article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
cell membranes contain large amounts of proteins within or attached to the lipid bilayer @xcite .
the distribution of the proteins is not necessarily homogeneous , which can have important functional consequences .
for example , proteins with an intrinsic curvature couple to the bilayer conformation @xcite ; on the one hand , such proteins are preferably found on similarly curved parts of the membrane @xcite , on the other hand , the proteins deform the membrane locally @xcite .
asymmetric , curved proteins can regulate the polymerization of the three - dimensional cytoskeleton of the cell @xcite and control intracellular transport via endocytosis @xcite .
virus endocytosis can occur via the same mechanism @xcite .
the conical inclusions in our model mimick asymmetric proteins within the bilayer @xcite , proteins or polymers attached to the bilayer @xcite , curved lipid domains @xcite , and viruses that bind to the membrane @xcite .
the interaction between the inclusions in a lipid bilayer is mediated by membrane deformations and thermal undulations @xcite , in addition to surface tension @xcite and possible direct interactions that we do not consider in this paper .
the deformation - induced , pairwise interaction of curved inclusions occurs in the absence of thermal membrane undulations and is usually repulsive @xcite ; in a planar membrane it is long - range @xcite .
however , the interactions can be strongly screened if the average curvature of the membrane and the protein curvature are similar @xcite .
one obvious example for strongly screened interactions are inclusions that are placed on a vesicle with similar curvature radius @xcite .
screening can also be achieved by many - body interaction in clusters of inclusions @xcite . at finite temperature ,
casimir - like interactions due to membrane undulations generate attraction @xcite .
curvature generation by inclusions and induced budding in lipid bilayer membranes has been reported in many experimental studies of biological and biomimetic systems @xcite .
computer simulations allow to study the membrane - mediated interaction between the inclusions in detail without the presence of other , direct interactions .
recently , bud formation by curved inclusions has been investigated with computer simulations @xcite .
it was found that the inclusions on the buds have a higher density than they had in the initially nearly flat membrane @xcite .
this might appear to be a result of undulation - induced attraction that in consequence leads to clustering of the inclusions and to budding .
such systems and processes can be studied theoretically on the basis of an elastic membrane that is characterized by its bending rigidity , @xmath0 , and gaussian saddle splay modulus , @xmath1 , with curved inclusions that consist of sections of a sphere with a given opening angle .
we demonstrate that bud formation can already be well understood on the basis of the membrane deformation alone .
we show that the higher inclusion density on the bud is a result of a screened repulsive interaction .
we further argue that the budding pathway plays an important role for bud size .
this allows us to predict a range of possible bud radii for a given system , which nicely agrees with recent simulation results @xcite . at finite temperature
, the inclusions can exist in a fluid and in a crystalline phase , which depends on the strength of their repulsive interaction .
we construct an approximate free - energy functional that takes into account for the bending energy as well as the translational entropy of the inclusions .
we calculate a phase diagram for the fission of a single vesicle of given size and for given number and geometry of the inclusions .
the inclusion entropy plays a decisive role for the sizes of the smaller vesicles into which a larger vesicle may split .
the bending energy @xmath2 of a lipid bilayer is given by the integral over the entire membrane area , @xmath3 where @xmath0 is the bending rigidity , @xmath1 is the saddle - splay modulus , @xmath4 is the mean curvature , @xmath5 is the gaussian curvature , and @xmath6 and @xmath7 are the principal curvatures at each point of the membrane .
the integral over the gaussian curvature is determined by the topology of the membrane and by the geodesic curvature at the boundary . in our case ,
the geodesic curvature is given by the geometry of the inclusions , so that in general this term of the integral over the membrane shape does not need to be calculated explicitly .
for bud formation , we neglect the constant contribution of the gaussian saddle splay modulus . in order to minimize the bending energy , the inclusions preferably order on a hexagonal lattice ( fig .
[ fig3 ] _ ( a ) _ ) ; therefore it is a natural assumption that the symmetry axis is oriented normal to the local tangent plane of the vesicle on which the inclusions are placed . to calculate the deformation energy , we approximate the hexagons with overlapping circles that have the same projected area ( fig . [ fig3 ] _ ( b ) _ ) . , like @xmath8 . to minimize the bending energy , the inclusions order in an hexagonal structure .
( b ) the hexagons are approximated by overlapping circles that have the same projected area . ]
if there are no overhangs , the membrane conformation can be described in monge parametrization by a height field , @xmath9 , over a planar reference surface . for an almost planar membrane ,
the bending energy of the membrane is @xmath10 ^ 2 \hspace{3ex } ( { \bm{\rho}}=(x , y ) ) \ , , \end{aligned}\ ] ] with @xmath11 the integral over the reference plane .
minimization of the bending energy gives the biharmonic euler - lagrange equation , @xmath12 in cylindrical coordinates , the general solution of eq .
( [ eq3 ] ) is @xmath13 with the four integration constants @xmath14 to @xmath15 @xcite .
the boundary conditions that are imposed on the membrane are sketched in fig .
the radius of the inner boundary , @xmath16 , and the slope of the membrane at the inner boundary , @xmath17 , are determined by the inclusion geometry . for @xmath18 inclusions on a vesicle with radius @xmath19 and surface number density @xmath20 of the inclusions , the radius of the outer boundary is @xmath21 with @xmath22 ; the slope of the membrane at the outer boundary is @xmath23 . for inclusions on a planar membrane ,
the latter expressions simplify to @xmath24 and @xmath25 .
the remaining two boundary conditions are given by fixing the membrane height at the inner ( or equivalently at the outer ) boundary and minimizing the energy with respect to the height of the inclusion above the vesicle ( i. e. the height difference between both boundaries ) , which implies @xmath26 at the inclusion and @xmath27 at the outer boundary .
, the opening angle , @xmath28 , and the projected inclusion radius , @xmath29 .
the size of the corresponding membrane patch is @xmath30 , the slope of the membrane at the inclusion is @xmath31 , and the slope of the membrane at the outer boundary is @xmath32 ( @xmath25 for inclusions on a planar membrane ) . ]
( [ eq7 ] ) together with the boundary conditions gives the shape of the deformation , @xmath33 and the corresponding bending - energy cost , @xmath34 ^ 2 = \frac{2 \pi \kappa ( b \rho_o - a \rho_i)^2}{(\rho_o^2-\rho_i^2 ) } \ , \ , \ , .
\label{eq10}\ ] ] the energy a function of @xmath30 and @xmath32 , which depend on the inclusion density , while all other quantities are intrinsic properties of membrane and inclusions . for a single inclusion in an infinite planar membrane , @xmath25 and @xmath35 , the bending energy vanishes and
the membrane deformation is catenoid - like , @xmath36 .
note that in a pairwise approximation , the interaction energy for two inclusions in a planar membrane ( @xmath25 ) decays like @xmath37 for large distances between the inclusions ( large @xmath38 ) .
[ sec1c ] for inclusions on a vesicle , the membrane shape and the minimal bending energy ( assuming that the inclusions have maximal mutual distances ) can be calculated using eqs .
( [ eq10a ] ) and ( [ eq10 ] ) . for @xmath39 ,
the membrane around the inclusion has almost catenoid shape @xcite ; the catenoid is a minimal surface without bending - energy cost .
if the entire vesicle is covered with inclusions and catenoids such that the bending energy is zero ( fig .
[ fig2 ] _ ( b ) _ ) , the inclusions have optimal density . , where @xmath40 is the area fraction of the vesicle that is covered with inclusions and catenoidal patches .
( b ) vesicle decorated with inclusions at optimal density ; the bending energy of the lipid bilayer membrane vanishes . ] for lower inclusion densities , in a first approximation the catenoidshape borders on a spherical shape with the curvature radius of the vesicle .
the bending energy of a vesicle that is decorated with curved inclusions is reduced by the fraction of the sphere s surface area that is covered by the inclusions and the catenoid - shaped membrane segments , see fig .
therefore for low inclusion density , the bending energy of the decorated spherical vesicle is in the range @xmath41 . in the full solution , which is given by eq .
( [ eq10a ] ) and will be used in the remainder of the paper , there is no jump in the mean curvature from @xmath42 to @xmath43 between the catenoid and a sphere as sketched in fig .
[ fig2 ] but rather a smooth transition from zero to finite mean curvature . for inclusion densities that are higher than the optimal density , due to the boundary conditions
no solution can be constructed by matching of catenoids . in this case
, the bending energy always has a finite value that can exceed the bending energy of a bare vesicle . the minimal bending energy of a vesicle with inclusions is shown in fig .
[ fig5 ] as function of the number of inclusions , @xmath44 , and the vesicle radius , @xmath19 .
we find degenerate zero - energy ground states that have optimal inclusion density with an approximately linear dependence @xmath45 , where @xcite @xmath46 the natural spontaneous curvature of the bilayer for given inclusion density and geometry is @xmath47 . , of a vesicle with radius @xmath19 with @xmath44 inclusions ( @xmath48 , @xmath49 ) .
there is a region of low inclusion density at large @xmath19 with @xmath50 , which is delineated by a line of zero - energy ground states from a region of high inclusion density at small @xmath19 , where also bending energies @xmath51 can be found .
( the high energies that are cut off at small @xmath44 and large @xmath19 mark the breakdown of the small - curvature expansion of the bending energy . ) ] the same value for the inclusion density can be high , optimal , or low , depending on the radius of the vesicle on which the inclusions are placed .
the smaller the radius of the vesicle , the larger the value of the optimal density . in a planar membrane , the slopes of two adjacent catenoid - like deformations can not be matched for any finite distance between the inclusions .
therefore the inclusions are always in the high - density regime in this case
. bud formation does not occur for a vesicle with low inclusion density and bending energy , @xmath50 , because this would lead to an increase of the total bending energy @xcite .
however , for high inclusion density ( @xmath52 , see eq .
( [ eq11 ] ) ) , the system can always reach a state of lower bending energy if small vesicles bud from the main vesicle .
the set of smaller vesicles into which a large vesicle with high inclusion density splits up is not uniquely determined from bending energy alone , because the states of vanishing bending energy are degenerate .
a natural assumption is that the vesicle will split into one large mother vesicle and one or more small daughter vesicle(s ) of equal size , such that the total bending energy vanishes and the membrane area is kept constant . in fig .
[ fig5a ] , we show the radii of the mother and daughter vesicles as function of the number of inclusions . for a given number of @xmath53 daughter vesicles ,
there is a maximal number of inclusions @xmath54 ( @xmath55 ) that still allows to obtain a zero energy state , for which mother and daughter vesicles have equal sizes . with @xmath44 inclusions
splits into one large mother-vesicle and several small daughter vesicles , in the figure the cases of 1 , 2 , 3 , and 4 daughter vesicles are shown .
the same parameters as in fig .
[ fig5 ] are used . for 100 inclusions
, the initial vesicle has vanishing bending energy . for more than 100 inclusions ,
the sizes of mother and daughter vesicles are plotted .
the upper branch always gives the radius of the mother vesicle , the lower branch is the size of the daughter vesicles .
for a fixed number @xmath56 of daughter vesicles ( as indicated ) , there is a maximum number of inclusions that allows the formation of a state with vanishing bending energy ( for which mother and daughter vesicles have equal sizes ; filled circles ) . ]
if the system can split up into @xmath57 smaller vesicles , it can also split up into a larger number of small vesicles @xcite . for a vesicle with total number of inclusions
@xmath58 and radius @xmath59 , bending energy minimization predicts for the radii and inclusion number on mother ( @xmath60 , @xmath61 ) and daughter ( @xmath62 , @xmath63 ) vesicles : @xmath64 because of the degeneracy of the states with vanishing bending energy , thermal fluctuations and the budding pathway play a decisive role to determine how a large vesicle with high inclusion density splits up into smaller vesicles .
note that the results of our analytical model so far do not depend on the value of the bending rigidity of the bilayer .
, describes the dependence of the optimal vesicle radius , @xmath65 , on the degree of aggregation of the inclusions in the membrane .
the total inclusion area is kept constant and the reference density @xmath66 corresponds to @xmath67 . ] a direct attractive interaction between inclusions can induce cluster formation @xcite . in this case , the preferred curvature radius , @xmath65 , not only depends on the number and geometry of the inclusions , but also on cluster size . for given inclusion curvature radius @xmath68 , opening angle @xmath69 , and fixed density @xmath66 , inclusion clusters with a larger opening angle @xmath28 and reduced density , @xmath70 , have a stronger effect on the curvature of the membrane than homogeneously distributed inclusions . the preferred curvature radius for clusters with opening angle @xmath28 is @xmath71 .
we call the normalized curvature radius @xmath72 the coagulation factor , because it multiplies the curvature radius for the reference inclusion density and opening angle @xmath69 @xcite , see fig .
[ fig6 ] . for a fixed number @xmath73 of inclusions , the preferred curvature radius decreases when aggregates are formed , i. e. the efficiency with which the inclusions influence the membrane curvature increases .
cluster formation therefore also shifts the high - density regime to smaller numbers of inclusions for the same vesicle radius , @xmath74 , and may cause a large vesicle to break up into smaller vesicles .
for finite temperature , the translational entropy of the inclusions contributes to the free energy .
we distinguish a crystalline hexagonal phase and a disordered fluid phase for which we construct free energy functionals .
phase diagrams have been calculated more rigorously in ref .
@xcite in the limiting case of an almost planar membrane and for inclusions that are slightly stiffer than the membrane and weakly curved but not in the context of budding .
we neglect the interaction between inclusions by thermal membrane undulations . for a pair of inclusions and an arbitrary orientation of the inclusion axis , to lowest order of @xmath75 the deformation energy is @xcite @xmath76 and the undulation - induced interaction energy @xcite
@xmath77 the ratio of the membrane deformation - induced repulsion to the undulation - induced attration in a planar membrane is @xmath78 .
the undulation - induced attraction can be neglected if it is one order of magnitude smaller than the deformation interaction ; for @xmath79 this is the case for @xmath80 , for @xmath81 already for @xmath82 . for low inclusion densities
, the undulation - induced interaction energy decays with the square of the density while the free energy due to the inclusion entropy is of the order of @xmath83 ; in case of the phase diagrams in fig .
[ fig12 ] , the undulation free energy is about @xmath84 . for inclusions on a hexagonal lattice
, each inclusion corresponds to three pair interactions and a radius of the deformation patch , @xmath30 , that is half the distance between the inclusions .
the effective pair potential , obtained from eq .
( [ eq10 ] ) , is thus @xmath85 for inclusions on a planar membrane ( @xmath25 ) and for large @xmath86 , i. e. for @xmath87 , the repulsive interaction potential decays with a power law , @xmath88 . to determine the free energy of this system , we employ the method developed for suspensions of repulsive colloids @xcite , i. e.we mimic the interaction potential by hard discs with an effective radius , @xmath89 .
the radius is calculated from a comparison of the membrane deformation energy with the thermal energy , @xmath83 .
we use a modified barker - henderson method that is appropriate for soft potentials @xcite , @xmath90 \right\ } d \rho \ , , \label{eq14}\ ] ] where the upper integral boundary is determined by @xmath91 @xcite .
the effective hard - disc radii therefore depend on the geometry of the inclusion , the bending rigidity of the membrane , and on the radius of the vesicle on which the inclusions are placed . in fig .
[ fig7 ] , the effective hard - disc radii , @xmath89 , are plotted for inclusions with various opening angles as function of the vesicle radius , @xmath19 ( an extremely large radius @xmath92 of the vesicle is used to describe inclusions in planar membranes ) . the hard disc radii increase with increasing vesicle radius
; the increase of @xmath93 with @xmath19 is the stronger , the larger the opening angle @xmath28 of the inclusion is .
for example , the inclusions with @xmath94 and @xmath95 can approach each other by diffusion about an order of magnitude closer on a vesicle with radius @xmath96 than this is possible on a planar membrane .
consequently , the translational entropy of the inclusions lowers the free energy on the vesicle compared with the planar membrane .
( see eq .
( [ eq14 ] ) ) .
all inclusions have the curvature radius , @xmath97 , and the opening angles @xmath98 , @xmath99 , ... , @xmath100 ( from top to bottom ) .
the projected inclusion radii are in the range @xmath101 , the membrane bending rigidity is @xmath102 . ] as discussed in sec .
[ sec3d ] , cluster formation of inclusions increases their effect on the membrane curvature .
the effect of cluster formation on the area fraction of effective hard discs is plotted in fig .
[ fig8 ] . for @xmath102 , which is a typical value for a lipid bilayer , clustering on vesicles with large radii , @xmath19 , strongly increases the area fraction of the effective hard discs . to illustrate the strong effect of the bending rigidity on the effective hard - disc radius , which enters the calculation of the radius via the exponential function in eq .
( [ eq14 ] ) , we plot the effective hard disc radii for @xmath103 ; the increase of the area fraction of the hard discs with @xmath28 is much smaller than for @xmath102 .
thus the translational entropy of the inclusions plays a much more important role for smaller @xmath0 @xcite . and opening angles @xmath104 .
for a fixed number of inclusions , cluster formation leads to a larger area fraction of the effective hard discs and finally to crystallization .
the area fraction of the effective hard discs with opening angle @xmath28 is normalized by the area fraction of effective hard discs with @xmath105 : an increase of @xmath28 corresponds to a decrease of @xmath20 ( compare fig .
[ fig6 ] ) .
the inclusions are placed on vesicles with @xmath102 and various radii @xmath106 ( dotted ) , @xmath107 ( short - dashed ) , @xmath108 ( long - dashed ) , and @xmath92 ( solid ) . for comparison ,
the area fraction of effective hard discs is also shown for @xmath109 and @xmath92 ( dashed - dotted ) . ]
the free energy of a fluid of hard discs can be very well described by the carnahan - starling free energy @xcite .
it is the sum of the ideal - gas free energy , @xmath110 , where @xmath111 is the thermal wavelength ( see e. g. ref .
@xcite ) , and the excess free energy @xmath112 @xcite .
the latter is calculated from the carnahan - starling equation of state @xcite , @xmath113 with the hard - disc area fraction , @xmath114 .
integration of the thermodynamic relation @xmath115 finally gives the excess free energy , @xmath116 the carnahan - starling excess free energy diverges at the crystallization transition of the effective hard discs .
because in the fluid as well as in the crystalline phase of the inclusions the squared thermal wavelength enters through the same constant and additive term , we consistently replace it in both cases without loss of information by the projected area of the inclusion , @xmath117 , such that @xmath118 .
usually the translational entropy favors a homogeneous distribution of particles .
however , because the effective hard - disc radius depends on the membrane curvature , on a deformable membrane , a homogeneous distribution of inclusion does not need to be the most favourable state . instead
, the inclusion density on the bud can be higher than on the mother vesicle because of the screened repulsive interactions .
[ fig9 ] shows the free energies of a fluid of effective hard discs with curvature radius @xmath97 for various opening angles in a lipid bilayer with @xmath102 . for nearly identical curvature radii of vesicle and inclusions
, the effective hard - disc radius almost coincides with the geometric hard - disc radius of the inclusion .
( solid line ) , as well as the free energies for effective hard discs for inclusions with curvature radius @xmath97 and various opening angles on vesicles with @xmath102 : @xmath119 and @xmath120 ( long dashed ) , @xmath119 and @xmath121 ( short dashed ) , @xmath122 and @xmath120 ( long - dashed dotted ) , @xmath122 and @xmath121 ( short - dashed dotted ) , @xmath123 and @xmath120 ( dotted ) , and @xmath123 and @xmath121 ( double dashed ) . ] , for inclusions with @xmath94 , @xmath123 in a membrane with @xmath124 for several vesicle radii : @xmath125 ( solid ) , @xmath126 ( long - dashed ) , @xmath127 ( short - dashed ) , @xmath128 ( dotted ) , @xmath120 ( long dashed - dotted ) , @xmath129 ( short dashed - dotted ) . for low densities ,
the inclusions are in a fluid phase and the free energy is given by the membrane bending energy plus the carnahan - starling excess free energy for the effective hard discs . for high densities ,
the inclusions are in a crystalline phase and the free energy is given by the membrane bending energy and the free energy of a harmonic crystal . ]
we construct the free energy per inclusion in the crystalline phase from the membrane bending energy and the fluctuation free energy of a harmonic crystal , and in the fluid phase from the sum of the membrane bending energy and the translational entropy of the inclusions @xcite . for the harmonic crystal , the spring
constant @xmath130 is obtained for a hexagonal lattice with the interaction potential in eq .
[ eq9new ] , @xmath131 the free energy contribution of the positional fluctuations of the inclusions is therefore @xmath132\ ] ] or after replacement of the thermal wavelength by the inclusion size @xmath133 \ , .\ ] ] whenever we use the free energy for the crystalline phase in this paper , the lindemann parameter remains below the critical value for melting of the inclusion crystal @xcite .
the transition between the fluid and the crystalline phase already occurs below the crystallization transition of the effective hard discs , @xmath134 . in fig .
[ fig10 ] , the free energy per inclusion is plotted for several vesicle radii . entropy reduces the optimal bud radius for a given inclusion density compared with eq .
( [ eq11 ] ) .
however , the bending energy alone still provides a good estimate for the optimal bud radius , because of the strong increase of the free energy per inclusion for low inclusion densities ( see fig .
[ fig10 ] ) . of inclusions with @xmath94 and @xmath122 on ( initially ) a single vesicle of radius @xmath19 . for small @xmath20 or @xmath19
, the energetically favorable state is the single vesicle ; fission , first into two , and finally into @xmath135 or more vesicles is expected to occur when @xmath20 and/or @xmath19 is increased ( more than 3 vesicles are not resolved by the calculation ) . _ ( a ) @xmath136 : _ in the two - vesicle regime , a region where the two vesicles have equal sizes grows with decreasing @xmath0 and bounds the three - vesicle regime .
the lines depict the border between equally and differently sized vesicles for @xmath137 , @xmath138 , @xmath139 , @xmath140 , @xmath79 .
( the vesicle sizes are not resolved in the @xmath141 region ) . _
( b ) @xmath142 : _ fission occurs already for smaller values of @xmath20 and @xmath19 . in the @xmath143-vesicle region
the two vesicles have different sizes for @xmath137 , @xmath138 , @xmath139 , and @xmath140 . for @xmath144 , in a small region
both vesicles have equal sizes . ] from the total free energy in sec . [ sec4 ]
, we calculate vesiculation diagrams starting with a single vesicle of radius @xmath19 and a given number of inclusions for several values of @xmath0 .
because the topology changes when buds detach , the value of @xmath1 plays an important role for vesiculation .
[ fig12 ] shows vesiculation diagrams for @xmath136 and for @xmath142 ( the ratio @xmath145 is still under debate , see ref .
we calculate whether the single vesicle is the most stable state or if fission in two or more smaller vesicles is favorable . with increasing initial vesicle radius and inclusion density ,
fission becomes more likely to occur first into two , at even larger @xmath19 or @xmath20 into three or more vesicles
. for @xmath146 , fission in the bending - energy dominated regime produces two smaller vesicles that in general have different sizes , see eq .
( [ eqvesbr ] ) .
if entropy is important , the two vesicles may have equal size . in fig .
[ fig12 ] ( a ) , it is shown that a regime of equally - sized vesicles develops , bordering the three - vesicle regime and increasing in size with decreasing @xmath0 . within the error bars of our calculation ,
we find that the boundaries between one , two , and three vesicles are independent of the value of the bending rigidity for @xmath147 . for @xmath148
, vesiculation takes place for smaller initial vesicle radii and inclusion densities than for @xmath149 , because each additional vesicle lowers the free energy by @xmath150 . in the entire two vesicle regime ,
both vesicles have different sizes for @xmath137 , @xmath138 , @xmath139 , and @xmath140 , while for @xmath144 a small region of equally sized vesicles is found , see fig .
[ fig12 ] ( b ) . note that for bud formation , which has to occur before vesiculation , the value of @xmath1 is irrelevant and the phase diagrams for @xmath136 apply ( assuming that the buds are connected by catenoidal necks with vanishing bending energy , and that the membrane area needed to form the neck is negligible ) . while the bud is being formed and has not yet detached , the integral over the gaussian curvature and therefore the contribution of the saddle splay modulus to the deformation energy stays constant .
however , a negative saddle splay modulus facilitates the neck between two vesicles to break . in this case
, the budded state can act as energy barrier for vesiculation that prevents fission , separating a high - energy single - vesicle state and a low - energy state of several smaller vesicles .
to shed more light on the role of the budding pathway , we compare the typical diffusion time of the inclusions with the relaxation time of the membrane conformation on the same length scale .
if the diffusion of inclusions is fast compared with the relaxation time of the membrane , the initial membrane shape is decisive ; for a fast membrane relaxation , the initial distribution of inclusions mainly determines the budding process .
the diffusion time is @xmath151 where @xmath152 is a characteristic length scale that the inclusion has diffused and @xmath153 is the diffusion coefficient of the inclusion .
a typical value is @xmath154 for the diffusion of lipids and the diffusion coefficient for proteins in cell membranes can be up to two orders of magnitude smaller @xcite .
the relaxation time of the membrane is @xmath155 @xcite , where @xmath156 is the viscosity of the surrounding water and @xmath152 is the wavelength of the membrane undulations . from the cubic versus the quadratic dependence on @xmath152
, we find that for small @xmath152 , @xmath157 , while for large @xmath152 , @xmath158 . for @xmath159 and @xmath160 ( which is an upper bound for the diffusion coefficient of proteins ) , we find that @xmath161 for @xmath162 .
this length is much larger than @xmath163 , which is the size of cells @xcite or giant unilamellar vesicles @xcite , thus the initial aggregation of inclusions is diffusion - limited .
we therefore assume that inhomogeneities in the protein distribution on the membrane will immediately lead to a membrane deformation that minimizes the bending energy . the larger the initial inclusion density , the larger the spontaneous curvature of the membrane ( compare eq .
( [ eq11 ] ) ) and the smaller the size of the buds that are formed .
bud formation is initialized in some regions of the membrane that have a noticeably higher inclusion density than others .
the relative protein density fluctuations decrease with the size of a membrane patch which is considered . if we assume a random distribution of inclusions , for a patch with @xmath164 inclusions the relative fluctuations of the inclusion number are of the order of @xmath165 .
thus for small patches , the inhomogeneities are strongest and budding will preferably occur on the smallest possible lengthscale .
a small average membrane curvature with appropriate sign further attracts proteins to those regions where the bending energy per inclusion is already reduced .
however , this clustering of inclusions during the budding process is hindered by a ring with opposite membrane curvature that forms the neck of a growing bud .
this ring acts as energetic barrier that prevents further inclusions to enter a patch of the membrane where budding has already started @xcite .
because of the neck formation and the diffusion - limited budding process , the bud size is roughly determined by the initial inclusion density on the membrane .
inclusions of size @xmath166 on a vesicle with @xmath129 , @xmath81 , and @xmath167 , as function of @xmath28 .
lines show bending energy ( long - dashed ) , bending energy and inclusion entropy ( short - dashed ) , saddle - splay energy ( dotted ) , and total energy ( solid ) . for comparison , we also plot the simulation data of ref .
@xcite ( symbols indicate different calculation methods @xcite ) , shifted by @xmath168 ( see main text ) .
the deviation of the simulations and our theory for @xmath169 might be due to the surface tension used in the simulation , which is not included in our model . ] budding due to membrane inclusions has been studied recently with computer simulations @xcite . in ref .
@xcite , entire vesicles with inclusions are simulated where the membrane is modelled as a dynamically triangulated surface . in ref .
@xcite , coarse - grained model lipids are used to study budding for planar bilayer patches . in fig .
[ fig11 ] , the different contributions to the free energy per inclusion needed to graft @xmath170 inclusions with given projected radius @xmath171 on a vesicle with radius @xmath172 are plotted as function of the opening angle , @xmath28 .
the inclusions are in the fluid phase . for comparison
, we also plot the simulation data of ref .
@xcite , shifted by @xmath168 because our model does not account for thermal undulations of the membrane conformation .
this energy difference is extracted from the simulation data for @xmath173 and for @xmath174 , see fig .
6a in ref .
@xcite . however , the very good match is somewhat fortuituous because we replace the thermal wavelength in the ideal gas free energy by the inclusion size , such that it is similar to the free energy obtained on a triangulated vesicle with a bond length that approximately equals to the inclusion diameter , compare @xcite .
we consider curvature radii of the inclusions that are both smaller and larger than the curvature radius of the vesicle . for @xmath175 and @xmath176 ,
such that @xmath171 , bending energy is needed to insert the flat inclusion into the curved vesicle .
this bending energy cost decreases with increased opening angle @xmath28 and is zero for @xmath177 where the curvature radius of the inclusion equals the curvature radius of the vesicle .
if @xmath28 is further increased , the bending energy per inclusion continues to decrease as more and more of the vesicle area consists of catenoidal patches around the inclusions . for @xmath178 , i. e. for even larger opening angles than plotted in the figure ,
the inclusions have optimal density and the bending energy gained by grafting all @xmath170 inclusions to the vesicle is @xmath179 .
in addition to the bending energy , there is an energy cost @xmath180 for grafting that arises due to the saddle - splay modulus , which has been chosen to be @xmath181 for consistency with ref .
the energy cost per inclusion only depends on the geodesic curvature of the membrane at the inclusion , i. e. on the opening angle @xmath28 , which implies @xmath182 .
therefore it is independent of vesicle radius and inclusion density and is only important to calculate the chemical potential for the inclusions on the surface ; the budding transition at given inclusion density in the membrane is independent of @xmath183 . in the simulations presented in ref .
@xcite , budding is induced by inclusions with @xmath97 that are initially placed in a regular array on a planar membrane with @xmath184 . under the assumption that the initial inclusion density , @xmath185 , determines the bud size ( see sec . [ sec5 ] )
, we can roughly predict the bud radius from eq .
( [ eq11 ] ) .
possible bud radii are estimated by comparing the free energies for different vesicle radii in fig .
[ fig10 ] at the initial inclusion density . the parameters for which the free energies are plotted in fig .
[ fig10 ] are chosen to match the bending rigidity and the inclusion geometry of the large inclusions in ref .
@xcite with @xmath186 @xcite . for an initial inclusion density in the planar membrane , @xmath187 , which is estimated by visual inspection from the simulation snapshots
, we find that the inclusions are in the crystalline phase on the vesicle with @xmath107 ( i. e. in a planar membrane ) .
the free energy per inclusion is about @xmath188 .
significantly smaller free energies per inclusion can be found for vesicle radii @xmath189 ; the optimal vesicle radii are @xmath190 with free energies per inclusion of about @xmath191 .
these radii agree well with the observed bud radius of @xmath192 that formes in the simulations as final state via an initially slightly larger bud . the optimal bud radius from eq .
( [ eq11 ] ) is approximately @xmath193 .
similarly , for the very large inclusions in ref .
@xcite ( @xmath194 ) we predict bud radii , @xmath195 , as observed in the simulations ; based on bending energy only we find from eq .
( [ eq11 ] ) @xmath196 . for the small inclusions in ref .
@xcite ( @xmath197 ) that are already in the planar membrane almost in the fluid phase , our model predicts for the @xmath198 inclusions studied in the simulations a maximal gain for the free energy per inclusion of @xmath199 for @xmath200 and a strong decrease of this energy gain for smaller vesicle radii .
already for @xmath201 and @xmath202 inclusions , the energy per inclusion on the bud and in the plane are approximately equal .
larger energy gains are possible for larger bud radii , for which many more than the simulated 36 inclusions and a larger area of the bilayer patch are needed . from these considerations ,
it is not surprising that no budding is observed for the small inclusions in the simulations of ref .
we have calculated the membrane - mediated interaction of conical inclusions in a lipid bilayer and the inclusion entropy , which allow the prediction of budding transitions and vesiculation .
our model is based on the membrane bending energy ; with this contribution alone , the spontaneous curvature of a bilayer with inclusions as well as budding can be predicted for many systems , ranging from protein inclusions to viral budding .
although the interaction between the inclusions by membrane deformation is repulsive , the screening of the repulsive interaction due to the average membrane curvature allows higher inclusion densities on a bud than in the initial vesicle .
translational entropy of the inclusions favors the formation of equally - sized daughter vesicles and lifts the degeneracy that is found for states with vanishing bending energy
. from our calculations , the following picture of the effect of the bilayer deformation by curved inclusions emerges . for low inclusion density
, the membrane around each inclusion assumes a catenoid shape of vanishing curvature energy . at optimal inclusion density ,
the catenoids are closely packed and the bending energy for the entire vesicle vanishes . for high inclusion density , the boundary conditions for
the membrane deformation around the inclusion do not allow the formation of catenoidal patches , and the inclusions always feel the membrane - mediated repulsive interaction with neighboring inclusions .
in this regime , bud formation can occur .
if we assume a specific biological mechanism that leads to formation of clusters with well - defined and limited size , we find that such a mechanism can induce bud formation and vesiculation without the need to insert additional conical proteins into the cell membrane : cluster formation reduces the preferred curvature radius of the membrane .
we quantify the effect of aggregation by the coagulation factor , which describes how the preferred curvature radius for a fixed amount of inclusions changes with the cluster size . in general , our analytical model is applicable for a wide range of length scales and inclusion geometries .
computer simulations are usually designed only for a specific length scale , e. g. a length scale comparable to the length scale of lipids in ref .
@xcite or the lengthscale of entire vesicles in ref .
the good agreement with the simulation results of refs .
@xcite suggests that the approximations used in our calculations are justified .
we argue that the undulation - induced attraction can be neglected compared with the deformation - induced repulsion and the translational entropy of the inclusions for @xmath203 , i. e. @xmath82 for @xmath204 .
for example , the bar protein induces a local membrane curvature with an opening angle @xmath205 @xcite .
clathrin can induce a variety of opening angles @xcite .
the number of the inclusions per bud is determined by the budding process . around a growing bud
, a neck forms that presents an energetic barrier for the diffusion of inclusions .
because the deformation of the lipid membrane typically occurs much faster than the diffusion of the inclusions within the membrane , the number of inclusions per bud is well determined by the initial inclusion density in the membrane . from this , we can estimate a range of possible bud radii , which agrees well with the simulations in ref .
. it would be interesting to test the validity of our model in the limits of ( a ) very small inclusions , such as lipids with large headgroups , when the description of the lipid membrane by curvature elastic constants may not be appropriate and ( b ) very floppy membranes , when neglecting the thermal membrane undulations may not be justified . | conical inclusions in a lipid bilayer generate an overall spontaneous curvature of the membrane that depends on concentration and geometry of the inclusions .
examples are integral and attached membrane proteins , viruses , and lipid domains .
we propose an analytical model to study budding and vesiculation of the lipid bilayer membrane , which is based on the membrane bending energy and the translational entropy of the inclusions .
if the inclusions are placed on a membrane with similar curvature radius , their repulsive membrane - mediated interaction is screened .
therefore , for high inclusion density the inclusions aggregate , induce bud formation and finally vesiculation . already with the bending energy alone our model
allows the prediction of bud radii .
however , in case the inclusions induce a single large vesicle to split into two smaller vesicles , bending energy alone predicts that the smaller vesicles have different sizes whereas the translational entropy favors the formation of equal - sized vesicles .
our results agree well with those of recent computer simulations . |
flows past a truncated cylinder are involved in a wide variety of problems dealing with the design of e.g. buildings , ship funnels or chimney stacks .
the typical values of the reynolds number @xmath1 considered in these problems are greater than @xmath2 .
in contrast , though relevant with probes used in intrusive measurement methods @xcite and heat transfer in electronic circuit boards @xcite , the flow dynamics at lower @xmath1 , _
i.e. _ up to @xmath3 , have been scarcely investigated . in these regimes ,
four main structures are generated in the cylinder wake : the horseshoe system , the free shear layer stretching from the cylinder free end , those stretching from both cylinder lateral faces and a system of trailing vortices . by convenience ,
we assume that the cylinder is mounted on the bottom wall .
the horseshoe pattern results from the interaction between the boundary layers arising respectively at the cylinder upstream face and at the bottom wall @xcite .
it is formed by a system of swirls generated by flow separation at the cylinder front that spirals around the cylinder .
it remains steady for @xmath4 without any restriction on the thickness of the wall boundary layer @xcite .
the system of trailing vortices consists of two pairs of counter - rotating streamwise vortices located below the cylinder tip and above the cylinder base , respectively . according to their location ,
they are referred to as _ tip _ and _ base vortices_. in both cases , their origin still remains an open question . from experimental flow visualisations in the wake of a truncated circular cylinder at @xmath5 , @xcite and @xcite suggested that the tip vortices were generated above the upper cylinder face from the rolling - up of the lateral ends of the upper free shear layer , while @xcite interpreted them as resulting from the tilting of the lateral free shear layers in the vicinity of the cylinder free end . from experimental measurements in the same configuration at @xmath5 , @xcite suggested that the tip vortices had nothing to do with the lateral free shear layers .
investigating experimentally the flow past a truncated square cylinder for @xmath6 , @xcite drew the same conclusions as @xcite . on the other hand ,
little information is available on the base vortices .
@xcite , @xcite and @xcite agree that they result from the tilting of the lateral free shear layers in the vicinity of the bottom wall .
@xcite furthermore indicated that the base vortices were initially aligned along the spanwise direction and then tilted along the streamwise axis between the mid - span and the free end .
the cylinder aspect ratio @xmath7 , defined as the ratio of the cylinder height @xmath8 to the characteristic length of its cross - section @xmath9 , determines which of the tip or the base vortices prevail in the wake @xcite . also , increasing the thickness of the boundary layer at the bottom wall strengthens the base vortices @xcite .
since the horseshoe structure remains steady for @xmath4 , any vortex shedding occurring in this regime is fed by the free shear layers only and the cylinder aspect ratio @xmath7 determines whether the vortex shedding is asymmetric , symmetric or absent . for high values of @xmath7 ,
the vortex shedding is governed only by the lateral free shear layers and an asymmetric krmn - like vortex street is observed @xcite . for intermediate values , the transition to unsteadiness leads to a symmetric vortex street formed by hairpin vortices aligned on the wake centreline @xcite .
finally , small values of @xmath7 prevent any kind of vortex shedding @xcite .
the critical values of @xmath7 separating these regimes depend on the shape of the cylinder cross - section .
the latter affects the shape of the free shear layers as well as the position of the wake structures with respect to one another and thereby the mode of vortex shedding @xcite .
also , the thickness of the boundary layer at the bottom wall has a noticeable influence on the determination of the critical values of @xmath7 @xcite .
the formation mechanism of the krmn vortices relies on the alternate roll - up and shedding of the lateral free shear layers as in non - truncated cylinder wakes .
in contrast , there has not been yet any clear agreement on the scenario leading to the symmetric hairpin vortex street . from experimental investigations
, @xcite suggested that both lateral free shear layers joined the top one to form a single entity in the near - wake and as the latter became unstable , an arch - type vortex was formed and released in the wake .
this view is supported by @xcite who performed numerical simulations of the flow past a truncated square cylinder with @xmath10 for @xmath11 and by experimental flow visualisations by @xcite at @xmath12 .
the latter authors also included the tip and base vortices within the arch - vortices released in the wake .
@xcite simulated the flow past a truncated square cylinder featuring @xmath13 at @xmath14 and claimed that hairpin vortices were originally vortices that detached from the free shear layer stretching from the cylinder free end and then grew into hairpin vortices . until now
though , the generation mechanism of hairpin vortices in the wake of a truncated cylinder has never been the object of any dedicated study so it remains rather unclear . in this article
, we investigate the wake of a truncated cylinder of square cross - section using three - dimensional direct numerical simulations .
our goal is to clarify the dynamics of the formation of hairpin vortices .
our approach relies on a detailed analysis of the flow structures in the steady regime .
we restrict our investigations to cylinder wakes confined in a duct of rectangular cross - section , in which the inlet velocity profile is a well defined paraboloid flow . under these conditions , the scenario we shall bring forward is new and differs from those inferred in previous works in several points .
in addition , we have detected a set of @xmath0-shaped vortices forming between two successive hairpin vortices .
we shall also give the details of the mechanism of their formation . the present flow configuration and the numerical set - up
are introduced in section [ sectionconfiguration ] . in section [ sectionsteadyregime ] , we identify the steady flow patterns and their dynamics throughout the steady flow regime .
section [ sectionunsteadyregime ] is dedicated to the description of the formation and release of hairpin vortices at the onset of vortex shedding .
the formation of the @xmath0-shaped vortices and the effects of increasing @xmath1 within the unsteady regime are also detailed in this section .
finally , in section [ sectionflowcoefficients ] , we show how the evolution of the flow coefficients is affected by the flow dynamics and , in particular , how the appearance of secondary recirculation regions on the cylinder lateral faces noticeably affects the viscous drag coefficient .
we consider the flow of an incompressible fluid ( density @xmath15 , kinematic viscosity @xmath16 ) past a truncated square cylinder in a rectilinear duct of rectangular cross - section as shown in figure [ configuration ] . a square cylinder ( width @xmath9 and height @xmath8 ) is mounted on the duct bottom wall at equal distance from both duct side walls .
the upstream cylinder face is normal to the streamwise direction taken along the @xmath17axis .
the cylinder axis is parallel to the @xmath18axis .
the origin of the frame of reference is located at the centre of the upper cylinder face . the duct height ( _ resp . _
width ) along the @xmath18axis ( _ resp . _
@xmath19axis ) is @xmath20 ( _ resp .
_ @xmath21 ) .
the present configuration features @xmath22 , @xmath23 and @xmath24 , which yields a transverse ( _ resp .
_ spanwise ) blockage ratio @xmath25 ( _ resp .
_ @xmath26 ) and a cylinder aspect ratio @xmath27 .
@xmath28 is low so that it does not have any noticeable effect on the horseshoe system @xcite , while the effects of the spanwise flow confinement shall be discussed throughout this study .
we study the flow using three - dimensional direct numerical simulations with a finite - volume code based on the openfoam 1.4.1 framework @xcite .
the flow motion is governed by the navier - stokes equations ( [ massequation ] ) and ( [ momentumequation ] ) , written in their non - dimensional form as : @xmath29 where @xmath30 and @xmath31 are the velocity and the pressure fields , respectively , and @xmath32 ( @xmath33 is the maximum velocity at the inlet ) .
the numerical domain is @xmath34 long upstream and @xmath35 long downstream the cylinder .
a no - slip impermeable boundary condition is imposed at all walls through a homogeneous dirichlet condition and a homogeneous neumann condition is applied for the velocity at the outlet .
this geometry ensures that the outlet boundary has no noticeable feedback onto the cylinder wake as it is located more than @xmath36 away from the cylinder @xcite .
preliminary numerical computations at a prescribed @xmath1 are performed in the duct with the cylinder absent .
the resulting velocity profile of the fully - established flow in the cross - section is used as the inlet boundary condition for the velocity in the three - dimensional simulations with the cylinder present . for the pressure field ,
a homogeneous neumann condition is imposed at all boundaries but at the outlet where a homogeneous dirichlet condition is applied .
the flow equations ( [ massequation ] ) and ( [ momentumequation ] ) are solved in a segregated way and the piso algorithm as detailed in @xcite is used to deal with the pressure - velocity coupling .
second - order discretisation using central difference schemes in space and a quadratic backward scheme in time has been used on a non - uniform cartesian mesh .
the numerical code has been successfully tested on the two - dimensional flow past a square cylinder at @xmath37 @xcite and by simulating the three - dimensional flow past a non - truncated square cylinder at @xmath38 as described in @xcite .
the latter test has been performed on a non - uniform mesh featuring @xmath39 points , _
i.e. _ slightly less points than in @xcite . the non - dimensional minimum grid size was set to 0.03 so that a better resolution of the boundary layer was achieved than in @xcite , whose computations were performed on a fully uniform mesh .
we have observed mode - a flow patterns together with dislocated krmn vortices as in @xcite .
our computed values of the strouhal coefficient @xmath40 and the pressure part of the drag coefficient @xmath41 ( @xmath42 is the vortex shedding frequency and @xmath43 the force obtained by the integration of the pressure field over the cylinder surface ) have recovered the values of @xcite within 1% and 3% respectively , which ensures that the present numerical code addresses the dynamics of three - dimensional cylinder wakes within a fully satisfactory accuracy . [
cols="<,^,^,^",options="header " , ] we have then implemented a test to estimate the required number of mesh nodes for the simulations with the truncated cylinder wake .
the main characteristics of the tested meshes are provided in table [ charmesh ] . in the plane normal to the cylinder axis ,
we have taken over the same mesh as that used in the test on the non - truncated cylinder wake which have yielded very satisfactory results .
we have then varied the number of points along the @xmath18axis and the distance @xmath44 between the cylinder top face and the grid point nearest to it .
we have simulated the flow at @xmath38 over 1000 turnover times @xmath45 and derived the total drag coefficient @xmath46 ( @xmath47 is the streamwise component of the total force exerted by the flow on the cylinder ) and the strouhal number @xmath48 in each case .
we have found that both the errors in @xmath49 and @xmath48 relative to mesh m3 decreased with the number of nodes , which shows good convergence . in order to save cpu time and keep a reasonable accuracy in our computations
, we shall perform all our simulations with m2 mesh .
a final validation test has been performed on the configuration of @xcite .
the authors investigated the flow past a truncated circular cylinder with aspect ratio @xmath50 between two infinite parallel impermeable walls placed at the cylinder bottom end and at 5 cylinder diameters above the cylinder free end .
we have simulated the flow at @xmath51 using the same geometry , computational domain and boundary conditions . for this test ,
we have designed a structured mesh consisting of a polar mesh embedded in a square of two cylinder diameter width and a rectangular cartesian grid covering the rest of the computational domain .
the mesh contains about @xmath52 nodes with a non - dimensional minimum grid size of 0.02 between the cylinder upper face and the grid point nearest to it .
we have computed the strouhal number at @xmath51 and found @xmath53 which represents an error of @xmath54 compared to the value of @xmath55 in @xcite .
in addition , our simulations have recovered a uniform spanwise distribution of the strouhal number along the cylinder span in agreement also with @xcite . in summary ,
our code has been successfully validated on two different configurations of flow past a truncated cylinder involving either a transverse or a spanwise flow confinement . in both cases the unsteady flow dynamics
have been accurately captured by the computation of the strouhal number .
we shall now consider the truncated cylinder wake in a rectangular duct in which the flow is confined along both transverse and spanwise directions .
to explain the formation mechanism of hairpin vortices , we need to identify flow structures and , in particular , vortices . establishing an objective definition of a vortex
has been the subject of several studies in the literature . @xcite
have assessed different approaches of identification and tracking of vortical structures .
one of them is based on the analysis of the eigenvalues of the symmetric tensor @xmath56 , where @xmath57 and @xmath58 are the respective symmetric and antisymmetric part of the velocity gradient tensor @xmath59 . in this approach
, a vortex core corresponds to a pressure minimum not induced by viscous effects nor unsteady straining .
it is defined as a connected region with two negative eigenvalues of @xmath56 .
a vortex is therefore detected at a given location in the fluid domain if the median eigenvalue , denoted @xmath60 , is locally negative .
this approach is particularly efficient at spotting ring - type vortices @xcite .
unfortunately , it delivers no information on the rotation directions of the vortical structures . for this reason
, flow patterns will also be characterised by iso - surfaces of vorticity , especially in the unsteady flow regime . in the steady one
, we will analyse the streamlines , since they match the flow trajectories @xcite .
we have performed nine successive simulations at increasing @xmath1 for @xmath61 .
unsteady flows have been computed over a total simulation time close to @xmath62 .
our computations have yielded a steady flow regime for @xmath63 and an unsteady regime for @xmath64 . in no case
does the resulting flow feature any closed streamlines , in line with the findings of @xcite . .
iso - surfaces of @xmath65 ( a ) and kinematic streamlines with horseshoe pattern in pink , trailing vortices in green ,
streamlines rejoining the stagnation points on the upstream cylinder face in cyan , head vortex in red , spanwise vortices in blue , black streamlines are only deflected by the cylinder wake without noticeably influencing it : three - dimensional ( b ) , side ( c ) and top ( d ) views.,scaledwidth=99.0% ] figure [ flowdyn](a ) shows the steady flow patterns at @xmath51 depicted by iso - surfaces of @xmath65 , _
i.e. _ less than @xmath66 of the absolute minimum in @xmath60 , and by kinematic streamlines in figures [ flowdyn](b - d ) . in figures [ flowdyn](b - c )
, one sees that the black streamlines that are simply deflected around the wake structure separate from those that enter it at saddle point @xmath67 located at @xmath68 in the centre - plane ( @xmath69 ) .
all streamlines other than the black ones enter the wake structure .
pink and green streamlines originate in the upstream centre - plane ( @xmath69 ) , follow a similar path around the cylinder and separate at saddle point @xmath70 located at @xmath71 in the centre - plane ( @xmath69 ) @xcite .
pink lines depict the horseshoe pattern as shown in experiments by e.g. @xcite .
the front horseshoe swirl , denoted hs in figure [ flowdyn](a ) , is observed upstream of the cylinder base at @xmath72 .
the green set of streamlines impacts the bottom duct wall downstream of a nodal point of attachment @xmath73 just in front of the upstream cylinder face , as observed in @xcite .
it then recirculates behind the cylinder within the horseshoe pattern where it reaches two foci , denoted @xmath74 and @xmath75 in figure [ flowdyn](d ) , and subsequently spirals upwards and eventually tilts along the @xmath17axis to create a pair of counter - rotating streamwise vortices .
this corresponds to the base vortices identified by @xcite and @xcite .
the foci were also detected in numerical simulations by @xcite and @xcite , but were not linked to the formation of base vortices . at @xmath51 , these vortices are very weak so that they only induce a weak pressure minimum that is difficult to capture with the iso - surfaces of @xmath60 . far upstream , in the centre - plane @xmath76 , the border between the pink and green sets of streamlines is located at @xmath77 and the upper border of the green set at @xmath78 .
interestingly , we did not isolate any streamwise tip vortices , whether in the iso - surfaces of @xmath60 or in kinematic
there are two probable reasons for this .
firstly , the spanwise flow confinement imposes a strong spanwise velocity gradient similar to that induced in thick boundary layers and @xcite showed that the thickening of the bottom boundary layer enhanced the base vortices at the expense of the tip ones .
secondly , the present cylinder aspect ratio is rather low and no tip vortices have been observed in simulations run with a low aspect ratio @xcite .
a third set of streamlines ( in cyan ) originating within the centre - plane @xmath76 just above the green set of streamlines at ( @xmath79 ) ends at a line of front stagnation points on the upstream cylinder face along the wake centreline . a half - saddle point @xmath80 , indicated in figure [ flowdyn](b ) ,
is detected along this line of stagnation points slightly below the cylinder tip .
blue and red streamlines rejoin the centre - plane @xmath76 downstream of the cylinder .
blue streamlines originate upstream , below those generating the horseshoe pattern [ see figure [ flowdyn](c ) ] but slightly off the wake centreline [ see figure [ flowdyn](d ) ] . on the one hand
, the streamlines located just upstream of @xmath70 head upwards and back upstream along the wake centre - plane behind the cylinder and separate into two substreams at half - saddle point @xmath81 on the downstream cylinder face [ see figure [ flowdyn](c ) ] .
the lower substream is deflected by this downstream face outwards the wake along the @xmath82-axis and eventually carried away by the free stream to form a pair of counter - rotating spanwise vortices .
the upper substream heads towards the cylinder tip where the free stream takes it away along the wake centreline . on the other hand , the blue streamlines located just downstream of @xmath70 head upwards and downstream
until they reach the free stream where they tilt along the @xmath17axis .
a gap between the blue streamlines moving upstream and downstream is observed at the rear of the cylinder slightly below the cylinder tip where a transverse vortex is seen [ red lines in figure [ flowdyn](c - d ) ] .
this head vortex consists of a pair of symmetric transverse vortices located at short distance off the wake centreline and connected to a single transverse vortex at the wake centreline .
only the right part of the head vortex , denoted hv , is visible on the @xmath60-iso - surfaces from figure [ flowdyn](a ) .
the streamlines enter this three - vortex structure at the periphery of the symmetric vortices , spiral towards the axis and exit at the periphery of the centre vortex onto the wake centreline .
note that the head vortex is not generated by streamlines curling from the cylinder free end as in @xcite , @xcite and @xcite , but from streamlines circulating upwards from the cylinder bottom half as in @xcite and @xcite . throughout this description of the flow topology at the bottom duct wall and cylinder
faces at @xmath51 , we have isolated three nodal points ( @xmath73 , @xmath74 , @xmath75 ) , two saddle points ( @xmath67 , @xmath70 ) and two half - saddle ones ( @xmath80 , @xmath81 ) .
this topology thus satisfies the mathematical criterion established by @xcite .
finally , two free shear layers arise from the lateral downstream edges of the cylinder and stretch on both sides of the wake ( see @xmath18vorticity contours in figure [ vortexstreetre200 ] ) . due to the presence of the bottom duct wall and the free shear layer stretching from the top trailing edge of the cylinder ,
their streamwise length is reduced at both their ends and reaches its maximum in the vicinity of the base vortices .
we insist that all the steady flow patterns detected in the present computations are generated exclusively by streamlines engulfed under these lateral free shear layers .
as @xmath1 is increased within the steady flow regime , the sets of streamlines rejoining the line of stagnation points at the cylinder upper face extends downwards , while the upstream spanwise extension of the set of streamlines generating the base vortices ( green lines ) shrinks and that feeding the horseshoe pattern ( pink lines ) broadens .
the increase of @xmath1 therefore implies an increase of @xmath83 and a shrinking of both @xmath84 and @xmath85 .
the overall shape of the horseshoe pattern however hardly changes : the main swirl hs [ see figure [ flowdyn](a ) ] drifts slightly upstream as observed also in @xcite @xcite and @xcite , but @xmath70 barely moves , _
i.e. _ @xmath86 increases and @xmath87 changes little .
the base vortices gain in strength and entangle with the spanwise vortices in the vicinity of the foci @xmath74 and @xmath75 , which get closer to the rear cylinder face as @xmath1 is increased . at @xmath37 ,
secondary recirculation regions appear on both lateral and top faces of the cylinder .
the lateral recirculations are made of stretched and twisted blue streamlines and the top one results from flow separation at the cylinder top face as in @xcite .
the top recirculation deflects the outer stream further up , which , combined with the increased strength of the head vortex , results in the latter shifting into a position higher than the cylinder tip in the late stages of the steady flow regime . in this regime
, the head vortex also moves farther downstream .
finally , at @xmath37 , the pair of symmetric vortices forming the head vortex merges into the centre - plane @xmath76 so that the streamlines are taken away downstream once they come out of the eye of this single remaining structure .
depicted by iso - surfaces of non - dimensional vorticity @xmath88 ( a - c ) and @xmath89 ( d - f ) at @xmath90 ( top ) , @xmath91 ( middle ) and @xmath92 ( bottom ) .
