ubv photometry spanning from its turnoff to its faint white dwarfs .
we have spectroscopically observed a sample of 10 white dwarf candidates in the young and rich ngc 2323 . while we only found that two of these white dwarfs are consistent with membership in ngc 2323 ,
they are both newly discovered high - mass white dwarfs at @xmath01.07 m@xmath1 . to supplement these new white dwarfs
we have reanalyzed all published , publicly available high - mass white dwarfs from the pleiades , ngc 2516 , ngc 2168 , ngc 2287 , and ngc 3532 star clusters and sirius b. at these higher masses , the inferred progenitor masses are increasingly dependent on the adopted star cluster ages .
therefore , we uniformly analyzed available high - quality ubv photometry for these clusters .
due to the difficulties of analyzing younger cluster ages , we have analyzed both color - color diagrams and color - magnitude diagrams .
the color - color diagrams provide a self - consistent manner to not only determined a cluster reddening but to identify peculiar stars in each cluster s turnoff . with the cleaned turnoffs for each cluster
, we can more precisely determine a cluster age .
in general , we find cluster ages moderately older than those adopted in previous ifmr analyses , and we attribute this both to cluster peculiar stars affecting previous age determinations and the systematically older ages derived from the y@xmath3 and parsec isochrones . application of our derived white dwarf cooling ages and cluster turnoff ages gives us the progenitor masses for each white dwarf .
the two massive white dwarfs from ngc 2323 at 1.07 m@xmath1 both find remarkably consistent progenitor masses of 4.69 m@xmath1 . with the addition of the publicly available high - mass white dwarfs ,
the derived ifmr is quite clean , and we find that with y@xmath3 ages the ifmr can be well defined by a linear relation from progenitor masses of 3 to 6.5 m@xmath1 .
in contrast , adoption of the only moderately younger cluster ages from the parsec isochrones gives a moderate turnover in ifmr slope near initial mass of 4 m@xmath1 . in our current analysis
we continue to adopt the y@xmath3 results , but this draws attention to the work that is needed on evolutionary timescales at these higher masses .
our ifmr also finds a significantly decreased scatter in comparison to recent ifmr work .
this shows that the typically observed scatter in the ifmr is likely not indicative of real variations in mass loss for singly evolved stars , but was the result of remaining systematics .
the ifmr relation is now reassuringly consistent across multiple star clusters .
metallicity , however , still may play an important role , in particular at these higher masses . to further improve the ifmr , not only
are newly discovered high - mass white dwarfs needed , but uniform ubv photometry and spectroscopic metallicity analyses are also needed .
this will further limit systematics that may still remain and give the foundation for detailed quantitative analysis of the metallicity dependence of the ifmr .
this project was supported by the national science foundation ( nsf ) through grant ast-1211719 .
this work was also supported by a nasa keck pi data award , administered by the nasa exoplanet science institute .
data presented herein were obtained at the w. m. keck observatory from telescope time allocated to the national aeronautics and space administration through the agency s scientific partnership with the california institute of technology and the university of california .
the observatory was made possible by the generous financial support of the w. m. keck foundation .
p - e.t . was supported during this project by the alexander von humboldt foundation and by nasa through hubble fellowship grant hf-51329.01 , awarded by the space telescope science institute , which is operated by the association of universities for research in astronomy , incorporated , under nasa contract nas5 - 26555 .
this research has made use of the keck observatory archive ( koa ) , which is operated by the w. m. keck observatory and the nasa exoplanet science institute ( nexsci ) , under contract with the national aeronautics and space administration .
lastly , we would like to thank c. deliyannis for his helpful discussions of open cluster color - color analysis . | we have observed a sample of 10 white dwarf candidates in the rich open cluster ngc 2323 ( m50 ) with the keck low - resolution imaging spectrometer .
the spectroscopy shows eight to be da white dwarfs , with six of these having high s
/ n appropriate for our analysis . two of these white dwarfs are consistent with singly evolved cluster membership , and both are high mass @xmath01.07 m@xmath1 , and give equivalent progenitor masses of 4.69 m@xmath1 . to supplement these new high - mass white dwarfs and analyze the initial - final mass relation ( ifmr ) ,
we have also looked at 30 white dwarfs from publicly available data that are mostly all high - mass ( @xmath20.9 m@xmath1 ) .
these original published data exhibited significant scatter , and to test if this scatter is true or simply the result of systematics , we have uniformly analyzed the white dwarf spectra and have adopted thorough photometric techniques to derive uniform cluster parameters for their parent clusters .
the resulting ifmr scatter is significantly reduced , arguing that mass - loss rates are not stochastic in nature and that within the ranges of metallicity and mass analyzed in this work mass loss is not highly sensitive to variations in metallicity .
lastly , when adopting cluster ages based on y@xmath3 isochrones , the slope of the high - mass ifmr remains steep and consistent with that found from intermediate - mass white dwarfs , giving a linear ifmr from progenitor masses between 3 to 6.5 m@xmath1 .
in contrast , when adopting the slightly younger cluster ages based on parsec isochrones , the high - mass ifmr has a moderate turnover near an initial mass of 4 m@xmath1 . |
optical scattering , induced by structural disorder , can increase the dwell time of light inside an absorbing material , thus enhancing the chance of absorption @xcite .
compared to other light - trapping schemes that have been developed for solar cells , the non - resonant nature of this scheme enables broadband enhancement for light with a wide angle of incidence @xcite .
however , the degree of absorption does not depend solely on the absorption coefficient and scattering properties of the medium , but also on the coherence and spatial characteristics of the input light .
one recent example confirming this is the demonstration of coherent perfect absorption , which is the time - reversed analog of laser emission @xcite .
similar to lasing , this effect is narrow - band and requires tuning of both the frequency of the input light and absorptivity of the material to achieve 100% absorption @xcite .
a related but distinct approach , termed coherently enhanced absorption , takes advantage of the large number of degrees of freedom available for multiple scattering in a lossy environment to control the absorption @xcite .
it has been shown theoretically an appropriate coherent superposition of input fields can significantly enhance the coupling of light into a scattering medium , increasing the absorption at any frequency over a broad range .
recent advances in wavefront shaping ( wfs ) techniques using spatial light modulators ( slms ) have enabled the generation of arbitrary wavefronts at the input of an optical system .
this technique is driving forward the development of optical imaging techniques through scattering media such as biological tissue and fog @xcite , and also enables the control of light intensity @xcite , polarization state @xcite , spectrum @xcite , or temporal profile @xcite at the scale of wavelength ( single or a few speckles ) .
while the wavelength - scale control is desired for imaging with high resolution @xcite , the ability to manipulate light distributions on a much larger scale in a disordered system is needed for many applications .
one example is photoelectrochemical cells , which often utilize light scattering to increase photon dwell time to enhance absorption .
as the photoelectrons are generated in a volume that contains hundreds of thousands of speckles , increasing the intensity of a single speckle even by a few thousand times makes little difference to the total photocurrent .
hence , global control of the light distribution is necessary , but this is known to be a much more challenging optimization problem than local control .
good progress has been made recently to enhance the total transmission through a lossless opaque medium via wfs of an incident laser beam @xcite .
however , coherent control of light distribution and absorption over a macroscale region in highly scattering media has never been demonstrated .
here we use wfs to both enhance and suppress the photocurrent in a dye - sensitized solar cell ( dssc ) , with a photoanode comprised of a highly scattering layer of dye - sensitized tio@xmath0 nanoparticles .
since the collection efficiency of charge carriers is highly dependent on where light is absorbed @xcite , it is advantageous to trap light inside the scattering layer as close as possible to the interface with the electrode , where charges can be efficiently separated , thus minimizing charge recombination @xcite .
however , strong light leakage through the interface makes it extremely difficult to confine light to a region close to the open boundary and enhance the absorption there .
but here we show a significant enhancement of photocurrent by optimizing the wavefront of an incident laser beam , and illustrate the underlying mechanism via numerical modeling .
in addition , we show that wfs can also _ suppress _ absorption by increasing the amount of light reflected .
this work demonstrates the great potential of wfs toward depositing energy into targeted regions inside turbid media , which will enable the spatially resolved characterization of light absorption and charge transport inside regenerative and artificial photosynthetic systems .
nm is split to two orthogonal polarizations and projected onto the slm .
the phase front of both polarizations are modulated and then recombined at a beam splitter .
the shaped wavefront is imaged onto the surface of the porous electrode in the dssc using a microscope objective .
the illuminated area is a disk of diameter @xmath1 @xmath2 m .
the short - circuit current of the solar cell is measured using a potentiostat . ]
a dssc comprises a porous layer of sintered tio@xmath0 nanoparticles covered by a molecular dye , in this case the ruthenium - based n719 dye @xcite , deposited on a conductive glass substrate .
it functions as the light - absorbing working electrode of the solar cell , while a platinum - loaded conductive glass slide acts as the counter electrode .
an electrolyte fills the gap between the two electrodes and shuttles charges between them .
the working electrode of a dssc constructed in this manner can be considered as a scattering sample with homogeneous absorption @xcite .
figure [ setup]a presents a schematic of the solar cells created in this study .
light is incident onto the porous tio@xmath0 layer through the glass substrate coated with transparent conducting oxide .
after light absorption , the excited dye molecules rapidly inject electrons into the tio@xmath0 .
these electrons diffuse through the sintered particle network to be collected at the front side electrode , while the dye is regenerated via reduction by a redox mediator in the electrolyte , in this case the iodide / triiodide redox couple .
the photo - induced current from a dssc is typically measured without applying a bias voltage , and is known as the short - circuit current @xmath3 .
not all photo - generated electrons can diffuse through the tio@xmath0 layer and reach the electrode , because electrons may recombine or get trapped on the way @xcite .
the farther the electrons are generated from the electrode , the lower their probability of reaching it and contributing to the current .
hence , @xmath3 depends not only on the amount of light being absorbed but also on the position where the light is absorbed inside the tio@xmath0 layer @xcite . in the wfs experiment ,
the solar cell is excited by a narrow band laser beam at wavelength @xmath4 = 532 nm .
the input wavefront is modulated using the experimental setup schematically shown in fig .
[ setup]b .
the two linear polarizations are modulated independently @xcite ( see methods ) and recombined into a single beam , which is projected onto the scattering medium in the solar cell using a microscope objective ( na = 0.95 ) .
we measure @xmath3 generated by the dssc for each input phase mask and use a genetic algorithm to find the wavefronts that maximize or minimize @xmath3 .
the genetic algorithm is known to be robust against noise in wfs experiments @xcite .
a cost function @xmath5 , which estimates the variation of the conversion efficiency of the solar cell , is defined as : @xmath6 where @xmath7 and @xmath8 are the average short - circuit current and input optical power , respectively , measured for a set of random phase masks .
@xmath9 is the small dark current measured at @xmath10w ( _ si _ ) . for the short - circuit current , normalized to the value with random input wavefront
( green dashed line ) , increases ( blue solid line ) or decreases ( red dotted line ) during the optimization of the input wavefront by a genetic algorithm .
the photo - induced current is enhanced @xmath11 times or suppressed @xmath12 times by wavefront shaping .
( b ) output electrical power of the dssc illuminated by a random phase pattern ( dashed green line ) , and the two optimized wavefronts for maximal @xmath5 ( solid blue line ) and minimal @xmath5 ( dotted red line ) .
the incident laser power is fixed at @xmath13w .
the asterisk indicates the maximal power , which is enhanced 4.7 times or reduced 5.7 times . ]
we constructed the dssc using standard procedures ( see methods and _ si _ ) .
the transport mean free path of light in the @xmath14 @xmath2 m thick tio@xmath0 layer , @xmath15 , is measured in a coherent backscattering experiment to be 3.2 @xmath2 m @xcite . to shorten the response time of the solar cell in order to perform wfs experiments on reasonable time scales , we use a concentrated non - volatile iodine - based electrolyte that also absorbs light at 532 nm .
the absorption of light by the electrolyte is proportional to that by the dye at any location , as both scale linearly with local field intensity .
thus the short - circuit current @xmath3 , which results from light absorption by the dye , also reflects the absorption of the electrolyte , which does not contribute directly to the current .
the ballistic absorption length @xmath16 is about 65 @xmath2 m , and the diffusive absorption length ( the typical distance light will travel via multiple scattering prior to being absorbed ) , @xmath17 , is 8.4 @xmath2 m @xcite . without absorption ,
the total transmission @xmath18 of the tio@xmath0 layer is estimated to be @xmath19 , and the reflection @xmath20 . in the presence of absorption , @xmath18 is reduced to @xmath21 , and @xmath22 to @xmath23 , with the remaining @xmath24 being absorbed ( in the absence of wavefront optimization ) .
these estimations agree well with the experimental measurement ( _ si _ ) . by optimizing the incident wavefront ,
the absorption cost function @xmath5 is gradually increased or decreased as shown in fig .
[ results1]a . with the optimal wavefronts ,
the photocurrent is enhanced by a factor of @xmath25 or suppressed by a factor of @xmath26 .
we are thus able to coherently vary the energy conversion efficiency by a ratio of @xmath27 . to quantify the variation of the output power due to the modification of the input optical pattern , we measure the current - voltage ( @xmath28 ) curves for the two optimized wavefronts corresponding to maximum and minimum @xmath5 , as well as for a random illumination pattern .
figure [ results1]b displays the light - induced electrical output power , given by @xmath29 , as a function of bias voltage for each scenario .
the maximum output power is enhanced 4.7 times or reduced 5.7 times , leading to total variation by a factor of 27 . since approximately @xmath24 of the light
is absorbed in the solar cell for a random illumination , the highest possible enhancement of the absorption would be a factor of 2.2 .
if @xmath3 were proportional to the total amount of light being absorbed @xcite , the experimental enhancement of 2.1 of the short circuit current would correspond to an almost perfect absorption of light in the absorbing medium .
however , achieving the maximum absorption requires complete control over the amplitude and phase of the light in all input channels @xcite .
experimentally we control only a limited fraction of the input channels due to the finite numerical aperture of the objective lens and the fact that the probe beam is incident only on one side of the scattering sample .
moreover , our slm setup only modulates the phase and not the amplitude of the incident light , as would be necessary to achieve the global optimum . consequently
, the enhancement of total absorption that could be achieved with wfs is significantly reduced @xcite .
to investigate the cause for such a high enhancement factor measured experimentally , we perform numerical modeling . without loss of generality , we simulate the scattering medium confined in a 2d waveguide with reflecting sidewalls to reduce the computation load ; the system dimensions , as well as the amount of scattering and uniform absorption , are chosen to be near to the experimental values ( see methods ) .
light with transverse - magnetic polarization ( @xmath30 field perpendicular to the waveguide plane ) is injected from one end of the waveguide , and its propagation in the random system is calculated with the recursive green s function method @xcite . the transmission and reflection of light for the systems with and without absorption , averaged over random input wavefronts , match the measured values of real samples . to model the effect of wfs in our experiment
, we use the empty waveguide modes as the basis to shape the input field . during the optimization process
, we modulate the phase of each empty waveguide modes while keeping their amplitude constant .
the ( linear ) space - dependent absorption of light is proportional to the local intensity @xmath31 , and the total absorption is proportional to the spatially integrated intensity . to take into account the spatial variation of the electron collection efficiency
@xmath32 $ ] , where @xmath33 is the electron diffusion length , we define a new cost function @xmath34 which models the relative change of the short - circuit current similarly to @xmath5 : @xmath35 where @xmath36 is the electric field distribution in the scattering medium for the optimized input wavefront , @xmath37 is the field distribution for a random input wavefront , and @xmath38 represents the average over random input wavefronts . as in the solar cell , the absorption of light that occurs far from the front surface of the random medium contributes little to @xmath34 .
we used both a genetic algorithm and a sequential algorithm @xcite to maximize or minimize @xmath34 , and found similar results , although the sequential algorithm converges faster than the genetic algorithm due to the absence of noise in simulation . , defined in eq.(2 ) , increases and then saturates when the input wavefront is optimized for @xmath32 $ ] ( red solid line ) and @xmath39 ( blue dashed line ) using a sequential algorithm .
@xmath33 is taken to be 3 @xmath2 m ( see methods ) .
( b ) the cross - section integrated intensity in the random system illuminated by random input wavefronts ( black dotted line ) and the optimized wavefront obtained with the non - uniform cost function for @xmath32 $ ] ( red solid line ) and the uniform cost function for @xmath39 ( blue dashed line ) .
the light is concentrated near the front side of the random system for @xmath32 $ ] , but penetrates deeper into the system for @xmath39 .
( c , d ) same as ( a , b ) for the minimization of @xmath34 . for @xmath32 $ ]
, the light intensity near the front side of the random system is reduced , but increased deeper into the system . for @xmath39 ,
the light intensity is reduced in the interior of the system but remains almost the same near the front surface . ]
figure [ rgfsim]a shows the enhancement of the cost function @xmath34 in the random medium with a uniform absorption coefficient . for comparison we also simulate the case of @xmath39 , which corresponds to a uniform cost function similar to the standard coherent enhancement of global absorption @xcite .
both the non - uniform ( @xmath32 $ ] ) and uniform cost functions increase continuously in the optimization process and eventually saturate ; however as expected , the final enhancement factor for @xmath32 $ ] significantly exceeds that for @xmath39 . figure [ rgfsim]b represents the integrated intensity @xmath40 inside the random medium . with random wavefront illumination , @xmath41 decays exponentially in the disordered system , as expected for light diffusion in the presence of absorption @xcite .
when the input wavefront is optimized for uniform cost function ( @xmath39 ) , @xmath41 is peaked in the interior of the random system .
the buildup of energy deep inside the scattering medium enhances the total absorption from @xmath24 to @xmath42 .
meanwhile , the reflection decreases from @xmath23 to @xmath43 , and the transmission slightly increases from @xmath21 to @xmath43 .
this result confirms that the global absorption is enhanced by wfs through causing the light to penetrate deeper into the lossy disordered medium .
in contrast , the optimization for the non - uniform cost function ( @xmath32 $ ] ) leads to a significant increase of light intensity close to the front surface of the random medium [ fig .
[ rgfsim]b ] , since only this region contributes significantly to the cost function @xmath34 .
naively , the proximity to the open boundary would be expected to facilitate light escape from the absorbing medium , thus reducing the absorption and increasing the reflection . instead , for the optimal wavefront , the total absorption is increased from @xmath24 to @xmath44 , and the reflection decreased from @xmath23 to @xmath45 .
this result suggests some partial trapping of the light near the surface , leading to a buildup of its intensity in the vicinity of the open boundary where the front side electrode is located .
hence , the numerical simulation illustrates that our experimental procedure , which does feedback optimization based on the short - circuit current , does not simply optimize total absorption , but optimizes absorption locally , near the working electrode , so as to preferentially enhance the collected current .
the measured current increase by a factor of 2.1 does not imply perfect absorption of all incident light and can be realized with the incomplete experimental control .
to understand how the leakage rate of the light close to the open boundary is reduced , we investigate the contribution of different quasi - modes in the system to the reflected light .
it has been shown previously that light propagation in random media can be well explained in terms of the interference effects of quasi - modes @xcite .
we take this approach by decomposing the electric field inside the random medium in terms of the quasi - modes of the system : @xmath46 where @xmath47 ( @xmath48 ) is the @xmath49-th mode with purely outgoing ( incoming ) wave boundary condition , and @xmath50 @xcite .
the reflected field at the front surface ( @xmath51 ) of the disordered medium can be expressed as @xmath52 .
the total intensity of reflected waves is @xmath53 , where @xmath54 represents the contribution of the @xmath49-th mode to the reflection .
while @xmath55 is a complex number , @xmath22 is a real number , and the contributions of different modes to @xmath22 interfere with each other .
the interference effect can be quantified by @xmath56 ( _ si _ ) .
if @xmath57 ( @xmath58 ) , the modes interfere constructively ( destructively ) at the front surface to enhance ( suppress ) reflection . of individual modes @xmath49 supported by the random system to the total intensity of reflected waves .
the modes are excited by a random input wavefront ( a ) , the optimized wavefront for maximal @xmath34 ( b ) and minimal @xmath34 ( c ) with @xmath32 $ ] . in ( a )
the modes have random phases and their interference is averaged out , @xmath59 . in
( b ) the modes interfere destructively , @xmath60 , to reduce reflection and enhance absorption . in ( c )
the modes interfere constructively , @xmath61 , to enhance reflection and reduce absorption . ] to reduce the computing time for the full modal analysis , the length and width of the system in the simulation is reduced by a factor of two .
the scattering and absorption parameters are adjusted so that the total transmission , reflection and absorption for random input wavefronts are essentially unchanged .
we use the finite - difference frequency - domain ( fdfd ) method to calculate the quasi - modes whose frequencies are close to the frequency of the input light .
the calculated values of @xmath55 for the different scenarios with @xmath32 $ ] are plotted in the complex plane in figs .
[ modepic]a - c .
the modes have uncorrelated phases for random wavefront illumination and their interference is averaged out ( and therefore @xmath62 is close to 1 ) .
we see that the optimized wavefront for enhanced local absorption causes a destructive interference among the modes , making @xmath62 much smaller than 1 , and more importantly , that reflection is minimized and absorption is maximized at the front surface . despite the complexity of the medium , wfs exploits intereferences to efficiently concentrate light in the targeted region of the system .
similarly , we show that wfs can also reduce the energy in a chosen region by minimizing the cost function @xmath34 [ figs .
[ rgfsim]c and d ] . for @xmath39 , the reflection
is greatly enhanced , decreasing the light intensity deep inside the system and thus the total amount of absorption . for @xmath32 $ ]
, the minimization of the cost function is less efficient as it requires us to _ simultaneously _ increase the reflection and decrease the absorption close to the front surface . by decomposing the fields into the quasi - modes of the system , we confirm that the increase in reflection is caused by the constructive interference of excited modes [ fig . [ modepic]c ] , which facilitates light escape and suppresses absorption .
in summary , we have experimentally demonstrated coherent control of light distribution and absorption over a length scale much larger than the wavelength of light inside an opaque mesoporous photoelectrochemical system that utilizes multiple scattering to improve light confinement .
our proof - of - concept experiment shows that it is possible to vary the photocurrent in a dye - sensitized solar cell ( dssc ) by tailoring the input wavefront of a laser beam . to maximize the dssc s short - circuit current
, the optimal wavefront concentrates the input light to the front side of the porous photoanode , where photo - generated electrons are most efficiently collected . despite the proximity to an open boundary
, we find that destructive interference among the excited modes minimizes light leakage and enhances absorption .
although in the current wfs experiment a narrow - band laser is used for coherent control of absorption , a recent theoretical study suggests that this approach is applicable to a relatively broadband light @xcite .
more generally , our method , assisted by a guidestar @xcite , will allow for spatially - resolved characterization of photo - induced processes inside opaque media , such as photoelectrochemical cells .
this approach opens the possibility of investigating photoelectrochemical processes in targeted regions inside dense scattering media , such as that used in regenerative and artificial photosynthetic solar cells .
those studies will provide physical insight into the operation of photovoltaic and photoelectrochemical systems , potentially leading to new methods to improve photoconversion performance .
* sample preparation . * to facilitate the wfs experiment
, we made a dssc with a thicker tio@xmath0 scattering layer , reduced loading of n719 dye , and a thin ( 175 @xmath2 m thick ) tin - doped indium oxide ( ito ) coated glass substrate for the working electrode .
the porous tio@xmath0 working electrode is made from @xmath63 nm tio@xmath0 nanoparticle paste ( p25 , degussa ) . the paste was doctor - bladed onto a 175 @xmath2 m thick ito - coated glass coverslip ( spi , product no .
06477b - ab ) 5 times .
the sample was heated to 200@xmath64c for 10 min then cooled to room temperature between each application of a layer of paste .
the final thickness of the tio@xmath0 layer is @xmath14 @xmath2 m .
the working electrode is sintered at 450@xmath64c for 1 hour , then immersed in a solution of n719 ( ditetrabutylammonium cis - bis(isothiocyanato)bis(2,2-bipyridyl-4,4-dicarboxylato ) ruthenium(ii ) for 30 minutes @xcite .
the counter electrode is made by heating hexachloroplatinic acid in ethanol ( 2 mg / ml ) to 400@xmath64c for 15 minutes on a fluorine - doped tin oxide ( fto ) coated glass slide .
the solar cells are assembled with a 25 @xmath2 m surlyn spacer hot - pressed between the electrodes and filled with a non - volatile iodine - based electrolyte ( mosalyte tde-250 , solaronix sa , switzerland ) .
* optical setup . * a shown schematically in fig . [ setup ] , the output beam from a continuous - wave nd : yag laser is expanded ( lenses l and l ) and projected onto a phase - only slm ( hamamatsu , x10468 - 01 ) . a half - wave - plate ( @xmath65 )
is used to balance the light intensities of two orthogonal linear polarizations . to control independently the two polarizations ,
a polarizing beam splitter ( pbs ) attached to a right - angle prism ( p ) separates the two polarizations of the laser beam , which are modulated by different areas of the slm .
the cube and the prism are mounted in the same holder to eliminate fluctuations from independent reflective elements in the two paths . a second half - wave - plate ( @xmath66 )
rotates the polarization to match that of the slm .
the two illuminated areas on the slm corresponding to the two polarizations of the input light are divided into @xmath67 macropixels .
the modulated wavefronts are then recombined and projected onto the pupil of a microscope objective ( mo ) with numerical aperture @xmath68 . a beam splitter ( bs@xmath0 ) along with a photo - detector ( pd ) are used to measure the optical power @xmath69 impinging on the solar cell .
the photo - induced current of the dssc is measured by a potentiostat ( digi - ivy dy2116b ) .
* numerical simulation . * in the 2d model , dielectric cylinders with radius 50 nm and refractive index 2.5 are randomly positioned inside a waveguide of width @xmath70 @xmath2 m and refractive index 1.5 .
the length of the random structure is @xmath71 @xmath2 m , and the effective index of refraction is 1.9 .
the wavelength of input light is @xmath72 nm , and the number of transverse guided modes in the waveguide is 114 .
the transport mean free path @xmath15 is @xmath73 @xmath2 m . without absorption and wavefront optimization ,
the transmission is @xmath74 and the reflection is @xmath22 = @xmath75 .
a uniform absorption coefficient is introduced via the imaginary part of the refractive index to the random system .
the diffusive absorption length is @xmath76 @xmath2 m .
@xmath18 is reduced to @xmath21 , and reflection to @xmath23 , thus the absorption @xmath77 .
these numbers match the measured values of our dssc .
the electron diffusion length @xmath33 is chosen to be @xmath78 @xmath2 m , close to the value of dsscs similar to ours @xcite . in the modal analysis to decompose the field distribution in the random medium with the quasi - modes , we keep the wavelength @xmath4 the same , but reduce the number of waveguide modes to 37 .
@xmath15 is reduced to 2 @xmath2 m , and the diffusive absorption length to 4.2 @xmath2 m . without absorption , @xmath79 and @xmath80 ; with absorption , @xmath81 , @xmath82 , and @xmath83 .
the electron diffusion length @xmath33 is also shortened to 1.5 @xmath2 m .
more details on the synthesis and optical characterization of the dssc .
additional numerical simulation results of wfs to maximize and minimize the cost function @xmath34
. detailed description of modal analysis on the electric field distribution inside the random medium .
s. f. liew and s. m. popoff contributed equally to this work .
we thank chia wei hsu for stimulating discussions . this work is supported by the nsf under grant no .
s.w.s . acknowledges the support of an nsf graduate research fellowship under dge 1122492 .
judkewitz , b. ; wang , y. m. ; horstmeyer , r. ; alexandre , m. ; yang , c. speckle - scale focusing in the diffusive regime with time reversal of variance - encoded light ( trove ) .
photon . _ * 2013 * , _ 7 _ , 300305 .
oekermann , t. ; zhang , d. ; yoshida , t. ; minoura , h. electron transport and back reaction in nanocrystalline tio@xmath0 films prepared by hydrothermal crystallization .
_ j. phys .
b _ * 2004 * , _ 108 _ , 22272235 .
sheehan , s. w. , noh , h. , brudvig , g. w. , cao , h. & schmuttenmaer , c. a. plasmonic enhancement of dye - sensitized solar cells using core
shell shell nanostructures .
_ j. phys .
c _ * 2013 * , _ 117 _ , 927934 .
hsu , c. w. ; goetschy , a. ; bromberg , y. ; stone , a. d. ; cao , h. broadband coherent enhancement of transmission and absorption in disordered media .
lett . _ * 2015 * , _ 115 _ , 223901 .
horstmeyer , r. ; ruan , h. ; yang , c. guidestar - assisted wavefront - shaping methods for focusing light into biological tissue .
photon . _ * 2015 * , _ 9 _ , 563571 . | a fundamental issue that limits the efficiency of many photoelectrochemical systems is that the photon absorption length is typically much longer than the electron diffusion length .
various photon management schemes have been developed to enhance light absorption ; one simple approach is to use randomly scattering media to enable broadband and wide - angle enhancement .
however , such systems are often opaque , making it difficult to probe photo - induced processes . here
we use wave interference effects to modify the spatial distribution of light inside a highly - scattering dye - sensitized solar cell to control photon absorption in a space - dependent manner . by shaping the incident wavefront of a laser beam , we enhance or suppress photocurrent by increasing or decreasing light concentration on the front side of the mesoporous photoanode where the collection efficiency of photoelectrons is maximal .
enhanced light absorption is achieved by reducing reflection through the open boundary of the photoanode via destructive interference , leading to a factor of two increase in photocurrent .
this approach opens the door to probing and manipulating photoelectrochemical processes in specific regions inside nominally opaque media . |
the _ kepler _ mission discovered thousands of candidate `` super - earths , '' planets between the size of earth and neptune ( e.g. , @xcite ) .
they exhibit a wide range of bulk densities , anywhere from @xmath110 g/@xmath2 to less than that of water ( e.g. , @xcite ) .
much of our knowledge of their orbital properties comes from the subset of systems each containing two or more transiting planets .
statistical studies have found that systems of multiple super - earth `` tranets '' ( planets that transit ; @xcite ) have small eccentricities @xcite and small mutual inclinations @xcite of a few percent or less .
the orbital spacings of tranets @xcite have been a particular focus for theoretical studies of formation @xcite , stability @xcite , dynamical excitation @xcite , and migration @xcite . in this paper
, we define the spacing @xmath3 as the semi - major axis difference @xmath4 between adjacent planets , normalized by their mutual hill radius : @xmath5 where @xmath6 is the semi - major axis , @xmath7 is the planet mass , and @xmath8 is the stellar mass . the tranet spacing distribution peaks at @xmath9 .
@xcite found a similar spacing distribution for the underlying planets , assuming a single population of planetary systems drawn from independent distributions of mutual inclinations , spacings , multiplicities , and planet sizes .
planet spacings of @xmath9 shrinking to @xmath10 for the highest multiplicity systems lie safely outside the empirical stability limit of @xmath11 @xcite .
studies of tranet multiplicity a property that depends on both spacing and mutual inclination have found a `` _ kepler _ dichotomy , '' i.e. , a need for two populations to account for an apparent excess of single tranet systems over multi - tranet systems @xcite . here
we explore how the orbital properties of super - earths their spacings , eccentricities , and inclinations originate from the circumstances of their formation from primordial disks of solids and gas .
we will uncover correlations between these orbital properties and planet compositions ( gas - enveloped vs. rocky ) , correlations that are not due to observational bias and that are currently neglected in _
kepler _ population studies ( e.g. , @xcite ) .
we work in the context of in - situ formation ( e.g. , @xcite ; see also @xcite ) , simulating the assembly of super - earths starting from isolation masses .
we devote especial attention to the origin of super - earth spacings .
the spacing distributions of planets and protoplanets have been explored theoretically from several vantage points .
one consideration is stability .
the timescale for orbit crossing is known empirically to increase exponentially with the hill spacing @xmath12 @xcite .
the spacing must be wide enough for the orbit crossing time to exceed the system s age .
@xcite presented another point of view from statistical mechanics .
other perspectives derive from considerations of planet formation . for isolation masses ( a.k.a .
oligarchs ) accreting small bodies in their feeding zones , a spacing equilibrium is achieved between mass growth , which decreases the separation in hill radii , and orbital repulsion driven by gravitational ( a.k.a .
viscous ) stirring @xcite . in the particle - in - a - box approximation for planet formation by coagulation of small bodies ( e.g. , @xcite ) ,
random velocities are limited to the surface escape velocity @xmath13 from the most massive body .
@xcite applied this principle to estimate the minimum disk masses required to form super - earths in situ . a planet that excites the random velocities of surrounding bodies to the maximum value of @xmath13
has a feeding zone of full width @xmath14 where @xmath15 is the planet s orbital angular frequency ( a.k.a .
mean motion ) , @xmath16 is the orbital period , and @xmath17 is the planet s bulk density .
this feeding zone width also corresponds to the expected maximum spacing between nascent planets .
we will take a related approach by finding the spacing that allows for an `` eccentricity equilibrium '' : one that balances gravitational scatterings between protoplanets , which excite eccentricities , with mergers , which damp them .
our paper , which explores how the orbital spacings of super - earths are fossil records of their formation in situ , and how orbital properties in general are correlated with planet composition , is organized as follows . in section 2 , we describe the setup for the @xmath0-body simulations that are the basis for all subsequent sections . in section 3 ,
we show how spacings , eccentricities , and inclinations evolve with time for a few illustrative , gas - free simulations .
in particular , we demonstrate how the evolution is sensitive to initial inclinations an input parameter that , to our knowledge , is not often highlighted as a controlling parameter in coagulation calculations .
we offer order - of - magnitude scalings to understand these results , explaining the dependence of spacings on inclinations in terms of an eccentricity equilibrium that balances scatterings with mergers .
section 4 gives a more complete overview of our gas - free simulations , detailing how final spacings depend on various initial conditions .
section 5 brings dynamical friction by disk gas into the mix ; we investigate how gas can establish those initial conditions that were assumed in preceding sections . including gas also enables us to connect planet compositions whether or not planets have volumetrically significant gas envelopes with orbital properties ; this is the focus of section 6 .
comparisons with observations are made in section 7 ; there we will find that we need a mixture of dynamically hot and dynamically cold populations to more faithfully reproduce _
kepler _ data .
we present our conclusions in section 8 .
to assess how orbital spacings depend on conditions during the late stages of planet formation , we simulate the growth of isolation mass embryos to super - earths via collisions ( a.k.a .
giant impacts ) . building on work by , e.g. , @xcite , @xcite , @xcite , and @xcite , we perform @xmath0-body integrations each lasting 27 myr using the hybrid symplectic integrator of mercury6 @xcite .
the time step is 0.5 days and a close encounter distance ( which triggers a switch from the symplectic integrator to the burlisch - stoer integrator ) is 1 @xmath18 .
we run several thousands of simulations , grouped into ensembles and summarized in table [ tab : ens ] .
the default number of simulations per ensemble is 80 .
our simulations begin with embryos each having an isolation mass @xmath19 : @xmath20 where @xmath21 .
we explore a large range of @xmath22400 g/@xmath23 . by default
, we use an initial spacing @xmath24 , similar to @xcite , and based on the @xcite simulations of embryos formed via the accretion of small bodies .
we list the range of the initial number of embryos per simulation , @xmath25 , and the final number of planets , @xmath26 , in table 1 .
some of our simulations include damping by residual disk gas ; these are detailed in section [ sec : damp ] .
we assign planets constant bulk densities @xmath17 that define their physical cross sections .
when two embryos touch , we assume perfect accretion with no fragmentation . initial eccentricities and inclinations are specified in table [ tab : ens ] .
depending on the ensemble , the magnitude of the initial eccentricity @xmath27 is set either to 0 ; to a constant fraction ( specified in the table ) of @xmath28^{1/3}}$ ] , where the subscripts `` 1 '' and `` 2 '' refer to adjacent inner and outer bodies ; or to a constant fraction of the orbital separation @xmath29}$ ] .
flat ensembles ( ef , eci0 ) are strictly 2d , beginning with and maintaining @xmath30 .
for other ensembles , the magnitude of the initial inclination @xmath31 is drawn either from a uniform distribution between 00.1@xmath32 or 00.001@xmath32 ; set to a constant value @xmath33 ( equipartition ) ; or set to @xmath34 . the initial mean anomaly ,
argument of periapse , and longitude of ascending node are drawn randomly from a uniform distribution spanning 02@xmath35 . in the remainder of this work , we compute the reported @xmath36 of each planet relative to the initial @xmath30 plane . following @xcite , the inner edge of the disk of embryos is drawn randomly from a uniform distribution spanning 0.040.06 au .
other embryos are spaced @xmath37 away from each other , up to a maximum @xmath6 of 1 au .
embryos are initially placed in order of increasing @xmath6 .
the semi - major axis @xmath38 and mass @xmath39 of the @xmath40th embryo are computed from the mutual hill radius of the previous embryo , i.e. , @xmath41 . as a result , the actual initial orbital spacings differ slightly from the values of @xmath37 listed in table 1 .
the surface density normalization @xmath42 for each simulation in the ensemble is drawn from a uniform log distribution spanning the range specified in table 1 . the surface density slope @xmath43 except for ensemble eh@xmath44 - 2 , which uses @xmath45 , as noted in table 1 . [
cols="<,^,^,^,^,^,^,^,^,^ " , ] inclinations drawn randomly from a uniform distribution between 0 and @xmath31 .
contains 500 simulations .
the default is 80 .
@xmath46^{1/3}$ ] is the hill parameter .
surface density power - law slope is @xmath47 instead of the default @xmath43 .
the planet bulk density @xmath17 is drawn from a uniform log distribution from 0.0214 g @xmath48 .
the default is a fixed @xmath49 g @xmath48 .
@xmath4 is the semi - major axis difference between neighboring embryos .
includes gas damping .
see section 5 .
during the giant impact stage of planet formation , planets scatter and merge , establishing the system s orbital spacings and eccentricities . through mutual gravitational interactions ,
planets convert keplerian shear into random velocity , leading to growth in @xmath50 and @xmath36 .
mergers ( inelastic collisions ) counter the growth of random velocities and stabilize the system by widening the spacings between bodies .
figure [ fig : avgevo ] ( row 1 ) compares the average eccentricity growth from two ensembles of simulations ( e , ef ) , two individual members of which are shown in figure [ fig : evo ] .
planets begin as moon - to - mars mass embryos on circular orbits and grow in mass , separation , and eccentricity .
[ fig : avgevo ] shows four stages for eccentricity growth : the initial growth at low @xmath50 ( stage 1 ) , a faster growth as @xmath50 approaches orbit crossing ( stage 2 ) , growth during orbit crossing ( stage 3 ) , and an eccentricity equilibrium ( stage 4 ) . for the non - flat ensemble ,
the average inclination is also plotted in figure [ fig : avgevo ] .
the second row of figure [ fig : avgevo ] shows the evolution of @xmath12 , the spacing in mutual hill radii , which is flat throughout the first two eccentricity growth stages when mergers do not yet occur .
in this section we use order - of - magnitude calculations to understand how eccentricities , inclinations , and orbital spacings are connected .
we focus exclusively on the latter phases of the evolution stages 3 and 4 when mergers occur and spacings @xmath12 evolve from their initial values .
inclinations matter for orbital spacings .
top : evolution of @xmath50 ( thick curves ) and @xmath36 ( thin curve ) , averaged over all planets in an ensemble .
averages from the non - flat ensemble e are plotted as red dashed curves and from flat ensemble ef as a black solid curve ; @xmath36 does not appear on the plot for ef because that ensemble is completely 2d @xmath51 . for both ensembles , the eccentricity undergoes four stages of evolution , as demarcated by vertical dotted lines .
middle : same for hill spacings @xmath12 for adjacent pairs of planets .
planet pairs in the flat ensemble ( black solid ) end up with tighter spacings . bottom : distributions of final @xmath12 .
, title="fig : " ] inclinations matter for orbital spacings .
top : evolution of @xmath50 ( thick curves ) and @xmath36 ( thin curve ) , averaged over all planets in an ensemble .
averages from the non - flat ensemble e are plotted as red dashed curves and from flat ensemble ef as a black solid curve ; @xmath36 does not appear on the plot for ef because that ensemble is completely 2d @xmath51 . for both ensembles , the eccentricity undergoes four stages of evolution , as demarcated by vertical dotted lines .
middle : same for hill spacings @xmath12 for adjacent pairs of planets .
planet pairs in the flat ensemble ( black solid ) end up with tighter spacings .
bottom : distributions of final @xmath12 .
, title="fig : " ] inclinations matter for orbital spacings .
top : evolution of @xmath50 ( thick curves ) and @xmath36 ( thin curve ) , averaged over all planets in an ensemble .
averages from the non - flat ensemble e are plotted as red dashed curves and from flat ensemble ef as a black solid curve ; @xmath36 does not appear on the plot for ef because that ensemble is completely 2d @xmath51 . for both ensembles , the eccentricity undergoes four stages of evolution , as demarcated by vertical dotted lines .
middle : same for hill spacings @xmath12 for adjacent pairs of planets .
planet pairs in the flat ensemble ( black solid ) end up with tighter spacings . bottom : distributions of final @xmath12 .
, title="fig : " ] we derive an order - of - magnitude formula for eccentricity growth for use in later subsections when we consider the specifics of stages 3 and 4 .
consider widely - spaced , shear - dominated pairs for which the relative velocities are @xmath52 where @xmath53 is the mutual hill velocity .
the relative velocity changes impulsively every conjunction .
an encounter between two planets each of mass @xmath7 at impact parameter @xmath54 produces an acceleration @xmath55 over a time interval @xmath56 , changing @xmath57 by @xmath58 it is assumed that @xmath59 is randomly directed ; over many encounters , @xmath57 random walks . for a given @xmath54 ,
the expected number of synodic periods @xmath60 ( i.e. , the number of conjunctions or random walk steps ) required to change @xmath57 by a fractional amount @xmath61 is given by @xmath62 eqn .
[ eqn : nsyn ] gives a rate of viscous stirring , which we now average over the geometry of possible impact parameters @xmath54 from @xmath63 to @xmath64 .
the probability of an encounter with impact parameter @xmath54 depends on the value of @xmath54 relative to the `` scale height '' @xmath65 , as depicted in figure [ fig : icartoon ] .
the scale height for a pair of bodies reflects the range of mutual inclinations resulting from their inclinations @xmath66 with respect to the initial plane and their range of nodal orientations .
let @xmath67 be the range of horizontal separations , determined by @xmath12 and the eccentricity vectors . for @xmath68 ,
the dynamics is essentially 2d ( i.e. , in a common orbital plane ) and independent of the inclinations .
if orbits are crossing ( figure [ fig : icartoon ] , right panel ) , the probability of a conjunction occurring within the interval @xmath54 and @xmath69 for @xmath68 is @xmath70 for non - crossing orbits ( figure [ fig : icartoon ] , left panel ) , this 2d probability is reduced by a factor of 2 . for @xmath71 ( the 3d case )
, the corresponding probability is @xmath72 reduced by a factor of @xmath73 relative to the 2d case and dependent on @xmath36 .
weighting eqn .
[ eqn : nsyn ] by these probabilities and integrating from @xmath63 to @xmath64 , we find @xmath74 } \frac{\pi dx}{ia x}\nonumber\\ & & + \int^{{x_{\rm max}}}_{\max [ { r_{\rm graze}},{x_{\rm min}},\min(ia,{x_{\rm max } } ) ] } \frac{2 dx}{x^2 } \bigg ) \nonumber\\\ & = & \frac{18 r_{\rm h}^2}{f^2 d\delta^4 } \nonumber \\ & & \times \bigg ( \frac{\pi}{2 ia } \ln\left [ \frac{\max \left[{r_{\rm graze}},{x_{\rm min}},\min(ia,{x_{\rm max}})\right]}{\max ( { r_{\rm graze}},{x_{\rm min}})}\right ] \nonumber \\ & & + \frac{1}{\max \left[{r_{\rm graze}},{x_{\rm min}},\min(ia,{x_{\rm max}})\right]}-\frac{1}{{x_{\rm max } } } \bigg ) , \nonumber\\\end{aligned}\ ] ] where the impact parameter for two bodies of radius @xmath75 to undergo a grazing collision is @xmath76^{1/2 } \nonumber \\ & = & \frac{12 r_{\rm h}}{\delta } \left(\frac{v_{\rm h}}{2 { v_{\rm esc}}}\right ) \left[1 + 4\left(\frac{\delta}{2 { v_{\rm esc}}/v_{\rm h}}\right)^2\right]^{1/2}. \nonumber \\\end{aligned}\ ] ] the @xmath77 lower limit of the first integral applies because when @xmath78 at conjunction , a merger occurs instead of a scattering . the @xmath79 $ ] upper limit of the first integral ensures that only the encounters with @xmath71 are taken into account . for @xmath80 , the first integral vanishes .
conversely , the second integral , which takes into account encounters with @xmath68 , vanishes if @xmath81 .
[ eqn : pstir ] applies to the growth of random velocity i.e. , both @xmath50 and @xmath36 but in the remainder of this section we will apply it to eccentricity growth .
most encounter geometries either predominantly excite @xmath50 ( @xmath82 ) or excite both @xmath50 and @xmath36 by comparable amounts ( @xmath83 ) .
in stage 3 , orbits cross and bodies merge ( fig .
1 : @xmath12 is increasing in row 2 during this stage ) .
eccentricity excitation by scatterings is tempered by eccentricity damping by mergers .
eccentricity damping follows from conservation of momentum : an inelastic head - on collision between two equal - mass bodies with relative velocity @xmath84 produces a single body with velocity @xmath85 . for @xmath86 ,
this argument yields @xmath87 .
empirically , we determine from our simulations ( sections 2 and 4 ) that @xmath88 is a more accurate rule of thumb .
( see @xcite for a detailed study of eccentricity damping via giant impacts . )
the takeaway point is that each merger lowers the eccentricity of the colliding pair by a factor of order unity .
thus the number of conjunctions ( i.e. , synodic periods ) required to reduce the median eccentricity by order unity is the inverse of the probability of a merger per conjunction , i.e. , the inverse of the probability that @xmath89 in a given encounter : @xmath90 the two cases correspond to 3d and 2d encounters , respectively .
we compare the above damping rate to the stirring rate ( eqn .
[ eqn : pstir ] ) , evaluated for @xmath91 and @xmath92 : @xmath93 + \frac{1}{\max \left(ia,{r_{\rm graze}}\right)}-\frac{1}{{x_{\rm max } } } \bigg \ } , \nonumber\\\end{aligned}\ ] ] where we have set @xmath94 because we are interested in order - unity changes to the eccentricity . as @xmath95 ,
the ratio of the stirring rate to the damping rate is @xmath96 & , \,\,ia > { r_{\rm graze}}\nonumber \\ \sim & \frac{36 r_{\rm h}^2 } { \delta^4{r_{\rm graze}}^2 } & , \,\,ia < { r_{\rm graze}}. \nonumber \\\end{aligned}\ ] ] note how @xmath67 conveniently divides out . in stage 3 ,
the ratio given by eqn .
[ eqn : equal ] is greater than unity : stirring exceeds damping for small @xmath12 , and eccentricities and inclinations rise . at a given @xmath12 , the stirring - to - damping ratio is larger by a logarithm for @xmath97 than for @xmath98 .
this is consistent with fig . 1 ( row 1 ) , which shows that eccentricity growth is somewhat faster in non - flat systems as compared to flat systems during stage 3 .
an eccentricity equilibrium can be achieved when the rate of eccentricity growth from scatterings matches the rate of eccentricity damping from mergers .
this eccentricity equilibrium is evident in stage 4 of fig . 1 ( row 1 ) .
it is distinct from the spacing equilibrium discussed by @xcite , whereby orbital repulsion between big bodies ( driven by scattering and dynamical friction ) keeps pace with the expansion of their hill radii caused by accretion of small bodies .
setting the stirring - to - damping ratio .
we do not treat the case @xmath81 . ] in eqn .
[ eqn : equal ] to unity gives the value of @xmath12 necessary to achieve an eccentricity equilibrium : @xmath99^{1/2 } = \nonumber\\ \frac{1}{2 } \sqrt { \ln\left ( \frac{ia}{{r_{\rm graze}}}\right ) \nonumber + \frac{2}{\pi } } , \,\,\,\,ia > { r_{\rm graze}}\nonumber \\ \frac{1}{2 } , \,\,\,\,ia < { r_{\rm graze}}\nonumber \\\end{aligned}\ ] ] or @xmath100 eqn .
[ eqn : deltaover ] can be re - cast as : @xmath101 at the risk of over - interpreting our order - of - magnitude calculations , we infer from eqns .
[ eqn : deltaover ] and [ eqn : delta ] that : ( a ) the spacing that gives an eccentricity equilibrium ( hereafter the `` equilibrium spacing '' ) does not depend explicitly on planet mass , embryo surface density , or ( for a given orbital period ) stellar mass ; ( b ) the equilibrium spacing does depend on planet bulk density because lower @xmath17 corresponds , at fixed mass , to larger @xmath75 and therefore larger @xmath102 ( eqn .
[ eqn : rgraze ] ) ; larger merger rates must be balanced by larger stirring rates which are obtained for smaller @xmath12 ( compare eqns .
[ eqn : pmerge ] and [ eqn : pstir3 ] ) ; ( c ) for practically all values of @xmath36 , the equilibrium spacing is less than the standard quoted value of @xmath103 ( section 1 ) , in agreement with our numerical simulations ( section 4 ) and observations ( e.g. , @xcite ) . of especial note is ( d ) higher @xmath36 s lead to wider spacings . while stirring and merger rates are each reduced by increasing @xmath36 because the volume of space in which bodies interact @xmath104 the stirring rate is reduced less severely , because encounters occurring at the largest impact parameters , at @xmath68 , unfold independently of @xmath36 .
for example , for @xmath105 , eqn .
[ eqn : delta ] yields a spacing about 50% wider than in a flat ( @xmath106 ) system .
our considerations of orbital spacings are qualitatively similar to those of @xcite .
both of our studies recognize that increasing mutual inclinations in packed multi - planet systems demands larger orbital spacings to maintain dynamical stability ( cf .
our eqn .
[ eqn : delta ] with their equation 14 ) . since
their study is numerical , it accounts for effects not captured by the crude arguments we have made in this section [ sec : analytic ] . in particular , our treatment above assumes that each conjunction between neighboring planets is a step in a random walk ; this assumption may fail at large orbital spacings where the dynamics is less chaotic .
hopefully this shortcoming does not compromise our goal to understand , if only qualitatively , how orbital spacings are determined during the formative stages of planetary systems , when spacings are comparatively small .
we offer eqn .
[ eqn : delta ] as a plausibility argument that can explain some of the results of our numerical simulations .
these simulations are the focus for the remainder of our paper .
here we use simulations to explore how orbital spacings depend on initial conditions in the purely @xmath0-body regime , i.e. , after damping from residual gas / planetesimals has ceased . in this section , we assign a range of ad hoc initial conditions . in section 5 , we will explore how gas damping might establish such initial conditions . in section 3.5 , we demonstrated with scaling arguments that the spacing necessary to achieve an eccentricity equilibrium via stirring and mergers depends on the planets inclinations .
we found that the final spacings of flat systems @xmath51 are tighter than those of systems where @xmath36 is free to grow ( fig . 1 ) .
here we compare several additional ensembles that all have @xmath36 free to grow but have different @xmath36 s during the orbit crossing stage , when @xmath36 starts to affect the stirring rate ( i.e. , eccentricity growth stage 3 , section 3.4 ) .
each ensemble of simulations begins with planets with eccentricities large enough to cross : @xmath107 where @xmath108 is the difference in semi - major axes between adjacent bodies .
the four ensembles differ in the magnitudes of their initial inclinations : @xmath109 ( eci0 , black curve in fig .
[ fig : icross ] ) ; @xmath110 ( eci0.001 , purple ) , @xmath111 ( eci0.1 , blue ) , and @xmath112 ( ec , red ) .
inclinations when orbits start to cross affect final spacings .
each of the ensembles ( red , blue , purple , black ) begins with an identical distribution of initial conditions excepting inclinations .
top : evolution of @xmath50 ( solid ) and @xmath36 ( dotted ) , averaged over all planets in an ensemble .
the eccentricity evolution is similar across ensembles except the @xmath30 ensemble ( black ) .
bottom : the smaller the initial @xmath36 , the tighter the final spacing .
each simulation begins with @xmath24 at @xmath113 .
, title="fig : " ] inclinations when orbits start to cross affect final spacings .
each of the ensembles ( red , blue , purple , black ) begins with an identical distribution of initial conditions excepting inclinations .
top : evolution of @xmath50 ( solid ) and @xmath36 ( dotted ) , averaged over all planets in an ensemble .
the eccentricity evolution is similar across ensembles except the @xmath30 ensemble ( black ) .
bottom : the smaller the initial @xmath36 , the tighter the final spacing .
each simulation begins with @xmath24 at @xmath113 . ,
title="fig : " ] we plot the evolution of @xmath50 , @xmath36 , and @xmath12 in fig . [ fig : icross ] . in all four ensembles ,
the slope of @xmath50 vs. time ( top panel ) is relatively shallow because stirring and merging tend to counter - balance each other .
it is evident from the bottom panel that mergers begin immediately .
excepting the @xmath30 case , the inclination grows and reaches equipartition with eccentricity within a fraction of a myr . as a result of this fast growth ,
the non - flat ensembles exhibit only modest differences in their final @xmath36 s and @xmath12 s . as expected from section 3.5 , the flatter the system , the tighter the final spacing .
the median final spacings are 20 , 26 , 27 , and 29 from ensembles eci0 , eci0.001 , eci0.1 , and ec , respectively .
next we explore the effects of the initial eccentricity on the final spacing .
although @xmath50 does not enter directly into the spacing equilibrium derived in section 3.5 , it can have several indirect effects that we discuss below .
for ensembles ec , ece0.3 , ece0.1 , and e , respectively , @xmath114 and @xmath115 ; these range from initially crossing orbits ( ec ) to initially circular orbits ( e ) . for case
e , the initial inclination magnitude is drawn from a uniform distribution from 00.1@xmath32 , while for the other cases @xmath116 rad .
all have the same initial spacings @xmath37 .
we plot the evolution of @xmath12 , @xmath50 , and @xmath36 in fig .
[ fig : cross ] , left panel .
the black curve ( @xmath117 , e ) represents the same ensemble plotted as the red - dashed curve in fig . 1 .
the purple curve ( @xmath118 , ece0.1 ) begins with the eccentricity growth stalled , even though mergers are not yet occurring ( bottom left panel ) .
it appears in individual simulations that the initial stalling occurs because we assigned random phases to the eccentricity vectors ; orbits need to precess so that their apsidal longitude differences @xmath119 before encounters become close and stirring begins in earnest .
the blue ( @xmath120 , ece0.3 ) and red ( @xmath121 , ec ) curves begin with nearly crossing or crossing orbits ; as seen for the crossing orbits in fig .
4 , the counteracting contributions of mergers and stirring keep the eccentricity evolution slow .
initial eccentricities affect final spacings : the median final spacings for the four ensembles ( ec , ece0.3 , ece0.1 , e ) are 29 , 27 , 26 , and 25 , respectively .
although the eccentricity does not explicitly factor into the spacing equilibrium in section 3.5 , we hypothesize that it has the following indirect effects .
the first is that our initial conditions assumed equipartition between @xmath50 and @xmath36 ( @xmath116 ) .
the larger initial @xmath27 leads to a larger initial @xmath31 which propagates to a larger @xmath36 during orbit crossing , increasing the spacing required for an eccentricity equilibrium ( sections 3.5 , 4.1 ) .
the ensembles that end up with wider spacings ( ec , ece0.3 , red and blue , fig . 5 , bottom left panel ) also have larger inclinations ( top left panel ) .
to test this idea further , we ran three additional ensembles of simulations ( eci0.1 , ece0.3i0.1 , ece0.3i0.1 ) that we plot along with ensemble e in the right panels of fig .
[ fig : cross ] . for these auxiliary runs , instead of
beginning with @xmath122 , we assign inclination magnitudes drawn from a uniform distribution from 00.1@xmath32 . whereas the original ensembles ec , ece0.3 , ece0.1 , and e ( each with different @xmath31 ) converged to different final eccentricities , inclinations , and spacings ( fig .
5 , left panel ) , the extra ensembles eci0.1 , ece0.3i0.1 , ece0.1i0.1 , and e converge to nearly identical final eccentricities and inclinations ( right panel ) .
the final spacing of ensemble ece0.3i0.1 ( 26 ) differs only slightly from those of ece0.1i0.1 and e ( 25 ) , despite their different initial eccentricities .
however , the ensemble with the largest initial eccentricities ( eci0.1 ) ends up with larger final eccentricities , inclinations , and spacings . therefore initial inclinations and equipartition are not the whole story .
two other effects may be contributing to large final spacings for the ensembles that began with large eccentricities ( eci0.1 , ec , @xmath123 , red curves ) .
first , systems that start with eccentricities close to crossing may be prone to mergers that overshoot the spacing value corresponding to eccentricity equilibrium .
overshooting the benchmark spacing value is possible because collisions are non - reversible .
second , we assumed in section 3 that relative velocities are shear - dominated .
but when @xmath124 , epicyclic motions contribute significantly to relative velocities .
increasing the relative velocity reduces @xmath102 , which reduces the merger rate and necessitates a wider spacing to achieve an eccentricity equilibrium .
last we explore the dependence of the final spacing on the initial hill spacing @xmath37 .
we compare seven ensembles of simulations : \{e@xmath123 , e@xmath125 , [email protected] , e@xmath129 , e , [email protected] , e@xmath1213 } , which have initial @xmath125 , respectively .
all have @xmath117 and @xmath31 drawn from a uniform distribution from 0@xmath126 .
we plot the evolution of @xmath50 , @xmath36 , and @xmath12 in fig .
[ fig : rh ] .
the final median spacings of the ensembles are 31 , 31 , 29 , 27 , 25 , 23 , and 17 , respectively .
the systems that are initially spaced more tightly end up with wider spacings .
( solid ) and @xmath36 ( dotted ) , averaged over all planets in an ensemble .
bottom : the tighter the initial spacing , the wider the final spacing . initial spacings differ slightly from those listed in the legend because of the way we lay down our embryos ; see the last paragraph of section 2 .
[ fig : rh ] , title="fig : " ] ( solid ) and @xmath36 ( dotted ) , averaged over all planets in an ensemble .
bottom : the tighter the initial spacing , the wider the final spacing .
initial spacings differ slightly from those listed in the legend because of the way we lay down our embryos ; see the last paragraph of section 2 .
[ fig : rh ] , title="fig : " ] our interpretation is that the initially tighter spacings allow eccentricities to enter the orbit crossing regime sooner ( i.e. , at a tighter spacing ) , further exciting eccentricities and inclinations ( see fig . 1 ) . during the subsequent evolution , at a given @xmath12 , systems with initially smaller @xmath37
have larger @xmath36 and @xmath50 than their counterparts that began with larger @xmath37 . as explored in sections 4.1 and 4.2 , larger @xmath50 and @xmath36 drive systems toward wider final spacings .
the equilibrium values of @xmath50 and @xmath36 achieved also seem to remember the initial spacing ( fig .
[ fig : rh ] , row 1 ) .
the final equilibrium @xmath50 and @xmath36 stratify according to initial spacing : tighter initial spacings ( light blue ) result in higher final equilibrium @xmath50 .
the assumed initial spacings of the @xmath0-body simulations presented here are intended to reflect possible endstates of oligarchic merging in the presence of residual gas and planetesimals . in the next section ,
we compute such endstates from scratch using damped @xmath0-body simulations .
before disk gas dissipates , it can damp planet eccentricities and inclinations . the competition between gas damping and mutual viscous stirring by protoplanets sets primordial orbital spacings , eccentricities , and inclinations , which in turn determine the subsequent gas - free evolution .
we present here the results of @xmath0-body simulations that include the effects of gas damping . the strength and persistence of gas damping depends on how the protoplanetary nebula clears
. observations of young stellar clusters show that transition disks disks with central clearings comprise @xmath110% of the total disk population @xcite .
we adopt the interpretation that all disks undergo a transitional phase that lasts @xmath127 10% of the total @xmath1520 myr disk lifetime ( @xcite ; see also @xcite , @xcite , and @xcite for their `` photoevaporation - starved '' model of disk evolution which reproduces the short final clearing timescale ) . for simplicity we model the gas disk as having a fixed density that persists for 1 myr and drops to zero thereafter .
the step function could represent the final myr of the slow phase ( prior to the formation of a central clearing ) followed by the fast transitional phase in which gas is assumed to vanish completely inside 1 au .
alternatively the step function could represent the act of clearing that takes place entirely during the fast transitional phase .
we implement gas damping in a customized version of mercury6 that incorporates user - defined forces as described in appendix a of @xcite , adding several corrections and modifications that allow us to use the hybrid symplectic integrator , which we benchmarked against the burlisch - stoer integrator .
we impose @xmath128 and @xmath129 @xcite .
following @xcite , we make use of three damping timescales @xmath130 , based on the regimes identified in @xcite , @xcite , @xcite , and @xcite : @xmath131 where @xmath132 is one solar mass , @xmath7 is the planet mass , @xmath133 is the random ( epicyclic ) velocity , @xmath134 with @xmath15 equal to the planet s mean motion , @xmath135 is the gas sound speed , and @xmath136 is a constant proportional to the degree of nebular gas depletion ( @xmath137 corresponds approximately to the minimum - mass solar nebula with gas surface density @xmath138 g/@xmath23 at 1 au , and @xmath139 corresponds to more depleted nebulae ) .
our simulations bear some resemblance to those of @xcite , who also include gas damping of planetary random velocities .
our study differs in having a damping timescale that varies with epicyclic velocity ( eqn .
[ eqn : taudamp ] ) ; solid surface densities @xmath42 that are up to an order of magnitude larger than theirs ( so as to reproduce _
kepler _ super - earths ) ; a gas disk evolution that behaves as a step function instead of their decaying exponential ; and hundreds more simulations that enable us to make statements with greater statistical confidence .
our aim is also broader in that we seek to understand how spacings , inclinations , and eccentricities inter - relate ; and later in section 6 we will bring planetary composition into this mix as we distinguish between purely rocky and gas - enveloped planets .
where we overlap with @xcite , we agree ; in disks of low @xmath42 , producing predominantly rocky @xmath140 planets , the final spacings @xmath12 tend to be large , exceeding @xmath120 ( compare our figure 8 with the results described in the text of their section 4 ) .
ensembles ed10@xmath141 , ed10@xmath142 , ed10@xmath143 , ed10@xmath144 , and ed10@xmath145 correspond to @xmath146 and @xmath147 , respectively .
the simulations all have initial @xmath148 , @xmath149 , @xmath150 , and @xmath151 g/@xmath2 .
the initial eccentricities are close to orbit crossing .
the small initial inclinations are assumed to have been damped by the undepleted nebula prior to the start of the simulations .
our choice of @xmath152 gives inclinations similar to those for @xmath153 ; the system is assumed to be initially nearly flat .
each ensemble comprises 80 simulations with solid surface densities @xmath42 spanning 38105 g/@xmath23 .
further details are listed in table [ tab : ens ] .
after 1 myr , we shut off gas damping , and integrate for an additional 27 myr .
the physical interpretation of the scenario simulated here is as follows .
before the simulation begins , a high gas surface density has kept the embryos separated by @xmath154
. then the gas density drops to the simulated depletion @xmath136 , allowing the embryos to scatter and merge in the presence of gas damping for 1 myr .
then the gas vanishes entirely and the system evolves for 27 myr without gas .
the left panels of fig .
[ fig : damp ] show the evolution of @xmath36 , @xmath50 , and @xmath12 in the presence of damping .
gas damping flattens and circularizes systems to low @xmath36 and @xmath50 on timescales that increase with @xmath136 ( top left ) . after 1 myr of evolution with gas damping , the smallest @xmath136 ( strongest damping , dark blue ) ensembles have the tightest spacings , smallest inclinations , and smallest eccentricities ( the @xmath137 ensemble is a bit of an outlier in eccentricity ) . the subsequent undamped
, gas - free evolution is shown in the right panels . in the more strongly damped ed@xmath155 , ed@xmath156 , and ed@xmath157 ensembles ,
the eccentricity and inclination ( fig .
[ fig : damp ] , top left ) initially grow .
the initial eccentricities are close to orbit crossing ( @xmath158/3 ) and so embryos begin to merge almost immediately ( bottom left ) . as the spacing increases , eccentricity and inclination growth rates decrease and
eventually @xmath50 and @xmath36 are damped .
inclinations are damped sooner and more steeply .
eccentricities are damped later and appear to plateau to fixed points .
when damping ends , the ensembles with the shortest damping timescales have the tightest spacings , but eventually their fortunes will reverse . as documented in section 4.3 , the tightest initial spacings yield the widest final spacings in the gas - free stage . among the ed@xmath155 , ed@xmath156 , and ed@xmath157 ensembles ,
all of which emerge from the damping stage with similarly small eccentricities , the ed@xmath155 ensemble ends up most widely spaced in the subsequent gas - free evolution , while ed@xmath157 ends up most tightly spaced .
we highlight the ed@xmath157 ensemble as producing typical final spacings of @xmath9 ( fig . 7 , bottom right panel ) in agreement with the observations ( e.g. , @xcite ) .
we will compare to the observations in greater detail in section 7 .
the less - damped ensembles ed@xmath159 and ed@xmath147 reach the widest spacings . at the end of gas damping , they begin ( right panels ) on initially nearly crossing orbits , like those explored in section 4.2 . in their subsequent evolution
, they reach a spacing of @xmath160 .
we also see the effects of the inclinations during orbit crossing , as described in section 4.1 . during the gas - free evolution
, the ed@xmath156 ensemble ends up with an average final eccentricity similar to those of the ed@xmath159 and ed@xmath147 ensembles , but the inclinations of ed@xmath156 when orbits begin to cross ( @xmath161 myr in the right panel ) are low .
the lower inclinations likely contribute to the narrower final spacing of ed@xmath156 compared to ed@xmath159 and ed@xmath147 .
we continued integrations of three ensembles ( ed@xmath156 , ed@xmath157 , ed@xmath159 ) up to a total time of 300 myr ( corresponding to four months of wall - clock time ) and found that their outcomes hardly changed from those shown in fig .
the eccentricities @xmath162 of these three ensembles remained approximately constant , while their median @xmath12 values grew by about 10% .
we estimate that the median @xmath12 may grow by another @xmath11015% if the integrations were extended to 10 gyr .
the ed@xmath157 ensemble retained its much tighter spacings and smaller eccentricities and inclinations as compared to the other two ensembles .
although the ensembles retained their qualitative character when integrated for longer , the quantitative changes , in particular to the spacings , can be significant when attempting to fit to the observations ( cf .
longer integrations of all ensembles could be run in a follow - up study .
here we demonstrate several ways in which the giant impact stage explored in sections 36 connects planets bulk densities to their orbital properties .
more rarefied planets are characterized by tighter orbital spacings , lower eccentricities , and lower mutual inclinations .
our results motivate a search for observational correlations between super - earths orbital and compositional properties , which statistical studies usually assume to be independent of each other ( e.g. , @xcite ) .
planets with lower bulk densities and planets with tighter spacings are each a consequence of higher solid surface density disks . in a higher solid surface density disk ,
a core of a given mass can form from mergers of embryos contained in a narrower zone of smaller @xmath12 .
figure [ fig : sig ] demonstrates the anti - correlation between the final spacing of super - earths and the solid surface density normalization .
furthermore , when cores grow from narrower feeding zones , they undergo mergers quickly ( see also the empirical study of orbit crossing timescales as a function of @xmath12 by @xcite ) , reaching a mass large enough to acquire gas envelopes before the gas disk dissipates . on surface density normalization @xmath42 from ensemble eh .
red points are planets with @xmath164 , and black points fall outside this mass range .
the spacing @xmath12 tends to decrease with @xmath42 , particularly for a restricted range of masses ( e.g. , the red points ) .
the trend is not sensitive to the initial conditions of the ensemble .
our result that @xmath165 for low @xmath42 agrees with @xcite .
[ fig : sig ] ] following @xcite , we apply the @xcite nebular accretion models to track the gas fractions of planets as they grow through mergers in the ed@xmath156 and ed@xmath157 ensembles , labeling planets that acquire a 1% atmosphere in less than 1 myr as `` gas - enveloped , '' and planets that do not as `` rocky . ''
we compute the final spacing between each pair of adjacent planets and plot the resulting spacing distributions in fig .
[ fig : deltadensesim ] .
gas - enveloped planets end up more tightly spaced than rocky planets . to test whether these differences are statistically significant , we apply a two - sample kolmogorov - smirnov ( k - s ) test to the gas - enveloped vs. rocky distributions of @xmath12 from ed@xmath157 ( fig .
[ fig : deltadensesim ] , left panel ) , obtaining a @xmath166-value of @xmath167 . for the distributions from ed@xmath156 ( right panel ) ,
we obtain @xmath168 .
these @xmath166-values are sufficiently small that for both the ed@xmath157 and ed@xmath156 ensembles , we reject the null hypothesis that the gas - enveloped and rocky planet spacings are drawn from the same underlying distribution .
for simulated rocky planets ( thin ) and gas - enveloped ( thick ) from ed@xmath157 ( left ) and ed@xmath156 ( right ) .
planets included have masses ranging from 230 @xmath169 .
planet pairs are labeled gas - enveloped when both members of the pair have accreted 1%-by - mass atmospheres within 1 myr ; all other pairs are labeled rocky .
the gas - enveloped planet pairs have tighter spacings .
[ fig : deltadensesim ] ] in the ed@xmath157 ensemble , about half of pairs with masses @xmath164 consist of two gas - enveloped planets .
these pairs have lower median eccentricities ( @xmath170 ) and inclinations ( @xmath171 ) than pairs with one or more rocky planets ( @xmath172 ) .
we apply a two - sample k - s test to the gas - enveloped vs. rocky distribution of @xmath50 ( @xmath36 ) and obtain a @xmath166-value of @xmath173 .
we therefore reject the null hypothesis that the gas - enveloped and rocky planet eccentricities ( inclinations ) are drawn from the same @xmath50 ( @xmath36 ) distribution .
we showed in section 5 that a depletion factor @xmath174 , corresponding to @xmath175 g @xmath176 , produced the tightest final spacings.@xmath157 relative to the mmsn corresponds to a factor of @xmath1@xmath159 depletion relative to the minimum - mass extrasolar nebula constructed by @xcite . ]
the planets in the ed10@xmath143 ensemble undergo most of their growth during the gas damping stage , as can be seen in fig .
[ fig : damp ] : @xmath12 grows most during the gas damping stage ( left panel ) rather than the gas - free stage ( right panel ) . for the range of solid surface densities explored in this ensemble ( 38105 g @xmath176 , table 1 )
, the gas - to - solid ratio is @xmath10.20.4 , so there is sufficient disk gas locally for planets to acquire low - mass ( 1% ) atmospheres without having to appeal to gas accreting inward from outside 1 au ( cf . * ? ? ? * ) .
the simulations with depletion factors that yield wider spacings ( @xmath177 ) also yield higher density planets ( as illustrated for @xmath137 compared to @xmath178 in fig .
[ fig : bothcomp ] ) .
on the one hand , when the depletion factor is small ( @xmath179 ) , the planets are largely assembled after the gas disk completely dissipates ( fig .
[ fig : damp ] , bottom right panel ) : they end up with lower gas fractions and higher densities . on the other hand , when the depletion factor is much higher ( gas - to - solid ratios of @xmath10.020.04 when @xmath180 or @xmath10.0020.004 when @xmath181 ) , then there is insufficient gas during the growth stage to create significant atmospheres ( assuming gas in the inner disk where cores reside is not replenished from the outer disk ; see @xcite who relax this assumption ) .
the case @xmath178 is a happy middle ground for making low - density planets . for simulated rocky pairs ( no shading ) , gas - enveloped pairs ( dark shading ) , and pairs of one rocky and one gas - enveloped planet ( light shading ) from the ed@xmath157 ( left ) and ed@xmath155 ( right ) ensembles .
planets included have masses ranging from 230 @xmath169 .
the ed@xmath157 ensemble has a much greater fraction of gas - enveloped planet pairs than the ed@xmath155 ensemble .
[ fig : bothcomp ] ] median eccentricities and inclinations in the most tightly spaced ed10@xmath143 ensemble tend to be low : @xmath182 and @xmath183 for @xmath184 .
in contrast , @xmath185 and @xmath186 for ensembles \{ed@xmath155 , ed@xmath156 , ed@xmath159 , ed@xmath147 } , respectively . figure [ fig : deltadamp ] compares the final spacing , eccentricity , and inclination distributions for three ensembles ( ed10@xmath141 , ed10@xmath143 , and ed10@xmath145 ) .
we apply two - sample k - s tests to test whether the differences between these three ensembles are statistically significant .
we reject the null hypotheses that the @xmath12 , @xmath50 , and @xmath36 values for ensemble ed10@xmath143 are drawn from the same distributions as for ed10@xmath145 or ed10@xmath141 ; the @xmath166-values are less than @xmath187 . in contrast
, we can not reject the null hypotheses that ed10@xmath141 and ed10@xmath145 are drawn from the same distribution with respect to @xmath12 ( @xmath188 ) and @xmath50 ( @xmath189 ) . for @xmath36 , we reject the null hypothesis that ed10@xmath141 and ed10@xmath145 are drawn from the same distribution ( @xmath190 ) . from the order - of - magnitude arguments in section 3 , we expect the spacing to scale approximately with planet bulk density as @xmath191 ( equation [ eqn : delta ] ) .
the dependence arises because the planet s size determines its collisional cross - section , which in turn affects the balance between mergers and scatterings and therefore the spacing required for an eccentricity equilibrium .
although the scaling of spacing with bulk density is weak , _
kepler _ super - earths span more than a factor of 10 in bulk density ( e.g. , @xcite ) .
we perform a suite of 80 simulations ( eh@xmath17 ) spanning a range of @xmath17 from 0.02 g/@xmath2 to 14 g/@xmath2 ( these are more extreme than those observed ) and plot the dependence of @xmath12 on @xmath17 in fig . [
fig : dense ] . overplotted in red is a @xmath192 line for comparison with the upper envelope ( 90% ) , median , and lower envelope ( 10% ) of the simulation points ( blue lines ; these are computed by quantile regression using the cobs package in r ; see @xcite ) . the red line is steeper than the upper envelope and median , but nearly matches the slope of the lower envelope for @xmath193 g/@xmath2 .
the upper envelope may reflect systems that overshoot their equilibrium spacings ; since all eh@xmath17 runs start with the same initial eccentricities and spacings , lower @xmath17 systems start closer to their equilibrium spacings and are therefore more prone to overshooting .
we conclude that smaller @xmath17 indeed produces tighter spacings .
planets with larger @xmath17 also have larger median @xmath50 and @xmath36 : @xmath194 and @xmath195 for planets with @xmath196 g @xmath48 vs. @xmath197 and @xmath198 for planets with @xmath199 g @xmath48 . vs. planet bulk density @xmath17 from the eh@xmath17 suite of simulations , where @xmath17 is the assumed constant bulk density that we assign and which defines the collisional cross section .
the red line is the @xmath200 scaling expected from eqn .
[ eqn : delta ] .
the blue lines model the upper envelope ( 90% ) , median , and lower envelope ( 10% ) of the simulation data , computed using quantile regression .
the red line is steeper than the upper envelope and median , but approximates the slope of the lower envelope for @xmath193 g/@xmath2 .
[ fig : dense ] ] the dependence of spacing on collisional cross section motivates a more realistic treatment of the planets compositional and collisional evolution .
here we assumed that the bulk density remains constant as the planets grow from embryos to super - earths , but ideally simulations should account for the change in density as material is compacted and atmospheres are accreted or eroded .
the collisional prescription used here is overly simplistic and does not account for the internal structure of the planet : some or all of the low - density planets from _
kepler _ are thought to be rocky cores with rarefied atmospheres and hydrodynamic simulations are needed to accurately assess the collisional outcomes , including the dependence on the collisional impact parameter . from eqn .
[ eqn : delta ] , we expect tighter spacings at smaller orbital periods . however , when we plot the final @xmath12 vs. semi - major axis in fig .
[ fig : per ] ( using ensemble eh ) , we see the opposite trend : wider spacings at smaller semi - major axes .
we hypothesize that this behavior arises because of the particular solid surface density power law we have chosen . for @xmath43 ( section 2 ) ,
embryo masses increase with @xmath6 , resulting in more massive distant planets .
these distant planets gravitationally stir closer - in planets , necessitating wider spacings to achieve an eccentricity equilibrium . vs. semi - major axis from simulations with our default surface density profile @xmath201 ( black , ensemble eh ) and with a steeper profile @xmath202 ( red , ensemble eh@xmath44 - 2 ) that results in a constant isolation mass with semi - major axis ( eqn .
[ eqn : memb ] ) . for @xmath43 ,
@xmath12 decreases with semi - major axis ; for @xmath47 , @xmath12 is constant with semi - major axis .
bottom : time evolution of average @xmath12 for @xmath43 ( black ) and @xmath47 ( red ) for @xmath203 au ( dashed ) and @xmath204 au ( solid ) .
the evolution differs between the @xmath43 and @xmath47 ensembles in the inner disk ( @xmath203 au ) but not in the outer disk ( @xmath205 au ) .
[ fig : per ] , title="fig : " ] vs. semi - major axis from simulations with our default surface density profile @xmath201 ( black , ensemble eh ) and with a steeper profile @xmath202 ( red , ensemble eh@xmath44 - 2 ) that results in a constant isolation mass with semi - major axis ( eqn .
[ eqn : memb ] ) . for @xmath43 ,
@xmath12 decreases with semi - major axis ; for @xmath47 , @xmath12 is constant with semi - major axis .
bottom : time evolution of average @xmath12 for @xmath43 ( black ) and @xmath47 ( red ) for @xmath203 au ( dashed ) and @xmath204 au ( solid ) .
the evolution differs between the @xmath43 and @xmath47 ensembles in the inner disk ( @xmath203 au ) but not in the outer disk ( @xmath205 au ) .
[ fig : per ] , title="fig : " ] to test this hypothesis , we run an additional ensemble of simulations ( eh@xmath44 - 2 ; table
1 ) with @xmath47 , resulting in an embryo mass that is constant with semi - major axis .
we find spacings similar to those of the eh ensemble at longer orbital periods , but tighter spacings than the eh ensemble at shorter orbital periods . in the bottom panel of fig .
[ fig : per ] , we plot the time evolution of the median @xmath12 . for @xmath204
au , the evolution proceeds similarly regardless of @xmath44 . for @xmath206 au ,
the mergers begin earlier in the @xmath43 ensemble , implying that eccentricities are more quickly excited and orbits cross earlier .
all of these results are consistent with our hypothesis that massive planets in the outer disk can contribute significant `` non - local '' stirring , particularly for shallow surface density profiles .
the results of this subsection are consistent with a recent study by @xcite who simulated the growth of super - earths in disks with @xmath207 .
they found tighter spacings , higher multiplicities , and smaller eccentricities and inclinations for planets formed in steeper @xmath44 disks .
the differences are most marked for the shortest period planets ( see their figs . 3 and 4 ) . in summary , for disks with solid surface density profiles less steep than @xmath47 , planets at shorter orbital periods form denser and more widely spaced . for the @xmath43 profile used throughout this work ,
shorter period planets tend to have smaller core masses ; they are expected to accrete less gas from the nebula and to more easily lose what little envelopes they may have gathered to photoevaporation .
short period planets also tend to have larger median eccentricities and inclinations ( @xmath208 and @xmath209 for @xmath206 au in the eh ensemble ) than do longer period planets ( @xmath210 and @xmath211 for @xmath205 au ) .
we make some basic comparisons to the _ kepler _ sample to assess how the conditions for late - stage planet formation explored in sections 36 manifest in _
kepler _ observables .
we convert our simulated planets to tranets by imposing a rough sensitivity cut of @xmath212 and @xmath213 days . for each simulated system , we generate @xmath147 systems randomly oriented in space with respect to an observer . the subset of planets that transit comprise our tranet sample . we plot spacings from the ed10@xmath143 and ed10@xmath145 ensembles in fig .
[ fig : tranet ] , row 1 , left panel .
compared to fig .
[ fig : deltadamp ] ( intrinsic planet spacings ) , observational selection effects slightly decrease the mode of the distribution of tranet spacings and enhance the tail . to compare our synthetic @xmath12 distribution with that based on observations
, we queried the cumulative _
kepler _ candidates from the nasa exoplanet archive ( march 16 , 2015 ) , restricting our sample to host stars with @xmath214 , @xmath215 , and _ kepler _ magnitude @xmath216 , and selecting for systems in which at least one planet has @xmath217 .
statistical modeling of the potential astrophysical false positive population has revealed that less than 10% of the candidates are false positives @xcite , allowing us to treat the sample as representative of true planets .
we use the @xcite probabilistic mass - radius relation to convert the observed candidate periods to a @xmath12 distribution .
the resulting distribution is not sensitive to the exact mass - radius relation used @xcite .
the observed spacing distribution has a mode @xmath218 , similar to our most tightly spaced distributions ( e.g. , ed10@xmath143 from our ensembles that include a gas damping stage ) , but also has a tail more consistent with our wider - spaced distributions .
following @xcite , we compare several other properties of our simulated tranets to the observed _
kepler _ planets . in the left panels of fig .
[ fig : tranet ] , we compare period ratio ( row 2 ) , number of tranets per system ( row 3 ) , and transit duration ratio ( row 4 ) distributions . the transit duration ratio of a pair of planets ( the inner labeled `` 1 '' and the outer labeled `` 2 '' ) is : @xmath219 where @xmath220 is the transit impact parameter and @xmath221 the argument of periapse . as argued by @xcite , the duration ratio distribution gives insight into the mutual inclinations .
in flatter systems , the inner planet in a pair tends to have a smaller impact parameter , skewing the @xmath222 distribution to values greater than 1 .
both the mutual inclinations and eccentricities affect the width of the @xmath222 distribution , but mutual inclination has a stronger effect because a small change in inclination has a strong effect on @xmath220 , whereas a small change in @xmath50 affects @xmath222 only modestly .
the ensembles that produce the widest spacings ( represented by ed10@xmath145 in fig .
[ fig : tranet ] ) also produce the largest period ratios , the lowest multiplicities , and the widest / least - skewed @xmath222 .
they overproduce single tranets and underproduce high multiplicity systems .
their spacing , period ratio , and @xmath222 distributions match the tails of the observed distributions but not the peak .
the @xcite initial conditions also result in spacings , period ratios , and @xmath222 distributions that are too broad .
in contrast , the ensembles that produce the tightest spacings ( exemplified by the intermediate damping ensemble , ed10@xmath143 , in fig . [ fig : tranet ] ) match the peaks of the observed spacing and period ratio distributions but not the tails .
they overproduce high multiplicity systems and underproduce single tranets .
they produce a @xmath222 distribution that is too narrow and skewed . rather than resembling a single one of our simulated ensembles , the observed distributions appear to comprise a mixture of ensembles .
the simulated distributions from one ensemble can either match the peaks or tails of the observed distributions , but not both . in the right column of fig .
[ fig : tranet ] , we combine the tranets from a tightly spaced , low mutual inclination ensemble ( ed@xmath157 ) with a widely spaced , larger mutual inclination ensemble ( ed@xmath147 ) . for the combined distribution , we use all the ed@xmath157 tranets and randomly draw 50% of the generated ed@xmath147 tranets ( where the 50% weighting is a free parameter that we chose to match the observations ) .
the resulting mixed tranet population reproduces well both the peaks and tails of the observed @xmath12 , period ratio , and @xmath222 distributions , in addition to the relative occurrence rates of single and multiple tranets in the observed multiplicity distribution . what remains to be explained
is the population of planets with the smallest period ratios @xmath111.5 and the tightest spacings @xmath223 .
we speculate that these close neighbors may reflect orbital migration , either of super - earths or their progenitor cores .
is a mixture of two populations truly necessary ?
the ed@xmath147 ensemble is clearly a poor match to the observed distribution , but the improvement of the combined population over ed@xmath157 is less obvious . we perform some statistical tests to help decide this issue .
a two - sample k - s test applied to the distribution of transit duration ratios ( fig .
14 , row 4 ) leads us to reject the null hypothesis that the observed distribution and the ed@xmath157 distribution are drawn from the same distribution ( @xmath166-value 0.0004 ) ; at the same time , we can not reject the null hypothesis that the observed and combined - group period ratios are drawn from the same distribution ( @xmath166-value 0.02 ) . applying the k - s test to the period ratio distributions ( row 2 )
, we reject the null hypothesis that the observed period ratios are drawn from any of the simulated distributions if we consider the full range of period ratios .
but if we restrict our attention to period ratios greater than that of the 3:2 resonance , we find that we can not reject the null hypothesis that the observed and combined - group period ratios are drawn from the same distribution ( @xmath166-value 0.06 ) , while still rejecting the null hypothesis that the observed period ratios and ed@xmath157 period ratios are drawn from the same distribution ( @xmath166-value 0.004 ) .
when we apply the k - s test to the @xmath12 distributions ( row 1 ) , we reject the null hypothesis that the observed spacings are drawn from any of the simulated distributions ( ed@xmath147 , ed@xmath157 , or combined ) even if we make a cut for @xmath224 .
finally , we consider the tranet multiplicity distribution ( row 3 ) , for which a k - s test is inappropriate because the distribution is spread across a small number of integers .
if we assume poisson uncertainties and add them in quadrature , the observed ratios of tranet multiplicities are @xmath225 , and @xmath226 for 1:2 , 2:3 , 3:4 , 4:5 , and 5:6 tranet systems , respectively . in ed@xmath157 , these ratios are 2.5 , 3.6 , 1.7 , 2.6 , and 2.3 ( chi - squared of 27 ) , whereas for the mixture model , these ratios are 3.6 , 4.3 , 2.0 , 2.7 , and 2.3 ( chi - squared of 23 )
. therefore the mixture model represents an improved match to the observations ; it matches the observed distribution well except that it overproduces two tranet systems relative to three tranet systems . in sum
, the mixture model offers quantitative improvements for matching three out of the four observables in fig .
the purpose of the k - s and chi - squared statistics computed here is to quantify the significance of features we see in the plots , such as a claimed difference between two distributions , or a putative improvement of a mixture model over a single ensemble .
our aim is to ensure that the features discerned by eye are statistically significant and not due to chance and/or insufficient sample size .
we caution that because of the simplifications in our simulations ( including their limited 30 myr timespan ) , these statistics should not be used to fine - tune our simulation parameters to match the observations . when we extended the simulated timescale of the ed10@xmath143 ensemble from 30 myr to 180 myr , the distributions of tranet spacing , period ratio , and duration ratio remained just as different from the observed distributions , but the distribution of tranet multiplicities changed to provide a better match to the observed distribution .
as discussed in section 5.2 , performing extended timescale simulations is an important next step for a detailed quantitative comparison to the observations .
even if a single one of our ensembles could be made to adequately match the observations , our argument that the data support two modes of planet formation reflecting gas - rich vs. gas - poor environments would still carry weight .
the ensemble ed@xmath157 is itself a mixture model to a degree , since it contains simulations having a range of solid surface densities .
core coagulation times are exponentially sensitive to solid surface density @xcite . for solid surface densities toward the upper end of the range in ed@xmath157 , cores coagulate sufficiently quickly that gas is still present at the end of coagulation . for solid surface densities toward the lower end of the range , cores coagulate more slowly , in environments drained of gas .
the thesis of our paper is that the observations implicate two modes of core coagulation for super - earths : a gas - rich mode ( but not too gas - rich ; see @xcite and @xcite ) and a gas - poor mode .
a mixture of two populations is consistent with the statistical study by @xcite of m - dwarf systems ; using a parametric mixture model for the distribution of tranets , they found evidence for two populations , one spawning single tranets and another of high tranet multiplicity .
@xcite gave more weight to systems producing single tranets ( 1:1 ) than we do ( 0.5:1 ) .
here we find that a mixture reproduces not only the multiplicity statistics but also the @xmath222 , @xmath12 , and period ratio distributions . in terms of the implications of our results for planet formation ,
our mixture represents a set of planets that underwent most of their growth in the presence of some residual gas ( ed@xmath157 ) , plus another set that assembled after the gas disk had completely dissipated ( ed@xmath147 ; this ensemble could be replaced with one or more of the ed@xmath159 , ed@xmath156 , ed@xmath155 ensembles with different weightings ) . among systems with two or more tranets ,
the median eccentricity and inclination in ensemble ed@xmath157 are 0.02 and 0.7@xmath32 , respectively . without invoking tidal circularization ,
the median eccentricity is in good agreement with the rayleigh parameter @xmath227 ( corresponding to a median @xmath228 ) measured by @xcite for planets with transit timing variations .
furthermore , among our simulated pairs in ed@xmath157 with period ratios less than 2 , the median eccentricity is as small as @xmath229 .
the ensemble ed@xmath157 is the one that we argued in section 6 should also produce the lowest density planets .
in contrast , the ensembles that we argued should produce higher density planets have median tranet eccentricities of @xmath230 and inclinations of @xmath231 for ensembles @xmath232ed@xmath155 , ed@xmath156 , ed@xmath159 , ed@xmath233 , respectively . for period ratios less than 2 , the corresponding median eccentricities are @xmath234 @xmath235 .
the lower eccentricities and lower bulk densities of the ed@xmath157 ensemble may account for the result of @xcite that larger planets ( @xmath12.54 @xmath236 ) having lower densities @xcite have smaller eccentricities than planets @xmath237 .
another observational application of our simulations is to the finding of @xcite that planets with masses measured via transit timing variations ( ttvs , which necessarily involve tight orbital spacings ) have lower bulk densities than those measured via radial velocities ( which are easier to measure for more widely separated planets ; see also @xcite , and in particular @xcite for how selection effects based on mass may contribute to the result of @xcite ) . for the mixed ensemble plotted in the right panel of fig .
[ fig : tranet ] , period ratios less than 2 ( i.e. , period ratios typical for ttv planets ) are dominated by planets in the ed@xmath157 ensemble , which are expected to have lower densities .
furthermore , as described in section 6.1 , for a particular gas depletion factor and a range of solid surface densities , planets of a given mass having tighter spacings will also typically have lower densities .
the circumstances of the giant impact era of planet formation link super - earths orbital properties to their compositions .
two types of systems are established : dynamically hot ( widely spaced , eccentric , mutually inclined ) planets with high bulk densities , and dynamically cold ( tightly spaced , circular , flat ) planets with gas envelopes and low densities .
our solar system s terrestrial planets fall into the former category .
typical orbital properties for the latter category hill spacings @xmath9 , @xmath238 , and @xmath239 are dynamically colder .
the hill spacing is set by an eccentricity equilibrium between gravitational scatterings , which tend to increase @xmath50 , and mergers , which tend to damp it ( section 3 ) .
this eccentricity equilibrium predicts that lower inclination systems end up with smaller hill spacings , a result that is confirmed numerically . the two most important disk properties for determining the orbits and compositions of planets are the disk s solid surface density and its late - stage residual gas surface density ( section 6 ) .
disks with higher solid surface density and moderate gas surface density ( depleted by a factor of @xmath1100 relative to the minimum - mass solar nebula ) tend to produce tightly spaced , low @xmath50 and @xmath36 , low density planets .
higher solid surface density enables cores to form faster fast enough for cores to accrete volumetrically significant atmospheres from residual gas before it dissipates completely ( section 6.1 ; see also @xcite who make the same point ) .
a moderate late - stage gas surface density has enough dynamical friction to flatten the system allowing for an eccentricity equilibrium to be achieved at tighter spacings ( section 3 ) but not so much that mergers up to the final core mass are prevented .
higher gas surface densities prevent mergers , so cores grow only after the disk gas dissipates sufficiently .
if gas surface densities are too low and if the inner disk gas is not replenished ( see @xcite for a discussion of replenishment ) then planets fail to acquire significant atmospheres .
hence a happy medium for the gas density is required to produce tightly spaced , low density planets .
we found that we need a mixture of dynamically hot and dynamically cold systems to reproduce the observed hill spacings , period ratios , tranet multiplicities , and transit duration ratios of _ kepler _ super - earths ( section 7 )
. we can match the observed distributions with a combination of moderate and lower ( at the time of core assembly ) gas surface densities , in concert with a range of solid surface densities .
however , we under - produce the tail of ultra tightly - spaced systems ( period ratios @xmath240 1.5 ; e.g. , kepler-36 , @xcite ) .
our findings are consistent with previous work that modeled the multiplicity distribution parametrically and found the need for two populations to match the observed tranet multiplicity distribution @xcite . without needing to invoke tidal circularization ,
our simulations with moderate residual gas surface density can account for the low eccentricities found by @xcite and @xcite .
our finding that lower planetary bulk densities are linked to tighter spacings and smaller eccentricities can potentially account for the discovery by @xcite that larger super - earths ( having lower densities , e.g. , @xcite ) have smaller eccentricities .
it can also help to explain the result of @xcite that planets with masses measured via transit timing variations ( which are necessarily more tightly spaced ) have lower bulk densities . over the past couple decades
, a variety of correlations have been uncovered between the orbital and compositional properties of planetary bodies and properties of their host stars .
these include the connection between giant planet occurrence and host star metallicity ( e.g. , @xcite ) ; the correlation between spin - orbit misalignment and host star temperature ( e.g. , @xcite ) ; the relation between planet radius and host star metallicity ( e.g. , @xcite ) ; the correlation between dynamical excitation and host star metallicity ( e.g. , @xcite ) ; and the connection between kuiper belt objects compositional and orbital properties ( e.g. , @xcite ) . the theoretical connections we found here between super - earths orbital properties and their compositions provide a new test of formation theories .
the simplified simulations and first - order comparisons to observations presented here can be improved upon .
we caution that we did not incorporate photo - evaporation of planetary atmospheres into our models so future studies that search for the predicted links between orbital and compositional properties should either account for photo - evaporation or be limited to periods beyond @xmath115 days .
eccentricity damping from residual gas could be incorporated explicitly into the scaling arguments in section 3 .
more realistic simulations could implement prescriptions for atmospheric growth through accretion @xcite and collisional outcomes @xcite into mercury6 .
longer integrations ( e.g. , to 10 gyr instead of the 30 myr reported here ) are important for detailed quantitative comparisons to the observed sample , because the median spacing increases by about 10% per decade in integration time ( section 5.2 ; see also @xcite ) .
recently @xcite showed that a combination of planets formed from disks of solids with steep and shallow surface density profiles ( a range of @xmath44 ; eqn .