@xmath93 ( _ resp . _
@xmath94 ) in yellow ( _ resp . _
cyan ) ; @xmath95 in green ; @xmath96 ( _ resp . _
@xmath97 ) in red ( _ resp . _
the red lateral free shear layer is not shown in ( b ) .
tv , lv and rv are vortical structures shed from the top , left and right free shear layers , respectively .
br and bl are the base vortices .
sr and sl are the secondary streamwise vortices clearly visible on @xmath60-iso - surfaces but not on vorticity iso - surfaces where their locations are indicated by dashed arrows .
red and blue arrows on ( d ) indicate the jets induced by the base and hairpin vortices respectively , whose interaction participates to the generation of @xmath0-shaped vortices .
supplementary movie 1 ( available at journals.cambridge.org/flm ) shows an animation of the vortex street at @xmath38.,scaledwidth=90.0% ] at @xmath38 , unsteadiness appears through a periodic laminar symmetric shedding of hairpin vortices aligned on a single row along the wake centreline .
this vortex shedding has been observed in experiments by @xcite and @xcite and numerical simulations by @xcite and @xcite .
the formation and release of hairpin vortices depicted by iso - surfaces of vorticity field and of @xmath89 ( @xmath98 of its absolute minimum ) are represented at three successive time instants during the formation and release of the hairpin vortices at @xmath38 in figures [ vortexstreetre200](a - f ) . at @xmath90
, one observes a hairpin vortex in the early stage of its formation in the near cylinder wake and a freshly released , fully formed , hairpin vortex in the far wake .
three distinct vortical structures denoted tv , lv and rv , respectively originating from the top , left and right free shear layers , are singled out in figure [ vortexstreetre200](d ) .
tv indirectly results from the mutual interaction of the base vortices br and bl , shown in figure [ vortexstreetre200](a ) , that generates an upwards jet between them and lifts the tail of the top free shear layer .
this results in an inversion of the curvature of the latter which becomes unstably curved @xcite , and leads to the breakaway of its tail ( tv ) . at @xmath99 ,
two secondary counter - rotating streamwise vortices , denoted sr and sl in figures [ vortexstreetre200](b - f ) , then rise just upstream of the structure shed from the top free shear layer and wrap around the base vortices , while tv , lv and rv have gathered into a single bow of vorticity .
the subsequent pairing between both the base and the secondary streamwise vortices eventually triggers the symmetric shedding of structures lv and rv from both lateral free shear layers [ blue and red iso - surfaces in figure [ vortexstreetre200](c ) ] observed at @xmath99 . at this moment , the head of the hairpin vortex is completely formed and taken away by the free stream , while the base vortices detach to join the head of the hairpin vortex and form its legs visible on the hairpin vortex present in the far wake in figure [ vortexstreetre200](d ) .
our computations show that the hairpin vortices are not generated only by the destabilization of the top free shear layer as suggested in @xcite and @xcite . instead
, they result from the smooth assembling of structures shed from the initially steady flow patterns .
the head of the hairpin vortex is formed from the smooth assembling of vortical structures detached from the top and lateral free shear layers .
the hairpin vortices observed in the present simulations differ from the arch - type vortices detected in experiments by @xcite and @xcite .
firstly , due to the spanwise velocity gradient imposed by our flow confinement , the hairpin head is located at the downstream end of the structure .
the free stream velocity is higher in the vicinity of the cylinder tip so that the hairpin head , located in this region , is carried away further downstream than the rest of the hairpin structure .
secondly , the hairpin legs are almost parallel to the streamwise axis and not aligned along the spanwise direction as those of the arch - type vortices . in our configuration ,
the hairpin legs are indeed the base vortices which are oriented along the streamwise axis , whereas those of the arch - type vortices are the spanwise vortices shed from the lateral free shear layers .
thirdly , in our configuration , neither the legs nor the spanwise vortical part of the hairpin vortex grow _ a posteriori _ from the structure shed from the top free shear layer , as claimed in @xcite . instead , both these structures are present in the very first steps of the formation of the hairpin vortices .
this shows that the base vortices play an active part in the vortex shedding mechanism as in the flow past a trapezoidal tab @xcite , but unlike in @xcite and @xcite . in the last two studies
though , the inlet velocity is almost spanwise invariant and the shedding results from the interaction between the top and lateral free shear layers only .
we have described how the hairpin vortices are generated and released in the cylinder wake . in the unsteady regime ,
the structures that take part in their formation were detected by plotting iso - surfaces of vorticity and @xmath60 for reasonably high values of these quantities .
we shall now refine our analysis and track weaker structures by bringing the threshold values of @xmath60 and of the vorticity closer to zero for the flow at @xmath38 . at @xmath38 and @xmath100 .
fields of non - dimensional ( b ) @xmath101-vorticity in plane a ( @xmath102 ) with @xmath103 , ( c ) @xmath18vorticity in plane b ( @xmath104 ) with @xmath105 and ( d ) @xmath19vorticity in plane c ( @xmath69 ) with @xmath106 ( minimum and maximum values in blue and red , respectively ) .
thick white lines on ( b - d ) are the intersected @xmath107iso - surfaces .
@xmath108 and @xmath109 are two successive hairpin vortices , @xmath110 , @xmath111 and @xmath112 three @xmath0-shaped vortices , sr and sl the respective left and right secondary streamwise vortices and br and bl the respective left and right base vortices ( see also figure [ vortexstreetre200]).,scaledwidth=95.0% ] figure [ omegavortex](a ) shows the three - dimensional vortex street at @xmath113 depicted by iso - surfaces of @xmath114 and two - dimensional vorticity iso - surfaces in three different planes are presented in figures [ omegavortex](b - d ) .
one observes a chain of @xmath0-shaped vortices between two successive hairpin vortices , denoted @xmath108 and @xmath109 .
three such @xmath0-shaped vortices , denoted @xmath110 , @xmath111 and @xmath112 , are detected in figure [ omegavortex](a ) .
the strength of these vortices decreases with their distance to the downstream hairpin vortex .
they are located in @xmath115-planes and their bottom ends are connected to the secondary streamwise vortices , denoted sr and sl in figure [ omegavortex](d ) .
the @xmath0-shaped vortices are mirror images of the hairpin vortex as they all rotate in the sense opposite to it , as shown by figures [ omegavortex](b - d ) .
vortex streets involving hairpin vortices and secondary reverse vortices ( _ i.e. _ rotating in the opposite sense to that of the hairpin head ) have been observed in the flow past a trapezoidal tab in experiments @xcite and numerical simulations @xcite .
reverse vortices detected in the latter configuration were however located below the line of hairpin vortices by contrast with our observations where hairpin and reverse vortices are all aligned .
the simulations by @xcite have yet clearly established the @xmath0-shape of the reverse vortices as in our case .
the formation of @xmath0-shaped vortices follows from a mechanism encountered in turbulent boundary layers in channel flows described in e.g. @xcite . during its downstream motion ,
the head of the hairpin vortex broadens and induces a backwards streamwise jet through the head [ blue arrow in figure [ vortexstreetre200](d ) ] .
the pair of streamwise vortices forming the legs of the hairpin vortex generates an upwards jet between them [ red arrow in figure [ vortexstreetre200](d ) ] .
the shearing between the latter jet and the one induced by the head of the hairpin vortex causes the formation of a bridging shear layer between both streamwise vortices .
this shear layer subsequently rolls up into an @xmath0-shaped vortex in the ( @xmath116)-plane .
this mechanism is active all along the legs of the hairpin vortex and leads to the formation of several @xmath0-shaped vortices between two successive hairpin vortices .
importantly , the generation of @xmath0-shaped vortices is a direct consequence of the streamwise orientation of the hairpin legs .
this mechanism is thus absent in configurations where arch - type vortices have been observed @xcite . and @xmath117 ( a ) and @xmath118 and @xmath119 ( b ) .
( _ left _ ) iso - surfaces of @xmath120 -0.5 ( a ) and -0.7 ( b ) and ( _ right _ ) fields of @xmath121 , @xmath122 and @xmath123 ( a ) and @xmath124 , @xmath125 and @xmath126 ( b ) in yellow and cyan , green , red and blue , respectively.,scaledwidth=85.0% ] at @xmath127 , the mechanism of formation and release of hairpin vortices described previously remains unchanged , but the shape of the hairpin vortices becomes irregular and weak spanwise oscillations of the base vortices are observed .
the latter oscillations disturb the shedding of the top and lateral free shear layers so that the vortex street turns slightly asymmetric as shown in figure [ irregularvortexstreetha0](a ) . for @xmath128 ,
these oscillations gain in intensity .
the legs of the hairpin vortices do not split at the same @xmath18coordinate anymore , while their head turns into a chaotic aggregate of structures shed from the free shear layers .
hairpin vortices are therefore released alternately on either side of the wake centreline and so a fully asymmetric vortex street is observed as seen in figure [ irregularvortexstreetha0 ] ( b ) .
the chain of @xmath0-shaped vortices is still observed in this regime , but their shape is irregular too and their orientation follows that of the hairpin vortex that creates them .
a similar evolution of the vortex street with increasing @xmath1 has been described by @xcite in experiments of the flow past a square plate . the gradual change from a regular to a chaotic vortex street for @xmath129
is illustrated in figure [ cdtimehistory ] by the respective time histories of the total drag coefficient @xmath49 , introduced in section [ sectionflowequationsnumericalsetup ] . at @xmath1=200 ,
250 , 300 and 400.,scaledwidth=65.0% ]
the variations of the total drag coefficient @xmath49 with @xmath1 are presented in figure [ coefre](a ) .
one observes that @xmath49 decreases within the steady flow regime and slightly increases in the unsteady one .
these variations reflect the dynamics of the vortex formation region as in the two - dimensional square cylinder wake @xcite . in steady regimes
, it encompasses the pair of spanwise recirculation regions , while in unsteady ones it includes the formation region of the hairpin vortices . in both cases
, this region induces an adverse pressure gradient .
the increase of @xmath1 within the steady regime results in the streamwise elongation of the symmetric steady recirculation regions .
this enhances the adverse pressure gradient which gradually counter - balances the streamwise pressure gradient induced by the duct flow and thus leads to the decrease in the total drag coefficient @xmath49 . in the unsteady flow regime ,
the mechanism is reversed as the formation region of the hairpin vortices shrinks , hence the increase in @xmath49 .
remarkably , the present values of @xmath49 obtained with a cylinder spanning over half the duct height are about half of those obtained with a non - truncated cylinder @xcite . the viscous component of the drag coefficient is defined as @xmath130 where @xmath131 is the viscous force acting on the cylinder .
we have reported its variations over the cylinder span at different @xmath1 in figure [ coefre](b ) .
one can identify three regions along the span : two narrow ones close to the respective cylinder ends and a larger one between them @xcite .
@xmath132 exhibits peak values just where the flow around the cylinder is fast , _
i.e. _ below the cylinder free end and at short distance of the bottom duct wall within the horseshoe region .
in contrast , @xmath132 remains rather constant in the mid - span region where the flow recirculates much more slowly .
the corresponding values are therefore significantly smaller than the peak values reached in both end regions .
one observes also that @xmath132 turns negative for ( @xmath133 ) at @xmath38 and that the @xmath18range with negative @xmath132 widens as @xmath1 is increased .
this feature results from the appearance and subsequent widening of the secondary recirculation regions at the cylinder lateral faces for @xmath134 .
the associated counter - flow induces advert friction at the side walls that reduces the overall value of @xmath132 to the point of making it negative .
this effect was detected in two - dimensional square cylinder wakes by @xcite .
we define the spanwise lift coefficient @xmath135 where @xmath136 is the @xmath18component of the force exerted by the flow on the cylinder .
the evolution of @xmath137 computed only on the upper cylinder face versus @xmath1 is reported in figure [ coefre](c ) . by definition , @xmath136 results from the integration of the pressure force on the cylinder upper face and is defined up to a constant determined by the value of the reference pressure . as a consequence ,
only the variations of @xmath137 reflect the flow dynamics , while its absolute values shall be regarded relatively to the reference pressure . in particular , the sign of @xmath137 bears no significance .
@xmath137 decreases in the steady regime and slightly increases in the unsteady one . as for the drag coefficient , the decrease of @xmath137 results from the lengthening of the spanwise recirculation regions at the back of the cylinder
the reason for its increase is the appearance of the secondary recirculation over the cylinder upper face that induces an adverse pressure gradient .
the destabilization of the top free shear layer in the unsteady regime for @xmath64 slightly enhances this trend throughout the unsteady regime .
the frequency of the symmetric ( _ resp .
_ asymmetric ) mode was obtained from the time history of the total drag coefficient @xmath49 ( _ resp .
_ lift coefficient @xmath138 where @xmath139 is the @xmath19component of the force exerted by the flow on the cylinder ) .
we were then able to calculate the strouhal numbers associated to the respective frequencies of the symmetric and asymmetric modes .
their dependence on @xmath1 is presented in figure [ coefre](d ) . at @xmath38 ,
the vortex street is indeed perfectly symmetric so the asymmetric mode is absent . at @xmath127
, a slight asymmetry is observed in the wake due to the appearance of the small vertical oscillations of the base vortices .
these oscillations are enhanced when @xmath1 is further increased until the wake becomes chaotic and completely asymmetric .
when the wake is symmetric at @xmath38 , only the symmetric mode is present with an associated @xmath140 .
when the asymmetric mode appears at @xmath127 , its associated @xmath48 is much lower ( @xmath141 ) but increases with @xmath1 .
as the wake becomes more asymmetric , the strouhal number associated to the symmetric mode collapses down to values below those of the asymmetric one . on the cylinder stem versus @xmath1 and ( b ) spanwise distribution of the viscous drag coefficient @xmath132 at different @xmath1 ( @xmath142 is the @xmath18average value and the time - average value of @xmath132 is used in the unsteady flow regime ) .
( c ) spanwise lift coefficient @xmath137 on the cylinder upper face and ( d ) strouhal numbers @xmath48 associated to the symmetric and asymmetric modes versus @xmath1 . the dashed line on ( a ) and ( c ) separates the steady and the unsteady flow regimes.,scaledwidth=99.0% ]
we have investigated the flow past a truncated square cylinder in a duct of rectangular cross - section using three - dimensional direct numerical simulations for @xmath143 .
figure [ summarysimulations ] summarises the main flow regimes we have encountered .
the steady regime has been thoroughly scrutinised and the flow patterns identified .
the present computations have shown that these patterns were all generated by sets of streamlines flowing around the cylinder foot and underneath the lateral free shear layers .
in particular , in our configuration , the spanwise flow confinement induces an important spanwise shear that promotes the development of the base vortices at the expense of the tip ones . also , we have singled out the appearance of secondary recirculation regions at the lateral and upper cylinder faces for @xmath134 . at @xmath38
, the flow unsteadiness leads to a symmetric vortex street formed by a single row of hairpin vortices aligned on the wake centreline .
we have detailed a formation mechanism of the hairpin vortices resulting from the smooth assembling of structures shed from the initially steady flow patterns .
this scenario has never been exposed so far and differs from those given in previous works @xcite . to partly explain this difference , we have stressed the effects of the vertical flow confinement which enhances the base vortices and induces a more efficient entrainment of the head of the hairpin vortices .
the latter is therefore located at the downstream end of the hairpin and the legs are almost parallel to the streamwise axis .
also , we have detected the formation of a chain of @xmath0-shaped vortices as a consequence of the streamwise orientation of legs of the hairpin vortices .
the @xmath0-shaped vortices are well known in turbulent boundary layers and their formation stems from the same mechanism as in this classic case . for @xmath144 ,
the appearance and enhancement of vertical oscillations of the base vortices causes the vortex street to turn asymmetric and irregular , while hairpin vortices become chaotic aggregates of structures .
we have reported the evolutions of a set of flow coefficients with @xmath1 .
firstly , the spanwise lift coefficient has been found to increase when the recirculation appeared on the top face .
secondly , the transition to asymmetry in the unsteady wake has been shown to induce a collapse in the strouhal number that reflected the disappearance of the symmetric mode associated with the hairpin vortex shedding we have identified . finally , although we have clearly identified the dynamics of the hairpin vortices , questions about the respective influences of the various parameters set in our numerical computations are still open .
firstly , the influence of the cylinder aspect ratio @xmath7 would require a parametric investigation .
although it is well accepted that @xmath7 determines whether the onset of vortex shedding gives rise to hairpin or krmn vortices , the effect of @xmath7 within any of these regimes is yet to be clarified . secondly , we have used physical walls at the outer boundaries of our domain .
though physically realistic , both the transverse and spanwise confinements should be subject to a thorough scrutiny to determine their effects on the stability and the shape of the hairpin vortices .
the transverse confinement in the present configuration is low enough to neglect its effects .
one may however expect valuable information on whether moderate transverse blockage improves the stability of the hairpin vortices .
by contrast , we have shown that the spanwise blockage had a significant influence on the flow dynamics .
one could expect a lighter spanwise confinement to minimise the role of the base vortices in these dynamics and result in a modification of the overall shape of the hairpin vortices which could be closer to the arch - type vortices observed in experiments by @xcite and @xcite . | we investigate the laminar shedding of hairpin vortices in the wake of a truncated square cylinder placed in a duct , for reynolds numbers around the critical threshold of the onset of vortex shedding .
we single out the formation mechanism of the hairpin vortices by means of a detailed analysis of the flow patterns in the steady regime .
we show that unlike in previous studies of similar structures , the dynamics of the hairpin vortices is entwined with that of the counter - rotating pair of streamwise vortices , which we found to be generated in the bottom part of the near wake ( these are usually referred to as _ base vortices _ ) . in particular ,
once the hairpin structure is released , the base vortices attach to it , forming its legs , so these are streamwise , and not spanwise as previously observed in unconfined wakes or behind cylinders of lower aspect ratios .
we also single out a trail of @xmath0-shaped vortices , generated between successive hairpin vortices through a mechanism that is analogous to that active in near - wall turbulence .
finally , we show how the dynamics of the structures we identified determine the evolution of the drag coefficients and strouhal numbers when the reynolds number varies . |
[ introduction ] _ combinatorial physics _ is a growing field , defining itself as the interdisciplinary domain overlapping between combinatorics and physics .
one can clearly identify nowadays several aspects of this larger and larger overlapping .
thus , the interference between combinatorics and quantum mechanics has been investigated in @xcite , @xcite ( and references within ) .
an important interplay area can be noticed between combinatorics ( bijective , enumerative and so on ) on one hand and statistical physics and integrable combinatorics on another hand ( see , for example , @xcite for a recent review on this topic ) .
our paper deals with yet another aspect of this interdisciplinary field of research , namely the interference between combinatorics and quantum field theory ( commutative or non - commutative ) .
this type of interference has already appeared in the pioneering work of alain connes and dirk kreimer , who have defined a hopf algebra encoding in an elegant way the combinatorics of the process of perturbative renormalization in commutative quantum field theory .
this structure has then been generalized for noncommutative quantum field theory @xcite , @xcite and for spin - foam quantum gravity models @xcite , @xcite . for general reviews of various interferences between algebraic and analytic ( and not only ) combinatorics and quantum field theory in general
, we invite the interested reader to consult @xcite and @xcite .
in this paper we consider appropriate characters of the the hopf algebra of isomorphic classes of matroids , algebra defined ( as a particularization of a more general construction of incidence hopf algebra ) in @xcite ( and then extensively studied in @xcite ) .
we also show that these characters are related to the tutte polynomial for matroids and we then use a quantum field theory renormalization group - like differential equation to prove the universality of the tutte polynomial for matroids .
more precisely , we show that a solution of such an equation is given by the characters that we have defined . as a by - product of our hopf algebraic approach , we give a new proof of a convolution formula for the tutte polynomial for matroids , formula exhibited in @xcite .
the paper is structured as follows . in the following section we introduce the renormalization group equation in quantum field theory and we then recall some useful notions related to matroids and to the tutte polynomial for matroids .
finally , we give the definition of the hopf algebra of isomorphic classes of matroids . in the third section
we define the hopf algebra characters that will be used in the sequel . in the following section
, we use this construction to give our new proof of the convolution formula for the tutte polynomial for matroids mentioned above .
the fifth section is dedicated to our main result , the proof of the universality of the tutte polynomial for matroids .
the last section presents some concluding remarks and perspectives for future work .
in this section we first briefly recall some quantum field theory notions , namely the renormalization group differential equation .
we then give the definition of matroids , of the associated tutte polynomial as well as some further properties which will be useful to prove the results of this paper .
finally , the hopf algebra of isomorphic classes of matroids is given . a qft model ( for a general introduction to qft see for example the books @xcite or @xcite ) is defined by means of a functional integral of the exponential of an _ action _ @xmath0
which , from a mathematical point of view , is a functional of the _ fields _ of the model . for the @xmath1 scalar model - the simplest qft model -
there is only one type of field , which we denote by @xmath2 . from a mathematical point of view , for an euclidean qft scalar model , one has @xmath3 , where @xmath4 is usually taken equal to @xmath5 ( the dimension of the space ) and @xmath6 ( real , respectively complex fields ) .
the quantities computed in qft are generally divergent .
one thus has to consider a real , positive , _ cut - off _ @xmath7 - the flowing parameter .
this leads to a family of cut - off dependent actions , family denoted by @xmath8 .
the derivation @xmath9 gives the _ renormalization group equation_. the quadratic part of the action - the _ propagator _ of the model - can be written in the following way [ propa ] c_,_0(p , q)=(p - q)_^ de^-p^2 , with @xmath10 and @xmath11 living in the fourier transformed space @xmath12 and @xmath13 a second real , positive cut - off .
in perturbative qft , one has to consider _ feynman graphs _ , and to associate to each such a graph a _
feynman integral _
( further related to quantities actually measured in physical experiments ) .
the contribution of an edge of such a feynman graph to its associated feynman integral is given by an integral such as .
one can then get ( see @xcite ) the polchinski flow equation [ eq : pol ] = _ ^2d 12 d^d p d^d q ( - ) , where @xmath14 represents the fourier transform of the function @xmath15 .
the first term in the right hand side ( rhs ) of the equation above corresponds to the derivation of a propagator associated to a bridge in the respective feynman graph .
the second term corresponds to an edge which is not a bridge and is part of some circuit in the graph .
one can see this diagrammatically in fig .
[ fig : pol ] .
this equation can then be used to prove perturbative renormalizability in qft .
let us also stress here , that an equation of this type is also used to prove a result of e. m. wright which expresses the generating function of connected graphs under certain conditions ( fixed excess ) . to get this generating functional ( see , for example , proposition @xmath16 the book @xcite ) , one needs to consider contributions of two types of edges ( first contribution when the edge is a bridge and a second one when not - see again fig .
[ fig : pol ] ) . in this subsection
we recall the definition and some properties of the tutte polynomial for matroids as well as of the matroid hopf algebra defined in @xcite . following the book @xcite
, one has the following definitions : a * matroid * m is a pair @xmath17 consisting of a finite set e and a collection of subsets of e satisfying the following set of axioms : @xmath18 is non - empty , every subset of every member of @xmath18 is also in @xmath18 and , finally , if @xmath19 and @xmath20 are in @xmath18 and @xmath21 , then there is an element @xmath22 in @xmath23 such that @xmath24 is in @xmath18 .
the set @xmath25 is the ground set of the matroid and the members of @xmath18 are the independent sets of the matroid
. one can define a matroid on the edge set of any graph - * graphic matroid*. the reciproque does not hold ( not every matroid is a graphic matroid ) .
let @xmath25 be an @xmath26element set and let @xmath27 be the collection of subsets of @xmath25 with at most @xmath28 elements , @xmath29 .
one can check that @xmath30 is a matroid ; it is called the * uniform matroid * @xmath31 .
[ remarca ] if one takes @xmath32 , there are only two matroids , namely @xmath33 and @xmath34 and both of these matroids are graphic matroids .
the graphs these two matroids correspond to are the graphs with one edge of fig .
[ graf1 ] and fig .
[ graf2 ] . .
] . ] in the first case , the edge is a loop ( in graph theoretical terminology ) or a tadpole ( in qft language ) . in the second case , the edge represents a bridge ( in graph theoretical terminology ) or a @xmath35-particle - reducible line ( in qft terminology ) - the number of connected components of the graphs increases by @xmath35 if one deletes the respective edge .
maximal independent sets of a matroid are called bases . the collection of minimal dependent sets of a matroid are called circuits .
let @xmath36 be a matroid and let @xmath37 be the collection of bases of @xmath38 .
let @xmath39 .
then @xmath40 is the collection of bases of a matroid @xmath41 on e. the matroid @xmath42 is called the dual of @xmath38 . [ def - rank ] let @xmath36 be a matroid .
the * rank * @xmath43 of @xmath44 is defined as the cardinal of a maximal independent set in @xmath45 .
@xmath46 [ def - nullity ] let @xmath47 be a matroid with a ground set @xmath25 .
the * nullity * function is given by @xmath48 [ def - loop ] let @xmath36 be a matroid .
the element @xmath49 .
is a * loop * iff @xmath50 is the circuit .
[ def - coloop ] let @xmath36 be a matroid .
the element @xmath51 is a * coloop * iff , for any basis @xmath52 , @xmath53 .
let us now define two basic operations on matroids .
let @xmath38 be a matroid @xmath54 ) and @xmath55 be a subset of @xmath25 .
let @xmath56 .
one can check that @xmath57 is a matroid .
we denote this matroid by @xmath58 - the * deletion * of @xmath55 from @xmath38 .
the * contraction * of @xmath55 from @xmath38 , @xmath59 , is given by the formula : @xmath60 .
let us also recall the following results : [ lm : res - del ] let @xmath38 be a matroid @xmath54 ) and @xmath55 be a subset of @xmath25 .
one has @xmath61 [ lm : coloop ] if @xmath62 is either a coloop or a loop of a matroid @xmath36 , then @xmath63 .
[ lm : rankcontra ] let @xmath36 be a matroid and @xmath64 , then , for all @xmath65 , @xmath66 let us now define the tutte polynomial for matroids : let @xmath47 be a matroid .
the * tutte polynomial * is given by the following formula : @xmath67 the sum is computed over all subset of the matroid s ground set .
let @xmath68 be a uniform matroid , @xmath69 .
the tutte polynomial of this matroid is given by t_u_k , n(x , y ) = _ i=0^k ( x-1)^k - i + _ i = k+1^n ( y-1)^i - k .
it is worth stressing here that one can define the dual of any matroid ; this is not the case for graphs , where only the dual of planar graph can be defined .
let us recall , from @xcite that [ reltdual ] t_m ( x , y)=t_m^(y , x ) .
let @xmath70 and @xmath71 be the matroids @xmath72 and @xmath73 where @xmath74 and @xmath75 are disjoint .
let @xmath76then @xmath77 is a matroid .
this matroid is called the direct sum of @xmath70 and @xmath71 . in @xcite , as a particularization of a more general construction of incidence hopf algebras , the following result was proved : [ prop - cs ] if @xmath78 is a minor - closed family of matroids ( and if we denote by @xmath79 the set of isomorphism classes of it ) , then @xmath80 is a coalgebra , with coproduct @xmath81 and counit @xmath82 determined by [ defc ] ( m ) = _ ae m|a m / a and respectively by @xmath83 for all @xmath84 . if , furthermore , the family @xmath78 is closed under formation of direct sums , then @xmath80 is a hopf algebra , with product induced by direct sum .
we refer to this hopf algebra as to the matroid hopf algebra .
we follow @xcite and , by a slight abuse of notation , we denote in the same way a matroid and its isomorphic class , since the distinction will be clear from the context ( as it is already in proposition [ prop - cs ] ) .
we denote the unit of this hopf algebra by @xmath85 ( the empty matroid , or @xmath86 ) .
( example 2.4 of @xcite ) let @xmath87 be a uniform matroid with rank @xmath88 .
its coproduct is given by @xmath89
let us give the following definitions : let @xmath90 be two mappings in @xmath91 .
the convolution product of @xmath92 and @xmath93 is given by the following formula : @xmath94 a matroid hopf algebra * character * @xmath92 is an algebra morphism from the matroid hopf algebra into a fixed commutative ring @xmath95 , i.e. such that f(m_1m_2 ) = f(m_1)f(m_2 ) , f ( ) = 1_. a matroid hopf algebra * infinitesimal character * @xmath93 is a linear morphism from the matroid hopf algebra into a fixed commutative ring @xmath96 , such that g(m_1m_2 ) = g(m_1)(m_2)+(m_1)g(m_2 ) .
since we work in a hopf algebra where the non - trivial part of the coproduct is nilpotent , we can also define an exponential map by the following expression [ def - expstar ] _ ( ) = + + 12 + where @xmath97 is an infinitesimal character .
as already stated above ( see remark [ remarca ] ) , there are only two matroids with unit cardinal ground set , @xmath33 and @xmath34 .
we now define two applications @xmath98 and @xmath99 .
[ def - dloop ] _ ( m ) = 1 _ m = u_0,1 , + 0_. [ def - dtree ] _ ( m ) = 1 _ m = u_1,1 , + 0_. one can directly check that these applications are _ infinitesimal characters _ of the matroid hopf algebra defined above .
we now define the following application : [ def - alpha ] ( x , y , s , m ) : = _ s\ { _ + ( y-1)_}_s\{(x-1)_+ _ } ( m ) .
let @xmath68 be a uniform matroid , @xmath69 .
one has ( x , y , s , u_k , n ) = _ i=0^k s^n(x-1)^k - i + _ i = k+1^n
s^n(y-1)^i - k = s^nt_u_k , n(x , y ) .
one then has : the application is a character .
the proof can be done by a direct check . on a more general basis , this is a consequence of the fact that @xmath98 and @xmath99 are infinitesimal characters and the space of infinitesimal characters is a vector space ; thus @xmath100 and @xmath101 are infinitesimal characters . since , @xmath102 is a character when @xmath103 is an infinitesimal character and since the convolution of two characters is a character , one gets that @xmath104 is a character .
let @xmath47 be a matroid .
one then has : [ lema - help ] let @xmath47 be a matroid .
one has @xmath105 using the definition , the lhs of the identity above reads : @xmath106 all the terms in the sum above vanish , except the one for whom @xmath88 is equal to @xmath107 . using the definition of the convolution product ,
this terms writes ( _ i=0^k a^k - ib^i _ _ ^(i_1)_^(j_1 ) _
^(i_n ) _ ^(j_m ) )
( _ ( i ) m^(1)
m^(k ) ) , where we have used the notation @xmath108 . using the definitions and respectively of the infinitesimal characters @xmath98 and respectively @xmath99
, implies that the submatroids @xmath109 ( @xmath110 ) are equal to @xmath33 or @xmath34 .
using the definition of the nullity and of the rank of a matroid concludes the proof .
let us illustrate lemma [ lema - help ] for the uniform matroid @xmath68 .
one has @xmath111 and @xmath112 .
we now use the definitions [ def - dloop ] and [ def - dtree ] of @xmath113 and @xmath114 to work out the lhs of identity [ help ] .
@xmath115 one has : [ pre - reiner ] ( x , y , s , m ) & = & _ ( s(_+(y-1)_))_(s(-_+ _ ) ) + & & _ ( s(_-_))_(s((x-1)_+ _ ) ) .
[ alpha - tutte ] let @xmath47 be a matroid .
the character @xmath104 is related to the tutte polynomial of matroids by the following identity : @xmath116 using the definition of the convolution product in the definition of the character @xmath104 , one has the following identity : @xmath117 we can now apply lemma [ lema - help ] on each of the two terms in the rhs of equation above .
this leads to the result . using and the proposition [ alpha - tutte ]
, one has : one has : ( x , y , s , m ) = ( y , x , s , m^ ) .
the proposition [ alpha - tutte ] allows to give a different proof of a matroid tutte polynomial convolution identity , which was shown in @xcite .
one has : ( theorem @xmath35 of @xcite ) the tutte polynomial satisfies @xmath118 taking @xmath119 , this is as a direct consequence of identity , and of proposition [ alpha - tutte ] .
let us define an application @xmath120 [ lemamor ] the application @xmath121 is a bialgebra automorphism .
one can directly check that the application @xmath121 is an algebra automorphism .
let us now check that this application is also a coalgebra automorphism .
using lemma [ lm : res - del ] and lemma [ lm : rankcontra ] , [ ranks ] r(m|_t ) + r(m / t ) = r(m ) .
thus , using the definitions of the application @xmath121 of the matroid coproduct , one has : @xmath122 using again the definition of the application @xmath121 leads to @xmath123 which concludes the proof .
let us now define : _ : = fg - gf . using the definition of the hopf algebra character @xmath104
, one can directly prove the following result : [ propalpha ] the character @xmath104 is the solution of the differential equation : @xmath124_\ast - \left[\delta_{\mathrm{loop}},\alpha\right]_\ast(m).\ ] ] it is the fact that the matroid tutte polynomial is a solution of the differential equation that will be used now to prove the universality of the matroid tutte polynomial . in order to do that , we take a four - variable matroid polynomial @xmath125 satisfying a multiplicative law and which has the following properties : * if @xmath62 is a coloop , then @xmath126 * if @xmath62 is a loop , then @xmath127 * if @xmath62 is a nonseparating point , then q_m(x , y , a , b ) = aq_m\e(x , y , a , b)+bq_m / e(x , y , a , b ) . note that , when one deals with the same problem in the case of graphs , a supplementary multiplicative condition for the case of one - point joint of two graphs ( _ i . e. _ identifying a vertex of the first graph and a vertex of the second graph into a single vertex of the resulting graph ) is required ( see , for example , @xcite or @xcite ) .
we now define the application : [ defbeta ] ( x , y , a , b , s , m):=s^|e|q_m(x , y , a , b ) .
one then directly checks ( using the definition above and the multiplicative property of the polynomial @xmath128 ) that this application is again a matroid hopf algebra character .
[ propbeta ] the character satisfies the following differential equation : @xmath129_\ast - a [ \delta_{\mathrm{loop}},\beta]_\ast\right ) ( m).\ ] ] applying the definition of the convolution product , the rhs of equation above writes @xmath130 using the definitions and respectively of the infinitesimal characters @xmath98 and respectively @xmath99 , constraints the sums on the subsets @xmath45 above .
the rhs of becomes : & ( x - b)_a , m/_a = u_1,1 ( m|_a ) + ( y - a)_a , m|_a= u_0,1 ( m/_a ) + & + b _ a , m|_a = u_1,1 ( m/_a ) + a _ a , m/_a = u_0,1(m|_a ) we now apply the definition of the hopf algebra character @xmath131 ; one obtains : [ int1 ] & s^|e|-1[(x - b)_a , m/_a = u_1,1 q(x , y , a , b , m|_a ) + ( y - a)_a , m|_a= u_0,1 q(x , y , a , b , m/_a ) + & + b _ a , m|_a = u_1,1 q(x , y , a , b , m/_a ) + a _ a , m/_a = u_0,1q(x , y , a , b , m|_a ) ] . we can now directly analyze the four particular cases @xmath132 , @xmath133 , @xmath134 and @xmath135 : * if @xmath132 , we can denote the ground set of @xmath136 by @xmath137 . note that @xmath62 is a coloop . from the lemma [ lm : res - del ] , one has @xmath138 .
one then has @xmath139 . * if @xmath140 , then @xmath141 and @xmath62 is a loop of @xmath38 .
thus , one has @xmath142 * if @xmath143 , then @xmath141 .
one has to distinguish between two subcases : * * @xmath62 is a coloop of @xmath38 .
then , by lemma [ lm : coloop ] , @xmath63 .
thus , one has @xmath139 .
* * @xmath62 is a nonseparating point of @xmath38 . * if @xmath144 , one can denote the ground set of @xmath136 by @xmath145 .
there are again two subcases to be considered : * * @xmath62 is a loop of @xmath38 , one has that @xmath146 . then one has @xmath147 . *
* @xmath62 is a nonseparating point of @xmath38 , then one has @xmath148 we now insert all of this in equation ; this leads to three types of sums over some element @xmath62 of the ground set @xmath25 , @xmath62 being a loop , a coloop or a nonseparating point : @xmath149 \notag\\\end{aligned}\ ] ] this rewrites as latexmath:[\[\begin{aligned } completes the proof .
we can now state the _ main result _ of this paper , the recipe theorem specifying how to recover the matroid polynomial @xmath128 as an evaluation of the tutte polynomial @xmath151 : one has : @xmath152 the proof is a direct consequence of propositions [ alpha - tutte ] , [ propalpha ] and [ propbeta ] and of lemma [ lemamor ] .
this comes from the fact that one can apply the automorphism @xmath153 defined in to the differential equation .
one then obtains the differential equation with modified parameters @xmath154 and @xmath155 .
finally , the solution of this differential equation is ( trivially ) related to the matroid tutte polynomial @xmath151 ( see proposition [ alpha - tutte ] ) and this concludes the proof .
let us end this section by stating that all the results obtained in this paper naturally hold for graphs ( instead of matroids ) , since graphs are a particular class of matroids ( the graphic matroids , see subsection @xmath156 ) .
we have thus given here the proofs of the graph results conjectured in @xcite .
we have used in this paper a quantum field theory renormalization group - like equation to prove the universality of the tutte polynomial for matroids . moreover
, we gave a new proof of the convolution identity established in @xcite by w. kook _ et .
al_. let us emphasize that the hopf algebra coproduct used in this paper is a so - called type @xmath157 coproduct , namely a coproduct using a _ selection - quotient _ rule .
examples of such coproducts are the connes - kreimer coproduct for commutative quantum field theory feynman graphs @xcite , its generalization to non - commutative quantum field theory @xcite and so on . as it was already noticed in @xcite or @xcite ,
this type of rule is fundamentally different of the one used to define a so - called type @xmath158 coproduct , namely a _
selection - deletion _ rule .
this latter rule has been extensively used in algebraic combinatorics ( see , for example , @xcite and references within ) .
let us also notice that for these type @xmath158 coproducts , explicit polynomial realizations have been recently obtained ( see @xcite and references within ) .
these polynomial realizations are particularly interesting in order to give new , straightforward , proofs of the coassociativity of the respective coproducts ( see , for example , j .- y .
talk @xcite ) .
it thus appears to us as a particularly interesting perspective for future work the investigation of the existence of connections between these type @xmath157 and @xmath158 combinatorial hopf algebra coproducts .
such connections could eventually be obtained by exhibiting explicit polynomial realizations for the type @xmath157 coproduct combinatorial hopf algebras , thus completing the picture of combinatorial hopf algebra polynomial realizations .
adrian tanasa , `` some combinatorial aspects of quantum field theory , '' sminaire lotharingien de combinatoire , b65 g ( 2012 ) [ arxiv:1102.4231 [ math.co ] ] .
jean - yves thibon . .
talk at the philippe flajolet seminar , 27th of january 2012 , available online at http://www-igm.univ-mlv.fr/ jyt / talks / flaj.pdf , 2012 .
@xmath159_centre de physique thorique , campus de luminy , _
+ + @xmath160_laboratoire de physique thorique , btiment 210 , universit paris 11 , _ + + @xmath161 _ horia hulubei national institute for physics and nuclear engineering , + p.o.b .
mg-6 , 077125 magurele , romania , eu _
+ @xmath162_universit paris 13 , sorbonne paris cit , 99 , avenue jean - baptiste clment + lipn , institut galile , cnrs umr 7030 , f-93430 , villetaneuse , france , eu _ + | using a quantum field theory renormalization group - like differential equation , we give a new proof of the recipe theorem for the tutte polynomial for matroids .
the solution of such an equation is in fact given by some appropriate characters of the hopf algebra of isomorphic classes of matroids , characters which are then related to the tutte polynomial for matroids .
this hopf algebraic approach also allows to prove , in a new way , a matroid tutte polynomial convolution formula appearing in w. kook _ et .
al . , j. comb .
series _
* b 76 * ( 1999 ) . -1.5
cm keywords : tutte polynomial for matroids , quantum field theory renormalization - group , combinatorial hopf algebras , hopf algebras characters . |
quantum spin models are indispensable tools for describing the complex behavior of quantum condensed matter systems .
they are a universal language for characterizing quantum magnetism and the behavior of quantum systems near phase transitions @xcite , and they can potentially shed light on the physics that underlie exotic new materials @xcite or high-@xmath0 superconductivity @xcite . yet , most spin models have not been solved , " meaning that it is not possible to write down analytic expressions describing the locations of critical points , the character of different phases , or how an arbitrary initial state will evolve in time .
while numeric simulations have made substantial progress investigating specific configurations @xcite , quantum many - body problems in general become intractable beyond only a few dozen spins due to the exponential scaling of hilbert space dimension with system size @xcite .
typically , such problems become even more computationally difficult as the dimensionality of the spin system is increased @xcite . quantum simulation , in which the many - body problem of interest is encoded within a well - controlled experimental quantum system @xcite ,
has proven an increasingly powerful technique for studying the behavior of interacting quantum spins .
such quantum simulators should be easily reconfigurable and contain widely tunable parameters , so that they may investigate a broad variety of problems in disparate physical regimes @xcite .
recent advances have used collections of trapped ions to investigate quantum phase transitions @xcite , explore open quantum systems @xcite , witness the growth of quantum correlations and entanglement @xcite , and directly measure the many - body energy spectrum @xcite in systems of up to @xmath1 fully - coupled spins . in all cases ,
however , experiments have been restricted to one - dimensional ion chains emulating one - dimensional spin models .
although effective 2d systems can be realized in a 1d chain by applying appropriate decoupling pulses and trotterized sequences @xcite , this approach scales very poorly due to the large number of required quantum gate operations . by instead constructing an ion trap quantum simulator with native 2d interactions
, one can begin to address many of the open topics in quantum many - body physics that become important in two dimensions , such as geometric frustration @xcite , exotic phases of matter ( such as spin glasses @xcite and liquids @xcite ) , and the relationship between entanglement , frustration , and high-@xmath0 superconductivity @xcite
. it would be strongly desirable for such a 2d trapped - ion quantum simulator to retain the traditional 1d ion - trap strengths : full control at the single - particle level , site - resolved measurements and readout , and spin - spin coupling rates that are fast compared to the decoherence rate . however , current efforts to build 2d trapped - ion systems in penning traps @xcite and microfabricated arrays @xcite have yet to solve issues of individual ion addressing and slow coupling rates , respectively . here ,
i propose the use of standard radio - frequency ( rf ) paul traps for use in 2d quantum simulation experiments .
i will show that it is possible to choose appropriate trap parameters so that the ions driven motion called micromotion will have a negligible effect on the outcome of a quantum simulation , even for hundreds of trapped ions . with such parameters
, the ions will remain individually addressable and resolvable , with spin - spin coupling rates comparable to those seen in 1d experiments .
the paper is organized as follows : sec .
ii reviews the standard rf paul trap , and introduces a choice of trap parameters that leads to a 2d triangular lattice of ions .
iii explicitly investigates the effects of micromotion on the crystal described in sec .
ii , calculating the shift in equilibrium ion positions , the change in normal mode structure , and an updated stability criterion for achieving a 2d structure within an rf trap . having developed this full analysis including micromotion , sec .
iv shows how the crystal can be used to perform quantum simulations of 2d spin models with strong coupling rates .
section v offers an outlook for future experiments and some concluding remarks .
an rf paul trap may be operated in a regime such that the trapped ion coulomb crystal self - assembles in a 2d plane .
we will investigate this regime here within linear paul traps with four segmented blades ( one of the most common trap designs @xcite ) , though the results are equally applicable for other trap geometries @xcite .
for a typical `` blade''-style trap , the central segments of two opposing blades are driven with an rf voltage @xmath2 at frequency @xmath3 , while the other two central segments are held at rf ground .
the outer electrode segments are biased with a dc voltage @xmath4 . near the center of the trap
, the potential can be written as @xcite : @xmath5 where @xmath6 and @xmath7 are the radial and axial trap dimensions and @xmath8 is a geometric factor of order unity . when cooled to millikelvin temperatures , ions trapped in this potential
behave as though they were in a 3d harmonic pseudopotential , @xmath9 where the radial and axial trapping frequencies are given by : @xmath10 with @xmath11 the ion charge , @xmath12 the ion mass , and @xmath13 the mathieu @xmath14 " parameter .
typically , a small asymmetry is introduced in the electrode structure to break the degeneracy of the @xmath15 and @xmath16 axes , thereby preventing a zero - frequency rotational mode and providing a unique minimum - energy configuration ; for the numeric simulations to follow , the trap frequency asymmetry between the @xmath15 and @xmath16 directions is set to be @xmath17 .
if the radial trap frequencies @xmath18 are much stronger than the axial frequency @xmath19 ( the typical regime for most experiments ) , the ions will form a 1d chain along the central trap axis .
as @xmath19 is increased ( while holding @xmath18 and the number of ions fixed ) , the linear chain passes through a series structural phase transitions into new configurations : zig - zag , helical , and ultimately a 2d triangular lattice in the radial plane @xcite .
this final arrangement , which is desired for the 2d quantum simulation experiments proposed here , requires that @xmath19 be large compared with @xmath18 , with the ratio scaling weakly with the number of ions @xmath20 in the trap @xcite : @xmath21 hence , a 100-ion system requires an axial frequency @xmath22 times larger than the radial frequencies . from eq .
[ eqn : trapfreqs ] , a large axial frequency can be achieved by increasing the voltages @xmath4 applied to the outer trap segments .
however , if the dc voltage @xmath4 is too large compared with the rf voltage @xmath2 , this has a destabilizing effect on the crystal : the radial frequency ( eq . [ eqn : trapfreqs ] ) becomes imaginary , and ions no longer have bound trajectories .
the ion trap voltages , frequencies , and sizes thus must be carefully chosen in order to achieve a stable and robust 2d planar crystal .
the full stability regime for a 2d ion lattice in an rf trap , which depends on the axial and radial trap frequencies as well as the number of ions , will be calculated and shown in sec .
iii(c ) . for the numeric simulations presented in upcoming sections ,
i choose the following set of trap parameters : @xmath23 and @xmath24 which result in radial and axial trap frequencies of @xmath25 khz and @xmath26 .
these parameters lie squarely within the stability regime for a 2d crystal of @xmath27yb@xmath28 ions and are all straightforward to achieve in the laboratory .
however , before implementing such a trap for quantum simulation experiments , we must first consider the effects of rf driven motion on the 2d ion crystal .
for a 2d crystal in a linear paul trap , each ion is subject to rf - driven micromotion with an amplitude proportional to the ion s distance away from the central trap axis . for several types of ion - trap experiments ,
even small amounts of micromotion can have significant harmful effects : it can lead to large systematic doppler shifts in ion - based atomic clocks @xcite , and it can substantially reduce the fidelities of quantum gates during a quantum computation ( in the absence of advanced micromotion - correcting protocols ) @xcite . nevertheless , many successful experiments do not depend sensitively on micromotion amplitude @xcite , and ion coulomb crystals of up to @xmath29 particles have been confined in rf paul traps @xcite .
this section will demonstrate that for carefully chosen parameters ( such as those introduced above ) , micromotion effects on 2d quantum simulation experiments are both predictable and small .
this result is enabled by the orientation of the 2d crystal : there is effectively no micromotion amplitude in the axial ( transverse ) direction , since the ions are compressed to a single plane at the axial trap center . by utilizing the axial modes of motion for quantum simulation protocols , one can thus sidestep the large effects of micromotion in the radial plane .
nevertheless , 4 residual effects in the axial plane must still be considered : ( 1 ) a shift in the equilibrium ion positions , ( 2 ) the resolvability of ions due to radial micromotion amplitude , ( 3 ) shifts in the axial normal mode frequencies , and ( 4 ) an altered stability region for maintaining a 2d crystal .
these effects are each explored in detail below . in the absence of micromotion , finding the equilibrium positions for a 2d ion crystal proceeds similarly to the 1d case . in the radial plane ,
the potential experienced by the ions has contributions from the trap voltages as well as the coulomb interaction : @xmath30 the equilibrium positions are the set of coordinates @xmath31 that minimize the energy .
although direct numerical minimization is possible for small system sizes , finding the equilibrium configuration for larger numbers often requires the use of molecular dynamics simulations with added dissipation @xcite ; this is the approached used for calculations here . in the presence of micromotion
, there is no longer an `` equilibrium '' position ; the radial coordinates of each ion vary in time as @xmath32 where @xmath33 is the ion s average position , and the higher order terms indicate the amplitude of the motion at the @xmath34 harmonic of the drive frequency @xmath3 . following @xcite ,
each of these amplitudes @xmath35 for each ion can be extracted by self - consistently solving the equations of motion within the full trap potential of eq .
[ eq : trappotential ] .
the calculated central positions @xmath33 for a 20- and 100-ion crystal , using the parameters introduced in sec .
ii and including micromotion , are shown in fig .
[ fig : micromotion]a - b . the crystals
self - organize into a 2d triangular lattice , with a 4.3 @xmath36 m average inter - ion spacing for the 100-particle case . when accounting for micromotion , the positions are found to shift by an average of only 0.08 @xmath36 m
a very small change that would not be noticeable with standard imaging techniques and will not have any foreseeable consequences for 2d quantum simulations .
the amplitude of the radial micromotion is also important to consider ; large amplitude excursions could obscure the individual ion positions , preventing quantum spin readout .
the amplitudes of the micromotion - induced terms in eq .
[ eq : mmposition ] may be written as @xmath37 and @xmath38 @xcite , where @xmath39 is the mathieu parameter . since @xmath40 for any stable rf trap , the first order amplitude dominates the time - dependent part of eq .
[ eq : mmposition ] , and smaller @xmath14 parameters are advantageous for minimizing the micromotion amplitude . for the trap settings proposed in sec .
ii , @xmath41 ; for the 100-ion case , this leads to a micromotion amplitude of 1.4 @xmath36 m for the most radially extended particle still small compared to the 4.3 @xmath36 m inter - ion distance ( see fig . [
fig : micromotion ] insets ) . since the radial extent of the 2d crystal scales as @xmath42 for @xmath20 particles with separation distance @xmath43 , the maximum micromotion amplitude scales as @xmath44 .
if we demand that this amplitude be smaller than half the inter - ion distance ( @xmath45 ) , this constrains the maximum number of trapped ions to be @xmath46 , which is @xmath47 for the chosen trap parameters . to calculate the normal mode frequencies of an ion crystal ,
one typically expands the coulomb potential around the equilibrium positions to second order , then diagonalizes the resulting matrix to find the eigenfrequencies and eigenmodes @xcite .
however , the presence of micromotion in the radial plane can have a notable effect on the axial mode frequencies and eigenfunctions ( even though there is no axial micromotion ) , and must be taken into account @xcite . as we will see , the primary effect of radial micromotion is to decrease the frequencies of all axial modes ( except for the center - of - mass ) , with the low - frequency modes experiencing the largest shift . to second order , the axial potential experienced by the ions is given by @xmath48 where @xmath49 .
due to the radial micromotion , the ion position differences @xmath50 are dynamic ( eq .
[ eq : mmposition ] ) , which implies that the axial normal mode frequencies will also inherit a time dependence . however , since @xmath51 , it is sufficient to consider the expectation value of the potential @xmath52 over one period of micromotion , and diagonalize this resulting matrix to find the normal mode eigenfrequencies .
the results of this calculation are shown in fig .
[ fig : modes](a ) , along with several representative mode eigenfunctions , for 100 trapped ions . here
, the highest frequency axial ( transverse ) motion corresponds to the center - of - mass ( com ) mode ( just as in the 1d case ) and remains unchanged in the presence of micromotion .
however , fig .
[ fig : modes](b ) shows that all other mode frequencies are suppressed under micromotion ; for the lowest ( zig - zag ) mode , the reduction is more than 0.5 mhz @xmath53 .
this change in mode structure will not impede 2d quantum simulations ; rather , its effect will be to increase the effective spin - spin interaction range between ions in the lattice ( as will be shown in sec .
when quantum simulating spin models that are sensitive to the specific value of interaction range , one must thus take into account this micromotion - induced frequency shift .
consider a 1d linear ion chain , which is characterized by a set of transverse normal mode frequencies .
if the axial frequency @xmath19 is increased while holding all other parameters fixed , the ions will be pushed closer together , and all transverse mode frequencies ( except for the com ) will shift downwards .
eventually , the chain will undergo a structural phase transition and `` buckle '' at the center ; this corresponds to the lowest transverse mode ( i.e. the zig - zag mode ) shifting downwards to zero frequency .
red points show the calculated boundary in the absence of micromotion , in agreement with eq .
[ eq : anisotropy ] ( red dashed line ) . when micromotion effects are included , the entire region below the blue points becomes unstable .
this stability boundary is also found to exhibit weak power - law scaling with the number of ions ( blue dashed line ) . ]
analogous effects occur for 2d planar ion crystals : when the radial confinement becomes too strong compared to the axial confinement , the zig - zag transverse mode crosses zero frequency , and the crystal buckles into 3d .
this effect has been well - described in @xcite , which led to the scaling law in eq .
[ eq : anisotropy ] .
however , this earlier analysis did not explicitly consider rf traps , where micromotion can have significant impacts on crystal stability . the previous section ( as well as fig .
[ fig : modes ] ) demonstrated that micromotion itself causes a downward shift in the transverse mode frequencies .
this observation implies that micromotion can have a destabilizing effect : the zig - zag mode can be pushed to zero - frequency , even for crystals that are `` stable '' according to eq .
[ eq : anisotropy ] .
thus , one must carefully calculate the mode structure including micromotion before concluding that the ions will remain in the desired planar geometry .
[ fig : stability ] shows the boundary between stable and unstable 2d planar crystals , using the trap parameters of sec .
ii with @xmath54 mhz , when micromotion is ignored ( red points ) and included ( blue points ) .
as argued above , the stability region including micromotion is reduced when compared with the predictions of eq .
[ eq : anisotropy ] . as before ,
the boundary scales weakly with the number of ions @xmath20 ; now , @xmath55 ( compared to @xmath56 previously ) .
the primary effect of micromotion is thus a multiplicative increase in the ratio @xmath57 required for stability ; a _ very _ rough rule would be to calculate the needed frequency ratio via eq .
@xmath58 , then add 45@xmath59 to account for effects of micromotion .
by loading @xmath27yb@xmath28 ions into the trap described above , one can engineer an effective 2d many - body spin system for quantum simulations . as in 1d ion trap experiments , effective spin qubits can be encoded in the hyperfine ground states @xcite , which are first - order insensitive to external magnetic field noise and can yield coherence times of over 10 minutes @xcite .
if the ions are irradiated with an appropriate frequency of laser light , the spin - dependent fluorescence from each ion allows for a high - fidelity measurement of the projected spin state @xcite .
[ eq : jij ] ) , which are shown for an edge spin ( a ) and a central spin ( b ) in a 100-ion array using the trap parameters detailed in the text . for these parameters , the couplings fall of algebraically with distance as @xmath60 , with @xmath61 tunable between 0 and 3 .
panel ( c ) shows this power - law decay for the edge spin chosen in ( a ) . ]
the effective spin qubits within each ion can be coupled together by applying spin - dependent optical dipole forces @xcite .
these forces are induced by global , far - detuned raman transitions at 355 nm that virtually excite the collective modes of ion motion .
if the wavevector difference between the two raman beams @xmath62 lies along the axial direction of the trap , this will excite the axial motional modes @xmath63 ( the desired direction to avoid problematic micromotion effects ) .
the two beams should contain a pair of beat - note frequencies symmetrically detuned from the hyperfine splitting by a frequency @xmath36 , which is comparable to the center - of - mass axial frequency .
when a resonant carrier interaction is added , this arrangement results in an effective transverse - field ising hamiltonian @xcite ; @xmath64 where @xmath65 has been set to 1 , @xmath66 is an effective transverse magnetic field , and the long - range spin - spin couplings are given by @xmath67 where @xmath68 is the carrier rabi frequency , and @xmath69 the normal mode eigenvector component of the @xmath70 ion in mode @xmath12 .
having calculated the mode frequencies @xmath63 and eigenfunctions @xmath69 using the techniques outlined in sec .
iii , the complete spin - spin coupling matrix @xmath71 follows immediately ( eq . [ eq : jij ] ) , and is shown in fig .
[ fig : jij](a)-(b ) for two different spins in the 2d lattice . here , the carrier rabi frequency is set to @xmath72 mhz , and the laser detuning @xmath36 was chosen to be blue of the center - of - mass axial mode by approximately @xmath73 khz ( ensuring that residual phonon effects are kept small @xcite ) . these proposed parameters
yield a nearest - neighbor antiferromagnetic spin - spin coupling of approximately @xmath74 khz ( which is nearly an order of magnitude faster than typical decoherence rates @xcite ) , as well as a long - range coupling that decays with distance @xmath75 as @xmath76 . in general , by choosing different values of @xmath36 , long - range interactions can be continuously tuned to decay with distance as any power between @xmath77 and @xmath78 @xcite .
this work has argued that rf paul traps with appropriately chosen parameters can serve as a scalable platform for developing 2d quantum simulation experiments . in such traps ,
the ions self - assemble into a triangular lattice with tunable , long - range couplings given by eq .
[ eq : jij ] . as pictured in fig .
[ fig : lattices ] , it would also be possible to shelve specific ions in electronic states outside of the qubit subspace , allowing for multiple types of effective lattice configurations @xcite .
the residual effects of micromotion in 2d traps can be well - characterized , leading to a slightly longer spin - spin interaction range and a reduced ( but still easily achievable ) 2d trap stability region
. it should be straightforward to achieve several hundreds of ions for use in these experiments , limited by the ability to individually resolve and address the atoms ( see sec .
iii(a ) ) and the ability to stably confine large numbers in a 2d configuration using reasonable laboratory voltages ( see fig .
[ fig : stability ] ) . using an appropriate combination of raman laser frequencies , amplitudes , and phases
, it will be possible to quantum simulate a diverse toolbox of spin - model hamiltonians .
ising couplings naturally occur as a result of the applied spin - dependent force @xcite , and have been demonstrated in numerous 1d quantum simulation experiments @xcite .
recently , 1d simulations have also demonstrated an xy spin model by applying a large resonant raman transition in conjunction with the ising interaction @xcite . while not yet
experimentally demonstrated , it should also be possible to realize full heisenberg spin - spin interactions of the form @xmath79 , following the ideas proposed in refs . @xcite .
these 1d advances can be directly applied to 2d quantum simulations , since they both fundamentally operate by coupling @xmath27yb@xmath28 ions together through collective motional modes . with the ability to apply a variety of different spin hamiltonians , along with tunable spin - spin coupling strengths @xmath71
, one can fully implement a quantum simulator to explore many of the important open questions in 2d quantum many - body physics .
the author is grateful to shengtao wang for illuminating discussions .
this work is supported by the air force office of scientific research award no .
fa9550 - 16 - 1 - 0277 , and the college of arts and sciences at indiana university . | the computational difficulty of solving fully quantum many - body spin problems is a significant obstacle to understanding the behavior of strongly correlated quantum matter .
experimental ion - trap quantum simulation is a promising approach for studying these lattice spin models , but has so far been limited to one - dimensional systems .
this work argues that such quantum simulation techniques are extendable to a 2d ion crystal confined in a radiofrequency ( rf ) trap .
using appropriately chosen parameters , driven ion motion due to the rf fields can be made small and will not limit the types of quantum spin models that can be experimentally encoded .
the rf - driven motion is calculated to modestly reduce the stability region of a 2d crystal and must be considered when designing the 2d trap .
the system will be scalable to 100 + quantum particles , far beyond the realm of classical intractability , while maintaining the traditional ion - trap strengths of individual - ion control , long quantum coherence times , and site - resolved projective spin measurements . |
the famous singularity theorem of penrose ( see @xcite ) states that if in addition to the dominant energy condition , the space - time has a trapped surface , namely a two dimensional space - like sphere whose outgoing and incoming expansions are negative , thus the space - time is future causally geodesically incomplete ( we usually say that the space - time contains a _ singularity _ ) . the _ weak cosmic censorship _
conjecture asserts there is no naked singularity under reasonable physical assumptions .
in other words , singularities need to be hidden from an observer at infinity by the event horizon of a black hole .
thus , by combining these two claims , if one can exhibit a trapped surface in a space - time , then one can predict the existence of black holes .
in other words , although many supplementary conditions are required , we regard the existence of a trapped surface as the presence of a black hole .
a major challenge in general relativity is to understand how trapped surfaces actually form due to the focusing of gravitational waves . in a recent remarkable breakthrough @xcite
, christodoulou solved this long standing problem .
he discovered a mechanism which is responsible for the dynamical formation of trapped surfaces in vacuum space - times . in the monograph @xcite , in addition to the minkowskian flat data on a incoming null hypersurface , christodoulou identified an open set of initial data ( this is the _ short pulse _ ansatz ) on a outgoing null hypersurfaces .
based on the techniques developed by himself and klainerman in the proof of the global stability of the minkowski space - times @xcite , he managed to understand the whole picture of how the various estimates on geometric quantities propagates along the evolution .
once those estimates are established in a large region of the space - time , the actual formation of trapped surfaces is easy to demonstrate .
christodoulou also proved a version of the same result for the short pulse data prescribed on past null infinity .
he showed that strongly focused gravitational waves , coming in from past null infinity , lead to a trapped surface .
this miraculous work provides the first global _ large data _ result in general relativity ( without symmetry assumptions ) and opens the gate for many new developments on dynamical problems related to black holes .
the methods undoubtedly have many future applications in both general relativity and other partial differential equations
. in @xcite , klainerman and rodnianski extend aforementioned result which significantly simplifies the proof of christodoulou ( from about six hundred pages to one hundred and twenty ) .
they enlarge the admissible set of initial conditions and show that the corresponding propagation estimates are much easier to derive .
the relaxation of the propagation estimates are just enough to guarantee that a trapped surface still forms . based on the trace estimates developed in a sequence of work @xcite , @xcite , @xcite , @xcite , @xcite and @xcite towards the critical local well - posedness for einstein vacuum equations ,
they reduce the number of derivatives needed in the argument from two derivatives on curvature ( in christodoulou s proof ) to just one .
more importantly , klainerman and rodnianski introduce a parabolic scaling in @xcite which is incorporated into lebesgue norms and sobolev norms .
these new techniques allow them to capture the hidden _ smallness _ of the nonlinear interactions among different small or large components of various geometric objects .
the result of klainerman and rodnianski can be easily localized with respect to angular sectors which leads to further developments , see @xcite for details .
we remark that klainerman and rodnianski only considered the problem on a finite region .
the question from past null infinity can be solved in a similar manner as in @xcite once one understand the picture on a finite region .
the problem from past null infinity has also been studied in a recent work @xcite by reiterer and trubowitz .
the aforementioned works are all investigating vacuum space - times . when some matter field is presenting , the formation of black holes has attracted a lot of interest .
we mention only two of them which are more related to the present paper . in @xcite , based on their work on positive mass theorem , especially the resolution of jang s equation , schoen and yau proved the existence of a trapped surface when matter is condensed in a small region , see @xcite for an improvement .
their result is restricted to the initial 3-slice hence is not dynamical .
another work is an earlier paper @xcite of christodoulou .
he studied the evolutionary formation of singularities for the einstein - scalar field system .
the radial symmetry is assumed and the mechanism of the formation of trapped surface is quite different to his most recent work @xcite . and moreover , he could get more precise information on the gravitational collapse . the purpose of the present paper is , under the framework of christodoulou and klainerman - rodnianski , to understand the formation of charged black holes , namely the evolutionary formation of trapped surfaces when there is an electromagnetic field coupled to the background space - time . on one hand , we will generalize the black - hole - formation results of christodoulou and klainerman - rodnianski .
in fact , by switching off the electromagnetic field in our main theorem , we can retrieve their results in vacuum . on the other hand
, we will show that the formation of black hole can purely due to condensation of maxwell field on the initial null hypersurface , namely , we can set the incoming gravitational energy to be zero .
we recall double - null - foliation formalism and we refer the reader to chapter 3 of @xcite for more precise descriptions .
we use @xmath1 to denote the underlying space - time and use @xmath2 to denote the background metric .
we assume that @xmath3 is spanned by a double null foliation generated by two optical functions @xmath4 and @xmath5 and we also assume that @xmath4 and @xmath5 increase towards the future , @xmath6 and @xmath7 .
we use @xmath8 to denote the outgoing null hypersurfaces generated by the level surfaces of @xmath4 and use @xmath9 to denote the incoming null hypersurfaces generated by the level surfaces of @xmath5 .
we use @xmath10 to denote the space - like two surface @xmath11 .
we denote by @xmath12 the region of @xmath8 defined by @xmath13 ; similarly , we can define @xmath14 .
we require that when @xmath15 , this part of initial null cone @xmath16 is a flat light cone in minkowski space - time .
the shaded region represents the domain @xmath17 .
the function @xmath5 is in fact defined from @xmath18 to @xmath19 .
when @xmath15 , this part of @xmath16 is assumed to be a flat light cone in minkowski space - time with vertex located at @xmath20 .
we shall show that the trapped surface forms when @xmath21 and @xmath22 .
we also require that @xmath23 is a fixed number which is at least @xmath24 , say , we may take @xmath25 .
let @xmath26 be the null geodesic generators of the double null foliation and we define the lapse function @xmath27 by @xmath28 .
we also need normalized null pair @xmath29 , i.e. @xmath30 , @xmath31 and @xmath32 . on a given two sphere
@xmath10 we choose an orthonormal frame @xmath33 .
we call @xmath34 a _
null frame_. to denote an index from @xmath35 to @xmath36 and latin letters @xmath37 to denote an index from @xmath35 to @xmath24 .
repeated indices should always be understood as summations . ]
we use @xmath38 to denote levi - civita connection of the metric @xmath2 and we define the _ connection coefficients _ , @xmath39 where @xmath40 . on @xmath41
we use @xmath42 to denote the induced connection and we use @xmath43 and @xmath44 to denote the projections to @xmath41 of @xmath45 and @xmath46 .
those @xmath42 derivatives are called _ horizontal derivatives _ ( we use the word _ horizontal tensor fields _ for tensor fields depending only on components along @xmath10 , see @xcite ) . for a weyl field @xmath47 , the null decomposition adapted to the null frame @xmath48 is is equal to weyl curvature tensor
, we use @xmath49 to denote its null components . ]
@xmath50 where @xmath51 is the space - time hodge dual of @xmath47 .
maxwell field _ shall always refer to a two form @xmath52 satisfying _ maxwell equations _ , @xmath53 } = 0 , \quad d^\alpha f_{\alpha\beta } = 0.\ ] ] the maxwell field @xmath52 is coupled to the background geometry , namely , in addition to , @xmath52 satisfies einstein field equations , @xmath54 where @xmath55 is ricci curvature , @xmath56 is scalar curvature and the @xmath57 is the energy - momentum tensor associated to @xmath52 defined as @xmath58 we remark that the energy - momentum tensor @xmath57 can also be written as @xmath59 where @xmath60 denotes the space - time hodge dual of @xmath61 . the symmetry between @xmath61 and @xmath62 in above expression plays an important role when we derive energy estimates .
we also decompose @xmath61 in a given null frame into null components , , @xmath63 , @xmath64 and @xmath65 to denote these components . ]
@xmath66 using null components , are equivalent to following null structure equations , see @xcite , @xcite or @xcite for details . to denote the energy momentum tensor of @xmath61 . ]
@xmath67 @xmath68 @xmath69 @xmath70 @xmath71 @xmath72 @xmath73 @xmath74 @xmath75 @xmath76 @xmath77 @xmath78 @xmath79 @xmath80 @xmath81 @xmath82 @xmath83 the second bianchi equations are equivalent to null bianchi equations , @xmath84 @xmath85 @xmath86 @xmath87 @xmath88 @xmath89 @xmath90 @xmath91 @xmath92 @xmath93 the maxwell equations are equivalent to null maxwell equations , @xmath94 @xmath95 @xmath96 @xmath97
@xmath98 @xmath99 we recall the concept of _ signature _ and _ scale _ for very horizontal tensor field introduced in @xcite .
let @xmath100 be a horizontal tensor , we use @xmath101 , @xmath102 and @xmath103 to denote the number of times @xmath104 , respectively @xmath105 and @xmath106 appearing in the definition of @xmath100 .
we define @xmath107 the _ signature _ of @xmath100 to be @xmath108 .
we also assign a _ scale _
@xmath109 for @xmath100 to be @xmath110 . for horizontal derivatives
, we also assign signatures via the following convention , @xmath111 , @xmath112 and @xmath113 . for product of two horizontal fields ,
we define the signature to be the sum , i.e. @xmath114 .
we introduce the _ scale invariant norms _ la klainerman and rodnianski . along the outgoing null hypersurfaces @xmath115 and incoming null hypersurfaces
@xmath116 , we define scale invariant @xmath117 norms , @xmath118 on two surface @xmath10 , we define the scare invariant @xmath119 norms , for this norm . ]
@xmath120 by definition , scale invariant norms come up naturally with a small parameter @xmath19 . roughly speaking , it reflects the _ smallness _ of the nonlinear interactions in our problem . one key aspect of this principle is the scale invariant hlder s inequality , @xmath121 where @xmath122 .
similar inequalities hold along incoming and outgoing hypersurfaces .
we emphasize the rule of thumb for treating the nonlinear terms .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ for a product of terms each use of hlder s inequality gains a @xmath123 .
_ _ _ _ _ _ _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ in this sense , nonlinear terms are better than linear terms
. other other hand , there are nonlinear terms where we do _ not _ gain any power in @xmath19 .
if @xmath124 is a bounded ( in usual sense ) scalar function ( bounded by a universal constant ) , the best we expect is @xmath125 .
appearing frequently in the paper . ]
we now introduce schematic notations @xmath126 ,
@xmath127 , @xmath128 , @xmath129 , @xmath130 , @xmath131 , @xmath132 and @xmath133 .
we use @xmath126 , @xmath128 and @xmath131 to denote any connection coefficient , , we always use @xmath134 defined by @xmath135 . ]
null curvature component and null maxwell component respectively ; we use @xmath127 , @xmath129 and @xmath132 to denote a _ good _ connection coefficient , null curvature component and null maxwell component respectively , i.e. @xmath136 , @xmath137 and @xmath138 .
we require @xmath139 and @xmath140 . for @xmath141 or @xmath36 , @xmath142 , @xmath143 and @xmath144
, we introduce a family of scale invariant norms for connection coefficients , to denote the sum for all the possible @xmath145 and @xmath146 @xmath147 for curvature components , @xmath148 for maxwell components , @xmath149 in those definitions , some terms come up with a @xmath123 , for example , @xmath150 in @xmath151 .
those terms are understood to _ cause a loss of @xmath152_. we call them _ anomalies _ or being _ anomalous_. by convention , we do not regard @xmath153 and @xmath63 as anomalies , since they are not anomalous on outgoing lightcones . in sequel , we shall encounter other terms which cause a loss of @xmath152 . by abuse of language
, we will call them _ anomalies _ too . and
, for those non - anomalous terms , in schematic notations , we usually use a @xmath2 as a subindex , for example , we use @xmath154 to denote @xmath155 . to rectify global @xmath156 estimate on anomalies , we need localized norms .
let @xmath157 be a patch of @xmath10 obtained by transporting a disc @xmath158 of radius @xmath123 along integral curves of @xmath159 and let @xmath160 be a piece of the hypersurface @xmath161 obtained by evolving a disc @xmath158 of radius @xmath123 along the integral curves of @xmath159 , we set @xmath162(u,\ub ) & = \sup_{{}^{\delta}\!s_{u,\ub } \subset s_{u,\ub } } \|\chih\|_{l^4_{(sc)}({}^{\delta}\!s_{u,\ub } ) } , \quad \ozerofour^{\delta}[\chibh](u,\ub ) = \sup_{{}^{\delta}\!s_{u,\ub } \subset s_{u,\ub } } \|\chibh\|_{l^4_{(sc)}({}^{\delta}\!s_{u,\ub})},\\ \rzero^{\delta}[\alpha](u ) & = \sup_{{}^{\delta}\!h_u \subset h_u } \|\alpha\|_{l^2_{(sc)}({}^{\delta}\!h_u ) } , \quad \fzero^{\delta}[\alpha_f](u ) = \sup_{{}^{\delta}\!h_u \subset h_u } \|\alpha_f\|_{l^2_{(sc)}({}^{\delta}\!h_u)}.\end{aligned}\ ] ] let @xmath163 and @xmath164 ; other norms are defined in a similar manner .
finally , we introduce total norms , @xmath165 we use @xmath166 , @xmath167 and @xmath168 to denote the corresponding total norms on initial surface @xmath16 .
let @xmath169 .
we define initial data norm , @xmath170.\end{aligned}\ ] ] our assumption on initial data is @xmath171 this ansatz ( when @xmath172 ) was discovered in @xcite and it is larger than those in @xcite .
unfortunately , the formation of trapped surfaces can not be derived from ansatz .
following @xcite , we shall present necessary modifications in * main theorem*. nevertheless , ansatz implies under the ansatz , along initial outgoing hypersurface @xmath16 , if @xmath19 is sufficiently small , we have @xmath173 thanks to this proposition , we can replace ansatz by @xmath174 we omit the proof of this proposition since the it is very much similar to the vacuum case .
in the vacuum case , the reader can find a proof in chapter 2 of christodoulou s book @xcite .
in general , the proof is mainly to chase null structure equations , null bianchi equations and null maxwell equations one by one along @xmath16 .
the proof actually shows that @xmath0 can be freely prescribed along @xmath16 for einstein - maxwell system .
this first result of the paper addresses the propagation of to later time , this is the content of * theorem c*. roughly speaking , modulo the electromagnetic field , curvatures are one derivative of connection coefficients .
thus , once one controls connection coefficients , one expects to control the curvature components . on the other hand
, one can gain the control of connection coefficients provided bounds on curvatures .
this later statement is recorded in * theorem a*. together with * theorem b * which , at the end of section [ section_trace_rotation ] , addresses estimates on angular momentums , * theorem b * serves as an intermediate step toward * theorem c*. we use @xmath175 to denote constants depending only on @xmath176 , @xmath177 , @xmath178 , @xmath179 and @xmath180 . [ theorema ] assume that @xmath176 , @xmath177 , @xmath178 , @xmath179 and @xmath180 are finite .
if @xmath19 is sufficiently small , we have @xmath181 \lesssim \ozero + c \delta^\frac{1}{4}.\ ] ] [ theoremc ] under ansatz , if @xmath19 is sufficiently small , we have @xmath182 finally , we state the main theorem of the paper and its immediate corollaries .
in addition to ansatz , we assume that @xmath183 for sufficiently small @xmath184 such that @xmath185 . on @xmath16
, we also assume @xmath186 where @xmath187 is a universal constant and @xmath188 is the maximal radius of the flat part of @xmath16 .
then , if @xmath19 is sufficiently small , @xmath16 is free of trapped surfaces and a trapped surface forms at @xmath189 .
we can set shear @xmath190 on @xmath16 but @xmath191 large enough to form trapped surfaces .
this can be regarded as formation of black hole purely due to the condensation of matter , there exists an initial data set for maxwell field such that the shear of @xmath16 is trivial , @xmath16 is free of trapped surfaces and a trapped surface forms at @xmath189 .
we can also switch off electromagnetic field to retrieve earlier results for vacuum space - times , ( _ christodoulou _ @xcite ; _ klainerman and rodnianski _ ,
@xcite ) for vacuum , there exists an initial data set on @xmath16 such that @xmath16 is free of trapped surfaces and a trapped surface forms at @xmath189 .
we prove * theorem a * in this section . the proof is based on following bootstrap assumption , @xmath192 where @xmath193 is a large positive number .
we will show that if @xmath19 is sufficiently small , all the statements in the theorem hold .
in particular , @xmath194 .
we also show @xmath195 . since @xmath175 is independent of @xmath193
, we obtain a better bound in therefore close the bootstrap argument .
once we prove * theorem a * , we can use the estimates derived in the proof throughout the paper .
we emphasize that _ everything _ in sequel is based on bootstrap assumption .
the proof is organized as follows : we first show some preliminary estimates , and then prove step by step @xmath196,@xmath197 , @xmath198 and second derivatives estimates on connection coefficients . in this section , we provide some preliminary estimates based on without detailed proof .
the reader can find details in section 4 of @xcite .
if @xmath19 is sufficiently small , more precisely if @xmath199 , we have @xmath200 the proof is straightforward by integrating @xmath201 . for @xmath202 or @xmath36 , we have following integral estimates , [ integral_estimates ] for a given horizontal tensor field @xmath126 , if @xmath203 is sufficiently small , we have the following scale invariant estimates : @xmath204 @xmath205 we define the transported coordinates along double null foliation . for a given local coordinate system @xmath206 on @xmath207 , , the metric is the round metric with constant gauss curvature .
] we parameterize points along the outgoing null geodesics starting from @xmath208 by corresponding @xmath206 and the affine parameter @xmath5 ; similarly , for a given local coordinate system @xmath209 on @xmath210 , we parameterize points along the incoming null geodesics starting from @xmath211 by the corresponding @xmath206 and the affine parameter @xmath4 .
corresponding metrics are denoted by @xmath212 and @xmath213 .
we have estimates on the geometry of @xmath10 , let @xmath214 be the standard metric on @xmath215 . if @xmath19 is sufficiently small , we have @xmath216 for christoffel symbols @xmath217 and @xmath218 whose signatures are @xmath219 , we have @xmath220 we can follow exactly strategy in section 4.5 and section 4.9 of @xcite to prove it .
the control of the geometry of @xmath10 at the level of metric and christoffel symbols allows us to derive the standard sobolev inequalities , elliptic estimates on hodge systems and trace inequalities .
[ sobolev_trace_estimates ] for a horizontal tensor @xmath126 , @xmath221 , @xmath222 and @xmath223 , if @xmath19 is sufficiently small , we have @xmath224 @xmath225 @xmath226 @xmath227 @xmath228 last inequality allows us to rectify the anomalies of @xmath229 , @xmath230 and @xmath191 in @xmath231 estimates .
we also have following interpolation estimates , [ interpolation_estimates ] if @xmath19 is sufficiently small , we have @xmath232 @xmath233 let @xmath234 for definitions . ]
be a hodge operator , we have following elliptic estimates , [ estimates_for_hogde_systems ] on @xmath235 , for a solution @xmath126 of the hodge system @xmath236 , if @xmath19 is sufficiently small , we have @xmath237 @xmath238 @xmath239 we provide localized estimates on @xmath150 and @xmath191 .
if @xmath19 is sufficiently small , then we have @xmath240 \lesssim \fzero^{\delta}[\alpha](0 ) + \f , \ , \rzero^{\delta}[\alpha ] \lesssim \rzero^{\delta}[\alpha](0 ) + \r + \f.\ ] ] by direct integration , we have @xmath241 we use to replace the last integrand , use hlder inequality and schematic form of . @xmath242
last term can be absorbed by gronwall s inequality .
this yields the desired estimates for @xmath191 .
the proof for @xmath150 is more or less the same .
the key is to use .
we also have to pay attention to @xmath131 terms form . by applying null maxwell equations to convert the null derivatives of a component into either tangential derivatives or lower order terms ,
those terms can be written as @xmath243 this allows one to use hlder s inequality and to conclude . in this section
, we derive @xmath196 as well as @xmath244 estimates on connection coefficients and maxwell field .
we first derive estimates on @xmath246 , @xmath247 , @xmath229 and @xmath248 , then we derive estimates on @xmath249 , @xmath250 , @xmath230 and @xmath134 . for @xmath246 , in view of , we have @xmath251 , therefore , @xmath252 thus , by integrating along outgoing directions , @xmath253 in view of , one concludes @xmath254 for @xmath247 , similarly , we can show @xmath255 for @xmath229 , according to , we have @xmath256 . in view of and the fact that @xmath229 has trivial data along @xmath257 , we have @xmath258 we multiply both sides by @xmath259 to derive @xmath260 \lesssim \delta^{\frac{3}{4 } } \delta_0 \ , \oszerofour + \rzero[\alpha]^{\frac{1}{2}}\rone[\alpha]^{\frac{1}{2}}+\rzero[\alpha].\ ] ] we also need localized estimates for @xmath229 .
the procedure is similar , @xmath261 hence , @xmath262 + \rone[\alpha].\ ] ] for @xmath263 , in view of , we derive @xmath264 combining and for @xmath191 , we derive @xmath265 for @xmath250 , in view , we have @xmath266 .
we bound @xmath267 by a constant to derive @xmath268 we also observe that @xmath269 , thus , @xmath270 thanks to gronwall s inequality , we remove @xmath247 on the right - hand side , @xmath271 for @xmath249 , similarly , we derive @xmath272 for @xmath230 , according to , @xmath273 , thus , @xmath274 as in , by gronwall s inequality , we ignore @xmath275 .
therefore , @xmath276 thus , @xmath277 \lesssim \ozero+ \delta^{\frac{3}{4 } } \delta_0 \oszerofour + c\delta^{\frac{1}{4}}.\ ] ] we repeat above procedure to derive localized estimates , @xmath278 for @xmath134 , since @xmath279 and @xmath280 , we have @xmath281 .
thus , implies , @xmath282 for the second term , we have @xmath283 , therefore , @xmath284 the double anomaly @xmath285 causes a loss of @xmath259 , @xmath286 combined with the bound of @xmath246 in , we have @xmath287 thanks to gronwall s inequality , we ignore @xmath288 on the right - hand side . thus , @xmath289 we add , , , , , , and together to derive @xmath290 if @xmath19 is sufficiently small , @xmath291 is absorbed by the left hand side .
next proposition summarizes the estimates derived in this subsection , [ ozero_four_estimates ] if @xmath19 is sufficiently small , we have @xmath292 moreover , we have @xmath293 \lesssim \rzero[\alpha]^{\frac{1}{2}}\rone[\alpha]^{\frac{1}{2}}+\rzero[\alpha ] + \delta^{\frac{1}{4 } } c , \,\ , \oszerofour[\chibh ]
\lesssim \ozero+ \delta^{\frac{1}{4 } } c,\ ] ] and their localized versions @xmath294 + \rone[\alpha ] , \,\,\|\chibh\|_{l^4_{(sc)}({}^\delta\!s_{u,\ub})}\lesssim \|\chibh\|_{l^4_{(sc)}({}^\delta\!s_{0,\ub } ) } + c.\ ] ] we introduce @xmath196 norms for maxwell field , @xmath295 ( u,\ub ) & = \delta^{\frac{1}{4}}\|\alpha_f\|_{l^4_{(sc)}(u,\ub ) } , \ !
\oszerofour[(\rho_f , \sigma_f , \alphab_f ) ] ( u,\ub ) = \|(\rho_f , \sigma_f , \alphab_f)\|_{l^4_{(sc)}(u,\ub)}.\end{aligned}\ ] ] for @xmath132 , in view of , and , we have @xmath296 . in view of and proposition [ ozero_four_estimates ] , we bound @xmath131 in @xmath231 and @xmath126 in @xmath297 .
thus , @xmath298 thus , by direct integration , @xmath299^{\frac{1}{2}}\fone[\upsilon]^{\frac{1}{2 } } + \delta^{\frac{1}{4}}\fone[\upsilon ] \lesssim c.\end{aligned}\ ] ] for @xmath191 , we rewrite as @xmath300 .
thus , @xmath301 we ignore @xmath302 on the right - hand side by gronwall s inequality to derive @xmath303^{\frac{1}{2}}\foneb[\upsilon]^{\frac{1}{2 } } + \delta^{\frac{1}{2 } } \foneb[\upsilon ] + c\delta^{\frac{1}{4 } } \delta_0.\end{aligned}\ ] ] this implies @xmath304(u,\ub ) \lesssim \oszerofour[\alpha_f](0,\ub ) + c \delta^{\frac{1}{4}}.\ ] ] we can also localize the proof to derive @xmath305 next proposition summarizes the estimates derived in this subsection , if @xmath19 is sufficiently small , we have @xmath306\lesssim c.\ ] ] moreover , we have @xmath307 we move on to @xmath244 estimates on @xmath126 , @xmath128 and @xmath131 .
it is straightforward since we have already established @xmath308 estimates .
since the procedure is standard , we will only sketch the proof .
if @xmath19 is sufficiently small , we have @xmath309 for @xmath150 , in view of , we have @xmath310 .
we proceed exactly as before : integrating along @xmath311 and removing @xmath312 by gronwall s inequality .
this yields @xmath313 for @xmath314 , we can bound it as follows , @xmath315(0),\ ] ] which yields the estimates on @xmath150 . for @xmath129 , in view of , , , and , @xmath316 .
this allows one to derive the bound for @xmath129 . for gaussian curvature @xmath317 , according to , we have @xmath318 .
therefore , @xmath319 ^ 2 \lesssim c.\ ] ] for @xmath320 , the corresponding null structure equations read as @xmath321 .
thus , @xmath322 we then integrate over @xmath8 to derive @xmath323 one key observation is , in above expressions , @xmath128 is not anomalous unless @xmath324 .
thus , @xmath325.\ ] ] we also prove estimates on @xmath230 , @xmath134 , @xmath249 and @xmath250 in the same manner .
the derivation of estimates on @xmath131 is also similar .
this completes the proof .
we use @xmath326 to denote one of pairs @xmath327 the definition of @xmath328 , @xmath329 , @xmath330 and @xmath331 will be given in the context .
we will show that @xmath326 satisfies a transport - hodge type system which has following schematic form , @xmath332 @xmath333 @xmath334 for @xmath335 , in view of , @xmath336 satisfies .
for @xmath337 , we commute @xmath42 with to derive @xmath338 \tr \chi + \nabla \nabla_4 \tr
\chi = \psi_g\cdot \nabla_4 \tr \chi + \psi \cdot \nabla \tr \chi + \nabla
\nabla_4 \tr \chi\\ & = \psi_g\cdot ( \psi \cdot\psi + \upsilon \cdot\upsilon ) + \psi\cdot \nabla \psi + \nabla ( \psi\cdot \psi + \upsilon \cdot\upsilon).\end{aligned}\ ] ] which is clearly in the form of . for @xmath339 , similarly , we can show that it satisfies and . for @xmath340
, we first define the mass aspect function @xmath341 in view of , we have @xmath342 .
we remark that we must keep tracking the exact coefficient for @xmath153 .
thus , we can derive @xmath343 \eta + \divergence \nabla_4 \eta = \psi \cdot \nabla \eta + \psi_g \cdot \nabla_4 \eta + ( \psi_g + \upsilon\cdot\upsilon)\cdot\eta + \psi_g \cdot \psi \cdot \eta + \divergence \nabla_4 \eta\\ & = \psi \cdot \nabla \psi + \psi \dot ( \psi_g + \upsilon \cdot\upsilon)+\psi_g \cdot\nabla_4 \eta + \divergence \nabla_4 \eta\\ % & = \psi \nabla \psi + \psi(\psi_g + \upsilon\upsilon)+\psi_g ( \psi_g \psi + \psi_g + \upsilon \upsilon ) + \divergence \nabla_4 \eta\\ % & = \psi \nabla \psi + \psi(\psi_g + \upsilon\upsilon)+\psi_g \psi \psi + \divergence \nabla_4 \eta\\ & = \psi\cdot \nabla \psi + \psi\cdot(\psi_g + \upsilon \cdot \upsilon)+\psi_g \cdot\psi\cdot \psi + \divergence ( \psi_g \cdot\psi -\beta + \upsilon\cdot\upsilon).\end{aligned}\ ] ] thus , @xmath344 in view of , we have @xmath345 we also record the exact coefficient for @xmath153 in . adding up , and
, @xmath346 hence , @xmath347 satisfies .
we also combine with to derive , @xmath348 so @xmath349 satisfies . for @xmath350 , we define the mass aspect function @xmath351 .
similarly , we can show that @xmath352 satisfies and . for @xmath353 or @xmath354 , we first define @xmath330 and @xmath331 .
we introduce two auxiliary functions @xmath355 and @xmath356 defined by following transport equations @xmath357 with initial data @xmath358 on @xmath16 and @xmath359 on @xmath360 .
we then define @xmath361 we enlarge the set of connection coefficients by adding two non - anomalous scalars @xmath355 and @xmath356 .
we also extend @xmath362-norms as well as to include them .
it is easy to show that the @xmath308 and @xmath363 estimates hold for @xmath355 and @xmath356 .
since @xmath364 acts as @xmath365 , @xmath366 , in view of , we derive @xmath367 \omegabpair\\ & = \donestar ( \frac{1}{2}\rho + \psi_g \cdot \psi_g + \upsilon \cdot \upsilon , \frac{1}{2}\sigma ) + \psi \cdot \nabla(\omegab , \omegabt ) + \psi_g \cdot(\nabla_4 \omegab + \nabla_4 \omegabt)\notag\\ & = \frac{1}{2}\,\donestar ( \rho , \sigma ) + \psi\cdot(\nabla \psi + \psi_g + \upsilon \cdot \upsilon ) + \psi_g \cdot \psi\cdot \psi + \upsilon\cdot \nabla \upsilon.\end{aligned}\ ] ] we have to keep tracking of the exact coefficient of @xmath368 . in view of
, @xmath369 if we introduce mass aspect function @xmath370 by suitably adding up and , we derive @xmath371 which is clearly of type . by definition
, we also have @xmath372 which is clearly of type .
hence , we showed that @xmath373 satisfies and .
similarly , by using the mass aspect function @xmath374 we can show that @xmath354 satisfies and . for @xmath375
, they satisfy . on @xmath10 , we bound @xmath126 or @xmath131 in @xmath231 norm to derive @xmath376 since @xmath377 s are trivial along @xmath360 , we have @xmath378 for @xmath379 , similarly , we use to derive estimates . the presence of @xmath380 in leads to a bit more complicated estimates , @xmath381 we turn to the hodge system , in view of lemma [ estimates_for_hogde_systems ] , on @xmath10
, we have @xmath382 hence , @xmath383 we combine , and to deduce , @xmath384 thanks to gronwall s inequality , we can remove the integral on the right - hand side .
combined with again , we deduce @xmath385 when @xmath19 is sufficiently small , this yields the following proposition , [ oone_estimates ] if @xmath386 is sufficiently small , for @xmath387 , we have @xmath388 as a byproduct of above argument , we have [ onederivativeonhhb ] we have @xmath389 for @xmath132 , in view of , and , we have @xmath390 .
we commute with @xmath42 to derive @xmath391 integrating this equation along @xmath8 , it is routine to derive @xmath392 combined with gronwall s inequality , similar argument along @xmath311 yields @xmath393 putting those estimates together , we have @xmath394 we move on to the @xmath231 bound on @xmath131 . for @xmath132 , in view of , @xmath395 according to and , @xmath396 thus , @xmath397 is bounded by @xmath175 . for @xmath398 , in view of and , we have @xmath399 to rectify the anomalous behavior of @xmath191 , we have to use localized sobolev estimates , @xmath400 thanks to , we have the desired estimates on @xmath191 . as conclusions , [ prop_l_infty_psi_f ] if @xmath19 is sufficiently small , we have @xmath401 moreover , @xmath402 this is an immediate consequence of null maxwell equations .
we remark that , compared to , we have obtained better bound on @xmath403 . for @xmath375
, we commute angular derivative with to derive @xmath404 + [ \nabla_4 , \nabla ] \theta\\ & = \psi \cdot ( \nabla \psi + \nabla^2 \psi+ \nabla\psi + \nabla \theta + \upsilon\cdot \nabla \upsilon ) + \nabla \psi \cdot ( \nabla \psi + \psi + \upsilon\cdot \upsilon ) \\ & \quad + \psi\cdot(\psi \cdot\nabla \psi + \psi \cdot\theta ) + ( \psi + \upsilon \cdot\upsilon)\cdot\theta + ( \upsilon\cdot \nabla^2 \upsilon + \nabla \upsilon \cdot\nabla \upsilon ) + \psi \cdot\nabla_4 \theta\\ & = t_1 + t_2 + t_3 + t_4 + t_5 + t_6.\end{aligned}\ ] ] similarly , for @xmath379 , we have @xmath405 where the additional term @xmath406 collects all terms involving @xmath407
. we also apply @xmath408 on hodge system to derive @xmath409 therefore , we obtain a transport - hogde systems , @xmath410 @xmath411 @xmath412 we start with which are slightly more difficult than because of the presence of @xmath407 .
estimates on can be derived in a similar way . in
, we replace @xmath377 by @xmath413 and replace @xmath414 by @xmath415 . by hlder s inequality and the estimates derived so far , on @xmath10 we have @xmath416 @xmath417 @xmath418 @xmath419 in view of lemma [ sobolev_trace_estimates ] , on @xmath10 we have @xmath420 thus , @xmath421 we observe that @xmath422 can be estimated by the corollary [ onederivativeonhhb ] .
thus , for @xmath379 , we have @xmath423 similarly , for @xmath375 , we can use to derive @xmath424 we make use of elliptic estimates on on @xmath10 to derive , @xmath425 we plug it in and to derive @xmath426 thus , gronwall inequality yields the following proposition , [ secondangularderivative ] if @xmath19 is sufficiently small , then for @xmath427 and @xmath428 , we have @xmath429 as a corollary of proposition [ secondangularderivative ] and lemma [ sobolev_trace_estimates ] , we have [ oonefour ] if @xmath19 is sufficiently small , then @xmath430 to predict formation of trapped surfaces , we need a slightly refined estimate on @xmath431 .
[ refined estimates on nabla 2 eta ] if @xmath19 is sufficiently small , then @xmath432 recall the following transport - hodge system for @xmath247 and @xmath328 , @xmath433 the first and second equations lead to @xmath434 the third equation leads to @xmath435 when we derive estimates , because each use of hlder s inequality gains @xmath123 , without loss of generality , we ignore most of the nonlinear terms in previous expression .
thus , we have @xmath436 equipped with and , we proceed exactly as we argued in proposition [ secondangularderivative ] .
the most dangerous term from is @xmath437 since @xmath128 can be @xmath150 . in this case , we use corollary [ oonefour ] and @xmath297 norms to save a @xmath259 . since @xmath438 has trivial data on @xmath360 , we can conclude @xmath439 combined with elliptic estimates on , we complete the proof .
for maxwell field , as an immediate consequence of - , we have [ first_null_derivative_maxwell ] if @xmath19 is sufficiently small , then @xmath440 for connection coefficients , we have [ remaininig_first_derivative_estimates ] if @xmath19 is sufficiently small , on @xmath10 we have @xmath441 @xmath442 @xmath443 those terms from and appear also on the left hand side of - . they can be estimated directly with the help of those equations .
to illustrate the idea , we estimate @xmath444 by virtue of , i.e. @xmath445 . therefore ,
@xmath446 the other terms in and can be estimated in the same way .
we omit the details .
is more delicate because @xmath447 and @xmath448 do not appear in the null structure equations . in order to derive estimates
, we have to commute derivatives .
we only estimate @xmath449 .
the remaining terms can be estimated exactly in the same way .
we commute with @xmath44 to derive , @xmath450 \etab + \nabla_4 ( -\frac{1}{2}\tr\chib \cdot ( \etab-\eta)-\chibh \cdot ( \etab -\eta)+\betab-\frac{1}{2}{\sigma_f}\cdot \,^*\!{\alphab_f } + \frac{1}{2}{\alphab_f } \cdot { \rho_f})\\ & = \tr\chib_0 \cdot \nabla_4 \etab + \nabla_4 \betab + e.\end{aligned}\ ] ] one checks easily that error term @xmath451 enjoys a better bound ( notice that @xmath452 in @xmath451 is controlled by proposition [ first_null_derivative_maxwell ] ) .
in fact , we can show @xmath453 .
this bound is good enough for our purpose .
we use to replace @xmath454 by @xmath455 .
it is easy to see that they can also be absorbed into error term @xmath451 , thus we have @xmath456 which implies @xmath457 therefore , @xmath458 finally , we use gronwall s inequality to complete the proof . as a byproduct of the proof , we have [ nabla_3_nabla_4_eta_etab ] if @xmath19 is sufficiently small , then @xmath459 we start with a lemma on commutator estimates which allows us to freely switch the order of differentiations in second derivative estimates .
for example , the order of differentiations in proposition [ 2 der est on max ] does not effect the estimates .
[ est_commutator ] if @xmath19 is sufficiently small , for all connection coefficients @xmath126 and null components of maxwell field @xmath131 , we have @xmath460 \psi , [ \nabla , \nabla_3 ] \psi)\|_{{l^2_{(sc)}(s_{u,\ub } ) } } + \|([\nabla , \nabla_4 ] \upsilon , [ \nabla , \nabla_3 ] \upsilon)\|_{{l^2_{(sc)}(s_{u,\ub } ) } } \lesssim c,\ ] ] @xmath461 \psi , [ \nabla , \nabla_3]\psi ) \|_{l^2_{(sc)}(h_u ) } + \|([\nabla , \nabla_4 ] \psi , [ \nabla , \nabla_3 ] \psi)\|_{l^2_{(sc)}(\hb_{\ub } ) } \lesssim c.\ ] ] @xmath462 \upsilon , [ \nabla , \nabla_3]\upsilon ) \|_{l^2_{(sc)}(h_u ) } + \|([\nabla , \nabla_4 ] \upsilon , [ \nabla , \nabla_3 ] \upsilon)\|_{l^2_{(sc)}(\hb_{\ub } ) } \lesssim c.\ ] ] we prove , while and are simple consequences of .
we use the following schematic formulas ( it is of the same form once we replace @xmath126 by @xmath131 ) , @xmath463 \psi & = \psi \cdot ( \nabla \psi + \psi_g + \upsilon \cdot\upsilon ) + \psi_g \cdot \nabla_4 \psi , \\ [ \nabla_3 , \nabla ] \psi & = \tr\chib_0 \cdot \nabla \psi + \psi \cdot ( \nabla \psi + \psi_g + \upsilon \cdot \upsilon ) + \psi_g \cdot \nabla_3 \psi,\end{aligned}\ ] ] these formulas allows one to use previously obtained @xmath464 estimates to bound the commutators .
we observe that possible anomalies are due to the presence of @xmath465 and @xmath466 at the right - hand side of the above formulas .
they may cause a loss of @xmath152 .
but hlder s inequality gains @xmath123 which compensates this loss .
this completes the proof .
we turn to the estimates on second derivatives of maxwell field .
[ 2 der est on max ] if @xmath19 is sufficiently small , we have @xmath467 @xmath468 the proof is straightforward .
we first apply @xmath42 on - .
the angular derivative cleans up all linear anomalies .
thus we can use the estimates derived so far to complete the estimates .
we omit the details .
next proposition collects remaining second derivative estimates on connection coefficients .
[ twoderivativericcicoefficients ] if @xmath19 is sufficiently small , then @xmath469 @xmath470 @xmath471 @xmath472 except for @xmath473 and @xmath474 , all the terms can be estimated directly from - . for @xmath475 ,
the corresponding null structure equations in @xmath44 direction read as @xmath476 .
thus , @xmath477 every term on the right - hand side is bounded along @xmath8 .
thus , @xmath478 similarly , along @xmath311 , we have @xmath479 since the norm @xmath480 does not include @xmath155 , @xmath481 is absent in .
we can proceed in the same way to prove the those estimates in @xmath43 directions and we omit the details .
it remains to control the most difficult terms @xmath473 and @xmath474 .
since the arguments are similar for both , we only derive estimates on @xmath473 .
recall the transport - hodge system for @xmath482 , @xmath483 @xmath484 we observe that @xmath249 and @xmath246 do not appear among @xmath485 s .
this is extremely important since we do not have estimates on @xmath486 and @xmath487 .
another observation is equally important : there is no double anomalous terms in the form @xmath488 , @xmath489 or @xmath490 .
we derive a transport - hodge system for @xmath491 . in view of
, we derive @xmath492 where @xmath493 .
we have replaced @xmath407 and @xmath494 by the constant 1 . without loss of generality
, we can drop some terms which enjoy easier and better estimates . thus above equation
is reduced to the following form , @xmath495 commuting @xmath44 with , we also have , @xmath496 we control @xmath497 . for @xmath498 , in view of , we have @xmath499 for @xmath500 , we have @xmath501 for @xmath502 , in view of , we also have @xmath503 . for @xmath504 ( we come back to @xmath505 later ) , it consists of four terms .
the one involving maxwell field is easy since it is a cubic nonlinearity .
each of the rest three terms can be estimated by virtue of sobolev inequality .
we only treat one of them to illustrate the idea .
the others can be derived in the same manner .
first of all , we have @xmath506 in view of proposition [ remaininig_first_derivative_estimates ] and lemma [ sobolev_trace_estimates ] , on @xmath10 we have @xmath507 @xmath508 thus , @xmath509 recall that in proposition [ twoderivativericcicoefficients ] we have just proved @xmath510 . since @xmath249 and @xmath246 are absent in and , we have @xmath511 .
hence , @xmath512 . for @xmath505
, we have to treat different components in different ways , @xmath513 hence , @xmath514 . for @xmath515 , it has two terms .
we estimate them one by one . according to null bianchi equations , @xmath516
, we derive @xmath517 thus , @xmath518 . putting all the @xmath519 s together
, we have @xmath520 thus , @xmath521 thanks to gronwall s inequality , we have @xmath522 we use elliptic estimates on hodge system on @xmath10 to derive @xmath523 combined with , we have @xmath524 thanks to gronwall s inequality , we have @xmath525 .
we now go back to again to derive @xmath526 this completes the proof . combining all estimates derived so far , we close the bootstrap argument for * theorem a * by showing next proposition . [ close_bootstrap ]
if @xmath19 is sufficiently small , we have @xmath527 for @xmath528 , in view of lemma [ sobolev_trace_estimates ] and proposition [ twoderivativericcicoefficients ] , along @xmath222 , @xmath529 for @xmath530 , similarly , we have @xmath531 .
thus , for a non - anomalous @xmath127 , @xmath532 for an anomalous @xmath533 , we use and the localized sobolev inequality to derive @xmath534 this completes the proof . in the rest of the section ,
we derive more @xmath297 estimates for later use . for curvatures ,
if @xmath19 is sufficiently small , we have @xmath535 for curvature component @xmath536 , in view of lemma [ sobolev_trace_estimates ] , we have @xmath537 to bound @xmath538 , we use , or to replace @xmath539 by @xmath540
. those terms can be easily bounded .
this yields the desired bound .
similarly , we can bound @xmath153 .
the estimates on @xmath317 is directly from .
it remains to control @xmath150 .
@xmath541 in view of , @xmath542 . in view of the anomalous estimates on @xmath150
, this gives the bound . for connection coefficients
, we have if @xmath19 is sufficiently small , we have @xmath543 @xmath544 @xmath545 in view of the proof of proposition [ close_bootstrap ] , we have estimates on @xmath546 ; for the remaining terms in and , they can be directly derived from null structure equations .
schematically , for @xmath547 or @xmath36 , they satisfy @xmath548 each term on the right - hand side can be bounded by the estimates derived so far .
the potential danger comes from @xmath549 .
when @xmath533 , it leads to anomalous behavior in . it remains to deal with @xmath550 and @xmath551 .
they can be estimated directly from corollary [ nabla_3_nabla_4_eta_etab ] , proposition [ twoderivativericcicoefficients ] and lemma [ sobolev_trace_estimates ] .
this completes the proof .
all the estimates derived in the bootstrap argument of * theorem a * will be valid throughout the paper . notice that we do not have the @xmath552 estimates on @xmath553 and @xmath448 .
we recall the definitions of trace norms introduced in @xcite . for a tensor field @xmath126 along @xmath554 , relative to the transported
coordinates @xmath555 , its scale invariant trace norm is defined as @xmath556 for a tensor field @xmath126 along @xmath144 , relative to the transported coordinates @xmath557 , its scale invariant trace norm is defined as @xmath558 we recall the sharp trace theorem below .
we refer to @xcite for a proof .
[ trace theorem ] if @xmath19 is sufficiently small , for any @xmath126 along @xmath554 , we have @xmath559 and for any @xmath126 along @xmath144 , we have @xmath560 if @xmath19 is sufficiently small , for @xmath561 , we have @xmath562 we first derive estimates on @xmath563 which relies on .
to avoid @xmath246 , we rewrite in following form @xmath564 where the modified term @xmath565 enjoys the same estimates as @xmath230 .
we can ignore @xmath566 since it enjoys better trace estimates . in view of proposition
[ trace theorem ] , we have @xmath567 to estimate @xmath568 , we differentiate to derive @xmath569 thus , @xmath570 this proves the desired estimates on @xmath571 .
we now derive estimates on @xmath572 .
it relies on .
the proof goes exactly as above .
similarly , we can derive estimates for other quantities .
this completes the proof .
we now turn to the trace estimates for maxwell components .
if @xmath19 is sufficiently small , we have @xmath573 we derive estimates for @xmath574 .
the bound for @xmath191 is similar so that we omit the details .
we start with @xmath575 . according to proposition [ trace theorem ] , we have @xmath576 it suffices to show that @xmath577 .
the idea is , by integration by parts , we can move bad derivatives to one component to save a @xmath259 . in fact , @xmath578 where error term @xmath451 comes from boundary terms of integration by parts and commutator of @xmath579 $ ] .
it is easy to see that in this form , we have @xmath580 . this yields the desired estimates . for the bound of @xmath581
, we introduce an auxiliary tensor @xmath582 @xmath583 by commuting derivatives , together with gronwall s inequality and all the estimates derived so far , one can easily show that @xmath584 we claim a more serious bound on @xmath582 , @xmath585 we use @xmath586 to denote terms which either are much easier to estimate or enjoy better estimates .
we commute @xmath587 with to derive @xmath588 we apply @xmath589 on and to replace @xmath590 by @xmath591 .
this allows us to do a renormalization to rectify @xmath592 as follows , @xmath593 hence , we derive @xmath594 this equation allows one to derive estimates on @xmath595 hence on @xmath596 .
finally , we use proposition [ trace theorem ] and lemma [ sobolev_trace_estimates ] to control @xmath597 . in view of the definition of @xmath582
, this completes the proof .
similarly , we can easily derive that , for @xmath598 = @xmath599 , @xmath600 , @xmath601 , @xmath602 , @xmath603 , @xmath604 , @xmath605 or @xmath606 , we have @xmath607 in order to control @xmath608 and @xmath609 , we introduce two auxiliary tensors @xmath100 and @xmath610 , @xmath611 [ trace_estimates_chih_chibh ] if @xmath19 is sufficiently small , we have @xmath612 @xmath613 @xmath614 @xmath615 [ remark nabla chih chihb trace ] as consequences of lemma [ sobolev_trace_estimates ] and proposition [ trace theorem ] , we have @xmath616 @xmath617 because of the presence of @xmath407 , we expect the estimates along incoming null hypersurfaces are more challenging .
we will only prove the second part of the proposition .
we observe that is a direct use of the estimates derived through the proof of * theorem a*. we turn to the proof of .
let @xmath618 . in particular ,
@xmath249 and @xmath246 are prohibited .
we commute derivatives to derive @xmath619 \phib = \chib \cdot \nabla\phib + \nabla \psi_3 \cdot \phib + \psi_3 \cdot \nabla_3\phib + \upsilon \cdot \upsilon \cdot \phib + \chib \cdot \psi_3 \cdot \phib,\ ] ] @xmath620 \phib & = \chib \cdot \nabla^2 \phib + \psi_3 \cdot ( \nabla_3 \nabla \phib + \nabla \nabla_3 \phib ) + \nabla \psi_3 \cdot ( \nabla \phib + \nabla_3 \phib ) + \nabla^2 \psi_3 \cdot \phib\\ & \quad + \upsilon \cdot \upsilon \cdot \nabla \phib + \upsilon \cdot \nabla \upsilon \cdot \phib + \chib \cdot \psi_3 \cdot \nabla \phib + \psi_3 \cdot \nabla \psi_3 \cdot \phib + \chib \cdot \nabla \psi_3 \cdot \phib.\end{aligned}\ ] ] hence , by taking trace , we have @xmath621 \phib & = \tr \chib_0 \cdot \nabla^2 \phib + \chibh \cdot \nabla^2 \phib + e_1.\end{aligned}\ ] ] we claim that error term @xmath622 satisfies @xmath623 .
the idea is to use either @xmath297 estimates on @xmath624 or @xmath231 estimates on @xmath610 . in view of lemma [ sobolev_trace_estimates ] , in either case
the quantity is bounded by @xmath625 .
we only deal with one term in @xmath622 , say @xmath626 , to illustrate the idea .
@xmath627 thus , @xmath628 \chih + e_1\end{aligned}\ ] ] since , on @xmath311 , @xmath629 \chih \|_{l^2_{(sc ) } } = \|k \nabla \phib + \nabla k \cdot \phib \|_{l^2_{(sc)}}\lesssim \delta^{\frac{1}{2}}(\|k\|_{l^4_{(sc)}}\|\phib\|_{l^4_{(sc)}}+\|\nabla k\|_{l^2_{(sc)}}\|\phib\|_{l^{\infty}_{(sc)}})\lesssim c,\ ] ] we derive @xmath630 with error term @xmath631 satisfying @xmath632 .
we then apply @xmath589 to to derive @xmath633 we differentiate the equation once more to derive @xmath634 + k \ , \nabla \chibh\\ & = \nabla^2 \betab + \nabla^3 \tr\chib + k \cdot \nabla_3 \phib + e_3,\end{aligned}\ ] ] with error term @xmath635 .
combined with , we have @xmath636 for @xmath637 , in view of , we transform it into a term involving @xmath43 derivative , @xmath638 with error term @xmath639 . combined with
, we have @xmath640 where @xmath641 .
since @xmath642 can be easily bounded according to , we have @xmath643 putting things together , we have derived the renormalized equation @xmath644 this formula allows us to derive @xmath645 thus , on @xmath221 and @xmath646 , we have @xmath647 standard elliptic estimates imply @xmath648 hence , @xmath649 we remove the last term by gronwall s inequality to derive @xmath650 we then integrate on @xmath311 and use the smallness of @xmath19 to complete the proof . if @xmath19 is sufficiently small , we have @xmath653 as a consequence , we have @xmath654 we bound the second term of .
the first one follows in the same manner . to avoid unpleasant derivatives on @xmath246
, we define @xmath655 and use the following form of , @xmath656 by a direct but tedious commutation of derivative , we can show that @xmath657 \tr\chib ' & = \chib \cdot \nabla^3 \tr \chib ' + \psi_3 \cdot ( \nabla_3 \nabla^2 \tr \chib ' + \nabla \nabla_3 \nabla \tr \chib ' + \nabla^2 \nabla_3 \tr \chib ' ) \\ & + [ ( \nabla_3 \psi_3 + \nabla \chib + \psi_3 \cdot \psi_3 + \upsilon \cdot \upsilon ) \cdot
\nabla^2 \tr \chib ' + \nabla \psi_3 \cdot ( \nabla_3 \nabla \tr \chib ' + \nabla \nabla_3 \tr \chib ' ) ] \\ & + [ ( \nabla^2 \psi_3 + \psi_3 \cdot \nabla \psi_3 + \upsilon \cdot \nabla \upsilon ) \cdot \nabla \tr \chib ' + \nabla^2 \psi_3 \cdot \nabla_3 \tr \chib']\\ & = \chib \cdot \nabla^3 \tr \chib ' + f_1.\end{aligned}\ ] ] we claim that error term @xmath658 enjoys @xmath659 to prove , we first observe that , since @xmath660 , we can simply ignore all @xmath27 s in estimates .
the second observation is the following sobolev estimates , @xmath661 the bound for @xmath658 is immediate by combining those two observations and the estimates derived so far .
hence , we derive @xmath662 with error term @xmath663 .
we turn to the renormalization of terms with three angular derivatives . for @xmath664 , in view of and , we have @xmath665 . after using @xmath364 and commuting derivatives , this equation is reduced to @xmath666 we differentiate the equation once more to derive @xmath667 with error term @xmath668 . in view of in the previous subsection
, we have @xmath669 hence , @xmath670 with error term @xmath671 . in the derivation of ,
we have ignored @xmath672 since it will be eventually eliminated by the gronwall s inequality .
we the estimate @xmath519 s along @xmath311 one by one . for @xmath498 , in view of proposition [ trace_estimates_chih_chibh ] , we have @xmath673 in view of , we have @xmath674 .
thus , we have @xmath675 for @xmath500 , in view of remark [ remark nabla chih chihb trace ] , we have @xmath676 for @xmath502 , we use two more derivative to bound @xmath677 , thus , @xmath678 for @xmath505 , we have @xmath679 for @xmath504 , we have @xmath680 thus implies @xmath681 thus , we have @xmath682 .
we then integrate on @xmath311 and use the smallness of @xmath19 to complete the proof . in order to obtain trace estimates on @xmath685 and @xmath686 , we introduce the following auxiliary tensors .
@xmath687 we also define @xmath688 and @xmath689 .
they can be estimated as follows .
if @xmath19 is sufficiently small , we have @xmath690 @xmath691 @xmath692 @xmath693 as consequences of lemma and proposition [ trace theorem ] , we have @xmath694 @xmath695 and are extremely easy to prove by using the estimates derived so far . for the remaining estimates , we shall prove ; similarly , one can derive .
first of all , we commute @xmath587 with to derive @xmath696 with error term @xmath697 .
secondly , we commute @xmath364 with the following hodge system @xmath698 to derive @xmath699 with error term @xmath700 .
similarly , we can derive @xmath701 with error term @xmath702 .
putting things together , we have @xmath703 with error term @xmath704 .
recall the following transport equation for @xmath705 and we keep the exact coefficients of @xmath706 and @xmath707 .