[ eqn : memb ] ) could reproduce the observed distributions of multiplicity , period ratio , and transit duration ratio . to distinguish between theories that rely on differences in surface density slope ( moriarty & ballard ) or on differences in the normalizations of both gas and solid surface densities ( this work )
, future studies should test how tranet properties vary with orbital period .
they should also probe for correlations between orbital and compositional properties correlations that our work predicts .
a follow - up study focused on planets near resonance could determine whether the gas damping treated here provides sufficient dissipation to produce the observed asymmetry in period ratios near the 3:2 and 2:1 resonances without invoking tides ( e.g. , @xcite ) .
finally , the origin of the most tightly spaced planets ( period ratios @xmath241 ) remains a mystery and may be a signature of migration .
we thank the apj referee and the apj statistics editor for constructive and thought - provoking reports that led to substantive improvements in our paper .
we also thank a mnras letters referee for helpful comments that inspired us to expand this work into a more detailed and comprehensive study .
we are grateful to sarah ballard , daniel fabrycky , brad hansen , edwin kite , eiichiro kokubo , renu malhotra , ruth murray - clay , hilke schlichting , and angie wolfgang for informative discussions and comments .
rid is supported by the berkeley miller institute and ejl by nserc under pgs d3 and the berkeley fellowship .
ec thanks nsf and nasa for financial support .
the valuable collection of planet candidates were discovered by nasa s _
kepler _ mission and compiled from nasa s exoplanet archive , operated by caltech , under contract with nasa under the exoplanet exploration program .
simulations were run on the savio computational cluster provided by berkeley research computing . | the giant impact phase of terrestrial planet formation establishes connections between super - earths orbital properties ( semimajor axis spacings , eccentricities , mutual inclinations ) and interior compositions ( the presence or absence of gaseous envelopes ) . using @xmath0-body simulations and analytic arguments ,
we show that spacings derive not only from eccentricities , but also from inclinations .
flatter systems attain tighter spacings , a consequence of an eccentricity equilibrium between gravitational scatterings , which increase eccentricities , and mergers , which damp them .
dynamical friction by residual disk gas plays a critical role in regulating mergers and in damping inclinations and eccentricities .
systems with moderate gas damping and high solid surface density spawn gas - enveloped super - earths with tight spacings , small eccentricities , and small inclinations .
systems in which super - earths coagulate without as much ambient gas , in disks with low solid surface density , produce rocky planets with wider spacings , larger eccentricities , and larger mutual inclinations .
a combination of both populations can reproduce the observed distributions of spacings , period ratios , transiting planet multiplicities , and transit duration ratios exhibited by _
kepler _ super - earths .
the two populations , both formed in situ , also help to explain observed trends of eccentricity vs. planet size , and bulk density vs. method of mass measurement ( radial velocities vs. transit timing variations ) .
simplifications made in this study including the limited timespan of the simulations , and the approximate treatments of gas dynamical friction and gas depletion history should be improved upon in future work to enable a detailed quantitative comparison to the observations . |
congruence is a fundamental concept in number theory .
two integers @xmath0 and @xmath1 are said to be congruent modulo a positive integer @xmath2 if their difference @xmath3 is integrally divisible by @xmath2 , written as @xmath4 @xcite .
congruence theory has been widely used in physics , biology , chemistry , computer science , and even music and business @xcite . because of the limited computational and storage ability of computers , congruence arithmetic is particularly useful and applicable to computing with numbers of infinite length @xcite .
significant and representative applications include generating random numbers @xcite , designing hash functions @xcite and checksumming in error detections @xcite . as a cornerstone of modern cryptography ,
congruence arithmetic has been successfully used in public - key encryption @xcite , secret sharing @xcite , digital authentication , and many other data security applications @xcite . despite the well - established congruence theory with a broad spectrum of applications , a comprehensive understanding of the congruence relation among natural numbers is still lacking .
our purpose is to uncover some intrinsic properties of the network consisting of natural numbers with congruence relations .
a link in the congruence network is defined in terms of the congruence relation @xmath5 , where @xmath1 is the reminder of @xmath6 divided by @xmath7 . for a fixed value of @xmath1
, we discern an infinite set of integer pairs @xmath8 . for each pair of such integers ,
a directed link from @xmath7 to @xmath6 ( suppose @xmath9 ) characterizes the congruence relation between @xmath7 and @xmath6 , giving rise to a congruence network for a given reminder @xmath1 .
let @xmath10 denote a congruence network , where @xmath11 is the largest integer considered .
note that congruence networks associated with different values of @xmath1 share the same set of nodes ( integers ) , thereby a multiplex network @xcite with a number of layers is formed , as shown in fig .
[ fig1](a ) . to our knowledge
, the multiplex congruence network ( mcn ) has not been explored in spite of some effort dedicated to complex networks associated with natural numbers @xcite .
we will demonstrate several unique and prominent properties of the mcn regarding some typical dynamical processes .
specifically , analytical results will show that all layers of the mcn are sparse with the same power - law degree distribution .
a counterintuitive property of the mcn is that every layer of the mcn has an extremely strong controllability , which significantly differs from ordinary scale - free networks requiring a large fraction of driver nodes . to steer the network in a layer , the minimum number of driver nodes is nothing but the reminder @xmath1 that is negligible as compared to the network size .
the controllability of the mcn is also very robust against targeted removal of nodes but relatively vulnerable to random failures , which is also in sharp contrast to ordinary scale - free networks .
this amazing robustness against attacks can be interpreted in terms of the multi - chain structure in mcn .
the mcn therefore sheds light on the design of heterogeneous networks with high searching efficiency @xcite and strong controllability simultaneously .
another application of the mcn is that it can graphically solve the simultaneous congruences problem in a more intuitive way than currently used methods , such as the garner s algorithm .
the solution of the simultaneous congruences problem is to locate common neighbors of relevant numbers in different layers .
this alternative approach by virtue of the mcn may have implication in cryptography based on simultaneous congruences . to our knowledge
, the multiplex congruence network ( mcn ) has not been explored in spite of some effort dedicated to complex networks associated with natural numbers @xcite
. we will demonstrate several unique and prominent properties of the mcn regarding some typical dynamical processes .
specifically , analytical results will show that all layers of the mcn are sparse with the same power - law degree distribution .
a counterintuitive property of the mcn is that every layer of the mcn has an extremely strong controllability , which significantly differs from ordinary scale - free networks requiring a large fraction of driver nodes . to steer the network in a layer , the minimum number of driver nodes is nothing but the number of reminder @xmath1 that is negligible as compared to the network size .
the controllability of the mcn is also very robust against targeted removal of nodes but relatively vulnerable to random failures , which is also in sharp contrast to ordinary scale - free networks .
this amazing robustness against attacks can be interpreted in terms of the multi - chain structure in mcn .
the mcn therefore sheds light on the design of heterogeneous networks with high searching efficiency @xcite and strong controllability simultaneously .
one of the most important applications of the mcn is that it allows to solve the simultaneous congruences problem in a much more efficient way than existing methods using the gauss algorithm . to tackle the problem with high computational efficiency
, the network of each layer should be stored in advance , e.g. , in a distributed storage system .
the solution of the simultaneous congruences problem is to locate common neighbors of relevant numbers in different layers .
this new efficient approach by virtue of the mcn offers deeper and broader insight into the modern cryptography and has potential applications in many other fields , such as communications and computer science .
mcn consists of a number of congrence networks ( layers ) @xmath12 , as shown in fig . [ fig1](a ) .
each layer contains all the natural numbers larger than @xmath1 but less than or equal to @xmath11 , so the size ( number of nodes ) of a layer is @xmath13 .
the remainder @xmath1 is a parameter that determines the structure of congruence network .
when @xmath14 , the congruence network reduces to a divisibility network @xcite , in which the dividend links to all of its divisors except itself .
the out - degrees of nodes are heterogeneous in each layer of the mcn .
we have analytically derived the distribution of the out - degrees in the thermodynamic limit ( see details in the section of * methods * ) : @xmath15 for large @xmath16 , the out - degree distribution becomes @xmath17 , thus @xmath10 is a typical scale - free network .
all @xmath12 have similar out - degree distributions , as shown in fig .
[ fig1](b ) , but the divisibility network @xmath18 has a different out - degree distribution . for small @xmath16 ,
@xmath19 of @xmath18 deviates from the other networks @xmath12 .
the main factor that accounts for the difference lies in that half of the nodes in @xmath18 have no outgoing links , but in @xmath12 there are only @xmath1 nodes without outgoing links .
analytical results demonstrate that the average degree of any layer increases logarithmically with the network size ( see fig .
[ fig1](b ) and the section of * methods * for details ) , and a larger value of @xmath1 corresponds to a sparser layer .
these results indicate that @xmath10 is always a sparse network .
hence , the mcn is compatible with a sparse storage , which is important for applying the mcn to solve real - world problems .
according to the definition of mcn , the numbers in each layer @xmath20 can be classified into @xmath1 arithmetic sequences : @xmath21 where @xmath22 ( @xmath23 denotes the largest integer not greater than @xmath24 ) .
the consecutive numbers in the sequence are linked from small to large , resulting in @xmath1 chains traversing all nodes in a layer , as shown in fig .
[ fig2](a ) .
the root node of a chain is the minimum number in the chain .
there are totally @xmath1 root nodes associated with @xmath1 chains .
the end of a chain is always the maximum number in the chain .
note that @xmath14 is a special case , because the arithmetic sequence does not exist in the layer @xmath18 , rendering the absence of the multi - chain structure in the divisibility network .
the above results indicate that although the divisibility network @xmath18 is a special case of @xmath10 , it has some fundamental differences from @xmath12 .
only when @xmath25 the multi - chain structure emerges , and the number of nodes without outgoing links is negligible . some evidence has suggested that the multi - chain structure and the absence of nodes with low out - degrees play an important role in the controllability of complex networks @xcite . in the next section
, we will further investigate the controllability properties of the mcn . in principle , the mcn composed of natural numbers is not a dynamical system , such that it can not be controlled .
however , because of the multi - chain structure , the mcn provides significant insight into the design of heterogeneous networked systems with strong controllability .
thus , we treat the mcn as a dynamical system and explore its unique and outstanding controllability properties .
the central problem of controlling complex networks is to discern a minimum set of driver nodes , on which external input signals are imposed to fully control the whole system .
let @xmath26 denote the minimum number of driver nodes and @xmath27 denote the fraction of driver nodes in a network .
in general , a network with a smaller value of @xmath27 is said to be more controllable . according to the exact controllability theory for complex networks @xcite and the sparsity of mcn
, we can prove that ( see in the section of * methods * ) @xmath28 and @xmath29 for large @xmath11 ( namely @xmath30 ) , @xmath31 and @xmath12 is considered as highly controllable .
furthermore , according to both the exact controllability theory @xcite and the structural controllability theory @xcite , the driver nodes are the @xmath1 root nodes of the chains in @xmath12 ( see details in the section of * methods * ) . meanwhile , the driver nodes are the hub nodes with the maximum degree . in comparison , due to
the absence of chains , the divisibility network @xmath18 with @xmath32 ( @xmath33 denotes the smallest integer not less than @xmath24 ) is hard to control for large @xmath11 .
it has been recognized that scale - free networks are often difficult to control @xcite . in particular , liu et al .
@xcite have analytically found that when the network size @xmath34 , one must control almost all nodes in order to fully control a scale - free network with scaling index @xmath35 .
the mcn is a scale - free network with scaling exponent @xmath36 , but one just needs to control @xmath1 root nodes to achieve full control .
such a strong controllability stems from the inherent multi - chain structure in mcn . from the perspective of structural controllability ,
all nodes in the chains in mcn are matched except the @xmath1 roots , which need to be controlled .
thus , the mcn is valuable for designing heterogeneous networks with strong controllability .
we also found that the mcn is strongly structurally controllable ( ssc ) because of the multi - chain structure ( see * methods * ) , which provides significant insight into the design of heterogeneous and controllable networks without exact link weights .
a network is said to be ssc if and only if its controllability will not be affected by the link weights in its adjacency matrix @xcite , or equivalently , for any distribution of link weights , the network will be fully controllable from the same set of driver nodes .
the ssc property implies that the mcn is robust against the fluctuation and uncertainty of link weights .
this is an outstanding feature with practical significance since sometimes link weights are hard to be exactly measured and they are sometimes time - varying in real situations .
the robustness against attacks is also a significant problem for the design of a controllable networked system @xcite .
we explore the robustness of the controllability of the mcn against attacks on nodes and find some unique properties , which is useful for the design of practical networks . on the one hand , due to the existence of chains rooted in @xmath1 driver nodes in mcn , targeted attacks to driver nodes will not destroy the multi - chain structure nor increase @xmath27 . here
, nodes critical for targeted attacks can be identified based on the rank of node degrees or their hierarchical structure @xcite . in mcn , driver nodes ( the @xmath1 root nodes )
become such critical nodes . in this
regard , mcn is robust against targeted attacks . on the other hand , random attacks to nodes
may cut some chains . as a result
, an additional driver is required to control each new breakpoint , leading to an increase of @xmath27 .
thus , the controllability of mcn is unusual in resisting attacks in the sense that it is robust against intentional attacks but vulnerable to random attacks , which significantly differs from general scale - free networks .
the results in comparing the mcn and sf networks generated by using static model @xcite are shown in fig .
[ fig2](b ) . to make an unbiased comparison , a scale - free network with the same scaling exponent @xmath37 and @xmath38 as the mcn is necessary .
however , because of the graphicality constraint , it is not possible to generate a random sf network with @xmath39 @xcite .
thus , we slightly release the requirement by using static model to generate sf networks with the same @xmath38 but with @xmath40 .
indeed , one can see that @xmath27 remains nearly unchanged under targeted attacks to driver nodes ; whereas random attacks to node causes clear increase of @xmath27 . this phenomenon is consistent with our analysis in terms of the multi - chain structure .
moreover , for the presence of random attacks , in a wide range of the fractions of failed nodes , the controllability of mcn is still better than sf networks in general .
one of the applications of mcn is that one can graphically solve the simultaneous congruences problem , which has implication in communication security and computer science . in particular
, one can find that mcn is exactly a topological representation of the system of simultaneous congruences @xcite .
a system of simultaneous congruences is a set of congruence equations : @xmath41 if the moduli @xmath42 are pairwise coprime , then a unique solution modulo @xmath43 exists .
this is the _
chinese remainder theorem _ ( crt ) , which has many applications in computing , coding and cryptography @xcite . a well - known algorithm to solve the simultaneous congruences in crt is the gaussian algorithm @xcite , also known as _
@xcite in ancient china . here , we present an intuitive approach based on mcn to solve the simultaneous congruences in eq .
( [ eq : sim ] ) .
firstly , we construct an mcn containing @xmath16 subnetworks with remainders @xmath44 , respectively . to focus on the minimum solution of eq .
( [ eq : sim ] ) , we set the maximum number in the network as @xmath45 .
then , we find the common successor neighbor of the nodes @xmath42 in this mcn , which is precisely the solution @xmath24 .
we use a well - known example of crt , recorded in _ sunzi suanjing _
@xcite , to demonstrate our approach .
the problem in this example is : ` suppose we have an unknown number of objects . when grouped in threes , 2
are left out , when grouped in fives , 3 are left out , and when grouped in sevens , 2 are left out .
how many objects are there ? '
this problem is equivalent to the following simultaneous congruences @xmath46 to solve the problem , we first construct an mcn of two layers , @xmath47 and @xmath48 , as shown in fig .
then , we find the common successor neighbor of the three moduli , 3 , 5 and 7 , and finally get the result @xmath49 .
it is noteworthy that the traditional algorithm for solving the simultaneous congruences problem , e.g. , the garner s algorithm @xcite , is more efficient than our algorithm based on the mcn that is essentially a brute - force search .
thus , it is infeasible to immediately use the graphical approach in data security .
however , the graphical algorithm offers new routes to the simultaneous congruences problem in the viewpoint of a complex network , which may be useful to improve the currently used algorithm .
is @xmath50 , and similarly @xmath51 and @xmath52 in the lower layer .
thus the common successor neighbor of the three nodes is @xmath53 , which is the solution of the simultaneous congruences problem described by in eq .
( [ eq : sz]).,width=340 ]
we have defined a multiplex congruence network composed of natural numbers and uncovered its unique topological features .
analytical results demonstrate that every layer of the multiplex network is a sparse and scale - free subnetwork with the same degree distribution .
counterintuitively , every layer with a scale - free structure has an extremely strong controllability , which significantly differs from ordinary scale - free networks . in general , a scale - free network with power - law degree distribution is harder to control than homogeneous networks .
this is attributed to the presence of hub nodes , at which dilation arises according to the structural control theory @xcite . as a result ,
downstream neighbor nodes of hubs are difficult to control . moreover , due to a large number of nodes connecting to hubs , scale - free networks are usually of weak controllability with a large fraction of driver nodes . in contrast , in spite of the scale - free structure of the congruence network , the long chains in each layer considerably inhibit dilation and reduce the number driver nodes .
furthermore , an interesting finding is that every layer is also strong structurally controllable in that link weights have no effect on the controllability .
this indicates that the controllability of the multiplex congruence network is extremely robust against the inherent limit to precisely accessing link weights in the real situation . to our knowledge ,
a scale - free network with strong structural controllability has not been reported prior to our congruence network .
an unusual controllability property is that the controllability of each layer is robust against targeted attacks to driver nodes , but relatively fragile to random failures of nodes , which is also different from common scale - free networks .
previously reported results @xcite demonstrate that targeted removal of high degree nodes and nodes in the top level of a hierarchical structure causes maximum damage to the network controllability .
under the two kinds of intentional removals , a network is easier to break to pieces , such that more driver nodes are required to achieve full control .
thus , targeted attacks are defined in terms of the two types of node removals . in the congruence network ,
high degree nodes and high level nodes are exactly identical , leading to the combination of the targeted attacks . strikingly , the congruence network is robust against the targeted attacks , because of the existence of the chains .
targeted attacks will not destroy the chains in the downstream of the attacked node . as a result , the number of driver nodes nearly does not increase , even when a large fraction of nodes has been targeted attacked .
the outstanding structural and controllability properties of the multiplex congruence network are valuable for designing heterogeneous networks with strong controllability and high searching efficiency rooted in the scale - free structure .
another application of the multiplex congruent network is to solve the simultaneous congruences problem in a graphical and intuitive manner .
the multiplex congruence network by converting the algebraic problem of solving simultaneous congruences equations to be a graphical problem of finding common neighbors in a graph , offers an alternative route to the traditional approaches . despite this property , the traditional algorithms , such as gaussian algorithm and
garner s algorithm , outperform the graphical method in computational efficiency .
hence , the graphical method is not applicable in data security at the present .
nevertheless , the graphical approach may inspire the combination of the graphical and algebraic method to improve the current algorithm , which is potentially valuable in communication security , computer science and many fields relevant to cryptography .
our work may also stimulate further effort toward studying of networks arising from natural relationships among numbers , with outstanding features and applied values .
many topological insights can be expected from complex networks consisting of natural numbers .
for a sub - network @xmath12 in mcn , the total number of nodes is @xmath13 and the number of nodes without out - links is @xmath1 .
the out - degree of a node labelled @xmath2 in the range of @xmath54 $ ] is 1 , because node @xmath2 can only link to one node i.e. node @xmath55 in the network ; similarly , the out - degree of a node labelled @xmath2 in the range of @xmath56 $ ] is 2 , because node @xmath2 can only link to two nodes i.e. nodes @xmath55 and @xmath57 ; similar scenarios appear to the other nodes .
thus , we can derive the distribution of out - degrees in the thermodynamic limit , as follows : @xmath58 in @xmath18 , the numbers larger than @xmath59 have no out - links , i.e. @xmath60 , and the numbers in the range of @xmath61 $ ] have only one out - link , i.e. @xmath62 .
analogously , as @xmath63 , @xmath64 this is the same as the in - degree distribution of a growing network with copying @xcite , which can be regarded as a random version of the divisibility network @xcite .
note that the minimum number @xmath65 in @xmath12 has the maximum out - degree @xmath66 and the second minimum number @xmath67 has the out - degree @xmath68 , and so on .
thus , the average degree @xmath69 of @xmath12 is @xmath70 where @xmath71 is the euler constant and @xmath72 is a constant , which is approximately @xmath73 when @xmath1 is very large .
( [ eq : av ] ) is not valid for the divisibility network @xmath18 . in @xmath18
, number 1 has the maximum out - degree @xmath74 , and 2 has the second maximum out - degree @xmath75 , and so on . in a similar way to eq .
( [ eq : av ] ) , we can obtain the average degree of @xmath18 , as @xmath76 which is consistent with the analytical results of the undirected divisibility network @xcite .
an arbitrary network with linear time - invariant dynamics under control can be described by @xmath77 where the vector @xmath78 stands for the states of @xmath11 nodes , @xmath79 denotes the coupling matrix ( transpose of the adjacency matrix ) of a network , @xmath80 is the vector of @xmath2 input signals , and @xmath81 is the input matrix . system ( [ eq : ct ] ) is said to be ( state ) controllable if the input signal @xmath82 imposed on a minimum number @xmath26 of driver nodes specified by control matrix @xmath83 can steer the state @xmath84 from any initial state to any target state in finite time . the level of controllability of the networked system ( [ eq : ct ] )
is defined by the fraction @xmath27 of driver nodes in the sense that a complex network is more controllable if a smaller fraction of driver nodes is needed to achieve full control . according to liu _
et al_. @xcite , the key is to find a matrix @xmath83 associated with the minimum number of controllers to ensure full control of system ( [ eq : ct ] ) .
because of the sparsity of mcn , @xmath26 of a layer @xmath10 is determined by @xcite @xmath85 because in @xmath12 each node - labelled number can only link to the node - labelled numbers larger than itself , the coupling matrix @xmath79 is a strictly lower - triangular matrix .
moreover , @xmath79 of @xmath12 is in a column echelon form because the minimum number linked from node @xmath2 is always @xmath55 .
in other words , the leading coefficient of the @xmath7th column in @xmath79 is precisely in the @xmath86th row . on the other hand ,
the last @xmath1 columns of @xmath79 are all zeros because the maximum @xmath1 nodes in @xmath12 have no out - links , so the rank of @xmath79 is exactly @xmath87 .
an example of matrix @xmath79 of @xmath88 is @xmath89 according to eq .
( [ eq : pbh ] ) , we can show that the minimum number of driver nodes @xmath26 for @xmath12 is @xmath1 and the value of non - zero elements in @xmath79 does not affect @xmath90 , which indicate that @xmath12 is ssc . according to the exact controllability theory @xcite , the control matrix @xmath83 to ensure full control of the congruence network @xmath12 should satisfy the following condition @xmath91 = n - r .
\label{eq : cond}\ ] ] notice that the rank of the matrix @xmath92 $ ] is contributed by the number of linearly independent rows , hence the input signals specified via @xmath83 should be imposed on the linearly dependence rows in @xmath79 so as to eliminate all linear correlations in eq .
( [ eq : cond ] ) .
apparently , the first @xmath1 rows in the coupling matrix @xmath79 of @xmath12 are all zero rows ( see eq.([eq : mtx1 ] ) ) , hence the @xmath1 driver nodes that need to be controlled to maintain full control are just the minimum @xmath1 nodes of the congruence network , i.e. the @xmath1 roots of the chains in the congruence network ( see fig . [ fig2](a ) ) .
the coupling matrix of the divisibility network @xmath18 is also a strictly lower - triangular matrix and in a column echelon form , but the rank of the matrix is @xmath93 , because in @xmath18 the node with labelled number larger than @xmath94 has no out - links , namely , the last @xmath95 columns of the matrix are all zeros .
an example of @xmath96 is @xmath97 therefore , according to eq .
( [ eq : pbh ] ) , we finally obtain the minimum number of driver nodes of @xmath18 as @xmath98 , indicating that one must control half of the nodes in order to control the whole divisibility network .
moreover , the value of non - zero elements in @xmath79 does not affect @xmath90 , which indicate that @xmath18 is ssc .
thus , the mcn composed of @xmath18 and @xmath12 is ssc .
* acknowledgments : * we thank dr .
zhengzhong yuan for useful discussions .
y.y . was supported by nsfc under grant nos .
61304177 , 71525002 and the fundamental research funds of bjtu under grant no .
2015rc042 . w .-
x.w . was supported by nsfc under grant no .
r.c . was supported by the hong kong research grants council under the grf grant cityu11208515 .
was supported by nsfc under grant no . | congruence theory has many applications in physical , social , biological and technological systems .
congruence arithmetic has been a fundamental tool for data security and computer algebra .
however , much less attention was devoted to the topological features of congruence relations among natural numbers . here
, we explore the congruence relations in the setting of a multiplex network and unveil some unique and outstanding properties of the multiplex congruence network .
analytical results show that every layer therein is a sparse and heterogeneous subnetwork with a scale - free topology .
counterintuitively , every layer has an extremely strong controllability in spite of its scale - free structure that is usually difficult to control .
another amazing feature is that the controllability is robust against targeted attacks to critical nodes but vulnerable to random failures , which also differs from normal scale - free networks . the multi - chain structure with a small number of chain roots arising from each layer accounts for the strong controllability and the abnormal feature .
the multiplex congruence network offers a graphical solution to the simultaneous congruences problem , which may have implication in cryptography based on simultaneous congruences .
our work also gains insight into the design of networks integrating advantages of both heterogeneous and homogeneous networks without inheriting their limitations . |
in recent years , distributed source coding ( dsc ) has received an increasing attention from the signal processing community .
dsc considers a situation in which two ( or more ) statistically dependent sources @xmath0 and @xmath1 must be encoded by separate encoders that are not allowed to talk to each other . performing separate lossless compression
may seem less efficient than joint encoding .
however , dsc theory proves that , under certain assumptions , separate encoding is optimal , provided that the sources are decoded jointly @xcite .
for example , with two sources it is possible to perform standard " encoding of the first source ( called _ side information _ ) at a rate equal to its entropy , and conditional " encoding of the second one at a rate lower than its entropy , with no information about the first source available at the second encoder ; we refer to this as asymmetric " slepian - wolf ( s - w ) problem .
alternatively , both sources can be encoded at a rate smaller than their respective entropy , and decoded jointly , which we refer to as symmetric " s - w coding .
dsc theory also encompasses lossy compression @xcite ; it has been shown that , under certain conditions , there is no performance loss in using dsc @xcite , and that possible losses are bounded below 0.5 bit per sample ( bps ) for quadratic distortion metric @xcite . in practice , lossy dsc
is typically implemented using a quantizer followed by lossless dsc , while the decoder consists of the joint decoder followed by a joint dequantizer .
lossless and lossy dsc have several potential applications , e.g. , coding for non co - located sources such as sensor networks , distributed video coding @xcite , layered video coding @xcite , error resilient video coding @xcite , and satellite image coding @xcite , just to mention a few .
the interested reader is referred to @xcite for an excellent tutorial .
traditional entropy coding of an information source can be performed using one out of many available methods , the most popular being arithmetic coding ( ac ) and huffman coding . conditional " ( i.e. , dsc ) coders are typically implemented using channel codes , by representing the source using the syndrome or the parity bits of a suitable channel code of given rate .
the syndrome identifies sets of codewords ( cosets " ) with maximum distance properties , so that decoding an ambiguous description of a source at a rate less than its entropy ( given the side information ) incurs minimum error probability . if the correlation between @xmath0 and @xmath1 can be modeled as a virtual " channel described as @xmath2 , with @xmath3 an additive noise process , a good channel code for that transmission problem is also expected to be a good s - w source code @xcite .
regarding asymmetric s - w coding , the first practical technique has been described in @xcite , and employs trellis codes .
recently , more powerful channel codes such as turbo codes have been proposed in @xcite , and low - density parity - check ( ldpc ) @xcite codes have been used in @xcite .
turbo and ldpc codes can get extremely close to channel capacity , although they require the block size to be rather large .
note that the constituent codes of turbo - codes are convolutional codes , hence the syndrome is difficult to compute . in @xcite
the cosets are formed by all messages that produce the same parity bits , even though this approach is somewhat suboptimal @xcite , since the geometrical properties of these cosets are not as good as those of syndrome - based coding . in @xcite a syndrome former is used to deal with this problem .
multilevel codes have also be addressed ; in @xcite trellis codes are extended to multilevel sources , whereas in @xcite a similar approach is proposed for ldpc codes .
besides techniques based on channel coding , a few authors have also investigated the use of source coders for dsc .
this is motivated by the fact that existing source coders obviously exhibit nice compression features that should be retained in a dsc coder , such as the ability to employ flexible and adaptive probability models , and low encoding complexity . in @xcite
the problem of designing a variable - length dsc coder is addressed ; it is shown that the problem of designing a zero - error such coder is np - hard . in @xcite a similar approach is followed ; the authors consider the problem of designing huffman and arithmetic dsc coders for multilevel sources with zero or almost - zero error probability .
the idea is that , if the joint density of the source and the side information satisfies certain conditions , the same codeword ( or the same interval for the ac process ) can be associated to multiple symbols .
this approach leads to an encoder with a complex modeling stage ( np - hard for the optimal code , though suboptimal polynomial - time algorithms are provided in @xcite ) , while the decoding process resembles a classical arithmetic decoder . as for symmetric s - w codes , a few techniques have been recently proposed .
a symmetric code can be obtained from an asymmetric one through time sharing , whereby the two sources alternatively take the role of the source and the side information ; however , current dsc coders can not easily accommodate this approach .
syndrome - based channel code partitioning has been introduced in @xcite , and extended in @xcite to systematic codes .
a similar technique is described in @xcite , encompassing non - systematic codes .
syndrome formers have also been proposed for symmetric s - w coding @xcite .
moreover , techniques based on the use of parity bits can also be employed , as they can typically provide rate compatibility .
a practical code has been proposed in @xcite using two turbo codes that are decoded jointly , achieving the equal rate point ; in @xcite an algorithm is introduced that employs turbo codes to achieve arbitrary rate splitting .
symmetric s - w codes based on ldpc codes have also been developed @xcite .
although several near - optimal dsc coders have been designed for simple ideal sources ( e.g. , binary and gaussian sources ) , the applications of practical dsc schemes to realistic signals typically incurs the following problems .
* channel codes get very close to capacity only for very large data blocks ( typically in excess of @xmath4 symbols ) . in many applications , however , the basic units to be encoded are of the order of a few hundreds to a few thousands symbols . for such block lengths ,
channel codes have good but not optimal performance . * the symbols contained in a block
are expected to follow a stationary statistical distribution .
however , typical real - world sources are not stationary .
this calls for either the use of short blocks , which weakens the performance of the s - w coder , or the estimation of conditional probabilities over contexts , which can not be easily accommodated by existing s - w coders . * when the sources are strongly correlated ( i.e. , in the most favorable case ) , very high - rate channel codes are needed ( e.g. , rate-@xmath5 codes )
. however , capacity - achieving channel codes are often not very efficient at high rate .
* in those applications where dsc is used to limit the encoder complexity , it should be noted that the complexity of existing s - w coders is not negligible , and often higher than that of existing non - dsc coders .
this seriously weakens the benefits of dsc . *
upgrading an existing compression algorithm like jpeg 2000 or h.264/avc to provide dsc functionalities requires at least to redesign the entropy coding stage , adopting one of the existing dsc schemes . among these issues , the block length is particularly important . while it has been shown that , on ideal sources with very large block length , the performance of some practical dsc coders can be as close as 0.09 bits to the theoretical limit @xcite ,
so far dsc of real - world data has fallen short of its expectations , one reason being the necessity to employ much smaller blocks .
for example , the prism video coder @xcite encodes each macroblock independently , with a block length of 256 samples .
for the coder in @xcite , the block length is equal to the number of 8x8 blocks in one picture ( 1584 for the cif format ) .
the performance of both coders is rather far from optimal , highlighting the need of dsc coders for realistic block lengths .
a solution to this problem has been introduced in @xcite , where an extension of ac , named distributed arithmetic coding ( dac ) , has been proposed for asymmetric s - w coding . moreover , in @xcite dac has been extended to the case of symmetric s - w coding of two sources at the same rate ( i.e. , the mid - point of the s - w rate region ) .
dac and its decoding process do not currently have a rigorous mathematical theory that proves they can asymptotically achieve the s - w rate region ; such theory is very difficult to develop because of the non - linearity of ac .
however , dac is a practical algorithm that was shown in @xcite to outperform other existing distributed coders . in this paper
, we build on the results presented in @xcite , providing several new contributions . for asymmetric coding ,
we focus on i.i.d .
sources as these are often found in many dsc applications ; for example , in transform - domain distributed video coding , dac could be applied to the bit - planes of transform coefficients , which can be modeled as i.i.d .
we optimize the dac using an improved encoder termination procedure , and we investigate the rate allocation problem , i.e. , how to optimally select the encoding parameters to achieve a desired target rate . we evaluate the performance of this new design comparing it with turbo and ldpc codes , including the case of extremely correlated sources with highly skewed probabilities .
this is of interest in multimedia applications because the most significant bit - planes of the transform coefficients of an image or video sequence are almost always equal to zero , and are strongly correlated with the side information . for symmetric coding ,
we extend our previous work in @xcite by introducing dac encoding and rate allocation procedures that allow to encode an arbitrary number of sources with arbitrary combination of rates .
we develop and test the decoder for two sources .
finally , it should be noted that an asymmetric dac scheme has been independently and concurrently developed in @xcite using quasi - arithmetic codes .
quasi - arithmetic codes are a low - complexity approximation to arithmetic codes , providing smaller encoding and decoding complexity @xcite .
these codes allow the interval endpoints to be only a finite set of points .
while this yields suboptimal compression performance , it makes the arithmetic coder a finite state machine , simplifying the decoding process with side information .
this paper is organized as follows . in sect .
[ sec : dac_decoder ] we describe the dac encoding process for the asymmetric case , in sect .
[ sec : dac_decoder ] we describe the dac decoder , and in sect .
[ sec : rate_sel ] we study the rate allocation and parameter selection problem . in sect .
[ sec : symm ] we describe the dac encoder , decoder and rate allocator for the symmetric case . in sect .
[ sec : results_asymm ] and [ sec : results_symm ] we report the dac performance evaluation results in the asymmetric and symmetric case respectively . finally , in sect .
[ sec : concl ] we draw some conclusions .
before describing the dac encoder , it should be noted that the ac process typically consists of a modeling stage and a coding stage . the modeling stage has the purpose of computing the parameters of a suitable statistical model of the source , in terms of the probability that a given bit takes on value 0 or 1 .
this model can be arbitrarily sophisticated , e.g. , by using contexts , adaptive probability estimation , and so forth .
the coding stage takes the probabilities as input , and implements the actual ac procedure , which outputs a binary codeword describing the input sequence .
let @xmath0 be a binary memoryless source that emits a semi - infinite sequence of random variables @xmath6 , @xmath7 , with probabilities @xmath8 and @xmath9 .
we are concerned with encoding the sequence @xmath10 $ ] consisting in the first @xmath11 occurrences of this source . the modeling and coding stages
are shown in fig .
[ fig : modeling]-a .
the modeling stage takes as input the sequence @xmath12 , and outputs an estimate of the probabilities @xmath13 and @xmath14 .
the coding stage takes as input @xmath12 , @xmath13 and @xmath14 , and generates a codeword @xmath15 .
the expected length of @xmath15 depends on @xmath13 and @xmath14 , and is determined once these probabilities are given . in order to use the dac , we consider two sources @xmath0 and @xmath1 , where @xmath1 is a binary memoryless source that emits random variables @xmath16 , @xmath7 , with probabilities @xmath17 and @xmath18 .
the first @xmath11 occurrences of this source form the side information @xmath19 $ ] .
we assume that @xmath0 and @xmath1 are i.i.d .
sources , and that @xmath6 and @xmath16 are statistically dependent for a given @xmath20 .
the entropy of @xmath0 is defined as @xmath21 , and similarly for @xmath1 .
the conditional entropy of @xmath0 given @xmath1 is defined as @xmath22 . for dac ,
three blocks can be identified , as in fig .
[ fig : modeling]-b , namely the modeling , rate allocation , and coding stages .
the modeling stage is exactly the same as in the classical ac .
the coding stage will be described in sect .
[ sec : dac_enc ] ; it takes as inputs @xmath12 , the probabilities @xmath13 and @xmath14 , and the parameter @xmath23 , and outputs a codeword @xmath24 . unlike a classical ac , where the expected rate is function of the source probabilities , and hence can not be selected _ a priori _ , the dac allows to select any desired rate not larger than the expected rate of a classical ac
this is very important , since in a dsc setting the rate for @xmath12 should depend not only on how much compressible " the source is , but also on how much correlated @xmath6 and @xmath16 are .
for this reason , in dac we also have a rate allocation stage that takes as input the probabilities @xmath13 and @xmath14 and the conditional entropy @xmath25 , and outputs a parameter @xmath23 that drives the dac coding stage to achieve the desired target rate . in this paper
we deal with the coding and rate allocation stages , and assume that the input probabilities @xmath13 , @xmath14 and conditional entropy @xmath25 are known _ a priori_. this allows us to focus on the distributed coding aspects of the proposed scheme , and , at the same time , keeps the scheme independent of the modeling stage .
we first review the classical ac coding process , as this sets the stage for the description of the dac encoder ; an overview can be found in @xcite .
the binary ac process for @xmath12 is based on the probabilities @xmath13 and @xmath14 , which are used to partition the @xmath26 interval into sub - intervals associated to possible occurrences of the input symbols . at initialization the current " interval is set to @xmath27 . for each input symbol @xmath28
, the current interval @xmath29 is partitioned into two adjacent sub - intervals of lengths @xmath30 and @xmath31 , where @xmath32 is the length of @xmath33 .
the sub - interval corresponding to the actual value of @xmath28 is selected as the next current interval @xmath34 , and this procedure is repeated for the next symbol . after all @xmath11 symbols have been processed ,
the sequence is represented by the final interval @xmath35 .
the codeword @xmath15 can consist in the binary representation of any number inside @xmath35 ( e.g. , the number in @xmath35 with the shortest binary representation ) , and requires approximately @xmath36 bits . similarly to other s - w coders , dac is based on the principle of inserting some ambiguity in the source description during the encoding process .
this is obtained using a modified interval subdivision strategy . in particular
, the dac employs a set of intervals whose lengths are proportional to the modified probabilities @xmath37 and @xmath38 , such that @xmath39 and @xmath40 . in order to fit the enlarged sub - intervals into the @xmath26 interval
, they are allowed to partially overlap .
this prevents the decoder from discriminating the correct interval , unless the side information is used .
the detailed dac encoding procedure is described in the following . at initialization the current " interval is set to @xmath41 . for each input symbol @xmath28 , the current interval @xmath42
is subdivided into two partially overlapped sub - intervals whose lengths are @xmath43 and @xmath44 .
the interval representing symbol @xmath28 is selected as the next current interval @xmath45 .
after all @xmath11 symbols have been processed , the sequence is represented by the final interval @xmath46 .
the codeword @xmath24 can consist in the binary representation of any number inside @xmath46 , and requires approximately @xmath47 bits .
this procedure is sketched in fig .
[ fig : dac ] . at the decoder side ,
whenever the codeword points to an overlapped region , the input symbol can not be detected unambiguously , and additional information must be exploited by the joint decoder to solve the ambiguity .
it is worth noticing that the dac encoding procedure is a generalization of ac . letting @xmath48 and @xmath49 leads to the ac encoding process described in sect .
[ sec : ac ] , with @xmath50 and @xmath51 .
it should also be noted that , for simplicity , the description of the ac and dac provided above assumes infinite precision arithmetic . the practical implementation used in sect . [
sec : results_asymm ] and [ sec : results_symm ] employs fixed - point arithmetic and interval renormalization .
the objective of the dac decoder is joint decoding of the sequence @xmath12 given the correlated side information @xmath52 .
the arithmetic decoding machinery of the dac decoder presents limited modifications with respect to standard arithmetic decoders ; a fixed - point implementation has been employed , with the same interval scaling and overlapping rules used at the encoder . in the following the arithmetic decoder state at the @xmath20-th decoding step
is denoted as @xmath53 .
the data stored in @xmath54 represent the interval @xmath55 and the codeword at iteration @xmath20 .
the decoding process can be formulated as a symbol - driven sequential search along a proper decoding tree , where each node represents a state @xmath54 , and a path in the tree represents a possible decoded sequence .
the following elementary decoding functions are required to explore the tree : * @xmath56_test - one - symbol_@xmath57 : it computes the sub - intervals at the @xmath20-th step , compares them with @xmath24 and outputs either an unambiguous symbol @xmath58 ( if @xmath24 belongs to one of the non - overlapped regions ) , or an ambiguous symbol @xmath59 . in case of unambiguous decoding , the new decoder state @xmath60 is returned for the following iterations .
* @xmath61_force - one - symbol_@xmath62 : it forces the decoder to select the sub - interval corresponding to the symbol @xmath63 regardless of the ambiguity ; the updated decoder state is returned . in fig .
[ fig : dec_tree ] an example of a section of the decoding tree is shown . in this example
the decoder is not able to make a decision on the @xmath20-th symbol , as _ test - one - symbol _ returns @xmath59 . as a consequence ,
two alternative decoding attempts are pursued by calling _ force - one - symbol _ with @xmath58 respectively . in principle , by iterating this process , the tree @xmath64 , representing all the possible decoded sequences , can be explored .
the best decoded sequence can finally be selected applying the _ maximum a posteriori _ ( map ) criterion @xmath65 . in general , exhaustive search can not be applied due to the exponential growth of @xmath64 .
a viable solution is obtained applying the breadth - first sequential search known as @xmath66-algorithm @xcite ; at each tree depth , only the @xmath66 nodes with the best partial metric are retained .
this amounts to visiting only a subset of the most likely paths in @xmath64 .
the map metric for a given node can be evaluated as follows : @xmath67 metric can be expressed into additive terms by setting : @xmath68 where @xmath69 and @xmath70 represent the additive metric to be associated to each branch of @xmath64 .
the pseudocode for the dac decoder is given in algorithm [ alg : dac_decoder ] , where @xmath71 represents the list of nodes in @xmath64 explored at depth @xmath20 ; each tree node stores its corresponding arithmetic decoder state @xmath72 and the accumulated metric @xmath73 .
initialize @xmath74 with root node ( @xmath75 ) set symbol counter @xmath76 @xmath56_test - one - symbol_@xmath57 @xmath61_force - one - symbol_@xmath77 @xmath78 insert @xmath79 in @xmath80 @xmath81 insert @xmath79 in @xmath80 sort nodes in @xmath80 according to metric @xmath82 keep only the @xmath66 nodes with best metric in @xmath80 output @xmath83 ( sequence corresponding to the first node stored in @xmath84 ) it is worth pointing out that @xmath66 has to be selected as a trade - off between the memory / complexity requirements and the error probability , i.e. , the probability that the path corresponding to the original sequence @xmath12 is accidentally dropped .
as in the case of standard viterbi decoding , the path metric turns out to be stable and reliable as long as a significant amount of terms , i.e. , number of decoded symbols @xmath63 , are taken into account . in the pessimistic case when all symbol positions @xmath20 trigger a decoder branching , given @xmath66 , one can guarantee that at least @xmath85 symbols are considered for metric comparisons and pruning . on the other hand , in practical cases ,
the interval overlap is only partial and branching does not occur at every symbol iteration .
all the experimental results presented in sect .
[ sec : results_asymm ] have been obtained using @xmath86 , while the trade - off between performance and complexity is analyzed in sect .
[ sec : perf_compl ] . finally , metric reliability can not be guaranteed for the very last symbols of a finite - length sequence @xmath12 . for channel codes , e.g. , convolutional codes , this issue is tackled by imposing a proper termination strategy , e.g. , forcing the encoded sequence to end in the first state of the trellis .
a similar approach is necessary when using dac .
examples of ac termination strategies are encoding a known termination pattern or end - of - block symbol with a certain probability or , in the case of context - based ac , driving the ac encoder in a given context . for dac
, we employ a new termination policy that is tailored to its particular features . in particular , termination is obtained by encoding the last @xmath87 symbols of the sequence without interval overlap , i.e. , using @xmath88 , for all symbols @xmath28 with @xmath89 . as a consequence ,
no nodes in the dac decoding tree will cause branching in the last @xmath87 steps , making the final metrics more reliable for the selection of the most likely sequence .
however , there is a rate penalty for the termination symbols .
the length of codeword @xmath24 is determined by the length @xmath90 of the final interval , which in turn depends on how much @xmath37 and @xmath38 are larger than @xmath13 and @xmath14 . as a consequence , in order to select the desired rate , it is important to quantitatively determine the dependence of the expected rate on the overlap , because this will drive the selection of the desired amount of overlap .
moreover , we also need to understand how to split the overlap in order to achieve good decoding performance . in the following we derive the expected rate obtained by the dac as a function of the set of input probabilities and the amount of overlap .
we are interested in finding the expected rate @xmath91 ( in bps ) of the codeword used by the dac to encode the sequence @xmath12 .
this is given by the following formula : @xmath92 this can be derived straightforwardly from the property that the codeword generated by an ac has an expected length that depends on the size of the final interval , that is , on the product of the probabilities @xmath93 , and hence on the amount of overlap .
the expectation is computed using the true probabilities @xmath94 .
we set @xmath95 , where @xmath96 , so that @xmath97 .
this amounts to enlarging each interval by an amount proportional to the overlap factors @xmath98 .
the expected rate achieved by the dac becomes @xmath99 where @xmath100 , and @xmath101 .
note that @xmath102 represents the rate contribution of symbol @xmath103 yielded by standard ac , while @xmath104 represents the decrease of this contribution , i.e. , the average number of bits saved in the binary representation of the @xmath103-th input symbol .
once a target rate has been selected , the problem arises of selecting @xmath105 . as an example
, a possible choice is to take equal overlap factors @xmath106 .
this implies that each interval is enlarged by a factor @xmath107 that does not depend on the source probability @xmath108 .
this leads to a target rate @xmath109 it can be shown that this choice minimizes the rate @xmath91 for a given total amount of overlap @xmath110 ; the computations are simple and are omitted for brevity .
this choice is not necessarily optimal in terms of the decoder error probability .
however , optimizing for the error probability is impractical because of the nonlinearity of the arithmetic coding process . in practice
, one also has to make sure that the enlarged intervals @xmath111 and @xmath112 are both contained inside the @xmath26 interval .
e.g. , taking equal overlap factors as above does not guarantee this .
we have devised the following rule that allows to achieve any desired rate satisfying the constraint above .
we apply the following constraint : @xmath113 with @xmath23 a positive constant independent of @xmath103 .
this leads to @xmath114 this can be interpreted as an additional constraint that the rate reduction for symbols 0 " and 1 " depends on their probabilities , i.e. , the least probable symbol undergoes a larger reduction . using ( [ eq : criterion ] )
, it can be easily shown that the expected rate achieved by the dac can be written as @xmath115 thus , the allocation problem for an i.i.d .
source is very simple .
we assume that the conditional entropy @xmath25 is available as in fig .