@xmath708 by commuting derivatives , we have @xmath709 with error term @xmath710 .
putting , and together , we have the renormalized equation @xmath711 with error term @xmath712 .
now we can combine gronwall s inequality , lemma [ sobolev_trace_estimates ] and the estimates derived so far to prove .
since this is standard , we omit the details .
if @xmath19 is sufficiently small , we have @xmath715 @xmath716 is a direct consequence of .
we only derive estimates on @xmath714 ; the other one can be derived in the same manner .
we start to derive a transport equation on @xmath717 .
we can not afford to any loss of information for highest derivative terms , since those terms are potentially dangerous .
they can only be handled through a renormalization process . commuting derivative with ,
we have @xmath718 we can use to eliminate @xmath719 to derive @xmath720 added up with and ignoring some irrelevant coefficients , it is reduced to @xmath721+(\nabla \eta + \nabla \etab)\cdot \chib + l.o.t.\end{aligned}\ ] ] for the bracket , we use the key fact @xmath722 of the energy - momentum tensor to derive @xmath723 thus , we derive @xmath724 thanks to , and , we have @xmath725 we replace @xmath563 and @xmath706 ( but keep @xmath572 ) by the right - hand side of in to derive @xmath726 in view of , we can regard @xmath727 as @xmath586 thus , we have @xmath728 where @xmath729 commuting again with @xmath587 , we derive @xmath730 with error term @xmath731 .
we now use to replace @xmath732 , thus @xmath733 & = ( \nabla^2\eta + \nabla^2 \etab)\cdot \betab + ( \eta + \etab)\cdot \nabla^3 \tr \chib + ( \nabla^2\eta + \nabla^2 \etab ) \cdot ( \nabla \tr\chib + \nabla \chibh)\\ & \quad + ( \nabla \triangle \eta + \nabla \triangle \etab)\cdot \chib + ( \nabla \eta + \nabla \etab ) \cdot ( \triangle \chibh + \nabla^2 \tr\chib ) + h_2,\end{aligned}\ ] ] with error term @xmath734 .
we still need to renormalize the terms involving @xmath735 , @xmath736 and @xmath737 . in view of and , if we define @xmath738 , we then have @xmath739 with error term @xmath740 .
we estimate @xmath519 s one by one . for @xmath498
, we have @xmath741 for @xmath500 , according to section [ l_2_estimates_for_nabla^3_tr_chi ] , we have @xmath742 for @xmath502 , thanks again to section [ l_2_estimates_for_nabla^3_tr_chi ] , we have @xmath743 for @xmath505 , in view of or , we have @xmath744 for @xmath504 , in view of , we replace @xmath745 by @xmath563 to derive @xmath746 for @xmath515 , in view of the estimates section [ tr_estimates_for_nabla_eta_etab ] , we have @xmath747 putting things together , we have @xmath748 last term can be removed by gronwall s inequality .
we then perform an integration on the incoming null hypersurfaces to complete the proof .
we now show the trace estimates on @xmath706 and @xmath707 . if @xmath19 is sufficiently small , we have @xmath749 we prove estimates on incoming null cones @xmath750 and the rest follows in the same way .
for @xmath751 , in view of , we have @xmath752 for @xmath753 , we first define @xmath754 and rewrite as @xmath755 we only keep the principal term @xmath756 since nonlinear terms enjoy better estimates .
thus , @xmath757 to estimate @xmath758 , we differentiate to derive @xmath759 thanks to , we have estimates on @xmath760 ; for those nonlinear terms , they can be easily estimated by the estimates derived so far .
hence , we complete the proof .
for @xmath761 , @xmath762 on @xmath763 denote the generators of @xmath764 satisfying the relation @xmath765 = \epsilon_{ijk}\,^{(k ) } \!o.\ ] ] in what follows , we shall suppress the index @xmath766 .
we extend @xmath767 to the entire domain @xmath3 by @xmath768 the deformation tensor @xmath769 of @xmath770 is defined by @xmath771 .
we also define @xmath772 and @xmath773 and we can associate signatures to them , @xmath774 we list the null components of @xmath775 , @xmath776 we observe that both @xmath777 and @xmath778 ( but not @xmath779 ! ) vanish on @xmath360 .
the goal of the current section is to derive estimates on @xmath775 . by virtue of the definitions of @xmath779 and @xmath778 , as well as , , and , we derive transport equations for @xmath779 and @xmath778 @xmath780 @xmath781 we observe that @xmath782 and @xmath246 do not appear in and . if @xmath19 is sufficiently small , we have @xmath783 we integrate to derive @xmath784 since @xmath785 and @xmath786 , we have @xmath787 .
this yields the estimate on @xmath779 .
the @xmath156 estimate for @xmath778 is similar in view of the key fact that @xmath778 vanishes on @xmath360 .
the @xmath788 estimates follows again directly from and and the triviality of the initial data of @xmath777 and @xmath778 .
we remark that @xmath788 estimate for @xmath779 is anomalous because @xmath789 .
we turn to one derivative of @xmath779 and @xmath778 .
commuting @xmath42 with and , we derive @xmath790 @xmath791 we can easily show that the error terms can be bounded as follows , @xmath792 if @xmath19 is sufficiently small , we have @xmath793 @xmath794 besides @xmath622 and @xmath631 , there are six terms appearing in and . for @xmath795 and @xmath796 ,
they can be absorbed eventually by gronwall s inequality .
we now indicate how to renormalize the rest four terms .
the details of the proof are routine . for @xmath797 , we split it as @xmath798 .
thanks to @xmath799 estimates on @xmath800 , we have @xmath801 to control @xmath802 , we use to derive @xmath803 where @xmath804 is much easier to control .
up to easier error terms , we use and to replace @xmath805 by @xmath806 . in this way , we can easily renormalize the system by moving @xmath807 to the left hand side of the equation . hence , we can derive estimates easily .
we also observe that @xmath808 can also be treated in this way . for @xmath809
, we renormalize it via transport equation @xmath810 . for @xmath811 , recall we defined @xmath812 in section [ tr_estimates_for_nabla_eta_etab ] , thus @xmath813 we then eliminate @xmath814 via the standard renormalization .
the other terms are easy to control thanks to the estimates on @xmath582 in section [ tr_estimates_for_nabla_eta_etab ] .
this completes the proof .
we commute @xmath43 with and to derive @xmath815 + ( \beta + \upsilon\cdot \upsilon)\cdot ( z + \chib \cdot o)\\ & \quad+(\nabla_3 \beta + \nabla_3 r_{b4})\cdot
o + \nabla_3 \chih \cdot z,\\ \nabla_4 \nabla_3 z & = ( \omega + \chi ) \cdot z + \omegab \cdot \nabla_4 z + ( \eta + \etab)\cdot \nabla_3 h + \nabla_3 \sigma \cdot o + \nabla_3 \chih \cdot z\\ & \quad + \nabla_3 ( \eta + \etab ) \cdot h + [ \sigma + ( \eta + \etab)\cdot(1 + \eta + \etab ) + \nabla(\eta + \etab ) + \nabla_3 ( \omega + \tr\chi)]\cdot z\\ & \quad + [ \nabla_3 \nabla ( \eta + \etab ) + \chib \cdot \nabla ( \eta + \etab ) + ( \eta + \etab)\cdot ( \nabla_3 ( \eta + \etab)+ ( \eta + \etab)\cdot \chib)+\chib \cdot\sigma ] \cdot o.\end{aligned}\ ] ] if @xmath19 is sufficiently small , we have @xmath816 we only derive estimates on @xmath817 and @xmath818 can be bounded exactly in the same manner .
we observe that the possible dangerous terms are @xmath819 , @xmath820 and @xmath821 .
the rest are much easier to control . for @xmath821 ,
@xmath778 is _ not _ anomalous , thus hlder s inequality gains an extra @xmath123 to compensate the anomaly of @xmath822 ; for @xmath819 , we use to convert the @xmath823 to angular derivative of curvature components ; for @xmath820 , we have @xmath824 which is still easy to estimate .
this completes the proof .
we have a slightly refined estimates on @xmath825 , @xmath826 } = \| \sup_{\ub ' \in [ 0,\ub]}|\nabla_3 z ( u , \ub')|\|_{l^{2}_{(sc)}(u,\ub)}\lesssim c.\ ] ] to prove this estimate , we observe that @xmath827}|\nabla_3 z ( u , \ub')| ] \leq |\nabla_4 \nabla_3 z ( u , \ub)|.\ ] ] thus , we can still use the transport equation for @xmath825 to derive .
now we state the main proposition of this section . if @xmath19 is sufficiently small , we have
@xmath828 we only sketch the idea .
the details can be easily carried out by the estimates derived so far .
if one computes the null component of @xmath829 one by one , we see immediately that @xmath779 comes either as @xmath830 or as @xmath831 .
we pay attention to @xmath779 since it does not enjoy non - anomalous @xmath788 estimates .
nevertheless , we still can proceed as follows , @xmath832 for @xmath833 , since @xmath777 has trivial data on the initial surfaces , this allows us to derive non - anomalous estimates on @xmath777 .
we summarize the estimates in this section in the following , [ theorem b]if @xmath19 is sufficiently small , we have @xmath834 and for all the components of @xmath835 except @xmath836 we have @xmath837 moreover , @xmath838 } \lesssim c.\ ] ]
we first take @xmath843 in to derive @xmath844.\end{aligned}\ ] ] in the scale invariant norms , we have @xmath845.\end{aligned}\ ] ] the integrand of the last term can be written schematically as @xmath846 with @xmath847 , where @xmath848 indicates the signature for the corresponding component . since latexmath:[\ ] ] | the purpose of the paper is to understand a mechanism of evolutionary formation of trapped surfaces when there is an electromagnetic field coupled to the background space - time .
based on the _ short pulse _ ansatz , on a given finite outgoing null hypersurface which is free of trapped surfaces , we exhibit an open set of initial data @xmath0 for einstein equations coupled with a maxwell field , so that a trapped surface forms along the einstein - maxwell flow . on one hand , this generalizes the black - hole - formation results of christodoulou @xcite and klainerman - rodnianski @xcite .
in fact , by switching off the electromagnetic field in our main theorem , we can retrieve their results in vacuum . on the other hand , this shows that the formation of black hole can be purely due to the condensation of maxwell field on the initial null hypersurface where there is no incoming gravitational energy . |
degradation of quantum superposed state by decoherence is an obstacle to quantum information processing . in order to proactively prevent errors ,
a multipulse control method has been proposed @xcite .
it is essential that the application of @xmath0 pulses causes time reversal in order to suppress the decoherence .
the method has attracted considerable attention ; it has been applied to suppress unwanted spontaneous emission@xcite , the magnetic - state decoherence by collisions in a vapor@xcite and the damping of vibrational mode of a chain of trapped ions@xcite . while the multipulse control method requires no ancillary bits and no accurate detection ,
its effectiveness has been shown when sufficiently short and strong pulses are periodically applied in a shorter interval than the characteristic time of the system - reservoir interaction@xcite .
the degree of suppression becomes larger as the pulse interval becomes shorter .
since these conditions are not easy to execute , a new approach to use a continuous control field instead of pulses has been proposed @xcite .
it is also shown that the control pulses do not always have to be ultra short for systems coupled to the reservoir with @xmath1 spectral density@xcite . in order to overcome the strict condition on the pulse application , it is desirable to seek a possibility to use the pulse trains with relatively long pulse interval . in previous paper@xcite , we have suggested a possibility to relieve the condition on pulse interval by formulating a theory of pulse control on the pure dephasing phenomena that is caused by the interaction with a boson reservoir . since the ordinary spin - boson model where a spin linearly interacts with the boson reservoir can not describe the irreversibility in the long time region except for the ohmic dissipation case
, we have extended the model to include a nonlinear interaction .
we have found that the multipulse control is effective for this model when the pulse interval is shorter than the reservoir correlation time .
we also found that the effective pure dephasing time shows a non - monotonic dependence on the pulse interval , that is , it has a peak when an application of @xmath0 pulse - train is synchronized with the oscillation of the reservoir .
this means that the pure dephasing phenomenon is also effectively suppressed by paying attention to the dynamics of the reservoir .
however , in many cases , the system - reservoir interaction is described with a strong linear interaction and a weak nonlinear interaction . in the time region where we want to discuss the effectiveness of suppression by synchronizing the pulse application and the dynamics of the reservoir , the linear interaction plays an essential role in the decay . in this paper , assuming that a two - level system linearly interacts with a boson reservoir that has a characteristic frequency , we discuss the effectiveness of the synchronization of a @xmath0 pulse train with the oscillation of the reservoir . for convenience
, we name this strategy as synchronized pulse control(spc ) in the following discussion .
as recognized in @xcite , spc also depends on the type of coupling(bath ) spectral density . in order to make clear the applicability of spc
, we study the effectiveness of spc on non - lorentzian and lorentzian coupling spectral density .
the outline of this paper is as follows : in sec.[sec:2 ] , we introduce the model of the decoherence and derive the basic formula for multipulse control on the linear spin - boson model .
next , we discuss the synchronized pulse control in sec.[sec:3 ] : the application of the basic formula to non - lorentzian ( lorentzian ) coupling spectral density is written in sec.[sec:3]a(b ) , respectively .
after discussing the effectiveness of the spc in sec.[sec:4 ] , we give concluding remarks in sec.[sec:5 ] .
we consider a two - level system composed of an excited state @xmath2 and a ground state @xmath3 with energy @xmath4 .
let us consider the decoherence of this two - level system , which is caused by a linear interaction between the excited state and a boson reservoir .
the system hamiltonian reads , @xmath5 in order to suppress the decoherence , we apply pulses that are sufficiently short and strong .
this indicates that the interaction with the reservoir is neglected during pulse application : @xmath6 where @xmath7 is the @xmath8-th applied pulse of external field .
we assume the pulse to be on resonance with the two - level system , which means @xmath9 .
when we apply @xmath10 pulses with a pulse interval @xmath11 and pulse duration @xmath12 , the time evolution of the density operator @xmath13 of the total system , is given by @xmath14 \nonumber \\ & & \hspace{-2 cm } \times \{\prod_{j=0}^{n-1 } e^{-i l_{r } ( \tau_{s}-\delta t ) } t_{+ } [ e^{-i \int _ { j \tau_{s } } ^{j \tau_{s } + \delta t } dt ' l_{p , j } ( t')}]\ } \rho(0 ) , \nonumber \\
\label{eqn:7}\end{aligned}\ ] ] where @xmath15 is the time ordering symbol from right to left and @xmath16 ( @xmath17 ) indicates the liouville operator during the @xmath8-th pulse ( the interaction with the reservoir ) which is defined as @xmath18 , \;\ ; ( \nu=\{p , j\ } \ ; or \ ; \{r\ } ) .
\label{eqn:8}\ ] ] we rewrite eq.([eqn:7 ] ) by using the following relation for an arbitrary operator @xmath19 , @xmath20 which consists of the operators as @xmath21 \nonumber \\ & = & ( u_{1}(t - t_0 ) \ee + u_{2}(t - t_0 ) \gg ) \label{eqn:10}\\ t_{+ } [ e^{-\frac{i}{\hbar } \int_{t_0}^{t } dt ' h_{p , j}(t ) } ] & = & e^{- \frac{i}{\hbar } h_{0 } ( t - t_0 ) } e^{- \frac{i}{\hbar } { \hat h_{p , j } } ( t - t_0 ) } \nonumber \\ & = & u_{2}(t - t_0 ) \cos(\frac{\theta_{j}}{2})(e^{-i \omega ( t - t_0 ) } \ee+\gg ) \nonumber \\ & & -i \sin(\frac{\theta_{j}}{2})(\eg+\ge ) , \label{eqn:11}\end{aligned}\ ] ] where @xmath22 here we have assumed the each applied pulse to be square whose strength is @xmath23 , which gives pulse area @xmath24 for the @xmath8-th pulse . in eq.([eqn:11 ] ) , we used the following definitions as , @xmath25 \;t_{+}[\exp[-i \int_{0}^{t } dt ' \sum_{k } h_{k } \epsilon_{k } ( b_{k } e^{-i\epsilon_{k } t ' } + b_{k}^{\dagger}e^{i\epsilon_{k } t ' } ) ] ] \nonumber \\ & = & \eta(t)\exp[-i(\omega + \sum_{k } \epsilon_{k } b_{k}^{\dagger}b_{k})t ] \exp[\sum_{k } \epsilon_{k } ( b_{k}^{\dagger } \xi_{k } ( t ) - b_{k } \xi_{k}^ { * } ( t ) ) ] , \label{eqn:12 } \\
u_{2}(t ) & = & \exp[-i \sum_{k } \epsilon_{k } b_{k}^{\dagger}b_{k } t ] , \label{eqn:13}\end{aligned}\ ] ] where @xmath26 , \;\ ; \xi_{k } ( t ) = \frac{h_{k}}{\epsilon_{k } } ( 1-e^{i \epsilon_{k } t } ) . \label{eqn:14}\ ] ] now we suppose the pulse area @xmath27 to be @xmath0 except for the first pulse whose pulse area is @xmath28 to generate a superposed two - level state at an initial time ( @xmath29 ) .
defining the intensity of off diagonal element of the density operator @xmath13 as @xmath30 where @xmath31 denotes the operation to trace over the reservoir variable , we obtain for even @xmath10 , @xmath32 |^2 , \label{eqn:15}\ ] ] and for odd @xmath10 , @xmath33 |^2 .
\label{eqn:16}\ ] ] here we have assumed the pulse duration @xmath12 to be infinitely small . in the case of the pulsed magnetic resonance or the transient nonlinear optics
, @xmath34 indicates the signal intensity .
next , we focus on the time evolution of the boson reservoir , eliminating the two - level system that periodically changes its state between @xmath2 and @xmath3 by the @xmath0 pulse train . denoting the displacement operator as @xmath35 , \label{eqn:18}\ ] ] where @xmath36 means a set of bosons in the reservoir , we obtain the off diagonal element of the density operator @xmath13 in the form for even @xmath10 , @xmath37|g\rangle|^2 = |\langle b_{n}(t)|a_{n}(t ) \rangle|^2,\label{eqn:19}\ ] ] with @xmath38 here we defined @xmath39 and @xmath40 as @xmath41 for odd @xmath10 , we obtain , @xmath42|g\rangle|^2 = |\langle \{\alpha_{n , k } ( t)\ } | \{\beta_{n , k } ( t)\ } \rangle |^2,\label{eqn:22}\ ] ] with @xmath43 in obtaining eqs.([eqn:19 ] ) @xmath44 ( [ eqn:23 ] ) , we assume that the boson reservoir is in the vacuum state and the two - level system is in the ground state at the initial time : @xmath45 eqs .
( [ eqn:19 ] ) and ( [ eqn:22 ] ) imply that the intensity @xmath34 is described with overlap between coherent states @xmath46 and @xmath47 of the reservoir .
these coherent states are biuniquely associated to @xmath48 and @xmath49 : for even @xmath10 , the state @xmath46 is associated to @xmath49 , whereas @xmath47 is with @xmath48 .
the @xmath0-pulse application alternately exchange the association between the reservoir states and the two - level system .
an actual evaluation of eqs.([eqn:19 ] ) @xmath44 ( [ eqn:23 ] ) requires us to rewrite the summation over @xmath50 into the energy integral , @xmath51 where we have defined coupling spectral density @xmath52 as , @xmath53
now we evaluate the time evolution of the intensity @xmath34 @xmath54 .
\label{eqn:27}\ ] ] in the following , we discuss the time dependence of @xmath34 for non - lorentzian and lorentzian coupling spectral density . while the spc can be effective for the former case , it is ineffective for the latter case . as the first example of the non - lorentzian coupling spectral density
, we consider a gaussian distribution with the mean frequency @xmath55 and the variance @xmath56 , @xmath57 setting @xmath58 in eq.([eqn:19 ] ) , we evaluate time evolution of @xmath34 after a single @xmath28 pulse at @xmath29 , which is shown in fig.[fig : fg1 ] . here and henceforth , we have used a scaled time variable as @xmath59 and set the parameters as @xmath60 , @xmath61 , which mean that the decay time of the interaction mode is relatively long , the average number of boson which interact with the spin is 3 .
we see a damped oscillation whose period is @xmath62 . without pulse control for @xmath63 , @xmath61 . ]
the dynamical decoupling method@xcite tells us that an application of @xmath0 pulse - train is sufficiently effective when the pulses are applied in a sufficiently small " interval .
when the spectral density has a cutoff frequency at @xmath64 as assumed in @xcite , the pulse interval @xmath11 is required to be much smaller than @xmath65 in order to control the decoherence . in the case of the gaussian distribution as eq.([eqn:28 ] ) with relatively small variance , we suppose that the pulse interval is required to be much smaller than @xmath66 for decoherence control . in fig.[fig : fg2 ] , we can see that the decay of @xmath67 is well suppressed for @xmath68 . for the pulse interval @xmath69 .
other parameters are same as in fig.[fig : fg1 ] . ] however , when the pulse interval becomes longer , we find that the pulse application makes things even worse than the damped oscillation in fig.[fig : fg1 ] .
this is shown in fig.[fig : fg3 ] where the pulse interval is @xmath70 .
for the pulse interval @xmath71 .
other parameters are same as in fig.[fig : fg1 ] . ]
now we plot the case where @xmath0 pulses are applied with the interval @xmath72 in fig.[fig : fg4 ] , where we find that the phase coherence recovers at the pulse application time .
the peak value asymptotically goes to be constant , which reflects that the dephasing in long time region can not be described by the linear interaction .
when we obtain a recovery of the intensity by synchronizing the pulse application with the characteristic period @xmath73 , we call this strategy for suppression of decoherence as synchronized pulse control . in this paper , we consider only the linear interaction between the spin and the original boson reservoir . while it is necessary to take into account the nonlinear interaction , which causes the pure dephasing phenomena ( irreversible processes in the long time region ) in many systems@xcite , the effect of the pure dephasing is not significant in the time region shown in fig.[fig : fg4 ] , since the nonlinear interaction is often much weaker than the linear one . for the pulse interval @xmath74 .
other parameters are same as in fig.[fig : fg1 ] . ]
the synchronized pulse application has been discussed in the context of transient optical nonlinear spectroscopy , called synchronized quantum - beat echoes(sqbe ) by tanigawa , et.al.@xcite .
they used two light pulse trains in order to cause an optical transition between a ground - state sublevel pair .
the repetition frequency of the pulse trains is equal to the separation of the sublevel pair , which is superficially similar to our approach .
however , the role of the pulse train in this paper is essentially different from sqbe , because each pulse area in sqbe is much smaller than @xmath0 . in the pulse train control of decoherence in the present work ,
it is essential that the pulse area of each pulse is @xmath0 , because the physical origin of the coherence recovery is the time reversal operation caused by _ each _
@xmath0 pulse . in tanigawa
s work , the maximum signal is generated when the total pulse area of the second pulse train is @xmath0 , and the pulse train is used to achieve the sublevel resonance .
next , we assume the coupling spectral density to have semi - elliptic distribution , @xmath75 defining @xmath76 as @xmath77 to have half width @xmath56 . the coupling function has been used to describe the coupling strength between phonons and a localized electron in a solid@xcite .
we show the time evolution of @xmath34 for the same parameters in fig.[fig : fg4 ] .
fig.[fig : fg5 ] shows similar behavior as in fig.[fig : fg4 ] except the fact that the degree of the suppression for semi - elliptic coupling spectral density is larger than the one for the gaussian spectral density . for semi - elliptic coupling spectral density with @xmath78 , @xmath79 , and @xmath80 ;
( a ) without pulse application , ( b ) pulse interval @xmath69 , ( c ) pulse interval @xmath71 , ( d ) for pulse interval @xmath74 . ] defining the asymptotic peak value of @xmath67 for a pulse interval @xmath81 as @xmath82 , we show the pulse - interval dependence of @xmath82 in fig.[fig : fg6 ] . the solid ( dotted ) line in fig.[fig : fg6 ] plots the @xmath83 for semi - elliptic ( gaussian ) coupling spectral density .
we find that @xmath82 has a local maximum when the pulse interval @xmath81 is close to @xmath84 .
since @xmath85 is nearly equal to @xmath86 as shown in fig.[fig : fg6 ] , we find that the same degree of decoherence suppression is obtained for much longer pulse interval by paying attention to the dynamical motion of the reservoir
. however , the effectiveness of the spc decreases with increasing of the width @xmath87 of the coupling spectral density .
for the same parameters as in fig .
[ fig : fg4 ] and [ fig : fg5 ] .
the solid ( dotted ) line shows the @xmath83 for semi - elliptic ( gaussian ) coupling spectral density . ]
now we assume the coupling spectral density to have the lorentzian distribution as , @xmath88 the lorentzian coupling function has been often used in quantum optics .
a relaxation process of an atomic system or quantum dots in a high - q cavities has been described with a structured reservoir .
the structure is determined by a distribution of coupling constants and often described with a lorentzian function@xcite .
we show the time evolution of @xmath34 for @xmath79 and @xmath80 in fig.[fig : fg7 ] . without @xmath0 pulse application , @xmath34 shows the damped oscillation as in fig.[fig : fg7](a ) .
contrary to the previous two cases of gaussian and semi - elliptic distribution , fig.[fig : fg7](b ) shows that we can not obtain the sufficient decoherence suppression for short pulse interval @xmath69 . increasing the pulse interval to @xmath71 , we find that the degree of suppression becomes worse(fig.[fig : fg7](c ) ) .
the time dependence under spc is shown in fig.[fig : fg7](d ) .
we find that the almost the same time evolution as the one without pulse control . for lorentzian coupling spectral density with @xmath80 ;
( a ) without pulse application , ( b ) pulse interval @xmath69 , ( c ) pulse interval @xmath71 , ( d ) pulse interval @xmath74 . ] it should be noted that the ineffectiveness of the spc does not come from the fact that the decay occurs faster for the lorentzian coupling spectral density than the non - lorentzian one . in order to make the point clear
, we show the time dependence for smaller @xmath89 @xmath90 , where we find the larger amplitude of oscillation in the time evolution without pulse application in fig.[fig : fg8 ] .
the spc makes the situation even worse , and is ineffective for the system with lorentzian coupling spectral density .
we will discuss the physical background of the ineffectiveness of the spc in the next section . for lorentzian coupling spectral density with @xmath91 .
the evolution with the spc is compared the one without pulse application . ]
let us consider the reason why the spc is ineffective for the case of lorentzian coupling spectral density .
here we use a picture which provides us an qualitative understanding of the physical process of the spc .
the picture is obtained by replacing the original boson reservoir by a two - step structured reservoir where a single harmonic oscillator is coupled to a new `` reservoir '' with a coupling function which is different from the original one .
( the `` reservoir '' means the newly introduced reservoir that causes the decay of the single harmonic oscillator . )
the single harmonic oscillator is called as a quasi mode for atom - cavity system@xcite or an interaction mode for electron - phonon system@xcite . in the following ,
we call the new harmonic oscillator as the interaction mode(fig.[fig : fg9 ] )
. the motion of the interaction mode , which is determined by the coupling to the `` reservoir '' , is characterized by the original coupling spectral density . the frequency ( decay constant ) of the motion of the interaction mode corresponds to the center frequency ( width ) of the original coupling spectral density , respectively .
the application of a @xmath0 pulse causes time reversal to the two - level system .
since the two - level system is coupled to the interaction mode , the degree of time reversal depends on the reversibility of the interaction mode . in the spc
, we equalize the pulse interval to the oscillation period of the interaction mode .
if the reversibility of the interaction mode partially remains at the pulse application times , the spc is effective for suppression of decoherence .
when the original spin - boson interaction is characterized by the lorentzian coupling spectral density , it has been shown that the interaction mode is coupled to the `` reservoir '' with a flat ( white ) coupling spectral density@xcite .
the interaction mode shows the markovian nature , which indicates the irreversibility of the motion of the interaction mode(see appendix ) .
the spc is ineffective for the lorentzian spectral density . for the case of the non - lorentzian coupling spectral density
, we can also use the two - step structured reservoir where the interaction between the interaction mode and the `` reservoir '' is characterized by a non - white coupling spectral density .
this implies that the time evolution of the interaction mode is non - markovian , and partially reversible at the pulse application time .
one should note that the non - lorentzian coupling spectral density may not always guarantee the effectiveness of pulse control when the coupling spectral density has a slow power - law tail at high frequencies as a recent study indicates@xcite .
however , in many systems such as localized - electron phonon systems , the shape of the coupling spectral density has clear maximum and minimum cut - off frequencies .
the detailed characteristics of the spectral density for an effective pulse control deserve further research .
we should remark that in the limit of short pulse interval , the pulse control is effective even for the lorenzian coupling function . in this case , the interaction mode does not oscillate , and the result is consistent with the ordinary dynamical decoupling@xcite .
we have proposed a new strategy to suppress decoherence by multipulse control , which is done by synchronizing a @xmath0 pulse train with the dynamical motion of reservoir .
we have discussed the effectiveness of spc for the non - lorentzian and the lorentzian coupling spectral density . for the former case
, we find a periodic recovery of a quantum superposition at the pulse application times , whereas we can not obtain the recovery for the latter case .
our scheme is somewhat similar to the synchronized quantum - beat echoes(sqbe)@xcite in the sense that pulses are applied synchronously with the dynamical feature of reservoir .
however , the sqbe is obtained by accumulating the response to each pulse whose area is much smaller than @xmath0 .
this means that the sufficient time reversal does not occur in the sqbe scheme and that the obtained echo is different from ours where essential physical origin is time reversibility caused by the each @xmath0 pulse application .
the spc provides us an another kind of method to suppress the decoherence by paying attention to the dynamical motion of the reservoir .
we hope that the synchronized pulse control might extend the possibility of the pulse control of decoherence .
especially , the drastic alteration in the feature of the quantum superposition by changing the pulse interval might indicate that the pulse application plays both roles to keep and erase a quantum memory .
this study is supported by the grant in aid for scientific research from the ministry of education , science , sports and culture of japan .
the authors are deeply thankful to dr . l. viola and the anonymous reviewer for their thoughtful comments and suggestions .
in this appendix , we briefly review how to introduce the two - step model with using the interaction mode(or the quasi mode ) .
the interaction mode is the single harmonic oscillator , and the two - level system couples only to this oscillator .
the annihilation operator of the interaction mode is defined by the linear combination of those for the original normal modes as @xmath92 where @xmath93 then , eq.([eqn:4 ] ) can be rewritten as @xmath94 the new `` reservoir '' modes are determined to be orthogonal to the interaction mode , and the many `` reservoir '' modes are orthogonal with each other .
these boson systems other than the interaction mode do not couple to the spin system , and are called the `` reservoir '' modes .
we should note that the `` reservoir '' here implies the environmental degrees of freedom for the interaction mode , and not for the spin system .
the interaction mode and the `` reservoir '' are coupled to each other as @xmath95 where @xmath96 ( @xmath97 ) is the annihilation ( creation ) operator of the oscillator for `` reservoir '' modes .
the transformation from the original normal modes to the interaction and `` reservoir '' modes provides a new picture where a two - level system interacts with a single harmonic oscillator which contacts with a `` reservoir '' .
when we consider a subsystem which consists of the interaction mode and the `` reservoir '' with @xmath98 , the markovian master equation which describes time evolution of the interaction mode is solved to give the exponential decay of the coherent state amplitudes@xcite , whose decay constant is determined by the coupling spectral density @xmath99 .
the center frequency of the coupling spectral density in the original normal modes corresponds to the frequency of the interaction mode , and the width of the coupling spectral density is associated with the decay of the interaction mode arising from the coupling between the interaction mode and the reservoir .
the lorentzian coupling spectral density expressed by eq .
( [ eqn:31 ] ) implies a markovian time evolution of the interaction mode , which indicates the irreversibility of the motion of the interaction mode .
the quasi mode in the quantum optics has been obtained by an analogous procedure@xcite .
we have another example to show that the two - step model with the white `` reservoir '' is equivalent to the normal mode picture with the lorentzian coupling function@xcite .
other than these examples , the two - step model has also been used to describe an effect of coupling between the nuclear reaction coordinate and the other coordinates on electron transfer in biomolecules@xcite .
99 l.viola and s.lloyd , phys .
a58 , 2733 ( 1998 ) .
m.ban , j. mod .
45 , 2315 ( 1998 ) .
lu - ming duan and guang - can guo , phys .
lett . a 261 , 139(1999 ) .
g.s.agarwal , m.o.scully and h.walther , phys .
86 , 4271 ( 2001 ) .
g.s.agarwal , m.o.scully and h.walther , phys .
a63 , 044101 ( 2001 ) .
c.search and p.r.berman , phys .
85 , 2272(2000 ) .
c.search and p.r.berman , phys .
a62 , 053405 ( 2000 ) . d.vitali and p.tombesi , phys .
a65 , 012305 ( 2001 ) . d.vitali and p.tombesi , phys .
a59 , 4178 ( 1999 ) .
l.viola , e.knill , and s.lloyd , phys .
82 , 2417 ( 1999 ) .
l.viola , s.lloyd and e.knill , ibid .
83 , 4888 ( 1999 ) .
l.viola and e.knill , phys .
90 , 037901(2003 ) ; quant - ph/0208056 . k.shiokawa and d.a.lidar , quant - ph/0211081 .
c.uchiyama and m.aihara , phys .
a66,032313(2002 ) .
m.tanigawa et .
al . , j. opt .
b9 , 313(1992 ) .
y.toyozawa , _ optical processes in solids _ , ( cambridge university press,2003 ) .
h.m.lai , p.t.leung , and k.young , phys .
a37,1597(1988 ) .
h.j.kimble , adv . at .
s2 , 203(1994 ) .
p.lambropoulos , g.m.nikolopoulos , t.r.nielsen and s.bay , rep.prog.phys.63,455(2000 ) .
r.lang , m.o.scully , and w.e.lamb,jr .
a7,1788(1973 ) . s.m.barnett and p.m.radmore , opt .
68,364(1988 ) .
b.j.dalton , s.m.barnett , and p.l.knight , j. mod .
46,1315(1999);ibid,46,1495(1999 ) .
b.j.dalton and p.l.knight , j. mod
46,1817(1999);ibid,46,1839(1999 ) b.j.dalton , s.m.barnett , and b.m.garraway , phys .
a64,053813(2001 ) . y.toyozawa and m.inoue , j. phys . soc .
21 , 1663(1966 ) .
s.j.d.phoenix , phys .
a 41 , 5132(1990 ) .
y.hamano and f.shibata , j. phys .
51 , 1727(1982 ) .
a.garg , j.n.onuchic and v.ambegaokar , j. chem .
83 , 4491(1985 ) .
d.vitali , p.allegrini and p.grigolini , chem . phys .
180 , 297(1994 ) . | we present a new strategy for multipulse control over decoherence . when a two - level system interacts with a reservoir characterized by a specific frequency , we find that the decoherence is effectively suppressed by synchronizing the pulse - train application with the dynamical motion of the reservoir .
we discuss the applicability of this strategy by studying the dependence of the decoherence suppression on the shape of the coupling spectral density .
we find that the effectiveness of this strategy arises from the non - markovian nature of dynamical motion of the reservoir . |
the hypothesis that elliptical galaxies may be formed from the merging of two disk galaxies was first proposed by toomre & toomre ( 1972 ) .
it is now known to be an essential part of galaxy formation , but there are few merging systems close enough to be studied in depth .
the merger - remnant systems ngc 3921 and ngc 7252 are the proto - typical products of major disk - disk merging , at the end of the merger evolutionary sequence ( read & ponman 1998 ) .
their nuclei have already coalesced , and their optical surface brightness profiles are close to the r@xmath8 de vaucouleurs law seen in typical elliptical galaxies ( schweizer 1996 ; 1982 )
. however , they are not yet fully relaxed , and still display the disturbed morphology of post - merger objects , with tails and loops , which are the classic signature of equal - mass mergers ( figure [ opt ] ) .
these two objects are nearby , and have been well - studied across a range of wavelengths .
however it is only with the recent advent of _ xmm - newton _ and _ chandra _ that it has become possible to study their x - ray emission in any detail .
our _ xmm - newton _ observations , and in the case of ngc 7252 , archival _ chandra _ data , in addition to the existing multi - wavelength observations , therefore provide a new opportunity to study the x - ray emission - from both extended and point - like sources - arising in post - merger , proto - elliptical galaxies .
ngc 3921 and ngc 7252 have different physical properties ( see table [ props ] ) , most likely as a result of their different post - merger ages , which are @xmath00.7 and @xmath01.0 gyr respectively ( schweizer 1996 ; hibbard & mihos 1995 ) and differences in their parent galaxies .
ngc 3921 is believed to be the product of a gas - poor@xmath9gas - rich , sa@xmath9sc ( or s0@xmath9sc ) merger ( hibbard & van gorkom , 1996 ) , whereas ngc 7252 is the product of gas - rich@xmath9gas - rich , sc@xmath9sc merger ( hibbard et al .
1994 ) . as ngc 7252
is more evolved than ngc 3921 , its optical emission is more relaxed and symmetrical .
the gas - rich nature of its progenitor galaxies has induced a vehement starburst , with more luminous young globular clusters arising from this starburst than are seen in the gas - poor@xmath9gas - rich merger remnant ngc 3921 ( see table [ props ] ) . these galaxies offer us an ideal laboratory to investigate how the different merger histories and merger ages of these galaxies are reflected in their x - ray emission .
hence , we investigate both the extended hot diffuse gas components and x - ray point source ( xrp ) populations in these galaxies .
[ cols="<,^,^",options="header " , ] read and ponman ( 1998 ) , in a study of the x - ray evolution of on - going mergers using rosat data , find that the late stages of mergers are under - luminous in x - rays compared with typical ellipticals .
although when the two nuclei coalesce , massive extensions of hot gas are observed ( up to 10@xmath10 , e.g. in the antennae , arp 220 , ngc 2623 ) , after this peak , the x - ray luminosity decreases .
in fact , in post - merger remnants , l@xmath5 / l@xmath11 is @xmath0 5 times less than that of a typical elliptical ( e.g. fabbiano et al .
if major mergers are indeed the progenitors of normal elliptical galaxies , the x - ray halo must be in some way regenerated following the merging incident .
in this section , we use our _ xmm - newton _ observations to investigate possible mechanisms for the regeneration of x - ray haloes in ngc 3921 and ngc 7252 .
as we have outlined in the introduction , there are three hypotheses proposed for rebuilding the hot halo : the infall of tidally stripped cold gas , mass loss from strong stellar winds , and the return of hot gas expelled earlier in the merging process . the hi gas in the two merger remnants is clearly associated with the debris from their gas - rich parent - galaxies , i.e. the tails and western loop in ngc 7252 and the southern tail in ngc 3921 .
hi gas is not , however , seen to the north of ngc 3921 ( m@xmath12/l@xmath13 0.02 / , hibbard & van gorkom 1996 ) . in ngc 7252 ,
the hi gas is not obviously associated with x - ray emission . in ngc 3921 ,
the picture is somewhat different .
the extremely luminous source to the south of the nucleus ( 3921-x2 ) is clearly associated with the hi tail , h@xmath14 emission and a large group of star clusters , suggesting that it is related to recent star formation .
the spatial resolution of _ xmm - newton _ is not sufficient to determine whether this source is point - like , and the possibility that this source is not an x - ray binary but a local region of shock - heated diffuse gas can not be dismissed on the basis of the spatial data .
however , it is unlikely that the region is being shock - heated via winds from close - by stellar clusters ; although these can reach temperatures @xmath7 2 kev ( the temperature of the best fitting mekal model to 3921-x2 ) , their x - ray luminosities are several orders of magnitude lower ( @xmath15 10@xmath16 ergs@xmath2 ) than in 3921-x2 ( e.g. stevens & hartwell 2003 ) .
the possibility that 3921-x2 is a shock region due to the infall of cold gas still remains .
however , we can estimate the upper limit to the post - shock temperature , and compare this with the temperature of the best - fitting mekal model of this source . from cavaliere et al .
( 1997 ) : @xmath17 where kt@xmath18 is the pre - shock temperature , kt@xmath19 is the post - shock temperature , m@xmath20 is the proton mass , and @xmath21 is the pre - shock velocity of the infalling cold gas . the upper limit of the gas velocity is the velocity it would gain from falling into the galaxy from rest at infinity .
we take the dynamical mass of the galaxy to be 1.06x10@xmath22 ( hibbard & van gorkom 1996 ) , and assume the pre - shock cold gas temperature is negligible .
putting these assumptions into equation 2 , we estimate the upper limit of the post - shock temperature to be @xmath00.96 kev ( @xmath010@xmath23k ) .
this is well below the lower limit of the best - fitting mekal model ( 1.94 kev ) , and suggests that shocked gas is not the origin of 3921-x2 .
however , it should be noted that freshly shocked gas may not be in either thermal or ionization equilibrium , and this can alter the observed properties of the post - shock gas .
takizawa ( 1999 ) discusses the thermal equilibrium timescale , noting that the ion temperature is initially much higher than the electron temperature in the post - shock region ( since most of the incoming kinetic energy is in the ions ) , and that the two temperatures equilibrate on an electron - ion interaction timescale .
however , this effect would lead to a _ low _ observed electron bremsstrahlung temperature , whereas we see a best fitting mekal temperature _ higher _ than the upper limit to the post - shock temperature .
moreover , the equilibration timescale at the temperature and gas density in the region of 3921-x2 ( @xmath00.4 myr ) is much shorter than the post - shock timescale , inferred from the time elapsed since star formation ( several hundred myr ) .
calculations of the thermal and dynamical evolution of a fast adiabatically expanding gas ( breitschwerdt & schmutzler 1999 ) show that a delay in recombination can lead to departures from ionization equilibrium .
a full set of dynamically and thermally self - consistent equations is needed to calculate the timescale for recombination , which is beyond the scope of this paper .
however , breitschwerdt & schmutzler suggest that the dynamical timescale for a hot plasma ( t@xmath24k ) is much shorter than the intrinsic timescales ( e.g. for recombination , ionization etc . ) .
the ionization non - equilibrium gives rise to the appearance of emission lines representative of higher temperatures than the actual gas temperature , and the over - estimation of the temperature if the spectrum is fitted with collisional ionization equilibrium models .
however , for 3921-x2 , the upper limit to the post - shock temperature is @xmath00.96 kev , and it is difficult to see how a temperature as high as the best fitting mekal temperature , 2.18 kev , could be mimicked by this process . in addition , we do not see any emission lines in the spectra ( figure [ 3921-blob - spec ] ) which we would expect from this scenario .
this lack of emission lines is consistent with the very low abundance of the best fitting mekal model .
hence , we conclude that the high temperature of the best fitting mekal model is inconsistent with a shock - heated gas model . in ngc 7252
, we do not see an x - ray emission region analogous to that of 3921-x2 , i.e. with associated clusters , hi and h@xmath14 emission .
however , in figure [ 7252hregs ] , we see a ` cavity ' in the [email protected] kev x - ray emission to the north , which appears to correspond to the north - western hi tail where it crosses the hot gas . since the tail comes across the _ front _ of the remnant body ( schweizer 1982 ; hibbard & van gorkom 1996 ) , the increased hi surface density in the cavity region ( 15@xmath920 x10@xmath25 @xmath26 , figs . 5c and 5d in hibbard et al .
1994 ) strongly suggests that the ` cavity ' is due to x - ray absorption by the neutral gas of the tidal tail in the foreground .
this effect may of course be enhanced by the positions of 7252-x1 and 7252-x7 . if x - ray halos are indeed regenerated by infall along tidal tails , then maybe insufficient time since peri - galacticon has passed for substantial regeneration to have occurred ( e.g. hibbard & mihos 1995 ) .
considerable hi is evident out at least as far as the optical tails in ngc 7252 and beyond the optical loops to the south of ngc 3921 .
hibbard & mihos ( 1995 ) model the kinematics of the hi gas in ngc 7252 using n - body simulations .
they calculate that , of the hi originally residing in the spiral progenitors of this object , at least 3x10@xmath27 has already fallen inwards . however , the central regions of ngc 7252 are hi - poor , so atomic gas falling into the remnant body from the tails must be converted to other forms , at a rate of 1@xmath92 yr@xmath2 .
the process of infall is a natural explanation for the conversion of hi to hot gas , stars and cold molecular gas .
this hypothesis is supported by observations of both ngc 3921 and ngc 7252 which show the presence of all these three phases in the remnant bodies ( e.g. whitmore et al .
1993 ; hibbard et al . 1994
; hibbard & van gorkum 1996 ; schweizer 1996 ; schweizer et al .
1996 ; yun & hibbard 2001 ) . using the results of the hibbard & mihos simulations
, we can estimate if there is sufficient hi still to fall back into the remnant body of ngc 7252 from the tidal tails , to increase the hot gas mass ( and hence luminosity ) to that seen in a typical elliptical galaxy .
the simulations indicate that @xmath06x10@xmath28 of hi will return to within 5r@xmath29 over the next 3@xmath30 gyr , i.e. one fifth as much again as has already fallen into the central regions .
we estimate the mass of hot gas already present in the remnant body from the parameters of the mekal components fitted to the nuclear region , and assuming spherical symmetry .
the mass is @xmath03x10@xmath31 , within the spectral extraction region .
` typical ' x - ray bright elliptical galaxies ( not including brightest group and massive cluster ellipticals , which may be considered a different class ) have hot gas masses of 10@xmath3210@xmath10 ( e.g. forman et al . 1985 ) . even if _ all _ the hi gas still to return to the remnant body is converted to hot gas , with none involved in star formation or conversion to molecular gas , the total hot gas mass would be just below 10@xmath33 , and it is unlikely that the conversion to hot gas is anywhere near this efficient . for comparison , @xmath03x10@xmath33 of hi has already fallen back into the body of the galaxy , but we see only @xmath03x10@xmath31 of hot gas , which , assuming all the hot gas has been converted from hi , with no contributions from stellar mass loss , represents an efficiency of only 10% .
it is hard to see that hi infall in this galaxy would create sufficient hot gas to reach the mass seen in typical elliptical galaxies . we know that the northern tail of ngc 3291 is hi - poor , but the southern tail contains @xmath0 2x10@xmath27 of hi , with the tail becoming more gas - rich along its length ( hibbard & van gorkom 1996 )
however , there are no simulations of the kinematics of the atomic gas in ngc 3921 in the literature , so we do not know what proportion of this gas will return to the remnant body .
in addition , the irregular x - ray structure of ngc 3921 invalidates the assumption of spherical symmetry used to calculate the hot gas mass in ngc 7252 , so we can not make the same calculation .
however , assuming its similar x - ray luminosity to ngc 7252 indicates a similar gas mass , this galaxy would require an additional mass of hot gas @xmath345x10@xmath31 to take it to typical values .
this only seems possible if the hi kinematics and/or conversion to hot gas in ngc 3921 are quite different from those in ngc 7252 .
given the differences between the predicted hot gas masses of typical ellipticals and those predicted from our calculations involving hi masses , it is difficult to believe that hi infall can be solely responsible for the regeneration of the x - ray halo to luminosities typical of elliptical galaxies .
additional support for the existence of another mechanism for halo regeneration comes from the work of osullivan et al .
( 2001a ) .
they see a long - term trend of increasing x - ray luminosity with mean stellar age , extending over 15 gyr , long after the hi tails seen in mergers have been completely assimilated into the body of the galaxy .
if the x - ray luminosity continues to rise after the completion of hi infall , then , even if hi conversion is responsible for the presence of some of the hot gas , there must be an additional source , which contributes to the hot gas mass after virialisation , for example , stellar mass loss , as discussed below .
the models of ciotti et al .
( 1991 ) investigate whether mass loss from stars could regenerate the x - ray halo .
they predict a super - sonic wind - driven flow in the early stages of post - merger elliptical evolution , arising from two sources : mass - loss from evolving stars , and secularly declining heating from type i supernovae . during this stage ,
a galaxy is x - ray faint , and the amount of hot x - ray gas , and hence the x - ray luminosity , decreases .
this stage can persist for @xmath0 2 gyr , and is followed by a sub - sonic outflow . as the flow velocity decreases , this flow allows the build - up of hot gas , and results in rising x - ray luminosity over timescales @xmath3410 gyr . in this model ,
ngc 3921 and ngc 7252 would be in the first , wind - driven stage , and indeed , they are x - ray faint .
the gas density and temperature are expected to fall off smoothly with radius , without the presence of shocks .
we see no evidence that this is not the case ( if we are correct in assuming that 3921-x2 is a ulx ) .
however , in order to model the behaviour of gas density and temperature with radius in detail , we would need both better spatial resolution than _ xmm - newton _ can offer , and more counts .
the sub - sonic outflow stage of this model , where x - ray luminosity increases gradually with time , is consistent with the observations of elliptical galaxies discussed by osullivan et al .
( 2001a ) .
the third possible mechanism for the regeneration of the x - ray halo is the return of the hot gas which was expelled from the main body of the galaxy during the nuclear merging stage .
this expelled gas can reach large radii before stalling .
it then falls back into the galaxy , and shocks are expected as it collides with the existing halo gas .
these outflows are typically asymmetrical ( e.g. read & ponman 1998 ) .
we see little evidence for the regeneration of the x - ray halo via this mechanism in ngc 3921 and ngc 7252 .
we see little x - ray emission at large radii in either galaxy .
this does not mean that we can reject the possibility that there is significant hot gas outside the main remnant bodies which is as yet too diffuse to have been detected in these observations .
however , although the infall of hot gas may be delayed with respect to the cold gas , as the hot gas reaches larger radii than the cold , it will still occur on a galaxy dynamical time - scale ( osullivan et al .
2001b ) . as with the infall of cold gas
, it is hard to reconcile this with the long - term ( @xmath3410gyr ) trend of increasing x - ray luminosity with mean stellar age observed by osullivan et al .
( 2001a ; 2001b ) .
we summarise our conclusions as follows : in both ngc 3921 and ngc 7252 , we observe ultra - luminous non - nuclear x - ray point sources .
where spectral fitting has been possible , the spectra are well - fitted by either power - law models ( 3921-x3 , 7252-x1 , 7252-x3 , 7252-x7 ) or a disk blackbody accretion disk model ( 3921-x2 , 7252-x4@xmath35 ) .
these energy distributions are typical of xrbs , and we suggest that these sources are hmxbs , associated with the recent star formation in these two galaxies . however , the possibility that the three off - nuclear sources in ngc 3921 are not point - like can not be ruled out at present .
the point - like nature of the four _ xmm - newton _ sources in ngc 7252 is confirmed by _
chandra _ observations , which also reveal a further two xpss , so that the total number of confirmed x - ray point sources observed in ngc 7252 ( assuming all are associated with the remnant ) is six .
7252-x8 has no corresponding _
chandra _ detection , and hence can not be confirmed as a point source .
the differences in the evolutionary stages and progenitor galaxies of ngc 3921 and ngc 7252 do not appear to be reflected in their ulx populations .
the numbers and luminosities of the candidate ulxs observed in ngc 3921 and ngc 7252 are compatible with the conclusion of swartz et al .
( 2003 ) , that more luminous ulxs are found in interacting and merging galaxies than in other morphological types . in ngc 7252 , the more evolved merger remnant ,
the diffuse x - ray gas is more relaxed : it is centred on the optical centre , is more symmetrical in the central regions and is cooler than the hot diffuse gas in ngc 3921 .
this is most likely a result of its greater dynamical age .
hot gas in ngc 7252 has a more compact core than that in ngc 3921 , reflecting differences between the progenitors of these two galaxies and their merger dynamics .
when compared to ngc 3921 , the more violent star formation in ngc 7252 is not reflected in a higher x - ray luminosity for the hot diffuse gas .
although substantial masses of hi are available to fall into the remnant bodies over the next several gyr , it seems unlikely that this is sufficient to increase the mass of hot gas in either ngc 3921 or ngc 7252 to the masses ( and hence luminosities ) seen in typical ellipticals .
we see no compelling evidence that massive x - ray halo regeneration is occurring , even in these late - stage merger remnants .
it is likely that regeneration occurs over much longer timescales , of up to 15 gyr , predominantly via mass loss from evolving stars , in a sub - sonic outflow stage commencing @xmath0 2 gyr after the merging event ( ciotti et al .
1991 ) .
* acknowledgements * + + we would like to thank the referee , andreas zezas , for his helpful comments .
we acknowledge use of the chandra data archive ( cda ) , part of the chandra x - ray observatory science center ( cxc ) , which is operated for nasa by the smithsonian astrophysical observatory .
the digitized sky survey was produced at the space telescope science institute under u.s .
government grant nag w-2166 .
the images of these surveys are based on photographic data obtained using the oschin schmidt telescope on palomar mountain and the uk schmidt telescope .
the plates were processed into the present compressed digital form with the permission of these institutions .
lan , tjp and amr acknowledge the support of pparc ; lan and amr through the award of pdras and tjp through a pparc senior fellowship .
fs gratefully acknowledges partial support from the national science foundation through grant ast-0205994 . + + + + + * references * + + awaki h. , matsumoto h. , tomida h. , 2002 , apj , 567 , 892 + angelini l. , loewenstein m. , mushotsky r.f . , 2001 , apj,557l , 35 + belczynski k. , kalogera v. , zezas a. , fabbiano g. , 2003 , submitted to apj , astro - ph/0310200 + blauw a. , 1964 , ara&a , 2 , 213 + brammer g. , whitmore b.c . , koekemoer a.m. , 2002 , in the 2002 hst calibration workshop : hubble after the installation of the acs and the nicmos cooling system , proceedings of a workshop held at the space telescope science institute , baltimore , maryland
arribas s. , koekemoer a. , whitmore b.c .. baltimore , md : space telescope science institute , 2002 . ,
p.331 + breitshwerdt d. , schmutzler t. , 1999 , a&a , 347 , 650 + bruzual g.a .
, charlot s. , 1993 , apj , 405 , 538 + cavaliere a. , menci n. , tozzi p , 1997 , apj , 484l , 21 + ciotti l. , pellegrini s. , renzini a. , dercole a. , 1991 , apj , 376 , 380 + colbert e.j.m .
, mushotsky r.f . , 1999 ,
apj , 519 , 89 + colbert e.j.m . , heckman t.m .
, ptak a.f .
, strickland d.k .
, 2004 , astro - ph/0305476 + coleman miller , m. , colbert e.j.m . , 2003
, invited review submitted to the international journal of modern physics , astro - ph/0308402 + cordes j.m .
, chernoff d.f . , 1998 ,
apj , 505 , 315 + dickey j.m . & lockman f.j . , 1990 , ara&a , 28 , 215 + deeg et al . , 1998 , a&a , 129 , 455 + ebisuzaki t. , makino j. , tsuru t.g . , funato y. , portegies zwart s. , hut p. , mcmillan s. , matsushita s. , matsumoto h. , kawabe r. , 2001 , apj , 562 , l19 + fabbiano g. , 1995 , aas , 186 , 4601 + fabbiano g. , kim d .- w . , trinchieri g. , 1992 , apjs , 80 , 531 + fabbiano g. , zezas a. & murray s.s . , 2001 ,
apj , 554 , 1035 + fabbiano g. , zezas a. , king a.r .
, ponman t.j . ,
rots a. , schweizer f. , 2003 , apj , 584 , l5 + faber s. m. , tremaine s. , ajhar e.a . ,
byun y - i . , dressler a. , gebhardt k. , grillmair c. , kormendy j. , lauer t.r . , richstone d. , 1997 , aj , 114 , 1365 + forman w. , jones c. , tucker w. , 1985 , apj , 293 , 102 + fryer c.l .
, kalogera v. , 2001 , apj , 554 , 548 + genzel r. , tacconi , l.j .
, rigopoulou d. , lutz , d. , tecza m. , 2001 , apj , 563 , 527 + helfand & moran , 2001 , apj , 554 , 27 + hibbard j.e . , guhathakurta p , van gorkom j.h .
, schweizer f. , 1994 , aj , 107 , 67 + hibbard j.e . , van der hulst j.m .
, barnes j.e . , rich r.m . , 2001 , aj , 122 , 2969 + hibbard j.e . , van gorkom j.h . , 1996 , aj , 111 , 655 + hibbard j.e . ,
mihos , j.c . , 1995 , aj , 110 , 140 + irwin j. , sarazin c. , 1998 , apj , 499 , 650 + irwin j.a . , bregman j.n . ,
athey a.e . , 2003 , accepted by apj letters ,
astro - ph/0312393 + kaastra , j.s .