[ fig : modeling]-b , modeling the correlation between @xmath0 and @xmath1 . in asymmetric dsc
, @xmath12 should be ideally coded at a rate arbitrarily close to @xmath25 . in practice , due to the suboptimality of any practical coder
, some margin @xmath116 should be taken .
hence , we assume that the allocation problem can be written as @xmath117 . since @xmath118 is a constant and @xmath119 and @xmath120
are given , one can solve for @xmath23 and then perform the encoding process .
finally , it should be noted that , while we have assumed that @xmath0 and @xmath1 are i.i.d . , the dac concept can be easily extended to a nonstationary source .
this simply requires to consider all probabilities and overlap factors as depending on index @xmath20 ; all computations , including the design of the overlap factors and the derivation of the target rate , can be extended straightforwardly .
a possible application is represented by context - based coding or markov modeling of correlated sources .
there is one caveat though , in that , if the probabilities and context of each symbol are computed by the decoder from past symbols , decoding errors can generate significant error propagation .
in many applications , it is preferable to encode the correlated sources at similar rather than unbalanced rates ; in this case , symmetric s - w coding can be used . considering a pair of sources , in symmetric s - w coding both @xmath0 and @xmath1
are encoded using separate dacs .
we denote as @xmath24 and @xmath121 the codewords representing @xmath0 and @xmath1 , and @xmath122 and @xmath123 the respective rates . with dac
, the rate of @xmath0 and @xmath1 can be adjusted with a proper selection of the parameters @xmath23 and @xmath124 for the two dac encoders .
however , it should be noted that , for the same total rate , not all possible choices of @xmath23 and @xmath124 are equally good , because some of them could complicate the decoder design , or be suboptimal in terms of error probability . to highlight the potential problems of a straightforward extension of the asymmetric dac ,
let us assume that @xmath23 and @xmath124 can be chosen arbitrarily .
this would require a decoder that performs a search in a symbol - synchronous tree where each node represents _ two _ sequential decoder states @xmath125 for @xmath0 and @xmath1 respectively .
if the interval selection is ambiguous for both sequences , the four possible binary symbol pairs ( 00,01,10,11 ) need to be included in the search space ; this would accelerate the exponential growth of the tree , and quickly make the decoder search unfeasible .
this example shows that some constraints need to be put on @xmath23 and @xmath124 in order to limit the growth rate of the search space . to overcome this problem
, we propose an algorithm that applies the idea of time - sharing to the dac .
the concept of time - shared dac has been preliminarly presented in @xcite for a pair of sources in the subcase @xmath126 , i.e. providing only the mid - point of the s - w rate region . in the following
we extend this to an arbitrary combination of rates , and show how this can be generalized to an arbitrary number of sources . for two sources ,
the idea is to divide the set of input indexes @xmath127 in two disjoint sets such that , at each index @xmath20 , ambiguity is introduced in at most one out of the two sources . in particular , for sequences @xmath12 and @xmath52 of length @xmath11 , let @xmath128 and @xmath129 be the subsets of even and odd integer numbers in @xmath130 respectively .
we employ a dac on @xmath12 and @xmath52 , but the choice of parameters @xmath23 and @xmath124 differs . in particular , we let the parameters depend on the symbol index @xmath20 , i.e. , @xmath131 and @xmath132 .
the dac of @xmath12 employs parameter @xmath133 for all @xmath134 , and @xmath135 otherwise .
vice versa , @xmath52 is encoded with parameter @xmath136 for all @xmath137 , and @xmath138 otherwise . as a consequence of these constraints , at each step of the decoding process , ambiguity appears in at most one out the two sequences . in this way
, the growth rate of the decoding tree remains manageable , as no more than two new states are generated at each transition , exactly as in the asymmetric dac decoder ; this also makes the map metric simpler .
the conceptual relation with time - sharing is evident . since , during the dac encoding process , for each input symbol the ambiguity is introduced in at most one out the two encoders , this corresponds to switching the role of side information between either source on a symbol - by - symbol basis . by varying the parameters @xmath23 and @xmath124 , all combinations of rates can be achieved .
the achieved rates can be derived repeating the same computations described in sect .
[ sec : rate_sel ] , and can be expressed as @xmath139 and @xmath140 .
the rate allocation problem amounts to selecting suitable rates @xmath122 and @xmath123 such that @xmath141 , @xmath142 , and @xmath143 . in practice one
will typically take some margin @xmath144 , such that @xmath145 ; for safety , a margin should also be taken on @xmath122 and @xmath123 with respect to the conditional entropy .
since the prior probabilities of @xmath0 and @xmath1 are given , one can solve for @xmath23 and @xmath124 , and then perform the encoding process .
thus , the whole s - w rate region can be swept . similarly to the asymmetric case
, the symmetric decoding process can be viewed as a search along a tree ; however , specifically for the case of two correlated sources , each node in tree represents the decoding states @xmath146 of two sequential arithmetic decoders for @xmath12 and @xmath52 respectively . at each iteration , sequential decoding
is run from both states .
the time - sharing approach guarantees that , for a given index @xmath20 , the ambiguity can be found only in one of the two decoders .
therefore , at most two branches must be considered , and the tree can be constructed using the same functions introduced in sect .
[ sec : dac_decoder ] for the asymmetric case .
this would be the same also for @xmath147 sources .
in particular , for @xmath148 , _ test - one - symbol_(@xmath149 ) yields an unambiguous symbol @xmath150 , whereas ambiguity can be found only while attempting decoding for @xmath12 with _ test - one - symbol_(@xmath149 ) .
in conclusion , from the node @xmath146 the function _ test - one - symbol _ is used on both states . if ambiguity is found on @xmath63 , _ force - one - symbol _
is then used to explore the two alternative paths for @xmath63 , whereas @xmath151 is used as side information for branch metric evaluation . in the case
that @xmath152 , the roles of @xmath12 and @xmath52 are exchanged .
therefore , algorithm [ alg : dac_decoder ] can be easily extended to the symmetric case by alternatively probing either @xmath12 or @xmath52 for ambiguity , and possibly generating a branching .
the joint probability distribution can be written as @xmath153 the symmetric encoder and decoder can be easily generalized to an arbitrary number @xmath147 of sources .
the idea is to identify @xmath147 subsets of input indexes @xmath127 such that , at each symbol index @xmath20 , ambiguity is introduced in at most one out of the @xmath147 sources . in particular , for sequences @xmath154 of length @xmath11 , let @xmath155 be disjoint subsets of @xmath156 .
we denote the dac parameters as @xmath157 .
the dac of @xmath158 employs parameter @xmath159 for all @xmath160 , and @xmath161 otherwise . as a consequence of these constraints , at each step of the decoding process , ambiguity appears in at most one out the @xmath147 sequences .
note that this formulation also encompasses the case that one or more sources are independent of each other and from all the others ; these sources can be coded with a classical ac , taking @xmath162 for this source .
the selection of the sets @xmath163 and the overlap factors @xmath164 , for @xmath165 , is still somewhat arbitrary , as the expected rate of source @xmath103 depends on both the cardinality of @xmath163 and the value of @xmath164 . in a realistic application
it would be more practical to fix the sets @xmath163 once and for all , and to modify the parameters @xmath164 so as to obtain the desired rate .
this is because , for time - varying correlations , one has to update the rate on - the - fly . in
a distributed setting , varying one parameter @xmath164 requires to communicate the change only to source @xmath103 , while varying the sets @xmath163 requires to communicate the change to all sources .
therefore , we define @xmath166 such that the @xmath147 statistically dependent sources take in turns the role of the side information .
any additional independent sources are coded separately using @xmath167 . in particular , we set @xmath168 , where @xmath169 denotes the remainder of the division between two integers , and @xmath170 .
the dac encoder for the @xmath103-th source inserts ambiguity only at time instants @xmath160 . at each node ,
the decoder stores the states of the @xmath147 arithmetic decoders , and possibly performs a branching if the codeword related to the only potentially ambiguous symbol at the current time @xmath20 is actually ambiguous .
although this encoding and decoding structure is not necessarily optimal , it does lead to a viable decoding strategy .
in the following we provide results of a performance evaluation carried out on dac . we implement a communication system that employs a dac and a joint decoder , with no feed - back channel ; at the decoder , pruning is performed using the m - algorithm @xcite , with m=2048 .
the side information is obtained by sending the source @xmath0 through a binary symmetric channel with transition probability @xmath171 , which measures the correlation between the two sources .
we simulate a source with both balanced ( @xmath172 ) and skewed ( @xmath173 ) symbol probabilities .
the first setting implies @xmath174 and @xmath175 , where @xmath25 depends on @xmath171 .
the closer @xmath171 to 0.5 , the less correlated the sources , and hence the higher @xmath25 . in the skewed case , given @xmath176 , @xmath120
is fixed , whereas both @xmath177 and @xmath25 depend on @xmath171 . unless otherwise specified , each point of the figures / tables presented in the following has been generated averaging the results obtained encoding @xmath178 samples . as a first experiment ,
the benefit of the termination policy is assessed .
stationary source @xmath0 emits sequences @xmath12 of @xmath179 symbols , with @xmath172 and @xmath180 , which are encoded with dac at fixed rate @xmath181 bps , i.e. , @xmath182 bps higher than the theoretical s - w bound .
for @xmath1 we assume ideal lossless encoding at average rate @xmath183 bps , so that the total average rate of @xmath0 and @xmath1 is 1.5 bps .
the bit error rate ( ber ) yielded by the decoder is measured for increasing values of the number of termination symbols @xmath87 . the same simulation is performed with @xmath184 . in all simulated cases , the dac overlap has been selected to compensate for the rate penalty incurred by the termination , so as to achieve the 1.5 bps overall target rate .
the overlap factors @xmath185 are selected according to .
the results are shown in fig .
[ fig : term ] ; it can be seen that the proposed termination is effective at reducing the ber .
there is a trade - off in that , for a given rate , increasing @xmath87 reduces the effect of errors in the last symbols , but requires to overlap the intervals more .
it is also interesting to consider the position of the first decoding error as , without termination , errors tend to cluster at the end of the block . for @xmath179 ,
the mean position value is 191 , 178 , 168 , 161 and 95 , with standard deviation 13 , 18 , 25 , 36 and 49 , respectively for @xmath87 equal to 0 , 5 , 10 , 15 and 20 .
for @xmath184 , the mean value is 987 , 954 , 881 , 637 and 536 , with standard deviation 57 , 124 , 229 , 308 and 299 .
the optimal values of @xmath87 are around 15 - 20 symbols .
therefore , we have selected @xmath186 and used this value for all the experiments reported in the following .
( number of termination symbols ) ; @xmath172 , total rate = 1.5 bps , rate of @xmath12 = 0.5 bps , @xmath180.,width=453 ] next , an experiment has been performed to validate the theoretical analysis of the effects of different overlap designs shown in sect .
[ sec : rate_sel]-b . in fig .
[ fig : overlap ] the performance obtained by using the design of equations and respectively is shown .
the experimental settings are @xmath179 , @xmath187 , fixed rate for @xmath12 of 0.5 bps , and total average rate for @xmath0 and @xmath1 equal to 1.5 bps , with ideal lossless encoding of @xmath1 at rate @xmath177 .
the ber is reported as a function of the source correlation expressed in terms of @xmath188 .
it is worth noticing that the performance yielded by different overlap design rules are almost equivalent .
note that the rule in consistently outperforms that in , confirming that this latter is only optimal for the rate .
there is some difference when @xmath188 is very high ( i.e. , for weakly correlated sources ) .
however , this case is of marginal interest since the performance is poor ( the ber is of the order of 0.1 ) . , total rate = 1.5 bps).,width=453 ] the performance of the proposed system is compared with that of a system where the dac encoder and decoder are replaced by a punctured turbo code similar to that in @xcite .
we use turbo codes with rate-@xmath189 generator ( 17,15 ) octal ( 8 states ) and ( 31,27 ) octal ( 16 states ) , and employ s - random interleavers , and 15 decoder iterations .
we consider the case of balanced source ( @xmath190 ) and skewed source ( in particular @xmath191 and @xmath187 ) . for a skewed source , as an improvement with respect to @xcite ,
the turbo decoder has been modified by adding to the decoder metric the _ a priori _ term , as done in @xcite .
block sizes @xmath192 , @xmath179 and @xmath184 have been considered ( with s - random interleaver spread of 5 , 11 and 25 respectively ) ; this allows to assess the dac performance at small and medium block lengths .
besides turbo codes , we also considered the rate - compatible ldpc codes proposed in @xcite .
for these codes , a software implementation is publicly available on the web ; among the available pre - designed codes , we used the matrix for @xmath193 , which is comparable with the block lengths considered for the dac and the turbo code .
the results are worked out in a fixed - rate coding setting as in @xcite , i.e. , the rate is the same for each sample realization of the source . fig .
[ fig:1 ] reports the results for the balanced source case ; the abscissa is @xmath188 , and is related to @xmath171 .
the performance is measured in terms of the residual ber after decoding , which is akin to the distortion in the wyner - ziv binary coding problem with hamming metric .
both the dac and the turbo code generate a description of @xmath12 at fixed rate 0.5 bps ; the total average rate of @xmath0 and @xmath1 is 1.5 bps , with ideal lossless encoding of @xmath1 at rate @xmath177 .
since @xmath183 , we also have that @xmath175 .
this makes it possible to compare these results with the case of skewed sources which is presented later in this section , so as to verify that the performance is uniformly good for all distributions .
the wyner - ziv bound for a doubly symmetric binary source with hamming metric is also reported for comparison . as can be seen
, the performance of dac slightly improves as the block length increases .
this is mostly due to the effect of the termination . as the number of bits used to terminate the encoder is chosen independently of the block length , the rate penalty for non overlapping the last bits weights more when the block length is small , while the effect vanishes for large block length . in @xcite , where the termination effect is not considered ,
the performance is shown to be almost independent of the block size .
it should also be noted that the value of @xmath66 required for near - optimal performance grows exponentially with the block size . as a consequence ,
the memory which leads to near - optimal performance for @xmath192 or @xmath179 limits the performance for @xmath184 .
we compared both 8-states and 16-states turbo codes .
the 8-states code is often used in practical applications , as it exhibits a good trade - off between performance and complexity ; the 16-states code is more powerful , and requires more computations .
it can be seen that , for block length @xmath192 and @xmath179 , the proposed system outperforms the 8-states and 16-states turbo codes . for block length @xmath184 ,
the dac performs better than the 8-states turbo code , and is equivalent to the 16-states code .
it should be noted that , in this experiment , only the channel coding performance " of the dac is tested , since for the balanced source no compression is possible as @xmath194 .
consequently , it is remarkable that the dac turns out to be generally more powerful than the turbo code at equal block length .
note that the performance of the 16-states code is limited by the error floor , and could be improved using an ad - hoc design of the code or the interleaver ; the dac has no error floor , but its waterfall is less steep . for @xmath195 , a result
not reported in fig .
[ fig:1 ] shows that the dac with @xmath179 and @xmath184 also outperform the 8-state turbo - coder with @xmath196 . in fig .
[ fig:1 ] and in the following , it can be seen that turbo codes do not show the typical cliff - effect .
this is due to the fact that , at the block lengths considered in this paper , the turbo code is still very far from the capacity ; its performance improves for larger block lengths , where the cliff - effect can be seen . in terms of the rate penalty , setting a residual ber threshold of @xmath197 , for @xmath179 the dac is almost 0.3 bps away from the s - w limit , while the best 16-state turbo code simulated in this paper is 0.35 bps away ; for @xmath184 the dac is 0.26 bpp away , while the best 8-state turbo code is 0.30 bps away .
the performance of the ldpc code for @xmath193 is halfway between the turbo codes for @xmath179 and @xmath184 , and hence very similar to the dac .
, total rate = 1.5 bps , rate for @xmath12 = 0.5 bps ) : dac versus turbo coding , balanced source .
dac : distributed arithmetic coding ; tc8s and tc16s : 8- and 16-state turbo code with s - random interleaver ; ldpc - r and ldpc - i : regular and irregular ldpc codes from @xcite.,width=453 ] the results for a skewed source are reported in fig . [ fig:3 ] for @xmath187 . in this
setting , we select various values of @xmath188 , and encode @xmath12 at fixed rate such that the total average rate for @xmath0 and @xmath1 equals 1.5 bps , with ideal lossless encoding of @xmath1 at rate @xmath177 . for fig .
[ fig:3 ] , from left to right , the rates of @xmath12 are respectively 0.68 , 0.67 , 0.66 , 0.64 , 0.63 , 0.61 , 0.59 , and 0.58 bps . consistently with @xcite , all turbo codes considered in this work perform rather poorly on skewed sources . in
@xcite this behavior is explained with the fact that , when the source is skewed , the states of the turbo code are used with uneven probability , leading to a smaller equivalent number of states . on the other hand ,
the dac has good performance also for skewed sources , as it is designed to work with unbalanced distributions .
the performance of the ldpc codes is similar to that of the best turbo codes , and slightly worse than the dac .
similar remarks can be made in the case of @xmath191 , which is reported in fig .
[ fig:2 ] . in this case
, we have selected a total rate of 1 bps , since the source is more unbalanced and hence easier to compress .
the rates for @xmath12 are respectively 0.31 , 0.34 , 0.37 , 0.39 , 0.42 , 0.44 , and 0.47 bps . in this case
the turbo code performance is better than in the previous case , although it is still poorer than dac .
this is due to the fact that the sources are more correlated , and hence the crossover probability on the virtual channel is lower .
therefore , the turbo code has to correct a smaller number of errors , whereas for @xmath187 the correlation was weaker and hence the crossover probability was higher . , total rate = 1.5 bps ) : dac versus turbo coding , skewed source .
dac : distributed arithmetic coding ; tc8s and tc16s : 8- and 16-state turbo code with s - random interleaver ; ldpc - r and ldpc - i : regular and irregular ldpc codes from @xcite.,width=453 ] , total rate = 1 bps ) : dac versus turbo coding , skewed source .
dac : distributed arithmetic coding ; tc8s and tc16s : 8- and 16-state turbo code with s - random interleaver ; ldpc - r and ldpc - i : regular and irregular ldpc codes from @xcite.,width=453 ] we also considered the case of strongly correlated sources , for which high - rate channel codes are needed .
these sources are a good model for the most significant bit - planes of several multimedia signals . due to the inefficiency of syndrome - based coders
, practical schemes often assume that no dsc is carried out on those bit - planes , e.g. , they are not transmitted , and at the decoder they are directly replaced by the side information @xcite .
the results are reported in tab .
[ tab : corr ] for the dac and the 16-state turbo code , when a rate of 0.1 bps is used for @xmath12 .
the table also reports the cross - over probability @xmath171 , corresponding , for a balanced source , to the performance of an uncoded system that reconstructs @xmath12 as the side information @xmath52 .
as can be seen , the dac has similar performance to the turbo codes and ldpc codes , and becomes better when the source is extremely correlated , i.e. , @xmath198 .
@xmath25 & @xmath171 & dac & tc16s + 0.1 & @xmath199 & @xmath200 & @xmath201 + 0.01 & @xmath202 & @xmath203 & @xmath204 + 0.001 & @xmath205 & @xmath206 & @xmath207 + + @xmath25 & @xmath171 & dac & tc16s + 0.1 & @xmath199 & @xmath208 & @xmath209 + 0.01 & @xmath202 & @xmath210 & @xmath211 + 0.001 & @xmath205 & @xmath212 & @xmath213 + + @xmath25 & @xmath171 & ldpc - r & ldpc - i + 0.1 & @xmath199 & @xmath214 & @xmath215 + 0.01 & @xmath202 & @xmath216 & @xmath217 + 0.001 & @xmath205 & @xmath218 & @xmath219 + [ tab : corr ] finally , the coding efficiency of dac is measured in terms of expected rate required to achieve error - free decoding .
this amounts to re - encoding the sequence at increasing rates , and represents the optimal dac performance if the encoder could exactly predict the decoder behavior .
since each realization of the source is encoded using a different number of bits , this case is referred to as variable - rate encoding .
this scenario is representative of practical distributed compression settings , e.g. , @xcite , in which one seeks the shortest code that allows to reconstruct without errors each realization of the source process . for this simulation ,
the following setup is used .
the source correlation @xmath25 is kept constant and , for each sample realization of the source , the total rate is progressively increased beyond the s - w bound , in steps of 0.01 bps , until error - free decoding is obtained .
this operation is repeated on 1000 different realizations of the source ; the mean value and standard deviation of the rates yielding correct decoding are then computed .
the results have been worked out for block length @xmath179 , with probabilities @xmath172 and @xmath191 . for @xmath172 , the conditional entropy @xmath25 (
i.e. , the s - w bound ) has been set to 0.5 bps . for @xmath191 , the joint entropy
@xmath188 has been set to 1 bps ; this amounts to coding @xmath1 at the ideal rate of @xmath220 bps , with a s - w bound @xmath221 bps .
the results are reported in tab . [
tab : vrate ] .
as can be seen , the dac has a rate loss of about 0.06 bps with respect to the s - w bound for both the symmetric and skewed source .
the turbo code exhibits a loss of about 0.2 bps and 0.13 bps .
the ldpc - r code has a relatively small loss , i.e. , 0.12 bps in the symmetric case and 0.10 in the skewed one .
the ldpc - i code has a slightly smaller loss , i.e. , 0.09 bps in the symmetric case and 0.075 in the skewed one .
however , the dac still performs slightly better .
it should be noted that , while for ldpc and turbo codes the encoding is done only once thanks to rate - compatibility , for the dac multiple encodings are necessary , leading to higher complexity .
.performance comparison for variable - rate coding : mean and standard deviation of rate needed for lossless compression .
[ cols="^,^,^,^,^ " , ] [ tab : complexity ]
in the following we provide results for the symmetric dac . we consider two sources with balanced ( @xmath172 ) and unbalanced ( @xmath191 ) distribution with arbitrary rate splitting , and use @xmath86 . for fixed rate
, we set the total rate of @xmath12 and @xmath52 equal to 1.5 bps .
we consider two cases of rate splitting . in the first case
the rate is equally split ; we choose @xmath222 so as to achieve a rate of 0.75 bps for each source . in the second case
we encode @xmath12 at 0.6 bps and @xmath52 at 0.9 bps .
the performance of the symmetric dac is worked out for @xmath179 and @xmath184 . since symmetric dsc coders typically reconstructs each sequence either without any errors or with a large number of errors @xcite , we report the frame error rate ( fer ) instead of the residual ber , i.e. the probability that a data block contains at least one error after joint decoding . for each point , we simulated at least @xmath178 bits .
[ fig : symm_fr ] shows the results for the symmetric dac .
comparisons with other algorithms can be done based on the following remarks . in @xcite , a symmetric s - w coder is proposed employing turbo codes , which can obtain any rate splitting . in the case that one source is encoded without ambiguity , this reduces to the asymmetric turbo - based s - w coder we have employed in sect .
[ sec : results_asymm ] .
in @xcite it is reported that this algorithm achieves its best performance in the asymmetric points of the s - w region , while it is slightly poorer in the intermediate points .
therefore , in fig . [
fig : symm_fr ] we report the fer corresponding to the best turbo code shown in fig .
[ fig:1 ] for @xmath179 and @xmath184 , as this lower - bounds the fer achieved by @xcite over the entire s - w region .
moreover , we also report the fer achieved by irregular ldpc codes with block length @xmath193 @xcite .
the asymmetric algorithm in @xcite has been extended in @xcite to arbitrary rate splitting , showing that the performance is uniformly good over the entire s - w region .
finally , we also report the fer curve of the asymmetric dac for @xmath184 . , total rate = 1.5 bps ) .
dac : distributed arithmetic coding ; tc16s : 16-state turbo code with s - random interleaver ; ldpc - i : irregular ldpc codes from @xcite.,width=453 ] in fig .
[ fig : symm_fr ] , the results for symmetric coding are very similar to what has been observed in the asymmetric case .
the dac achieves very similar ber for @xmath179 and @xmath184 ; hence , the fer is smaller for @xmath179 .
the results are almost independent of the rate splitting between @xmath12 and @xmath52 , as can be seen by comparing the two rate - splitting cases as well as the asymmetric dac .
the turbo codes for @xmath179 and @xmath184 , and the irregular ldpc code , exhibit poorer performance than dac . for variable rate coding
, we consider the same two settings as in sect .
[ sec : res_vr ] , i.e. , block length @xmath179 , with probabilities @xmath172 and @xmath191 ; in the first case the conditional entropy has been set to 0.5 bps , while in the second case the joint entropy @xmath188 has been set to 1 bps .
the results are shown in fig .
[ fig : symm_vr ] . as can be seen , the performance of the symmetric dac is uniformly good over the entire s - w region , and is significantly better than turbo codes and ldpc codes . in particular ,
the dac suboptimality is between 0.03 - 0.06 bps , as opposed to 0.07 - 0.09 for the irregular ldpc code , and 0.14 - 0.21 for the turbo code .
it should be noted , however , that variable rate coding requires feedback , while the s - w bound is achievable with no feedback , with vanishing error probability as @xmath223 . in our simulations we re -
encode the sequence at increasing rates ( in steps of 0.01 bps ) , which represents the optimal dac performance if the encoder could exactly predict the decoder behavior . , and those in the bottom - left corner to @xmath191 .
dac : distributed arithmetic coding ; tc16s : 16-state turbo code with s - random interleaver ; ldpc - i : irregular ldpc codes from @xcite .
the solid curves represent the s - w bound.,width=453 ]
we have proposed dac as an alternative to existing dsc coders based on channel codes .
dac can operate in the entire s - w region , providing both asymmetric and symmetric coding .
dac achieves good compression performance , with uniformly good results over the s - w rate region ; in particular , its performance is comparable with or better than that of turbo and ldpc codes at small and medium block lengths .
this is very important in many applications , e.g. , in the multimedia field , where the encoder partitions the compressed file into small units ( e.g. , packets in jpeg 2000 , slices and nalus in h.264/avc ) that have to be coded independently . as for encoding complexity , which is of great interest for dsc
, dac has linear encoding complexity , like a classical ac @xcite .
turbo codes and the ldpc codes in @xcite also have linear encoding complexity , whereas general ldpc codes typically have more than linear , and typically quadratic complexity @xcite . as a consequence ,
the complexity of dac is suitable for dsc applications .
a major advantage of dac lies in the fact that it can exploit statistical prior knowledge about the source very easily .
this is a strong asset of ac , which is retained by dac .
probabilities can be estimated on - the - fly based on past symbols ; context - based models employing conditional probabilities can also be used , as well as other models providing the required probabilities .
these models allow to account for the nonstationarity of typical real - world signals , which is a significant advantage over dsc coders based on channel codes .
in fact , for channel codes , accounting for time - varying correlations requires to adjust the code rate , which can only be done for the next data block , incurring a significant adaptation delay .
moreover , with channel codes it is not easy to take advantage of prior information ; for turbo codes it has been shown to be possible @xcite , employing a more sophisticated decoder .
another advantage of the proposed dac lies in the fact that the encoding process can be seen as a simple extension of the ac process . as a consequence , it is straightforward to extend an existing scheme employing ac as final entropy coding stage in order to provide dsc functionalities .
liveris , z. xiong , and c.n .
georghiades , `` distributed compression of binary sources using conventional parallel and serial concatenated convolutional codes , '' in _ proc . of ieee data compression conference _ , 2003 , pp . 193202 .
a. majumdar , j. chou , and k. ramchandran , `` robust distributed video compression based on multilevel coset codes , '' in _ proceedings of thirty - seventh asilomar conference on signals , systems and computers _ , 2003 , pp . | distributed source coding schemes are typically based on the use of channels codes as source codes . in this paper
we propose a new paradigm , named distributed arithmetic coding " , which extends arithmetic codes to the distributed case employing sequential decoding aided by the side information .
in particular , we introduce a distributed binary arithmetic coder for the slepian - wolf coding problem , along with a joint decoder . the proposed scheme can be applied to two sources in both the asymmetric mode , wherein one source acts as side information , and the symmetric mode , wherein both sources are coded with ambiguity , at any combination of achievable rates .
distributed arithmetic coding provides several advantages over existing slepian - wolf coders , especially good performance at small block lengths , and the ability to incorporate arbitrary source models in the encoding process , e.g. , context - based statistical models , in much the same way as a classical arithmetic coder .
we have compared the performance of distributed arithmetic coding with turbo codes and low - density parity - check codes , and found that the proposed approach is very competitive .
distributed source coding , arithmetic coding , slepian - wolf coding , wyner - ziv coding , compression , turbo codes , ldpc codes . |
we acknowledge support through the nsf ( dmr-1305691 ) for measurements , and the gordon and betty moore foundation ( grant gbmf4420 ) , keck foundation , the nsf mrsec ( dmr-1420541 ) , and the doe bes ( de - fg02 - 00-er45841 ) for sample fabrication .
we thank r.n .
bhatt and l.w .
engel for illuminating discussions . | when the kinetic energy of a collection of interacting two - dimensional ( 2d ) electrons is quenched at very high magnetic fields so that the coulomb repulsion dominates , the electrons are expected to condense into an ordered array , forming a quantum wigner crystal ( wc ) .
although this exotic state has long been suspected in high - mobility 2d electron systems at very low landau level fillings ( @xmath0 ) , its direct observation has been elusive . here
we present a new technique and experimental results directly probing the magnetic - field - induced wc .
we measure the magneto - resistance of a bilayer electron system where one layer has a very low density and is in the wc regime ( @xmath0 ) , while the other ( probe ) layer is near @xmath1 and hosts a sea of composite fermions ( cfs ) .
the data exhibit commensurability oscillations in the magneto - resistance of the cf layer , induced by the periodic potential of wc electrons in the other layer , and provide a unique , direct glimpse at the symmetry of the wc , its lattice constant , and melting .
they also demonstrate a striking example of how one can probe an exotic many - body state of 2d electrons using equally exotic quasi - particles of another many - body state .
interacting 2d electrons subjected to high perpendicular magnetic fields ( @xmath2 ) and cooled to low temperatures exhibit a plethora of exotic quasi - particles and states @xcite . at @xmath1 landau level filling factor , for example ,
the interacting electrons capture two flux - quanta each and create new quasi - particles , the so - called cfs @xcite , which behave as essentially non - interacting particles .
the cfs offer an elegant explanation for the fractional quantum hall effect ( fqhe)@xcite . furthermore , because of the flux attachment , the effective magnetic field ( @xmath3 ) felt by the cfs vanishes at @xmath1 so that cfs occupy a fermi sea and exhibit fermi - liquid - like properties , similar to their zero - field electron counterparts @xcite .
in particular , in the presence of an imposed periodic potential , cfs resistance oscillates as a function of @xmath4 , where @xmath5 is the value of the external field at @xmath1 @xcite .
these oscillations are a signature of the commensurability of cfs quasi - classical cyclotron orbit diameter with an integer multiple of the period of the imposed potential , and their positions in @xmath3 reflect the symmetry and period of the potential .
another example of a collective state is the wc @xcite , an ordered array of electrons , believed to form at very small fillings ( @xmath6 ) when the coulomb repulsion between electrons dominates @xcite .
the wc , being pinned by the ubiquitous residual disorder , manifests as an insulating phase in dc transport @xcite , and exhibits resonances in its ac ( microwave ) transport which strongly suggest collective motions of the electrons @xcite .
so far , however , there have been no direct measurements of the wc order or its lattice constant .
here we present high magnetic field data in a bilayer electron system with unequal layer densities .
one layer has a very low density and is in the wc regime , while the adjacent layer is near @xmath1 and contains cfs ( fig .
1 ) . the cfs feel the periodic electric potential of the wc in the other layer and exhibit magneto - resistance maxima whenever their cyclotron orbit encircles certain integer number of the wc lattice points .
the positions of the maxima are consistent with a triangular wc and yield a direct measure of its lattice constant , while the disappearance of the resistance maxima as the temperature is raised signals the melting of the wc .
our samples were grown via molecular beam epitaxy and consist of two , 30-nm - wide , gaas quantum wells ( qws ) separated by a 10-nm - thick , undoped al@xmath7ga@xmath8as barrier layer .
the qws are modulation - doped with si @xmath9-layers asymmetrically : the bottom and top spacer layer thicknesses are 300 nm and 80 nm respectively .
this asymmetry leads to very different as - grown 2d electron densities in the qws ; the top - layer has a density of @xmath10 1.5 @xmath11 @xmath12 , much higher than the bottom - layer density @xmath13 @xmath12 . based on the growth parameters and our data for other , single - layer samples
, we estimate the top and bottom layers to have low - temperature mobilities of @xmath14 and @xmath15 @xmath16/vs , respectively . here
we report data for a hall bar sample ( 200 @xmath17 m wide and 800 @xmath17 m long ) with insn ohmic contacts .
the sample was thinned down to about 120 @xmath17 m and then fitted with an in back - gate which covers its entire back surface . applying a negative voltage bias ( @xmath18 ) to the back - gate
reduces the bottom - layer density .
the measurements were carried out in a dilution refrigerator with a base temperature of @xmath19 mk , and using low - frequency ( @xmath20 40 hz ) lock - in technique . the longitudinal resistance ( @xmath21 ) vs @xmath2 traces presented in fig .
2 capture the highlight of our findings .
data are shown only in the high-@xmath2 region of interest . in fig .
2(a ) , we present traces for the bilayer sample at different @xmath18 . in the top trace , the bottom - layer is completely depleted so that the trace represents @xmath21 for the top - layer .
this trace exhibits the usual characteristics seen in high - quality 2d electron systems ( 2dess ) : @xmath21 has minima at numerous odd - denominator fillings such as 3/5 , 4/7 , etc .
, indicating fqhe states , and is essentially featureless close to @xmath22 where it shows a broad , smooth minimum .
the features in the lowest trace , are also characteristic of a simple , single - layer 2des .
they , too , essentially reflect the properties of the top - layer , including its density and filling factors as a function of @xmath2 .
the bottom - layer , although present , is at a very small filling factor ( @xmath23 ) , has very high resistance , and appears not to contribute to @xmath21 at high fields near @xmath22 .
( based on measurements on single - layer samples , we esitmate the bottom - layer resistance to be several @xmath24 k@xmath25 at high _ b _ , about 100 times larger than the top - layer resistance . ) the middle trace in fig .
2(a ) which is taken just before the bottom - layer is depleted , however , is unusual as near @xmath22 there are small oscillatory features superimposed on the smooth background .
moreover , as illustrated in fig .
2(b ) , these features disappear when we warm up the sample .
as we discuss below , these oscillations near @xmath22 reflect the commensurability of the cyclotron orbits of the cfs in the top - layer with the periodic potential induced by a wc formed in the bottom - layer ( see fig .
1 ) . to test our hypothesis
, we carefully analyze the positions of the anomalous @xmath21 maxima observed near @xmath22 in the middle trace of fig .
2(a ) ; this trace is shown enlarged near @xmath22 in fig .
2(b ) ( lower trace ) . a quantitative prediction of the expected positions of the cf commensurability features induced by the wc potential requires accurate knowledge of both cf and wc layer densities , i.e. , both top- and bottom - layer densities ( @xmath26 and @xmath27 ) at high magnetic fields .
fortunately , we can experimentally determine these densities quite precisely .
our determination of @xmath28 is based on the magnetic field positions of the well - defined fqhe minima observed near @xmath22 . for @xmath27
, we subtract @xmath26 from the total density of the bilayer system ( @xmath29 ) , which we can also determine experimentally as described later in the manuscript . using @xmath26 for the density of cfs ( @xmath30 ) and @xmath27 for the density of wc electrons ( @xmath31 )
, we can predict the positions of commensurability features expected when the cf cyclotron orbit circles around @xmath32 wc lattice points ( fig .
1(b ) ) . assuming that each such orbit has equal distance to the nearest lattice points inside and outside the orbit ( fig .
1(b ) ) , the diameters for the @xmath32 orbits have values of @xmath33 in units of the period @xmath34 of the triangular lattice . in the case of _ electrons _ near zero magnetic field moving in a triangular anti - dot lattice , strong magneto - resistance maxima consistent with such cyclotron diameter values have indeed been reported @xcite .
the detailed nature and strength of the potential imposed by the wc on the cfs is of course unclear .
however , experimental data for both electrons @xcite and cfs @xcite indicate that , even in the case of a _ weak _ 2d periodic potential , the @xmath35 cf commensurability resistance peak is observed when the cyclotron orbit diameter equals @xmath34 . in fig .
2 , we use small triangles to mark the positions of the _ expected _ commensurability resistance maxima for the @xmath32 orbits .
these positions are based on the relation @xmath36 , where @xmath37 is the cf cyclotron radius and @xmath38 is the cf fermi wavevector .
we use @xmath39 for @xmath40 and @xmath41 for @xmath42 , namely we assume that @xmath30 is equal to the minority carriers density in the lowest landau level of the top layer ; see ref .
@xcite .
this leads to a slight ( @xmath43 5% ) asymmetry in the expected positions of the two cf resistance maxima for @xmath40 and @xmath42 . in fig
2(b ) , despite some clear discrepancies , there is overall a good agreement between the positions of the observed and expected resistance maxima .
we emphasize that the positions of the triangles are based on our _ measured _ top- and bottom - layer densities and do not rely on any fitting parameters .
note also the disappearance of the oscillatory features above @xmath44 mk , which we associate with the melting of the wc .
we believe that the data in fig . 2 provide evidence that the cfs in the top - layer are indeed dancing to the tune of a wc formed at low temperatures in the bottom - layer .
we now describe details of our determination of top- and bottom - layer densities . besides providing @xmath26 and @xmath27
, the data also reveal subtle interlayer charge transfers that occur near zero and at intermediate _
b_. in our experiments , for each value of @xmath18 , we first measure the shubnikov - de haas oscillations at very low fields ( @xmath45 t ) .
the fourier transform of these oscillations typically exhibits two peaks whose frequencies directly yield the _ low - field _ densities of the bottom and top layers ( @xmath46 and @xmath47 ) .
the sum of @xmath46 and @xmath47 gives @xmath29 , which we assume is independent of the magnetic field . in fig .
3 , we show the measured @xmath29 ( triangles ) and @xmath47 ( squares ) vs @xmath18 .
as expected , @xmath29 changes linearly with @xmath18 ( with a slope that is consistent with the capacitance of our sample ) , and @xmath47 equals @xmath29 for @xmath48 v when the bottom - layer is completely depleted .
however , as @xmath18 is decreased from zero , @xmath47 _ increases _ initially before decreasing at smaller @xmath18 .
this is because of the negative compressibility of an interacting 2des , and can be explained quantitatively using self - consistent ( poisson - schroedinger equations ) calculations that include exchange and correlation energies via local - density - approximation @xcite . in fig .
3 , we show the results of such calculations for the parameters of our bilayer electron sample , indicating good agreement between the measured and calculated @xmath47 ( black squares and curve in fig .
3 , respectively ) .
surprisingly , however , in fig .
3 we find that the top - layer density at high magnetic fields ( @xmath26 ) is significantly larger than its low - field value @xmath47 .
our data imply that there is an _
additional _ charge transfer which is of the order of @xmath49 @xmath12 and occurs from the bottom- to the top - layer at intermediate _
b_. the origin of this charge transfer is the mismatch between the ground - state energies of the two layers that have unequal densities but are in thermal equilibrium , and the requirement that the lowest landau levels of the two layers have to align at high magnetic fields when both layers are in the extreme quantum limit . in a simple , capacitive model ,
the magnitude of this charge transfer is @xmath50 , where @xmath51 is the difference between the layer densities , @xmath52 is the distance between the charge distribution peaks of the two layers , and @xmath53 is the gaas / alas dielectric constant . while this simple model gives the correct order of magnitude for the observed charge transfer , it does not quantitatively explain its size nor its dependence on @xmath18 .
the details of this subtle charge transfer are very interesting on their own as they reflect a delicate balance between interaction at intermediate _
b _ and the capacitive energy associated with interlayer charge transfer .
we emphasize that this charge transfer has no bearing on our analysis of fig .
2 data or our conclusions because our determination of the cf and wc layer densities is based on direct measurements of @xmath26 ( from the positions of fqhe minima near @xmath22 ) and @xmath29 ( from shubnikov - de haas oscillations at low field ) , and the reasonable assumption that @xmath54 . returning to the commensurability data , in fig .
2(c ) we show additional traces taken near the depletion of the bottom - layer . for each trace
we indicate @xmath27 and also mark the expected positions of commensurability resistance maxima based on @xmath26 and @xmath27 for that trace . in a narrow range of @xmath27 ,
oscillatory features are seen . for these traces ,
the position of the expected @xmath35 maximum matches remarkably well the strong peak observed to the left of @xmath22 ( @xmath55 ) .
it is noteworthy that experiments on commensurability of cfs in an anti - dot lattice also show that the resistance maximum is more pronounced for @xmath56 compared to @xmath57 @xcite .
the reason for this asymmetry is , however , not known .
similar to the data of fig .
2(b ) , in some traces of fig .
2(c ) there are also hints of weak resistance maxima , closer to @xmath22 compared to the @xmath35 commensurability maxima .
these could signal the commensurability of larger cf cyclotron orbits ( @xmath58 and 7 ) with the wc periodic potential . in fig .
2(c ) traces , for @xmath59 @xmath12 , there are also strong @xmath21 maxima observed to the left of the @xmath35 triangles ( see up - arrows in fig .
2(c ) ) . we do not know the origin of these anomalous maxima ; they might indicate the resonance of a ( small ) cf cyclotron orbit that fits _ inside _ three adjacent lattice points .
we note that anomalous shoulders at fields highers than the @xmath35 resistance maximum are also seen in some commensurability data in anti - dot samples ; see , e.g. , traces 1 and 2 * in fig . 2 of ref .
@xcite .
so far we have assumed that the wc attains a triangular lattice .
this is indeed generally conjectured for a fully spin - polarized wc @xcite . near the 114-mk - trace in fig .
2(b ) , we have included marks ( diamonds ) for the expected positions of resistance maxima if the wc lattice were square instead of triangular .
it appears that the positions of the triangles agree with the experimental data better than those of the squares .
finally , the temperature dependent data of fig .
2(b ) indicate the disappearance of the oscillatory resistance features above @xmath60 mk , signaling a melting of the wc .
a melting temperature of @xmath61 mk is comparable to but somewhat smaller than those reported from various measurements @xcite .
it is worth noting , however , that the wc density in our sample ( @xmath62 1.7 @xmath63 @xmath12 at @xmath64 v ) is about an order of magnitude smaller than the densities in previous measurements . also , it is possible that the cfs in the top - layer , which is only @xmath65 nm away , screen and reduce the coulomb interaction between the electrons in the wc layer @xcite .
such screening would lead to a lower wc melting temperature . before closing , we remark on data we obtained from large ( 4 mm x 4 mm ) van der pauw samples .
such samples exhibit more pronounced resistance maxima near @xmath22 , some of which can be assigned to the commensurability of top - layer cf cyclotron orbits with the potential of a wc formed in the bottom - layer .
the wc layer density deduced from this assignment , however , is typically much smaller ( by a factor of up to 4 ) than the bottom - layer density we obtain from our layer density measurements .
the discrepancy might stem from the non - uniformity of the layer densities in van der pauw samples whose area is about 100 times larger than the hall bar samples presented here .
we note that density uniformity is especially important in our experiments because the wc layer density has to be extremely small ( of the order of @xmath66 @xmath12 ) so that the cf commensurability condition is satisfied very near @xmath22 , i.e. , far from the fqhe minima .
our results offer a direct look at the microscopic structure of the magnetic - field - induced wc in 2dess .
they also raise several intriguing questions .
for example , what is the exact shape of the periodic potential modulation that the wc layer imposes on the cfs ?
a related question is the optimum interlayer distance that would lead to the most pronounced @xmath21 commensurability maxima .
more generally , the weak periodic potential imposed by the wc and the homogeneous _ effective _ magnetic field @xmath3 can conspire to also lead to a fractal energy diagram ( the hofstadter s butterfly ) @xcite for the of cfs , similar to what has been studied for _ electrons _ in an externally imposed periodic potential @xcite .
finally , our technique might find use in studying other possibly ordered , broken - symmetry states of 2dess .
for example , we envision measurements where the cfs in one layer are used to probe , in an adjacent layer , the anisotropic phases that are observed at half - filled landau levels with high index @xcite .
these anisotropic phases are believed to signal many - body states , consisting of stripes of electrons with alternating density ( filling ) .
if the stripes are periodic , they would generate a one - dimensional periodic potential in the nearby cf layer , and the cfs should exhibit commensurability features . |
irc + 10 216 (= afgl 1381 = cw leo ) is the best - studied carbon star , because it is intrinsically bright and nearby .
one aspect of study has traditionally been the properties of the circumstellar shell : e.g. , what is the mass loss rate , has the mass loss rate changed with time , what kind of chemistry takes place , what is the geometry of the shell ? in two recent papers we have made an extensive study of the circumstellar dust shell of irc + 10 216 ( groenewegen 1996 , 1997 , hereafter g97 ) . a much larger set of observational constraints was used than in any previous study .
this allowed the derivation of the present - day dust mass loss rate , the ratio between the luminosity at maximum light divided by the distance squared , the dominant grain size , the effective temperature and the inner dust radius . in another paper
we have mapped the 1.3 mm continuum emission around the star ( groenewegen et al .
it was detected out to 50at a noise level of 3@xmath13 or better .
there is evidence for phases of enhanced mass loss , but a quantitative analysis is hampered by the significant contribution of molecular lines to the continuum emission , and the uncertainty in the radial distribution of that contribution . another way to study
the circumstellar shell is by observing the gas component .
a good tracer of the ( h@xmath12 ) gas is the co molecule , and numerous co observations exist of irc + 10 216 ( see the catalog of loup et al .
( 1993 ) for a compilation ) .
a fair number of detailed models have been proposed to explain various co observations .
the pioneering work was done by kwan & hill ( 1977 ) . for an assumed distance of 200 pc they deduced a mass loss rate of 2 @xmath10 10@xmath14 m@xmath2 yr@xmath7 , a co abundance relative to h@xmath12 of about 8 @xmath10 10@xmath15 ( both uncertain to a factor of 2 ) , and a co/@xmath1co isotope ratio of 35 @xmath5 7 . in a second paper
kwan & linke ( 1982 ) deduced a mass loss rate of 4 @xmath10 10@xmath14 m@xmath2 yr@xmath7 for a distance of 200 pc and a co abundance of 6 @xmath10 10@xmath15 .
sahai ( 1987 ) finds , based on the modeling of high - j lines , that the temperature in the inner region is much larger than that predicted by the kwan & linke model , and that the mass loss rate in the inner 6 region is 3.2 @xmath10 10@xmath14 m@xmath2 yr@xmath7 and 50% higher in the extended envelope , assuming a distance of 300 pc and a co abundance of 8 @xmath10 10@xmath15 .
truong - bach et al . ( 1991 ) also find evidence for a hot inner core : a mass loss rate of 2.5 @xmath10 10@xmath14 m@xmath2 yr@xmath7 in the inner 4.2 and a mass loss rate of 4 @xmath10 10@xmath14 m@xmath2 yr@xmath7 in the outer region ( for a distance of 200 pc ) .
kastner ( 1992 ) found a best fit to some co data for a distance of 150 pc and a mass loss rate of 2 @xmath10 10@xmath14 m@xmath2 yr@xmath7 . in a recent paper , crosas & menten ( 1997 ) also presented j = 6 - 5 @xmath0co and @xmath1co data , and combined those with other data from the literature , and modelled the data . in this respect
the philosophy behind that paper is similar to ours .
the major differences with that paper are that the effective absorption coefficient is taken into account in a consistent way ( see below and sect .
7.2 ) and that the parameter space is scanned consistently no a - priori fixing any parameter . in crosas & menten ,
the dust - to - gas ratio is _ fixed _ at a value of 0.01 .
they find a good fit using a distance of 150 pc , a mass loss rate of 3.25 @xmath10 10@xmath14 m@xmath2 yr@xmath7 , a co abundance of 6 @xmath10 10@xmath15 , and a dust - gas momentum transfer rate @xmath16 of 0.025 for the inner region ( @xmath17 cm ) , and 0.018 for the outer region .
they conclude there is no need for a hot inner core . in all the model calculations mentioned above the effective absorption coefficient @xmath16
is assumed to be a free parameter .