1992 , an x - ray spectral code for optically thin plasmas ( internal sron - leiden report , updated version 2.0 ) + king a.r . , 2002 , mnras , 335l , 13 + king a.r . , 2003 , accepted by mnras , astro - ph/0309450 + king a.r . , davies m.b . ,
ward m.j . , fabbiano g. , elvis m. , 2001 , apjl , 552 , 109 + kulkarni s.r . , hut p. , mcmillan s. , 1993 , nature , 364 , 421 + laycock s. , corbet r.h.d .
, coe m.j . , marshall f.e . ,
markwardt c. , edge w. , 2003 , mnras 338 , 211 + lee h.m . , 1995 , mnras , 272 , 605 + lee m.h .
, 1993 , apj , 418 , 147 + liedahl d.a .
, osterheld a.l . , goldstein w.h .
1995 , apjl , 438 , 115 + maccarone t.j . , kundu a. , zepf s.e . , 2003 ,
apj , 586 , 814 + maddox , sutherland , efstathiou , loveday , 1990 mnras , 243,692 + makishima k. , maejima y. , mitsuda k. , bradt h.v .
, remillard r.a . ,
tuohy i.r .
, hoshi r. , nakagawa m. , 1986 , apj , 308 , 635 + makishima k. et al . 2000 , apj , 535 , 632 + matsushita k. et al . , 1994 ,
apj , 436l , 41 + mewe r. , gronenschild e.h.b.m . , van den oord g.h.j .
1985 , a&as , 62 , 197 + mewe r. , lemen j.r .
, van den oord g.h.j .
1986 , a&as , 65 , 511 + miller b.w .
, whitmore b.c . , schweizer f. , fall m.s . , 1997 ,
aj , 114 , 2381 + miller j.m .
, zezas a. , fabbiano g. , schweizer f. , 2003 , submitted to apj , astro - ph/0305488 + mitsuda k. , inoue h. , koyama k. , makishima k. , matsuoka m. , ogawara y. , suzuki k. , tanaka y. , shibazaki n. , hirano t. , 1994 , pasj , 36 , 741 + osullivan e.j .
, forbes d.a .
, ponman t.j . , 2001a , mnras , 328 , 462 + osullivan e.j . , forbes d.a . ,
ponman t.j .
, 2001b , mnras , 324 , 420 + quinlan g.d . , 1996 , newa , 1 , 35 + rangarajan f.v.n . , fabian a.c .
, forman w.r . , jones c. , 1995 , mnras , 272 , 665 + read a.m. , 2003 , mnras , 342 , 715 + read a.m. , ponman t.j . , 1998 , mnras , 297 , 143 + roberts t.p . ,
warwick r.s . , 2000 ,
mnras , 315 , 98 + schweizer f. , 1982 , apj , 252 , 455 + schweizer f. , 1996 , aj , 111 , 109 + schweizer f. , miller b.w . , whitmore b.c . , fall m.s.,1996 , aj 112 , 1839 + sigurdsson s. , hernquist l. , 1993 , nature , 364 , 423 + stauffer j.r . , 1982a , apj , 262 , 66 + stauffer j.r .
, 1982b , apjs , 50 , 517 + stevens i.r . , hartwell j.m . , 2003 , mnras , 339 , 280 + swartz d.a . , ghosh k.k . , tennant a.f . , 2003 , aas , 202 , 1109 + takizawa m. , 1999 , apj , 520 , 514 + tozzi p. et al . , 2001 , apj , 562 , 42 + webbink r.f . , 1985 , in dynamics of star clusters , iau symposium 113 , ed .
j. goodman & p. hut ( dordrecht : reidel ) , 541 + whitmore b.c .
, schweizer f. , leitherer c. , borne k. , robert c. , 1993 , aj , 106 , 1354 + white r.e . ,
sarazin c.l . ,
kulkarni s.r . , 2002 ,
apj , 571l , 23 + yun m.s .
, hibbard j.e . , 2001 , apj , 550 , 104 + zezas a. , fabbiano g. , rots a.h . ,
murray , s.s . , 2002 ,
apj , 142,239 + zwicky f. , 1956 , ergebnisse der exakten naturwissenschaften , 29 , 344 + | using the high sensitivity of _ xmm - newton _ , we have studied the x - ray emission of the two proto - typical late - stage merger remnants , ngc 3921 and ngc 7252 . in the case of ngc 7252 , this is complemented by archival _
chandra _ data .
we investigate the nature of the discrete x - ray point source populations and the hot diffuse gas components in these two galaxies , and compare them in the light of their different merger ages and histories .
we detect 3 candidate ultra - luminous x - ray point sources in ngc 3921 and at least 6 in ngc 7252 , for which we have high spatial resolution _
chandra _ data . these have luminosities ranging from @xmath01.4x10@xmath1 erg s@xmath2 ( for h@xmath3 75 km s@xmath2 mpc@xmath2 ) .
we expect these ulxs to be high mass x - ray binaries , associated with the recent star formation in these two galaxies .
extended hot gas is observed in both galaxies .
we have sufficient counts in the _ xmm - newton _ data to fit two - component hot plasma models to their x - ray spectra and estimate the x - ray luminosities of the hot diffuse gas components to be 2.75x10@xmath4 erg s@xmath2 and 2.09x10@xmath4 erg s@xmath2 in ngc 3921 and ngc 7252 , respectively .
these luminosities are low compared with the luminosities observed in typical mature elliptical galaxies ( l@xmath5@xmath010@xmath6 erg s@xmath2 ) , into which these merger remnants are expected to evolve .
we do not see evidence that the x - ray halos of these galaxies are currently being regenerated to the masses and luminosities seen in typical elliptical galaxies .
the mass of atomic gas available to fall back into the main bodies of these galaxies and shock - heat to x - ray temperatures is insufficient for this to be the sole halo regeneration mechanism .
we conclude that halo regeneration is most likely a long - term ( @xmath710 gyr ) process , occurring predominantly via mass loss from evolving stars , in a sub - sonic outflow stage commencing @xmath0 2 gyr after the merging event .
galaxies : individual : ngc 3921 - galaxies : individual : ngc 7252 - galaxies : evolution - x - rays : galaxies - x - rays : ism - x - rays : binaries |
the celebrated gauss - lucas theorem claims that for any univariate polynomial @xmath1 with complex coefficients , all roots of @xmath2 belong to the convex hull of the roots of @xmath1 , see theorem 6.1 of @xcite .
many generalizations have been obtained over the years see , e.g. , @xcite and references therein . in the present note , motivated by problems in potential theory in @xmath3
we extend the gauss - lucas theorem to sequences of polynomials of increasing degrees whose almost all zeros lie in a given convex bounded domain in @xmath0 .
namely , given a convex bounded domain @xmath4 let @xmath5 be a sequence of univariate polynomials with the degrees @xmath6 such that @xmath7 .
assume that @xmath8 where @xmath9 is the number of zeros of @xmath10 lying in @xmath11 ( counted with multiplicities ) .
[ prob : main ] following the above notation we now ask whether there exists @xmath12 where @xmath13 denotes the number of zeros of @xmath14 lying in @xmath11 ?
it turns out that the answer to problem [ prob : main ] formulated verbatim as above , is , in general , negative . ]
let @xmath15 be the open square @xmath16 .
if @xmath17 is the @xmath18-th chebyshev polynomial of the first kind , then the derivative of @xmath19 has all its zeros in the upper half plane .
therefore , if we replace @xmath20 by @xmath21 for some sufficiently small @xmath22 ( i.e. , shift all zeros downward by @xmath22 ) , then we obtain polynomials of degrees @xmath23 with @xmath18 zeros in @xmath15 , but whose derivatives have no zeros in @xmath15 .
choosing @xmath22 appropriately for each @xmath18 , we get a sequence of polynomials with all but one zeros in @xmath15 whose derivatives have no zeros in @xmath15 .
strict convexity ( e.g. , as in the case of the open unit disk ) will not be of much help either .
just replace above @xmath20 by @xmath24 with some large @xmath25 and then make a vertical translation so that after all these operations the image of @xmath26 $ ] becomes a tiny secant segment of the unit circle .
( this example was suggested to the third author by professor v. totik . )
however with slightly weaker requirements problem 1 has a positive answer .
[ th : main ] given a polynomial sequence @xmath27 as above and any @xmath28 , @xmath29 where @xmath30 is the number of zeros of @xmath14 lying in @xmath31 and @xmath32 is the open @xmath33-neighborhood of @xmath34 .
two illustrations of theorem [ th : main ] are given in figure 2 .
_ acknowledgements . _
the third author want to thank professor v. totik for discussions of the problem .
( larger dots ) and @xmath35 ( smaller dots ) for @xmath36 ( left ) and @xmath37 where @xmath38 has @xmath39 uniformly placed zeros on each side of the unit square ( right).,title="fig : " ] ( larger dots ) and @xmath35 ( smaller dots ) for @xmath36 ( left ) and @xmath37 where @xmath38 has @xmath39 uniformly placed zeros on each side of the unit square ( right).,title="fig : " ] @xmath63 hence @xmath64 where @xmath65 is chosen so that @xmath66 the disk of radius @xmath65 centered at @xmath67 contains @xmath68 .
the integral in converges for all @xmath69 and follows , hence does .
to finish the proof for the general case of an arbitrary convex domain @xmath11 observe that we only used some properties of a disk to get a convenient foliation of a neighborhood of the unit disk by concentric circles and when applying gauss - lucas theorem .
both these facts are readily available for an arbitrary bounded convex domain . finally , the hardy spaces of analytic functions in the disk are replaced by the smirnov classes @xmath89 of functions representable by cauchy integrals with @xmath90-densities ( with respect to arclength ) . in the domains with piecewise smooth boundaries , e.g. , convex domains ,
the latter behave in the very same manner as hardy spaces cf . @xcite . | in this note we extend the gauss - lucas theorem on the zeros of the derivative of a univariate polynomial to the case of sequences of univariate polynomials whose almost all zeros lie in a given convex bounded domain in @xmath0 . |
the tidal disruption of stars is an inevitable consequence of the presence of supermassive black holes ( smbhs ) at the cores of galaxies .
stars in close approach can be ripped apart by the tidal forces of the smbh .
a significant fraction of the stellar material will subsequently be accreted , producing a luminous flare of electromagnetic radiation .
the occurrence of tdes was first predicted based on theoretical considerations in the seventies . at that time , many important theoretical foundations were laid , and first ideas emerged , on the relevance and usage of tdes in astrophysical context ( hills 1975 , frank & rees 1976 , young et al .
1997 , kato & hoshi 1978 , gurzadian & ozernoi 1979 , carter & luminet 1982 , luminet & marck 1985 , rees 1988 ) .
questions which were raised included : do tdes provide a fuel source for quasars ? how does the occurrence of tdes constrain the presence of a smbh at our galactic center ?
can extreme tdes explain variants of gamma - ray bursts ( grbs ) ?
are tdes a unique signpost of the presence of dormant smbhs at the cores of inactive galaxies ? events were discovered in the nineties in form of luminous , soft x - ray outbursts from otherwise quiescent galaxies during the _ rosat _ all - sky survey , and have more recently been found with other x - ray missions , and at longer wavelengths .
the _ swift _ mission has been a game changer in this field .
its discovery of the first tde which launched a relativistic jet , swiftj1644 + 57 , has triggered many theoretical studies on the formation of radio jets , and swiftj1644 + 57 has now the best - covered lightcurve of any tde to date .
further , _
swift _ has been used for follow - ups of tdes discovered by other missions , providing excellent lightcurves and contributing to the discovery of the first candidate binary smbh in a quiescent galaxy .
a star is disrupted , once the tidal forces of the smbh exceed the self - gravity of the star ( hills 1975 ) .
the distance at which this happens , the tidal radius , is given by @xmath1 a fraction of the stellar material will be on unbound orbits and escape , while the rest will eventually be accreted ( fig .
the events appear as luminous transients .
their emission peaks in the uv or soft x - rays , declining on the timescale of months to years ( e.g. , rees 1990 , evans & kochanek 1989 , cannizzo et al .
because tidal radius and schwarzschild radius depend differently on bh mass , solar type stars are swallowed whole for smbh masses exceeding @xmath2 10@xmath3 m@xmath4 ( smbh spin can raise this limit ; beloborodov et al .
white dwarfs can be tidally disrupted for smbh masses below @xmath2 10@xmath5 m@xmath4 .
computations of the stellar evolution , including the early stages of stellar deformation , the actual disruption , accretion and ejection of material , the development of a disk wind , and the evolution and late stages of the accretion phase , and for different types of stars , are very challenging .
recent state of the art modelling has addressed the different stages of tde evolution under various conditions ( for the most recent results , see , e.g. , bonnerot et al .
2015 , guillochon & ramirez - ruiz 2015 , hayasaki et al .
2015 , miller 2015 , piran et al .
2015 , shiokawa et al .
2015 , and references therein ) .
other recent studies have addressed the disruption rates for different galaxy morphologies ( e.g. , brockamp et al .
2011 , vasiliev 2014 , stone & metzger 2014 , zhong et al . 2014 ) , and for recoiling and binary smbhs ( e.g. , komossa & merritt 2008 , chen et al . 2009 , 2011 ,
stone & loeb 2011 , 2012a , li et al . 2012 , liu & chen 2013 ) and spinning bhs ( kesden 2012 ) .
these have shown that rates are strongly boosted in some phases of binary smbh evolution and the early phase of smbh recoil , and that rates strongly depend on bh spin for the most massive bhs ( @xmath6 m@xmath4 ) .
tidally disrupted stars will also produce a gravitational wave signal along with the electromagnetic emission ( e.g. , east 2014 , and references therein ) .
there is now increasing evidence that smbhs reside at the centers of many massive galaxies ( review by graham 2015 ) .
tdes likely significantly contributed to growing smbhs at low masses ( @xmath7 m@xmath8 ; freitag & benz 2002 ) , while stars swallowed whole may have contributed to smbh growth at high masses ( zhao et al .
one initial idea was to use tde flares as tracers of dormant smbhs in quiescent galaxies ( e.g. , rees 1988 ) .
this will continue to be an important topic in the future , since the luminous accretion flares reach out to large cosmic distances , and tdes will be detected in large numbers in future transient surveys ( sect .
in addition , many other applications have been suggested , making use of the characteristic tde properties and rates : * in x - rays , tdes probe relativistic effects ( via emission - line profiles or precession effects in the kerr metric ) and the extremes of accretion physics at high rates and near the last stable orbit , and provide us with a new means of measuring bh spin . *
jetted tdes provide new insight into the formation and early evolution of radio jets , and may shed new light on related issues like the cause of the radio - loud radio - quiet dichotomy of active galactic nuclei ( agn ) . *
tdes , once detected in large numbers , will unveil the population of imbhs in the universe .
* tde rates depend on , and therefore trace , stellar dynamics in galaxy cores on spatial scales which can not be resolved directly .
* tdes are signposts of binary smbhs and recoiling bhs , because their rates are strongly enhanced under these conditions , and tdes will occur off - nuclear if the smbh is recoiling . *
tdes in gas - rich environments will illuminate the circum - nuclear material , so that the reprocessed emission lines and their temporal evolution provide us with an unparalleled opportunity of reverberation mapping the cores of quiescent galaxies . here , we present an overview of the status of observations of tdes , highlighting the important role plaid by the _ swift _ mission ( gehrels et al .
2004 , gehrels & cannizzo 2015 ) . an accompanying review by g. lodato ( 2015 ) will focus on theoretical aspects of tidal disruption .
law ( dashed lined ) .
this point was first made based on _ rosat _ data of ngc5905 ( komossa & bade 1999 ) , and later for the overall luminosity evolution of the sources displayed above ( e.g. , fig . 1 of komossa 2004 ) .
rxj1242 - 1119 shows a further drop in x - rays at late times ( not shown here ) , deviating from the early phase decline law , implying a total amplitude of decline of a factor @xmath21000 ( komossa 2005 ) .
, width=359 ]
the _ rosat _ observatory ( trmper 2001 ) with its high sensitivity , long lifetime , all - sky coverage in its first year of operation , and soft x - ray response ( 0.12.4 kev ) , was well suited for the detection of tdes .
luminous , high - amplitude x - ray flares from quiescent galaxies , matching all order - of - magnitude predictions of the tidal disruption scenario ( e.g. , rees 1990 ) , have first been discovered during the _ rosat _ x - ray all - sky survey ( rass ; fig . 2 , 3 ) .
four main events were identified , from the galaxies ngc5905 ( bade et al .
1996 , komossa & bade 1999 ; see also li et al .
2002 ) , rxj1242@xmath91119 ( komossa & greiner 1999 ) , rxj1624 + 7554 ( grupe et al .
1999 ) , and rxj1420 + 5334 ( greiner et al . 2000 ) . among these , ngc5905 and
rxj1242@xmath91119 are the best - covered events in terms of their _ long - term _ x - ray lightcurves , spanning time intervals of more than a decade , with amplitudes of decline _ larger than a factor of 1000 _ ( komossa et al . 2004 , halpern et al .
2004 , komossa 2005 ) .
ngc5905 was first noticed due to its luminous , soft ( @xmath10 kev ) x - ray emission with peak luminosity in the soft x - ray band of @xmath11 erg / s during the rass .
it remained bright for at least @xmath25 days ( the time interval its position was repeatedly scanned during the rass ) increasing in luminosity to the observed peak .
x - rays then declined on the timescale of months to years ( fig .
3 ) . within the errors , the x - rays came from the center of this nearby
barred spiral galaxy ( @xmath12 ; fig .
2 ) . while the x - ray spectrum was initially very soft , it had hardened significantly ( @xmath13 ) 3 years later , when re - observed with _ rosat_. the decline of its x - ray lightcurve is well consistent with the predicted @xmath14 law , as first reported based on its _ rosat _ observations ( komossa & bade 1999 ) and confirmed with _
( halpern et al . 2004 ) .
all observations of this event are in very good agreement with tidal disruption theory ( bade et al .
1996 , komossa & bade 1999 ) .
whenever enough data exist , the _ rosat _ events , and most of the more recent soft x - ray tdes ( next section ) , follow a similar trend in spectral and lightcurve evolution as ngc5905 , providing independent evidence that the same mechanism was at work in all cases . more recently , similar x - ray events have been detected with _
chandra _ and _ xmm - newton _ , based on dedicated searches or serendipitous discoveries . the _ xmm - newton _ slew survey has been used to identify new bright tdes based on a comparison with the _ rosat _ data base , and a few events have been found so far ( esquej et al .
2007 , 2008 , saxton et al .
2012 , saxton et al .
2015 in prep ) . among these ,
sdssj120136.02 + 300305.5 has the best - covered first - year lightcurve ( saxton et al .
2012 ) , based on follow - ups with _ xmm - newton _ and _ swift_. overall , the x - rays continue fading after high - state .
additional large - amplitude variability is apparent on the timescale of weeks ( fig .
the x - ray spectrum of sdssj120136.02 + 300305.5 , observed with _
xmm - newton _ weeks and months after high - state is very soft ( no photons detected beyond 2 - 3 kev ) , but is not well fit with black - body emission .
it is consistent with a broken powerlaw or a bremsstrahlung - like spectral shape .
a few tdes were identified in clusters of galaxies ( cappelluti et al .
2009 , maksym et al .
2010 , 2013 , donato et al .
the most likely counterpart of the source wingsj1348 in abell 1795 is a dwarf galaxy , and the disrupting black hole is of relatively low mass , @xmath15 m@xmath4 ( maksym et al .
2013 , 2014a , donato et al .
a second candidate tde hosted by a dwarf galaxy was reported by maksym et al .
( 2014b ) .
other events emerged through systematic searches of the _ xmm - newton _ data base ( lin et al .
2011 , 2015 ) and new searches of the _ rosat _ data base ( khabibullin & sazonov 2014 , maksym et al .
the events cover x - ray luminosities in the range ( 10@xmath16 several 10@xmath17 ) erg / s , and arise in relatively nearby galaxies ( @xmath18 ) which are optically quiescent ( i.e. , they lack the characteristic optical narrow emission lines of agn ) .
the _ swift _ mission has been essential in providing rapid follow - ups of several of these events , confirming the fading x - rays , and providing tight constraints on the luminosity evolution .
overall , the salient properties of the soft x - ray tdes detected with _ rosat _ , _ xmm - newton _ and _ chandra _ can be summarized as follows : * peak luminosities are large , up to several 10@xmath17 erg / s in the soft x - ray band .
* amplitudes of decline reach factors up to @xmath19 ( the _ rosat _ events ) , more than a decade after the observed high - states .
* x - ray spectra are very soft during the high - states ( @xmath20 kev ) , followed by a spectral hardening on the time scale of years .
* host galaxies show essentially no evidence for _ permanent _ activity as it is seen in agn .
years after the flare ( and before , when data exist ) , the galaxies are optically inactive , radio inactive , and x - ray inactive .
* x - ray lightcurves decline on the timescale of months years , and are overall consistent with the law @xmath21 predicted by the fall - back model of tidal disruption theory .
* smbh masses ( derived from applying scaling relations or multi - wavelength correlations ) are mostly on the order of 10@xmath22 m@xmath4 . * only a small amount of the stellar material needed to be accreted to power the observed emission , typically @xmath23 10% m@xmath8 .
the lightcurve of the tde from sdssj120136.02 + 300305.5 ( sect .
2.2 , fig .
4 ) does show the overall downward trend expected after tidal disruption .
however , one month after the peak , the x - ray emission suddenly dropped by a factor @xmath24 within a week and the source was no longer detected by _
swift_. x - rays re - appeared after 115d , and then dropped a second time .
if the intermittence of the x - ray emission was due to absorption , the first disappearance in x - rays would require a huge column density of gas of order @xmath25 @xmath26 .
while in agn , a broad - line cloud could be the source of such absorption , the optical spectrum of sdssj120136.02 + 300305.5 is that of a non - active galaxy , without any detectable emission lines neither before nor after the outburst .
the chances of molecular clouds in the very core repeatedly crossing our line - of - sight toward the x - ray source are therefore very low .
further , no jet was launched by this tde , since no radio emission was detected at all ( saxton et al .
2012 ) .
however , the characteristic intermittence and recovery of the lightcurve of sdssj120136.02 + 300305.5 is reminiscent of predictions by liu et al .
( 2009 ) , who computed tde lightcurves in binary smbhs . in that case
, the second smbh acts as a perturber and temporarily interrupts the accretion stream on the primary .
simulations by liu et al .
( 2014 ) have shown , that the lightcurve of sdssj120136.02 + 300305.5 is well consistent with a binary smbh model ( fig .
4 ) with a primary mass of @xmath27 m@xmath8 , a mass ratio @xmath28 and semi - major axis of 0.6 milli - pc ( mpc ) .
this is the first supermassive binary bh ( smbbh ) candidate identified in a non - active host galaxy , and the one with the most compact orbit among the known smbbh candidates ( review by komossa & zensus 2015 ) .
it has overcome the `` final parsec problem '' ( e.g. , colpi 2014 ) .
upon coalescence , it will be a strong source of gravitational wave emission in the sensitivity range of space - based gravitational wave detectors . if significant numbers of smbbhs exist at the cores of non - active galaxies , we expect to see more such events in _ swift _ lightcurves of tdes .
a good lightcurve coverage is essential for constraining the system parameters .
the _ galaxy evolution explorer _ ( _ galex _ ; martin et al .
2005 ) mission performed the first sky survey in the uv from space , and was well suited for a search of tdes at uv wavelengths .
three events were first identified in the uv .
the first one was detected in the _ galex _ groth field and showed uv variability by a factor of a few ( gezari et al .
an elliptical galaxy at redshift @xmath29=0.370 was identified as host .
its spectrum shows only faint emission from [ oiii ] and is dominated by stellar absorption features .
two more events were identified with _ galex _
( gezari et al .
2008 , 2009 ; our tab .
1 ) . in the optical bandpass ,
several photometric sky surveys have become online recently , performing wide - field transient surveys at high cadence [ e.g. , ptf ( law et al . 2009 ) , pan - starrs ( kaiser et al . 2002 ) , and for selected areas also sdss ( york et al .
these have found several tde candidates based on their optical variability , with lightcurves ( or spectra ) which look different from those of known supernovae . the first event identified optically was from sdss ,
associated with the galaxy sdssj095209.56 + 214313.3 ( komossa et al .
it shows only mild optical variability of @xmath21 mag , but was noticed for its transient optical emission lines which imply a luminous high - energy flare ( sect .
4 ) . van velzen et al . (
2011 ) performed a search for optical flares using the sdss stripe 82 data base , and presented two probable tdes , optically variable by factors 2 2.5 .
one of them showed evidence for variable broad - line emission .
further optical flares , some followed up with _ swift uvot _ , were reported by cenko et al .
( 2012a ) , gezari et al .
( 2012 ) , holoien et al .
( 2014 ) , vinko et al .
( 2015 ) , arcavi et al .
( 2014 ) , and chornock et al .
several of them come with variable optical emission lines ( further discussed in sect .
4 ) , while the spectrum of ps1 - 11af ( chornock et al . 2014 ) is featureless except for two deep uv absorption features . overall , the uv and optical events have spectral energy distributions that are peaked at much lower temperatures ( @xmath30 k ) than the x - ray events ( @xmath31 k ) , and the majority of the low - temperature events did not have detectable x - ray emission ( even though for the higher - redshift uv sources limits on x - ray emission , including in low - state , are not as tight as those for the more nearby _ rosat _ , _ chandra _ and _ xmm - newton _ x - ray events ) .
when travelling through any gaseous material of the galaxy core , the bright euv - x - ray continuum of a tde is reprocessed into emission lines .
these emission lines provide us a with a powerful new opportunity of measuring the physical conditions in the circum - nuclear material .
when the luminous electromagnetic radiation travels across the galaxy core , it will photoionize any circum - nuclear material ( rees 1998 , ulmer 1999 ) and the tidal debris itself ( bogdanovic et al . 2004 , strubbe & quataert 2011 ) and is reprocessed into line radiation .
this emission - line signal enables us to perform reverberation mapping of any gaseous material in the galaxy core , including , in principle , high - density gas like the broad - line region ( if present ) , the molecular torus , and the interstellar medium
. we will also be able to address , which of these regions are permanently present in non - active galaxies .
the emission - line fluxes , line widths , line shifts , and their evolution with time , tightly constrain the amount , density , composition , dynamics and geometry of the circum - nuclear material .
while the highest - ionization iron lines were very bright initially ( sdss spectrum , red ) , they were much fainter in a _ xinglong _ spectrum ( black ) taken several years later ( komossa et al .
2008 ; see wang et al .
2012 for several more cases ) . for examples which lack the narrow , high - ionization lines , but still show bright and broad emission of balmer lines and/or heii , see , wang et al .
2011 ( their fig .
1 ) , gezari et al . 2012 ( their fig .
1 ) , and arcavi et al . 2014 ( their fig .
4 ) . , width=336 ] recently , large optical spectroscopic surveys ( sdss ) , and follow - ups of sources identified in photometric transient surveys ( panstarrs , ptf , asas - sn ) have enabled the discovery of several well - observed cases of transient optical emission lines , arising from otherwise quiescent galaxies ( as judged from their low - ionization emission - line ratios ; or from the absence of any narrow emission lines ) .
all of them exhibit bright , broad , fading emission from helium and/or hydrogen ( komossa et al .
2008 , 2009 , wang et al .
2011 , 2012 , van velzen et al .
2011 , gezari et al .
2012 , gaskell & rojas lobos 2014 , holoien et al .
2014 , arcavi et al .
2014 ) , while some of them additionally show transient super - strong iron coronal lines up to ionization stages of fe@xmath32 including transitions of [ fevii ] , [ fex ] , and [ fexiv ] ( komossa et al .
2008 , 2009 , wang et al .
2011 , 2012 , yang et al .
all events have been interpreted as candidate tdes .
lcll source name & redshift & discovery mission & reference + + ngc5905 & 0.011 & _ rosat _ & bade et al .
1996 , komossa & bade 1999 + rxj12421119 & 0.050 & _ rosat _ & komossa & greiner 1999 + rxj1624 + 7554 & 0.064 & _ rosat _ & grupe et al .
1999 + rxj1420 + 5334 & 0.147 & _ rosat _ & greiner et al .
2000 + ngc3599 & 0.003 & _ xmm - newton _ & esquej et al . 2007 , 2008
+ sdssj1323 + 4827 & 0.087 & _ xmm - newton _ & esquej et al .
2007 , 2008 + tdxf13473254 & 0.037 & _ rosat _ & cappelluti et al .
2009 + sdssj1311 - 0123 & 0.195 & _ chandra _ & maksym et al .
2010 + 2xmmi18476317 & 0.035 & _ xmm - newton _ & lin et al .
2011 + sdssj1201 + 3003 & 0.146 & _ xmm - newton _ & saxton et al .
2012 + wingsj1348 & 0.062 & _ chandra _ & maksym et al .
2013 , donato et al .
2014 + rbs1032 & 0.026 & _ rosat _ & maksym et al . 2014b , khabibullin & sazonov 2014 + 3xmmj1521 + 0749 & 0.179 & _ xmm - newton _
& lin et al . 2015
+ + swiftj1644 + 57 & 0.353 & _ swift _ & bloom et al .
2011 , burrows et al .
2011 , + & & & levan et al . 2011 ,
zauderer et al .
2011 + swiftj2058 + 0516 & 1.186 & _ swift _ & cenko et al . 2012b
+ + j1419 + 5252 & 0.370 & _ galex _ & gezari et al .
2006 + j02250432 & 0.326 & _ galex _ & gezari et al .
2008 + j2331 + 0017 & 0.186 & _ galex _ & gezari et al .
2009 + + sdssj0952 + 2143@xmath33 & 0.079 & sdss & komossa et al .
2008 + sdssj0748 + 4712@xmath33 & 0.062 & sdss & wang et al .
2011 + sdssj2342 + 0106 & 0.136 & sdss & van velzen et al .
2011 + sdssj23230108 & 0.251 & sdss & + ptf10iya & 0.224 & ptf & cenko et al .
2012a + sdssj1342 + 0530@xmath33 & 0.034 & sdss & wang et al .
2012 + sdssj1350 + 2916@xmath33 & 0.078 & sdss & + ps1 - 10jh & 0.170 & pan - starrs & gezari et al .
2012 + asassn-14ae & 0.044 & asas - sn & holoien et al .
2014 + ptf09ge & 0.064 & ptf & arcavi et al .
2014 + ptf09axc@xmath34 & 0.115 & ptf & + ptf09djl & 0.184 & ptf & + ptf10nuj@xmath35 & 0.132 & ptf & + ptf11glr@xmath35 & 0.207 & ptf & + ps1 - 11af & 0.405 & pan - starrs & chornock et al .
2014 + luminous , transient high - ionization lines with ionization potential in the soft x - ray regime implied the presence of a high - amplitude high - energy outburst ( even though simultaneous x - ray observations did not exist ) .
the most luminous coronal lines are those from sdssj095209.56 + 214313.3 .
the highest ionization lines have faded by a factor @xmath36 several years after high - state and are no longer detectable .
the broad h@xmath37 line is asymmetric and comes with an initial redshift of 600 km / s , and additional multi - peaked narrow components are present ( komossa et al .
the highest - ionization lines , as well as broad h and he lines , continue fading over several years , while the lower - ionization lines like [ oiii ] increase in strength ( komossa et al .
2009 , wang et al .
2012 , yang et al .
2013 ) , consistent with light travel time effects , i.e. , flare emission reaching more extended regions of gaseous material / ism as time goes by .
the broad emission lines from h and he may arise from the tidal debris itself which forms the accretion disk ( see guillochon et al .
2014 , bogdanovic et al .
2014 , and gaskell & rojas lobos 2014 for models of ps1 - 10jh ) , while line emission from the unbound stellar streams is expected to be very faint . repeated spectral coverage shortly after peak is available for two of the events identified in the ptf archive ( arcavi et al .
2014 ) , and one event identified in the asas - sn survey ( holoien et al .
2014 ) along with optical lightcurves .
the balmer emission lines show complex , variable profiles with kinematic shifts as high as several 1000 km / s .
likely , all events are related , and those with fainter broad lines present a later stage in evolution ( while strong narrow high - ionization lines only appear in gas - rich environments ) . a future test of the tde interpretation will consist of searching for similar events at off - nuclear locations .
while recoiling smbhs and ongoing galaxy mergers will produce off - nuclear tdes ( komossa & merritt 2008 , liu & chen 2013 ) , these events would be rare in the local universe , and the vast majority of nearby tdes should therefore arise from the very cores of their host galaxies .
a new chapter in the study of tdes was opened when _ swift _ detected the unique transient grb110328a/ swiftj164449.3 + 573451 ( swiftj1644 + 57 hereafter ) .
it was first noticed when it triggered the _ swift _ burst alert telescope ( bat ) in march 2011 ( cummings et al .
2011 ) , and initially shared similarities with a grb .
however , the x - rays by far did not fade as quickly as expected for grbs . the rapid rise , huge x - ray peak luminosity , long duration , compact and variable associated radio emission , and optically inactive host galaxy all contributed to the interpretation of this event as the launch of a powerful jet following the tidal disruption of a star ( bloom et al .
2011 , burrows et al .
2011 , zauderer et al .
2011 , levan et al .
2011 ) .
the early evolution of the lightcurve is characterized by recurrent high - amplitude flares with ( isotropic ) peak luminosities exceeding 10@xmath38 erg / s ( burrows et al .
after the first few days , the lightcurve shows an overall decline ( fig . 6 , 7 ) , but continues to be complex and highly variable , with variability timescales as short as @xmath2100 s. optical imaging and spectroscopy revealed the host galaxy at redsift @xmath39 ( levan et al .
the optical emission lines imply that the host is not an agn , but an hii - type galaxy .
variable emission was also detected in the nir ( not in the optical ; likely due to the excess extinction seen in the optical spectrum ) .
various estimates of the smbh mass , based on scaling relations or the most rapid variability timescale all give @xmath40 @xmath23 @xmath41 m@xmath4 .
swiftj1644 + 57 was also associated with bright , compact , variable radio synchrotron emission ( e.g. , zauderer et al .
2011 ) coincident with the nucleus of the host galaxy .
radio emission was already detected during the first radio follow - up observation of swiftj1644 + 57 four days after the first x - ray detection with _
tidal disruption of a star powering a relativistic jet pointing in our direction has been the widely favored explanation of this unique event .
given the relatively low mass of the smbh , the observed peak luminosity of swiftj1644 + 57 is up to a few orders of magnitude above the eddington limit , implying that either the accretion at peak was highly super - eddington , and/or that the emission was boosted due to a collimated relativistic jet .
a variety of follow - up observations of swiftj1644 + 57 have been carried out at all wavelengths ( x - rays : saxton et al .
2012 , reis et al .
2012 , castro - tirado et al .
2013 , zauderer et al .
2013 , gonzales - rodriguez et al .
2014 , mangano et al .
2015 ; radio : berger et al .
2012 , zauderer et al .
2013 , cendes et al . 2014 ; ir and radio polarimetry : wiersema et al .
non - detections were reported with _
veritas _ and _ magic _ at energies @xmath42100 gev ( aliu et al .
2011 , aleksic et al .
2013 ) . in x - rays ,
evidence for a quasi - periodicity of 200s was reported by reis et al .
( 2012 ; see also saxton et al .
unlike the long - term x - ray lightcurve , the radio emission from swiftj1644 + 75 continued to rise ( berger et al .
2012 , zauderer et al . 2013
; our fig .
the _ swift _ x - ray lightcurve of swiftj1644 + 57 continued declining ( fig .
7 ) , but then showed a sudden drop at @xmath2500 days by a factor of @xmath2170 ; no longer detectable with _ swift _ but still with _ chandra _ ( zauderer et al .
2013 ) and _ xmm - newton _ in deeper pointings ( gonzalez - rodriguez et al .
the rapid decline is not seen in the radio ( fig .
8) , implying that x - ray and radio emission have different sites of origin also at late times .
zauderer et al .
( 2013 ) interpreted the drop in x - ray emission as evidence for a change in accretion mode , turning off the jet production ; while the faint late - stage x - rays themselves ( and the ongoing radio emission ) are consistent with arising from the forward shock related to the jet .
curves_. ( see also zauderer et al . 2013 , and see mangano et al .
2015 for a detailed spectral and timing analysis of the whole _ swift _
lightcurve.),width=332 ] the discovery of swiftj1644 + 57 has triggered a large number of theoretical studies , addressing the trajectory and type of the disrupted star and mass and spin of the smbh ( shao et al .
2011 , miller & gltekin 2011 , lei & zhang 2011 , cannizzo et al .
2011 , krolik & piran 2011 , abramowicz & liu 2012 ) , the nature of the radio emission and general implications for the formation of radio jets ( metzger et al .
2012 , wang & cheng 2012 , krolik & piran 2012 , de colle et al .
2012 , cao & wang 2012 , gao 2012 , barniol duran & piran 2013 , kumar et al .
2013 , zou et al .
2013 , tchekhovskoy et al .
2014 , wang et al .
2014 , kelley et al .
2014 , parfrey et al .
2015 , liu et al .
2015 , mimica et al . 2015
; see also comments by van velzen et al .
2011b , giannios & metzer 2011 ) , the evolution of the accretion disk including effects of precession ( stone & loeb 2012b , socrates 2012 , lei et al .
2013 , kawashima et al .
2013 , coughlin & begelman et al .
2014 , shen & matzner 2014 ) , and a few alternative interpretations of the event ( ouyed et al .
2011 , quataert & kasen 201 , woosley & heger 2012 ) . a key question which has been raised ( first mentioned in application to swiftj1644 + 57 by bloom et al .
2011 ) , regards the role of magnetic fields in magnetohydrodynamic models of jet formation , and specifically the question , whether the required large - scale magnetic field is generated _ in situ _ in the disk , or is rather advected in with the flow .
the magnetic field strength estimated for swiftj1644 + 57 is much higher than that of a main - sequence star , and must have then been generated locally in the disk , or requires the presence of a fossil disk .
regarding the source of the long - term radio emission there is overall concencus that it is very likely synchrotron radiation from the shock which is formed when the jet interacts with the interstellar medium .
another focus of attention has been the nature of the x - ray emission and its rapid variability and early epochs of high - amplitude flaring .
most models have linked it with dissipation in the inner jet ( and/or effects of jet precession or nutation ) , and/or the forward shock , rather than directly with emission from the accretion disk .
finally , a topic of close consideration has been the trajectory and type of the disrupted star .
while several studies have assumed , or explored , the disruption of a main - sequence star , arguments have also been made in favor of the disruption of a white dwarf ( krolik & piran 2011 ) , a giant star ( shao et al .
2011 ) , or a star with a deeply plunging orbit ( cannizzo et al .
2011 ) .
evidence for a possible second jetted tde discovered by _
swift _ was reported by cenko et al ( 2012b ) .
swiftj2058.4 + 0516 ( swiftj2058 + 05 hereafter ) exhibited a luminous , long - lived x - ray outburst with an ( isotropic ) peak luminosity of @xmath43 erg / s .
its x - ray lightcurve shows rapid variability as fast as 1000s .
the event is accompanied by strong radio emission . the likely host galaxy of swiftj2058 +
05 is at redshift @xmath44 , and is optically inactive . because of the many similarities with swiftj1644 + 57 , cenko et al .
suggested a similar outburst mechanism , consistent with multi - wavelength follow - up observations ( pasham et al .
notably , like swiftj1644 + 57 , the x - ray lightcurve of swiftj2058.4 + 0516 shows an abrupt drop ( between days 200 - 300 ) .
since the x - rays vary rapidly before the drop , an origin near the smbh rather than in a forward shock is implied ( pasham et al . 2015 ) .
the possibility that tdes launch radio jets , came up with the detection of the first few x - ray tdes with _ rosat_. in order to test , whether tdes do launch radio jets ( and to exclude a peculiar mini - blazar at the galaxy core ) , radio follow - ups of ngc5905 were performed with the vla ( komossa 2002 ) 6 years after the initial x - ray outburst .
no radio emission was detected ( @xmath45 mjy at 8.46 ghz ; see sect .
2.3 of komossa 2002 for an extended discussion of the radio properties of ngc5905 ) .
more recently , deep radio follow - ups have been carried out of some of the newly identified ( candidate ) tdes ( e.g. , van velzen et al . 2011 , saxton et al .
2012 , cenko et al .
2012 , saxton et al .
2014 , chornock et al .
no radio emission was detected .
radio observations were also carried out for several of the previously known tdes , covering time intervals of up to more than a decade ( the _ rosat _ events ) , and years ( bower et al . 2013 , van velzen et al . 2013 ) .
most sources were undected with the vla in either study , with upper limits on the flux density in the range @xmath46jy .
the two exceptions ( bower et al . 2013 ) are the seyfert galaxy ic3599 [ the radio emission in this agn is most likely linked to the long - lived seyfert activity of this galaxy , and the nature of its ( repeat ) x - ray flaring remains debated ; e.g. , komossa et al .
2015 , see our sect . 7 ] and a source in the x - ray error circle of the _ rosat _ event rxj1420 + 5334 .
jets similar to the one of swiftj1644 + 57 should have been detected in the deepest radio follow - ups , even if not pointed at us .
therefore , most tdes do not launch powerful radio jets .
the fact that swiftj1644 + 57 does , has re - raised the question , which parameter(s ) determine jet formation . in application
to swiftj1644 + 57 , bh spin or the pre - existence of a fossil disk or other mechanisms to produce a large magnetic flux to power the jet have been considered ( sect .
several approaches have been followed to estimate tde rates from observations , making use of the _ rosat _ all - sky survey ( donley et al .
2002 , khabibullin & sazonov 2014 ) , the _ xmm - newton _ slew survey ( esquej et al . 2008 ) , upper limits from _
deep fields ( luo et al . 2008 ) , clusters of galaxies ( maksym et al . 2010 ) , galex ( gezari et al . 2008 ) , and sdss ( e.g. , wang et al .
2012 , van velzen & farrar 2014 ) .
all rates are in the range @xmath47/yr / galaxy , and consistent with order - of - magnitude theoretical predictions which are on the same order ( e.g. , rees 1990 , merritt 2009 , review by alexander 2012 ) .
tdes might well occur in agn , too , and rates are expected to be high ( karas & subr 2007 ) .
however , it is generally more challenging to identify tdes in agn because of the much lower contrast of any tde - related accretion flare relative to the permanently bright accretion disk .
further , it is more difficult to uniquely associate any particular flare in classical agn ( i.e. , galaxies which show long - lived activity in x - rays , and/or radio , and characteristic bright emission - lines from the narrow - line region ) with a tde , because processes in the long - lived accretion disk itself can potentially cause high - amplitude variability .
nevertheless , a few cases of unusually high - amplitude variability of agn have been considered in the context of a tidal disruption scenario .
\(1 ) the seyfert galaxy ic3599 showed an x - ray outburst during the _ rosat _ all - sky survey , accompanied by a temporary remarkable brightening and fading of the optical emission lines ( brandt et al .
1995 , grupe et al .
1995a , komossa & bade 1999 ) .
the cause of the outburst remained unknown , and high - amplitude narrow - line seyfert 1 ( nls1 ) variability , a disk instability , or a tde have all been considered .
ic3599 was already known as an active galaxy before the outburst ( based on its narrow emission lines like [ oiii ] ) .
recently , a second outburst of ic3599 was discovered by _
swift _ in 2010 ( komossa et al . 2015 , campana et al . 2015 , grupe et al .
recurrent flaring within decades is uncommon in tde scenarios , even though not impossible , since , for instance , rates are strongly boosted under some conditions in galaxy mergers ( e.g. , chen et al .
2009 ) , and repeat tidal stripping can occur for orbiting stars surviving the first encounter .
this latter possibility was explored by campana et al .
( 2015 ) , while processes related to the agn accretion disk ( and smbbhs ) where discussed by komossa et al .
( 2015 ) and grupe et al .
( 2015a ) .
variability of similar magnitude as ic3599 ( a factor @xmath48 ) is rare , but has also been observed in a few other agn4422 ( factor 60 , grupe et al .
2015b ) ] , likely related to processes in the long - lived agn accretion disk ( and this therefore is a likely explanation of ic3599 , too ) , or to absorption .
\(2 ) in the uv , renzini et al . ( 1995 ) observed faint , enhanced emission from the core of the elliptical radio galaxy ngc4552 with _
hst _ , and suggested a mild accretion event from an interstellar cloud or the tidal stripping of a star ; or fluctuations in the accretion rate of the very faint agn in ngc4552 ( cappellari et al . 1999 , renzini 2001 ) . the uv spike varied by a factor of @xmath49 on the timescale of several years .
very generally , maoz et al .
( 2005 ) found , that nuclear uv variability in liner galaxies with radio cores is very common .
( 3 ) meusinger et al .
( 2010 ) observed a high - amplitude ( factor 20 ) uv flare from the quasar j004457 + 4123 , which they interpreted as either a microlensing event or else a tde .
( 4 ) nikolajuk & walter ( 2013 ) reported the discovery of a hard x - ray flare with _ integral _
, associated with the seyfert galaxy ngc4845 .
because of the high amplitude of variability , the authors strongly favored a tidal disruption scenario ; in this case the disruption of a giant planet .
( 5 ) the galaxy xmmj061927.1@xmath9655311 was found in a high - amplitude flaring state ( factor @xmath42140 ) with _ xmm - newton _
( saxton et al . 2014 ) at @xmath50 erg / s .
optical follow - up spectroscopy revealed low - level agn activity ( even though the spectrum looks peculiar and narrow emission lines are very faint ) . the x - rays and uv , observed with _ swift _ , declined subsequently .
the event was interpreted as either a tde or a change in accretion rate in a long - lived agn .
upcoming sky surveys will find tdes in the hundreds or thousands , including in the radio with ska ( donnarumma et al .
2015 ) , in the optical with lsst ( van velzen et al . 2011 ) , and in hard x - rays with loft ( rossi et al . 2015 ) .
further , a dedicated x - ray transient mission is under consideration in china , which has among its prime goals the detection and study of tdes in large numbers ; the soft x - ray mission _
einstein probe _ ( zhao et al . 2014 , yuan et al . 2015 ) .
_ einstein probe _ is designed to carry out an all - sky transient survey at energies of 0.54 kev .
the concept is based on a wide - field micro - pore lobster - eye imager ( 60@xmath51 @xmath52 60@xmath51 ) , and a more sensitive narrow - field instrument for follow - ups .
the large number of new events and the well - covered lightcurves will enable a wealth of new science .
in particular , x - rays will be sensitive to relativistic effects and accretion physics near the last stable orbit .
it is my pleasure to thank andrew levan , shuo li , and ashley zauderer for providing figures , fukun liu for numerous discussions on tdes , eduardo ros for a critical reading of the manuscript , and the anonymous referee for useful comments .
i would like to thank the aspen center for physics for support and hospitality during a workshop in 2012 , and the participants for illuminating discussions on tdes .
the aspen center for physics is supported by the national science foundation under grant no . | stars in the immediate vicinity of supermassive black holes ( smbhs ) can be ripped apart by the tidal forces of the black hole .
the subsequent accretion of the stellar material causes a spectacular flare of electromagnetic radiation . here
, we provide a review of the observations of tidal disruption events ( tdes ) , with an emphasis on the important contributions of _ swift _ to this field .
tdes represent a new probe of matter under strong gravity , and have opened up a new window into studying accretion physics under extreme conditions .
the events probe relativistic effects , provide a new means of measuring black hole spin , and represent signposts of intermediate - mass bhs , binary bhs and recoiling bhs .
luminous , high - amplitude x - ray flares , matching key predictions of the tidal disruption scenario , have first been discovered with _ rosat _ , and more recently with other missions and in other wavebands .
the _ swift _ discovery of two @xmath0-ray emitting , jetted tdes , never seen before , has provided us with a unique probe of the early phases of jet formation and evolution , and swiftj1644 + 75 has the best covered lightcurve of any tde to date .
further , _
swift _ has made important contributions in providing well - covered lightcurves of tdes discovered with other instruments , setting constraints on the physics that govern the tde evolution , and including the discovery of the first candidate binary smbh identified from a tde lightcurve . |
manganites rmn@xmath0o@xmath1 ( r are rare earth ions , y and bi ) are typical representatives of multiferroics in which ferroelectric ( with @xmath8 k ) and magnetic ( with @xmath9 k ) orderings coexist @xcite . it has been assumed until now that rmn@xmath0o@xmath1 has a centrosymmetric sp .
. _ pbam _ which forbids the existence of electric polarization .
the ferroelectric ordering along the @xmath10 axis at low temperatures is induced by charge and magnetic orders which break the central symmetry of the crystal due to the exchange - striction mechanism @xcite .
it has recently been shown @xcite that the real rmn@xmath0o@xmath1 symmetry at room temperature is a noncentral symmetry which means that the electric polarization should exist at room temperature .
the authors could not give preference to one of two possible space groups : _ p2 _ ( allowing polarization along the @xmath11 axis ) and _ pm _ ( allowing polarization in the @xmath12 plane ) on the basis of structural studies . by assuming from the physical considerations that the _ homogeneous _ electric polarization could exist at all temperatures only in one direction ( i.e. , along the @xmath10 axis ) , they preferred the sp .
_ pm_. the authors @xcite did not discuss the nature of this polarization . in @xcite it was reported that the hysteresis loops of electric polarization and remanent polarization were detected in gdmn@xmath0o@xmath1 in a wide temperature range 5 330 k. we interpreted this polar state as a frozen superparaelectric state induced by _
local polar domains _ formed in the main crystal matrix due to phase separation and charge carriers self - organization .
theoretically , the frozen superparaelectric state was considered in the system of isolated ferroelectric nanoscale domains in a dielectric matrix @xcite .
this paper presents a comparative study of the electric polarization hysteresis loops which were revealed in two crystals of the rmn@xmath0o@xmath1 family with @xmath2 and @xmath3 in temperature interval from 5 k up to 330 k. there are significant structural distortions in bimn@xmath00@xmath1 caused by bi ions due to the presence of alone pairs of 6s@xmath13 electrons @xcite .
the bi@xmath6 ions are nonmagnetic ones but they strongly distort the lattice , while strongly magnetic gd@xmath6 ions with the ground state ( @xmath14s@xmath15 ) interact weakly with the lattice
. of particular interest is the correlation between the properties of phase separation domains and electrical polarizations in these two model crystals .
magnetization , electric polarization , permittivity , and conductivity studies and also features of high - resolution x - ray diffraction scattering of rmn@xmath0o@xmath1 ( @xmath2 , @xmath3 ) are reported .
a characteristic feature of rmn@xmath0o@xmath1 is the presence of equal amounts of manganese ions with different valences mn@xmath6 ( containing 3t@xmath16 , 1e@xmath17 electrons in the @xmath18 shell ) and mn@xmath7 ( with 3t@xmath16 , 0e@xmath17 electrons ) , which provides conditions for charge ordering .
the mn@xmath7 ions have an octahedral oxygen environment and are arranged in the layers with @xmath19 and @xmath20 .
the mn@xmath6 ions have a local non - central surrounding in the form of a pentagonal pyramid and are arranged in the layers with @xmath21 .
the r@xmath6 ions have the environment similar to that of the mn@xmath6 ions and are arranged in the layers @xmath22 @xcite .
charge ordering in rmn@xmath0o@xmath1 and the e@xmath17 electron transfer between mn@xmath6 -mn@xmath7 ion pairs ( double exchange @xcite ) are key factors responsible for polar electric states of these multiferroics at all temperatures .
the low - temperature ferroelectric ordering in rmn@xmath0o@xmath1 is caused by charge ordering of alternating mn@xmath6 - mn@xmath7 ion pairs with antiferromagnetically and ferromagnetically oriented spins along the @xmath10 axis .
different values of indirect exchange between the antiferromagnetic mn@xmath6 - mn@xmath7 ion pairs and considerably stronger double exchange between the ferromagnetic pairs of these ions lead to the exchange striction that breaks the centrosymmetricity of the lattice , thus giving rise to the ferroelectric polarization along the @xmath10 axis at @xmath23 @xcite .
the double exchange between mn@xmath6-mn@xmath7 ions induces phase separation in rmn@xmath0o@xmath1 which is similar to phase separation in lnamno@xmath24 ( a- sr , ba , ca ) manganites containing mn@xmath6 and mn@xmath7 ions as well .
phase separation exists at all temperatures and makes the formation of local conductive domains containing mn@xmath6 - mn@xmath7 ion pairs with ferromagnetically oriented spins energetically favorable .
the phase separation domains are located in a dielectric antiferromagnetic ( paramagnetic ) matrix of the original crystal . the size and geometry of the local phase separation domains
are determined by a dynamic balance of interactions which leads to the attraction of charge carriers ( double exchange , the jahn - teller effect ) and coulomb repulsion @xcite .
the effect of phase separation domains on the rmn@xmath0o@xmath1 properties was studied from 5 k up to the temperatures above room temperature . at @xmath25
the phase separation domains manifested themselves as @xmath26 superlattices with alternating ferromagnetic layers containing different numbers of charge carriers .
it was supposed that these superlattices were multiferroic domain walls ( with their own polarizations ) located between ferroelectric domains of the basic crystal matrix .
a set of ferromagnetic resonances from individual superlattice layers was detected for a number of multiferroics rmn@xmath0o@xmath1
( @xmath27 , @xmath28 , @xmath29 , @xmath30 , @xmath3 ) @xcite . in the paramagnetic and paraelectric temperature range , up to temperatures above room temperature ,
the polar nature of local phase separation domains was found to be responsible for dielectric properties and the features of the high - resolution x - ray diffraction spectra of rmn@xmath0o@xmath1 ( @xmath27 and @xmath28 ) single crystals @xcite .
the 2d superstructures of alternating layers of the initial rmn@xmath0o@xmath1 crystals and phase separation domains were revealed at sufficiently high temperatures .
most distinctly these superstructures ( perpendicular to the c axis ) manifested themselves in the x - ray studies of eumn@xmath0o@xmath1 and eu@xmath31ce@xmath32mn@xmath0o@xmath1 at room temperature .
the superstructure layer thicknesses were 900 and 700 , respectively @xcite .
single crystals of rmn@xmath0o@xmath1 were grown by the spontaneous crystallization technique described in @xcite .
the as - grown single crystals were in the form of 2 3 mm thick plates with areas of 3 5 mm@xmath13 . to measure the polarization , capacitors with a thickness of 0.3 0.6 mm and area of 3
4 mm@xmath13 were used .
the electric polarization was measured by two methods , i.e. , the thermo - stimulated pyrocurrent method and the so - called positive - up negative - down ( pund ) method of hysteresis loops measuring @xcite . in the first case the polarization was measured by a keithly 6514 electrometer and the measurements were produced during sample heating with a constant temperature variation rate after the preliminary sample cooling in a polarizing electric field which was switched off at the lowest temperature . in the second case we employed the version of the pund method presented in @xcite which was adapted to our measurements ( see fig . 1 in @xcite ) . if the sample had a relatively high conductivity ( which is important for rmn@xmath0o@xmath1 containing restricted polar phase separation domains with local conductivities ) , the shape of the polarization electric field (
p - e ) hysteresis loop was distorted and did not give correct information on the intrinsic polarization p. in the pund method only the hysteresis of p can be extracted by applying a series of voltage pulses to the sample . during successive positive p1-p2 and negative n1-n2 pulses , independent curves
( p1-p2 and n1-n2 ) of effective polarization p changes are registered .
the pund method is based on the difference between polarization and conductivity responses to variations in the field e. the time intervals between p1-p2 and n1-n2 pulses should be chosen such that the intrinsic polarization p is still unrelaxed , while the conductivity relaxation is complete . in the conventional volume ferroelectrics with the domain structure those time intervals may be up to seconds . in our case the intrinsic p response was determined by the restricted polar phase separation domains which rather rapidly restored after the field e was switched off .
the reason for this will be discussed below .
o@xmath1 ( a ) and bimn@xmath0o@xmath1 ( b).,title="fig:",scaledwidth=45.0% ] o@xmath1 ( a ) and bimn@xmath0o@xmath1 ( b).,title="fig:",scaledwidth=45.0% ] as a result , the time intervals between the p1-p2 and n1-n2 pulses were chosen so that the responses to the p1 and n1 pulses were irreversible ( due to the intrinsic polarization contribution ) and the conductivity responses to the p2 and n2 pulses were closed and reversible ( see fig . [ fig1 ] in @xcite ) .
this could be achieved if the time intervals between the p1-p2 and n1-n2 pulses did not exceed 0.8 ms .
the p1 and n1 curves characterized the total polarization and conductivity contributions . to obtain the actual p - e loop , we subtracted the p2 and n2 curves from the p1-n1 curves .
the durations of pulses in the p1-p2 and n1-n2 series were 2 ms and 4.5 ms , the intervals between the pulse series were 4 ms and 9 ms for gdmn@xmath0o@xmath1 and bimn@xmath0o@xmath1 , respectively .
the permittivity and conductivity were measured by a good will lcr-819 impedance meter in the frequency range 0.5 50 khz at 5 350 k. magnetic properties were investigated by ppms ( quantum design ) .
the intensity distributions of bragg reflections were studied with a high - sensitivity 3-crystal x - ray diffractometer .
complicated magnetic structures with wave vector @xmath33 and a number of low - temperature magnetic phase transitions at @xmath34 at which incommensuration parameters @xmath35 and @xmath36 change in a step - like fashion are typically observed in rmn@xmath0o@xmath1 with different r ions @xcite . as temperature decreases ,
an incommensurate magnetic phase is appeared near t@xmath37 . at a lower temperature this phase is transformed into a commensurate structure .
as temperature further decreases , one more transition into the incommensurate phase is possible .
ferroelectric ordering in rmn@xmath0o@xmath1 is typically occurs in the intermediate commensurate phase .
magnetic structures of gdmn@xmath0o@xmath1 and bimn@xmath0o@xmath1 differ from those commonly observed in rmn@xmath0o@xmath1 . in gdmn@xmath0o@xmath1 , a commensurate collinear antiferromagnetic structure with wave vector @xmath38
occurs in the temperature range 0 30 k @xcite . at these temperatures there
is a ferroelectric ordering with @xmath39 k. in the temperature range 35 k@xmath40 k there is an incommensurate phase . in bimn@xmath0o@xmath1 the commensurate noncollinear antiferromagnetic structure with @xmath41
is observed at t@xmath42 k @xcite . along the @xmath10 axis at @xmath43 for gdmn@xmath0o@xmath1 and bimn@xmath0o@xmath1 ( in the inset).,scaledwidth=45.0% ]
[ fig1 ] shows temperature dependence of magnetization in magnetic field @xmath44 koe along different axes of the gdmn@xmath0o@xmath1 ( fig .