@xmath16 describes the momentum transfer efficiency between the gas and the dust .
however , we claim , and this is a key point in this paper , that effective absorption coefficient @xmath16 is not a free parameter but is related to the properties of the dust .
this had been realised before but it requires the simultaneous fitting of the spectral energy distribution ( sed ) and the co data in order to estimate @xmath16 in a self - consistent way .
this approach was taken in the fitting of two oh / ir stars in groenewegen ( 1994b ) and is proposed here for irc + 10 216 .
the outline of the paper is as follows . in sects . 2 and 3
the model is briefly described . in sect .
4 the @xmath0co and @xmath1co j = 6 - 5 observations as well as the other co data to which the models are compared are presented . in sect .
5 the results of the model calculations are described .
so - called visibility curves , to be compared with future interferometric observations , are presented in sect .
the paper ends with discussion and conclusions in sect .
[ cols= " < , < , < , < , < , < " , ] the uncertainty in the mass loss rate and dust - to - gas ratio derived from the dust shell is about 8% and 6% due to uncertainties in the assumed mass ( varied to 0.6 m@xmath2 ) and @xmath18 ( varied to 0 km s@xmath7 ) . in the ideal situation
the mass loss rates and dust - to - gas ratios from the two methods would be identical to indicate a perfectly consistent model .
the best overall agreement based on this analysis is achieved for a luminosity of 15 000 l@xmath2 . given all these considerations
it is likely that the luminosity of irc + 10 216 is between 10 000 and 15 000 l@xmath2 and its distance is between 110 and 135 pc . the present - day mass loss rate is ( 1.5 @xmath5 0.3 ) @xmath6 m@xmath2 yr@xmath7 and the gas - to - dust ratio is about 700 @xmath5 100 .
the dust opacity at 60 @xmath8 m is found to be of order 250 @xmath9gr@xmath7 .
the co abundance is 1.1 @xmath10 10@xmath11 relative to h@xmath12 .
compared to the previously proposed models mentioned in the introduction this represents a somewhat lower mass loss rate , partly because these other models adopted larger distances as well in many cases .
the co abundance is larger than in most previous models , but in agreement with theoretical expectations .
the gas - to - dust ratio we find is larger that the canonical value of 200 often quoted . on the other hand
there have been other recent studies that found larger gas - to - dust ratios for carbon stars as well , e.g. a gas - to - dust ratio of 550 for irc + 10 216 for a distance of 135 pc ( ivezi & elitzur 1996 ) , a value of 1230 for irc + 10 216 ( winters et al .
1997b ) , a value of 530 for afgl 3068 ( winters et al .
note that in the model of crosas & menten ( 1997 ) the gas - to - dust ratio is fixed at 100 , and they mention that a larger value would result in lines that are consistently too low .
cernicharo j. , barlow m.j .
, gonzlez - alfonso e. , et al . , 1996 ,
a&a 315 , l201 crosas m. , menten k.m . , 1997 ,
apj 483 , 913 fleischer a.j . ,
gauger a. , sedlmayr e. , 1995 , a&a 297 , 543 groenewegen m.a.t . , 1994a , a&a 290 , 531 ( g94 ) groenewegen m.a.t .
, 1994b , a&a 290 , 544 groenewegen m.a.t . , 1996 ,
a&a 305 , l61 groenewegen m.a.t . , 1997 ,
a&a 317 , 503 ( g97 ) groenewegen m.a.t . , de jong t. , 1994 , a&a 282 , 127 groenewegen m.a.t . , ludwig h
, 1998 , a&a , accepted groenewegen m.a.t . , whitelock p.a . , 1996 , mnras 281 , 1347 groenewegen m.a.t . , baas f. , de jong t. , loup c. , 1996 , a&a 306 , 241 groenewegen m.a.t . , van der veen w.e.c.j . , lefloch b. , omont a. , 1997 , a&a 322 , l21 groenewegen m.a.t . , whitelock p.a . ,
smith c.h . , kerschbaum f. , 1998 , mnras 293 , 18 habing h.j . , tignon j. , tielens a.g.g.m . , 1994 ,
a&a 286 , 523 hills , r. , 1996 , the jcmt newsletter 6 , 32 huggins p.j . , olofsson h. , johansson l.e.b .
, 1988 , apj 332 , 1009 ivezi z. , elitzur m. , 1995 , apj 445 , 415 ivezi z. , elitzur m. , 1996 , mnras 279 , 1011 kahane c. , cernicharo j. , gomez - gonzalez j. , gulin m. , 1992 , a&a 256 , 235 kastner j.h . , 1992 ,
apj 401 , 337 knapp g.r . , morris m. , 1985 , apj 292 , 640 knapp g.r .
, chang k.m . , 1985 , apj 293 , 281 koepf g.a . , buhl d. , chin g. , et al .
, 1982 , apj 260 , 584 kwan j. , hill f. , 1977 , apj 215 , 781 kwan j. , linke r.a . , 1982 , apj 254 , 587 kwan j. , webster z. , 1993 , apj 419 , 674 le bertre t. , 1992 , a&as 94 , 377 loup c. , forveille t. , omont a. , paul j.f .
, 1993 , a&as 99 , 291 mamon m. , glassgold a.e .
, huggins p.j . , 1988 , apj 328 , 797 morris m. , lucas r. , omont a. , 1985 , a&a 142 , 107 netzer n. , elitzur m. , 1993 , apj 410 , 701 olofsson h. , eriksson k. , gustafsson b. , carlstrm u. , 1993 , apjs 87 , 267 sahai r. , 1987 , apj 318 , 809 stanek k.z . , knapp g.r .
, young k. , phillips t.g . , 1995 , apjs 100 , 169 truong - bach , morris d. , nguyen - q - rieu , 1991 , a&a 249 , 435 wang y. , jaffe d.t . ,
graf u.u . , evans ii n.j . , 1994 ,
apjs 95 , 503 wannier p.g . , sahai r. , 1986 , apj 311 , 335 wannier p.g . , sahai r. , andersson b .- g . , johnson h.r . , 1990 ,
apj 358 , 251 wattenbach r. , krgel e. , rser h.p .
, et al . , 1988 , a&a 202 , 133 williams p.g . , white g.j . , 1992 ,
a&a 266 , 365 winters j.m .
, dominik c. , sedlmayer e. , 1994a , a&a 288 , 255 winters j.m . , fleischer a.j . , gauger a. , sedlmayer e. , 1994b , a&a 290 , 623 winters j.m . ,
fleischer a.j . ,
le berte t. , sedlmayer e. , 1997a , a&a 326 , 305 winters j.m .
, fleischer a.j .
, sedlmayer e. , 1997b , in : `` science with the vlt interferometer '' , springer verlag , berlin , p. 403 | @xmath0co and @xmath1co j = 6 - 5 observations of irc + 10 216 with the jcmt are presented .
a spherically symmetric radiative transfer code is used to model these and other co observations of the carbon star irc + 10 216/cw leo . compared to previously published model calculations
a much larger set of observational data is used as constraints ; on - source @xmath0co and @xmath1co j = 1 - 0 up to j = 6 - 5 and mapping data taken with various telescopes , most of which are obtained from the literature .
the gas temperature in the envelope is calculated taking into account heating and cooling .
the heating by dust - gas collisions and various other parameters ( such as luminosity over distance squared ) are constrained from our previous modeling of the circumstellar dust shell .
photoelectric heating is taken into account .
a grid of models is calculated with the following parameters : luminosity ( in the range 10 000 30 000 , in steps of 5 000 l@xmath2 ) , mass loss rate , dust - to - gas ratio , dust opacity and co abundance . for each of the considered luminosities a good fit to the on - source data
can be obtained . a comparison with co j = 1 - 0 data obtained from the literature points towards a preferred luminosity of 10 - 15 000 l@xmath2 . notwithstanding the overall good agreement
, there remain discrepancies .
the different observed @xmath0co j = 3 - 2 observations appear to be always larger than the model predictions .
the observed @xmath1co j = 6 - 5 is almost flat - topped , while the model gives a slight double - peaked profile .
there is a large discrepancy with the single existing @xmath0co j = 7 - 6 observation .
the best fitting models ( for each of the considered luminosities ) can not accommodate the more extended emission seen in the mapping data .
this is not due to an underestimate of the photoelectric effect . to fit the data for radial offsets @xmath350 the mass loss rate must be a factor of 5 higher in the outer envelope . because the various sets of data for offsets 150 are not consistent with each other
it is unclear if the enhancement in the mass loss rate extends beyond that radius .
visibility curves are calculated for comparison with future interferometric observations .
these appear to be insensitive to luminosity and mass loss variations but should be good tracers of the geometry of the co shell .
a comparison is made between the mass loss rates and dust - to - gas ratios derived from the co modelling and those derived from our previous dust modelling .
to do this we make use of the relation @xmath4 , which is valid for radiation pressure driven outflows .
the best agreement is obtained for the model with 15 000 l@xmath2 .
this agrees well with the luminosity range 7 700 - 12 500 l@xmath2 based on the period - luminosity relation for carbon miras . in summary
, we conclude that the likely luminosity of irc + 10 216 is between 10 000 and 15 000 l@xmath2 and that its distance is between 110 and 135 pc .
the present - day mass loss rate is ( 1.5 @xmath5 0.3 ) @xmath6 m@xmath2 yr@xmath7 and the gas - to - dust ratio is a 700 @xmath5 100 .
the dust opacity at 60 @xmath8 m is found to be of order of 250 @xmath9gr@xmath7 .
the co abundance is 1.1 @xmath10 10@xmath11 relative to h@xmath12 . |
fluids that are brought into contact with a disordered porous matrix drastically change their physical properties . over the past two or three decades numerous experimental and theoretical investigations
have been dedicated to studying this phenomenon @xcite .
the remarkable efforts that have been undertaken to obtain a deeper understanding of the underlying mechanisms are not only motivated by academic interest .
fluids confined in disordered porous environments play a key role in a wide spectrum of applied problems ranging from technological applications over chemical engineering to biophysical systems .
thus , a deeper understanding of _ why _ fluids change their properties under such external conditions is of great practical relevance . from a theoretical viewpoint ,
a quantitative description of such systems is a formidable challenge for several reasons .
for one , it is difficult to reliably and realistically represent the matrices found in experiments or technological applications ( such as aerogels or vycor ) .
having chosen some model for the confinement , an even more demanding task is to formulate a theoretical framework that is able to appropriately describe the interplay of connectivity and confinement of the pores pertinent to the matrix . even after reducing such system to the so - called quenched - annealed ( qa ) model , where for simplicity the matrix is assumed to be an instantaneously - frozen configuration of an equilibrated fluid ,
an appropriate theory is highly intricate .
madden and glandt @xcite , and later given and stell @xcite , derived a formalism based on statistical mechanics that considers the system to be a peculiar mixture with the matrix being one of the components . within their framework ,
given and stell derived ornstein - zernike - type integral equations ( the so - called replica ornstein - zernike equations , roz ) that in combination with a suitable closure relation allow to determine the static correlators and the thermodynamic properties of a qa system @xcite . their formalism is based on a twofold thermodynamic averaging procedure : one over all possible fluid configurations for a fixed matrix realization , the other one over all possible matrix configurations .
this double - averaging procedure is adapted to the letter in computer simulation studies of fluids confined in disordered porous materials ; in terms of computing time this is a demanding enterprise . until few years ago most theoretical and simulation studies focused on the _ static _ properties of qa systems such as structure and phase behavior .
in contrast , comparatively little effort had been devoted to the _ dynamic _ properties .
the main obstacles inhibiting such investigations were , on one side , the lack of a suitable theoretical framework that would have allowed to reliably evaluate the dynamic correlation functions of the fluid particles , and , on the other hand , limitations in computational power for simulations when explicitly performing the above - mentioned double - averaging procedure . to the best of our knowledge , the first simulation studies dedicated to the dynamic properties of fluids in disordered porous confinement were performed by gallo and rovere @xcite , followed shortly after by kim @xcite . in this context
it is worth mentioning similar investigations on the lorentz gas model @xcite and on binary mixtures in which particles are characterized by a large disparity in size @xcite or mass @xcite , i.e. , systems that can be viewed to be closely related to qa systems . over the past years
, major breakthroughs have been achieved in describing theoretically the dynamic properties of fluids in confined in porous media .
in particular , use was made of mode - coupling theory ( mct ) @xcite and of the self - consistent generalized langevin equation ( scgle ) theory @xcite , both of which rely on inferring the dynamics of a system from its static structure .
mct has played over the last decades a central role in describing the dynamic properties of fluids @xcite , including in particular the glass transition .
krakoviack succeeded in deriving mct - type equations for the dynamic correlation functions , which require as an input the static correlation functions of the annealed fluid , which in turn are obtained from the roz equations .
subsequently , krakoviack solved these equations for the specific case of a hard - sphere fluid confined in a disordered matrix of hard - sphere particles of equal size and evaluated single - particle and collective correlation functions of the system . based on this knowledge ,
a kinetic diagram was traced out in the parameter space spanned by the packing fraction of the fluid and that of the matrix , indicating the regions in which the system is in an arrested or in a nonarrested state .
the resulting kinetic diagram is very rich and quite surprising : the regions of arrested and nonarrested states are separated by a line along which at small matrix packing fractions type - b transitions and at intermediate matrix packing fractions type - a transitions occur . as a peculiar feature of this line ,
a re - entrant glass transition is predicted for intermediate matrix and small fluid packing fractions .
additionally , in the region of nonarrested states a continuous diffusion - localization transition is observed .
however , we emphasize that the combined roz+mct framework predicts _ ideal _ transitions , whereas in experiment or simulation transitions will always be smeared out . throughout this contribution
we employ the type - a / type - b terminology that is conventionally used to characterize the behavior of a dynamic correlator as an ideal transition occurs .
this terminology refers to the decay pattern of the correlator as well as to the behavior of its long - time limit upon crossing a dynamic transition . in a `` type - a '' transition
a correlator relaxes in a single step , and its long - time limit may assume arbitrarily small nonzero values . in a transition of `` type - b , '' on the other hand , a correlator decays in a distinct two - step pattern ; upon crossing the transition the second decay step is delayed to infinity and the long - time limit of the correlator jumps from zero to a nonzero value .
the type - b behavior is well - known from bulk glass formers @xcite and is usually attributed to a `` cage effect '' imposed upon fluid particles by their neighbors .
this remarkably rich phase behavior has motivated simulation studies which as a consequence of the considerable increase in available computation power have meanwhile come within reach .
short accounts have been given in @xcite and @xcite .
in this contribution we elaborate on our investigations on the dynamic behavior of a fluid confined in a disordered porous matrix within the qa picture . for simplicity , and for congruence with the theoretical studies @xcite , we consider a fluid of hard spheres confined in a disordered matrix formed by hard spheres of the same size .
investigations are based on event - driven molecular dynamics simulations .
within the two - dimensional parameter space we focus mainly on three representative pathways to put the theoretical predictions to a thorough test : ( i ) a path at constant low matrix packing fraction , probing the existence of the continuous type - b transition , ( ii ) a path at constant intermediate matrix packing fraction , probing the discontinuous type - a transition and the re - entrant scenario , and ( iii ) a path at intermediate , constant fluid packing fraction , probing the diffusion - localization transition .
results are summarized in kinetic diagrams , where we consider complementing our previous communication different criteria to define dynamical arrest .
we compare our simulation results with the predictions of mct , which , however , is not straightforward due to the ideal nature of the latter .
the paper is organized as follows : in we present the model and our simulation method , transferring the rather lengthy description of the system setup to .
results are compiled in : we start with the static structure of the system and then discuss various methods of how to define slowing down and the ensuing kinetic diagrams . in the subsequent three subsections we discuss the intermediate scattering functions and the mean - squared displacement along the three aforementioned paths through parameter space . the final subsection of shall deal with the van hove correlation function in order to discuss trapping . in we
extend the analysis of the intermediate scattering functions to discuss the arrest scenarios in more detail .
the paper closes with concluding remarks alongside with an outlook on future work required to settle questions yet open .
the aforementioned is succeeded by in which we elaborate on the issue of the effects of system size on the dynamical correlation functions .
[ width = ] figure01.eps for the purpose of this work we opted for the `` quenched - annealed '' ( qa ) protocol since it is well - defined , well - tractable by theory , and involves only a small number of parameters . in this protocol both the porous confinement and the confined fluid are represented by a set of particles , conventionally called `` matrix '' and `` fluid '' particles or borrowing from the typology of metallurgy as `` quenched '' and `` annealed '' component . in this , qa systems bear resemblance to mixtures ; however , the matrix particles are pinned at particular positions and act as obstacles for the fluid particles . in this work ,
we exclusively consider systems containing one matrix and one fluid species ; quantities pertaining to those species will be designated with the subscripts `` m '' and `` f , '' respectively . adopting the concept of immobile obstacle particles , there are still numerous schemes for their arrangement @xcite .
the qa protocol specifies the positions of the matrix particles to be frozen from an equilibrium one - component fluid ; the fluid particles merely move within this matrix and do not exert influence upon it .
notably , the fluid particles may also dwell in `` traps''finite spatial domains entirely bounded by infinite potential walls formed by the confining matrix .
since for this work we are interested in a minimal approach to the dynamics of a fluid moving in disordered confinement , we decided to exclusively employ hard - sphere ( hs ) interactions among the particles .
this deprives the system of any inherent energy scale , rendering the only relevant system attributes to be of geometrical nature .
we further reduced the parameter space by requiring all particles involved to be of the same ( monodisperse ) diameter @xmath2 .
this introduces @xmath2 as the unit of length used throughout this study , and leaves the system to be governed by mere packing fractions , namely the matrix packing fraction @xmath0 and the fluid packing fraction @xmath1 . as mentioned before ,
this particular choice of system was examined in @xcite . a snapshot of a typical hs - qa system at large matrix density is shown in .
in this work we study qa systems utilizing molecular simulations . since we are primarily interested in dynamical features , we applied molecular dynamics ( md ) simulations to track the physical time evolution of a system .
all md computations were performed using the event - driven algorithm described in @xcite ; modifications to that algorithm were essentially limited to fixing particles in space in order to adapt to the qa protocol .
( notably , this adjustment invalidates the conservation of momentum and angular momentum . ) as is conventional , we employed periodic ( toroidal ) boundary conditions and the minimum - image convention in order to mimic an infinite system .
in addition to the usual thermodynamic averaging , @xmath3 , the qa protocol imposes another averaging procedure over _ disorder _ , @xmath4 , so as to restore homogeneity and isotropy . depending on the state point @xmath5 and the number of particles in the system
, we averaged our data over at least ten matrix configurations ; for selected state points , we increased that number to as much as @xmath6 .
a nontrivial task is to find an initial system configuration given some state point . in this work
, we are interested in systems in which the @xmath7 fluid particles comprise most of the space that is left accessible to them by the @xmath8 immobile matrix particles i.e . ,
_ high - density _ systems .
generating instances of the porous confinement is effortless , since for all state points of interest , @xmath0 is situated well within the fluid regime when used for a one - component system of mobile particles .
the problem rests in determining permissible initial positions for the @xmath7 fluid particles . creating an amorphous high - density configuration of bulk monodisperse hard - sphere particles already represents an interesting task , with various elaborate algorithms having been developed to accomplish this task @xcite
however , none of these algorithms are suited for adaption to the setup of hs - qa systems .
this unfortunate finding prompted us to devise a custom routine which is described in detail in .
after successful setup of a system instance , the simulation time was set to @xmath9 and an attempt to equilibrate the system using the md routine was made . the decision whether or not the particular system realization was considered equilibrated was based on the mean - squared displacement @xmath10 where @xmath11 is the location of a tagged ( superscript `` @xmath12 '' ) fluid particle at time @xmath13 .
note that the above definition is the one used in and hence involves an average over matrix configurations ; of course this average has to be omitted when using @xmath14 to characterize a single system realization .
also , note that the notion of a `` tagged particle '' does _ not _ imply the presence of an additional tracer particle .
we defined a system realization to be in equilibrium if @xmath15 was fulfilled .
we allowed at least @xmath16 time units for the system to equilibrate ; for certain state points we extended @xmath17 to as much as the tenfold value . the conventional unit of time @xmath18 used throughout this work is based upon the mass @xmath19 of the fluid particles , their diameter @xmath2 ( see above ) , and their temperature @xmath20 @xcite .
data were only compiled after equilibration was attempted , using the final configuration of the equilibration run as initial configuration of the data run .
generally , the final time for data collection , @xmath21 , was chosen to be equal to the final time of the equilibration attempt , @xmath22 ; in particular in some nonequilibrated systems we also used @xmath23 to investigate particular features of the system . for most state points
, we chose @xmath24 to be the total particle number and/or @xmath7 to accurately approximate the desired fraction @xmath25 resulted in slight deviations from the standard @xmath26 . ] . in order to retain reasonable statistics even close to @xmath27
, we increased @xmath28 to as much as @xmath29 for elevated @xmath0 , such that @xmath30 for all state points .
[ width = ] figure02.eps [ width = ] figure03.eps the most obvious properties to inspect in our systems are simple static structural properties ; this investigation was already carried out in great detail in @xcite . in the present context static properties
are of use for a different reason : since the particles were chosen to be of monodisperse diameter @xmath2 , crystallization might take place in the case of dilute matrices .
inspection of the radial distribution function @xmath31 and the static structure factor @xmath32 provides a means to determine whether or not this is the case .
in fact , it is sufficient to investigate fluid - fluid correlations , @xmath33 and @xmath34 , since only for @xmath35 systems are prone to crystallization . therefore , the index `` @xmath36 '' will henceforth be dropped .
we first consider the static structure factor @xmath37 where @xmath38 , and @xmath39 denotes the location of fluid particle @xmath40 .
displays this function for moderate ( @xmath41 ) and for low ( @xmath42 ) matrix packing fractions , i.e. , for systems particularly prone to crystallization . clearly , all features of @xmath32 evolve smoothly upon increasing @xmath1 , suggesting that no phase transition takes place .
moreover , the absence of sharp peaks for all state points indicates that no simple crystalline long - range order is present .
related arguments hold for the radial distribution function @xmath43 where @xmath44 is the system volume .
shows this function for the same matrix packing fractions as above ( @xmath45 ) . as for @xmath32 , the features of @xmath31 develop continuously , which again supports the notion that a phase transition is absent .
additional peaks in @xmath31 at specific positions would be a hallmark of crystallization . for fcc- or bcc - like short - range order
, the most pronounced additional peaks would be located at @xmath46 or @xmath47 , where @xmath48 and @xmath49 is the volume fraction of close - packed monodisperse hard spheres .
for instance , for @xmath50 and @xmath51 , crystallites would cause a peak at @xmath52 or @xmath53 .
clearly , peaks at these positions are absent in . on the other hand , in both panel ( a ) and
( b ) an additional peak at @xmath54 emerges upon increasing @xmath1 .
however , this peak is a well - known feature of glass - forming systems @xcite and thus provides further evidence for the absence of crystallization .
we conclude that crystallites , albeit possibly temporarily present , do not play a major role in the systems under investigation .
[ width = ] figure04.eps [ width = ] figure05.eps many features of the system under investigation can be understood by means of kinetic diagrams in which the `` state '' of a dynamic property is indicated in the plane spanned by @xmath0 and @xmath1 .
in an earlier work @xcite we presented a kinetic diagram based on the mean - squared displacement @xmath14 . in order to classify the state of the system ,
we chose a mean - squared displacement @xmath55 and a simulation time @xmath16 ; if the system obeyed @xmath56 then it was considered `` nonarrested , '' and otherwise `` arrested . ''
the so - constructed kinetic diagram qualitatively confirmed the behavior of the single - particle dynamics of the system at hand as predicted by mct @xcite . the arrest line determined from this criterion
is indicated in .
in we present a kinetic diagram of the system under investigation based on the self - intermediate scattering function @xmath57 where @xmath58 is the density of a tagged ( superscript `` @xmath12 '' ) fluid particle in @xmath59 space after some time @xmath13 has passed .
the symbols in the figure indicate the times @xmath60 needed for @xmath61 to decay below the value @xmath62 .
the value @xmath63 was chosen to be close to the main peak in @xmath32 in a typical high - density qa system for instance , for @xmath50 and @xmath64 ( cf . )
the first peak in @xmath32 is located at @xmath65 ; for @xmath66 and @xmath67 it is found at @xmath68 . if we let @xmath69 , discriminating whether for a state point @xmath70 ( `` arrested '' ) or @xmath71 ( `` nonarrested '' ) yields the thick solid ( blue ) line in .
obviously , the latter is only slightly different from the thick dotted ( green ) line that is obtained from the @xmath14 criterion described above .
this is not unexpected since both @xmath72 and @xmath14 are single - particle properties , and in the gaussian limit it is even @xmath73 @xcite .
thus , the validity of the approach chosen in @xcite is confirmed . in order to complement the kinetic diagrams based upon single - particle properties , in we present for the first time a kinetic diagram for a collective property of the system . for comparison with the theory developed in @xcite we chose to operate with the connected part of the intermediate scattering function @xmath74 where @xmath75 is the fluctuating part of @xmath76 ( cf . ) . very much like for the self - intermediate scattering function @xmath72 , in this figure we display the times @xmath77 needed for @xmath78 to decay below the value @xmath79 .
the thick dash - dotted ( red ) line in interpolates through points for which @xmath80 . the shape of this iso-@xmath77 line clearly contradicts the mct scenario @xcite , which predicts a re - entrance regime ( a `` bending - back '' of the collective glass transition line ) for large values of @xmath0 . since the arrest line obtained from @xmath60
is situated at considerably larger @xmath0 than the diffusion - localization line predicted by mct , it is reasonable to expect a similar shift for the glass transition line .
unfortunately , an estimate for such shift suggests that the re - entrant scenario may take place at combinations of @xmath0 and @xmath1 at which systems can not be set up due to geometric constraints ( cf . ) .
however , one should also bear in mind that the available mct calculations are based on structure factors obtained from integral equations @xcite as it is still difficult to extract reliable blocked correlations from simulations @xcite .
thus , it remains hard to disentangle the inherent deficiencies of mct from those related to the input structural data , especially at high matrix densities .
based on simple arguments , in @xcite it is concluded that the glass transition line has to coincide with the diffusion - localization line in the limit of vanishing fluid density .
the data presented in and [ fig : kinetic_diagram / fkt_coll.con ] are ostensibly inconsistent with this expectation .
we remark , however , that our simulations do not provide for this special case since we decided to retain a nonvanishing number of fluid particles in every system realization so as to always have finite fluid - fluid correlations .
more importantly , in we will provide evidence that using the present definition of the decay times @xmath77 and @xmath60 , the two quantities are not quite comparable to one another . [ width = ] figure06.eps [ width = ] figure07.eps in this subsection , as well as in and [ subsec : path.iii .. constant.fluid.density ] , we refine our findings from @xcite concerning @xmath72 , @xmath81 , and @xmath14 . additionally , we computed the logarithmic derivative @xmath82}}{d { \left[{}}\log t { \right]}}\ ] ] which facilitates the discussion of @xmath14 since in an assumed subdiffusive law of the form @xmath83 it represents the momentary value of the power @xmath84 . the first set of data we present for state points at various @xmath1 at a common @xmath42 , which corresponds to a vertical line in the kinetic diagrams and [ fig : kinetic_diagram / fkt_coll.con ] that shall henceforth be referred to as `` path i. '' in , the time dependence of @xmath81 and @xmath72 is depicted .
for both observables , the wave vector was chosen to be @xmath85 , which is close to the first peak in the static structure factor ( cf . ) .
for elevated @xmath1 an intermediate - time plateau can be identified in both @xmath72 and @xmath81 ; as @xmath1 is increased , this plateau extends to longer and longer times but does not change in height .
moreover , @xmath72 and @xmath81 decay on the same time scale as they approach this type - b transition .
this is a well - known phenomenon in bulk glass formers @xcite , where the collective glass transition drives the arrest of the individual particles .
the bulklike behavior is also prevalent in @xmath14 , as is evident from .
for the lowest value of @xmath1 depicted , @xmath14 crosses over directly from ballistic ( @xmath86 ) to diffusive ( @xmath87 ) behavior . upon increasing @xmath1 the ballistic range
is followed by a distinct regime in which @xmath14 grows only very slowly , which is reflected by a drop of @xmath88 below unity and as low as @xmath89 for the highest @xmath1 considered .
this is another manifestation of the cage effect ( cf . ) . for all such @xmath1 ,
ensuingly diffusive behavior is recovered in the long - time limit , with the time to approach @xmath87 increasing enormously with @xmath1 .
[ width = ] figure08.eps [ width = ] figure09.eps in this subsection we present data for state points at various @xmath1 at a common @xmath90 , i.e. , along another vertical line in the kinetic diagrams presented in . in the following
we will refer to this line as `` path ii . '' in
, the time dependence of @xmath91 and @xmath92 is depicted . as can easily be seen , the decay pattern of both correlators is quite unlike that in and therein .
moreover , @xmath81 and @xmath72 deviate substantially from one another .
most strikingly , the time scales on which the correlators decay to their long - time value are vastly disparate , and increasingly so for larger values of @xmath1 . at @xmath93 , for instance , the decay times read @xmath94 and @xmath95 .
this represents a difference of almost three decades in time , with the _ collective _ correlator decaying faster .
@xmath91 seems to approach a transition since the time it needs to decay to its long - time value increases substantially as @xmath1 is increased .
however , it is difficult to pinpoint for sure whether @xmath81 exhibits a transition of type - a ( single - step decay ) or type - b ( two - step decay , cf . ) .
for one , in simulations a type - a transition may show features that differ from the theoretical predictions @xcite .
further , it is possible that an intermediate - time plateau would emerge upon increasing @xmath1 above @xmath96 ( the largest value of @xmath1 depicted in ) .
note , however , that beyond @xmath97 geometry strongly inhibits to realize system instances ( see ) .
we conclude that using the current scheme of simulation and analysis , no definite statement can be made about the type of transition in @xmath81 in this part of the kinetic diagram .
@xmath92 shows an intermediate - time plateau for large values of @xmath1 , just as in .
the times at which such plateau may be identified range from @xmath98 for @xmath99 up to @xmath100 for @xmath97 .
in contrast to , @xmath72 attains an additional nonzero long - time value that increases with @xmath1 .
a tail can be identified even at the lowest fluid packing fraction considered ( @xmath101 ) , whereas in this case _ no _ intermediate - time plateau is present .
therefore , for the sequence of state points at hand , two superposing decay behaviors exist in @xmath72 : a type - a transition responsible for the long - time value , and another relaxation mechanism at higher @xmath1 that leads to the intermediate - time plateau and is likely to be caused by a weak collective cage effect . the type - a transition is probably connected to the continuous diffusion - localization transition predicted by mct @xcite and is discussed in more detail in and [ subsec : trapping ] .
although not as obvious , evidences that the mean - squared displacement @xmath14 differs from bulklike behavior as well . as is the case for @xmath50 , for @xmath90 ballistic behavior is followed by a regime for which @xmath84 drops well below unity .
contrary to the bulklike case , this observation holds also for very low @xmath1 , suggesting that caging by fluid particles is not the only mechanism effective in this regime .
most likely , at low @xmath1 the drop in @xmath84 is entirely due to the porous matrix , while at larger @xmath1 both the pores and caging play a role . for low @xmath1
, the range of decreasing @xmath84 is followed by an almost straight increase to @xmath87 , i.e. , a direct approach of the diffusive regime . upon increasing @xmath1 ,
a distinct intermediate - time plateau with @xmath102 emerges in @xmath88 , which corresponds to a subdiffusive regime in @xmath14 .
the estimated value @xmath103 in the subdiffusive regime is only weakly dependent on @xmath1 ; merely for @xmath104 the value of @xmath84 seems to systematically decrease .
this decline , however , may be due to insufficient equilibration of the respective systems . [ width = ] figure10.eps [ width = ]
lastly we turn to a selection of state points at varying @xmath0 for a relatively low fixed @xmath67 , i.e. , along a horizontal straight line in the kinetic diagrams , perpendicular to paths i and ii . this line will henceforth be called `` path iii . '' in we present @xmath91 and @xmath92 for these state points .
the correlators differ substantially from each other and from their bulklike counterparts even more than was the case in .
most notably , @xmath72 attains a sizeable nonzero value as time approaches infinity , and does so more pronouncedly as @xmath0 increases .
by contrast , for all @xmath0 represented in @xmath81 decays strictly to zero . on the other hand
the actual pattern of the decay to the long - time value is quite similar in @xmath81 and @xmath72 : unlike in , both correlators decay in a single step , suggesting only type - a transitions to be involved . comparison of with and [ fig : const_mpf0.05/msd ] reveals the functional form of @xmath14 to be more sensitive to @xmath0 than to @xmath1 . as observed in , at low matrix densities the initial ballistic behavior
is directly followed by diffusive behavior . for intermediate @xmath105 ( cf . ) a region of very slowly increasing particle displacement ( @xmath102 ) succeeds the ballistic range .
the value of the exponent @xmath84 in this region decreases as @xmath0 increases .
subsequently , diffusive behavior is recovered as @xmath84 raises in an almost linear fashion . upon further increasing @xmath0 , at some @xmath106
a geometric percolation transition takes place in the space accessible to the fluid particles ( `` voids '' ) , with a void stretching through the whole system at @xmath107 .
this transition is intimately connected with the diffusion - localization transition predicted by mct @xcite and observed in our simulations . as a hallmark of this transition , @xmath14 does not approach diffusive behavior for @xmath108 but instead remains at an approximately constant @xmath109 for a window covering about three decades in time .
since for that matrix density @xmath88 ultimately increases beyond @xmath110 , one can expect the diffusion - localization transition to take place at slightly higher @xmath0 .
an upper bound to that value is @xmath111 , the smallest value presented that is larger than @xmath112 . at this packing fraction of obstacles the window of constant @xmath84
extends over roughly two decades in time , with @xmath84 ultimately decreasing below that plateau .
judging from these observations , the subdiffusion exponent @xmath84 at the diffusion - localization transition might be slightly lower than 0.5 .
nevertheless , the observed value of @xmath84 is in striking agreement with mct , which predicts that @xmath113 @xcite .
an additional agreement of numbers calls for investigation : it has long been known in theory @xcite and also verified in experiment @xcite that in a one - dimensional random walk
so - called `` single - file diffusion''the mean - squared displacement of the particles grows with precisely the same exponent as in our case .
however , it remains an open question whether in qa systems the main effect in single - file diffusion nonovertaking particles is prevalent , or if other ( possibly compensating ) reasons lead to a coincidental agreement in @xmath84 .
[ width = ] figure12.eps [ width = ] figure13.eps in this subsection we will elucidate some aspects of the geometrical structure imposed by the quenched matrix upon the fluid immersed therein .
the key concept involved is that of `` voids , '' i.e. , domains of space that may be fully explored by a fluid particle if placed within . in hs - qa systems
distinct voids are separated by infinite potential walls , and there exist voids of finite volume .
particles located in such finite void can not travel infinitely far away from their initial location ; such particles will henceforth be denoted as `` trapped particles , '' and the corresponding void as a `` trap . ''
due to the statistical nature of the matrix structure , at any nonzero matrix density there exist traps .
likewise , for any nonzero density of traps there will be trapped particles since the initial locations of the fluid particles are randomly distributed .
two types of nontrivial questions remain to be assessed .
first of all , there exists the aforementioned distinct @xmath106 above which all fluid particles are trapped .
differently phrased ( cf . ) , upon varying @xmath0 , a percolation transition of the space accessible to the fluid particles takes place : for @xmath107 , there exists an infinitely - large void , whereas for @xmath114 any void is of finite size .
naturally , the question arises what the precise value of @xmath106 is . in the following
we will investigate the behavior of fluid particles that explore the voids by studying the distribution of their displacements , as was done before for other types of confinement @xcite .
our quantity of choice for the desired analysis is the self - part of the direction - averaged van hove correlation function @xmath115 where @xmath116 and by construction @xmath117 . strictly speaking , to assess the void structure only knowledge about the infinite - time limit @xmath118 is required .
however , since data are readily available , in the following we will also discuss some features of the time evolution of @xmath119 in order to corroborate and extend our findings from the previous sections . using the self - part of the van hove function a number of quantities of interest
can be assessed at least qualitatively .
for instance , the matrix density at percolation , @xmath106 , can be estimated by exploiting the fact that as the system control parameter ( here @xmath0 ) is varied the average size of the `` aggregates '' ( in this case the traps ) diverges at the transition .
for sufficiently small @xmath120 , the distribution of trap sizes is reflected in @xmath118 since in the infinite - time limit the nontrapped fluid particles have diffused far away . concentrating on a particular distance @xmath121 , the function
@xmath122 will exhibit two maxima as @xmath0 is varied , one for @xmath123 and one for @xmath124 .
the corresponding integral @xmath125 represents an estimate for the fraction of fluid particles located in traps .
of course , in simulations @xmath126 has to be approximated by some finite time @xmath127 greater than the structural relaxation time of the system . in and [ fig : trapping / lower - right ] the van hove function
is displayed at various @xmath13 as a function of @xmath120 . for visual guidance ,
curves are highlighted for selected times .
the times of successive light ( gray ) lines differ by @xmath128{10}$ ] , i.e. , there are about six curves per time decade . in
( a ) @xmath119 is shown for @xmath50 and @xmath51 , a bulklike state point like those discussed in . at short times
@xmath119 exhibits a single , approximately gaussian peak .
as @xmath13 increases , the maximum value of this peak decreases , but notably its @xmath120 position remains nearly constant even for large @xmath13 .
this is another manifestation of the cage effect prevalent in glassy systems @xcite .
further support for this picture arises from the second peak in @xmath119 that emerges at @xmath129 for @xmath130 , which indicates the presence of hopping processes . for very long times
even a moderate third peak appears at @xmath131 as @xmath119 recovers a gaussian - like shape centered at @xmath132 .
the validity of these arguments becomes clear when considering ( b ) which depicts @xmath119 for @xmath90 and @xmath93 .
for this state point , representing intermediate matrix density , similar features can be observed for short and intermediate times , only with the first peak decreasing slower in height , the second peak being less pronounced , and the third peak missing .
however , for large @xmath13 and small @xmath120 the shape of @xmath119 is markedly different from that in ( a ) : even for @xmath133 ( the largest time considered ) we find @xmath134 to be nonzero for small @xmath120 .
since curves of @xmath119 very much resemble each other for @xmath135 and @xmath136 , it is reasonable to assume that in this range @xmath134 represents a reliable approximation to @xmath118 .
hence we can infer the presence of trapped particles , with their fraction being @xmath137 using @xmath138 .
moreover , since @xmath134 is close to zero for @xmath139 , there likely exist only few traps with a spatial extent exceeding @xmath140 .
the trend seen from ( a ) to [ fig : trapping / upper - left](b ) is continued by the systems represented in .
panel ( c ) of that figure shows the state point at @xmath66 and @xmath67 , that is , close to the diffusion - localization transition ( discussed in ) ; the state point in panel ( d ) is situated well in the localized phase at @xmath141 and @xmath101 .
the change in shape over time becomes less pronounced as @xmath0 increases , turning the second peak into a mere shoulder .
the largest times depicted in again represent a sound approximation of @xmath118 .
for the state point in panel ( d ) we numerically obtain @xmath142 using @xmath143 and @xmath144 , which is in excellent agreement with the figure @xmath145 that would be expected in an infinitely - large system residing beyond the percolation transition . for the state point in ( c ) , using the same @xmath140 and @xmath127 we obtain @xmath146 , which , however , is rather imprecise since @xmath147 is considerably greater than zero . in ( c ) and ( d ) the correlator @xmath119 differs markedly from that in ( a ) and ( b ) . for one , in both ( c ) and ( d )
a second peak is present even for @xmath148 , suggesting that on average traps are large enough to allow for hopping .
another feature is less obvious but more important to the percolation transition : for large distances @xmath118 is _ greater _ in ( c ) than in ( d ) for the same @xmath120 .
this implies that the mean trap size is larger in ( c ) than in ( d ) , which is not unexpected in fact this confirms the transition to take place between ( b ) and ( d ) .
thus , using @xmath119 we find @xmath149 as an estimate for the matrix packing fraction at the percolation transition , which is in accordance with our findings in the previous sections .
[ width = ] figure14.eps [ width = ] figure15.eps [ width = ] figure16.eps [ width = ] figure17.eps in our presentation of the global aspects of the system s dynamics , as well as of its behavior along selected paths in the @xmath150 plane , we have focused our attention on self and connected density correlators as indicators of the single - particle and collective dynamics .
this choice is motivated by the key role played by these dynamic correlation functions within the mct framework for confined fluids , which provides detailed predictions for @xmath72 and @xmath81 . before discussing in more detail
the comparison between the theoretical scenario of mct and our simulations , it is instructive to study correlation functions _ related _ to @xmath72 and @xmath81 .
namely , we will investigate here the total intermediate scattering function @xmath151 as well as a modification of the self - intermediate scattering function in which the long - time plateau ( as observable , for instance , along paths ii and iii ) is subtracted out .
this will allow to clarify aspects related to the kinetic diagram of the system and to facilitate the interpretation of the numerical results in the light of the mct predictions . by construction ,
fluids confined in porous media possess nonzero average density fluctuations , @xmath152 , induced by the matrix structure .
this purely - static background becomes visible in the long - time limit of the total intermediate scattering function @xmath153 .
collective relaxation phenomena in confined fluids are therefore more conveniently described by the connected correlator , , in which only the _ fluctuations _ of the microscopic density are considered .
nevertheless it is interesting to also inspect the shape of @xmath153 as a function of the state parameters . in
we show @xmath153 for some selected densities along path i ( @xmath50 ) for the same value of @xmath59 considered in . except for the lowest fluid density , @xmath101
, the @xmath153 correlator attains a finite , sizeable plateau at long times . to check whether this long - time plateau is due to the purely - static background mentioned above or actually
reflects a nonergodic behavior of the system , we also include in this figure the function @xmath154/s(k)$ ] , where @xmath155 is the so - called blocked structure factor , and @xmath156 is the connected structure factor .
it is easy to show that , if the system is ergodic , the blocked structure factor is the infinite - time limit of the un - normalized total intermediate scattering function @xcite and that @xmath157 holds as an equality . from
we see that this equality is indeed fulfilled for @xmath158 , while slight deviations occur at the largest fluid packing fractions .
note , however , that the calculation of @xmath159 is also affected by statistical noise ( see error bars in ) and systematic effects @xcite .
nonetheless provides an effective criterion to discriminate nonequilibrated samples as far as the collective relaxation is concerned .
we also note that if we employ the criterion @xmath160 to define the relaxation time @xmath161 , the latter strongly overestimates the slowness of the collective relaxations compared to @xmath77 due to the emergence of the finite plateau in @xmath153 upon increasing @xmath1 .
this should be born in mind when discussing decoupling phenomena between single - particle and collective properties .
a similar , yet physically different long - time plateau is observed in the self - intermediate scattering functions along path ii and iii ( see and [ fig : const_fpf0.10/fkt ] ) . as discussed in ,
the finite value of @xmath162 reflects the trapping of particles in disconnected voids of the matrix structure .
since the existence of a rising plateau affects the evaluation of the relaxation times @xmath60 , it is sensible to define a modified correlator @xmath163 , where the long - time limit @xmath164 is identified by some construction for fixed @xmath59 is as follows : first , the global minimum @xmath165 of the self - intermediate scattering function is identified .
then , the standard deviation , @xmath166 , of @xmath72 is computed over different time ranges @xmath167 , keeping the upper bound @xmath168 of the range fixed at the maximum time available while progressively increasing the lower bound @xmath169 .
if @xmath170 is below some preset tolerance threshold then @xmath72 is considered to be constant in the corresponding time range . if this condition is met while @xmath171 then @xmath172 .
otherwise , @xmath72 has not reached its long - time plateau within the time range considered . ] .
the results of this procedure are collected in for selected state points at @xmath67 ( path iii ) and compared to the corresponding @xmath72 .
at the two largest @xmath0 the relaxation to the finite plateau is not complete within the time range considered , and the correlators are left unshifted .
evaluation of the single - particle relaxation time @xmath173 , defined by @xmath174 as for the other correlators , allows to filter out the contribution to the single - particle dynamics due to trapping .
we constructed two alternative kinetic diagrams for the system at hand using the relaxation times @xmath161 and @xmath173 obtained from @xmath153 and @xmath175 .
as in , lines drawn in and [ fig : kinetic_diagram / fkt_self.no_ltv ] correspond to a fixed , constant value of the relaxation times ( isorelaxation times lines ) .
let us first focus on , displaying the kinetic diagram based on @xmath161 . in this case
, the range of state points available to determine the isorelaxation times lines is more limited due to the incipient growth of the plateau in the region of large @xmath1 and large @xmath0 .
nevertheless it is clear that the shape of the estimated arrest line is non - re - entrant , confirming the analysis based on the connected correlators ( see ) .
a more interesting effect is observed in where we show the kinetic diagram based on @xmath173 . for better comparison with the analysis in
we also include the isorelaxation times lines obtained from the self and the connected correlators .
interestingly , the shape of the kinetic diagram for @xmath173 strongly resembles that obtained from the connected correlator .
in particular , the iso-@xmath173 lines do not bend rapidly toward @xmath27 as @xmath0 is increased but rather stretch to larger @xmath0 values , following the trend of the iso-@xmath77 line .
this effect can be understood as follows : after subtracting the long - time plateau , which is due to trapping , the residual relaxation of @xmath72 is governed by a weak , collective caging effect .
the relaxation of @xmath72 _ toward the long - time limit _ occurs , in fact , on a _
similar _ time scale as the relaxation of @xmath81 .
this supports our interpretation of the complex relaxation pattern of @xmath72 for large matrix packing fractions ( see ) as a superposition of trapping and caging effects , and confirms the connection between the diffusion - localization transition and the appearance of a nontrivial decoupling between single - particle and collective properties @xcite .
in this contribution we have investigated the dynamic properties of a fluid confined in a disordered porous matrix . in our investigations
we have reduced such system to a simple qa model where the matrix is a quenched configuration of a liquid and the fluid particles equilibrate within that matrix .
both matrix and fluid particles are chosen to be hard spheres of equal size .
thus , the parameter space characterizing the system reduces to a two - dimensional plane spanned by the packing fraction of the matrix particles and that of the fluid particles .
investigations have been carried out via event - driven molecular dynamics simulations , observables were evaluated using the double - averaging recipe characteristic for qa systems .
specifically , we have evaluated the connected and the single - particle intermediate scattering functions , the mean - squared displacement , and the self - part of the van hove function .
the data and conclusions summarized here represent a counterpart to the recently presented results obtained from a theoretical framework @xcite . the latter approach an extension of mct to the qa protocol had predicted a number of intriguing features in the kinetic diagram .
our investigations have complemented this scenario of _ ideal _ transitions , unveiling the properties of the _ real _ transitions : ( i ) at low matrix packing fractions we have undoubtedly confirmed the predicted type - b glass transition scenario .
( ii ) for intermediate matrix packing fractions we have found at least two arrest mechanisms of different nature , with the self - intermediate scattering function displaying a more complex behavior than its connected counterpart .
this scenario can be rationalized as a superpositioning of the effect of trapping in disconnected voids and of a weak caging mechanism collective in nature .
( iii ) varying at low fluid packing fractions the density of the matrix we found evidence for a diffusion - localization transition , reflected in the self - intermediate scattering function as well as in the mean - squared displacement and in the self - part of the van hove function .