[ fig1]a ) and bimn@xmath0o@xmath1 ( fig . [ fig1]b ) crystals studied .
it can be seen that the magnetic moment in gdmn@xmath0o@xmath1 is much higher due to the contribution of gd@xmath6 ions with the @xmath45 spin which have their own ordering at @xmath46 k. on the background of magnetization of gd ions , the antiferromagnetism of manganese ions along the a axis ( with @xmath47 k ) manifests itself only slightly ( see the left inset in fig .
[ fig1]a ) .
the right inset in fig .
[ fig1]a shows that the curie - weiss temperature for gdmn@xmath0o@xmath1 nearly coincides with t@xmath37 .
this means that there is a distinct phase transition near t@xmath37 in gdmn@xmath0o@xmath1 ( the magnetic state is not frustrated ) .
a quite different magnetic state is observed in bimn@xmath0o@xmath1 ( fig .
[ fig1]b ) .
the antiferromagnetism along the @xmath11 axis manifests itself only slightly , while it is pronounced along the @xmath48 axis .
unlike in gdmn@xmath0o@xmath1 , but similar to rmn@xmath0o@xmath1 with other r ions , the curie - weiss temperature exceeds @xmath49 k in 6 - 7 times .
this means that the magnetic state in bimn@xmath0o@xmath1 is highly frustrated , and spin correlations in local domains are observed at @xmath50 .
specific magnetic and structural properties of bimn@xmath0o@xmath1 in the rmn@xmath0o@xmath1 family are caused by the presence of electron alone pairs in bi@xmath6 ions .
high - resolution neutron powder diffraction studies @xcite show that bimn@xmath0o@xmath1 is characterized at room temperature by the orthorhombic sp .
. _ pbam _ , similar to other rmn@xmath0o@xmath1 .
however , mn@xmath7-o6 octahedra , mn@xmath6-o5 pyramids , and bi@xmath6-o units are more deformed as compared with those in other rmn@xmath0o@xmath1 . as a result ,
the mn - o distances and magnetic exchange bonds differ as well . as will be shown below , differences in magnetic and structural states of gdmn@xmath0o@xmath1 and bimn@xmath0o@xmath1 lead to different electric polarizations at all temperatures .
let us consider at first the low - temperature polar states .
[ fig2 ] shows dispersion - free anomalies of real permittivity @xmath51 along the b axis which are characteristic of the low - temperature ferroelectric phase transition . as one can see , the value of @xmath51 at the maximum in the vicinity of t@xmath52 in gdmn@xmath0o@xmath1 considerably exceeds the maximum in bimn@xmath0o@xmath1 .
[ fig3 ] shows electric polarizations along the @xmath10 axis ( p@xmath53 ) in gdmn@xmath0o@xmath1 ( fig.[fig3]a ) and bimn@xmath0o@xmath1 ( fig.[fig3]b ) .
the polarizations are measured by the pyrocurrent method on the assumption that p@xmath54 at t@xmath55 .
the maximum polarization p@xmath53 ( in our case @xmath56 @xmath57c / cm@xmath13 ) for the rmn@xmath0o@xmath1 family is observed in gdmn@xmath0o@xmath1 @xcite . in bimn@xmath0o@xmath1 ,
polarization p@xmath53 is an order of magnitude lower and is closer to the values typically observed in other rmn@xmath0o@xmath1 @xcite . and pyrocurrents along the @xmath10 axis at @xmath59 and @xmath60 for gdmn@xmath0o@xmath1 ( a ) and bimn@xmath0o@xmath1 ( b).,title="fig:",scaledwidth=45.0% ] and pyrocurrents along the @xmath10 axis at @xmath59 and @xmath60 for gdmn@xmath0o@xmath1 ( a ) and bimn@xmath0o@xmath1 ( b).,title="fig:",scaledwidth=45.0% ] note that a fully polarized state is often observed in structurally perfect ( according to x - ray diffraction ) rmn@xmath0o@xmath1 single crystals with different r ions in the absence of an applied external electric field e @xcite .
this indicates that this fully polarized p@xmath53 polarization is formed in the internal field .
we suppose that this is a staggered - like field caused by charge ordering of mn@xmath6 - mn@xmath7 ion pairs along the @xmath10 axis , and the polarization has an exchange - striction nature . this situation is observed for both crystals we studied . in gdmn@xmath0o@xmath1 ( fig .
[ fig3]a ) , the maximum polarization is detected in @xmath59 .
a uniform external field @xmath61 which is much weaker than the internal staggered - like field , only slightly reduces polarization p@xmath53 and can not reverse its orientation ( fig .
[ fig3]a ) .
the application of a strong field @xmath62 ( comparable to the internal field ) enhances the transfer of e@xmath17 valence electrons between the mn@xmath6 -mn@xmath7 ion pairs along the @xmath10 axis , which leads to the electric breakdown of the crystal .
therefore , measurements of the polarization hysteresis loops p@xmath53 ( which are the response to an applied external field e ) is not effective when the striction contribution to the gdmn@xmath0o@xmath1 polarization at @xmath63 is measured .
o@xmath1 ( left panels ) at different temperatures at e@xmath64 ( a ) , e@xmath65 ( b ) , and e@xmath66 ( c ) .
temperature dependences of remanent polarization are shown in right panels.,title="fig:",scaledwidth=45.0% ] o@xmath1 ( left panels ) at different temperatures at e@xmath64 ( a ) , e@xmath65 ( b ) , and e@xmath66 ( c ) .
temperature dependences of remanent polarization are shown in right panels.,title="fig:",scaledwidth=45.0% ] o@xmath1 ( left panels ) at different temperatures at e@xmath64 ( a ) , e@xmath65 ( b ) , and e@xmath66 ( c ) .
temperature dependences of remanent polarization are shown in right panels.,title="fig:",scaledwidth=45.0% ] in bimn@xmath0o@xmath1 ( which are also structurally perfect single crystals ) , bi ions distorts the uniformityof the internal staggered - like field , thus reducing its value and , hence , the exchange - striction polarization p@xmath53 . in this case application of field
@xmath61 changes the internal field to a greater extent and can even reverse the bimn@xmath0o@xmath1 polarization ( see fig . [
as can be seen , the application of e in the same direction in which the polarization was observed at @xmath59 enhances the polarization by the value close to that detected in e of the opposite orientation ( fig .
[ fig3]b ) .
there is good reason to believe that the internal exchange - striction polarization is the polarization observed at @xmath59 .
there are two additional contributions to the polarization of bimn@xmath0o@xmath1 in the case of application of field @xmath62 .
there is the contribution originating from a weak change in the internal staggered - like - field and that due to the response of phase separation domains to the applied field e. note that the polarization measured by the pyrocurrent method includes the contribution of conductivities of phase separation domains .
as noted above , it was shown in @xcite that electric polarization hysteresis loops in gdmn@xmath0o@xmath1 were observed in a wide temperature interval ( from the lowest up to room temperature ) .
we put forward the hypothesis that this polarization was due to the frozen superparaelectric state of restricted polar phase separation domains formed in the initial crystal matrix .
phase separation must exist in rmn@xmath0o@xmath1 with any r ions at all temperatures , but the phase separation domain states depend on the r ion type and temperature .
thus , two electric polarizations having different natures ( the long - range ferroelectric order with polarization p@xmath53 caused by exchange striction and the polarization due to polar phase separation domains ) must coexist in rmn@xmath00@xmath1 along the b axis at @xmath23 .
the previous subsection was concerned with the polarizations p@xmath53 at @xmath23 measured by the pyrocurrent method at @xmath59 in gdmn@xmath0o@xmath1 ( fig.[fig3]a ) and bimn@xmath0o@xmath1 ( fig .
[ fig3]b ) .
this polarization gave no contribution to the hysteresis loops which were responses to the applied field e. on the contrary , the polarization caused by phase separation domains could be determined only by measuring the hysteresis loops by the pund method that excluded the contribution of conductivity . by using these two methods , we could separate the contributions of the polarizations considered above at low temperatures .
let us consider in more detail the hysteresis loops for bimn@xmath0o@xmath1 in a wide temperature interval and compare them with the loops for gdmb@xmath0o@xmath1 .
[ fig4 ] shows the p - e hysteresis loops of bimn@xmath0o@xmath1 in e oriented along the @xmath48 , @xmath10 , and @xmath11 axes ( left panels in figs .
[ fig4]a , [ fig4]b , and [ fig4]c , respectively ) . the right panels in these figures demonstrate temperature dependences of the remanent polarization ( @xmath67 ) .
o@xmath1 ( left panels ) at different temperatures at e@xmath64 ( a ) , e@xmath65 ( b ) , and e@xmath66 ( c ) .
temperature dependences of remanent polarization are shown in right panels.,title="fig:",scaledwidth=45.0% ] o@xmath1 ( left panels ) at different temperatures at e@xmath64 ( a ) , e@xmath65 ( b ) , and e@xmath66 ( c ) .
temperature dependences of remanent polarization are shown in right panels.,title="fig:",scaledwidth=45.0% ] o@xmath1 ( left panels ) at different temperatures at e@xmath64 ( a ) , e@xmath65 ( b ) , and e@xmath66 ( c ) .
temperature dependences of remanent polarization are shown in right panels.,title="fig:",scaledwidth=45.0% ] figs .
[ fig5]a , [ fig5]b , and [ fig5]c present the p - e hysteresis loops and prem for gdmn@xmath0o@xmath1 .
as one can see , the hysteresis loops are observed for both crystals along all three axes from 5 k up to some temperatures t@xmath68 which depend on the axis direction .
the maximum @xmath67 polarizations are observed along the @xmath11 axis from 5 k up to @xmath69 k in bimn@xmath0o@xmath1 and up to @xmath70 k for gdmn@xmath0o@xmath1 .
the minimum @xmath67 along the @xmath10 axis which exists up to @xmath71 k is observed for gdmn@xmath0o@xmath1 , while a twice higher @xmath67 along the @xmath10 axis for bimn@xmath0o@xmath1 manifests itself up to @xmath72 k. the @xmath67 along the @xmath48 axis are observed for gdmn@xmath0o@xmath1 and bimn@xmath0o@xmath1 up to @xmath73 k and @xmath69 k , respectively .
thus , @xmath67 and hysteresis loops for bimn@xmath0o@xmath1 and gdmn@xmath0o@xmath1 which demonstrate a strong anisotropy are revealed in the paramagnetic phase .
it is important to note that the polarizations of both crystals occurred along all crystal axes and disappeared at different temperatures along these axes .
this fact means that the polarizations measured in the hysteresis loops were not caused by the ferroelectric phase transition in the homogeneous ferroelectric crystal states .
as noted above , we assumed that these polarizations were induced by the restricted polar phase separation domains .
it turned out that these polarizations were significantly lower than the exchange - striction ferroelectric low - temperature polarization along the @xmath10 axis at @xmath23 .
the remanent polarization @xmath67 along the @xmath10 axis ( fig .
[ fig5]b ) in gdmn@xmath0o@xmath1 is two orders of magnitude lower than the exchange - striction contribution of the basic crystal matrix polarization @xmath74 at @xmath75 k measured by the pyrocurrent method in the same sample at @xmath59 ( fig . [ fig3]a ) .
the contribution of @xmath74 into the hysteresis loops manifests itself only in the form of maxima on the background of @xmath76 near @xmath77 when the internal staggered - like - field which gives rise to polarization @xmath74 begins to decrease and disappears and the @xmath74 fluctuations increase ( see fig .
[ fig5]b ) .
there is no exchange - striction polarization of the main matrix in the direction of the @xmath48 and @xmath11 axes at @xmath23 , and therefore contributions to the hysteresis loops come only from phase separation domains at all temperatures .
the fact that @xmath67 is independent of temperature along all axes of gdmn@xmath0o@xmath1 in a wide temperature range ( for the @xmath10 axis this relates to the background value minus the maxima near @xmath77 ) indicates that the contribution into the polarization from phase separation domains in gdmn@xmath0o@xmath1 is anisotropic but temperature - independent . at @xmath78
the remanent polarizations rather abruptly disappear along different crystal axes .
of interest is the fact that , unlike in gdmn@xmath0o@xmath1 , the @xmath67 values along all axes of the bimn@xmath0o@xmath1 crystal abruptly increase in the vicinity of @xmath79 k and then rapidly reduce to zero near the @xmath78 .
such @xmath67 jumps near 125 k indicate that phase separation domain states in bimn@xmath0o@xmath1 should change in a step - like fashion near this temperature .
the absence of such jumps in gdmn@xmath0o@xmath1 shows that most probably these jumps in bimn@xmath0o@xmath1 result from structural lattice distortions caused by the bi ions .
the effect of the longitudinal magnetic field h on the hysteresis loops was also studied .
as one can see from the right panels in fig .
[ fig5 ] , the field @xmath80 t increases in gdmn@xmath0o@xmath1 both the remanent polarization and temperature @xmath78 along the @xmath10 axis and only @xmath78 along the @xmath48 and @xmath11 axes .
characteristic activation barriers e@xmath81 must exist at the interfaces between phase separation domains and the main matrix of the crystal .
information on e@xmath81 along different crystal axes can be obtained from frequency - temperature dependences of real permittivity @xmath51 and conductivity @xmath82 . and conductivity @xmath82 ( in insets ) at different frequencies for bimn@xmath0o@xmath1 along the @xmath48 axis ( a )
, @xmath10 axis ( b ) , and @xmath11 axis ( c).,title="fig:",scaledwidth=45.0% ] and conductivity @xmath82 ( in insets ) at different frequencies for bimn@xmath0o@xmath1 along the @xmath48 axis ( a ) , @xmath10 axis ( b ) , and @xmath11 axis ( c).,title="fig:",scaledwidth=45.0% ] and conductivity @xmath82 ( in insets ) at different frequencies for bimn@xmath0o@xmath1 along the @xmath48 axis ( a ) , @xmath10 axis ( b ) , and @xmath11 axis ( c).,title="fig:",scaledwidth=45.0% ] fig .
[ fig6 ] shows the @xmath51(t ) and @xmath82(t ) dependences for bimn@xmath0o@xmath1 at different frequencies along the @xmath48 axis ( fig .
[ fig6]a ) , @xmath10 axis ( fig .
[ fig6]b ) , and @xmath11 axis ( fig .
[ fig6]c ) .
it is evident that the values of @xmath51 and @xmath82 along all axes of the crystal are nearly independent of temperature at @xmath83 k. near @xmath79 k a frequency - dependent step - like @xmath51 increase to some constant level ( most pronounced along the @xmath48 and @xmath10 axes ) begins .
as temperature further grows , @xmath51 starts to increase sharply again .
the frequency - dependent growth of @xmath82 also begins near 125 k , and two the different temperature intervals of @xmath82 increase are observed as well .
note that the temperature independence of @xmath51 and @xmath82 at @xmath83 k and similar steps in the @xmath51(t ) and @xmath82(t ) dependences near 125 k correlate with the temperature dependences of @xmath67 in bimn@xmath0o@xmath1 along different crystal axes ( see fig .
[ fig4 ] ) .
indeed , near @xmath79 k @xmath67 increases in a step - like fashion along all axes of the crystal . the temperatures at which the second high - temperature growth in @xmath51 and @xmath82 start are close to @xmath78 for different crystal axes .
the temperatures of @xmath51 jumps at different frequencies at @xmath84 k are well described by the arrhenius law , which allows one to estimate characteristic activation barriers @xmath85 at the phase separation domain boundaries in bimn@xmath0o@xmath1 .
these values are close to @xmath86 ev along different crystal axes ( see fig . [ fig6 ] ) .
let us turn to a more detailed examination of the conductivity of the crystals .
we deal with the real conductivity @xmath87 @xcite which is calculated from dielectric losses @xmath88 ( @xmath89 is the angular frequency , @xmath90 is the permittivity @xmath51 at @xmath91 ) .
this conductivity depends on both the frequency and temperature .
the low - frequency conductivities are dispersion - free ( percolation conductivity @xmath92 ) .
the conductivity @xmath93 has a frequency dispersion : the higher the frequency , the higher the conductivity .
the frequency dispersion of this type is typical of local conductivity ( i.e. , dielectric losses ) in restricted domains @xcite . in our case
, we attributed this local conductivity to phase separation domains .
the percolation conductivity ( leakage ) is attributable to the initial crystal matrix .
the relative local conductivity @xmath94 characterizes the ratio between the phase separation domain local conductivity and matrix leakage .
[ fig7 ] shows @xmath95 for bimn@xmath0o@xmath1 along the @xmath48 , @xmath10 , and @xmath11 axes ( figs .
[ fig7]a , [ fig7]b , and [ fig7]c , respectively ) . along the @xmath10 axis only one , most intense , maximum in @xmath95 in the temperature interval 125 220 k is observed .
below 100 k @xmath95 also exists , but at @xmath96 k it is transformed into leakage , and near 125 k it again becomes a local conductivity ( see the inset in fig .
[ fig7]b ) . the activation barrier calculated from the shifts of the local conductivity maxima at @xmath97 k ( fig .
[ fig7]b ) exceeds @xmath98 derived from the @xmath51 jumps only slightly ( figs . [ fig6]b and [ fig7]b ) . at @xmath99
k the conductivities along the @xmath48 and @xmath11 axes were low and could not be measured correctly by our device .
there are two @xmath95 maxima along these axes at @xmath97 k , the first is in the same temperature interval 125 220 k as along the @xmath10 axis ; the second is in the temperature region 225 350 k. the presence of two maxima split in temperature along the @xmath48 and @xmath11 axes in bimn@xmath0o@xmath1 means that the leakage sharply increases in the temperature interval between the maxima .
such behaviors of @xmath95 along the @xmath48 and @xmath11 axes correlate with the temperature behaviors of @xmath51 along these axes ( fig .
[ fig4 ] ) . however , the activation barriers calculated from the shifts of the first maxima of @xmath95 are almost twice as high as the barriers derived from the @xmath51 jumps ( compare fig .
[ fig6 ] and fig .
[ fig7 ] ) . at different frequencies for bimn@xmath0o@xmath1 along the @xmath48 axis ( a )
, @xmath10 axis ( b ) , and @xmath11 axis ( c ) .
inset shows the same dependences along the @xmath10 axis at @xmath100 k at a larger scale.,title="fig:",scaledwidth=45.0% ] at different frequencies for bimn@xmath0o@xmath1 along the @xmath48 axis ( a ) , @xmath10 axis ( b ) , and @xmath11 axis ( c ) .
inset shows the same dependences along the @xmath10 axis at @xmath100 k at a larger scale.,title="fig:",scaledwidth=45.0% ] at different frequencies for bimn@xmath0o@xmath1 along the @xmath48 axis ( a ) , @xmath10 axis ( b ) , and @xmath11 axis ( c ) .
inset shows the same dependences along the @xmath10 axis at @xmath100 k at a larger scale.,title="fig:",scaledwidth=45.0% ] tte inset in fig .
[ fig8 ] shows temperature dependences of conductivities of gdmn@xmath0o@xmath1 along the @xmath10 and @xmath11 axes at different frequencies .
the conductivity along the @xmath48 axis is close to that along the @xmath11 axis .
of interest is the fact that the frequency - dependent conductivity jumps along the @xmath48 and @xmath11 axes in gdmn@xmath0o@xmath1 also emerge near 125 k , like in bimn@xmath0o@xmath1 .
the activation barrier @xmath85 in gdmn@xmath0o@xmath1 calculated from the conductivity jumps is @xmath101 ev .
[ fig8 ] shows the temperature intervals in gdmn@xmath0o@xmath1 in which @xmath95 of the phase separation domains exceeds considerably the matrix leakage .
conductivity @xmath95 along the @xmath10 axis manifests itself up to @xmath102 k. at @xmath97 k there is only the leakage which grows with temperature . at @xmath79
k @xmath95 along the @xmath48 and @xmath11 axes rises abruptly and exists up to room temperature without a noticeable change .
the hysteresis loops and @xmath67 along all axes in both crystals are screened by leakage rather sharply at the temperatures @xmath78 at which the leakage and local conductivities become comparable .
this means that the @xmath78 values correspond to the temperatures at which the potential barriers of the restricted polar domain reorientations become equal to the kinetic energy of the itinerant electrons ( leakage ) .
differences between polar properties of phase separation domains in gdmn@xmath0o@xmath1 and bimn@xmath0o@xmath1 are observed along all axes of the crystals at @xmath103 k. at @xmath100 k the ground states of phase separation domains are similar for both crystals , which means that these states are formed inside the mn subsystems under the barriers of any origin . near 125 k the leakage and local conductivities along
the @xmath10 axes become comparable in both crystals . the itinerant electrons which appear along the @xmath10 axes at @xmath97
k are localized anew in deeper potential wells thus passing into @xmath95 of the phase separation domains . in gdmn@xmath0o@xmath1
these @xmath95 exceed leakage up to @xmath78 equal to 295 k and 325 k along the @xmath48 and @xmath11 axes , respectively ( see fig . [ fig5 ] ) . in bimn@xmath0o@xmath1 @xmath95 of phase separation domains
emerges at @xmath97 k along all crystal axes , including the @xmath10 axis , due to the effect of bi ions .
the localization of itinerant electrons in potential wells leads to deepening of these wells and increases the activation barriers of phase separation domains .
these @xmath85 barriers can be calculated from the shifts of the temperature maxima of @xmath95 at different frequencies . indeed , these barriers in bimn@xmath0o@xmath1 turned out to be twice as high as @xmath85 calculated from the @xmath51 jumps .
the latter barriers correspond to the condition when the kinetic energy of the itinerant electrons emerging along the @xmath10 axis becomes comparable to @xmath85 calculated from the @xmath51 jumps .
the values of @xmath78 in bimn@xmath0o@xmath1 along different axes correlate with the temperatures at which the conductivity at the lowest frequency ( leakage ) begins to increase sharply in the first @xmath95 maxima ( compare fig .
[ fig4 ] and fig [ fig7 ] ) . for gdmn@xmath0o@xmath1 along all axes ( the @xmath48 axis unfilled symbols , the @xmath10 axis
filled symbols , the @xmath11 axis lines ) for different frequencies .
the inset shows temperature dependences of conductivity along the @xmath10 and @xmath11 axes for different frequencies.,scaledwidth=45.0% ] note that near @xmath104 k dispersion - free anomalies typical of the ferroelectric phase transition manifest themselves in gdmn@xmath0o@xmath1 on the background of @xmath95 and @xmath93 along the @xmath10 axis ( see fig .
[ fig8 ] ) .
this is due to the fact that a maximum in @xmath88 should be observed near @xmath77 , while in bimn@xmath0o@xmath1 a low polarization @xmath74 does not manifest itself on the background of @xmath95 ( fig . [
fig7]b ) .
both the intrinsic polarization and @xmath95 of restricted polar phase separation domains give contributions into the p1 and n1 curves in measurements of hysteresis loops for gdmn@xmath0o@xmath1 and bimn@xmath0o@xmath1 at @xmath105 by the pund method , while @xmath95 contributes significantly to the p2 and n2 curves in the same temperature interval . at these temperatures @xmath67
emerges in these crystals . near @xmath78
at which @xmath67 tends to zero , the polarization and local conductivity contributions to the p1-p2 and n1-n2 curves start to decrease simultaneously . at the temperatures at which @xmath106 ,
these contributions are transformed into linear dependences on e. at @xmath107 the loop restores as an inverted loop , and the leakage contribution begins to dominate in it .
the experimental evidence of the existence of restricted polar phase separation domains of gdmn@xmath0o@xmath1 and bimn@xmath0o@xmath1 was obtained in the x - ray high - sensitivity diffraction study carried out at room temperature ( see figs .
[ fig9]a and [ fig9]b , respectively ) .
the angular intensity distributions of * ( 004 ) * , * ( 060 ) * and * ( 600)*@xmath108 bragg reflections were detected in the 3-crystal regime with the @xmath109 scan . as a monochromator and an analyzer , germanium crystals in the * ( 004 ) * reflection
were used , which allowed conditions of nearly dispersion - free high - resolution ( @xmath110 ) survey geometry to be realized .
[ fig9]a shows a single diffraction maximum of * ( 060 ) * and two diffraction maxima of the * ( 004 ) * bragg reflections recorded from different planes of a single crystal of gdmn@xmath0o@xmath1 .
these planes are perpendicular to the @xmath10 and @xmath11 axes , respectively .
the * ( 004 ) * bragg reflection positions are characterized by slightly different interplanar spacings @xmath111 ( @xmath112 ) .
these * ( 004 ) * reflections , which have comparable intensities and half - widths ( @xmath113 ) clearly point to the coexistence of two phases with slightly differing @xmath11 lattice parameters which differ in the third decimal place .
the bragg peak along the @xmath10 axis with a similar half - width is not split , i.e. , it is identical to these two phases .
the positions of all bragg peaks nearly coincide with those for gdmn@xmath0o@xmath1 with the generally accepted _ pbam _ symmetry .
this means that two phases we detected are high - quality gdmn@xmath0o@xmath1 single crystal phases with similar large correlation lengths @xmath114 .
the phase with a higher intensity of the * ( 004 ) * bragg peak can be attributed to the original matrix .
the phase with a lower intensity of such a peak can be attributed to the phase separation domains . a similar picture of the angular intensity distributions of the * ( 004 )
* , * ( 060 ) * and * ( 600)*@xmath108 bragg reflections is also observed for bimn@xmath0o@xmath1 , but splitting into two peaks for the reflexes along all axes of the crystal is detected ( fig .
[ fig9]b ) .
it is natural to attribute this to lattice distortions caused by bi ions along all crystal axes .
note that the structural quality of both crystals is similar .
bragg reflections as functions of interplanar distances ( d @xmath115 ) in gdmn@xmath0o@xmath1 ( a ) and bimn@xmath0o@xmath1 ( b ) .
parameters of gdmn@xmath0o@xmath1 lattice are @xmath116 ; @xmath117 ; @xmath118 .,title="fig:",scaledwidth=45.0% ] bragg reflections as functions of interplanar distances ( d @xmath115 ) in gdmn@xmath0o@xmath1 ( a ) and bimn@xmath0o@xmath1 ( b ) .
parameters of gdmn@xmath0o@xmath1 lattice are @xmath116 ; @xmath117 ; @xmath118 .,title="fig:",scaledwidth=45.0% ] let us discuss the shape and structure of phase separation domains in gmn@xmath0o@xmath1 and bimn@xmath0o@xmath1 . as noted above , the phase separation domains in the form of @xmath26 superlattices
were observed below the temperatures of magnetic and ferroelectric orderings in a number of rmn@xmath0o@xmath1 with various @xmath119 ions , including @xmath2 and @xmath3 @xcite .
a set of ferromagnetic resonances was detected from individual layers of superlattices .
the intensity of these resonances sharply decreased when @xmath120 k was approached .
since the phase separation domains should exist at all temperatures , this fact means that the structure of the phase separation domains must change at @xmath121 k. we manage to study the temperature evolution of feromagnetic resonances of layered superlattices at @xmath121 k only in eu@xmath31ce@xmath32mn@xmath0o@xmath1 , in which the phase separation domains were similar to those in eumn@xmath0o@xmath1 , but their concentration was much high @xcite .
in temperature interval @xmath122 k only one ferromagnetic resonance signal was observed in eu@xmath31ce@xmath32mn@xmath0o@xmath1 .
this resonance corresponded to the low - temperature superlattice layer having the maximal barrier and conductivity @xcite .
near @xmath123 k an intensity of this resonance became indistinguishable due to growth of conductivity .
at room temperature the structure of phase separation domains was installed in eu@xmath31ce@xmath32mn@xmath0o@xmath1 in x - ray diffraction study @xcite .
the two @xmath124 well defined bragg peaks were accompanied by the characteristic oscillation peaks independent on the interplanar spacings @xmath111 , which pointed to existence of the layered superstructure in this crystal . as shown in @xcite ,
the layered superstructure consisting of the alternating original crystal eumn@xmath0o@xmath1 layers and conductive layers of phase separation emerged in eu@xmath31ce@xmath32mn@xmath0o@xmath1 at all temperatures above 40 k. no similar oscillation peaks on the background of the bragg reflexes were observed in x - ray studies of gdmn@xmath0o@xmath1 and bimn@xmath0o@xmath1 at room temperature ( fig .
[ fig9 ] ) , i.e. , there were no layered superstructures in these crystals .
most likely the phase separation domains in gdmn@xmath0o@xmath1 and bimn@xmath0o@xmath1 had the shape of nanoscale droplets at @xmath121 k. the possibility of different geometries of the phase separation domains in coulomb - frustrated systems , including manganites and high - temperature superconductors , was shown theoretically in @xcite .
the properties of the phase separation domains in the form of nanoscale droplets for manganites were discussed in @xcite . now consider in more detail what are polar phase separation domains .
there is a probability that @xmath125 electrons of some mn@xmath6 ions tunnel to mn@xmath7 ions in the original rmn@xmath0o@xmath1 matrix .
these valence electrons and recharged mn@xmath6-mn@xmath7 ion pairs are accumulated in the restricted phase separation domains inside the original matrix because phase separation is energetically favorable @xcite .
since mn@xmath6 ions appear in the mn@xmath7 ion positions ( i.e. , in the octahedral oxygen surrounding ) , they become jahn - teller ions and give rise to local deformations of these octahedra . in turn
, mn@xmath7 ions appear in the mn@xmath6 ion positions ( in noncentrosymmetrical pentagonal pyramids ) and local distortions arise near these ions as well . as a result
, structural distortions caused by both these factors occur inside restricted phase separation domains .
we believe that phase separation domains are noncentrosymmetric , and their sizes are large enough to give rise to ferroelectric single - domain states inside them . the restricted polar phase separation domain states existing in gdmn@xmath0o@xmath1 and bimn@xmath0o@xmath1 original matrices are analogs of the superparaelectric state formed by an ensemble of spherical ferroelectric nanoparticles in a dielectric matrix theoretically studied in @xcite .
it was shown in @xcite that at low temperatures a homogeneous polarization could exist in these particles ( in our case restricted polar phase separation domains ) if their sizes @xmath119 were less than the correlation radius @xmath114 but larger than the critical radius @xmath126 of the size - driven ferroelectric - paraelectric phase transition . under these conditions
all dipole moments inside particles are aligned due to correlation effects .
surface screening of depolarization fields makes single - domain states energetically favorable .
if @xmath127 , separate paraelectric dipoles are uncorrelated and represent local polar defects which can only increase the width of the bragg peak original matrix .
the fact that the well - defined bragg reflections related to the restricted polar phase separation domains are observed in our study together with similar bragg peaks of the original matrix indicates that the conditions for the ferroelectric single - domain states of phase separation regions given in @xcite are fulfilled .
it was also found in @xcite that a frozen superparaelectric phase could emerge for an ensemble of ferroelectric nanoparticles in a dielectric matrix . in this phase remanent polarization and
hysteresis loop arose at the temperatures lower than the freezing temperature @xmath128 .
@xmath128 was defined from the condition that the potential barriers of nanoparticle polarization reorientation became equal to the thermal activation energy @xmath129 . at @xmath130
the conventional superparaelectric state occurred for which there are no hysteresis loops and remanent polarization .
it was also accepted in @xcite that the temperature of thermal destruction of intrinsic nanoparticle ferroelectricity was @xmath131 .
so , the temperature behavior of the gdmn@xmath0o@xmath1 and bimn@xmath0o@xmath1 polarization is also analogous to that of the frozen superparaelectric state discussed in @xcite . @xmath78
considered above can be regarded as @xmath128 . as noted above
, the @xmath78 values correspond to the temperatures at which the potential barriers of the restricted polar domain reorientations become equal to the kinetic energy of the itinerant electrons ( leakage ) .
the restricted phase separation domains still exist at @xmath107 due to energetic favorability of the phase separation caused by strong interactions .
but the frozen superparaelectric state turns into the conventional superparaelectric one near @xmath78 .
since @xmath132 hysteresis loops are measured under the field e applied along different axes , electric polarizations inside phase separation domains are induced along these axes .
indeed , the application of field e along any crystal axis initiates a drift of valence @xmath125 electrons localized inside phase separation domains in this direction .
these electrons recharge @xmath133 and @xmath134 ions inside the phase separation domains . as a result , the spatial distribution of the @xmath133 and @xmath134 ions and structural distortions inside phase separation domains are bound to change , thereby giving rise to the polarization along the e direction .
thus , the actual symmetry of rmn@xmath0o@xmath1 at room temperature can be established only in polarization measurements in @xmath135 .
we believe that the additional structural reflexes typical of the noncentral monoclinic structure reported in @xcite can be related to the frozen superparaelectric state arising in rmn@xmath0o@xmath1 at @xmath136 .
note that in the cases when temperatures @xmath78 are sufficiently high , the correlations between the phase separation domains can occur in the entire crystal volume .
this situation was observed earlier in eu@xmath31ce@xmath32mn@xmath0o@xmath1 at room temperature @xcite .
as noted above , phase separation and charge carrier self - organization give rise to a dynamic equilibrium of the phase separation domain states with a balance between attraction ( double exchange , jahn - teller effect ) and coulomb repulsion of charge carriers @xcite .
the formation of the phase separation domains due to a balance between strong interactions leads to specific features in rmn@xmath0o@xmath1 properties .
first , polar phase separation domains are bound to emerge up to high temperatures , thus giving rise to high - temperature polarization .
second , changes in polar phase separation domains under varying e rapidly relax to the dynamic equilibrium states after e is switched off .
these features manifested themselves in our experiments and necessitated the use of the pund method for measuring the polarization hysteresis loops which were discussed above .
the application of magnetic field h increases the barriers at the phase separation domain boundaries due to the double exchange growth , thus increasing the @xmath78 temperatures in gdmn@xmath0o@xmath1 ( see the right panels in fig .
[ fig5 ] ) .
the field h oriented along the @xmath10 axis in gdmn@xmath0o@xmath1 also enhances the polarization induced by the restricted polar domains due to the increasing probability of charge transfer between @xmath137 ion pairs with the greatest distance between them ( see the caption to fig .
[ fig9]a ) .
it was found that application of magnetic field @xmath80 t to bimn@xmath0o@xmath1 had a much weaker effect on phase separation domain properties and polarization than in gdmn@xmath0o@xmath1 because of strong structural distortions of the crystal field by bi ions .
thus , remanent polarizations and hysteresis loops originating from the frozen superparaelectric state of similar ferroelectric restricted polar phase separation domains arranged inside the initial matrix of rmn@xmath0o@xmath1 ( @xmath2 and @xmath3 ) have been revealed in wide temperature intervals from 5 k up to @xmath78 depending on the crystal axis .
the restricted polar phase separation domains emerge due to phase separation and charge carrier self - organization . at @xmath138
the potential barriers of the restricted polar domain reorientations become equal to the kinetic energy of the itinerant electrons ( leakage ) . in gdmn@xmath0o@xmath1 ,
the polarization of this type emerges along the @xmath11 axis up to @xmath70 k. in bimn@xmath0o@xmath1 , the maximum temperature @xmath72 k is observed along the @xmath10 axis . at @xmath139 the frozen superparaelectric state turns into the conventional superparaelectric one .
the polarizations observed can be attributed to the magnetically induced polarizations since the magnetic double exchange between mn@xmath6-mn@xmath7 ion pairs is the key interaction giving rise to the phase separation domain formation .
the effect of magnetic field on these polarizations in gdmn@xmath0o@xmath1 demonstrates that the magnetoelectric coupling exists in the paramagnetic phase as well .
application of electric field e modifies the structure and shape of the phase separation domains , giving rise to the polarization along the e direction .
additional structural reflexes typical of the noncentral monoclinic structure reported in @xcite can be related to the frozen superparaelectric state arising in rmn@xmath0o@xmath1 at @xmath136 . in the cases
when temperatures @xmath78 are sufficiently high , the correlations between the phase separation domains can emerge in the entire crystal volume thus giving rise the noncentosymmetricity of the whole crystals .
the work was supported by the government of russian federation ( project n0.14.b25.31.0025 ) .
n. hur , s. park , p.a .
sharma , j.s .
ahn , s. guba , and s - w .
cheong , electric polarization reversal and memory in a multiferroic material induced by magnetic fields , nature ( london ) 429 ( 2004 ) 392 - 395 .
doi : 10.1038/nature02572 .
y. noda , h. kimura , m. fukunaga , s. kobayashi , i. kagomiya , and k. kohn , magnetic and ferroelectric properties of rmn@xmath0o@xmath1 , j. phys . : condens .
matter 20 ( 2008 ) 434206 .
doi : http://dx.doi.org/10.1088/0953-8984/20/43/434206 .
v. baledent , s. chattopadhyay , p. fertey , m.b .
lepetit , m. greenblatt , b. wanklyn , f.o .
saouma , j.i .
jang , and p. foury - leylekian , evidence for room temperature electric polarization in rmn@xmath0o@xmath1 , phys .
( 2015 ) 117601 .
doi : http://dx.doi.org/10.1103/physrevlett.114.117601 .
khannanov , v.a .
sanina , e.i .
golovenchits , m.p .
scheglov , room - temperature electric polarization induced by phase separation in multiferroic gdmn@xmath0o@xmath1 , jetpletters 103 ( 2016 ) 248 - 253 .
doi : 10.1134/s002136401604007x .
glinchuk , e.a .
eliseev , and a.n .
morozovska , superparaelectric phase in the ensemle of non - interacting ferroelectric nanoparticles , phys .
b. 78 ( 2008 ) 134107 .
doi : http://dx.doi.org/10.1103/physrevb.78.134107 .
sanina , e.i .
golovenchits , v.g .
zalesskii , spin - wave excitations in superlattices self - assembled in multiferroic single crystals , j. phys .
24 ( 2012 ) 346002 .
doi : doi : http://dx.doi.org/10.1088/0953-8984/24/34/346002 .
khannanov , v.a .
sanina , e.i .
golovenchits , giant polarization and magnetic dynamics in gdmn@xmath0o@xmath1 , j. phys .
572 ( 2014 ) 012046 .
doi : http://dx.doi.org/10.1088/1742-6596/572/1/012046 .
sanina , e.i .
golovenchits , v.g .
zalesskii , s.g .
lushnikov , m.p .
scheglov , s.n .
gvasaliya , a. savvinov , r.s .
katiyar , h. kawaji , and t. atake , phase separation and charge carrier self - organization in semiconductor- multiferroic eu@xmath31ce@xmath32mn@xmath0o@xmath1 , phys .
b 80 ( 2009 ) 224401 .
doi : 10.1103/physrevb.80.224401 .
sanina , e.i .
golovenchits , v.g .
zalesskii , and m.p .
scheglov , magnetic properties of multiferroics - semiconductors eu@xmath31ce@xmath32mn@xmath0o@xmath1 , j. phys . : cond .
23 ( 2011 ) 456003 .
doi : http://dx.doi.org/10.1088/0953-8984/23/45/456003 .
sanina , e.i .
golovenchits , b.kh.khannanov , m.p .
scheglov , v.g .
zalesskii , temperature evolution of the polar states in gdmn@xmath0o@xmath1 and gd@xmath31ce@xmath32mn@xmath0o@xmath1 , jetplett . ,
100 ( 2014 ) 407 - 412 .
doi : 10.1134/s002136401418009x .
scott , l. kammerdiner , l.m .
parris , s. traynor , v. ottenbacher , a. shawabkeh , and w.f .
oliver , switching kinetics of lead zirconate titanate submicron thin - film memories , j.appl.phys 64 ( 1988 ) 787 - 792 .
doi : http://doi.org/10.1063/1.341925 .
feng , y.s .
chai , j.l .
zhu , n. manivannan , y.s .
wang , y.s .
yang , c.q .
jin , and kee hoon kim , determination of the intrinsic ferroelectric polarization in orthorombic homno@xmath24 , new j. phys .
12 ( 2010 ) 073006 .
doi : http://iopscience.iop.org./1367-2630/12/7/073006 .
n. lee , c. vecchini , y.j .
choi , l.c .
chapon , a. bombardi , p.g .
radaelli , and s.w .
cheong , giant tunability of ferroelectric polarization in gdmn@xmath0o@xmath1 , phys .
( 2013 ) 137203 .
doi : 10.1103/physrevlett.110.137203 .
a. munoz , j.a .
alonso , m.t .
cosais , m.j .
martinez - lope , j.l .
martinez , m.t .
fernandez - diaz , magnetic structure and properties of bimn@xmath0o@xmath1 oxide : a neutron diffraction study , phys .
b 65 ( 2002 ) 144423 .
doi : http://dx.doi.org/10.1103/physrevb.65.144423 .
m. fukunago and y. noda , measurement of complicated temperature - dependent polarization of multiferroics rmn@xmath0o@xmath1 , j. phys .
( 2011 ) 012090 .
doi : http://iopscience.iop.org/1742-6596/320/1/012090 .
c. ortix , j. lorenzana , and c. di castro , coulomb - frustrated phase separation in systems with short - range negative compressibility , phys .
( 2008 ) 246402 .
doi : http://dx.doi.org/10.1103/physrevlett.100.246402 .
kugel , a.l .
rakhmanov , a.o .
sboychakov , f.v .
kusmartsev , nicola poccia , and antonio bianconi , a two - band model for the phase separation induced by the chemical mismatch pressure in different cuprate superconductors , semicond .
technol . 22
( 2009 ) 014007 .
doi : http://dx.doi.org/10.1088/0953-2048/22/1/014007 . | the electric polarization hysteresis loops and remanent polarization were revealed in multiferroics rmn@xmath0o@xmath1 with @xmath2 and @xmath3 at wide temperature interval from 5 k up to 330 k. until recently , the long - range ferroelectric order having an exchange - striction magnetic nature had been observed in rmn@xmath0o@xmath1 only at low temperatures ( @xmath4 35 k ) .
we believe that the polarization we observed was caused by the frozen superparaelectric state which was formed by the restricted polar domains resulting from phase separation and charge carriers self - organization . at some sufficiently high temperatures @xmath5 the frozen superparaelectric state
was destroyed , and the conventional superparaelectric state occurred .
this happened when the potential barriers of the restricted polar domain reorientations become equal to the kinetic energy of the itinerant electrons ( leakage ) .
the hysteresis loops were measured by the so - called pund method which allowed us to correctly subtract the contribution of conductivity from the measured polarization .
the correlations between properties of the phase separation domains and polarization were revealed and studied .
the high - temperature polarization also had a magnetic nature and was controlled by the magnetic field because the double exchange between pairs of mn ions with different valences ( mn@xmath6 and mn@xmath7 ) in rmn@xmath0o@xmath1 was the basic interaction resulting in phase separation .
multiferroic , polarization hysteresis loop , phase separation , charge carrier self - organization 75.85.+t , 77.22.ej , 77.80.dj |
massive stars ( @xmath4 ) play an important role in astrophysics .
they are the progenitors of blue supergiants ( bsg ) , red supergiants ( rsg ) , wolf - rayet ( wr ) and luminous blue variable ( lbv ) stars . at the end of their life
, they explode as type ii or ib , c supernovae ( sn ) and maybe also as long soft gamma - ray bursts ( grb ) .
their cores after collapse become neutron stars ( ns ) or black holes ( bh ) .
they are one of the main sites for nucleosynthesis , which takes place during both pre - sn ( hydrostatic ) burnings as well as during explosive burnings .
a weak s - process occurs during h - burning and r - process probably occurs during the explosion .
radioactive isotopes like @xmath5al and @xmath6fe detected by the integral satellite are produced by massive stars .
massive stars even though they are much less numerous than low mass stars contribute significantly to the integrated luminosity of galaxies . at high redshifts ( @xmath7 , or low metallicities @xmath8 ) , their importance grows .
the first stars formed are thought to be all massive or even very massive ( bromm 2005 ) and to be the cause of the re - ionisation of the universe .
finally , recent surveys of metal poor halo stars provide many constraints for the early chemical evolution of our galaxy ( beers et al .
1992 , beers 1999 , israelian et al .
2004 , christlieb et al .
2004 , cayrel et al .
2004 , spite et al .
2005 ) .
the measured abundances of many elements present a very small scatter down to very low metallicities ( [ fe / h]@xmath9 , see cayrel et al . 2004 ) .
this tends to prove that the interstellar medium ( ism ) was already well mixed at an early stage .
however , r - process elements show a scatter in their abundances contradicting this trend ( ryan ... ) . there is no significant enrichment by pair - instability supernovae ( pisn , see heger and woosley 2002 , umeda and nomoto 2002 , 2003 for more details ) .
production of primary nitrogen in massive stars ( that died before the formation of the halo metal poor star ) is necessary in order to explain the high abundance of nitrogen observed . about a quarter of the extremely metal poor ( emp )
stars show a strong enrichment of carbon ( carbon - rich , cemp ) , with [ c / fe]@xmath1024 , where @xmath11= { \rm log_{10}}(x_{\rm a}/x_{\rm b})- { \rm log_{10}}(x_{\rm a}/x_{\rm b})_\odot$ ] .
the cemp stars can probably be divided into two subclasses , the ba - rich and the ba - normal stars ( see ryan et al . 2005 and references therein ) . according to ryan et al .
( 2005 ) , the ba - rich cemp stars have accreted enriched matter from an agb stars in a binary system in the course of their life . on the other hand
the ba - normal cemp stars would have formed from a cloud already enriched in carbon and neutron capture elements .
the cloud itself would have been enriched by wind or supernova ejecta from massive stars .
the two most metal poor halo stars found up to now , he0107 - 5240 ( [ fe / h]=-5.3 , see christlieb et al . 2004 ) and he1327 - 2326 ( [ fe / h]=-5.5 , see frebel et al .
2005 and aoki et al 2005 ) belong to this last group , the ba - normal cemp stars . these two stars both present strong but different enrichment in nitrogen and oxygen . in the light of these recent observational results and following the exploratory work of meynet et al ( 2005 )
, we explore the impact that rotation can have in massive very low metallicity stars . in sect .
2 , we present the pre - sn models , their evolution and stellar yields for different initial masses , rotation velocities and metallicities and compare them with emp stars . in sect .
3 , we describe the recent update in the explosion models and the corresponding nucleosynthesis in 1d and the latest development of a new 3d mhd model . in sect .
4 we give our conclusions .
the computer model used to calculate the stellar models is described in detail in hirschi et al ( 2004 ) .
convective stability is determined by the schwarzschild criterion .
convection is treated as a diffusive process from oxygen burning onwards .
the overshooting parameter is 0.1 h@xmath12 for h and he burning cores and 0 otherwise .
instabilities induced by rotation taken into account are meridional circulation and secular and dynamical shears .
the reaction rates are taken from the nacre ( angulo et al 1999 ) compilation for the experimental rates and from the nacre website ( http://pntpm.ulb.ac.be/nacre.htm ) for the theoretical ones . at low metallicities
the initial chemical composition is calculated in the following way . for a given metallicity @xmath8 ( in mass fraction )
, the initial helium mass fraction @xmath13 is given by the relation @xmath14 , where @xmath15 is the primordial helium abundance and @xmath16 the slope of the helium to metal enrichment law .
@xmath15 = 0.24 and @xmath16 = 2.5 were used according to recent determinations ( see izotov & thuan 2004 for example ) . for the mixture of the heavy elements
, we adopted the same mixture as the one used to compute the opacity tables for weiss 95 s alpha enriched composition ( iglesias & rogers 1996 ) .
the mass loss rates are described and discussed in meynet , ekstrm & maeder ( 2005 ) . the mass loss rates ( and opacities )
are rather well determined for chemical compositions which are similar to solar ( or alpha enriched mixing ) composition or similar to a fraction of the solar composition . however
, very little was known about the mass loss of very low metallicity stars with a strong enrichment in cno elements until recently .
vink & de koter ( 2005 ) study the case of wr stars but a crucial case , which has not been studied in detail yet , is the case of red supergiant stars ( rsg ) .
as we shall see later , due to rotational and convective mixing , the surface of the star is strongly enriched in cno elements during the rsg stage .
awaiting for future studies , it is implicitly assumed in this work ( as in meynet , ekstrm & maeder 2005 ) that cno elements have a significant contribution to opacities and mass loss rates .
therefore the mass loss rates depend on metallicity as @xmath17 , where @xmath8 is the mass fraction of heavy elements at the surface of the star , also when the iron group elements content is much lower the cno elements content .
a specific treatment for mass loss was applied at break - up ( see meynet , ekstrm & maeder 2005 ) . at break - up , the mass loss rate adjusts itself in such a way that an equilibrium is reached between the envelope extension and the removal of gravitationally unbound mass . in practice , however , since the critical limit contains mathematical singularities , we considered that during the break - up phase , the mass loss rates should be such that the model stays near a constant fraction ( around 0.95 ) of the limit .
the value of 300 kms@xmath18 used for the initial rotation velocity at solar metallicity corresponds to an average velocity of about 220kms@xmath18 on the main sequence ( ms ) which is very close to the average observed value ( see for instance fukuda 1982 ) .
it is unfortunately not possible to measure the rotational velocity of very low metallicity stars since they all died a long time ago .
nevertheless , there is indirect evidence that stars with a lower metallicity have a higher rotation velocity .
observational evidence is the higher ratio of be to b stars in the magellanic clouds compared to our galaxy ( maeder et al 1999 ) .
this can be due to the difficulty of evacuating angular momentum during the star formation , which is even more important at lower metallicities ( see abel et al 2002 for simulations of star formation in the early universe ) . finally , a low metallicity star containing the same angular momentum as a solar metallicity star has a higher surface rotation velocity due to its smaller radius ( one quarter of @xmath19 radius for 20 @xmath1 stars ) .
it is therefore worth studying models with an initial rotational velocities faster than 300 kms@xmath18 . in order to compare the models at different metallicities and with different initial masses with another quantity than the surface velocity ,
the ratio @xmath20 is used ( see table [ table1 ] ) .
@xmath21 , where @xmath22 is the equatorial radius at break up and @xmath23 is the ratio of the luminosity to the eddington luminosity .
@xmath20 increases only as @xmath24 for models with the same angular momentum ( @xmath25 ) but lower metallicity , whereas the surface rotational velocity increases as @xmath26 ( @xmath27 ) .
the angular momentum could be used as well but it varies significantly for models of different initial masses . finally , @xmath20 is a good indicator for the impact of rotation on mass loss . in the first series of models , the aim is to scan the parameter space of rotation and metallicity with 20 @xmath1 models since a 20 @xmath1 star is not far from the average massive star concerning stellar yields .
for this series , two initial rotational velocities were used at very low metallicities .
the first one is the same as at solar metallicity , 300kms@xmath18 .
the ratio @xmath20 decreases with metallicity ( see table [ table1 ] for the 20 @xmath1 models ) for the initial velocity of 300kms@xmath18 . the second @xmath28 is 500kms@xmath18 at z=10@xmath29 ( [ fe / h]@xmath30 - 3.6 ) and 600kms@xmath18 at z=10@xmath31 ( [ fe / h]@xmath10 - 6.6 ) .
these values have ratios of the initial velocity to the break up velocity , @xmath32 around 0.55 , which is only slightly larger than the solar metallicity value ( 0.44 ) .
the 20 @xmath1 model at z=10@xmath31 and with 600kms@xmath18 has a total initial angular momentum @xmath33ergs which is the same as for of the solar metallicity 20 @xmath1 model with 300kms@xmath18 ( @xmath34ergs ) .
so a velocity of 600kms@xmath18 , which at first sight seems extremely fast , is probably the average velocity at z=10@xmath31 . in the second series of models ,
we follow the exploratory work of meynet , ekstrm & maeder ( 2005 ) and compute models at z=10@xmath31 with initial masses of 40 , 60 and 85 @xmath1 and initial rotational velocities of 700 , 800 and 800kms@xmath18 respectively .
note that , for these models as well , the initial total angular momentum is similar to the one contained in solar metallicity with rotational velocities of 300kms@xmath18 .
since this is the case , velocities between 600 and 800kms@xmath18 are considered in this work as the average rotational velocities at these very low metallicities .
the evolution of the models was followed until core si burning .
note that the non rotating and fast rotating 20 @xmath1 models at z=10@xmath31 have a strongly degenerate core such that si
burning does not occur in the central percent of the star .
it is planned to improve the model to better follow strongly degenerate core .
this does nevertheless not affect the results for presented in this work .
the 60 @xmath1 model was evolved until ne burning .
this means that , for this model only , the yields of the heavy elements can still vary .
the stellar yields are calculated as in hirschi et al ( 2005 ) .
therefore the contribution from explosive nucleosynthesis is not included and the remnant mass is determined from the co core mass ( see maeder 1992 ) .
the impact on the results are discussed in hirschi ( 2004 ) and we only present yields for elements which are not significantly affected by the evolution beyond our calculations .
the main characteristics of the models are presented in table [ table1 ] .
.initial parameters of the models ( columns 15 ) : mass , metallicity , rotation velocity [ kms@xmath18 ] , total angular momentum [ 10@xmath36ergs ] and @xmath20 .
total lifetime [ myr ] and various masses [ @xmath1 ] ( 710 ) : final mass ( 3 ) , masses of the helium and carbon oxygen cores and the remnant mass .
[ cols=">,>,>,>,>,>,>,>,>,>",options="header " , ] [ wind ] in fig .
[ m85 ] ( _ right _ ) , we compare the chemical composition of the wind material with abundances observed in non - evolved carbon rich extremely and ultra ( frebel et al 2005 ) metal poor stars .
the ejected masses of the wind material are also given in table [ wind ] .
it is very interesting to see that the wind material can reproduce the observed abundance in two ways . either , the wind material is richer than necessary and dilution ( by a factor 100 for example for the 40 @xmath1 models and he1327 - 2326 ) with the ism is needed or the wind has the right enrichment ( for example the 60 @xmath1 and he1327 - 2326 ) and the low mass star could form from pure wind material .
the advantage of the pure wind material is that it has a ratio @xmath38c/@xmath39c around 5 ( meynet , ekstrm & maeder 2005 ) and it can explain li depletion . with or without dilution , the wind material has the advantage that it brings the initial metallicity of the low mass star above the critical value for its formation ( bromm 2005 ) .
supernovae of type ii , ib / c are triggered by the gravitational collapse of the inner stellar core when it reaches the chandrasekhar mass , i.e. the maximum iron core mass supported by the dominant electron pressure .
early spherically symmetric simulations ( at that time adiabatic or based on local neutrino emission approximations ) ( van riper & arnett 1978 and references therein ) suggested the separation of the collapsing material into an inner and an outer core .
the inner core fails only marginally to be pressure supported and the sound speed stays faster than the fluid velocity at all times . in this regime ,
the infall velocity increases linearly with the radius . because the matter density and sound speed decrease with the distance from the center , there is a sonic point where the sound speed approaches the infall velocity .
outside of the sonic point , the infall velocity is larger than the sound speed and information about the collapsed core reach the outer layers only in form of a rarefaction wave .
the essentially polytropic behavior of the equation of state allows a non - relativistic analytic investigation of collapse ( goldreich & weber 1980,yahil 1983 ) ( note that general relativistic effects decrease the mass of the inner core by @xmath40 ( liebendrfer et al .
2001 , hix et al . 2003 ) ) .
the analytic investigation shows that the mass of the inner core is well approximated by@xmath41 where @xmath42 refers to the core mass and @xmath43 to the coefficient in the polytropic equation of state @xmath44 with @xmath45 .
values with index @xmath46 belong to the marginally stable stage immediately before collapse . with the coefficient for the degenerate ultra - relativistic electron gas ( shapiro & teukolsky 1983)@xmath47 one
readily reproduces the conclusion that the mass of the inner core , @xmath48 , evolves proportionally to the squared electron fraction , @xmath49 .
the evolution of @xmath50 and the entropy during collapse are determined by the interplay between three different physical processes : ( i ) the transition of protons to neutrons by electron capture and neutrino emission , @xmath51 , ( ii ) isoenergetic neutrino scattering off nucleons , @xmath52 , and ( iii ) neutrino thermalization by neutrino scattering off electrons , @xmath53 .
each of these processes is rather straightforward to describe in a gas of free nucleons ( bruenn 1985 ) . in reality , however , the by far most abundant nuclear species in the cold collapsing matter are neutron - rich heavy nuclei .
while the described weak interactions basically stay the same , they rather involve nucleons bound in nuclei than free ones .
this leads to the interesting expectation that the nuclear structure is probed in many nuclei that are difficult or impossible to explore under terrestrial conditions .
see ( martnez - pinedo et al .
2004 ) for a recent review .
\(i ) as the density , @xmath54 , increases , the electron chemical potential increases with @xmath55 ( bethe 1990 ) .
the electrons fill higher energy levels and electron captures on free or bound protons become more frequent . as long as the density is lower than @xmath56 g/@xmath57 ,
the neutrinos escape freely and the deleptonization rate is determined by the electron capture rate .
the electron capture rate on free protons is by two orders of magnitude larger than the one on nuclei .
but the nuclei are by orders of magnitude more abundant targets .
because of a highly floating magnitude of the free proton fraction , it is difficult to decide from first principles whether electron captures on free protons or nuclei would dominate in core collapse : due to the repulsive positive charges of the protons , nuclei are most stable in a neutron - rich configuration and the ratio of bound protons to neutrons is mainly determined by intrinsic properties of the nuclei and the entropy .
hence , small changes in the electron fraction or entropy of the fluid element cause significant changes in the small abundance of unbound free protons .
a rising abundance of free protons is immediately followed by increased electron captures , lowering the electron fraction , and thus driving back the free proton abundance .
the deleptonization caused by this strong negative feedback sets an upper limit to the sustainable electron fraction in the inner core , even in the absence of other open channels for electron capture ( messer 2000 , liebendrfer et al .
2002 ) .
however , improved electron capture rates on heavy nuclei ( langanke & martnez - pinedo 2002 ) that overcome the idealized blocking of gamov - teller transitions in the traditionally applied single - particle model allow the deleptonization to occur always faster than with captures on free protons alone .
in fact , the most recent simulations with improved electron capture rates demonstrate that electron capture on heavy nuclei is always dominating over electron capture on free protons and that the deleptonization proceeds to significantly lower electron fractions at the center of the core than with the previous `` standard '' nuclear physics input , resulting in a @xmath58 smaller inner core at bounce ( langanke et al . 2003 , hix et al . 2003 , marek et al . 2005 ) .
\(ii ) the basic neutrino opacity in core collapse is provided by neutrino scattering on nucleons .
depending on the distribution of the nucleons in space and the wavelength of the neutrinos , various important coherence effects can occur : most important during collapse is the binding of nucleons in nuclei with a density contrast of several orders of magnitude to the surrounding nucleon gas .
the cross section for coherent scattering of low energy neutrinos on nuclei scales with @xmath59 for nuclei with atomic number @xmath60 .
coherent scatterings off nuclei therefore easily dominate the scattering opacity of neutrinos on nucleons of the surrounding neutron gas ( freedman 1974 ) .
a useful comparison of inverse mean free paths at the important density @xmath61 g/@xmath57 is given in ( bruenn & mezzacappa 1997 ) .
they find @xmath62 , @xmath63 , @xmath64 , and @xmath65 .
further corrections are necessary : with an increasing ratio between the coulomb potential of the positively charged ions and their thermal energy , the average separation between nuclei will more strongly peak around the value of most efficient packing .
the neutrino opacities are then to be corrected by an ion - ion correlation function @xmath66 ( itoh 1975 , 1979 , horowitz 1997 ) .
its consideration in core collapse simulations based on a representative nucleus lowers the trapped lepton fraction at bounce by @xmath67 and increases the central entropy per baryon by @xmath68 ( bruenn & mezzacappa 1997 ) .
it is found that the sizable entropy increase is not only due to the increased deleptonization , but also to the fact that the correlation effect is most pronounced for the low energy neutrinos with long wavelengths . as current core collapse models proceed toward the inclusion of a full ensemble of nuclei , it becomes rather non - trivial how to adequately determine correlation effects in the ion mixture , see for example ( sawyer 2005 ) and references therein .
the situation is even more complicated in the phase transition from isolated nuclei to bulk nuclear matter where the nuclei or the holes in between them are strongly deformed .
various pasta - like shapes may be assumed .
correlation effects in this phase could also affect the neutrino opacities ( horowitz et al .
2004 , watanabe et al . 2004 ) .
for an immediate effect on core collapse , however , it would be required that this transition phase would reach to fairly low densities in order to affect the opacities at the neutrinospheres where the neutrino luminosities and spectra are set . a detailed overview over neutrino opacities in nuclear matter has recently been given by ( burrows , reddy & thompson 2004 ) .
an extensive quantitative overview over most reaction rates has also been provided in the literature ( bruenn & haxton 1991 ) .
\(iii ) neutrinos are produced by the capture of degenerate electrons from high energy levels . to some extent depending on the q - value of the capturing nucleus ,
the emitted neutrino starts with a high energy of the order of the electron chemical potential .
as the neutrino opacities scale with the squared neutrino energy , the initially trapped neutrinos will down - scatter to lower energies until the diffusion time scale becomes comparable to the thermalization time scale .
the thermalization in current collapse models occurs through neutrino - electron scattering because the energy transfer per collision with the light electron is more efficient than with the heavier nucleons .
the contribution of inelastic scattering of neutrinos off heavy nuclei depends on the individual nuclei and remains to be explored in detail .
once nuclear densities are reached at the center of the collapsing core , repulsive nuclear forces dominate the stiffness of the equation of state .
the collapse is halted by an outgoing pressure wave .
different snapshots of the velocity profile around bounce are shown in fig .
[ fig_velocity.ps ] .
the pressure wave travels through the inner core , where infall velocities are subsonic , and turns into a shock wave when it meets supersonic infall velocities at its edge .
the matter in the inner core experiences a rather adiabatic pressure wave and remains at low entropy @xmath69 kb per baryon . the shock wave in the outer core , however , heats matter to entropies larger than @xmath70 kb per baryon so that heavy nuclei are dissociated .
if the bounce - shock were to dynamically propagate through the core to expel outer layers in a prompt explosion , it would have to provide the energy to dissociate the material between the edges of the inner core and the iron core at a rate of @xmath71 erg per @xmath72 m@xmath73 of dissociated material ( the shock additionally suffers from unavoidable neutrino losses ) .
numerical simulations based on standard input physics and accurate neutrino transport exclude the possibility that the kinetic energy of the hydrodynamical bounce at nuclear densities drives a prompt supernova explosion ( liebendrfer et al .
2001 , bruenn et al .
2001 , rampp & janka 2002 , thompson et al .
2003 , sumiyoshi et al . 2005 ) .
while the trajectory through core collapse determines the state of the cold nuclear matter inside the protoneutron star ( pns ) , the mass of the hot mantle surrounding the pns grows by continued accretion .
the infalling matter is heated and dissociated by the impact at the fairly stationary accretion front .
before about @xmath74 ms after bounce , the entropy achieved by this shock - heating is higher than the maximum entropy obtained by infinite exposure to the prevailing neutrino field . with time
ongoing , the neutrinospheres recede to smaller radii and produce a harder neutrino spectrum . on the other hand ,
the accretion front hydrostatically moves to larger radii where the dissipated kinetic energy of the infalling matter is smaller .
hence , only after @xmath74 ms , when the entropy of shocked material is smaller and the neutrino luminosity higher , neutrino heating becomes effective behind the shock . behind the shock , matter continues to drift inward and to converge to higher densities .
the increasing electron chemical potential blocks the electron phase space and favors antineutrino capture in the gain region below the accretion front , and electron capture in the cooling region above the pns surface .
only expanding matter can lead to preferential neutrino capture as will be discussed in section [ sec : sniinucleosynth ] .
the gain radius separates the gain region from the cooling region .
it is the radius where cooling by neutrino and antineutrino emission balances heating by neutrino and antineutrino absorption , i.e. the point where the infalling fluid element reaches equilibrium entropy in the neutrino field produced at the pns surface .
the reaction time scale in the cooling region is comparable to or faster than the infall time scale so that an infalling fluid element adheres to the equilibrium entropy on its further journey toward the surface of the ns .
one half to two thirds of the neutrino luminosity in the heating region stems from the accreting matter in the cooling region , the smaller part diffuses out of the hot pns . because the accretion rate reacts to neutrino heating and the neutrino luminosity to the accretion rate , the conditions at the surface of the pns
are strongly coupled to the evolution of the matter behind the accretion front . in spherical symmetry
, the important neutrino transport between these layers can be numerically tackled in a fully consistent way by the accurate solution of the boltzmann transport equation for the three neutrino flavors ( rampp & janka 2002 , thompson et al .
2003 , liebendrfer et al .
2004 , sumiyoshi et al .
2005 ) , where the latter two references implemented all equations in general relativity .
one result of these calculations was that all progenitor stars between main sequence masses of @xmath75 and @xmath76 m@xmath73 showed no explosions in simulations of their postbounce evolution phases ( see fig .
[ fig : explosions ] ) ( liebendrfer et al . 2002 ) .
this indicates that the neutrino flux emanating from the pns has not the ample strength to blow off the surrounding layers for a vigorous explosion without the consideration of more details . among the reasons
are the strong deleptonization during collapse , the dissociation of heavy nuclei by the shock , a slow penetration of the neutrinos from layers at high density through layers with densities between @xmath77 g/@xmath57 , efficient neutrino cooling at the surface of the pns , and fast infall velocities through the gain region ( see e.g. janka 2001 ) . however , spherically symmetric simulations ignore fluid instabilities that are known to exist between the protoneutron star surface and the stalled shock , as well as deep in the protoneutron star ( herant et al . 1994 ) .
there is consensus that the convective overturn in the heating region increases the heating efficiency .
heated uprising plumes free the electron phase space for neutrino absorptions and convect the energy toward the stalled shock front instead of balancing the heating in loco by neutrino emission .
additionally , cool and narrow downflows can feed the heating by continued neutrino accretion luminosity .
however , first simulations in axisymmetry with energy - dependent neutrino transport do still not obtain vigorous explosions ( see buras et al .
2005 and references therein for a discussion of convection in the heating region ) . with respect to the convection in the pns ,
the luminosities are not significantly boosted because the crucial region around the neutrinospheres appears to remain convectively stable ( see bruenn et al .
2004 and references therein for a discussion of pns convection ) .
as two - dimensional axisymmetric calculations are now emerging as the new standard for simulations with reliable neutrino transport , several newly appreciated phenomena are currently explored in these computationally challenging calculations ( buras et al .
2005 , walder et al .
2004 , swesty & myra 2005 , burrows et al .
the consistent coupling of the cooling and heating regions with the feedback between accretion and neutrino luminosity is much more involved in multiple dimensions than under the assumption of spherical symmetry .
although axisymmetric supernova models add an essential new dimension to the world of spherically symmetric models , namely the possibility to accrete and expand matter at the same time at different locations , there are still important degrees of freedom missing
. hot and cool domains in the heating and cooling regions can only assume toroidal shapes around the axis of symmetry so that convection will always imply the motion of the whole torus .
axisymmetric simulations can not resolve small convective volumes or funnels linking the outer layers with the surface of the pns .
not only the fluid instabilities may bear an inherently three - dimensional topology , also the presence of magnetic fields may introduce new degrees of freedom that can only be explored in three dimensions .
first estimations of magnetic fields at the progenitor stage lead to rather low values around @xmath78 gauss ( heger et al . 2005 ) and stars with magnetic fields are subject to magnetic breaking of differential rotation ( maeder & meynet 2005 ) .
the uncertainties are yet large , but nevertheless , several explosion mechanisms have been suggested based on the influence of magnetic fields : large magnetic fields could collimate the matter outflow after the postbounce evolution and lead to jet - like explosions ( leblanc & wilson 1970 ) .
smaller fields could grow by winding through differential rotation and add magnetic pressure to the fluid pressure ( akiyama et al .
2003 , ardeljan et al .
in combination with differential rotation the magnetic fields could lead to the magneto - rotational instability so that the dissipation of the turbulent energy adds to neutrino heating ( thompson et al .
moreover , the magnetic field has been thought to possibly induce asymmetries to the neutrino heating ( kotake et al .
finally , it has been suggested that magnetic loops on the magnetized pns would heat the hot mantle in analogy to the solar corona ( ramirez - ruiz & socrates .
for all these reasons it is interesting to complement the axisymmetric simulations and accurate neutrino transport with simulations that support all degrees of freedom of three - dimensional space . already in one- and two - dimensional supernova models
it becomes clear that the computation time spent on hydrodynamics is negligible .
most time is spent on energy - dependent neutrino transport .
a systematic improvement of the neutrino transport from one to two dimensions alone required a substantial increase of computation time in yet incomplete implementations ( buras et al .
2005 , livne et al . 2004 ) and three - dimensional neutrino transport has not yet been attempted with a reliable resolution of the neutrino phase space . following three - dimensional approaches of ( fryer & warren 2002 , scheck et al .
2003 ) , we try to balance the computation time spent on hydrodynamics and neutrino transport by maximizing the degrees of freedom in the fluid dynamics in combination with approximations in the neutrino transport .
the effect of magnetic fields on the dynamics of the nucleons in the hot mantle has not yet been studied in three - dimensional numerical simulations with neutrino transport approximations . here
we continue to report on efforts in this direction ( liebendrfer , pen & thompson 2005 ) .
a simple and fast three - dimensional magneto - hydrodynamics code ( pen , arras & wong 2003 ) provides the core of our simulations .
it spans a central region of @xmath79 km@xmath80 with an equidistant resolution of @xmath81 km in cartesian coordinates .
this covers the hot mantle and part of the infalling layers .
the code has received a parallelization with mpi for cubic domain decomposition that minimizes the resources occupied on distributed memory machines by a simple and efficient reuse of buffer zones during the directional sweeps .
the finite differencing is second order accurate in time and space and handles discontinuities in the conservation equations with a total variation diminishing scheme .
furthermore , a specific choice of the finite differencing for the update of the magnetic field conserves its divergence to machine precision .
the computational domain of the mhd code is embedded in spherically symmetrically infalling outer layers evolved by a one - dimensional hydrodynamics code .
our collapse simulations are launched from a @xmath82 m@xmath73 progenitor model ( woosley & weaver 1995 ) .
the lattimer - swesty equation of state ( lattimer & swesty 1991 ) has been used .
we imposed a rotation with angular velocity @xmath83 rad / s along the z - axis with a quadratic cutoff at @xmath84 km radius .
along the same axis , we added a poloidal magnetic field of @xmath85 gauss .
the deleptonization during collapse has been parameterized in a very simple but carefully tested way ( liebendrfer 2005 ) .
an investigation of the spherically symmetric model g15 ( liebendrfer et al . 2005 ) with boltzmann neutrino transport reveals that the electron fraction during infall can roughly be approximated as a function of density @xmath54 . in our three - dimensional simulation ,
we update the electron fraction with @xmath86 $ ] , where the function @xmath87 has been read out of model g15 at the time of core - bounce . based on this parameterized deleptonization ( which also approximates the phenomenology of the various nuclear and weak interaction processes discussed in section [ sec : collapse.bounce.pbe ] ) ,
the changes of the entropy and the momentum by neutrino stress are also estimated during the collapse phase .
the effective gravitational potential of ( marek et al .
2005 ) has been used to implement general relativistic effects .
[ fig : parameterization.ps ] compares the velocity and entropy profiles in a three - dimensional collapse calculation without rotation and magnetic fields with the corresponding general relativistic spherically symmetric calculation based on boltzmann neutrino transport .
it becomes evident that the scheme reproduces the collapse phase very accurately , while important features of the postbounce phase are not yet captured , for example the neutrino burst , or neutrino heating . due to the rotation ,
the polar infall velocities are slightly larger than the equatorial infall velocities .
when the central density reaches @xmath88 g@xmath89 the magnetic field lines become visibly distorted . in the very early collapse phase
this is due to matter inflow along the rotation axis , meeting at the stellar center , and pushing field lines outward into the equatorial plane .
moreover , due to the centrifugal forces , the projection of the velocities onto the plane orthogonal to the rotation axis is largest at about @xmath84 km above and below the gravitational center .
these are then the locations where the magnetic field lines condense most rapidly , bending slightly outward around the center . with ongoing collapse , this effect shifts to smaller radii and becomes more pronounced . at bounce ,
the magnetic field exceeds @xmath90 gauss in these hot spots located @xmath91 km above and below the center .
the field lines run along double cones aligned with the z - axis , except for the small deviation that circumvents the center . in the early shock expansion until @xmath92 ms after bounce , the shock front is almost spherically symmetric .
afterward , the simulation becomes unrealistic , because the dynamically important neutrino burst is not implemented . behind the expanding accretion front ,
entropy variations due to variations in the shock strength induce fluid instabilities that entangle the magnetic field lines .
[ fig : collapse.ps ] and [ fig : bounce.ps ] show a subdomain of the three - dimensional simulation at @xmath93 ms before bounce and at @xmath92 ms after bounce , respectively .
the simulations have been performed on @xmath94 processors of the @xmath95 processor mckenzie cluster at cita .
they required a wall clock time of @xmath96 hours .
if the main computational effort can be spent on magneto - hydrodynamics , three - dimensional simulations allow a spatial resolution comparable to the resolution used in spherically symmetric simulations with boltzmann neutrino transport .
of course , this is only possible as long as the computationally expensive neutrino transport can be replaced by adequate and efficient parameterizations .
the complexity of neutrino transport and the frequent failure of self - consistent models for core collapse supernovae to produce explosions have generally divorced modeling of core collapse supernova nucleosynthesis from modeling of the central engine . despite the ( fundamental ) problem that the supernova mechanism is still not understood , supernova nucleosynthesis predictions have a long tradition .
all of these predictions rely on artificially introduced explosions , replacing the central engine either with a parameterized kinetic energy piston ( woosley & weaver 1995 , rauscher et al .
2002 , chieffi & limongi 2004 ) or a thermal bomb ( thielemann et al .
1996 , nakamura et al .
the explosion energy and the placement of the mass cut ( separating ejected matter from matter which is assumed to fall back onto the neutron star ) are tuned to recover the desired ( observed ) explosion energy and ejected @xmath97ni mass .
both approaches are largely compatible ( aufderheide et al . 1991 ) and justifiable for the outer stellar regions .
it is this inner region , where most of the fe - group nuclei are produced , which is most affected by the details of the explosion mechanism , especially the effects of the interaction of nuclei with the large neutrino flux .
the question arising is how more realistic supernova nucleosynthesis predictions could be made , given the existing problem with self - consistent explosions .
discussed improvements possibly leading to explosions are rotation and magnetic fields ( thompson 2000 , thompson et al .
2004 ) or uncertainties in neutrino opacities ( see e.g. burrows et al . 2004 ) or other microphysics properties .
they would introduce additional mixing at the neutrino sphere and convective transport or change the neutrino luminosity via improved opacities .
this indicates two options for successful explosions : ( a ) enhanced neutrino luminosities or ( b ) enhanced deposition efficiencies for neutrino captures in the convective layers .
we use two different methods to enforce explosions in otherwise non - explosive models . in a first approach ,
we parameterize the neutral current neutrino scattering opacities .
this helps to artificially increase the diffusive fluxes in regions of very high matter density , resulting in a faster deleptonization of the proto - neutron star such that the neutrino luminosities are boosted in the heating region . in a second approach ,
explosions are enforced by multiplying the reaction rates for forward and backward reactions in @xmath98 and @xmath99 in the heating region by equal factors .
this reduces the time scale for neutrino heating as it is expected in combination with overturn in this convectively unstable domain .
both approaches allow successful explosions with a consistently emerging mass cut without the need to add external energy .
figure [ fig : explosions ] shows the position of the shock front as function of time for different core collapse supernova simulations . in all these models the shock front runs outwards , stalls at @xmath100s , and subsequently retrieves , leaving us with a non - explosive model .
the dashed lines stem from models with modified neutrino - nucleon elastic scattering as described above . in the trajectories from these exploding models
three phases can be distinguished : an initial outwards propagating shock , which stalls around 0.1s after bounce .
but shortly before 200ms after bounce , the accretion front moves outwards again .
while the importance of neutrino interactions is manifest and well documented in the @xmath2__-process _ _ and the _ r_-process , neutrinos potentially impact all stages of supernova nucleosynthesis .
because of the impact of the neutrinos on the nucleosynthesis , the nucleosynthesis products from future explosion simulations ( utilizing multi - group neutrino transport ) will be qualitatively different from either parametrized bomb or piston nucleosynthesis models . the dominant process are @xmath2/@xmath101 and @xmath102 captures on shock dissociated free nucleons , though at later times the more poorly known @xmath2/@xmath101 captures on heavy nuclei may contribute significantly . an indispensable quantity to describe explosive nucleosynthesis in the innermost ejecta
is the electron fraction @xmath103 ( the number of electrons per nucleon ) .
this @xmath103 is set by weak interactions in the explosively burning layers , i.e. electron and positron capture , beta - decays , and neutrino or anti - neutrino captures .
we examined the effects of both electron and neutrino captures in the context of recent multi - group supernova simulations .
these models are based on fully general relativistic , spherically symmetric simulations ( liebendrfer et al .
pruet et al .
( 2005 ) have performed similar simulations using tracer particles from two dimensional simulations ( buras et al .
2003 ) . in both cases , artificial adjustments to the simulations were needed to remedy the failure of the underlying models of central engines to produce explosions ( see also section [ sec : snmodelling ] for details on how to invoke an explosion in a non - explosive model ) .
also in both cases , the neutrino transport could not be run to later times and the simulations were mapped to a more simple model at later times . despite these shortcomings
, these simulations nevertheless reveal the significant impact of neutrino interactions on the composition of the ejecta .
we find that all our simulations that lead to an explosion by neutrino heating develop a proton - rich environment around the mass cut with @xmath104 ( frhlich et al .
2006 ) , as it is required from galactic evolution and solar abundances .
several phases can be identified in the evolution of the electron fraction of the matter that will become the innermost ejecta . at early times matter
is degenerate and electron capture dominates . at the same time matter
is being heated by neutrino energy deposition and subsequently , the degeneracy is lifted . while the ratio between electron captures and positron captures significantly decreases , neutrino absorption reactions start to dominate the change of @xmath103 . as the matter expands the density decreases and eventually the electron chemical potential drops below half the mass difference between the neutron and proton .
with both neutrino absorption and emission processes favoring a higher electron fraction , @xmath103 rises markedly in this phase , reaching values as high as 0.55 .
the global effect of this proton - rich ejecta is the replacement of previously documented overabundances of neutron rich iron peak nuclei ( woosley & weaver 1995 , thielemann et al .
production of @xmath105ni is suppressed while @xmath106sc and @xmath107ti are enhanced .
the results ( fig .
[ fig : elemental ] ) for the elemental abundances of scandium , cobalt , copper , and zinc are closer to those observed ( see frhlich et al .
2006 for details ) .
we have found that the transformation of protons into neutrons by neutrino captures allows ( n , p)-reactions to take the place of the @xmath108-decays of waiting point nuclei ( with lifetimes longer than the expansion timescale ) , allowing significant flow to @xmath109 .
we termed this process the @xmath110-process ( see frhlich et al .
2005 for more details ) due to the essential role of the neutrinos in producing these light p - nuclei .
these results clearly illustrate the need to include the full effect of the supernova neutrino flux on the nucleosynthesis if we are to accurately calculate the nucleosynthesis from core collapse supernovae .
two series of pre - sn models and their yields were presented in this paper .
the first series consists of 20 @xmath1 models with varying initial metallicity ( solar down to @xmath0 ) and rotation ( @xmath111kms@xmath18 ) .
the second one consists of models with an initial metallicity of @xmath0 , masses between 20 and 85 @xmath1 and average rotation velocities at these metallicities ( @xmath112kms@xmath18 ) .
the most interesting models are the models with @xmath0 ( [ fe / h]@xmath113 ) . in the course of helium burning , carbon and oxygen are mixed into the hydrogen burning shell .
this boosts the importance of the shell and causes a reduction of the size of the co core . later in the evolution ,
the hydrogen shell deepens and produces large amount of primary nitrogen . for the most massive models ( @xmath114@xmath1 ) ,
significant mass loss occurs during the red supergiant stage .
this mass loss is due to the surface enrichment in cno elements via rotational and convective mixing .
the yields of the rotating 20 @xmath1 models can reproduce the observed abundances at the surface of extremely metal poor ( emp ) stars and the metallicity trends .
the wind of the massive models can also reproduce the cno abundances of the carbon rich umps , in particular for the most metal - poor star known to date , he1327 - 2326 .
the inclusion of neutrino interactions in explosive nucleosynthesis changes significantly the final abundances by changing @xmath103 to values above 0.5 in the innermost ejecta .
neutrino interactions also enables nucleosynthesis to go beyond p rich waiting point nuclei .
the new results for iron group nuclei are in much better agreement with observations of emp stars ( sc , ti , ni , zn , ... ) .
m. liebendrfer , o. e. b. messer , a. mezzacappa , r. w. hix , f .- k .
thielemann , k. langanke , in _ proceedings of the 11th workshop on `` nuclear astrophysics '' _ , edited by w. hillebrandt , e. mller ( ringberg castle , tegernsee , germany , 2002 ) , pp . | _ recently , measurements of abundances in extremely metal poor ( emp ) stars have brought new constraints on stellar evolution models .
indeed , these stars are believed to have been enriched only by one or a few stars .
the abundances observed in emp stars can therefore be almost directly compared with the yields of very metal poor or metal free stars .
this removes the complex filter of chemical evolution models . in an attempt to explain the origin of the abundances observed , we computed pre
supernova evolution models , explosion models and the related nucleosynthesis . in this paper , we start by presenting the pre - sn models of rotating single stars with metallicities ranging from solar metallicity down to almost metal free .
we then review key processes in core - collapse and bounce , before we integrate them in a simplistic parameterization for 3d mhd models , which are well underway and allow one to follow the evolution of the magnetic fields during collapse and bounce .
finally , we present explosive nucleosynthesis results including neutrino interactions with matter , which are calculated using the outputs of the explosion models .
_ the main results of the pre - sn models are the following .
first , primary nitrogen is produced in large amount in models with an initial metallicity @xmath0 .
this large production is due to mixing of carbon and oxygen in the h - burning shell and the main production occurs between core helium and carbon burnings .
second , at the same metallicity of @xmath0 and for models with an initial mass larger than about 60 @xmath1 , rotating models may experience heavy mass loss ( up to more than half of the initial mass of the star ) .
several solar masses are lost during the main sequence when the star reaches break - up velocities .
the largest amount of mass is lost during the red supergiant stage after rotational mixing and convection have enriched the surface in cno elements and therefore increased its metal content .
the chemical composition of these winds can qualitatively reproduce the abundance patterns observed at the surface of carbon - rich emp stars .
third , our models predict an upturn of c / o at very low metallicities .
explosive nucleosynthesis including neutrino - matter interactions produce improved abundances for iron group elements , in particular for scandium and zinc .
it also opens the way to a new neutrino and proton rich process ( @xmath2p - process ) able to contribute to the nucleosynthesis of elements with a@xmath364 .
* stellar evolution of massive stars + at very low metallicities * + r. hirschi , c. frhlich , m. liebendrfer and f .- k . thielemann + dept . of physics and astronomy , university of basel + klingelbergstr .
82 , 4056 basel , switzerland + raphael.hirschiunibas.ch |
understanding detector response to low energy nuclear recoils is imperative to the interpretation of experimental results from a detector designed to search for wimps ( weakly interacting massive particles ) , a dark matter candidate .
direct detection of low mass wimps occurs in the low energy region of detectors with a threshold down to sub kev .
since the threshold energy represents the visible energy in the detector , understanding the ionization efficiency of the detector response to low energy nuclear recoils is crucial to calculating the recoil energy .
an example of the need for this requirement is the claim of experimental evidence for dark matter by cogent @xcite that has been unverified by cdms ii @xcite .
both experiments use germanium as the target material .
thus , similar results are expected if the detection thresholds for both experiments were determined using a standardized ionization efficiency , which accurately accounts for all processes that occur at a low energy range .
this standardized ionization efficiency must also be validated with measurements in combination with reliable monte carlo simulations .
two different approaches can be used for modeling ionization efficiency in germanium detectors @xcite .
one model , traditionally used for a number of different detector materials , is proposed by lindhard _ et .
@xcite : @xmath4 where @xmath1 = 0.133@xmath5 , @xmath6 , and @xmath7 = 11.5@xmath8 for a given atomic number , @xmath9 , mass number , @xmath10 , and recoil energy , @xmath11 .
however , this model has not been proved accurate at low energies as the theoretical derivation has uncertainties in this region @xcite . another model designed for low energy interactions in germanium
was proposed by barker and mei @xcite .
this model takes into account the fraction of nuclear stopping power that contributes to the ionization efficiency at low energies @xcite .
the barker - mei model can be expressed as : @xmath12}.\ ] ] this model is valid for recoil energies , @xmath13 , from 1 kev to 100 kev .
however , the barker - mei model has not been experimentally verified .
the purpose of this paper is to address that issue .
a comparison between the two models and existing data was performed @xcite , and fig .
[ fig : bmlindwdata ] shows both models together with available experimental data . as shown in fig .
[ fig : bmlindwdata ] , the two models agree in the low energy region but disagree with the available data points in the same region
. further validation of the lindhard and barker - mei models , using more measurements and more accurate monte carlo simulation , was necessary .
taking exact nuclear recoil measurements can be challenging at lower energies when systematic errors are introduced by a variety of sources .
for example , thermal neutrons and elastic / inelastic scattering have their own uncertainties . utilizing thermal neutrons requires the full absorption of out - going gamma rays measured by another detector in coincidence . without this additional measurement ,
compton scattering from out - going gamma rays within the germanium detector can contaminate the visible energy . in a measurement of neutron elastic scattering , a monte carlo simulation
must be incorporated to exclude multiple scatters .
it is also necessary to precisely measure the scattering angle and time - of - flight of the out - going neutrons .
with inelastic scattering , the compton scattering due to de - excitation of gamma rays in the detector can contaminate the signal .
all of these techniques need to be implemented with an accurate monte carlo that reduces systematic errors in order to obtain a reliable ionization efficiency
. however , current popular simulation tools often need to be tuned to simulate inelastic scattering processes . therefore , we desired a simple method to accurately measure nuclear recoils in germanium .
the @xmath0 transition of germanium-72 ( @xmath14ge(n , n@xmath15e ) ) , which is the internal conversion process for this nucleus , was chosen .
the @xmath0 transition of @xmath14ge(n , n@xmath15e ) is induced when neutrons inelastically scatter off a @xmath14ge nucleus .
after this collision , the @xmath14ge nucleus is left in an excited state .
when the nucleus returns to ground state , it does not directly produce a gamma - ray ( as is common for other nuclei ) , but interacts electromagnetically with the inner shell electrons and causes one to be emitted from the atom @xcite .
the physics process for the internal conversion of @xmath14ge can be generalized as : @xmath16 where @xmath14ge@xmath17 denotes the excited 0@xmath18 state .
the total detectable energy from the @xmath14ge@xmath17(0@xmath18 ) to @xmath14ge(0@xmath18 ) transition is 691.6 kev , which includes energy from the x - ray .
this energy is well within the visible range of traditional detectors , and both electron and x - ray can be detected with a consistency of approximately 100% .
the characteristic energy , 691.6 kev , is superimposed with the nuclear recoil of a @xmath14ge nucleus to form a quasi - triangular shape , which is distinguishable from other processes in the spectrum .
thus , we can observe low energy nuclear recoils without taking low energy nuclear recoil measurements by extrapolating with the known quasi - triangular fit .
this quasi - triangular shape , starting at 691.6 kev , has been studied in great detail @xcite .
the tail of the quasi - triangular shape comes from a combination of the 691.6 kev energy deposition and nuclear recoil energy due to neutron scattering .
depending on the incident neutron energy and scattering angle @xcite , nuclear recoil energy can be expressed as : @xmath19 where e@xmath20 is the nuclear recoil energy , e@xmath21 is the neutron energy , m is the nucleus mass , m is the neutron mass , and @xmath22 is the scattering angle between the incident neutron and the recoil nucleus .
the quasi - triangular shape is created by adding the energy from the nuclear recoils to the 691.6 kev energy deposition : @xmath23 where @xmath24 is the observed energy and @xmath7 is the ionization efficiency .
the nuclear recoil energies can be determined using a monte carlo simulation with applied ionization efficiency , if an agreement with the measurements is obtained . to validate the ionization efficiency models proposed by lindhard and barker - mei , a geant4.9.2-based monte carlo simulation package @xcite , corrected for the internal conversion processes , was used .
this simulation package was verified using a well - calibrated @xmath25co radioactive source with the same experimental setup . utilizing the spectrum measurements with a substantiated monte carlo simulation
, we compared the unique quasi - triangular shape induced by the 691.6 kev electrons and nuclear recoils in the data to the monte carlo simulations for the two models .
shape analysis was used to verify the quasi - triangular shape of spectra in the data and monte carlo simulations by analyzing data points in the region of interest .
we found a good agreement between the measurements and the lindhard ( with @xmath1 = 0.159 ) and barker - mei models . in this paper
, we corroborate the ionization efficiency models proposed by lindhard _
et al _ and barker - mei with a neutron induced @xmath0 transition for @xmath14ge .
the experimental design is discussed in section 2 , followed by data analysis in section 3 . the monte carlo simulation is demonstrated and validated in section 4 , and the comparison with data described in section 5 .
section 6 summarizes our results .
the germanium detector used in our experiment was an old coaxial detector from princeton gamma tech , model rg11b / c @xcite .
its linearity of energy response between consecutive calibrations was within 0.35% .
however , the full width at half maximum ( fwhm ) was 7.1 kev at the 1173 kev @xmath25co peak , a factor of two worse in energy resolution than a new germanium detector . nevertheless , the detector was sufficient for measuring neutron induced internal conversion .
we used an @xmath26am-@xmath27be ( ambe ) source with neutron energies ranging from @xmath281.0 to 11.2 mev , at a frequency of 100 hz @xcite .
an ambe neutron source produces neutrons in four discrete groups : @xmath29 , @xmath30 , @xmath31 , and @xmath32 , which populate the ground state , the 4.443 mev level , the 7.65 mev level , and the 9.64 mev level of the @xmath33c product nucleus , respectively @xcite .
the @xmath30 group neutrons , accompanied by 4.443 mev gamma rays , dominate the production of neutrons in an ambe source @xcite .
this feature allowed us to set up a coincidence measurement using sodium - iodide ( nai ) detectors ( bicron model number 3m3/3 @xcite ) with a threshold of above 1 mev . utilizing this coincidence method , the nai detectors measured gamma rays while neutrons were detected with the germanium detector . this coincidence pattern
required that the nai and ge detectors both trigger in order for an event to be recorded , suppressing random background events generated by gamma rays from surrounding materials .
we took approximately 22 days of data with three nai detectors and 13 days of data with two nai detectors .
due to low neutron emission , the source was placed on the center of the ge detector cap and held in place by electrical tape . when using three nai detectors , they were placed on the right , the left , and directly in front of the germanium detector .
when using two nai detectors , they were placed on the right and on the left of the germanium detector .
[ fig : expsetup ] shows the experimental set - up .
the data acquisition was performed using a national instruments pxi-1031 system @xcite and igor pro 4.07 software @xcite .
each run lasted approximately 4.5 days .
the data from each individual run was added consecutively off - line using analysis code from the root software package @xcite .
evidence of the 691.6 kev @xmath0 transition peak can be seen in fig .
[ fig : expdata ] .
after 35 days , the number of events in the region of interest ( 675 @xmath28 765 kev ) was approximately 500 . this is consistent with the known neutron emission from the ambe source and the coincidence method , which suppresses random background events from gamma rays , as verified by the monte carlo simulation .
because of the small sample size of valid data in the region of interest , the bin size in the histogram was set to 6 kev in order to mitigate statistical fluctuation . as shown in fig .
[ fig : expdata ] , there are several peaks near the region of interest .
we identified these peaks in order to help understand the processes that took place within our detector as well as possible sources of contamination .
the first peak can be fitted using a standard gaussian distribution and linear background distribution .
this is given by : @xmath34-p_3 + p_4e,\ ] ] where @xmath35 = 37 @xmath2 1.4 is the normalization constant , @xmath36 = 508 @xmath2 0.2 kev is the center value , @xmath37 = 4.7 @xmath2 0.2 kev is the gaussian width , @xmath38 = -65 @xmath2 12 is a constant , @xmath39 = 0.2 @xmath2 0.02 is the slope , and @xmath24 is the energy in kev .
the peak at 511 kev is mainly from the annihilation of e@xmath40e@xmath18 pairs induced by cosmic rays passing through the surrounding materials with other minor contributions .
most notable is the 4.443 mev induced e@xmath41e@xmath42 in the surrounding materials .
the positrons can annihilate with electrons producing 511 kev gamma rays , which enter the germanium detector .
the fitted central value of 508 kev is slightly lower than the expected 511 kev , but still within the margin of error for the given bin size .
the remaining peaks can be fitted using a moyal distribution and linear background .
the moyal distribution used is @xcite : @xmath43{\frac{1}{2\pi}exp\left[-\left(r\left(e - e_{mpv}\right)+exp\left[r\left(e - e_{mpv}\right)\right]\right)\right ] } , \label{moyalfit}\ ] ] where @xmath44 is a constant and @xmath45 is the most probable value of energy deposition in the detector . @xmath44 and @xmath45 are physically significant parameters . the value of @xmath44 is related to the reciprocal of the stopping power and can be used to calculate average path length for particles in the medium .
the most probable value of energy , @xmath45 , is the most common energy deposition in that region .
these parameters are further discussed in section 3.3 and used to interpret the experimental data .
the peak around 560 kev , is likely caused by events from @xmath46ge@xmath47 inelastic scattering . because of the small number of events , the fitting function was only partially accurate .
the second peak at around 600 kev has the fitted parameters , @xmath44 = 0.06 @xmath2 0.02 and @xmath45 = 609 @xmath2 3.6 kev .
this peak is likely the combination of inelastic scattering from @xmath48ge@xmath47 and the neutron capture on @xmath49ge , @xmath49ge@xmath50 .
finally , the peak of interest , 691.6 kev , exhibits the internal conversion of @xmath14ge@xmath51 .
the fitted parameters are , @xmath44 = 0.06 @xmath2 0.02 and @xmath45 = 705 @xmath2 3.0 kev .
all fits in the region of interest are shown in fig . [
fig : expdatafitall ] .
we also took a background spectrum using the coincidence method to identify any random coincidence .
this spectrum is also shown in fig .
[ fig : expdatafitall ] .
the background data was taken using two sodium iodide detectors ( see right of fig . [
fig : expsetup ] ) for 4.6 days . after the background spectrum was collected , we subtracted it from the ambe data sets as shown by fig .
[ fig : bkrdsubtracted ] .
the 511 kev peak was significantly reduced , verifying our conjecture on the main origin of this peak as induced pair production from cosmic rays in the surrounding materials .
the subtracted data set was also fitted using a moyal distribution and linear background .
it was found that the most probable value of energy deposition in the detector for the @xmath0 transition was 696 @xmath2 1.4 kev .
thus the percent difference of the ambe data with and without background subtraction is 1.3% .
we have identified the following sources of systematic error for our data analysis : the energy scale and the energy resolution .
the associated energy scale error is 3 kev , which is from the fit of the known 511 kev peak .
the center value of the fitted function was 508 kev which is a difference of 3 kev from the known value of the peak .
this shift in energy is from the energy calibration used to convert to the energy scale . as stated previously , there was a 0.35% error in successive energy calibrations . using a more accurate energy scale to produce a peak at 511 kev caused other energy regions to become less accurate . for energy resolution ,
the value from the 662 kev peak of @xmath52cs was used , it is nearest of the calibration sources to our region of interest .
the energy resolution for this peak was 3.45 kev . adding two errors in quadrature , since they are independent of each other , causing the resulting systematic error to be 4.6 kev .
thus , the larger bin size of 6 kev was used to accommodate for this error .
statistically , there are approximately 500 events in the region of interest ( 675 kev @xmath28 765 kev ) which gives a statistical error of 4.5% . for data with the background subtracted out ,
the error associated with the most probable value is 696 @xmath2 4.6 ( sys ) @xmath2 1.4 ( stat ) .
electrons and nuclear recoils that travel through germanium with a high momentum lose energy by exciting and ionizing the germanium atoms .
the average amount of energy lost can be calculated with the bethe - bloch equation @xcite .
however , the energy transfer is not a continuous process .
it occurs through random collisions during which various amounts of energy can be transferred .
the energy loss in the detector can be described by the moyal distribution , eq .
[ moyalfit ] , as shown in fig .
[ fig : expdatafitall ] . as shown in eq.[moyalfit ]
, @xmath44 is the reciprocal of the density , @xmath53 , times the average path - length of charged particles in the detector , @xmath54 , times the parameter , @xmath55 , as given by the bethe - block formula : @xmath56 where @xmath55 = 4@xmath57 is related to the stopping power , with constants @xmath58 , avogadro s number ; @xmath59 , the mass of electrons ; @xmath60 , the speed of light ; @xmath61 , the bohr radius ; @xmath62 , the electron charge ; @xmath9 , atomic number of target ; @xmath10 , the atomic mass number of target ; and @xmath63 , the speed of charged particles divided by @xmath60 .
the density of germanium is @xmath53 = 5.323 @xmath64 .
since electronic stopping power is different from nuclear stopping power , the value of @xmath44 is very different for pure electronic recoils than nuclear recoils .
thus , the value of @xmath44 can be used to derive the average path - length for a given electronic recoil or nuclear recoil . however
, the value of @xmath44 from our measurements is a combination of electronic and nuclear recoils , and it can not be used to directly determine the average path length for either .
nevertheless , we can use the most probable value of energy deposition , which is related to the stopping power multiplied by the average path length , to determine the average path length .
given @xmath45 = 696 @xmath2 4.6 ( sys ) @xmath2 1.4 ( stat ) kev , obtained from the fitted function in fig .
[ fig : bkrdsubtracted ] , is a convolution of the 691.6 kev energy deposition and the nuclear recoils , then we contend that 696 kev = @xmath65 @xmath66 , where @xmath65 is the mass stopping power in kev @xmath67/g .
since there are both electronic and nuclear recoils , we can rewrite this as : @xmath68 where @xmath69 = 1301 kev @xmath70/g @xcite and @xmath71 = 691.6 kev ; thus @xmath72 = 691.6 /(@xmath73 ) = 0.1 cm , which is the average path length in germanium for electrons with an energy of 691.6 kev .
therefore , we have @xmath74 and @xmath75 from eq.[eq : recoil ] , we can conclude that the most probable nuclear recoils that we measured in the detector have an average of 4.4 @xmath2 0.007 ( sys ) @xmath2 0.002 ( stat ) kev electronic equivalent energy , which corresponds to approximately 17.5 @xmath2 0.12 ( sys ) @xmath2 0.035 ( stat ) kev nuclear recoil energy from ionization efficiency in the barker - mei model . to determine @xmath76
, we can use @xmath77 = 2341720 kev @xmath70/g for a nuclear recoil of 17.5 kev @xcite .
thus , @xmath76 = 17.5 kev /(@xmath78 ) = 0.014 @xmath3 m is the average path length in germanium for 17.5 kev nuclear recoils .
this determines the range of low energy nuclear recoils in a germanium detector .
an accurate monte carlo simulation is needed to determine the validity of the lindhard and barker - mei models when compared to collected data .
two crucial steps were taken before creating a monte carlo simulation that would determine nuclear recoil energy : 1 ) the @xmath0 transition in geant4.9.2 @xcite was fixed and 2 ) the monte carlo simulation was verified with a well - known gamma - ray source . by studying the inelastic scattering processes that contribute to internal conversion
, we found that the @xmath0 transition code is included in geant4 , but is missing neutron data for @xmath14ge .
specifically , geant4 does not provide @xmath79/@xmath80 transition data in each energy level ( data type = 12 ) and has no cross section data corresponding to the @xmath0 transition . in order to make the internal conversion process available in our geant4 simulation
, we created a @xmath79/@xmath80 ratio in the database for @xmath14ge with geant4.9.2 several years ago ( see our correction in ref .
@xcite ) . in addition
, we calculated all the cross - sections of the @xmath14ge@xmath81 reaction using talys @xcite , a reliable software for the simulation of nuclear reactions .
we then converted the transition cross - section data into the transition ratio in geant4.9.2 as shown in fig .
[ fig : transmap ] ( we utilize geant4.9.2 for this work because the e@xmath82 transition problem was amended several years ago ) .
the independence of this ratio can be cross - checked by looking at the reactions in different energy scales .
eventually , the talys data was converted to geant4 format where it could be used in the internal conversion process ( see the supplemental data used in geant4.9.2 from ref .
@xcite ) . during the @xmath0 transition , a characteristic x - ray , or auger electron ,
is produced simultaneously with the conversion electron .
this process was not included in the geant4 code .
however , since this is a complicated process that involves binding energy , it would require editing for all the elements in the geant4 package . for the sake of simplicity , we combined the x - ray and conversion electron together , since they both contribute to the total energy deposition ( in a solid germanium detector only ) .
a more general solution will be required when considering other materials .
a simulation was created from the geant4.9.2 and g4ndl3.13 packages that includes the additional transition ratio data . for further information
, readers can refer to ref .
@xcite for detailed geometry and experimental framework .
after the missing data was added , the @xmath0 transition of @xmath14ge@xmath83 could be simulated using geant4.9.2 .
this makes the geant4.9.2 simulation more reliable for dark matter searches utilizing germanium detectors . to obtain reliable results from our monte carlo
, it was essential to verify the simulation .
@xmath25co , with an original radioactivity of 1.0 @xmath3ci , was used to validate the simulation .
the @xmath25co source was mounted on the center of the ge detector cap as shown in fig .
[ fig : expsetup ] ( right ) . the corresponding geometry used in the monte carlo simulation
is shown in fig .
[ fig : gegeometry ] .
since the germanium detector is aged , the monte carlo geometry was modified to include a dead layer of thickness 2.5 mm ( see fig .
[ fig : gegeometry ] , left ) .
we also implemented a smearing process by applying energy resolution to the spectrum of deposited energy in the active germanium . when fitting the energy resolution ,
these four peaks were identified : 1173 kev , 1332 kev , 2506 kev and 662 kev .
the first three peaks are gamma rays from the @xmath25co source and the last peak is a gamma ray from @xmath52cs source . the best fit function is presented in eq . [ enresfit ] : @xmath84 where @xmath24 is the energy in kev
. under the square root , the first term , 0.3 , is the percentage of energy resolution , for a new ge detector ; the second and third terms are due to noise and the age of the ge detector .
the percentage of relative energy resolution as a function of energy is plotted in fig . [ fig : enres ] using this fitted function , we applied the energy resolution to the range 600 kev @xmath28 2510 kev and obtained agreement between the experimental data and monte carlo simulation as shown in fig .
[ fig : co60 ] .
this validates that our monte carlo can be reliably used for gamma - ray simulation .
using the modified geant4.9.2 package , we performed simulations to verify the accuracy of the lindhard and barker - mei models with an ambe neutron source . the experimental setup for our ge and nai detectors , as shown in fig .
[ fig : expsetup ] , was simulated according to the dimension and material information provided by the manufacturer @xcite . since the @xmath14ge@xmath51 reaction causes a quasi - triangular shape in the data , we expect to see this same feature in our monte carlo simulation if the lindhard and barker - mei models are accurate . in order to provide an accurate simulation ,
the geometry , as well as the ambe source generator , need to be implemented correctly .
because the ambe source was placed very close to the germanium detector , gamma rays emitted from the source have a greater chance of entering the germanium detector and causing contamination in the region of interest ( 675 kev @xmath28 765 kev ) .
thus , in the simulation it is necessary to account for all potential gamma rays emitted from the ambe source .
two reactions , shown in eqs . [ am241 ] and [ be9 ] , occur in the ambe source : @xmath85 @xmath86 eq .
[ am241 ] shows the decay of @xmath26am to @xmath87np , which causes the emission of alpha particles and gamma rays .
[ be9 ] shows the reaction between an alpha particle and @xmath27be . in eq .
[ be9 ] , the energy of gamma rays emitted depends on the resulting state of @xmath33c , which is 4.443 mev ( for the 1st excited state ) , 7.65 mev ( for the 2nd excited state ) and 9.64 mev ( for the third excited state ) . using the gamma ray energy and recoil energy of @xmath33c ,
the resulting neutron energy can be calculated by applying energy conservation to eq .
[ be9 ] .
fully absorbed gamma rays ( from eq .
[ be9 ] ) will not be in the region of interest ( 675 kev @xmath28 765 kev ) since their energies are at a few mev .
however , the compton continuum of their interaction can contribute to the region of interest .
in addition , gamma rays ( from eq . [ am241 ] ) with energies 26.34 kev ( a branching ratio of 2.4% ) , 59.54 kev ( a branching ratio of 35.9% ) and 722.01 kev ( a branching ratio of 0.000196% ) @xcite can contribute to the region of interest by occurring in coincidence with the e@xmath82 transition ( 26.34 kev and 59.54 kev ) or by becoming fully absorbed ( 722.01 kev ) .
we generated these three gamma rays in our monte carlo simulation along with gamma rays at energies of 4.443 mev , 7.65 mev , and 9.64 mev caused by the transitions in the excited @xmath33c nucleus .
the simulated neutron energy spectrum from the ambe neutron source is shown in fig .
[ fig : ambeneutron ] .
the spectrum agrees with the prediction from marsh et .
al . @xcite .
after this validation , the ambe source generator was implemented in the monte carlo simulation .
the simulated results are presented as a spectrum of energy deposited in the germanium detector after the application of smearing , which is the process of accounting for energy resolution in a germanium detector .
gaussian and moyal distributions with fitted parameters from our data ( section 3 ) have been incorporated into the smearing of different peaks in the region of interest .
we used the model proposed by lindhard _
et al . _
@xcite ( @xmath1 = 0.159 ) to determine the ionization efficiency for a germanium detector .
[ fig : dephpgeafter ] shows the simulated energy deposition spectrum after smearing .
the @xmath0 , 691.6 kev , transition is visible , as shown in fig .
[ fig : dephpgeafter ] .
appearance of this peak indicates that the internal conversion process has been successfully implemented in our geant4.9.2 .
after successfully simulating the ambe neutrons in our geant4.9.2 , we collected the energy deposited in the germanium detector .
this collected energy spectrum was compared with our data after the application of smearing ( eq . [ enresfit ] ) .
two models utilizing ionization efficiency functions were applied to the monte carlo spectrum : lindhard @xmath1 = 0.159 and the barker - mei model .
normalization was applied to the energy range 675 to 765 kev . by overlaying the monte carlo energy spectrum and the collected data
, we were able to perform a shape analysis on the characteristic @xmath0 transition .
this is demonstrated in fig .
[ fig : aftbksub ] .
the shape analysis was performed on a bin - to - bin basis by comparing the data and two monte carlo simulations , which correspond to the two models .
the difference in shape between the data and the two monte carlo simulations was calculated using @xmath88 .
table [ tab : table1 ] shows the results of this comparison in which the largest difference is shown to be less than 4% .
.the percent difference between the collected data and the monte carlo simulations with two models . [ cols="^,^,^,^,^,^,^,^,^,^",options="header " , ] using @xmath89 , fig .
[ fig : efficiency ] shows the extracted ionization efficiency from the data and monte carlo simulation .
note that this is not a direct measurement of ionization efficiency , but an extraction using the shape analysis .
we took measurements using an ambe neutron source , incident on a germanium detector , for a total of 35 days . the characteristic quasi - triangular shape located at 691.6
kev represents the e@xmath82 transition of germanium-72 .
all peaks in the region of interest have been identified , thereby confirming source related and environmental backgrounds . the unique quasi - triangular shape induced by neutrons
can be described using the moyal distribution with the fitted parameters @xmath44 = 0.06 @xmath2 0.01 ( stat ) and @xmath45 = 696 @xmath2 4.6 ( sys ) @xmath2 1.4 ( stat ) kev , after background subtraction . utilizing these parameters
, we derived the most probable value for nuclear recoils as 4.4 @xmath2 0.007 ( sys ) @xmath2 0.002 ( stat ) kev electronic equivalent energy , which corresponds to a nuclear recoil energy of 17.5 @xmath2 0.12 ( sys ) @xmath2 0.035 ( stat ) kev .
the average path length in the germanium detector for 17.5 kev nuclear recoils was approximately 0.014 @xmath3 m .
a monte carlo simulation employing a corrected geant4.9.2 package was modified to duplicate the same experimental setup and ambe neutron source .
the lindhard ( @xmath1 = 0.159 ) and barker - mei models were used to apply ionization efficiency to the energy spectrum and were compared to the experimental measurements .
a bin - to - bin shape analysis was performed , and the difference between the measurements and two models were calculated .
we obtained a percent difference that was less than 4% for the collected data and two monte carlo simulations ( see table 1 ) . using the shape analysis
, we calculated the most probable values for visible nuclear recoils ( as shown in table 1 ) , extracted the average visible nuclear recoil energy ( see table 2 ) , and obtained the corresponding nuclear recoil energy from the monte carlo simulation ( see table 2 ) . the extracted average ionization efficiency is shown in fig .
[ fig : efficiency ] .
these values are in agreement with the lindhard and barker - mei models in the energy range of 1 to 100 kev .
therefore , the lindhard model ( with @xmath1 = 0.159 ) and barker - mei model can be used to determine the ionization efficiency in germanium detectors for 1 to 100 kev nuclear recoil energy .
the authors wish to thank iseley marshall and angela a. chiller for their careful reading of this manuscript .
additionally , the authors would like to thank rupak mahapatra for his comments and suggestions .
this work was supported in part by nsf phy-0758120 , doe grant de - fg02 - 10er46709 , the office of research at the university of south dakota and a 2010 research center support by the state of south dakota .
al . _ ( cogent ) , phys .
lett . * 107 * , 141301 ( 2011 ) .
z. ahmed _ et .