( iv ) despite considerable effort we were unable to support the predicted re - entrant transition scenario . in order to obtain deeper insights into the open questions we have additionally considered along path ii the total intermediate scattering function , and along path iii we have subtracted from the self - intermediate scattering functions the long - time plateau as there is evidence that the latter is due to trapping .
our analysis of these data supports the previously established notion of an interplay between the two phenomena dominating the arrest transition at intermediate matrix packing fractions , namely trapping and caging .
work to disentangle the effects of the latter on the dynamic correlation functions is currently underway .
the authors gratefully acknowledge helpful discussions with v. krakoviack and k. miyazaki .
this work was funded by the austrian science foundation ( fwf ) under project no .
[ width = ] figure18.eps as mentioned in the initial setup of a hs - qa system represents a formidable challenge . in this appendix
we will point out algorithms known to succeed in setting up disordered one - component hs systems at high density , comment on problems in extending those algorithms to the qa protocol , and report on our custom solution to the problem .
the challenge is the following : given the positions of the @xmath8 immobile obstacle particles , permissible positions are sought for the @xmath7 particles that during md will be allowed to move . unfortunately , to this respect multiple complications arise in hs - qa systems , the three most important of which are : ( 1 ) particle overlaps are strictly forbidden , ( 2 ) in the resulting configuration the matrix particles have to occupy precisely - specified locations , and ( 3 ) all voids notably also the disconnected ones have to be considered as locations for the fluid particles .
the simplest method trial - and - error insertion of the fluid particles is useless already for setting up _
high - density systems of hard spheres . for the latter task a useful straight - forward method
( although prone to crystallization ) is compressing a low - density system , for instance by coordinate re - scaling combined with random particle displacements , or by applying a unidirectional force in a bounded system .
however , there also exist elaborate algorithms to maximize the density in disordered hs systems ; they employ for instance serial deposition @xcite or overlap elimination @xcite .
unfortunately , the complications mentioned before render these methods difficult to adapt and/or ineffective .
therefore , in order to efficiently achieve high densities in hs - qa systems we devised a custom algorithm , which consists of the following three steps : \(1 ) upon insertion of a fluid particle its `` real '' diameter ( the diameter to be used during md simulations ) is reduced to some minute value ( `` deflated '' ) so that the probability of an insertion via trial - and - error is greatly increased . this way , throughout the search of an allowed configuration all particles can be present in the system simultaneously .
\(2 ) in order to find such allowed configuration a simple metropolis monte - carlo algorithm @xcite is employed using the hs potential . since using
this algorithm with the minute diameters is meaningless , at the beginning of each sweep ( one displacement attempt for each fluid particle ) the diameter of each deflated particle is increased ( `` inflated '' ) to the maximum value possible without overlapping another particle . if for a particle the latter value is greater than its real diameter then inflating is done for this particular particle and it is assigned its real diameter ( `` fully inflated '' ) .
\(3 ) a disconnected void may be filled with ( almost ) any number of deflated particles ; however , the particular void may be too small to accommodate all of these particles if they were fully inflated . to remedy this problem , before certain sweeps
all particles that are not yet fully inflated are removed from the simulation box and then re - deflated and re - inserted according to step 1 .
the number of sweeps separating two such removal - and - reinsertion procedures is gradually increased so as to allow for any number of sweeps to fully inflate a particle while still quickly `` draining '' crammed voids .
obviously , as soon as all particles are fully inflated the setup routine is completed and a configuration fulfilling the qa requirements has been found . in step 2
the fluid particles are inflated serially .
this leads to the distribution of diameters covering a wide range at all times during iteration , except when close to the approach of an allowed configuration .
since this introduces a `` fluctuating polydispersity , '' we found the generation of ordered states at high densities to be strongly suppressed even for bulk monodisperse systems .
figure [ fig : kinetic_diagram / inflation ] gives an overview over the state points at which , according to the above - described steps , a system instance with @xmath24 particles may be set up with probability @xmath176 within @xmath177 cpu minutes ( intel core 2 duo e8400 ) .
comparing the so - found line of discrimination ( red solid line in ) with the @xmath14 dynamic arrest criterion ( green dotted line in ; see and @xcite ) and other features of the kinetic diagrams , one can see that even using our custom routine the system setup represents a challenge to access the state points of interest .
note that in practice a probability of 50% is unsatisfactory .
ideally , none of the setup attempts should be rejected since as little bias as possible should be exerted on the statistics
. however , already at relatively low @xmath0 achieving @xmath178 is strictly impossible , the reason being that there exist peculiar matrix configurations that prohibit all fluid particles to be inserted .
nonetheless , by prolonging the cpu time @xmath179 it is possible to increase @xmath180 since more theoretically possible system configurations can actually be realized . throughout this paper
only state points have been considered for which @xmath181 was achieved .
[ width = ] figure19.eps as mentioned in , throughout this work we chose to assign the same constant total number of particles @xmath182 to most systems simulated .
the purpose of this appendix is to determine for which state points this number is _ not _ sufficient to rule out the presence of finite - size effects , and to describe our pertinent solution . strictly adhering to the approach of constant @xmath28 , state points with a large disparity of @xmath0 and @xmath1 attain low values of either @xmath8 or @xmath7 .
neither is desirable since it will render the calculation of properties pertaining to either the matrix or the fluid inaccurate . in this work ,
the most extreme case for low @xmath8 is the state point ( @xmath183 , @xmath184 ) , with @xmath26 entailing @xmath185 .
the corresponding extremal case for @xmath7 is the state point ( @xmath186 , @xmath187 ) , leaving only _ eight _ fluid particles to move in a matrix of @xmath188 particles .
hence , in the present work the much more severe problem lies in low values of @xmath7 . to remedy this problem we chose to adjust @xmath28 such that neither @xmath8 nor @xmath7 assume a value below a lower limit of @xmath189 .
the previous brief analysis shows that in none of our systems @xmath190 , which limits the discussion to state points with low @xmath7 . in the following ,
we assess whether or not the choice of @xmath191 is reasonable . in order to limit the effort , we consider a single state point representative of the problem of low @xmath7 .
specifically , we inspect various observables at @xmath192 and @xmath193 , i.e. , at a state point close to both @xmath27 and the percolation transition in the voids ( cf . and [ subsec : trapping ] ) .
using @xmath26 this system attains @xmath194 which is reasonably close to @xmath191 . to investigate finite - size effects we computed @xmath81 , @xmath72 , @xmath14 , and @xmath88 for @xmath26 , @xmath195 , and @xmath196 while in each case averaging over ten matrix configurations . in the dependence of @xmath197 and @xmath198 upon @xmath28
is visualized .
the intermediate scattering functions are shown for selected wave vectors , where any finite - size effects should be most prominent at small @xmath59 .
indeed , for @xmath199 and @xmath200 there are noticeable discrepancies between @xmath201 , @xmath202 , and @xmath203 ; however , this is hardly surprising considering that even for @xmath204 the corresponding length scale @xmath205 is only slightly smaller than half the edge length of the simulation cell .
in fact , already for @xmath206 , there is no distinguishable difference between @xmath207 , @xmath202 , and @xmath208 . on the other hand , @xmath198 displays small consistent discrepancies between @xmath26 , @xmath195 , and @xmath196 ; however , the worst deviation between @xmath209 and @xmath210 ( which is at @xmath211 and @xmath212 ) is no more than @xmath213 .
it is not surprising that the @xmath198 curves in this case exhibit larger deviations from one another than the @xmath197 curves considering that the latter approach zero rather quickly whereas @xmath198 does not completely relax .
also , we point out that any discrepancies in @xmath198 for different @xmath28 values are well within the range of the error bars . similarly , illustrates the dependence of @xmath214 and @xmath215 upon @xmath28 .
the mean - squared displacement displays only minor differences for @xmath26 , @xmath195 , and @xmath196 : the discrepancy between @xmath216 and @xmath217 at @xmath218 amounts to @xmath219 ; however , since throughout this work we are interested in @xmath220 on _ logarithmic _ scales , the above deviation can be considered sufficiently accurate . @xmath221 , on the other hand , is clearly more susceptible to stochastic errors than @xmath220 , leading those errors to dominate finite - size effects in @xmath222 . for @xmath223
, we find a difference of mere @xmath224 between @xmath225 and @xmath226 , which permits almost quantitative interpretation .
we conclude that @xmath227 is sufficient to yield reliable results for all observables considered .
none of the latter exhibited a deviation of more than @xmath228 between @xmath26 and the tenfold value for @xmath28 , which validates all conclusions presented in this work . | using molecular dynamics simulations we study the slow dynamics of a colloidal fluid annealed within a matrix of obstacles quenched from an equilibrated colloidal fluid .
we choose all particles to be of the same size and to interact as hard spheres , thus retaining all features of the porous confinement while limiting the control parameters to the packing fraction of the matrix , @xmath0 , and that of the fluid , @xmath1 .
we conduct detailed investigations on several dynamic properties , including the tagged - particle and collective intermediate scattering functions , the mean - squared displacement , and the van hove function .
we show the confining obstacles to profoundly impact the relaxation pattern of various quantifiers pertinent to the fluid . varying the type of quantifier ( tagged - particle or collective ) as well as @xmath0 and @xmath1 , we unveil both discontinuous and continuous arrest scenarios .
furthermore , we discover subdiffusive behavior and demonstrate its close connection to the matrix structure .
our findings partly confirm the various predictions of a recent extension of mode - coupling theory to the quenched - annealed protocol . |
intermetallic compounds like ybincu@xmath2 @xcite and euni@xmath3(si@xmath4ge@xmath5 ) @xcite exhibit valence - change transitions with dramatic effects on their thermodynamic and transport properties at ambient pressure and low temperatures .
one characteristic behavior is seen in the magnetic susceptibility which increases with a curie - like or curie - weiss - like law from high temperatures up to the transition temperature @xmath0 where it changes to a pauli - like law with one to two orders of magnitude drop at the transition .
similarly , a magnetic field , for @xmath6 , induces a metamagnetic transition where the magnetization increases rapidly in a field on the order of a few tesla . in the high - temperature phase
the yb ions are close to the 3 + ( 4f@xmath7 ) configuration with one paramagnetic ( j=7/2 , l=3 , s=1/2 ) f - hole , while the eu ions are close to the 2 + ( 4f@xmath8 ) configuration which has one paramagnetic ( j=7/2 , l=0 ) f - electron .
x - ray edge and mossbauer spectroscopies indicate a valence change occurs at @xmath0 which is typically about 0.1 for ybincu@xmath2 and 0.6 for euni@xmath3(si@xmath4ge@xmath5 ) .
the volume change of the lattice at the transition roughly agrees with the change in the size of the rare - earth ion assuming that the low - temperature phase of yb ions contains a mixture of magnetic 3 + and non - magnetic 2 + ( 4f@xmath9 ) configurations , and in the eu compounds a mixture of the magnetic 2 + and non - magnetic 3 + ( 4f@xmath10 ) states .
the surprising feature is that despite the presence of well defined local moments there is no sign of the kondo effect in the high temperature phase : the electrical resistance is large and with a positive ( temperature ) slope , the thermoelectric power is small , the magnetoresistance is positive , and the hall coefficient is large @xcite .
hence , the high - temperature phase seems to be a semi - metal consisting of conduction electrons that scatter off of f - electron ions with well defined moments that behave nearly independently .
as the temperature is reduced , the local moments become unstable , and the system undergoes a transition to a low - temperature valence - fluctuating metal that no longer possesses local moments .
magnetic and neutron scattering experiments@xcite show that the loss of moment in the low - temperature phase is not due to the onset of any long - range order .
the low - temperature phase is metallic with a low resistivity and a small hall coefficient , but with a large peak in the thermopower .
this loss of f - electron moment occurs not due to a complete transfer of the f - electron into the conduction band , but rather occurs due to the additional formation of a hybridized valence - fluctuating phase with a quenched magnetic moment . but
this valence - fluctuating phase is not an ordinary one , because it is sensitive to external perturbations like a magnetic field which can easily restore the local moments via a metamagnetic transition or hydrostatic ( and chemical ) pressure which stabilizes the f - electron shell with a local moment .
finally , optical conductivity measurements @xcite show a complete reconstruction of the many - body excitations as one passes through @xmath0 , where a mid - infrared peak instantly appears in the low - temperature phase and a drude peak sharpens and increases its spectral weight .
a complete theoretical description of these systems poses quite a challenge .
the creation of a valence - fluctuating state in a conventional periodic anderson model is robust and insensitive to external perturbations like temperature , magnetic field , or pressure .
transitions from the valence - fluctuating state to any other phase are all believed to be continuous crossovers that are never first - order - like . on the other hand ,
the stabilization of intermediate - valence states in the falicov - kimball@xcite ( fk ) model is problematic in the limit as @xmath11 , since the f - electrons are usually completely filled or empty in the homogeneous ground state phases .
this is true even though the falicov - kimball model is known to possess discontinuous valence - change transitions .
a proper treatment of these valence fluctuating systems requires combining the periodic anderson model with the falicov - kimball model , by introducing d - f coulomb interactions in the former or hybridization in the latter . solving such a model , and evaluating the phase diagram as a function of all of its parameters is difficult , even with the advanced techniques of dynamical mean field theory .
here we choose an alternate , and hence incomplete approach , which nevertheless appears to capture the majority of the interesting behaviors of these remarkable compounds .
we choose to describe these materials by a spin one - half falicov - kimball model .
we will see below what properties can be described by this model , and more importantly , we point out all of the pitfalls that arise due to the neglect of the hybridization . we give reasonable arguments about how we expect hybridization to modify the behavior below .
the falicov - kimball hamiltonian is @xmath12 where @xmath13 ( @xmath14 is the electron creation ( annihilation ) operator for an electron at site @xmath15 with spin @xmath16 , @xmath17 is the energy level of the localized electrons ( not to be confused with a fermi energy ) , @xmath18 labels the corresponding f - electron creation and annihilation operators , and @xmath19 is the interaction strength .
the total f - electron occupation is restricted to be less than or equal to one at each lattice site , so that this system is always strongly correlated ( @xmath20 ) .
the hopping integral is scaled with the spatial dimension @xmath21 so as to have a finite result in the limit@xcite @xmath22 ; we measure all energies in units of @xmath23 .
we work on a hypercubic lattice where the noninteracting density of states is a gaussian @xmath24 .
a chemical potential @xmath25 is used to maintain the total electron filling @xmath26 at a constant value .
the solution of the falicov - kimball model has been described before in the literature@xcite and wo nt be repeated here .
instead , we focus on the physical behavior of the exact solution .
we present the results for the case where the total filling is @xmath27 but the electrons can change from localized to itinerant ( i.e. , we fix the total electron concentration not the individual electron concentrations ) .
we choose values of the parameters@xcite where the system has a well defined crossover from a state at high temperature that has large f - occupancy ( @xmath28 for @xmath29 ) , to a state at low temperature with no f - electrons ( for definition of the transition temperature @xmath0 , see below ) . to open a gap in the dos at high temperatures
we need a large f - f correlation and we choose @xmath30 ( see below ) . to model the low - temperature transition
, we need to tune the renormalized f - electron level to lie just above the chemical potential as @xmath11 , and to describe ybincu@xmath2-like systems we choose several values of @xmath17 .
we find that the qualitative features of the results do not depend too strongly on the parameters in this regime ( i.e. , if we choose another total electron filling @xmath31 , then we find similar results when we vary @xmath17 ) . as @xmath32 increases ,
the large entropy of the local moments stabilizes the f - electron number and produces a large magnetic response of the system .
we choose the coulomb interaction of an f - electron with a conduction electron to be such that the density of states splits into two bands@xcite .
the shape and the relative weight of the two bands depends on the band filling and temperature but the peak - to - peak distance between the lower and the upper band is always about u. the location of the f - level @xmath17 , is the one parameter that we will adjust to control the high - temperature number of f - electrons ( in the case of eu ) or f - holes ( in the case of yb ) .
note , the single - particle excitations are strongly renormalized by the coulomb interaction , and the bare value of @xmath17 has no physical meaning .
since the fk model neglects the hybridization , we use the variation of @xmath17 to represent the change of the system due to the pressure or doping ( chemical pressure ) .
the gap in the density of states , which appears for @xmath33 , defines a characteristic energy scale @xmath34 .
the gap shrinks with the reduction of temperature but is fairly insensitive to the value of @xmath17 which , on the other hand , has a drastic effect on @xmath0 . for @xmath30 ( @xmath35 )
we have @xmath36 ( @xmath37 ) and @xmath38 ( @xmath39 , a pseudogap ) at very high and low temperatures , respectively . in this paper
we consider the system with @xmath38 ( @xmath30 ) and discuss the temperature dependence of the spin susceptibility of f - electrons @xmath40 , the f - occupancy @xmath41 , the electrical resistance @xmath42 , the thermoelectric power @xmath43 , and the thermal conductivity @xmath44 , for five different values of @xmath17 .
we also discuss the frequency dependence of the optical conductivity @xmath45 , the raman response @xmath46 , the magnetic field dependence of the f - electron magnetization @xmath47 and the electrical resistance @xmath48 , at several characteristic temperatures .
the finite - frequency results are presented for @xmath49 , and the finite - field data for @xmath50 , which are fairly typical values .
the first thing we show is a plot of @xmath40 and @xmath51 , which is a curie - like law , but with a temperature - dependent concentration of spins @xmath52 .
one can see from panel ( a ) , in figure 1 that as @xmath17 is made more negative , the peak in the susceptibility moves to lower @xmath32 and the curves sharpen .
we define the valence transition temperature @xmath0 to be the temperature below the peak in the susceptibility where @xmath53 , i.e. , it is at the temperature corresponding to half the peak height of the susceptibility .
this definition provides a well - defined benchmark for the value of @xmath0 that is similar to choosing the inflection point of the susceptibility curves .
all of the transitions we see here are actually smooth crossovers centered around @xmath54 .
we believe there is a region of parameters where the transition can be made discontinuous at low @xmath32 , but the iterative algorithm normally used to solve this problem becomes unstable in the region close to such first - order transitions , and alternate methods need to be used to solve the problem . in panel ( b ) we plot the inverse susceptibility which shows the characteristic curie or curie - weiss law at moderate temperatures ( as @xmath17 increases , the intercept of the curie - weiss law becomes more negative ) .
note that in our case there never is a low temperature turnover of the inverse susceptibility , since we plot only the f - electron susceptibility here , which vanishes as @xmath55 . in ybincu@xmath2 and euni@xmath3(si@xmath4ge@xmath5)-like systems
the ground state is a valence fluctuator with an enhanced pauli susceptibility , such that the curie plot shows a characteristic low-@xmath32 maximum .
[ fig : chi ] = 3.5 in [ fig : n_f ] = 3.5 in in figure 2 we plot @xmath41 , which is relatively constant over a large temperature range and then suddenly drops toward zero at the transition . as @xmath17 is made more negative , and the high - temperature f - filling is increased , the transition temperature @xmath0 is reduced and the @xmath41-curves sharpen .
note , the magnetic susceptibility , the particle number and the charge susceptibility , defined as @xmath56 , reveal a characteristic scale @xmath0 but do not provide any hint on the magnitude of @xmath34 .
the inset shows the correlation between @xmath0 and @xmath17 , which is our analog of experimental plots of @xmath0 versus doping ( for chemical pressure ) or versus pressure .
the behavior seen here , where applying pressure makes the @xmath17 less negative and reduces the number of f - electrons , is similar to what would occur in experiment on eu compounds ; @xmath0 increases along with a smoothing out of the susceptibility . in yb - based systems
, pressure has the opposite effect@xcite on the properties of the system , since it increases the number of f - holes . as mentioned already , and discussed in detail in refs .
, because of the fk interaction the temperature- or field - induced change in @xmath57 leads to the mi transition in the conduction band , and is accompanied by substantial changes in the density of states . in the low - temperature phase , which is found for @xmath6 ,
the weight of the upper conduction band is very small and the chemical potential is within the lower conduction band . in the high - temperature phase , which sets in for @xmath58 ,
the spectral weight of the lower and upper band are about the same , and the chemical potential lies within the gap .
contrary to @xmath0 , the characteristic temperature @xmath34 is only weakly dependent on @xmath17 .
[ fig : transport ] = 4.5 in the dc transport properties in zero magnetic field , obtained by an exact evaluation of the current - current , current - heat current , and heat current - heat current correlation functions , are summarized in figure 3 .
we show @xmath42 , @xmath43 , and @xmath44 , using the normalization that is usual for infinite dimensional lattices@xcite .
the thermopower is plotted for the hole - like yb case ; for the eu compounds , the sign would have to be changed .
the high temperature phase is a poor metal with a minimum in @xmath43 at about @xmath59 and small hump in @xmath42 at about @xmath60 ( do not confuse the hump at @xmath61 with the sharp low-@xmath32 decrease in @xmath62 that occurs near @xmath0 ) .
note , the thermopower does not exhibit any additional features around @xmath0 .
the minimum of @xmath43 decreases from @xmath63 to @xmath64 , as we change from @xmath65 ( @xmath66 ) to @xmath67 ( @xmath68 ) , i.e. the relevant temperature scale for the thermoelectric power is set by @xmath34 and not by @xmath0 . in the @xmath35 case , which has the high - temperature gap @xmath37 , the minimum of @xmath43 shifts from @xmath69 for @xmath70 ( @xmath71 ) to @xmath72 for @xmath50 ( @xmath73 ) .
the electronic part of the thermal conductivity shows a featureless decrease as temperature is reduced and no @xmath17-dependence .
thus , as long as @xmath74 , the temperature scales revealed by the thermoelectric data and the thermodynamic data appear to be fundamentally different .
the temperature dependence of various correlation functions discussed above is remarkably similar to what one observes in ybincu@xmath2 and euni@xmath3(si@xmath4ge@xmath5)-like systems for @xmath75 .
the thermoelectric data obtained recently for yb@xmath4y@xmath5incu@xmath2 and ybag@xmath5in@xmath4cu@xmath2 for various concentrations of y and ag dopants@xcite can be explained by the fk model , if we take the initial parameters for ybincu@xmath2 such that @xmath76 , and assume that yttrium doping makes @xmath17 more negative , while silver doping does the opposite .
note , the thermoelectric power of the low - temperature valence fluctuating phase is much larger than of the high - temperature semimetallic phase@xcite , i.e. the heating of real systems across @xmath0 always leads to a smaller @xmath43 .
the y - doping pushes @xmath0 ( as defined by the susceptibility anomaly ) to lower temperatures but does not change the position of the high - temperature minimum of @xmath43 , since @xmath77 , and this does not depend on @xmath78 or @xmath17 .
the thermoelectric power of yb@xmath4y@xmath5incu@xmath2 exhibits two minima : a deep one at the temperature of the valence transition and shallow one at the temperature of the pseudogap .
the first one at @xmath0 is doping - dependent , and the second one is doping - independent . in ybag@xmath5in@xmath4cu@xmath2 one
finds the valence transition for @xmath79 , and the silver doping enhances @xmath0 , i.e. @xmath80 . the slope of @xmath43 in the valence - fluctuating phase is characterized by the kondo scale @xmath81 , and as the temperature increases up to @xmath82 , @xmath43 can reach large values . above @xmath0
, @xmath43 drops to values typical of the high - temperature phase but since @xmath83 , there is no high - temperature minimum of @xmath43 .
thus , as long as @xmath84 , the thermopower has a single minimum at a temperature that coincides with the susceptibility maximum , and is strongly doping ( or @xmath17 ) dependent .
however , ag doping not only enhances @xmath0 but also reduces @xmath85 , and for @xmath86 the ybag@xmath5in@xmath4cu@xmath2 becomes a heavy fermion with @xmath87 . for @xmath88 , the kondo effect stabilizes the paramagnetic state before the valence transition takes place and the thermoelectric power exhibits the usual kondo minimum at about @xmath89 , that is not much changed by further doping .
the effect of hydrostatic pressure on these compounds@xcite follows from similar considerations , if we assume that pressure shifts @xmath17 and reduces ( increases ) the average number of f - electrons ( f - holes ) .
[ fig : optical ] = 4.5 in the transport properties of euni@xmath3(si@xmath4ge@xmath5)@xmath3 compounds , which show@xcite ( for @xmath90 ) a valence transition at @xmath91 k , can also be explained by the fk model , but with f - electrons rather than holes .
the reduction of the ge concentration shrinks the lattice , reduces the average value of f - electrons , increases @xmath0 , and makes the susceptibility transition less sharp . in this electron - like system ,
the thermoelectric power is positive@xcite .
since the maximum of @xmath43 is concentration dependent and @xmath92 , we would classify the high - temperature phase of euni@xmath3(si@xmath4ge@xmath5)@xmath3 for @xmath90 , as a small - gap fk state , with @xmath93 . in figure 4
, we plot the optical conductivity@xcite @xmath45 and the @xmath94 raman response@xcite for the case @xmath49 at a number of different temperatures . in the high temperature phase
, the optical conductivity has the characteristic poor metal behavior , with a low - energy drude like feature and a large weight charge - transfer peak around @xmath95 .
as @xmath32 is lowered , the scattering processes become sharpened , and both peaks grow in temperature .
the charge - transfer peak is frozen in at around @xmath96 , but the drude feature continues to increase .
as @xmath0 is approached , we see a transfer of weight from the charge - transfer peak to the drude peak , which then becomes complete as @xmath32 is lowered well below @xmath0 . in this lowest temperature region ,
the results are not representative of real materials , as we expect there to still be a charge - transfer peak since the f - electrons have not gone into the conduction band , they have only hybridized with them .
we also expect midinfrared excitations to appear , because the hybridization with the f - electrons will allow them to participate in the optical scattering , and the broadened f - spectral function should have weight in the midinfrared region .
these latter features are seen in experiment on the yb compounds .
the raman response behaves similarly , with a sharpening of the charge - transfer peak and the growth of a lower fermi - liquid peak as @xmath32 is lowered to @xmath0 , where the spectral weight shifts become too large .
thus , the optical experiments can be used to estimate the value of u. we expect one would also see the turn on of the midinfrared response in the raman scattering as @xmath32 is lowered below @xmath0 for the exact same reason it is seen in the optical conductivity .
if we estimate the f - d correlation from the high - frequency peak in the optical conductivity and use the susceptibility results to define @xmath0 , we obtain from figs . 4 and 2 the ratio @xmath97 . if we use model with @xmath35 and @xmath50 , the ratio is @xmath98 . the optical data for ybincu@xmath2 crystals , with @xmath99 k , exhibit the peak at 12000 @xmath100@xcite , which gives the experimental ratio @xmath101 . [ fig : field ]
= 4.5 in the field dependence of @xmath47 and @xmath48 is shown in figure 5 for @xmath30 , @xmath50 ( @xmath102 ) .
each curve corresponds to a different temperature , as indicated in the caption . for low temperatures and fields , @xmath47 and @xmath48
behave typically of a normal fermi liquid . at high temperatures and high fields
, however , @xmath47 behaves as expected of a non - interacting localized moment , while @xmath48 indicates a poor metal or a semiconductor .
figure 5 shows that the low - temperature metallic state is destroyed by large enough magnetic field . at a critical field
@xmath1 we find a metamagnetic transition in the f - subsystem and a metal - insulator transition in the conduction band .
the h - t phase boundary between the non - magnetic ( metallic ) phase and the paramagnetic ( semimetallic ) phase can be approximated by the expression @xmath103 , where the critical field @xmath104 is estimated from the midpoint of the magnetization curves shown in fig . 5 , and @xmath105 is the critical field at @xmath106 .
the ratio of the zero - temperature zeeman energy at @xmath105 and the zero - field thermal energy at @xmath0 is @xmath107 , where the gyromagnetic factor is @xmath108 , since we are interested in the yb or eu compounds which have the f - electrons in their lowest spin - orbit state .
thus , the high - field and the low - field data show the same characteristic energy scale .
we discuss the properties of ybincu@xmath2 and euni@xmath3(si@xmath4ge@xmath5)-like compounds using the exact solution the falicov - kimball model in infinite dimensions .
in addition to the usual fk interaction , we include an infinitely large repulsion between f - particles of the opposite spin , which leads to @xmath109 ( @xmath57 describes electrons for eu and holes for yb ) .
we consider the total filling @xmath27 , and choose the parameters such that the ground state has a large d - electron spectral weight around the chemical potential @xmath25 and a negligible f - occupancy .
the mass of f - particles is assumed to be infinite but the f - level has a finite width because of the fk coupling to the density fluctuations in the conduction band .
( not surprising , this feature is similar to what one finds in the x - ray edge problem , defined by a single impurity version of the fk hamiltonian . )
the spectral functions of the f- and d - states are gapped and the spectral weight is strongly temperature- and field - dependent . at low temperatures ,
there are many more conduction states below than above the gap , while for the f - states just the opposite is true . at high temperatures ,
the chemical potential is in the gap , and the weight of the occupied and unoccupied the d- and f - states are about the same .
the fk model predicts a valence transition at @xmath0 , a metamagnetic transition for the f - electron magnetization , and a metal - insulator transition in the conduction band , such that the high - temperature , high - field transport is dominated by a pseudogap or a gap @xmath110 .
the paramagnetic , semimetallic phase sets in because of thermal population of highly - degenerate f - states , which can be understood as an entropy driven transition .
the proximity of the renormalized f - state to the chemical potential explains the sensitivity of the transition to an external field , doping or pressure . by tuning the position of the f - level
, the transition can be pushed to low temperatures , where it becomes very sharp . for @xmath111 ,
the number of f - particles is large and only weakly temperature dependent .
the zero - field susceptibility is given by the curie law , @xmath112 , rather than the curie - weiss law , i.e. the curie plot gives only a small curie - weiss temperature and a curie constant " that depends on the thermal occupation of the excited f - states .
note , the presence of the local moment in the high - temperature phase is not accompanied by the usual kondo anomalies in the electronic transport , since the model has no f - d mixing .
the calculated electrical resistance is large with a positive ( temperature ) slope , the magnetoresistance is positive , and the thermoelectric power is small with a peak at about @xmath113 .
the optical conductivity has a pronounced peak at @xmath114 and a suppressed drude peak , as expected of a strongly correlated semimetal .
most of these features are observed in the high - temperature , high - field phase of ybincu@xmath2 and euni@xmath3(si@xmath4ge@xmath5)-like compounds and they seem to characterize a new fixed point of strongly correlated electrons .
the low - temperature , low - field properties of the fk model correspond to a metal with a pauli susceptibility , usual dc transport with a small thermoelectric power , and a large drude peak in the optical conductivity .
however , the model neglects the d - f hybridization and its low - temperature metallic phase without the f - particles , can not describe the low - temperature state of intermetallic compounds with eu and yb ions .
the physical ground state is a valence fluctuator with a large average f - occupancy , an enhanced pauli susceptibility , huge thermoelectric power , and an additional midinfrared peak in the optical conductivity .
all these features follow from the periodic anderson model with an infinite f - f correlation and a d - f hybridization . in that model ,
the proximity of the d and f spectral weight to the chemical potential leads to d - f inter - band transitions and a mid - infrared peak in @xmath45 .
it appears that the low - temperature phase of the above mentioned compounds can be explained by the pam , and the f - d correlation can be neglected . on the other hand , the semimetallic high - temperature phase of ybincu@xmath2 and euni@xmath3(si@xmath4ge@xmath5 ) is difficult to describe in terms of the usual solutions of the pam , which can lead to a fluctuating local f - moment but always keeps the d - electrons metallic .
the absence of the kondo effect in the electrical resistance , the magnetoresistance and the thermoelectric power data is difficult to reconcile with the susceptibility data which is typical of local moments with a small curie - weiss temperature .
such a behavior is not easy to explain in terms of conduction electrons that exchange scatter on the localized f - moments . as long as the f - state overlaps the continuum of metallic states , the quantum mixing leads to the kondo effect , and that is not seen in the high - temperature phase of ybincu@xmath2 and euni@xmath3(si@xmath4ge@xmath5)-like compounds .
also , it is not clear that the pam could provide a sharp transition from a metallic ( mixed valence , non - magnetic ) state to a semimetallic ( integral valence , paramagnetic ) state , as observed experimentally . in our approach , we argue that the high - temperature width and the weight of the f - spectral function below the gap is very large , such that the additional broadening due to the f - d mixing can be neglected . in that sense , the fk model can be considered as an effective model for the high - temperature phase of a generalized periodic anderson model which , in addition to the usual f - f correlations and f - d hybridization , also has the f - d correlation .
it would be interesting to study the low temperature fixed point of such a generalized periodic anderson model with the falicov - kimball term , to see if it reduces to the usual valence fluctuating fixed point of the periodic anderson model .
j. freericks and v. zlati , phys .
b * 58 * , 322 ( 1998 ) ; v. zlati and j.k . freericks , in _ open problems in strongly correlated electrons _ ( eds . j. bona , a. ramak , p. prelovek and s. sarkar ) nato science series ii : mathematics , physics and chemistry - vol . 15 ( kluwer academic publishers , dordrecht 2001 ) | the exact solution of the multi - component falicov - kimball model in infinite - dimensions is presented and used to discuss a new fixed point of valence fluctuating intermetallics with yb and eu ions . in these compounds , temperature , external magnetic field , pressure , or chemical pressure
induce a transition between a metallic state with the f - ions in a mixed - valent ( non - magnetic ) configuration and a semi - metallic state with the f - ions in an integral - valence ( paramagnetic ) configuration .
the zero - field transition occurs at the temperature @xmath0 , while the zero - temperature transition sets in at the critical field @xmath1 .
we present the thermodynamic and dynamic properties of the model for an arbitrary concentration of d- and f - electrons . for large u
, we find a mi transition , triggered by the temperature or field - induced change in the f - occupancy . |
the mean value has long been used as a measure of location of the center of a distribution . however , in many applications , there are occasions when analysts are interested to observe and analyze different points of the distribution . the distribution function @xmath0 of a random variable
can be characterized by infinite number of points spanning its support .
these points are called quantiles .
thus , quantiles are points taken at regular intervals from @xmath0 .
the @xmath1th quantile of a data distribution , @xmath2 , is interpreted as the value such that there is @xmath3 of mass on its right side and @xmath4 of mass on its left side . in particular , for a continuous random variable @xmath5 , the @xmath4 quantile of @xmath0 is the value @xmath6 which solves @xmath7 ( we assume that this value is unique ) , where @xmath8 .
thus , the population lower quartile , median and upper quartile are the solutions to the equations @xmath9 , @xmath10 and @xmath11 , respectively . compared to mean value , quantiles are useful measures because they are less susceptible to skewed distributions and outliers .
this fact form the building block of quantile regression which unlike its standard mean regression counterpart , lies in its flexibility and ability in providing a more complete investigation of the entire conditional distribution of the response variable distribution given its predictors . to this end
, quantile regression has become increasingly popular since the pioneering research of @xcite . after its introduction ,
quantile regression has attracted considerable attention in recent literature .
it has been applied in a wide range of fields such as agriculture @xcite , body mass index @xcite , microarray study @xcite , financial portfolio @xcite , economics @xcite , ecology @xcite , climate change @xcite , survival analysis @xcite and so on .
a comprehensive account of other recent applications of quantile regression can be found in @xcite and @xcite .
longitudinal data is encountered in a wide variety of applications , including economics , finance , medicine , psychology and sociology .
it is repeatedly measured from independent subjects over time and correlation arises between measures from the same subject . since the pioneering work by @xcite ,
the mixed models with random effects have become common and active models to deal with longitudinal data .
a number of books and a vast number of research papers published in this area have been motivated by laird and ware s mixed models .
the majority of these books and research papers focuses on standard mean regression .
see for example , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite and @xcite , among others .
in contrast , limited research papers have been conducted on quantile regression for longitudinal data .
for example , @xcite proposed the @xmath12 regularization quantile regression model , @xcite studied quantile regression for longitudinal data in different contexts and developed resampling approaches for inference .
@xcite suggested a bayesian quantile regression method for longitudinal data using the skewed laplace distribution ( sld ) for the errors , @xcite proposed a flexible bayesian quantile regression method for dependent and independent data using an infinite mixture of normals for the errors , @xcite studied quantile regression for longitudinal data with nonignorable intermittent missing data and @xcite proposed a method for regularization in mixed quantile regression models .
longitudinal data with ordinal responses routinely appear in many applications , including economics , psychology and sociology .
existing approaches in classical mean regression are typically designed to deal with such data . at present time , the most common method in classical mean regression for modelling such data employs the cumulative logit model .
there exists a large literature on the analysis of longitudinal data with ordinal responses , and we refer to @xcite for an overview . in contrast ,
quantile regression approaches for estimating the parameters of ordinal longitudinal data have not been proposed , yet .
the goal of this paper is to fill this gap by introducing an ordinal random effects quantile regression model that is appropriate for such data . in section
[ methods ] , we present a random effects ordinal quantile regression model for analysis of longitudinal data with ordinal outcome by using a data augmentation method .
we also discuss prior elicitation . in section [ gibbs sampler ] , we outline the bayesian estimation method via gibbs sampler . in section [ simulation_studies ] ,
we carry out simulation scenarios to investigate the performance of the proposed method , and in section [ rd ] , we illustrate our proposed method using a real dataset .
we conclude the paper with a brief discussion in section [ conclusion ] .
an appendix contains the gibbs sampler details .
given training data @xmath13 with covariate vector @xmath14 and outcome of interest @xmath15 .
the @xmath1th quantile regression model for the response @xmath16 given the covariate vector @xmath17 takes the form of @xmath18 where @xmath19 is the inverse cumulative distribution function ( cdf ) and @xmath20 is the unknown quantile coefficients vector .
this is what makes the quantile estimators can be considered nonparametric maximum likelihood estimators @xcite . unlike the standard mean regression
, the error term does not appear in ( [ eq1 ] ) because all the random variation in the conditional distributions is accounted for by variation in the @xmath1th quantile , @xmath2 .
consequently , quantile regression does not require any assumption about the distribution of the errors and , unlike standard mean regression , is more robust to outliers and non - normal errors , offering greater statistical efficiency if the data have outliers or is non - normal .
it belongs to a robust model family , which can provide a more complete picture of the predictor effects at different quantile levels of the response variable rather than focusing solely on the center of the distribution @xcite
. one attractive feature of quantile regression is that the linear quantile regression model @xcite can be used to estimate the parameters of the nonlinear model because the quantile regression estimators are equivariant to linear or nonlinear monotonic transformations of the response variable , i.e. , @xmath21 . in general , this is a very important property since it tells us that quantile regression provides consistent back transformation and easy in interpretation in the case of transformations such as the logarithm and the square root .
the quantile estimators have the same interpretation as those of a standard mean regression model except for the indexed quantile levels where each is estimated @xcite .
for example , if the slope is -0.78 for the response variable @xmath6 given the predictor @xmath22 in the @xmath23th quantile would indicate that the @xmath23th quantile of the response variable decreased by 0.78 for each 1 unit increase in @xmath22 .
the unknown quantity @xmath20 is estimated by @xmath24 where @xmath25 is the check loss function ( clf ) at a quantile @xmath1 , @xmath26 , and @xmath27 is the indicator function .
by contrast , standard mean regression method based on the quadratic loss @xmath28 .
@xcite observed that the clf ( [ eq2 ] ) is closely related to the skewed laplace distribution ( sld ) and consequently the unknown quantity @xmath20 can be estimated through exploiting this link .
this observation opens new avenues when dealing with quantile regression and its applications .
the density function of a sld is @xmath29 minimizing the clf ( [ eq2 ] ) is equivalent to maximizing the likelihood function of @xmath30 by assuming @xmath31 from a sld . by utilizing the link between the clf ( [ eq2 ] ) and the sld , @xcite proposed a bayesian framework for quantile regression using the sld for the error distribution and show the propriety of the conditional distribution of @xmath20 under an improper prior distribution .
unfortunately , the joint posterior distribution under this framework does not have a known tractable form and consequently @xcite update the unknown quntity @xmath20 from its posterior using the metropolis - hastings ( m - h ) algorithm . in this context
, @xcite proposes a simple and efficient gibbs sampling algorithm for updating @xmath20 by motivating the sld as a member of the scale mixture of normals . if @xmath32 , then the sld for @xmath33 arises when @xmath34 has an exponential distribution with rate parameter @xmath35 . under this formulation
, @xcite presented a bayesian framework for structured additive quantile regression models , @xcite developed bayesian quantile regression for longitudinal data and @xcite presented bayesian lasso mixed quantile regression .
let @xmath36 , where @xmath37 denote the response for the @xmath38th subject measured at the @xmath39th time .
then , the @xmath1th quantile regression model for ordinal longitudinal data can be formulated in terms of an ordinal latent variable @xmath40 as follows : @xmath41 where @xmath42 is the inverse cdf of the unobserved latent response @xmath40 conditional on a location - shift random effect @xmath43 , @xmath44n@xmath45 , and @xmath46 is a vector of predictors .
the observed ordinal response @xmath37 is assumed to be related to the unobserved response @xmath40 by @xmath47 + + where @xmath48 are cut - points whose coordinates satisfy @xmath49 here , @xmath50 and @xmath51 are respectively defines the lower and upper bounds of the interval corresponding to observed outcome @xmath52 .
assuming that the error @xmath53 of the unobserved response @xmath40 has a sld as in ( [ eq3 ] ) , we have @xmath54 @xcite . here ,
the latent variable @xmath55 follows an exponential distribution with rate parameter @xmath35 , and @xmath56 follows the standard normal distribution .
then , the cdf for the c category of the observed response @xmath37 is : @xmath57 where @xmath58 is the standard normal cdf .
using @xmath58 , we can calculate @xmath59 as follows : @xmath60 prior distribution selection is an essential step in any bayesian inference ; however , in the bayesian paradigm it is particularly crucial as issues can arise when default prior distributions are used without caution @xcite . for the fixed effects ,
a typical choice is to assign a zero mean normal prior distribution on each @xmath61 , which leads to the ridge estimator .
however , this prior performs poorly if there are big differences in the size of fixed effects @xcite .
an generalization of the ridge prior is a laplace prior , which is equivalent to the lasso model @xcite .
this prior has received considerable attention in the recent literature ( for example see , @xcite ) . in this paper
, we assign a laplace prior on each @xmath62 takes the form of @xmath63 according to @xcite , the prior ( [ eq8 ] ) can be written as @xmath64 from ( [ eq9 ] ) , it can be seen that we assign a zero - mean normal prior distribution with unknown variance for each @xmath62 .
we specify an exponential prior distributions with rate parameter @xmath65 for the variances assuming they are independent .
then , we put a gamma prior on @xmath66 with shape parameter @xmath67 and rate parameter @xmath68 .
since @xmath44n@xmath45 , this motivates us to consider an inverse gamma prior on @xmath69 with shape parameter @xmath70 and scale parameter @xmath71 .
following @xcite and @xcite , we consider an order statistics from @xmath72 distribution , for the @xmath73 unknown cut - points : @xmath74 where @xmath75 and @xmath76 because @xmath77n@xmath78 and we observe @xmath79 if @xmath80 , the posterior distribution of all the parameters and latent variables is given by @xmath81 where , @xmath82 , @xmath83 , @xmath84 , @xmath85 and @xmath86 .
the full conditional distributions for @xmath87 , @xmath88 and @xmath69 are summarized below and details of all derivations are provided in appendix [ appendix ] .
using the data augmentation procedure as in @xcite , a gibbs sampling method for the ordinal quantile regression model with longitudinal data is constructed by updating @xmath87 , @xmath89 , @xmath66 , and @xmath69 from their full conditional distributions . from ( [ eq12 ] ) , we can construct a tractable algorithm for efficient posterior computation that works as follows : 1 .
sample @xmath55 @xmath90 from the generalized inverse gaussian distribution gig@xmath91 , where @xmath92 and @xmath93 .
sample @xmath94 from n@xmath95 where @xmath96 3 .
sample @xmath97 from gig@xmath98 , where @xmath99 and @xmath100 .
4 . sample @xmath66 from gamma distribution with shape parameter @xmath101 and rate parameter @xmath102 .
sample @xmath103 from n@xmath104 , where @xmath105 6 .
sample @xmath69 from inverse gamma distribution with shape parameter @xmath106 and scale parameter @xmath107
sample @xmath108 from truncated normal ( tn ) distribution tn@xmath109 .
sample @xmath51 from a uniform distribution on the interval @xmath110 $ ] .
the details of the proposed gibbs sampler algorithm and fully conditional posterior distributions are given in appendix a.
we carry out a monte carlo simulation studies to assess the performance of the proposed method .
we compared the proposed bayesian quantile regression method for ordinal longitudinal data , referred to as bqol " , with bayesian quantile regression for ordinal models ( bqror ) reported by @xcite .
the results of bayesian logistic ordinal regression ( blor ) for longitudinal data and the maximum likelihood logistic ordinal regression ( mle ) were also reported .
models were assessed based on the relative average bias and the estimated relative efficiency .
suppose that we are interested in the estimation of a vector of parameters @xmath111 .
then , the relative average bias of @xmath112 is given by @xmath113 and the estimated relative efficiency @xmath114 where @xmath115 denotes the number of replications , @xmath116 is the parameter estimate for the @xmath117th replication , @xmath118 is the true value , @xmath119 and @xmath120 . here
, we follow the same simulation strategy introduced by @xcite .
specifically , we simulated data from the following liability : @xmath121 where @xmath122 and @xmath123 were sampled independently from a uniform distribution on the interval @xmath124 $ ] , @xmath125 and @xmath53 were sampled independently from a logistic distribution with location parameter @xmath126 and scale parameter @xmath127 .
three values of @xmath128 were considered , @xmath129 and @xmath130 .
the cut - points used were @xmath131 and @xmath132 .
then the outcome @xmath37 was sampled according to : @xmath133 for each choice of @xmath128 @xmath134 and @xmath135 , we generated 200 data sets .
we ran the proposed gibbs sampler algorithm for 20,000 iterations , after a burn - in period of 2000 iterations .
cccccccccccccc & & + @xmath128 & parameter & bias & eff & & bias & eff & & bias & eff & & bias & eff & + & @xmath136 & -0.051 & 1.000 & & -0.079 & 1.103 & & 0.082 & 1.012 & & -0.068 & 1.009 + & @xmath137 & -0.002 & 1.000 & & 0.082&1.113 & & 0.093 & 1.034&&-0.134 & 1.071 + & @xmath138 & 0.019 & 1.000 & & 0.169&1.350 & & 0.110 & 1.131&&0.167 & 1.025 + 5 & @xmath139 & -0.012 & 1.000 & & -0.011 & 1.091 & & -0.036 & 1.113&&-0.031 & 1.004 + & @xmath140 & 0.003 & 1.000 & & -0.010&1.014 & & -0.009 & 1.013&&-0.016 & 1.037 + & @xmath141 & 0.003 & 1.000 & & -0.024&1.213 & & -0.001 & 0.944&&0.002 & 1.056 + & @xmath142 & 0.012 & 1.000 & & -0.011 & 1.203 & & 0.013 & 1.147&&0.016&1.007 + + & @xmath136 & -0.058 & 1.000 & & -0.099 & 1.128 & & -0.165 & 1.211 & & -0.119 & 1.018 + & @xmath137 & 0.022 & 1.000 & & 0.132 & 1.059 & & -0.033&1.067&&-0.100 & 1.009 + & @xmath138 & 0.020 & 1.000 & & 0.026 & 1.056 & & -0.055&1.029&&-0.032&1.035 + 10 & @xmath139 & -0.002 & 1.000 & & 0.009 & 1.007 & & -0.001&1.004&&-0.002 & 1.021 + & @xmath140 & -0.001 & 1.000 & & 0.004 & 0.999 & & -0.003&0.998&&0.002 & 1.048 + & @xmath141 & -0.003 & 1.000 & & 0.038 & 1.006 & & 0.002&1.003&&-0.003 & 1.025 + & @xmath142 & -0.001 & 1.000 & & 0.035 & 1.014 & & 0.000&0.998&&0.006 & 1.008 + + & @xmath136 & -0.011 & 1.000 & & -0.032 & 1.117 & & -0.030 & 1.071 & & -0.107 & 1.051 + & @xmath137 & -0.010 & 1.000 & & 0.077 & 1.019 & & -0.162&1.131&&-0.116 & 1.119 + & @xmath138 & 0.004 & 1.000 & & 0.042 & 1.107 & & 0.029&1.224&&0.108 & 1.103 + 20 & @xmath139 & -0.005 & 1.000 & & -0.016 & 1.008 & & -0.001&0.999&&-0.003&0.999 + & @xmath140 & 0.003 & 1.000 & & -0.017 & 1.015 & & 0.014&1.007&&-0.004 & 1.028 + & @xmath141 & 0.000 & 1.000 & & -0.029 & 1.073 & & -0.002&0.996&&-0.016 & 1.008 + & @xmath142 & 0.003 & 1.000 & & -0.036 & 1.088 & & 0.002&0.999&&-0.006 & 1.034 + in table [ table.infer1 ] , we present the simulation results of simulation study 1 for @xmath143 @xmath144 and @xmath142 , including the estimated relative bias and the estimated relative efficiency .
in general , it can be seen that the absolute bias obtained by the proposed model ( bqol ) when @xmath145 is much smaller than its competing models . in most cases ,
bqol was better than the other methods in terms of bias and the relative efficiency .
the results suggest that our method performs well compare to other approaches .
we see that the bayesian quantile regression approach for ordinal data ( bqror ) performs poorly compared to the other methods because it ignores the nature of the longitudinal data .
we also see that as @xmath128 increases , the bayesian logistic ordinal regression ( blor ) for longitudinal data yields low bias and more efficiency . in table [ table.infer2 ] , we present the simulation results of simulation study 1 for bayesian quantile regression methods when @xmath146 and 0.75 . again , in most cases , bqol was better than bqror in terms of bias and the relative efficiency .
cccccccccccccc & & + & & + @xmath128 & parameter & bias & eff & & bias & eff & & bias & eff & & bias & eff & + & @xmath136 & 0.128 & 1.000&&0.194&1.738&&-0.174&1.000&&-0.191 & 1.223 + & @xmath137 & -0.057&1.000&&-0.058&2.014&&-0.070&1.000&&0.091 & 2.031 + & @xmath138 & 0.016 & 1.000&&0.018&1.374&&0.083&1.000&&0.066 & 1.319 + 5 & @xmath139 & 0.035 & 1.000&&-0.030&1.238&&-0.003&1.000&&-0.014 & 1.011 + & @xmath140 & 0.018 & 1.000&&-0.039&1.098&&0.005&1.000&&-0.059 & 1.015 + & @xmath141 & 0.013 & 1.000&&-0.021&1.043&&0.010&1.000&&-0.040 & 1.056 + & @xmath142 & 0.020 & 1.000&&-0.025&1.051&&0.022&1.000&&-0.027 & 1.037 + + & @xmath136 & -0.132&1.000&&0.137&1.035&&-0.182&1.000&&-0.183 & 1.159 + & @xmath137 & -0.049&1.000&&0.064&1.836&&-0.054&1.000&&-0.074 & 1.037 + & @xmath138 & -0.003&1.000&&0.002&1.005&&0.011&1.000&&0.083 & 1.215 + 10 & @xmath139 & -0.016 & 1.000&&0.037&1.016&&-0.008&1.000&&0.020 & 1.117 + & @xmath140 & -0.009 & 1.000&&0.040&1.142&&-0.008&1.000&&0.031 & 1.027 + & @xmath141 & -0.008 & 1.000&&0.033&1.115&&-0.007&1.000&&0.004 & 1.006 + & @xmath142 & 0.019 & 1.000&&0.036&1.003&&-0.018&1.000&&0.028 & 1.014 + + & @xmath136 & -0.026&1.000&&-0.037&1.731&&0.013&1.000&&0.018 & 1.005 + & @xmath137 & -0.156 & 1.000&&0.210&1.079&&-0.177&1.000&&0.235 & 1.371 + & @xmath138 & 0.037 & 1.000&&-0.044&1.117&&0.013&1.000&&0.018 & 1.009 + 20 & @xmath139 & -0.014&1.000&&0.064&1.008&&-0.012&1.000&&-0.009 & 1.004 + & @xmath140 & -0.007 & 1.000&&0.005&1.116&&-0.009&1.000&&0.016 & 1.032 + & @xmath141 & 0.000 & 1.000&&0.030&1.138&&-0.008&1.000&&0.016 & 1.098 + & @xmath142 & 0.008 & 1.000&&0.070&1.129&&-0.006&1.000&&0.010 & 1.007 + cccccccccccccc & & + @xmath128 & parameter & mean & sd & & mean & sd & & mean & sd + & @xmath136 & -5.169&0.738&&-5.310&1.003&&-5.661&1.069 & & & + & @xmath137 & -10.028&0.771&&-9.838&1.035&&-9.682&1.036 & & & + & @xmath138 & 15.259&0.688&&15.392&0.935&&14.273&1.053 & & & + 5 & @xmath139 & -0.862&0.061&&-0.842&0.083&&-0.871&0.088 & & & + & @xmath140 & -0.249&0.066&&-0.237&0.064&&-0.231&0.071 & & & + & @xmath141 & 0.251&0.063&&0.261&0.079&&0.261&0.069 & & & + & @xmath142 & 0.849&0.073&&0.857&0.083&&0.881&0.093 & & & + + & @xmath136 & -5.132&0.935&&-5.116&0.998&&-5.842&1.014 & & & + & @xmath137 & -10.239&0.722&&-10.117&1.028&&-9.773&0.998 & & & + & @xmath138 & 15.024&0.776&&15.317&0.993&&15.235&1.037 & & & + 10 & @xmath139 & -0.848&0.064&&-0.842&0.083&&-0.863&0.087 & & & + & @xmath140 & -0.258&0.061&&-0.244&0.089&&-0.255&0.073 & & & + & @xmath141 & 0.258&0.073&&0.246&0.076&&0.259&0.079 & & & + & @xmath142 & 0.848&0.059&&0.845&0.037&&0.868&0.081 & & & + + & @xmath136 & -4.992&0.739&&-5.162&0.893&&-5.337&0.982 & & & + & @xmath137 & -10.039&0.812&&-10.337&0.926&&-10.241&1.013 & & & + & @xmath138 & 15.040&0.699&&14.893&0.794&&14.753&1.045 & & & + 20 & @xmath139 & -0.856&0.066&&-0.853&0.081&&-0.863&0.079 & & & + & @xmath140 & -0.263&0.085&&-0.257&0.063&&-0.247&0.066 & & & + & @xmath141 & 0.249&0.081&&0.249&0.066&&0.261&0.071 & & & + & @xmath142 & 0.842&0.076&&0.853&0.087&&0.857&0.081 & & & + the setup for this simulation study is the same as simulation 1 , except we sampled the latent variable @xmath40 as follows : @xmath147 this allows us to examine the performance of the proposed model in the case of random effects . in this simulation study , we only consider the performance of the bayesian methods for longitudinal data with ordinal outcome ( bqol and blor ) . in table
[ table.infer3 ] , we present the estimates of the parameters @xmath143 @xmath144 and @xmath142 , when @xmath145 and 0.25 . from table [ table.infer3 ] we can see that , our approach tends to give less biased parameter estimates for @xmath143 @xmath144 and @xmath142 compared to blor .
the convergence of the proposed gibbs sampling algorithm in this simulation study was monitored using the multivariate potential scale reduction factor ( mpsrf ) reported by @xcite . from figure ( [ fig1 ] ) , it can be observed that the mpsrf becomes stable and very close to 1 after about the first 5000 iterations for each quantile level under consideration .
hence , the convergence to the posterior distribution was quick and the mixing was good .
in this section , we consider a data set from the national institute of mental health schizophrenia collaborative ( nimhsc ) study previously analysed by @xcite .
the objective of this study is to assess treatment - related changes in illness severity over time . specifically , we studied item 79 ( imps79o ; severity of illness ) of the inpatient multidimensional psychiatric scale .
this item was measured on a seven point scale as in table [ seven1 ] : [ pointscale ] .
seven point scale for the nimhsc study [ seven1 ] [ cols="<,<",options="header " , ] table [ rd ] lists parameter estimations obtained using the bayesian methods ( bqol and blor ) .
the methods are assessed based on @xmath148 credible intervals and the deviance information criterion ( dic ; * ? ? ?
clearly , it can be seen that the credible intervals ( 95% ci ) obtained using the bqol when @xmath149 are generally shorter than the credible intervals obtained using the blor , suggesting an efficiency gain and stable estimation from the posterior distributions .
in addition , dic was computed for our model when @xmath149 and @xmath146 as well as for blor and the numbers were 3311.32 , 3615.48 and 3417.39 , respectively .
hence , under @xmath149 , model comparison using dic indicates that quantile ordinal models can give a better model fit compared to the bayesian logistic ordinal regression ( blor ) for longitudinal data .
this shows that the model uesd for the errors in ( [ eq3 ] ) is a working model with articial assumptions , employed on the outcome variable to achieve the equivalence between maximising sld and the minimising proplem in ( [ eq2 ] ) .
since bayesian quantile methods for estimating ordinal models with longitudinal data have not been proposed , yet .
this paper fills this gap and presents a random effects ordinal quantile regression model for analysis of longitudinal data with ordinal outcome of interest .
an efficient gibbs sampling algorithm was derived for fitting the model to the data based on a location - scale mixture representation of the skewed double exponential distribution .
the proposed approach is illustrated using simulated data and a real data example .
results show that the proposed approach performs well .
one of the most desirable features of the proposed method is its model robustness in the sense that makes very minimal assumptions on the form of the error term distribution and thus is able to accommodate non - normal errors and outliers , which are popular in many real world applications .
the full conditional distribution of each @xmath55 , denoted by @xmath150 is proportional to @xmath151 .
thus , we have @xmath152\big\ } \\ & \propto & v_{ij}^{-1/2 } \exp\big\ { -\frac{1}{2}\big[\frac{(l_{ij}-\mbox{\boldmath$x$}_{ij}'\mbox{\boldmath$\beta$}-\alpha_i)^2}{2}v_{ij}^{-1}+\big(\frac{1}{2 } \big ) v_{ij}\big]\big\ } \\ & \propto & v_{ij}^{-1/2 } \exp\big\ { -\frac{1}{2}\big[\varrho_1 ^ 2 v_{ij}^{-1}+\varrho_2 ^ 2 v_{ij}\big]\big\ } , \\\end{aligned}\ ] ] where @xmath153 and @xmath154 .
thus , the full conditional distribution of each @xmath55 is a generalized inverse gaussian distribution gig @xmath91 , where @xmath92 and @xmath93 . recall that if @xmath155 gig @xmath98 then the pdf of @xmath22 is given by @xcite @xmath156 where @xmath157 and @xmath158 is so called _ modified bessel function of the third kind _ " .
the full conditional distribution of each @xmath62 , denoted by @xmath159 is proportional to @xmath160 , where @xmath161 is the vector @xmath20 excluding the element @xmath62 .
thus , we have @xmath162 \big\ } , \\\end{aligned}\ ] ] where @xmath163 and @xmath164 .
then the full conditional distribution for @xmath62 is normal with mean @xmath165 and variance @xmath166 , where @xmath167 the full conditional distribution of each @xmath168 , denoted by @xmath169 is @xmath170 thus , the full conditional distribution of @xmath168 is a gig@xmath98 , where @xmath99 and @xmath100 .
the full conditional distribution of each @xmath43 , denoted by @xmath173 is proportional to @xmath174 .
thus , we have @xmath175 \big\ } , \\\end{aligned}\ ] ] where @xmath176 .
then the full conditional distribution for @xmath43 is normal with mean @xmath177 and variance @xmath178 , where @xmath179 the full conditional distribution of @xmath69 , denoted by @xmath180 , is proportional to @xmath181 .
thus , we have @xmath182 that is , the full conditional distribution of @xmath69 is a inverse gamma distribution .
the full conditional distribution of each @xmath40 , denoted by @xmath183 is proportional to @xmath184 @xmath185 .
thus , we have @xmath186 that is , the full conditional distribution of @xmath40 is a truncated normal distribution . at last ,
the full conditional posterior distribution of @xmath51 , denoted by @xmath187 is proportional to @xmath188 .
thus , we have @xmath189 following @xcite and @xcite , the full conditional distribution of @xmath51 is @xmath190 that is , the full conditional distribution of @xmath51 is a uniform distribution .
bottai , m. , frongillo , e. a. , sui , x. , oneill , j. r. , mckeown , r. e. , burns , t. l. , liese , a. d. , blair , s. n. , and pate , r. r. ( 2014 ) .
`` use of quantile regression to investigate the longitudinal association between physical activity and body mass index . '' , 22(5 ) : e149e156 .
lipsitz , s. r. , fitzmaurice , g. m. , molenberghs , g. , and zhao , l. p. ( 1997 ) .
`` quantile regression methods for longitudinal data with drop - outs : application to cd4 cell counts of patients infected with the human immunodeficiency virus . '' , 46(4 ) : 463476 .
montesinos - lpez , o. a. , montesinos - lpez , a. , crossa , j. , burgueo , j. , and eskridge , k. ( 2015 ) . `` genomic - enabled prediction of ordinal data with bayesian logistic ordinal regression . '' , 5(10 ) : 21132126 . | since the pioneering work by @xcite , quantile regression models and its applications have become increasingly popular and important for research in many areas . in this paper , a random effects ordinal quantile regression model is proposed for analysis of longitudinal data with ordinal outcome of interest .
an efficient gibbs sampling algorithm was derived for fitting the model to the data based on a location - scale mixture representation of the skewed double exponential distribution .
the proposed approach is illustrated using simulated data and a real data example .
this is the first work to discuss quantile regression for analysis of longitudinal data with ordinal outcome . |
biomembrane is mainly comprised of phospholipid molecules . due to the amphiphilic property of phospholipid
, these molecules assemble into bilayers in water with their hydrophobic tails facing inward , and form vesicles called liposomes @xcite . to investigate a model cell system , it is important to establish a method for creating giant vesicles ( gv ) with a diameter above 1 @xmath1 m , which are comparable in size to living cells @xcite .
phospholipid molecules usually form small unilamellar vesicles ( suv ) or multi - lamellar vesicles ( mlv ) , as described in the literature @xcite , through the use of sonication and/or mechanical agitation . on the other hand ,
a method for forming gvs has not been fully established , and especially the fundamental mechanism that is involved in the formation of gvs is not yet understood @xcite .
a natural swelling method , in which gvs are spontaneously formed by hydrating multi - stacked dry phospholipid layers on a solid substrate , is quite useful for creating a model cell because no physical or chemical stresses are applied to gvs @xcite . in this method ,
each lipid bilayer should peel off from the stack of bilayers on a substrate and form a vesicle .
therefore , it is expected that the mechanism of this natural swelling is closely associated with the unbinding transition @xcite .
it is to be noted that this mechanism should be discussed as a kinetic unbinding process that occurs through hydration , although the unbinding transition is commonly discussed in the framework of a temperature - induced transition @xcite .
recently , the present authors verified the conditions needed to effectively form giant vesicles from dry phospholipid films through natural swelling @xcite .
the terrace - like morphology of dry phospholipid film and the adequate thermal fluctuation of bilayers in the liquid crystalline phase are necessary for obtaining gvs . on the other hand ,
a rough morphology is seen for dry lipid films when the lipid is in the gel phase , and no gvs are formed under this condition .
these results indicated that the terrace - like morphology of dry lipid film and the liquid - crystalline phase are essential for the formation of gvs .
although the initial and final structures in the vesicle formation are clarified , the hydration process that leads to the formation of gvs from dry films remains to be solved .
time - resolved small angle x - ray scattering ( tr - saxs ) is a powerful tool for the in situ observation of structural changes on the order of nm .
several tr - saxs studies have been performed on the kinetics of vesicle formation or the unbinding transition .
thimmel et al .
investigated the process of swelling from a stack of lipid bilayers .
however , they examined this process by increasing temperature ( not swelling with water ) @xcite .
hartung et al . and pabst et al .
have shown that the intensity of bragg peak due to the regular stacking of bilayers decrease slowly with time when well - aligned phospholipid bilayers on a solid substrate are hydrated by excess water @xcite .
although their results suggested that the bilayers peel off from the substrate and form vesicles , the process of hydration has not been clarified yet .
further , the kinetics has not been discussed quantitatively @xcite . in the present letter ,
we report the hydration process of phospholipid dry film in detail as examined by tr - saxs .
the process was found to have two stages in the formation of giant vesicles .
water molecules penetrate between lipid bilayers in the early stage , and each bilayer peels off from the stack of bilayers in the late stage .
the kinetics of the late stage is discussed quantitatively in terms of kramers formulas .
1,2-dioleoyl - sn - glycero-3-phosphocholine ( dopc , sigma - aldrich ) and 1,2-dipalmitoyl - sn - glycero-3-phosphocholine ( dppc , wako pure chemical industries ) were obtained in powder form and used without further purification .
dopc is in the liquid - crystalline phase in both dry and wet conditions at room temperature , while dppc is in the gel phase under the same conditions @xcite .
these lipids were dissolved in an organic solvent composed of dehydrated chloroform and dehydrated methanol ( 2:1 v / v ) as mother solutions ( both from nacalai tesque ) with molecular sieves ( nacalai tesque ) . these mother solutions ( 10 mm ) were stored at -30 @xmath2 before use . each mother solution
( 10 @xmath1l ) was dropped onto a capillary glass tube that had an outer diameter of 1.5 mm . by evaporating the organic solvent overnight in a vacuum ,
dry phospholipid film 8 mm long remained on the surface of the glass tube . under these conditions ,
about a hundred lipid bilayers were stacked on the glass with the total thickness being 400 - 500 nm which was estimated by x - ray reflectivity measurement @xcite . to observe the hydration process in detail ,
tr - saxs experiment was carried out at bl40b2 , spring-8 , jasri ( the japan synchrotron radiation research institute ) , japan .
the wavelength of the x - ray was 1 @xmath0 and the diameter of x - ray at the sample was 0.4 mm .
the scattered x - ray beam was detected by ccd .
the sample - to - detector distance was about 1 m , which was calibrated using a standard sample ( lead stearate ) .
the observed momentum transfer ranged over @xmath3 .
the samples were kept in a special cell , as shown in fig .
the capillary glass tube with dry phospholipid film was hung in an acrylic cell with windows for x - ray made of kapton film .
pure water ( milliq ) was automatically supplied into the cell and the water hydrated the dry lipid film .
tr - saxs measurements were started just before hydration .
all the measurements were performed at room temperature .
the time - dependence of the saxs profiles in the hydration of the dopc film are shown in figs .
[ fig2 ] and [ fig3 ] . the early stage ( 0 - 100 s after hydration ) and the late stage ( 100 - 600 s ) of hydration are shown separately in fig .
[ fig2 ] and fig .
[ fig3 ] , since the time scales of these processes are different , i.e. , the time - step of the measurement in the early stage was 72.7 ms and that in the late stage was 727 ms .
a large difference in behavior was noted between the early stage and the late stage .
a sharp peak corresponding to the regular stacking of lipid bilayers tends to shift to a lower-@xmath4 in the early stage . before hydration ( 0
s ) , the peak is at @xmath5 . within 0 - 20 s after hydration , the intensity of the bragg peak from the regular stacking of dry lipid layers decreases and the position of the peak shifts slightly and a broadening occurs . next , from 20 to 100 s , the peak position moves inverse - exponentially down to @xmath6 , which corresponds to the lamellar repeat distance @xmath7 , and the peak intensity increases simultaneously .
the value @xmath7 is 1 - 2 @xmath0 larger than the lamellar repeat distance of dopc mlv @xcite .
although the reason on the difference is not clear yet , it may be attributed to any experimental errors , or an effect of the substrate , or some intrinsic origins as the vapor pressure paradox @xcite .
anyway , it is apparent that the final value of @xmath8 is almost the same as that of mlv . in this stage ,
the regularity of the bilayer stacking initially breaks , and then recovers at about 100 s after hydration .
the result means that , in the early stage , the lipid film on a solid substrate swells and reaches a quasi - equilibrium state , due to the penetration of water molecules among lipid bilayers .
the integrated intensity of the peak corresponding to the number of layers in the film does not change with time ( data not shown ) .
this indicates that multi - stacked dopc bilayers do not peel off from the substrate in this process .
figure [ fig3 ] shows the time - dependence of the saxs profile in the late stage , from 100 s to 600 s after hydration began , for the dopc film . in this process , the lamellar repeat distance @xmath8 for the lipid bilayers on the substrate remains almost constant and the intensity of the peak decreases monotonically .
in addition , the peak width does not change and the integrated intensity decreases as the peak intensity .
therefore , the peak intensity corresponds to the number of stacked bilayers .
the time dependence of the number of stacked bilayers is shown in fig .
[ fig4 ] ( b ) ( @xmath9 ) , which estimated from the number of stacked bilayers in the initial condition ( dry film ) .
this behavior suggests that the lipid bilayers are unbound and peel off from the stack in this stage .
based on these results , we can conclude that the hydration process of dry dopc film proceeds in two stages . in the early stage from 0
s to about 100 s after hydration begins , water molecules penetrate between bilayers and the multi - lamellar structure reaches a quasi - stable state . in the following stage , the lipid bilayers gradually peel off from the multi - lamellar stack ( maybe one by one ) , and form giant vesicles .
the behavior of the late stage is consistent with the results by hartung et al . and
pabst et al .
@xcite . however , the present process is much faster than their cases .
it should be noticed that the early stage is observed for the first time . to compare the hydration process with that in the gel phase
, dppc dry film was investigated by the same method . in this case , only the early stage , a bragg peak corresponding to the regular stacking of bilayers shifts to lower-@xmath4 , is observed .
the final repeat distance @xmath8 is 66 , which is also 1 - 2 @xmath0 larger than the lamellar repeat distance of mlv @xcite .
the peak intensity does not change until 350 sec after hydration begins .
this means that the number of stacked bilayers does not change and no bilayers peel off from the stack on the substrate .
this result corresponds to the fact that no gvs is formed by the hydration of dry dppc film in the gel phase at room temperature @xcite .
the equilibrium distance @xmath8 of lipid bilayers has been interpreted in terms of the balance of interactions between bilayers .
since the present systems do not include salts , van der waals attractive interaction and the repulsive force due to the binding of water molecules at hydrophilic headgroups of lipid molecules , and steric repulsion that originates from thermal fluctuations of lipid bilayers should be taken into consideration @xcite .
therefore , the free energy per unit area of bilayers depending on @xmath8 can be written as , @xmath10 @xmath11\ ] ] @xmath12 @xmath13 where @xmath14 is the hamaker constant , @xmath15 is the thickness of a bilayer , @xmath16 is a prefactor of @xmath17 , @xmath18 is the decay length of a hydration layer , @xmath19 is the bending rigidity of a bilayer , and @xmath20 is the boltzmann constant @xcite .
this free energy profile is calculated for a layer in regularly multi - stacked bilayers , and is also applicable to the bilayers in mlv @xcite .
the calculated free energy profile of dopc multi - layers as a function of @xmath8 is shown in fig .
[ fig4 ] ( a ) ( dashed line .
the values of the coefficients are taken from @xcite ) .
the observed early stage is interpreted as a relaxation process to the local minimum at @xmath21 and the late stage is a diffusion ( unbinding ) process from the local minimum . in this case , the energy barrier @xmath22 ( energy difference between the local minimum and maximum ) is @xmath23 , and the unbinding of bilayers due to thermal fluctuation , i.e. peeling off of bilayers , should be too slow to observe .
thus , each phospholipid bilayer in multi - stacked bilayers is bound tightly as in a stable state .
it should be noted that the present authors previously showed that a collaboration with other repulsive forces is necessary for unbinding @xcite .
the case of the outermost layer of a stack of lipid bilayers could be different from that in the inner layers , since one side of the outermost bilayers is not restricted by other bilayers , the edge of the layer is exposed to water , and water can flow into bilayers .
thus , here we assume that the effective helfrich repulsion of the outermost bilayer is slightly larger than the theoretical value : @xmath24 where @xmath25 accounts for the effective repulsion of the outermost bilayer ( @xmath26 ) .
such an assumption allows us to explain the experimental evidence of the spontaneous unbinding to form vesicles in the late stage .
the one - dimensional smoluchowski equation is @xmath27 where @xmath28 . @xmath29 and @xmath30 are a resistance coefficient and the mass of a lipid bilayer , respectively .
the outermost bilayer is assumed to be disc - shaped with diameter @xmath31 , and the resistance coefficient @xmath29 is derived as @xmath32 using stokes law , where @xmath33 is the viscosity of water @xcite .
in the present explanation , it is reasonable to assume that the diffusion behavior of the outermost bilayer remains almost constant because the number of stacked bilayers is sufficient within the range of this experiment , i.e. , we can consider the next outermost bilayer after a neighboring bilayer has left to form a gv .
this assumption is consistent with the experimental evidence that the peak width , which corresponds to the distribution of the mean repeat distance of the stacked bilayers , remained constant in the late stage and the peak height decreased monotonically at an almost constant rate .
thus , we can calculate the smoluchowski equation analytically by kramers method @xcite and the number of bilayers passing through the energy barrier per unit time is @xmath34 where @xmath35 and @xmath36 are curvature radii at the minimum and maximum of the free energy , respectively .
we can obtain @xmath35 , @xmath36 , @xmath22 from the calculated free energy . with these coefficients , by substituting @xmath37 $ ] , @xmath38 $ ] , @xmath39 $ ] , and @xmath40 $ ] , we can estimate @xmath41 numerically . in the case of @xmath42 ,
the time - dependence of the number of stacked bilayers calculated with this assumption is consistent with the experimental result , which is depicted in fig .
[ fig4 ] ( b ) by the solid line . the free energy profile in this case is shown as a solid line in fig .
[ fig4 ] ( a ) . from this estimation
, it is concluded that a few - percent increase in the helfrich repulsion only for the outermost bilayer triggers the unbinding of a bilayer from a stack of bilayers .
the same discussion can be applied in the case of the dppc .
since it is in the gel phase at room temperature , the helfrich repulsion is much smaller than in the case of dopc ; @xmath22 is @xmath43 even if the effective helfrich repulsion of the outermost bilayer is assumed to be two times larger than the normal helfrich repulsion ( @xmath44 ) .
this means that bilayers can not be spontaneously unbound by thermal fluctuation .
we summarize the process of natural swelling from a stack of lipid bilayers on a solid substrate as follows . in the early stage of the hydration ,
water molecules penetrate within the stack of lipid bilayers and the multi - stacked bilayers reach a quasi - equilibrium state . following this process
, bilayers gradually peel off and form vesicles . in this late stage ,
only the outermost bilayer is unbound due to thermal fluctuation and peels off from the stack .
this process can occur only when the lipid is in the liquid - crystalline phase with a terrace - like morphology on a solid substrate . in the gel phase ,
lipid bilayers remain stacked and gvs are not formed .
the authors are indebted to dr .
k. inoue and dr .
s. sasaki at jasri and dr .
m. takenaka and dr .
h. kitahata at kyoto univ . for their supports with the experiment and the data analysis .
the saxs experiments were performed under the approval of jasri ( no .
2004b0520-nl2b - np ) .
this work was supported by a grant - in - aid for scientific research ( no .
17540382 ) , by a grant - in - aid for the 21st century coe `` center for diversity and universality in physics '' from the ministry of education , culture , sports , science and technology ( mext ) of japan , and by the yamada science foundation . | the hydration process of multi - stacked phospholipid bilayers to form giant vesicles was investigated by time - resolved small angle x - ray scattering .
the hydration of lipids in the liquid - crystalline phase was found to proceed in two stages . in the early stage , until about 100 s after hydration , lipid bilayers on a solid substrate swell about 12 @xmath0 and reach a quasi - stable state . in the late stage , after several hundreds of seconds , lipid bilayers gradually peel off from the stack . on the other hand , in the case of a lipid in the gel phase , only the early stage is observed .
these behaviors correspond to the capacity for the giant vesicle formation depending on the lipid phases .
the kinetics of the peeling - off process is discussed in terms of kramers formulas . |
it has long been known @xcite that bounded plasmas support a special type of collective electrostatic and electromagnetic excitations
the surface plasma waves whose field and energy are concentrated near , and propagate along , the plasma boundaries .
the surface waves ( sw ) in various ( classical ) plasmas have been extensively studied , and they have found many applications ( see ref .
@xcite and references therein ) . yet
recently , due to a remarkable progress in nanotechnology , the interest in surface waves supported by various nanostructures ( especially metallic structures such as thin films and tiny metallic particles ) , and in their interaction with light , has been revived .
it is believed that light - induced surface excitations in such structures may offer a route to faster , smaller , and more efficient electronics , as well as new technology @xcite .
in particular , one could note such recent advents in the new and promising area of quantum nanoplasmonics as the development of the concept of spaser @xcite followed by its further development into a lasing spaser @xcite , and the experimental demonstration of a spaser - based nanolaser @xcite . in view of these developments , understanding the properties of surface waves in various metallic ( and semiconductor ) structures , bounded by vacuum or dielectric , is thus important .
such understanding requires using models describing the dynamic response of such structures to self - consistent electromagnetic fields , that also appropriately take into account the relevant quantum effects arising from quantum nature of free charge carriers in such structures and , in general , from their ( quantum ) interaction with each other and with the underlying ion lattice .
these quantum effects may significantly alter the properties of the surface waves ; see , e.g. , refs .
@xcite .
recently , the dispersion relation of surface waves in one of the basic structures modeling a nanoplasmonic device a semi - bounded collisionless quantum plasma with degenerate electrons was obtained in ref .
@xcite using the quantum fluid theory ( qft ) approach @xcite . in the electrostatic limit ,
the authors of ref .
@xcite obtained for the frequency of surface waves in this structure : @xmath5 where @xmath6 , @xmath7 , @xmath8 , @xmath9 is the electron plasma frequency , @xmath10{3\pi^2 n_e}/m_e$ ] is the electron fermi velocity , @xmath11 is the thomas - fermi length , @xmath0 is the sw frequency , @xmath12 is the sw wave vector along the plasma boundary , @xmath13 , @xmath14 and @xmath15 are electron charge , mass and number density , respectively , and @xmath16 is the reduced planck constant .
however , the validity of the dispersion relation ( [ eq : lazar_disp ] ) obtained in ref . @xcite ( as well as of the similar dispersion relation obtained in ref .
@xcite ) is limited by the validity of the qft approach itself @xcite , and is restricted only to long waves , @xmath17 . moreover
, the qft approach by its nature completely ignores the purely kinetic effect of collisionless damping of surface waves , which is known to be significant , e.g. , for potential sw in warm electron plasma at short wavelengths @xcite . to overcome these limitations , a kinetic approach to the problem of sw in semi - bounded plasma with degenerate electrons
is needed . in this paper
, we study potential surface waves in semi - bounded collisionless quantum plasma with fermi - degenerate electrons , using the initial value problem solution for the semi - classical vlasov - poisson system @xcite .
we obtain the dispersion and collisionless damping rate of these waves , valid for both long and short wavelengths , and report on a surprising result that these waves remain weakly damped for all values of @xmath18 ( i.e. , for all wavelengths ) , unlike , e.g. , in plasma with maxwellian electrons in which the sws are weakly damped only for @xmath17 and quickly become strongly damped as @xmath18 increases .
we also report a nontrivial nonmonotonic dependence of the damping rate on @xmath18 , featuring a maximum damping at @xmath19 ( corresponding to the surface wave length of @xmath3 ) .
we consider a uniform nonrelativistic quantum plasma consisting of mobile electrons with charge @xmath13 , mass @xmath14 and number density at equilibrium @xmath20 , and immobile uniform background of singly charged ions with number density @xmath21 that neutralizes the electron charge at equilibrium .
the plasma is assumed to be confined to the region @xmath22 by a sharp perfectly reflecting wall located at @xmath23 .
we will be interested in evolution of an initial perturbation to the equilibrium state of the system s electronic component applied at time @xmath24 . in general
, this system of many interacting quantum particles ( electrons ) can be described by the quantum analog of the bogoliubov - born - green - kirkwood - yvon ( bbgky ) hierarchy @xcite of equations for the electron @xmath25-particle quantum distribution functions @xmath26 ( also called the @xmath25-particle wigner functions ) , where @xmath27 , @xmath28 and @xmath29 are the @xmath30-dimensional vectors denoting the sets of coordinates and canonical momenta of system particles , and @xmath31 is the total number of electrons in the system . in this hierarchy , each of the equations for the @xmath25-particle quantum distribution function contains the @xmath32-particle distribution function , making the whole set of @xmath31 equations coupled , and thus prohibitively large to solve . in practice , however , this hierarchy can be truncated by making a physically justified assumption about correlation of particles due to their interaction . in particular , for a system of weakly interacting particles , with a small plasma coupling parameter @xmath33 ( here @xmath34 is the characteristic potential energy of particle interaction , and @xmath35 is the characteristic kinetic energy of plasma particles ) , the two- and higher - order particle correlations can be ignored , leading to the collisionless mean - field approximation @xcite involving only one equation for the one - particle quantum distribution function @xmath36 , where @xmath37 and @xmath38 are now the 3-dimensional vectors of particle coordinate and momentum . in the quasi - classical approximation , with the effect of
quantum recoil ignored ( see sec . [
sec : applicability ] ) , this equation reduces to the vlasov equation for the one - particle classical distribution function @xmath39 , where @xmath40 is electron velocity . for a system of particles with electrostatic interaction , in the chosen geometry , the vlasov equation for @xmath41 ( where @xmath42 and @xmath43 are , respectively
, the components of @xmath37 and @xmath44 parallel to the boundary ) reads @xmath45 where the electrostatic potential @xmath46 is defined by the poisson s equation @xmath47 .
\label{eq : poisson}\ ] ] in the absence of fields , the equilibrium distribution function of plasma electrons @xmath48 is defined by the pauli s exclusion principle for fermions , and for low electron temperatures @xmath49 ( where @xmath50 is the electron temperature , @xmath51 is the electron fermi energy ) it becomes @xmath52 corresponding to fully degenerate electron distribution .
the condition of specular reflection of plasma electrons off the boundary at @xmath23 implies @xmath53 we now introduce a small initial perturbation @xmath54 to the equilibrium electron distribution function @xmath55 , @xmath56 , and study the resulting evolution of the system s charge density @xmath57 $ ] , and hence of the electrostatic potential @xmath46 defined by ( [ eq : poisson ] ) . introducing the dimensionless variables @xmath6 , @xmath58 , @xmath59 , @xmath60 , @xmath61 , @xmath11 , and following guernsey @xcite , the solution of the formulated initial value problem for @xmath62 with the boundary condition ( [ eq : boundary ] ) is @xmath63 where @xmath64 the integration in ( [ eq : inv_laplace ] ) is performed in complex @xmath65 plane along the horizontal contour that lies in the upper half - plane @xmath66 above all singularities of the function @xmath67 .
the function @xmath67 , defined as the laplace transform of @xmath68 : @xmath69 is given by @xmath70\int_{-\infty}^{+\infty}{\frac{dk_x'}{{k^\prime}^2\ \varepsilon(\omega , k')}\int{d^3\mathbf{v}\frac{g(\mathbf{v},\mathbf{k^\prime})}{\omega-\sqrt{3}\mathbf{k'\cdot v } } } } , \label{eq : rho(w , k ) } \end{aligned}\ ] ] where the fourier transforms @xmath71 and @xmath72 of the ( dimensionless ) initial perturbation are defined by @xmath73 with @xmath74 where @xmath75 , @xmath76 , @xmath77 , @xmath78 , and @xmath79 .
the functions @xmath80 and @xmath81 in ( [ eq : rho(w , k ) ] ) are defined ( for @xmath82 ) as follows : @xmath83 with @xmath84 for fully degenerate plasma with electron distribution ( [ eq : f0 ] ) , the function @xmath80 becomes @xcite : @xmath85 , \label{eq : epsilon_degenerate}\ ] ] where @xmath86 is the principal branch of the complex natural logarithm function .
note that the solution ( [ eq : rho(w , k ) ] ) differs from the corresponding solution of the transformed vlasov - poisson system for infinite ( unbounded ) uniform plasma only in the second term involving @xmath81 ; indeed , this term appears due to the boundary at @xmath23 .
the definition ( [ eq : laplace ] ) of the function @xmath87 of complex @xmath65 has a sense ( i.e. , the integral in ( [ eq : laplace ] ) converges ) only for @xmath82 .
yet the long - time evolution of @xmath88 is obtained from ( [ eq : inv_laplace ] ) by displacing the contour of integration in complex @xmath65 plane from the upper half - plane @xmath82 into the lower half - plane @xmath89 @xcite .
this requires the definition of @xmath87 to be extended to the lower half - plane , @xmath90 , by analytic continuation of ( [ eq : rho(w , k ) ] ) from @xmath82 to @xmath90 .
hence , the functions @xmath91 @xmath80 , and @xmath81 that make up the function @xmath87 , must also be analytically continued into the lower half - plane of complex @xmath65 , thus extending their definition to the whole complex @xmath65 plane . with thus continued functions , the contributions to the inverse laplace
transform ( [ eq : inv_laplace ] ) are of three sources @xcite : 1 .
contributions from the singularities of @xmath92 in the lower half of complex @xmath65 plane ( defined solely by the initial perturbation @xmath71 ) ; with some simplifying assumptions about the initial perturbation @xcite these contributions are damped in a few plasma periods and can be ignored .
contribution of singularities of @xmath93 in the lower half of complex @xmath65 plane , of two types : ( i ) residues at the poles of @xmath93 , which give the volume plasma oscillations @xcite , and ( ii ) integrals along branch cuts ( if any ) of @xmath93 in the lower half - plane of complex @xmath65 , which can lead to non - exponential attenuation of the volume plasma oscillations @xcite .
3 . contribution of singularities of @xmath94 in the lower half of complex @xmath65 plane , of two types : ( i ) residues at the poles of @xmath94 , corresponding to the surface wave solutions of the initial value problem in the considered system @xcite , and ( ii ) integrals along branch cuts ( if any ) of @xmath94 in the lower half - plane of complex @xmath65 , which will be discussed elsewhere . in this paper , we will only consider the surface wave solutions due to the contribution of the residues at the poles of @xmath94 .
the dispersion and damping properties of these surface wave solutions are defined by the dispersion relation for plasma surface waves @xcite @xmath95 which in case of stable plasma ( with velocity distribution of electrons having only one maximum ) only has non - growing solutions @xmath96 with @xmath90 @xcite . out of all complex - valued solutions of ( [ eq : zeta=0 ] ) ,
@xmath97 ( here @xmath98 is the dimensionless frequency , and @xmath99 is the dimensionless damping rate of the surface wave ) , only those corresponding to _ weakly damped _
surface waves , with @xmath100 , are of physical interest .
such solutions can be obtained by solving , instead of the actual dispersion equation ( [ eq : zeta=0 ] ) , the following set of approximate equations ( that follow from ( [ eq : zeta=0 ] ) for @xmath100 ) : @xmath101 & = & 0,\text { yielding } \omega_s=\omega_s(k_\parallel)\in\mathbb{r } , \label{eq : omega_s}\\ \gamma_s & = & -\left.\frac{{\rm im}\left[\zeta(\omega_s , k_\parallel)\right]}{\partial{\rm\ re}\left[\zeta(\omega_s , k_\parallel)\right]/\partial\omega_s}\right|_{\omega_s=\omega_s(k_\parallel ) } , \label{eq : gamma_s}\end{aligned}\ ] ] which only involve @xmath81 as a function of _ real _ @xmath65 .
thus , instead of performing analytic continuation of @xmath81 from @xmath82 to @xmath89 , it suffices to only analytically continue @xmath81 from @xmath82 onto the real axis @xmath102 of the complex @xmath65 plane , thus defining it for @xmath103 .
this analytic continuation of @xmath81 , required for the numerical solution of eqs ( [ eq : omega_s])([eq : gamma_s ] ) , is discussed in appendix [ sec : zeta ] .
the numerical solution of eqs ( [ eq : omega_s ] ) and ( [ eq : gamma_s ] ) for a given @xmath104 , with @xmath81 defined for @xmath103 as discussed in appendix [ sec : zeta ] , yields the dispersion @xmath105 and damping @xmath106 of surface waves ; below we discuss them in greater detail .
the dispersion curve of surface waves @xmath105 is shown in fig .
[ fig : dispersion ] . the sw frequency increases monotonically with @xmath18 , starting from the well - known long - wavelength limit of @xmath107 at @xmath108 ( which also corresponds to the cold - plasma limit of sw frequency ) , and increasing as @xmath109 for @xmath17 .
we note the discrepancy between ( [ eq : omega_kz<<1 ] ) and the corresponding small-@xmath18 asymptote of lazar et al.s result ( [ eq : lazar_disp ] ) of ref .
@xcite : @xmath110 .
this discrepancy can not be attributed to the effect of quantum recoil , which is ignored here but included in the qft model of ref .
@xcite , as the quantum recoil only gives the higher - order contribution @xmath111 to the dispersion ( [ eq : lazar_disp ] ) at small @xmath18 .
in fact , the mentioned discrepancy is due to the error in the coefficient of the classical pressure gradient term used in ref .
@xcite : instead of the three - dimensional pressure gradient term @xmath112
@xcite , the one - dimensional term @xmath113 was used . with
this error corrected , the qft model of ref .
@xcite yields the small-@xmath18 asymptote of @xmath114 that matches our asymptote ( [ eq : omega_kz<<1 ] ) . at short wavelengths , @xmath115 ,
the sw dispersion can be approximated as ( see appendix [ sec : dispersion_kz>>1 ] ) @xmath116\right ) , \label{eq : dispersion_kz>>1}\ ] ] and quickly approaches the line @xmath117 that corresponds to the dispersion of _ volume _ zero sound mode in degenerate fermi gas @xcite propagating along the boundary , with @xmath118 .
note that for @xmath115 the sw frequency @xmath119 becomes practically indistinguishable from the corresponding frequency @xmath120 of the volume plasma wave propagating along the boundary ( with @xmath118 ) @xcite : @xmath121\right).\ ] ] ( color online ) the dispersion curve of surface waves @xmath105 ( solid blue line ) , and the asymptotes : the long - wave limit @xmath122 ( dotted line ) , the long - wave asymptote ( [ eq : omega_kz<<1 ] ) ( dot - dashed blue line ) , and the short - wave asymptote ( [ eq : dispersion_kz>>1 ] ) corresponding to zero sound propagating along the boundary ( dashed red line).,width=384 ] the @xmath18-dependence of the surface wave damping rate @xmath123 is shown in fig .
[ fig : damping ] . in the long - wave limit @xmath17 , the damping rate increases linearly with @xmath18 as @xmath124 similar to the damping rate of sw in plasma with maxwellian electrons @xcite .
however , as @xmath18 increases , an important difference between sw damping in plasma with degenerate electrons and in plasma with maxwellian electrons becomes obvious .
in maxwellian plasma , the sw damping rate increases monotonically with @xmath18 , quickly exceeding the sw frequency , so that the sw become strongly damped at short wavelengths @xcite . in degenerate electron plasma , however , the sw damping rate has a non - monotonic dependence on @xmath18 , as seen in fig .
[ fig : damping ] : at small @xmath18 , the damping rate increases almost linearly with @xmath18 , reaching a distinct maximum of @xmath125 at @xmath19 , and then _ decreases _ monotonically with @xmath18 for @xmath126 , approaching zero at large @xmath18 . most importantly , as seen from fig .
[ fig : damping ] , the sw damping rate in degenerate electron plasma remains small ( @xmath127 ) at all values of @xmath18 , and hence the electrostatic surface waves in degenerate collisionless electron plasma are weakly damped at all wavelengths ( yet the surface waves with @xmath128 are preferentially damped , as compared to other wavelengths ) .
( color online ) the absolute value of surface wave damping rate @xmath123 ( solid blue line ) , and the long - wave asymptote ( [ eq : gamma_kz<<1 ] ) ( dashed line ) , as functions of @xmath18.,width=384 ] the following note on the nature of damping of surface waves in collisionless plasma is in order .
it is well known that in a maxwellian plasma the collisionless ( landau ) damping of surface plasma waves is not exponentially small and significantly exceeds that of the volume plasma waves in unbounded plasma , even at small value of @xmath18 when their phase velocity greatly exceeds the electron thermal velocity @xcite .
the reason is that , as seen from eq .
( [ eq : zeta ] ) , the surface wave field is essentially a result of plasma response to a collection of `` virtual '' plasma waves [ i.e. , having frequency @xmath65 and wave vector @xmath129 , but not having a dispersion relation @xmath130 , unlike the `` real '' plasma waves with dispersion relation @xmath130 following from @xmath131 , with the same frequency @xmath132 but with different wave vectors @xmath133 with absolute values @xmath134 ranging from @xmath118 ( for @xmath135 ) to @xmath136 ( for @xmath137 ) .
each of these virtual waves with its own @xmath133 interacts resonantly with plasma electrons whose velocities along @xmath133 is close to the phase velocity of the virtual wave ( and such electrons always exist in maxwellian plasma ) , and thus is subject to a non - zero landau damping . as a result ,
the surface wave , consisting of both weakly damped virtual waves ( with @xmath138 , where for maxwellian plasma @xmath139 , @xmath140 is the electron debye length ) and strongly damped virtual waves ( with @xmath141 ) , interacts resonantly with a significant part of the plasma electron distribution , and hence is strongly damped , compared to the `` real '' plasma waves with the same wavelength . in the degenerate plasma considered here
, the difference in damping of surface and the `` real '' volume plasma waves is even more striking .
indeed , the `` real '' volume waves in such plasma , defined by the dispersion relation @xmath142 with @xmath80 given by eq .
( [ eq : epsilon_degenerate ] ) , are not subject to landau damping at all , as their phase velocity exceeds the maximum possible velocity of degenerate plasma electrons , @xmath143 ( here @xmath144 , @xmath11 is the electron fermi length ) , for all values of @xmath129 @xcite .
the `` virtual '' plasma waves are , however , still subject to landau damping by plasma electrons if their phase velocity is less than the electron fermi velocity , @xmath145 , i.e. , if their @xmath146 is large enough , @xmath147 ; such virtual waves always give a contribution to the surface wave field , as @xmath146 in ( [ eq : zeta ] ) attains arbitrarily large values .
this also results in the finite , albeit small ( yet infinitely large compared to the zero damping of the `` real '' volume plasma waves ) landau damping of surface waves in degenerate plasma .
the collisionless mean - field approximation used here is justified for a system of weakly interacting particles , when two - particle correlations ( and higher - order particle correlations as well ) can be ignored .
physically , this corresponds to a system in which the collective effects dominate over the effects of particle collisions , which happens when the plasma coupling parameter @xmath148 is small ( see sec . [
sec : model ] ) . with @xmath149 and @xmath150 for degenerate electrons , this implies that the collisionless mean - field approximation for degenerate electron plasma is justified when @xmath151 moreover , when the condition ( [ eq : kappa<<1 ] ) is satisfied , the effect of quantum recoil on the dispersion properties of electrostatic plasma waves is also negligible , at least for @xmath152 @xcite , and the only quantum effects come from the fermi - dirac statistics of plasma electrons . thus the mean - field quasi - classical kinetic model ( [ eq : vlasov ] ) , in which particle correlations and quantum recoil are neglected , is internally consistent under the condition ( [ eq : kappa<<1 ] ) .
we note that for degenerate electron plasma , the coupling parameter @xmath153 scales as @xmath154 and thus decreases as the plasma density @xmath21 increases ; hence the condition ( [ eq : kappa<<1 ] ) is satisifed for sufficiently dense plasma , with @xmath155 , where @xmath156 is the effective electron mass in such plasma .
however , as @xmath21 increases , the electron fermi velocity increases as @xmath157 and may become comparable to the speed of light @xmath158 , in which case the relativistic effects may become important ; to avoid this , we also require that @xmath159 .
thus the nonrelativistic mean - field quasi - classical approximation ( [ eq : vlasov ] ) is valid for quantum plasmas with degenerate electrons with densities @xmath160 the condition ( [ eq : joint_condition ] ) may be satisfied for some semiconductors , as pointed out in ref .
@xcite . for electron gas in metals , however ,
the condition ( [ eq : kappa<<1 ] ) ( as well as the first part of the condition ( [ eq : joint_condition ] ) ) is not satisfied ( in fact in metals @xmath161 ) , which suggests that in metals the collisionless mean - field approximation ( as well as the qft model used in ref .
@xcite , that is derived from the mean - field kinetic model @xcite ) is formally not justified , as the two- and higher - order particle correlations ( due to their coulomb and exchange interactions ) become important .
instead , a kinetic model of moderately coupled plasma , accounting for particle correlations and quantum recoil , should be developed a difficult task , given the lack of a small parameter characterizing the electron interactions in metals . nevertheless , there are reasons , discussed below , to believe that our idealized model can still work well ( at least qualitatively ) in metals , despite being formally unjustified there . in the experimental work of watanabe @xcite ,
the relation between energy loss and scattering angle of 25 kev electrons passing through thin metallic films was measured , and it was found , in particular , that the value of energy loss increases with the scattering angle of electrons . an empirical formula representing this relation for be , mg , and al films was found to be in _ good agreement _ with the corresponding theoretical formula that follows from the dispersion relation of electron plasma waves ( excited by the energetic electrons traversing the metallic film ) , derived using the mean - field kinetic model of free electrons in metal , despite this model being formally unjustified in metals due to violation of ( [ eq : kappa<<1 ] ) there . moreover ,
the effect of quantum recoil , although detectable , is small compared to the effect of fermi - dirac statistics in the whole range of measured scattering angles , especially at small scattering angles ( corresponding to small values of plasma wave number ) .
these facts can be perceived as an experimental evidence supporting the claim that the mean - field quasi - classical kinetic model can work well even in metals , despite there being @xmath162 .
there is some theoretical evidence of this as well . in ref .
@xcite , the effect of electron correlations due to their exchange interaction was taken into account under the assumption of characteristic plasma perturbation wavelengths being large compared to the correlation length , using the hartree - fock approximation .
in particular , it was found that the exchange correlations affect the spectrum of volume plasma waves by modifying their dispersion relation ( in the limit of small @xmath129 ) , which becomes : @xmath163 where the third term on the right is due to the exchange correlations , while the second term is due to the quantum statistics resulting in the fermi - dirac distribution ( [ eq : f0 ] ) of electron velocities in degenerate plasma .
we see from ( [ eq : spectrum_exchange ] ) that the term due to exchange correlations is small compared to the term due to quantum statistics even for @xmath162 .
this suggests that in metals the exchange correlations of electrons only lead to minor modification of plasma wave spectra , and neglecting them does not lead to a serious error , while greatly simplifying the model . as for the quantum recoil
it modifies the spectrum ( [ eq : spectrum_exchange ] ) by adding the higher - order term @xmath164 . even for @xmath165 , the term due to quantum recoil
remains small compared with the term due to quantum statistics for @xmath166 , and can be safely neglected . finally , it has beem pointed out in ref .
@xcite that in plasmas with low electron temperatures , @xmath49 ( e.g. , for metals at room temperature ) , the effect of electron - electron coulomb collisions is negligible .
in fact , the typical electron - electron collision frequency @xmath167 for metals at room temperature is of order @xmath168 @xcite , which is many orders of magnitude smaller than the typical plasma oscillation frequency @xmath169 ; hence the electron - electron collisions are not expected to play a significant role for processes occuring at the characteristic collective plasma time scale @xmath170 ( such as the surface waves studied here , with @xmath171 ) , and can thus be neglected .
indeed , the dimensionless surface wave damping rate due to the electron - electron collisions , @xmath172 with @xmath173 , is of the order @xmath174 , which is much smaller than the characteristic collisionless damping rate @xmath175 obtained in this work ( see fig . [
fig : damping ] ) , and thus can be safely neglected . beside colliding with each other
, the electrons can also collide with the ions of metal lattice , yet these collisions were also neglected in our model . to justify this approximation ,
let us estimate the electron - ion collision frequency @xmath176 in typical metals used in plasmonic applications , and compare the characteristic surface wave damping rate @xmath177 associated with these collisions with the collisionless damping rate @xmath175 obtained above .
the experimentally measured electric resistivity of metals such as gold and aluminium at room temperature is of order @xmath178@xcite . using the definition of resistivity @xmath179 , where @xmath180 is the applied electric field , and @xmath25 is the current density in metal ( which is assumed to be entirely due to the free electrons of conductivity ) , and the ohm s law @xmath181 for electrons , where @xmath182 is the dielectric permittivity of vacuum
, we have ( in si units ) : @xmath183 for metals , @xmath184 , and ( [ eq : nu_ei ] ) yields @xmath185 , which is small compared to the characteristic surface oscillation frequency , and thus is not expected to have a significant effect on the dispersion of surface waves . on the other hand ,
the dimensionless surface wave damping rate due to electron - ion collisions , @xmath186 with @xmath187 , is of order @xmath188 , which is comparable to the collisionless damping rate @xmath189 obtained here ( recall that the maximum collisionless damping rate at @xmath190 is @xmath191 ) .
thus the electron - ion collisions , neglected in our model , may lead to additional collisional damping of surface waves that is comparable to the collisionless damping obtained from the collisionless kinetic model .
the above discussion suggests that the quasi - classical mean - field kinetic model used here is adequate for describing surface plasma waves even in moderately coupled plasmas with @xmath165 such as the electron plasma in metals . to verify this , as well as to assess the qualitative and quantitative importance ( or otherwise ) of the effects associated with moderate values of the coupling parameter @xmath192 , a comparison of the model s predictions with experiments in various quantum plasmas , including the electron plasma in metals , is needed
in this paper , electrostatic surface waves in semi - bounded plasma with degenerate electrons were studied using the nonrelativistic collisionless mean - field kinetic model . the dispersion relation for the waves
is obtained from the initial value problem , and its solution corresponding to weakly damped surface waves is presented , yielding dispersion and collisionless damping of the waves for an arbitrary wave number @xmath18 .
it is shown that the surface waves in the semi - bounded plasma with degenerate electrons are weakly damped at all wavelengths , and their damping rate exhibits nonmonotonic dependence on @xmath18 , linearly increasing with @xmath18 at @xmath193 , then reaching maximum at @xmath190 , then falling off rapidly to zero as @xmath18 increases .
this is in contrast with the strong damping of surface waves in semi - bounded plasma with maxwellian electrons , and is the consequence of the effect of quantum statistics ( leading to fermi - dirac velocity distribution ) for plasma electrons .
this work , using the more general kinetic model , extends the results of ref .
@xcite obtained using the quantum fluid theory , in two ways : ( i ) the range of the dispersion relation @xmath105 is extended from @xmath193 in ref .
@xcite to @xmath166 in this paper , and ( ii ) the collisionless damping , absent in the model used in ref .
@xcite , is obtained and discussed here .
to perform the required analytic continuation of @xmath81 onto the real axis of the complex @xmath65 plane , we start from the definition ( [ eq : zeta ] ) of @xmath81 , with @xmath80 given by ( [ eq : epsilon_degenerate ] ) , for @xmath82 ( where @xmath81 is an analytic function of @xmath65 ) , and then reduce @xmath194 down to zero , taking the limit @xmath195 while ensuring that the analyticity of @xmath81 is preserved in the process . for @xmath82 ,
@xmath81 is defined in terms of the integral @xmath196 where @xmath104 , @xmath197 , @xmath134 , and @xmath80 is given by eq .
( [ eq : epsilon_degenerate ] ) .
the function @xmath198^{-1}$ ] under the integral in ( [ eq : integral_kx ] ) is an elementary function of @xmath134 , which in turn is an elementary function of @xmath146 .
thus the function @xmath198^{-1}$ ] can be extended to complex @xmath146 plane by analytic continuation from the real axis @xmath199 of the complex @xmath146 plane , which is achieved by taking the principal branches of the complex square root function @xmath134 and of the complex logarithm function @xmath200 $ ] in ( [ eq : epsilon_degenerate ] ) , considered as functions of complex @xmath146 .
the resulting function @xmath198^{-1}$ ] of complex @xmath146 has the following singularities in the complex @xmath146 plane : 1 .
branch cut of the complex square root @xmath201 , taken along the negative real axis of the argument @xmath202 .
this branch cut maps into two branch cuts of @xmath198^{-1}$ ] in the complex @xmath146 plane , given by two parametric equations : @xmath203 2 .
branch cut of the complex logarithm in ( [ eq : epsilon_degenerate ] ) , taken along the negative real axis of the argument @xmath204 .
this branch cut maps into two branch cuts of @xmath198^{-1}$ ] in the complex @xmath146 plane , given by two parametric equations : @xmath205 3 .
two poles @xmath206 ( @xmath104 ) at the roots of @xmath207 , lying symmetrically above and below the real axis of the complex @xmath146 plane .
two poles @xmath208 at the roots of @xmath142 .
note that for any @xmath209 , @xmath142 does not have roots with @xmath82 , if the plasma equilibrium is stable @xcite , which is the case considered here ; therefore , for any @xmath82 the poles @xmath210 are located _ away _ from the real axis of the complex @xmath146 plane , and thus do not lie on the integration contour in ( [ eq : integral_kx ] ) .
( color online ) singularities ( branch cuts and poles ) of the function @xmath198^{-1}$ ] in the complex @xmath146 plane , and their modification from @xmath82 ( left panel ) to the limit @xmath195 ( right panel ) , for @xmath211 , with @xmath212 defined from ( [ eq : omega0 ] ) .
the branch cuts of the square root and the logarithm are shown with the black dashed lines and the blue dot - dashed lines , respectively .
the poles @xmath210 at the roots of @xmath142 are shown with the filled circles .
the arrows show the direction of motion of the singularities in the process of @xmath195 .
the contour of integration over @xmath146 in ( [ eq : zeta ] ) is shown with the solid red line , with the arrow showing the direction of the contour.,title="fig:",width=288 ] ( color online ) singularities ( branch cuts and poles ) of the function @xmath198^{-1}$ ] in the complex @xmath146 plane , and their modification from @xmath82 ( left panel ) to the limit @xmath195 ( right panel ) , for @xmath211 , with @xmath212 defined from ( [ eq : omega0 ] ) .
the branch cuts of the square root and the logarithm are shown with the black dashed lines and the blue dot - dashed lines , respectively .
the poles @xmath210 at the roots of @xmath142 are shown with the filled circles .
the arrows show the direction of motion of the singularities in the process of @xmath195 .
the contour of integration over @xmath146 in ( [ eq : zeta ] ) is shown with the solid red line , with the arrow showing the direction of the contour.,title="fig:",width=288 ] ( color online ) same as fig .
[ fig : kx_plane_omegar < omega0 ] , but for @xmath213 .
the branch cuts of the square root and the logarithm are shown with the black dashed lines and the blue dot - dashed lines , respectively .
the poles @xmath210 at the roots of @xmath142 are shown with the filled circles .
the arrows show the direction of motion of the singularities in the process of @xmath195 .
the contour of integration over @xmath146 in ( [ eq : zeta ] ) is shown with the solid red line , with the arrow showing the direction of the contour.,title="fig:",width=288 ] ( color online ) same as fig .
[ fig : kx_plane_omegar < omega0 ] , but for @xmath213 .
the branch cuts of the square root and the logarithm are shown with the black dashed lines and the blue dot - dashed lines , respectively .
the poles @xmath210 at the roots of @xmath142 are shown with the filled circles .
the arrows show the direction of motion of the singularities in the process of @xmath195 .
the contour of integration over @xmath146 in ( [ eq : zeta ] ) is shown with the solid red line , with the arrow showing the direction of the contour.,title="fig:",width=288 ] ( color online ) same as fig .
[ fig : kx_plane_omegar < omega0 ] , but for @xmath214 .
the branch cuts of the square root and the logarithm are shown with the black dashed lines and the blue dot - dashed lines , respectively .
the poles @xmath210 at the roots of @xmath142 are shown with the filled circles .
the arrows show the direction of motion of the singularities in the process of @xmath195 . the contour of integration over @xmath146 in ( [ eq : zeta ] )
is shown with the solid red line , with the arrow showing the direction of the contour ; for @xmath215 , this contour is `` squeezed '' between the poles @xmath210 , leading to a singularity in @xmath81 for @xmath195 , when the poles @xmath216 and @xmath217 merge.,title="fig:",width=288 ] ( color online ) same as fig .
[ fig : kx_plane_omegar < omega0 ] , but for @xmath214 .
the branch cuts of the square root and the logarithm are shown with the black dashed lines and the blue dot - dashed lines , respectively .
the poles @xmath210 at the roots of @xmath142 are shown with the filled circles .
the arrows show the direction of motion of the singularities in the process of @xmath195 .
the contour of integration over @xmath146 in ( [ eq : zeta ] ) is shown with the solid red line , with the arrow showing the direction of the contour ; for @xmath215 , this contour is `` squeezed '' between the poles @xmath210 , leading to a singularity in @xmath81 for @xmath195 , when the poles @xmath216 and @xmath217 merge.,title="fig:",width=288 ] the location of these singularities in the complex @xmath146 plane is shown in figs [ fig : kx_plane_omegar < omega0][fig : kx_plane_omegar = omega0 ] . for @xmath82 ( the left panels of figs [ fig : kx_plane_omegar
< omega0][fig : kx_plane_omegar = omega0 ] ) , none of the singularities of @xmath198^{-1}$ ] lie on , or intersect with , the real axis @xmath199 of the complex @xmath146 plane , along which the integration in ( [ eq : integral_kx ] ) is carried out ; the function ( [ eq : integral_kx ] ) and , consequently , the function @xmath81 are thus analytic functions of @xmath65 for @xmath82 . however , as @xmath195 in the process of analytic continuation of @xmath81 to the real axis of complex @xmath65 plane , some of the singularities of @xmath198^{-1}$ ] move about in the complex @xmath146 plane as shown in figs [ fig : kx_plane_omegar < omega0][fig : kx_plane_omegar = omega0 ] , and may collapse onto , or cross with the real axis , thus requiring deformation of the integration contour in ( [ eq : integral_kx ] ) to avoid crossing these singularities and to preserve analyticity of @xmath81 .
below we consider three cases : ( i ) @xmath218 , ( ii ) @xmath219 , and ( iii ) @xmath220 , where @xmath221 are found from the equation @xmath222 for @xmath218 , the logarithm branch cuts ( [ eq : log_cuts ] ) collapse onto the real axis , while the poles @xmath210 collapse onto the imaginary axis of the complex @xmath146 plane as @xmath195 .
the resulting integration contour in @xmath146 is shown in fig .
[ fig : kx_plane_omegar
< omega0 ] . for @xmath219 , both the logarithm branch cuts ( [ eq : log_cuts ] ) and the poles @xmath210 collapse onto the real axis of the complex @xmath146 plane as @xmath195 . the resulting integration contour in @xmath146 is displaced to avoid the poles @xmath210 , as shown in fig . [
fig : kx_plane_omegar > omega0 ] .
finally , for @xmath220 ( the boundary between the above two cases ) , the two poles @xmath210 both collapse towards zero as @xmath194 decreases , as shown in fig .
[ fig : kx_plane_omegar = omega0 ] , `` squeezing '' the @xmath146 integration contour between them . in the limit @xmath195 , the poles @xmath210 merge at @xmath135 , and the integration contour passes through both of them , resulting in @xmath81 being singular ( non - analytic ) at the point @xmath215 .
the function @xmath81 , defined for @xmath223 as described above , was calculated numerically by performing integration over @xmath146 along the appropriate one of the integration contours shown in figs [ fig : kx_plane_omegar < omega0][fig : kx_plane_omegar > omega0 ] . the characteristic @xmath65-dependence of thus defined @xmath81 is shown in fig .
[ fig : zeta(omega ) ] , for a fixed value of @xmath104 .
( color online ) the characteristic @xmath65-dependence of @xmath81 analytically continued to @xmath102 , for a fixed @xmath224 .
@xmath225 $ ] and @xmath226 $ ] are shown with the solid blue and dashed purple lines , respectively .
the frequency @xmath212 defined by eq .
( [ eq : omega0 ] ) , at which @xmath81 is singular , is shown with the dotted vertical line.,width=384 ]
consider the function @xmath81 defined by eq .
( [ eq : zeta ] ) , with @xmath80 given by ( [ eq : epsilon_degenerate ] ) , whose definition is extended to @xmath103 as described in appendix [ sec : zeta ] . for
any given @xmath103 , one can find @xmath18 large enough so that @xmath227 , where @xmath212 is defined by eq . ( [ eq : omega0 ] ) ( this follows from the fact that @xmath228 is a monotonically growing function of @xmath18 ) .
hence for sufficiently large @xmath18 , all poles of the function @xmath229 under the integral in ( [ eq : zeta ] ) lie on the imaginary axis of the complex @xmath146 plane , symmetrically above and below @xmath135 , as shown in fig .
[ fig : kx_plane_omegar
< omega0 ] .
therefore , the `` projection '' of @xmath229 onto the integration contour along the real axis of complex @xmath146 plane is peaked at @xmath135 ( the point on the integration contour closest to the poles ) and decreases away from @xmath135 ( as the distance from the poles increases ) . at @xmath115
the poles @xmath230 of @xmath231 are located closer to the real axis of complex @xmath146 plane than the poles @xmath232 of @xmath93 , hence the variation of the integrand in @xmath81 with @xmath146 is mainly defined by the variation of @xmath233 along the real axis of complex @xmath146 plane . neglecting the variation of @xmath93 with @xmath146 ( and using the peak value of @xmath93 at @xmath135 )
, we can thus approximate the integral over @xmath146 as @xmath234 with @xmath235 defined by ( [ eq : epsilon_degenerate ] ) with @xmath118 .
the corresponding approximation for @xmath81 is then @xmath236^{-1 } \right\},\ \ k_\parallel\gg 1 .
\label{eq : zeta_approx}\ ] ] substituting ( [ eq : zeta_approx ] ) into the dispersion equation ( [ eq : omega_s ] ) for weakly damped surface waves , we have @xmath237 assuming the solution of ( [ eq : disp_tmp ] ) to be of the form @xmath238,\ \text{with } \lim_{k_\parallel\rightarrow\infty}|\delta\omega(k_\parallel)|=0 , \label{eq : omega_assumption}\ ] ] we obtain for @xmath239 : @xmath240,\ ] ] which tends to zero at large @xmath18 , in agreement with the assumption ( [ eq : omega_assumption ] )
. thus we arrive at the approximation ( [ eq : dispersion_kz>>1 ] ) for the frequency of surface waves at @xmath241 . | potential ( electrostatic ) surface waves in plasma half - space with degenerate electrons are studied using the quasi - classical mean - field kinetic model .
the wave spectrum and the collisionless damping rate are obtained numerically for a wide range of wavelengths . in the limit of long wavelengths ,
the wave frequency @xmath0 approaches the cold - plasma limit @xmath1 with @xmath2 being the plasma frequency , while at short wavelengths , the wave spectrum asymptotically approaches the spectrum of zero - sound mode propagating along the boundary .
it is shown that the surface waves in this system remain weakly damped at all wavelengths ( in contrast to strongly damped surface waves in maxwellian electron plasmas ) , and the damping rate nonmonotonically depends on the wavelength , with the maximum ( yet small ) damping occuring for surface waves with wavelength of @xmath3 , where @xmath4 is the thomas - fermi length . |
the recent observations @xcite of bose - einstein condensation ( bec ) of molecules formed from fermionic atoms are extremely exciting . because of a feshbach resonance , in which fermions couple to molecular bosons
, it is possible @xcite , via application of magnetic fields , to obtain an arbitrarily strong attraction between fermions , and to probe the crossover @xcite from bec to bcs .
the resultant superfluidity of pre - formed pairs has a natural counterpart in some theories @xcite of high @xmath2 superconductors .
indeed , a striking , and important feature of the cuprate superconductors is their pronounced precursor superconductivity , or
pseudogap " effects , the origin of which is still under active debate . a second , very important motivation for these experiments is based on the theoretical observation that a bcs - like ground state wavefunction is capable @xcite of describing both fermionic and bosonic - based superconductors , provided that the chemical potential of the fermions , @xmath3 , is determined self consistently . given the vast success of weak coupling or bcs theory , it is extremely important to formulate this extended theory at all @xmath1 , and confront it with controlled experiments .
in this paper we explore the implications of this specific ground state wavefunction @xcite in detail to address both @xmath4 and @xmath5 , with particular emphasis on the near - bec " regime .
our goals are ( i ) to emphasize how to include finite temperature effects in a manner consistent with the ground state ; this necessitates the introduction of non - condensed fermion pairs which lead to pseudogap " effects @xcite , ( ii ) to provide analytic calculations and insights by working in a ( near - bec ) regime where calculations are more tractable , ( iii ) to discuss in some detail the differences between the one channel " and two channel " models for the feshbach resonance , and ( iv ) to compare with well established theories of weakly interacting bose superfluid @xcite , as well as ( v ) with the measured density distributions in a trap .
we begin with the homogeneous case and then consider the trap configuration ; from this we infer the ratio between the effective bosonic and fermionic scattering lengths , which is found to be strongly dependent on the magnetic detuning @xmath6 , and generally less than 2 , as a consequence of many - body combined with feshbach resonance effects .
the hamiltonian used in the cold atom and high @xmath2 crossover studies consists , in its most general form , of two types of interaction effects : those associated with the direct interaction between fermions parameterized by @xmath7 , and those associated with fermion - boson " interactions , whose strength is governed by @xmath8 .
@xmath9 here the fermion and boson kinetic energies are given by @xmath10 and @xmath11 , respectively , and @xmath12 is an important parameter which represents the detuning " .
the variational ground state which we will consider here is a product of both fermionic and bosonic contributions @xmath13 where the normalized fermionic wave function is the standard crossover state @xcite @xmath14 and the normalized molecular or feshbach boson contribution @xmath15 is represented by @xmath16 the variational parameters are , thus , @xmath17 , @xmath18 and @xmath19 . applying standard variational techniques on
the ground state wavefunction leads to eqs .
( [ eq : extra2 ] ) and ( [ eq : extra ] ) , and the @xmath4 limit of eq .
( [ number_equation ] ) . along with the result that @xmath20 . whether both forms of interactions ( @xmath7 and @xmath8 ) in the hamiltonian are needed in either system is still under debate . the bosons ( @xmath21 ) of the cold atom problem @xcite will be referred to as feshbach bosons .
these represent a separate species , not to be confused with the fermion pair ( @xmath22 ) operators .
thus we call this a two channel " model . for the sake of clarity we begin by ignoring the feshbach resonance - induced interactions .
the variationally induced constraints on the fermionic degrees of freedom @xcite are given by @xmath23 with the fermion density @xmath24 , where @xmath25/2 $ ] .
we define the quasiparticle dispersion , @xmath26 , below .
one can anticipate that the excited states of this system are associated with their natural extension @xcite following bcs theory . for @xmath27 ,
the gap equation and number equations now involve fermi functions @xmath28 and are given by @xmath29\ , , \label{eq : gap_equation}\\ n & = & \sum _ { \bf k } \left [ 1 -\frac{{\epsilon_{\mathbf{k}}}- \mu}{{e_{\mathbf{k } } } } + 2\frac{{\epsilon_{\mathbf{k}}}- \mu}{{e_{\mathbf{k}}}}f({e_{\mathbf{k } } } ) \right ] \ , , \label{eq : number_equation}\end{aligned}\ ] ] where @xmath3 is now evaluated at finite @xmath1 and @xmath30 .
here we assume a contact potential where @xmath31 is the isolated two body scattering length , which is simply related to @xmath7 , ( see , for example , eq . ( [ eq : a6 ] ) ) .
two important observations need to be made . here
@xmath32 is associated with the excitation gap for fermions . _ this excitation gap reflects both condensed and non - condensed pairs and is not the same as the order parameter @xmath33 , except at @xmath4_. eqs .
( [ eq : gap_equation ] ) and ( [ eq : number_equation ] ) are widely cited in the literature , but with an important difference . throughout the literature @xmath32
is taken to be the order parameter .
by contrast , to determine @xmath33 here we will first need to characterize the contribution of non - condensed pairs to @xmath32 .
our second important observation is that for @xmath27 , the fermi factors in the above equations are essentially negligible @xcite .
it follows that _ both @xmath34 and @xmath35 are temperature independent in this near - bec regime .
_ indeed , this is consistent with the physical picture of well established , pre - formed pairs in the bec limit , so that the fermionic energy scales are unaffected by @xmath1 below @xmath2 . the simple physics in eq . ( [ eq : gap_equation ] )
may be schematically represented by plots of @xmath32 versus temperature .
figure [ fig:3 ] contrasts the behavior in the weak coupling or bcs and strong coupling or bec regimes . in the bcs and bec regimes .
similarly , in the bec regime @xmath3 is a constant , so that all fermionic energy scales are @xmath1 independent , as expected.,width=326 ] in the bcs limit @xmath34 follows the behavior of the order parameter , whereas in the bec regime , pairs are pre - formed and there is no temperature dependence in @xmath34 on the scale of @xmath2 .
we now extend these qualitative observations to a more quantitative level .
the self consistent equations in the bec limit for general temperature @xmath1 can then be written as it follows from the above equations that , just as in the @xmath4 limit @xcite , we have for general @xmath37 , @xmath38 where the coefficient of proportionality @xmath39 we arrive at an important physical interpretation .
even though @xmath32 or @xmath40 is a constant in @xmath1 , this constant must be the sum of two temperature dependent terms .
just as in the usual theory of bec these two contributions correspond to condensed and non - condensed components @xmath41 so that we may decompose the excitation gap into two contributions @xmath42 where @xmath43 corresponds to condensed and @xmath44 to the non - condensed ( or pseudo ) gap component .
each of these are proportional to the respective number of condensed and non - condensed pairs with proportionality constant @xmath45 . just as in bec , at @xmath2 , @xmath46 where @xmath47 is the usual bose - einstein function and
@xmath48 is the dispersion of the non - condensed pairs , which will be self consistently determined below .
thus @xmath49 we may deduce directly from eq .
( [ eq:14 ] ) that @xmath50 , if we presume that below @xmath2 , the non - condensed pairs have propagator @xmath51 this lead to a key question : how can one deduce the contribution from _ non - condensed _
we now work backwards to infer the dispersion @xmath48 for these pairs .
a fundamental requirement on non - condensed pairs in equilibrium with a bose condensate is that their effective chemical potential satisfy @xmath52 equation ( [ eq : gap_equation ] ) can be shown to be consistent with eq .
( [ eq : chempot ] ) provided that the propagator for non - condensed pairs is given by @xmath53 where @xmath54 and @xmath55 represents the fermionic green s function which has a self energy @xmath56 . here
@xmath57 is the bare propagator .
the details of this analysis are presented in appendix [ app:1 ] .
another important point should be noted .
this pair propagator or t - matrix differs from that first introduced by nozieres and schmitt - rink @xcite because here there is one dressed and one bare green s function . in the approach of ref .
both are taken as bare green s functions .
by contrast there are other schemes in the literature @xcite where both green s functions are dressed .
we end by noting that at small four - vector @xmath58 ( and moderately strong coupling ) we may expand eq .
( [ eq : t - matrix_pg ] ) to obtain @xmath59 now we can deduce directly from eq .
( [ eq : t - matrix_pg ] ) that the dispersion of non - condensed pairs is of the form @xmath60 in summary , this quadratic dispersion can be derived from the pair susceptibility @xmath61 . in turn , the particular form for @xmath61 shown in eq .
( [ eq : chi ] ) is chosen in order to be consistent with eqs .
( [ eq : gap_equation ] ) and ( [ eq : number_equation ] ) . in this sense
the usual bec constraint expressed by eq .
( [ eq : chempot ] ) is intimately connected to the bcs - like gap equation of eq .
( [ eq : gap_equation ] ) .
the details of this analysis are shown in appendix [ app:1 ] .
we now extend this analysis to include feshbach bosons @xcite .
for this situation we can write down an equation @xcite equivalent to eq .
( [ eq : gap_equation ] ) with the direct fermion interaction @xmath7 replaced by @xmath62 and effective scattering length @xmath63 . here
we define @xmath64 where @xmath65 is dependent on @xmath3 .
we thus have @xmath66 \ , , \label{eq:19}\end{aligned}\ ] ] where @xmath67 , and again @xmath32 is to be distinguished from the order parameter @xcite .
alternatively one can rewrite eq .
( [ eq:18 ] ) as @xmath68 \ , .
\label{eq:18a}\ ] ] in eq .
( [ eq:19 ] ) @xmath69 represents the number of fermions , but the number equation for the total number of particles involves both condensed and uncondensed bosons as well @xmath70 where @xmath71 is the number of molecular bosons in the condensate .
the number of noncondensed molecular bosons is given by @xmath72 where the bose propagator is @xmath73 and we choose the self energy @xcite
@xmath74 \label{eq : sigmab}\ ] ] to be consistent with the hugenholtz - pines condition that bosons in equilibrium with a condensate must necessarily have zero chemical potential .
this is equivalent to @xmath75 where @xmath76 .
it follows after some simple algebra that eq .
( [ eq : chempot2 ] ) is then consistent with eq .
( [ eq:18 ] ) .
we have introduced quantities ( @xmath77 ) which characterize the two body scattering in vacuum and correspond to the scattering lengths @xmath31 or @xmath65 .
how are these related @xcite to the parameters ( @xmath78 ) which enter into the hamiltonian ?
to begin we ignore feshbach effects .
the way in which the two body interaction @xmath7 enters to characterize the scattering ( in vacuum ) is different from the way in which it enters to characterize the n - body processes leading to superfluidity . in each case , however , one uses a t - matrix formulation to sum an appropriately selected but infinite series of terms in @xmath7 .
for the two - body problem in vacuum , there is a measurable characteristic of the scattering , the scattering length , @xmath31 , @xmath79 which is related to @xmath7 via the lippmann - schwinger equation @xmath80 in this way one may solve for the unknown @xmath7 in terms of @xmath31 or @xmath81 .
the two - body t - matrix equation ( eq.([eq:11c ] ) ) can be rewritten as @xmath82 where we define the quantity @xmath83 as @xmath84 here @xmath83 is the critical value of the potential associated with the binding of a two particle state in vacuum
. specific evaluation of @xmath83 requires that there be a cut - off imposed on the above summation , associated with the range of the potential .
now , for the feshbach problem , the calculation proceeds in a similar fashion and one finds @xcite @xmath85 here we define the parameter @xmath6 which is directly related to the difference in the applied magnetic field @xmath86 and @xmath87 @xmath88 where @xmath89 is the difference in the magnetic moment of the two paired hyperfine states . to connect the various energy scales , which appear in the problem , typically @xmath90 for @xmath91k .
there is an important difference in the nature of the condensate for the two cases , with and without feshbach bosons . in the latter case
there are two components to the condensate associated with cooper pairs ( @xmath33 ) and condensed molecular bosons .
moreover , the cooper condensate enters only into the gap equation , whereas the molecular bose condensate enters also into the number equation .
as long as the bosonic condensate @xmath92 is non - negligible , this difference leads to essentially different physics between the one and two channel problem , as will be demonstrated below . in this section
we enumerate the different relationships between the cooper condensate and bosonic condensate terms . the order parameter associated with eq .
( [ eq:0c ] ) represents a linear combination of both paired fermions ( cooper condensate ) and condensed molecules .
it is given by @xcite @xmath93 where , the boson order parameter @xmath94 .
we have @xmath95 therefore , @xmath96 can be written as @xmath97 \sum _ { \bf k } \langle a_{-{\bf k}\downarrow}a_{{\bf k}\uparrow}\rangle \label{eq : c6}\ ] ] where we have used the fact @xcite that @xmath98 the number of condensed feshbach bosons which enters the number equation [ eq .
( [ number_equation ] ) ] is given by @xmath71 .
thus we have @xmath99 ^ 2}= \frac{g^2 \tilde{\delta}_{sc}^2 } { [ ( 2\mu-\nu)u+g^2]^2 } \label{eq : c2}\ ] ] using eqs .
( [ eq:200 ] ) we can also write @xmath100 where @xmath101 is defined in eq .
( [ eq : c7 ] ) .
we now rewrite our central equations ( [ eq : gap_equation ] ) , ( [ eq : number_equation ] ) and ( [ eq:14 ] ) in the near - bec limit to compare more directly with the case of a weakly interacting bose gas , described by gross - pitaevskii ( gp ) theory .
it can be shown that ( in the absence of fb ) @xmath102 which , in conjunction with the expansion of eq .
( [ eq : gap_equation ] ) , @xmath103 , \label{eq : gap_eq_exp}\ ] ] yields @xmath104 these equations hold at all @xmath27 . at @xmath4 ,
these equations have been shown @xcite to be equivalent to the results of gp theory where one identifies an effective inter - pair scattering length @xmath105 via @xmath106 . here
@xmath107 represents the number density of pairs , @xmath108 is the bare " chemical potential of the pairs , and @xmath109 the pair mass .
we now show that in the presence of feshbach bosons this equation of state is no longer that of gp theory and , moreover , there are important differences in the ratio of the bosonic to fermionic scattering lengths . as a result of the bose condensate @xmath92 in eq .
( [ number_equation ] ) one finds an extra term in the number equation @xmath110 .
\label{eq : density}\end{aligned}\ ] ] combining the gap and number equation yields @xmath111 .
\label{eq : combined}\end{aligned}\ ] ] solving for @xmath3 in terms of @xmath65 one finds a new equation of state @xmath112 to lowest order in @xmath65 .
the second term in the above equation derives from the bose condensate term . the first term in eq .
( [ eq : density ] ) contributes a term of order @xmath113 to this correction @xcite .
this behavior should be contrasted with the situation when fb are absent , where @xmath114 .
it should also be emphasized that the fermionic scattering length in a model without fb is an independent experimental parameter , while here @xmath65 depends on @xmath3 , and must be obtained self - consistently .
we turn next to a calculation of @xmath2 , which requires that we determine @xmath48 [ via the t - matrix of eq .
( [ eq : t - matrix_pg ] ) ] as a function of the scattering length @xmath31 .
we address the one - channel case first .
the general expression for @xmath115 in the near bec limit is given by @xmath116 \
, , \label{eq : b}\ ] ] where we have used eqs .
( [ eq : mass ] ) and ( [ eq : expandt ] ) . after expanding to lowest order in @xmath117 @xmath118 from eq .
( [ eq:14 ] ) we then conclude that , at @xmath2 , @xmath119 equation ( [ eq : totaln ] ) reflects the fact that , in the near - bec limit , and at @xmath2 , all fermions are constituents of uncondensed pairs . from the above equation it follows that @xmath120 which , in conjunction with eq .
( [ eq : mass_change ] ) implies @xmath121 here @xmath122 is the transition temperature of the ideal bose gas with @xmath123 .
this downward shift of @xmath2 follows the effective mass renormalization , much as expected in a hartree treatment of gp theory at @xmath2 . here
, however , in contrast to gp theory for a homogeneous system with a contact potential @xcite , there is a non - vanishing renormalization of the effective mass .
we turn now to the analysis of the behavior of @xmath2 in the presence of feshbach bosons .
the ( inverse ) residue in the @xmath1-matrix is replaced by @xmath124 ^ 2 } \approx \frac{n}{2\delta^2}\ ] ] and the @xmath125 coefficient @xmath126 in @xmath127 is such that @xmath128 ^ 2}}{z_1 } \approx \frac{1}{4m}\left[1-\frac{2g_0 ^ 4}{u_0 ^ 4}\pi^3 n a_s^{*9}\right].\ ] ] here @xmath45 and @xmath87 are the appropriate counterparts when fb are absent . since @xmath45 is proportional to the fermionic contribution to the density , it is very small in the bec limit . using the same reasoning as in the previous case ,
we conclude that the ratio @xmath129 is constant with varying coupling . _ thus , again , @xmath2 follows the behavior of the inverse effective mass _ with , to leading order , @xmath130 .
the differences between the equations of state for the one and two channel models will have physical implications in the density profiles of particles in a trap .
we now introduce the harmonic trapping potential @xmath131 , which is treated in the thomas - fermi ( tf ) approximation . in this approximation , one replaces @xmath3 with @xmath132 .
in contrast to the uniform case , here @xmath133 becomes non - zero beyond a critical radius @xmath134 , where @xmath135 . in this way at @xmath2 , only the center of the trap is superfluid , while at @xmath4 all of the trap contains condensed states . to obtain the @xmath4 density profile , @xmath136
, we insert @xmath137 into eq .
( [ eq : combined ] ) and solve for @xmath136 [ here @xmath136 refers to the sum of both fermion and molecular boson contributions ] .
the solution is @xmath138 \nonumber \\ & & \times { } \left[\frac{1}{16 \pi \sqrt{|\mu(r)|}}(\frac{2 m}{\hbar^2})^{\frac{3}{2}}+2 \frac{(1-\frac{u_0}{u^*})^2}{g_0 ^ 2 } \right ] , \label{eq : nr_fb}\end{aligned}\ ] ] where we use @xmath139 to self - consistently determine @xmath3 . here
it should be noted that @xmath65 is itself a function of @xmath137 . and ( b ) @xmath140 .
solid and dashed lines are at @xmath4 with and without feshbach bosons ( fb ) , respectively , and the dotted lines are at @xmath5 .
inset to ( a ) plots the fb weight in the condensate .
inset to ( b ) shows the ratio of the inter - boson to inter - fermion scattering lengths at trap center .
units are chosen such that @xmath141 , and @xmath142 .
hence the units for @xmath81 , @xmath143 , and for @xmath144 , @xmath145 , are both set to unity . for @xmath146 the interaction parameter @xmath147.,width=288 ] in fig .
[ fig : profiles](a ) , we plot @xmath136 for the parameters @xmath148 , @xmath149 , @xmath150 and @xmath151 . for this value of @xmath6
we are somewhat away from the deepest bec regime as is necessary to ensure the validity of the tf approximation and the fermionic contribution to the density is no longer negligible compared to its bosonic counterpart .
indeed , the percentage weight of the molecular boson condensate ( shown in the inset ) indicates that by @xmath152 the cooper pair condensate is beginning to dominate . in this figure
we also show the density profile as computed in the absence of fb ( dashed line ) , as well as the behavior at @xmath5 ( dotted line ) .
our @xmath5 curves were computed in the absence of fb for both panels .
we were unable , thus far to find important differences between this distribution and that of an ideal bose gas .
here one has to solve self consistently for @xmath153 , as well .
to arrive at a meaningful comparison of the two @xmath4 cases ( with and without feshbach bosons ) we used the same value for the effective two - body scattering length @xmath65 , obtained from the self - consistent calculations of @xmath136 in the presence of fb .
the profile without fb bosons is then calculated using the familiar tf result near the bec limit @xcite at a fixed fermionic scattering @xcite length @xmath65 .
the same plots are presented with @xmath140 in fig .
[ fig : profiles](b ) , which is further from the bec limit
. an important consequence of the one _ vs _ two channel problems is that for the latter , as the extreme bec limit is approached , the number of fermions is diminished in favor of fb .
this occurs due to the self consistent adjustment of the fermionic chemical potential .
thus the interaction between the bosons mediated by the fermions is weaker than what might have been expected without consideration of fb .
this weakened interaction is reflected in fig .
[ fig : profiles ] through a comparison between the solid and dashed lines , which shows that the trap profiles are narrower in the presence of feshbach effects . indeed ,
this could have been anticipated from the above calculations in the homogeneous case , at @xmath4 and @xmath2 , which deduced a very weak dependence of the bosonic scattering length and effective pair mass on the fermionic scattering length @xmath65 .
this finding mirrors the result in a trapped atomic bose gas@xcite where one sees that a weaker repulsive inter - boson interaction leads to a narrowed density profile .
comparison between the lower and upper panels of fig .
[ fig : profiles ] shows that in both cases ( with and withour fb ) , the profiles become narrower as the bec limit is approached .
we may use the results in fig .
[ fig : profiles ] to obtain a semi - quantitative estimate of the bosonic scattering length @xmath154 , based on a phenomenology used in experimental analysis @xcite .
we compare our results to the thomas - fermi approximated gp equation at @xmath4 which yields @xmath155 where the inter - boson interaction @xmath156 is connected to the scattering length @xmath154 via @xmath157 .
if we fit the profiles of fig . [ fig : profiles ] to the inverted parabola ( @xmath158 ) , we may infer that @xmath159 .
the ratio @xmath160 is plotted vs @xmath6 in the inset to the lower panel of fig .
[ fig : profiles]b .
the same analysis applied to the profile without fb yields the familiar 2:1 ratio of the bosonic to fermionic scattering lengths .
the parameter @xmath154 is an important quantity which appears in experiment to be considerably less than a factor of 2 times its fermionic counterpart .
there are very detailed calculations @xcite of four - body atomic scattering processes which yield a fixed length ratio of @xmath161 for the bosonic to fermionic scattering lengths . presumably in the asymptotic bec regime
this constraint should be imposed on a more exact theory of the ground state .
indeed , a weakness of the mean field approach taken here ( based on a generalized bcs ground state ) is that the inter - boson " scattering length is treated only approximately .
this can be viewed as related to the inclusion of only one fermion and two fermion propagators ( _ i.e _ , t - matrix ) with no higher order terms .
a strength of this approach , over the counterpart atomic calculations is that many body physics associated with the broken symmetry or superfluid state is included . _
a comparison of the one and two channel calculations presented here shows that , through many body effects , the nature of the condensate enters in an important way to determine the equation of state and , thereby , the ratio of the scattering lengths_. it should also be stressed that the many body context with which the number @xmath161 is to be compared is different from @xmath162 , which is the number generally assumed in the literature .
this difference appears once one includes feshbach bosons .
indeed , while it depends upon the magnetic field or detuning , in the near - bec regime this ratio is found to be significantly less than @xmath162 .
this is a central point of the present paper .
future experiments will be required to determine , whether , as some have presumed , the ratio of @xmath161 applies to all detunings , or whether this number is variable as argued here .
in this paper we have shown that superfluidity of fermionic atoms in the near - bec limit is in general different from bose superfluidity ( as described by gp theory ) .
we have compared one and two channel models and find that differences from the gp picture for each are associated with the underlying fermionic character of the system .
this shows up at @xmath163 for the one channel system and at all @xmath1 for a hamiltonian which includes feshbach bosons .
essential to our approach is the presence ( at all @xmath164 ) of non - condensed pairs which contribute separately to the fermionic gap parameter @xmath32 , in addition to the condensate term .
noncondensed pairs or
pseudogap effects " have been generally ignored in the literature @xcite and we have shown elsewhere @xcite that they are negligible only in the strict bcs regime .
our inclusion of these effects corresponds @xcite to a self - consistent hartree treatment of pairing fluctuations in bcs theory , in which condensation naturally occurs in the presence of a finite excitation gap ( pre - formed pairs " ) .
our approach also makes evident the strong analogies between this generalization of bcs theory and bose condensation , where both condensed and non - condensed bosons must be properly characterized . _ in the two channel model , non - condensed bosons should be viewed as consisting of a strongly hybridized mixture of fermion pairs and feshbach bosons . in the one channel model they result exclusively from fermion
pairs_. a key aspect of this work is in the comparison we have presented between one and two channel models .
differences arise because of the nature of the condensate .
a fermionic or cooper pair condensate ( @xmath33 ) enters into the gap equation while a bosonic condensate ( @xmath92 ) enters also into the number equation .
when there is an appreciable fraction of bosonic condensate ( as in the near - bec and bec regimes ) these differences will be apparent .
a distinction between the one and two channel models is relatively unimportant once the bosonic condensate is negligible ( usually when the scattering length is large and positive or alternatively negative ) .
these differences are associated with physical properties such as the equations of state and the density profiles in a trap .
feshbach bosons can , in effect , collapse much more completely to the center of a trap than can fermion pairs .
this effect can be inferred from the ratio of the scattering lengths , and is related to the fact that in this limit , fermions are essentially absent , since the number equation constraint can be entirely satisfied by populating the bosonic state . as a result , these bosons are close to ideal . for the relatively broad feshbach resonances currently under study in lithium and potassium , differences between the one and two channel models are presumably only important in the bec and near - bec limits studied here , where there is an appreciable bosonic condensate .
it will be interesting to study narrower feshbach resonances where these differences may persist into the unitary scattering regime .
we acknowledge useful discussions with r. hulet , d. jin and z. tesanovic .
this work was supported by nsf - mrsec grant no .
dmr-0213765 ( js and kl ) , nsf grant no .
dmr0094981 and jhu - tipac ( qc ) .
the expansion of the t - matrix at small four - vector @xmath165 can be written in the following form : @xmath166 therefore , the condition for the divergence of the t - matrix at zero @xmath58 is equivalent to @xmath167 or @xmath168 we now show that for a proper choice of the pair susceptibility @xmath169 eq .
( [ eq : a1 ] ) is equivalent to the bcs - like gap equation of eq .
( [ eq : gap_equation ] ) , provided also the greens function is of the bcs form @xmath170 just as in bcs theory , we associate this dressed green s function with self - energy @xmath171 we calculate @xmath172 by performing the appropriate matsubara sums following standard procedure @xcite . at @xmath173 , we have @xmath174 and eq .
( [ eq : sfl_condition ] ) becomes @xmath175 which , in conjunction with the two - body scattering equation @xmath176 gives @xmath177 \label{eq : a5}\ ] ] thus , we have demonstrated that , if we presume eqs .
( [ eq : a2 ] ) and ( [ eq : a3 ] ) , then eqs . ( [ eq : a1 ] ) and ( [ eq : a5 ] ) are equivalent .
this scattering length should be distinguished from the isolated two - body scattering parameter which unlike @xmath65 does not reflect the interactions responsible for superfluidity .
indeed this isolated two - body parameter appears to be inappropriate , in that it becomes negative deeper into the bec regime ( @xmath180 ) .
this sign change should occur quite generally bcs - like behavior at large positive @xmath6 . | we show how to describe the @xmath0 behavior associated with the usual bcs- bose einstein condensation ( bec ) crossover ground state .
we confine our attention here to the bec and near - bec regime where analytical calculations are possible . at finite @xmath1 ,
non - condensed fermion pairs must be included , although they have been generally ignored in the literature . within this bec regime
we compute the equations of state for the one and two channel models ; these two cases correspond to whether feshbach resonance effects are omitted or included .
differences between these two cases can be traced to differences between the nature of a cooper pair and bosonic condensate .
our results are also compared with the gross pitaevskii equations of state for true bosons .
differences found here are associated with the underlying fermionic character of the system . finally , the particle density distribution functions for a trap containing superfluid fermionic atoms are computed using a thomas - fermi approach .
the one and two channel behavior is found to be very different ; we find a narrowing of the density profile as a result of feshbach resonance effects .
importantly , we infer that the ratio between bosonic and fermionic scattering lengths depends on the magnetic detuning and is generally smaller than 2 .
future experiments will be required to determine to what extent this ratio varies with magnetic fields . |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.