( cdms ) , arxiv:1203.1309 ( 2012 ) .
j. lindhard _ et .
. selsk . * 33 * , 1 ( 1963 ) .
d. barker and d .-
mei , astroparticle physics , volume 38 , p. 1
jones and h.w .
kraner phys .
c , * 4 * 125 ( 1971 ) .
barbeau , j.i . collar and o. tench jcap09(2007)009 ; p. barbeau , phd .
thesis , u. chicago ( 2009 ) .
xichao ruan , _ nuclear recoil quenching factor measurment for hpge detector _
( cdex - texono collaboration ) , beijing , china , 24 - 26 march 2011 .
y. messous _ et .
_ astrophys . * 3 * ( 1995 ) 361 - 366 . c. chasman _ et .
* 21 * 1430 ( 1968 ) .
t. shutt _ et .
. lett . * 69 * 3425 ( 1992 ) .
phys . rev . *
143 * 588 ( 1966 ) .
l. baudis _
methods phys .
res . a * 418 * ( 1998 ) 348 - 354 .
e. simon _ et .
methods phys .
a * 507 * ( 2003 ) 643 - 656 .
kenneth s. krane , introductory nuclear physics , 1987 , p. 341
published by john wiley & sons , inc .
a. dragic et .
al . , 2013 j. phys . :
409 012054 doi:10.1088/1742 - 6596/409/1/012054 .
geant4 , s. agostinelli et al .
instrum . meth .
a506 , 250 ( 2003 ) .
princetion gamma tech inc .
sources , qsa global inc . saint - gobain crystals .
r. wordel et .
al . , nim a369(1996)557 .
g. fehrehbacher , r. meckbach and h. g. paretzke , nim a372(1996)239 .
g. fehrehbacher , r. meckbach and h. g. paretzke , nim a372(1996)391 .
e. gete et .
al . , nim a388(1997)212 . k. w. geiger and l. van der zwan , nuclear instruments and methods 131 ( 1975 ) 315 - 321 .
f. de guarrini and r. malaroda , nuclear instruments and methods 92 ( 1971 ) 277 - 284 .
x - ray instruments associates , 8450 central ave . ,
newark ca 94560 , usa .
wavemetrics inc .
, po box 2088 , lake oswego , or 97035 , usa .
rene brun and fons rademakers , nucl .
instr . and meth .
* 389 * , 81 ( 1997 ) ; http://root.cern.ch .
moyal , theory of ionization fluctuations , phil . mag .
46 ( 1955 ) 263 .
h. bethe und j. ashkin in " experimental nuclear physics , ed .
e. segr , j. wiley , new york , 1953 , p. 253 .
berger , j.s .
coursey , m.a .
zucker , and j. chang ( 2005 ) , estar , pstar , and astar : computer programs for calculating stopping - power and range tables for electrons , protons , and helium ions ( version 1.2.3 ) .
[ online ] available : @xmath90 .
national institute of standards and technology , gaithersburg , md .
refer to `` problem 957 '' in the geant4 problem tracking system : @xmath91 .
a. j. koning , s. hilaire and m. c. duijvestijn , aip conf . proc . , 769 , 11541159 ( 2005 ) .
leong ying , princeton gamma - tech instruments ( private communication ) .
refer to @xmath92 .
j. w. marsh , d. j. thomas and m. burke , nuclear instruments and methods in physics research a 366 ( 1995 ) 340 - 348 .
mei , s. r. elliott , a. hime , v. gehman and k. kazkaz , phys .
c77 , 054614 ( 2008 ) . | we used the internal conversion ( @xmath0 transition ) of germanium-72 to indirectly measure the low energy nuclear recoils of germanium . together with a reliable monte carlo package , in which we implement the internal conversion process ,
the data was compared to the lindhard ( @xmath1=0.159 ) and barker - mei models .
a shape analysis indicates that both models agree well with data in the region of interest within 4% .
the most probable value ( mpv ) of the nuclear recoils obtained from the shape analysis is 17.5 @xmath2 0.12 ( sys ) @xmath2 0.035 ( stat ) kev with an average path - length of 0.014 @xmath3 m . , , , and nuclear recoil , ionization efficiency , dark matter detection 95.35.+d , 07.05.tp , 25.40.fq , 29.40.wk |
many people consider the discovery of dark energy to be the main finding of present cosmology .
they are sure that an existence of dark energy has been proved with observations of new , precise , era of cosmology , and it is necessary only to clarify what it adds up . because of this
, new cosmological centers are created and addicted to this main goal .
it seems to me that a new scientific myth has risen in our eye ; it is nice , almost commonly accepted , with global consequences for physics , but it is really based on nothing . what was a base for its rising ?
in 1998 , two teams of astrophysicists reported about dimming remote sn 1a @xcite ; the one can not be explained in the standard cosmological model on a basis of the doppler effect if the universe expands with deceleration .
their conclusion that the universe expands with acceleration since some cosmological time served a base to endenizen dark energy .
but this conclusion is not a single possible one ; if the model does not fit observations , probably , the one may simply be wrong .
if we stay on such the alternative point of view , what should namely be doubt in the standard cosmological model ?
i think that it should be at first its main postulate : a red shift is caused with an expansion of the universe . if this postulate is wrong , then the whole construction of the model will wreck : neither the big bang nor inflation , nor a temp or character of expansion would not be interested . in the model of low - energy quantum gravity by the author @xcite , the alternative redshift mechanism is quantum and local .
i review here a few possibilities to verify its predictions .
in my model @xcite , any massive body must experience a constant deceleration @xmath0 , where @xmath1 is the hubble constant and @xmath2 is the light velocity , of the same order of magnitude as observed for nasa deep - space probes pioneer 10/11 ( the pioneer anomaly ) @xcite .
this effect is an analogue of cosmological redshifts in the model .
their common nature is forehead collisions with gravitons .
if my conjecture about the quantum nature of this acceleration is true then an observed value of the projection of the probe s acceleration on the sunward direction @xmath3 should depend on accelerations of the probe , the earth and the sun relative to the graviton background
. it would be very important to confront the considered model with observations for small distances when pioneer 11 executed its planetary encounters with jupiter and saturn . in this period , _ the projection of anomalous acceleration may change its sign _ @xcite . how to verify the main conjecture of this approach about the quantum gravitational nature of redshifts in a ground - based laser experiment ? if the temperature of the background is @xmath4 , the tiny satellite of main laser line of frequency @xmath5 after passing the delay line will be red - shifted at @xmath6 ev / h and its position will be fixed @xcite .
it will be caused by the fact that on a very small way in the delay line only a small part of photons may collide with gravitons of the background .
the rest of them will have unchanged energies .
the center - of - mass of laser radiation spectrum should be shifted proportionally to a photon path @xmath7 .
then due to the quantum nature of shifting process , the ratio of satellite s intensity to main line s intensity should have the order : @xmath8 where @xmath9 is an average graviton energy .
an instability of a laser of a power @xmath10 should be only @xmath11 if a photon energy is of @xmath12 .
it will be necessary to compare intensities of the red - shifted satellite at the very beginning of the path @xmath7 and after it . given a very low signal - to - noise ratio , one could use a single photon counter to measure the intensities .
when @xmath13 is a quantum output of a cathode of a used photomultiplier , @xmath14 is a frequency of its noise pulses , and @xmath15 is a desired ratio of a signal to noise s standard deviation , then an evaluated time duration @xmath16 of data acquisition would have the order : @xmath17 assuming @xmath18 we would have the estimate : @xmath19 years , that is unacceptable . but given @xmath20 , we get : @xmath21 days , that is acceptable for the experiment of such the potential importance . of course
, one will rather choose a bigger value of @xmath7 by a small laser power forcing a laser beam to whipsaw many times between mirrors in a delay line - it is a challenge for experimentalists .
maybe , it will be more convenient to work with high - energy gamma rays to search for this effect in a manner similar to the famous pound - rebka experiment @xcite .
the luminosity distance in this model is @xcite : @xmath22 where @xmath23 @xmath24 is a redshift .
the theoretical value of relaxation factor @xmath25 has been found in the assumption that in any case of a non - forehead collision of a graviton with a photon , the latter leaves a photon flux detected by a remote observer : @xmath26 .
it is obvious that this assumption should be valid for a soft radiation when a photon deflection angle is big enough and collisions are rare .
it is easy to find a value of the factor @xmath25 in another marginal case - for a very hard radiation .
due to very small ratios of graviton to photon momenta , photon deflection angles will be small , but collisions will be frequent because the cross - section of interaction is a bilinear function of graviton and photon energies in this model .
it means that in this limit case @xmath27 for an arbitrary source spectrum , a value of the factor @xmath25 should be still computed , and it will not be a simple task .
it is clear that @xmath28 and in a general case it should depend on a rest - frame spectrum and on a redshift .
it is important that the hubble diagram is a multivalued function of a redshift : _ for a given @xmath29 @xmath25 may have different values . _
theoretical distance moduli @xmath30 are shown in fig . 1 for @xmath26 ( solid ) ,
@xmath31 ( dot ) and @xmath32 ( dash ) @xcite .
if this model is true , all observations should lie in the = 8.0 cm stripe between lower and upper curves . for fig .
1 , supernova observational data ( circles , 82 points ) are taken from table 5 of @xcite , gamma - ray burst observations are taken from @xcite ( x , 24 points ) and from @xcite ( + , 12 points for @xmath33 ) . as it was recently shown by cuesta et al .
@xcite , the hubble diagram with @xmath31 ( in the language of this paper ) gives the best fit to the full sets of gamma - ray burst observations of @xcite and it takes place in the standard flrw cosmology plus the strong energy condition .
twelve observational points of @xcite belong to the range @xmath33 , and one can see that these points peak up the curve with @xmath32 which corresponds in this model to the case of very hard radiation in the non - expanding universe with a flat space . in a frame of models without expansion ,
any red - shifted source may not be brighter than it is described with this curve . in this model , the galaxy number counts / magnitude relation is @xcite : @xmath34 to compare this function with observations by yasuda et al .
@xcite , let us choose the normalizing factor from the condition : @xmath35 where @xmath36 is the function assuming `` euclidean '' geometry and giving the best fit to observations @xcite , @xmath37 depends on the spectral band ; an upper limit is @xmath38 in this case , we have two free parameters - @xmath39 and @xmath40 - to fit observations , and the latter one is connected with a constant @xmath41 if @xmath42 we have for @xmath43 by @xmath44 ( it is a theoretical estimate of @xmath1 in this model @xcite ) : @xmath45 where @xmath46 is the sun luminosity .
matching values of @xmath39 shows that @xmath47 is the closest to @xmath48 in the range @xmath49 by @xmath50 the ratio @xmath51 is shown in fig . 2 for different values of @xmath43 by this value of @xmath39 .
= 8.0 cm if we compare this figure with figs .
6,10,12 from @xcite , we see that the considered model provides a no - worse fit to observations than the function @xmath48 if the same k - corrections are added for the range @xmath52 that corresponds to @xmath53 _ observations prefer a rising behavior of this ratio up to @xmath54 and the model demonstrates it . _ for quasars , i computed the galaxy number counts / redshift relation @xmath55 with a different ( than for galaxies ) luminosity function @xmath56 @xcite . in fig 3
, there are a couple of = 8.0 cm curves for each case : the left - shifted curve of any couple ( 1 - 5 ) corresponds to the range @xmath57 another one ( 1 - 5 ) corresponds to @xmath58 these ranges are chosen the same as in the paper by croom et al .
@xcite , and you may compare this figure with fig . 3 in @xcite .
we can see that the theoretical distributions reflect only some features of the observed ones but not an entire picture .
perhaps , it is necessary to consider some theoretical model of a quasar activity to get a distribution of `` instantaneous '' luminosities ( a couple of simple examples is considered in @xcite ) .
one can verify the quantum and local redshift mechanism of this model in different ways , but i think that the most cogent one would be the described prompt measurement of a possible length - dependent red shift of radiation spectrum in the laboratory experiment .
a negative result of this experiment would be a very strong support of the standard cosmological model ; a positive one might open the door not only for new cosmology , but for otherwise quantum gravity , too .
references riess , a.g .
aj _ 1998 , _ 116 , _ 1009 .
perlmutter , s. et al .
_ apj _ 1999 , _ 517 , _ 565 .
ivanov , m.a . in the book ``
focus on quantum gravity research '' , ed .
moore , nova science , ny , 2006 , pp .
89 - 120 ; [ hep - th/0506189 ] , [ http://ivanovma.narod.ru/nova04.pdf ] .
anderson , j.d .
_ 1998 , _ 81 , _ 2858 .
anderson , j.d .
d65 , _ 082004 ; [ gr - qc/0104064 v4 ] .
ivanov , m.a .
[ arxiv:0711.0450v2 ] [ physics.gen-ph ] .
ivanov , m.a .
[ gr - qc/0410076 ] .
pound , r. v. ; rebka jr . ,
g. a. _ phys .
_ 1959 , _ 3 _ , 439 ; 1960 , _ 4 _ , 337 .
ivanov , m.a .
[ astro - ph/0606223 ] .
riess , a.g .
_ apj , _ 2004 , _ 607 , _ 665 ; [ astro - ph/0402512 ] .
bloom , j.s . ,
frail , d.a . , and kulkarni , s.r .
_ apj , _ 2003 , _ 594 , _ 674 .
schaefer , b.e .
_ baas _ , 2005 , _ 37 _ , 1418 ; 207 aas meeting , 2006 .
cuesta , h.j.m .
[ astro - ph/0609262 ] .
ivanov , m.a .
[ astro - ph/0606223 ] .
n. yasuda et al . , _ aj _ 2001 , _ 122 , _ 1104 .
croom , s.m .
mnras , 2004 , 349 , 1397 . | in the model of low - energy quantum gravity by the author , the redshift mechanism is quantum and local , and it is not connected with any expansion of the universe .
a few possibilities to verify its predictions are considered here : the specialized ground - based laser experiment ; a deceleration of massive bodies and the pioneer anomaly ; a non - universal character of the hubble diagram for soft and hard radiations ; galaxy / quasar number counts . |
the gravitational stability of a fluid against convective motion has been extensively studied over the past century .
pioneering examinations of the stability of unmagnetized and magnetized compressible fluids have been conducted by @xcite , @xcite , and @xcite .
these papers have had a profound influence on diverse subjects from the dynamics of astrophysical objects to the confinement of plasma in a fusion device . here
we examine the effects of shear in the magnetic field and of an applied shear plasma flow on stability against gravitational interchange .
these effects change the stability properties and our results are important for many applications .
we construct a simple model to study the effects of magnetic shear and shear flow on the stability properties of a magnetized plasma in a gravitational field .
we derive the equations which determine the behavior of this model in the boussinesq limit .
the equations depend on two parameters , the plasma flow alfvn mach number and the entropy gradient .
we conduct a numerical study of the normal modes of instability and summarize the behavior of the unstable modes by a stability diagram .
the stability diagram demonstrates three important characteristics .
first , the entropy gradient must exceed a threshold value for unstable mode growth to occur .
therefore , the shear magnetic field can stabilize a nonzero entropy gradient .
second , as expected , shear flow does act to suppress unstable mode growth when the system is at a substantially unstable point in parameter space .
but , surprisingly , near marginal stability , shear flow actually enhances the growth rates of the instability and also lowers the threshold entropy gradient necessary for instability .
the effect of magnetic shear to stabilize the plasma and increase the threshold entropy gradient is diminished by the addition of shear flow .
the system can extract energy from the shear flow to further drive the system to instability .
third , as the alfvn mach number approaches one , the unstable growth rate is suppressed ; the normal modes of the instability are completely stabilized when the plasma flow exceeds the alfvn speed . here , the unstable region in space where a mode can grow moves faster than any perturbation in the system ;
any disturbance will be swept downstream out of the unstable region , leaving behind a stable plasma .
analytical work helps us to understand the mechanisms responsible for these characteristics .
first , the stability criterion for the case without shear flow is derived demonstrating that a shear magnetic field can support a positive entropy gradient .
next , asymptotic solutions demonstrate analytically that stabilization occurs as the alfvn mach number approaches one , yet the threshold entropy gradient for instability goes to zero in the same limit .
in addition , a complete wkb solution in the limit of a large growth rate demonstrates both the stabilization by flow at large growth rates and destabilization near marginal stability .
then , a bounded straight field case is solved exactly to show that the lowering of the threshold entropy gradient with increased shear flow is a characteristic of plasma flow along the field lines and not dependent on the magnetic shear in the general model .
finally , energy conservation relations are derived and analyzed . in section
[ sec : derive ] , we describe the model under consideration and derive the governing system of equations .
the numerical stability analysis for the general model and the stability diagram are presented in section [ sec : stab ] .
section [ sec : anal ] contains the analytical results illuminating the characteristics of the stability diagram .
we describe an exactly soluble , bounded , straight - field case in section [ sec : str8 ] .
finally , in section [ sec : disc ] , the implications of this work on galactic physics , accretion disk physics , solar physics , and tokamak confinement are discussed .
in this section , we derive the equations for linear perturbations of a vertically stratified atmosphere with shear magnetic field and shear flow in the high-@xmath0 , or boussinesq , limit . we motivate and apply a coordinate transformation which casts the problem in its most natural form . investigating the limit of the most unstable modes , we derive a system of three coupled first - order ordinary differential equations which capture the lowest order behavior of the model .
consider a stationary state for an ideal plasma with mass density @xmath1 and thermal pressure @xmath2 and an embedded horizontal shear magnetic field given by @xmath3 we then impose a shear flow on this plasma given by @xmath4 and include gravitational acceleration in the vertical direction given by @xmath5 .
equilibrium force balance yields @xmath6 figure [ fig : geom ] shows the geometry of this shear magnetic field as well as the imposed shear flow on the system .
the instabilities of this plasma are expected to have a short perpendicular wavelength and a long parallel wavelength ( with respect to the magnetic field ) so as to maximize motion in the vertical direction and minimize field line bending @xcite .
short perpendicular wavelengths , however , are rapidly sheared apart by the perpendicular shear flow .
we would thus like to transform to a coordinate system with two properties : first , that the flow is along the magnetic field lines ; and , second , that field lines are coordinate lines .
the application of sheared coordinate systems to simplify a problem of this nature is well documented .
@xcite employed a coordinate system in which the field lines are coincident with the coordinate lines to describe the rayleigh - taylor instability of a fluid supported by a shear magnetic field ; @xcite used a shearing coordinate system to attack the problem of local gravitational instabilities in a system with shear flow arising from differential rotation .
we transform the shear velocity to a parallel velocity by transforming to a frame moving in @xmath7 @xmath8 where @xmath9 . in this transformed frame
, the velocity becomes @xmath10 _
i.e. _ parallel to @xmath11 .
we construct field line coordinates by transforming the @xmath12 coordinate to @xmath13 the @xmath14 coordinate is left unchanged , @xmath15 .
it is easy to verify that @xmath16 and @xmath17 are constant along field lines ( @xmath18 ) and that @xmath19 .
the surfaces of constant @xmath17 twist from vertical at @xmath20 to almost horizontal as @xmath21 . this geometry is shown in figure [ fig : geom ] . at first , introducing the field line coordinates seems unhelpful since it introduces explicit @xmath22 dependence into the equations .
the problem also has @xmath14 dependence that arises from the variation of @xmath23 , @xmath2 , and @xmath1 .
indeed , one way to tackle this problem is to fourier transform in @xmath12 and @xmath7 and solve for the @xmath14 dependence .
however , the lowest order solution in the twisting coordinate system is a superposition of these fourier modes , or , complementarily , a fourier solution can be constructed by a superposition of these twisting modes @xcite .
we summarize the relationship between these representations in appendix [ app : reps ] .
we consider the solutions in the twisting coordinates to be more physically relevant since they are localized in @xmath22 .
the basic equations of ideal mhd include the momentum equation written in terms of the gradient of total pressure ( thermal and magnetic ) , the magnetic tension force , and the gravitational force , @xmath24 the induction equation in the limit of zero resistivity , @xmath25 the continuity equation , @xmath26 and the adiabatic equation of state , @xmath27 here @xmath28 denotes the lagrangian derivative , @xmath29 represents the plasma velocity , and @xmath30 is the adiabatic index .
these equations must be evolved subject to the constraint that @xmath31 equations ( [ eq : mhdmom])-([eq : mhdstate ] ) , linearized about the equilibrium ( equations ( [ eq : b0])-([eq : equilib ] ) ) , yield : @xmath32 @xmath33 @xmath34 @xmath35 where we have taken all quantities to vary in time as @xmath36 . in a straight field @xcite and
a shear field @xcite without flow , the most unstable perturbations are incompressible to lowest order and have a small horizontal perpendicular wavelength .
such perturbations maximize vertical motion which extracts energy from the gravitational potential energy and minimize horizontal motions which extract no energy . to isolate these motions , we impose the ordering @xmath37 where @xmath38 is the ordering parameter or our problem .
clearly , all perturbed quantities vary as @xmath39 .
it is also convenient to define the vectors , @xmath40 the perturbed plasma velocity and magnetic field are projected along these directions , _
@xmath41 @xmath42 note that the basis vectors @xmath43 , @xmath44 , and @xmath45 are neither unit vectors nor constant in space_e.g .
_ , @xmath46 .
we expand all perturbed quantities in powers of @xmath47 and denote order as a superscript for example , @xmath48 .
the ordered , perturbed quantities and operators are substituted into equations ( [ eq : mhdmomlin])([eq : mhdstatelin ] ) .
equations ( [ eq : mhdindlin])([eq : mhdstatelin ] ) at @xmath49 and the @xmath44 projection of equation ( [ eq : mhdindlin ] ) yield @xmath50 thus , the dominant motion is along the constant @xmath17 surfaces in the @xmath51 direction . equations ( [ eq : mhdindlin])([eq : mhdstatelin ] ) at @xmath52 produce @xmath53 thus , the perpendicular motion is incompressible to lowest order . at @xmath49 ,
the @xmath44 component of equation ( [ eq : mhdmomlin ] ) gives @xmath54 equation ( [ eq : pbal ] ) expresses the fact that , on the time scales of interest , pressure balance is achieved across the convective eddies ( in the @xmath44 direction ) . in a subsidiary expansion , we take the high-@xmath0 , or boussinesq , limit ( @xmath55 ) . thus , equation ( [ eq : pbal ] ) reduces to @xmath56 , and we find from equation ( [ eq : mhdstatelin ] ) that @xmath57 at @xmath52 , the @xmath44 component of equation ( [ eq : mhdmomlin ] ) yields @xmath58 , or taking the high-@xmath0 limit , @xmath59 .
the final stability equations are obtained from the sum and difference of the @xmath60 projections of equations ( [ eq : mhdmomlin ] ) and ( [ eq : mhdindlin ] ) at @xmath61 and from equation ( [ eq : mhdcontlin ] ) using equation ( [ eq : div ] ) to substitute for @xmath62 .
after some tedious but straightforward algebra , we obtain @xmath63 @xmath64 @xmath65 where @xmath66 and @xmath67 are the elssser variables defined by @xmath68 @xmath69 and the entropy is given by @xmath70 we have normalized so that @xmath71 and @xmath72 . equations ( [ eq : a+b])([eq : sb ] ) contain two parameters : the mach number of the plasma flow with respect to the alfvn speed ( @xmath73 ) , @xmath74 and the entropy gradient , @xmath75 where the primes denote differentiation by @xmath14 . with the boundary conditions that @xmath76 , @xmath77 , and @xmath78 as @xmath79 , equations ( [ eq : a+b])([eq : sb ] )
define an eigenvalue problem for @xmath80 .
solution of the stability equations yields @xmath81 .
equations ( [ eq : a+b])([eq : sb ] ) have a simple physical interpretation .
@xmath66 , the alfvn wave going in the negative @xmath7 ( upstream ) direction , travels at the ( normalized ) speed @xmath82 .
the @xmath66 wave is coupled to the @xmath67 wave by magnetic and velocity shear ( the @xmath67 term in equation [ [ eq : a+b ] ] ) .
the @xmath66 wave is driven by gravity via the @xmath83 term in equation ( [ eq : a+b ] ) .
@xmath67 , the alfvn wave going in the positive @xmath7 ( downstream ) direction , travels at speed @xmath84 , is coupled to @xmath66 , and is driven by @xmath83 .
the variable @xmath83 is proportional to the density or entropy perturbation and it is driven by both alfvn waves , as seen in equation ( [ eq : sb ] ) .
in this section , we discuss the unstable eigenvalues ( @xmath85 ) and eigenfunctions of equations ( [ eq : a+b])([eq : sb ] ) .
we have not examined the stable part of the spectrum in detail although numerical results indicate a continuum along the imaginary @xmath80 axis .
two properties of equations ( [ eq : a+b])([eq : sb ] ) show that it is sufficient to examine stability in the region @xmath86 .
first , note that @xmath87 since we can map equations ( [ eq : a+b])([eq : sb ] ) onto themselves by the changes @xmath88 , @xmath89 , @xmath90 , @xmath91 , and @xmath92 .
second , note that the three asymptotic solutions as @xmath93 are : @xmath94 if @xmath95 , there are no acceptable ( decaying ) asymptotic solutions as @xmath96 for @xmath97 .
thus , @xmath95 has no unstable eigenmodes .
physically , this is because all solutions , even the upstream propagating alfvn wave @xmath66 , are swept downstream . without flow ,
mhd stability equations are self - adjoint @xcite and @xmath98 is real . with flow ,
no such property is known and @xmath98 can be complex . however , in all our solution methods , @xmath98 has been found to be real for this problem , although we have not been able to prove that this is rigorously true .
the discrete positive real values for @xmath80 correspond to unstable growing modes , and the continuum along the imaginary @xmath80 axis represents traveling alfvn waves . in the following subsection , we present the numerical solution of equations ( [ eq : a+b])([eq : sb ] ) .
various analytical limits that illuminate the numerical solutions are treated in subsection [ sec : anal ] . an exactly soluble model with a straight magnetic field that demonstrates qualitatively similar behavior
is presented in subsection [ sec : str8 ] .
finally , in subsection [ sec : energy ] , energy constraints on the instability are discussed .
we looked for normal mode growth in the system defined by equations ( [ eq : a+b])([eq : sb ] ) using three different numerical methods .
we directly solved for the eigenvalues @xmath80 of this system by matrix solution of the corresponding finite difference equations using the commercial numerical routine package lapack .
we also found the eigenvalues of the equations to high precision using a 1-d shooting code in @xmath7 driven by an adaptive step - size , fourth order runge - kutta method with fifth order correction ( rk45 ) .
finally , for equations ( [ eq : a+b])([eq : sb ] ) with @xmath80 replaced by @xmath99 , an initial - value code employing barton s method @xcite for second order accuracy in time was written to determine the fastest growing mode for any choice of parameters .
results from all three codes were consistent . a stability diagram of unstable normal - mode growth rates and stable regions over the parameter space defined by @xmath100 and @xmath101 is presented in figure [ fig1 ] .
as we have already noted , no growing mode exists for @xmath95 .
as well , it is obvious that for the non - positive values of the entropy gradient , @xmath102 , there can be no unstable mode growth since the atmosphere is stably or neutrally stratified ; this is demonstrated by the energy arguments presented in section [ sec : energy ] .
hence , figure [ fig1 ] need only cover the region of @xmath103 parameter space defined by @xmath104 and @xmath105 to include all possible unstable mode growth .
several features of figure [ fig1 ] are important to emphasize .
first , for a system without plasma flow ( along the line @xmath106 ) , we see that the entropy gradient must exceed a threshold value , @xmath107 , in order to become unstable when the fluid is supported by a shear magnetic field .
second , the qualitative effect of increasing the plasma flow ( increasing @xmath100 ) on the instability growth rate depends on both the growth rate and the plasma flow . away from marginal stability
( @xmath108 ) , an increase of the plasma flow equivalent to moving along a horizontal line to the right on the stability diagram decreases the instability growth rate as one may expect . here , the growing perturbation is sheared out horizontally so that less gravitational potential energy is extracted by motions along the constant @xmath17 surface , reducing the instability growth rate . but near marginal stability ( @xmath109 ) , an increase in the shear flow effects the instability growth rate differently depending on the relative orders of the growth rate , @xmath80 , and of one minus the alfvn mach number , @xmath82 .
the diagonal dotted line in figure [ fig1 ] denotes @xmath110 . in region ( ii ) of the diagram , @xmath111 and flow suppresses the instability .
but , in region ( i ) , where @xmath112 , an increase in the flow actually enhances the instability growth .
this unexpected result can be explained with some physical insight .
the point where the instability can grow is localized at @xmath20 in our transformed coordinates ; this is where the constant @xmath17 surface is vertical and motions along that surface can extract the most gravitational potential energy with which to drive the instability . when a shear plasma flow is introduced into the system , this is manifested in our transformed system by a plasma flow along the field lines in the @xmath113 direction .
this causes the alfvn modes in the @xmath113 and @xmath114 directions to propagate at different speeds in our transformed system .
the counter - propagating mode ( @xmath66 in our model ) is slowed down , spending more time in the region around @xmath20 where unstable mode growth occurs . hence , the instability is enhanced by a shear flow in the plasma . the final point to be gleaned from figure [ fig1 ] concerns the behavior as @xmath115 , or as the alfvn mach number approaches one . in this region , every contour corresponding to a finite growth rate asymptotes to @xmath116 ; hence , the instability is stabilized as the alfvn mach number approaches one .
the stability diagram , figure [ fig1 ] , prominently displays the three main characteristics discussed in section [ sec : stab ] : a threshold entropy gradient , @xmath107 , necessary for instability in the system without plasma flow ; for increasing plasma flow , a reduction of the unstable growth away from marginal stability , but an enhancement of that unstable mode near marginal stability including a decrease in the threshold entropy gradient necessary for instability ; and the stabilization of unstable normal modes as the alfvn mach number approaches one .
each of these characteristics is relevant in a different region of the @xmath103 parameter space of the diagram . by examining the model in each of these regions of parameter space
, we can confirm and explain our results analytically .
first , we examine the criterion for instability in the absence of plasma flow in section [ sec : stabcrit ] ; this corresponds to the left vertical axis of the stability diagram where @xmath106 .
next , in section [ sec : m1lim ] , we conduct an asymptotic analysis in the @xmath115 limit region ( ii ) of the stability diagram to show that the plasma is indeed stabilized as the alfvnic mach number approaches one . in section [ sec : g0 ] , we investigate the reduction of the threshold for instability with plasma flow ; this covers the lower , right - hand side of region ( i ) of the stability diagram . a wkb analysis for a large instability growth rate , presented in section [ sec : wkb ] ,
yields the behavior of the system in the central and upper portion of the stability diagram ; the suppression of the growth rate by flow in region ( ii ) and its enhancement in region ( i ) are verified by the eigenvalue condition @xmath117 obtained in this analysis . here
we obtain the stability criterion for a magnetized fluid supported by a shear magnetic field in the boussinesq limit with no shear flow . by using the substitution @xmath118 , the equations without plasma flow ( @xmath106 ) can be simplified to a sturm - liouville equation of the form @xmath119 - \left [ \gamma^2 ( 1 + z^2 ) - s'_0 \right ] v = 0 \label{eq : vm0}\ ] ] over the interval @xmath120 .
the boundary conditions on this system necessitate that @xmath121 as @xmath122 . from sturm
s first comparison theorem @xcite , we know that , as the eigenvalue @xmath98 is increased , the solution will oscillate less rapidly with zeros of the function @xmath123 leaving the interval @xmath124 at the boundaries .
thus , if the solution with @xmath125 oscillates , we can increase @xmath98 until the boundary conditions are satisfied , so there will be an unstable solution .
note also that the fastest growing mode has no zeros in the interval and must be even in @xmath7 .
let w(z ) satisfy equation ( [ eq : vm0 ] ) with @xmath125 . substituting a series solution of the form @xmath126 in equation ( [ eq : vm0 ] ) ( with @xmath125 )
, we obtain the recurrence relation , @xmath127 with @xmath128 when @xmath129 , solution ( [ eq : vsoln ] ) oscillates for @xmath130 , and clearly an unstable solution exists .
let us therefore consider stability when @xmath131 .
we take the positive sign in equation ( [ eq : alpha ] ) ; then @xmath132 is integrable for @xmath130 .
we note that all @xmath133 are positive if @xmath134 is positive .
we choose @xmath135 such that @xmath136 .
the series solution for @xmath137 given by equation ( [ eq : vsoln ] ) is non - differentiable at @xmath138 ( as we see below ) .
thus , we can not use the solution from equation ( [ eq : vsoln ] ) over the whole interval and must restrict its use to @xmath139 .
let us suppose ( for contradiction ) that there exists at least one unstable solution of equation ( [ eq : vm0 ] ) .
further , let @xmath140 be the most unstable solution as noted above , @xmath140 must be even in @xmath7 and have no zeros in the interval @xmath124 . we therefore choose @xmath141 everywhere .
it is straight forward to show that @xmath142 where the limit @xmath143 is infinitesimally above @xmath138 .
since the integral and @xmath144 in equation ( [ eq : zero ] ) are positive , we have stability , @xmath145 ( a contradiction ) , if @xmath146 .
since every term in the series equation ( [ eq : vsoln ] ) is a monotonically decreasing function of @xmath7 we expect @xmath147 , but since the limit is nonuniform we take a more careful approach .
we determine the sign of the limit @xmath148 from an examination of the convergence of the series .
it is straightforward to show that @xmath149 as @xmath150 with @xmath151 a positive constant .
thus , the series for @xmath137 ( equation [ [ eq : vsoln ] ] ) converges for @xmath152 .
however , the series for @xmath153 converges for @xmath139 but not for @xmath138 .
let us write for @xmath154 @xmath155 where @xmath156 is a positive constant and @xmath157 is a large number in the range @xmath158 . using @xmath159
, we obtain @xmath160 thus , the limit of @xmath153 as @xmath161 is @xmath162 , _
i.e. _ negative . from equation ( [ eq : zero ] )
we conclude that @xmath145 and there are no unstable modes for @xmath131 .
thus , the necessary and sufficient condition for instability is @xmath129 .
this criterion can also be written @xmath163 this confirms the result in figure [ fig1]that a threshold value of the entropy gradient , given by equation ( [ eq : bouscriterion ] ) , must be exceeded in order to cause instability when the fluid is supported by a shear magnetic field . clearly , without magnetic shear ( @xmath164 ) , the usual criterion , @xmath165 , holds and the motion is the simple interchange of field lines . with magnetic shear , the field lines _ must _ be bent since the interchange of field lines is impossible with finite displacements thus magnetic shear is stabilizing .
an asymptotic , boundary layer analysis can be carried out in the limit that @xmath166 .
this asymptotic expansion is described in appendix [ app : mto1 ] .
the eigenvalue condition derived for this limit in that appendix is @xmath167 this relation explains the upturn towards infinity of the constant growth rate contours in region ( ii ) of figure [ fig1 ] as @xmath115 . we demonstrate that the threshold entropy gradient for instability ( the limit that @xmath168 ) decreases as the plasma flow velocity is increased and that the threshold value of @xmath101 approaches zero linearly as @xmath166 .
letting @xmath169 , we can redefine the following variables in terms of @xmath170 : @xmath171 , @xmath172 , @xmath173 and @xmath174 . for @xmath166
, we can drop terms of order @xmath175 and cancel @xmath170 from each remaining term to yield the simplified set of equations @xmath176 these equations are now independent of @xmath170 , or , equivalently , are independent of @xmath100 .
the equations will hold true for constant values of @xmath177 and @xmath178 . in this case , if we have a negligibly small value @xmath179 , we obtain the corresponding threshold value of @xmath180 , where @xmath178 is a constant .
thus , the threshold entropy gradient for stability must linearly approach zero as @xmath166 , as seen in the lower right - hand corner of figure [ fig1 ] .
a complete solution of the model for a large growth rate @xmath80 can be constructed if we assume an ordering , for a small parameter @xmath170 , such that @xmath181 , @xmath182 , @xmath183 , and @xmath184 . in this case
, we can neglect the second term on the right hand side of both equations ( [ eq : a+b ] ) and ( [ eq : a - b ] ) .
converting back from elssser variables to @xmath185 and @xmath186 notation , combining the three equations into a single second order equation , and neglecting the term @xmath187 ( which can be shown to be small ) , yields @xmath188 \delta v_\perp = 0 .
\label{eq : wkb1}\ ] ] changing variables with an integrating factor to @xmath189 , we obtain @xmath190 v = 0 .
\label{eq : wkb2}\ ] ] we assume a wkb solution of the form @xmath191 and find @xmath192^{1/2 } dz ' } \right ) .
\label{eq : wkbsols}\ ] ] the turning points are at @xmath193 where @xmath194 for @xmath195 , the wkb solutions are decaying exponentials ; in the region @xmath196 , the wkb solution is an oscillatory function . in the usual way @xcite , we obtain the bohr - sommerfeld quantization condition @xmath197^{1/2 } dz = 2 n \pi ( 1-m^2 ) .
\label{eq : bscond}\ ] ] the growth rate of the @xmath198th mode is then given , for small @xmath199 , by @xmath200 ( 1-m^2 ) .
\label{eq : gammawkb}\ ] ] for a large growth rate @xmath201 and the lowest , nontrivial eigenmode @xmath202 , we can solve this eigenvalue condition for @xmath101 to obtain @xmath203 this condition agrees with the behavior of the constant growth rate contours in region ( ii ) of figure [ fig1 ] . although we do not expect the eigenvalue condition in the wkb approximation to be precise in the limit of small growth rate , @xmath204 , we do find to lowest order the qualitatively correct form , @xmath205 that the contours in region ( i ) decrease like @xmath206 .
the general model defined by this paper has the characteristic that the shear in the magnetic field localizes the region of instability around where the constant @xmath17 surface is vertical ( this corresponds to @xmath20 in our transformed coordinate system ) . with this characteristic as our guide ,
an exactly soluble , simplified model can be constructed which demonstrates the same qualitative behavior displayed in figure [ fig1 ] .
we construct a case with a straight magnetic field and a plasma flow along the field lines which has boundaries at @xmath207 .
since the explicit @xmath7 dependence drops out of the equations in the straight - field limit ( @xmath208 ) , we fourier transform in @xmath7 to obtain the algebraic dispersion relation @xmath209 = 0 .
\label{eq : simpdisp}\ ] ] if we remove the plasma flow from the system by setting @xmath106 , this dispersion relation agrees with the results of @xcite for a straight - field without flow in the boussinesq limit .
we solve for the three solutions for @xmath210 from equation ( [ eq : simpdisp ] ) and find the eigenvectors corresponding to each @xmath210 .
constructing general solutions for @xmath185 , @xmath186 , and @xmath83 from these eigenvectors , we find the eigenvalue condition @xmath81 that must hold in order to satisfy the three necessary boundary conditions on the system .
the boundary conditions we apply are @xmath211 at @xmath212 and the upstream boundary condition @xmath213 at @xmath214 .
the eigenvalue condition thus obtained is @xmath215 ( 1-m^2 ) \label{eq : simpeig}\ ] ] for the @xmath198th order unstable mode where @xmath216 .
a plot of constant @xmath80 contours is displayed in figure [ fig : simp ] .
note that the qualitative behavior pointed out in the text in section [ sec : stab ] is demonstrated by this simplified model .
therefore , the magnetic shear is not responsible for the unexpected decrease in the stability threshold with shear flow ; only a localization of the instability is necessary to demonstrate this characteristic .
a quantitative comparison of figures [ fig1 ] and [ fig : simp ] shows that , to yield equivalent growth rates , a much larger entropy gradient must be supplied in the unbounded case with magnetic shear than in the bounded , straight - field model .
we can understand the difference as follows .
the energy required to bend the magnetic field lines slows the growth of the instability . in the bounded case
, this energy is needed to bend the field line only within the bounded domain .
but , in the unbounded model , the bending of the field lines occurs over a larger extent in @xmath7 , thus requiring more energy and so more effectively suppressing the instability . in the standard way ,
equations ( [ eq : a+b])([eq : sb ] ) can be combined , replacing @xmath80 with the time derivative @xmath217 , to obtain an energy integral for the model .
converting back to @xmath186 and @xmath185 using equations ( [ eq : a+def ] ) and ( [ eq : a - def ] ) , we find @xmath218 dz \right\ } = m \int^\infty_{- \infty } \frac { s \delta b_\perp } { 1 + z^2 } dz .
\label{eq : bousenergy}\ ] ] adopting the terminology of @xcite , we define the integral on the left - hand side of equation ( [ eq : bousenergy ] ) as the wave energy of the perturbation .
note that , in the absence of flow ( @xmath106 ) , the wave energy is constant in time .
equation ( [ eq : bousenergy ] ) supplies a limit on the value of @xmath101 necessary for instability : since the @xmath219 and @xmath220 terms are both positive definite , an instability can only develop for @xmath221 . in this case , gravitational potential energy from the @xmath222 term can be harnessed to drive the kinetic energy and field line bending of the instability . by the same method as above , we find that the wave energy integral for the straight - field model bounded at @xmath223 , in section [ sec : str8 ] , has the form @xmath224 dz \right\ } = \frac{m}{2 } \left\ { \delta b_\perp^2(-l ) - \delta b_\perp^2(l ) + \frac{s^2(l)}{s_0 ' } \right\ } .
\label{eq : straightenergy}\ ] ] without flow , again , we find that the necessary condition for instability to develop is @xmath221 and that the wave energy integral is constant in time . in the presence of flow ,
we interpret the terms on the right - hand side of equation ( [ eq : straightenergy ] ) as follows : @xmath225 represents the flow of magnetic energy into the region , @xmath226 represents the flow of magnetic energy out of the region , and @xmath227 represents the flow of gravitational potential energy out of the the region .
to study the effect of magnetic shear and shear flow on local gravitationally induced instabilities , we have constructed a simple model in the boussinesq limit of ideal mhd .
numerical solutions to this model yield a stability diagram of the @xmath103 parameter space .
this stability diagram has three important characteristics .
first , there exists a threshold entropy gradient for unstable mode growth , demonstrating that magnetic shear is a stabilizing influence .
second , flow serves to suppress mode growth in a substantially unstable regime , but near marginal stability it lessens the stabilizing effect of magnetic shear , enhancing unstable mode growth rates and lowering the threshold entropy gradient necessary for instability .
third , normal modes of instability are stabilized completely as the alfvn mach number approaches one because the disturbance is swept downstream out of the unstable region .
analytical work corroborates these characteristics in the different regions of @xmath103 parameter space . in a shear magnetic field without flow , the characteristic shape of the unstable mode
is such that the field lines remain on the constant @xmath17 surface shown in figure [ fig : geom ] ; hence , in the lab frame , the field lines must twist as they fall under gravity to remain on this surface .
this occurs because the perturbed field line at any vertical height @xmath14 must align with the direction of the unperturbed field at that height to facilitate interchange .
unlike the ordinary interchange of straight field lines , if magnetic shear is present , the field line _ must _ be bent to allow interchange ; this is the root of the stabilizing influence of the shear magnetic field .
energy extracted from gravitational potential energy as the field line falls must supply both the kinetic energy of the moving plasma , which is frozen to the field line , and the energy required to bend the field line .
unstable motions are localized about the point where the constant @xmath17 surface is vertical ( @xmath138 in figure [ fig : geom ] ) , because this point is where motions on that surface extract the most gravitational potential energy while minimizing field line bending . when plasma shear flow is included in the problem , this unstable region moves in the lab frame with the flow velocity @xmath228 in the @xmath229 direction .
therefore , this characteristic twisting geometry propagates along @xmath7 at a speed @xmath230 determined by the local flow speed @xmath144 , flow shear length @xmath231 , and magnetic field shear length @xmath232 .
typical eigenfunctions for an entropy gradient of @xmath233 and alfvn mach numbers @xmath106 and @xmath234 are displayed in figure [ fig : eigfunc ] . the modifications of the gravitational stability of a magnetized plasma due to the presence of magnetic shear and shear flow impact several diverse subjects .
in solar physics , one of the key processes necessary for the success of the interface dynamo @xcite is the storage of an intense toroidal field in the solar tachocline @xcite until an instability causes an isolated flux tube to rise into the base of the convection zone @xcite .
shear flow is present due to the large differential rotation of the sun in the tachocline region @xcite , and some magnetic shear probably exists here as well ; the impact of shear flow and magnetic shear may alter gravitational stability within this region .
there are two regions in galactic physics where the conditions in our model may apply : near the center of the galaxy , where there likely exists a region of magnetic shear in which the large - scale magnetic field changes from primarily azimuthal to primarily vertical @xcite and shear flow is present through differential rotation , and in the disk of the galaxy , where both magnetic shear and shear flow may be present in the vertical direction away from the mid - plane . in examining the mechanisms for turbulence in accretion disks , @xcite pointed out that a previously discovered @xcite but unappreciated linear mhd instability , driven by differential rotation coupled with magnetic tension , would occur if the accretion disk were threaded by a weak magnetic field .
but the balbus - hawley instability bends the magnetic field lines extensively ; thus , it is stabilized by magnetic tension for large magnetic field strengths @xcite .
although the twisting interchange instability studied here depends not on rotation but on an entropy gradient , it has a characteristic geometry that minimizes field line bending and so may be important in regions of large field strength . as is well known within the fusion community @xcite ,
employing a sheared toroidal magnetic field in a tokamak can help to stabilize , or at least suppress , ballooning instabilities ( buoyant interchange instabilities driven by pressure and curvature forces ) .
the idea of employing shear flow to further stabilize ballooning instabilities has gained much attention in the past decade @xcite .
our work demonstrates that , near marginal stability , shear flow actually lessens the stabilizing effect of magnetic shear , lowering the threshold entropy gradient required for instability and enhancing the unstable mode growth rate .
full stabilization of the plasma will occur only if the alfvn mach number of the plasma flow ( as defined in our transformed coordinates ) exceeds one .
the local nature of the instability examined in this work means that our treatment may apply locally in more complicated environments .
the instability may behave as a traveling `` wave packet '' which moves with the intersection of shear magnetic field lines ( where the constant @xmath17 surface is vertical ) ; the disturbance may move in and out of unstable regions , with the perturbation growing where conditions are unstable and decaying in stable regions .
we have extended the model to include compressibility . in this case , the behavior is governed by a system of five coupled , first - order ordinary differential equations .
five parameters are necessary to describe the system : the alfvn mach number , the plasma @xmath0 , the density gradient , the pressure gradient , and the magnetic field gradient . over some portions of this five - dimensional parameter space
, the growth rate eigenvalue @xmath80 does indeed become complex . producing a simple answer from this more complicated model is quite difficult .
our current research is addressing this difficulty .
but , the magnetic - buoyancy instability @xcite , which depends on compressibility , can not be investigated without employing this more detailed treatment .
we will then be able to relate our work quantitatively to applications such as the stability of magnetic fields in the solar tachocline .
one final possible extension of this research is an investigation of the nonlinear behavior using an initial - value code .
the research was performed under appointment of greg howes to the fusion energy sciences fellowship program administered by the oak ridge institute for science and education under a contract between the u.s .
department of energy and the oak ridge associated universities .
it was supported in part by the national science foundation under grant no .
phy94 - 07194 .
we are grateful to b. albright , a. cumming , b. dorland , e. quataert , s. tobias , j. toomre , and e. zweibel for useful discussions .
in the sheared magnetic field , the local unstable modes can be written in two ways @xcite ; we call them twisting and fourier modes . for the reader
s convenience , we summarize the essence of the argument here . in the twisted field line
coordinates , we obtain the twisted mode where a perturbed quantity @xmath235 is given by @xmath236 with @xmath237 localized in @xmath7 and varying weakly in @xmath14 compared to @xmath238 .
but since the origin in @xmath7 is arbitrary , we can also write @xmath239 thus , there are an infinite number of twisting modes , each with a different origin of the twist .
we can construct a mode that does not depend on @xmath7a fourier mode by integrating equation ( [ eq : t2 ] ) over @xmath199 @xmath240 where @xmath241 thus , the fourier modes and the twisting modes are related by a fourier transform .
clearly , the fourier mode can be made by `` adding '' twisting modes together ( equation [ [ eq : t3 ] ] ) , or vice versa , using the fourier inversion theorem on equation ( [ eq : t4 ] ) ; see @xcite for pictures of this superposition .
the fourier modes are narrowly localized in @xmath14typically @xmath242 where @xmath243 is the @xmath14 width of the fourier mode and @xmath244 is the @xmath7 width of the twisting mode .
note this @xmath14 localization of the fourier mode is narrow compared to the @xmath14 variation of the twisting mode . in this paper
, we have taken the twisting mode representation for two reasons : first , the role of the flow , we believe , is more intuitive in this picture ; and , second , the twisting modes are finite in @xmath7 extent and therefore represent more easily the evolution of an initial value problem .
to demonstrate stabilization as the alfvn mach number approaches one , as seen of region ( ii ) of figure [ fig1 ] , we perform an asymptotic analysis of our model in the limit @xmath115 . for a small dimensionless parameter @xmath170
, we quantify the order of this limit as @xmath245 .
we expect the instability growth rate to be @xmath246 . identifying the terms in the dimensionless system of equations ( [ eq : a+b])([eq : sb ] ) for reference
, we have @xmath247 @xmath248 @xmath249 the boundary conditions demand @xmath250 as @xmath251 .
asymptotic solutions in the @xmath166 limit can be found in the four regions along @xmath7 displayed in figure [ fig : asymp ] .
below we find the solutions for each of these regions and , by matching the solutions between these regions , we obtain an eigenvalue condition on the growth rate demonstrating stabilization as the alfvn mach number approaches one .
first , we obtain two reductions of equations ( [ eq : a+])([eq : s ] ) ; one over a boundary layer where @xmath252 , and the other over an outer region where @xmath253 . then , we present the solutions in each of the four regions in figure [ fig : asymp ] . as the alfvn mach number approaches one , regions ( 1 ) , ( 2 ) , and ( 3 ) of figure [ fig : asymp ] behave like a boundary layer region : we expect derivatives to be large and thus take @xmath254 .
we treat @xmath255 over these three regions .
balance of the dominant terms ( 9 ) and ( 11 ) in equation ( [ eq : s ] ) shows that @xmath256 . in turn
, this demands that terms ( 5 ) and ( 8) must balance in equation ( [ eq : a- ] ) , yielding the ordering @xmath257 . adopting the specified ordering allows us to drop terms ( 6 ) , ( 7 ) , ( 10 ) , and ( 12 ) from equations ( [ eq : a+])([eq : s ] ) ; term ( 3 ) , although one order in @xmath170 smaller than the other terms in equation ( [ eq : a+ ] ) , will contribute in the regions for @xmath258 , so we retain it in order to be certain that our first order correction in region ( 3 ) is valid .
the remaining equations can be combined to a single third - order equation in @xmath67 and simplified by the substitution @xmath259 to obtain @xmath260 in region ( 4 ) , we find the smoothly varying outer solution over which @xmath261 .
we expect that @xmath262 here . by insisting that terms ( 9 ) and ( 10 ) balance with term ( 11 ) in equation ( [ eq : s ] ) , we find that @xmath263 .
similarly , by balancing terms ( 5 ) and ( 6 ) with term ( 8) in equation ( [ eq : a- ] ) , we obtain @xmath264 .
this ordering allows us to drop terms ( 1 ) and ( 7 ) in equations ( [ eq : a+])([eq : s ] ) .
approximating @xmath265 , substituting @xmath266 , and combining the equations , we obtain the second - order equation @xmath267 \frac{d b_-}{dz } \nonumber \\ & + \left [ \gamma^2 z^2 + ( 1 + 3 m)\gamma z + m(1+m ) - s_0 ' \right ] b_- = 0 .
\label{eq : a2}\end{aligned}\ ] ] to find the solution in region ( 1 ) , we assume an eikonal solution for equation ( [ eq : a1 ] ) . neglecting the trivial constant solution , we find two independent solutions of the form @xmath268^{1/2 } } \exp \left ( \frac{\gamma } { 2(1-m ) } \left[z + ( 1+z^2)^{1/2}\right]\right ) \label{eq : a-1a}\ ] ] @xmath269 + z\left[z-(1+z^2)^{1/2}\right ] \right ) .
\label{eq : a-2a}\ ] ] to get the behavior for @xmath253 , we can expand @xmath270 .
for region ( 1 ) , we note that @xmath271 and , retaining only the dominant terms , we obtain solutions of the form @xmath272 @xmath273 the boundary conditions impose that @xmath77 as @xmath274 , so our solution in region ( 1 ) must be entirely of the form of equation ( [ eq : a-2reg1 ] ) , a growing solution in the @xmath275 direction . to determine the behavior of equations ( [ eq : a-1a ] ) and ( [ eq : a-2a ] ) in the overlap with region ( 2 ) , we take the limit @xmath276 and approximate @xmath277 to get the two solutions @xmath278 \right ) \label{eq : a-1reg1small}\ ] ] @xmath279 \right ) .
\label{eq : a-2reg1small}\ ] ] one of these solutions must smoothly match onto the solution for region ( 2 ) .
region ( 1 ) s solution will be valid as we move in the @xmath275 direction until the eikonal approximation , @xmath280 , breaks down .
the failure of this condition occurs in region ( 2 ) of figure [ fig : asymp ] where @xmath281 . for region ( 2 )
, we expand equation ( [ eq : a1 ] ) about @xmath138 . for @xmath282 , @xmath283 and @xmath284 where @xmath285 , so we can drop the last term of equation ( [ eq : a1 ] ) . letting @xmath286 and using the integrating factor @xmath287 to simplify the result
, we obtain the equation @xmath288
\label{eq : b - ho}\ ] ] neglecting the central term in the coefficient of @xmath289 because it is an order @xmath170 smaller than the other terms , we can cast equation ( [ eq : b - ho ] ) in the form of hermite s equation for which the solutions are well known . therefore , for the @xmath290 hermite polynomial , the solution in region ( 2 ) is @xmath291 and the eigenvalue condition on the growth rate imposed by hermite s equation is @xmath292 for the @xmath198th hermite polynomial .
thus , we find that the solution in region ( 1 ) given by equation ( [ eq : a-2reg1small ] ) matches smoothly onto our solution in region ( 2 ) . in region ( 3 )
, we once again assume an eikonal solution for equation ( [ eq : a1 ] ) and find the two solutions given by equations ( [ eq : a-1a ] ) and ( [ eq : a-2a ] ) . to match with region ( 2 ) , find the @xmath276 limit of these equations , yielding once more equations ( [ eq : a-1reg1small ] ) and ( [ eq : a-2reg1small ] ) ; we observe that equation ( [ eq : a-2reg1small ] ) for the small @xmath7 limit of region ( 3 ) matches solution ( [ eq : a - reg2 ] ) for region ( 2 ) . in the @xmath293 limit of equations ( [ eq :
a-1a ] ) and ( [ eq : a-2a ] ) for region ( 3 ) , we obtain the solutions ( for @xmath139 ) @xmath294 @xmath295 to continue our asymptotic solution , we must smoothly match one of these region ( 3 ) solutions to the solution for region ( 4 ) in the overlap around @xmath296 . in region
( 4 ) , we assume eikonal solutions for equation ( [ eq : a2 ] ) in the limit @xmath154 .
the two solutions found are @xmath297 @xmath298 hence , we can match the solution given by equation ( [ eq : a-2reg3 ] ) in region ( 3 ) with the solution given by equation ( [ eq : a-2reg4 ] ) in region ( 4 ) if @xmath299 but , this is identical to the lowest order of the eigenvalue condition , equation ( [ eq : eigcond ] ) . to complete our asymptotic solution , we must find that a solution to equation ( [ eq : a2 ] ) in the limit @xmath130 which satisfies the boundary condition that @xmath77 as @xmath130 . in this limit ,
the two solutions take the form @xmath300 @xmath301 both of these solutions satisfy the boundary condition as @xmath302 . now that we have seen that it is possible to construct a complete asymptotic solution in the limit @xmath166 ,
let us examine this solution more closely .
beginning in region ( 1 ) at the left of figure [ fig : asymp ] , the boundary conditions demand that the solution must be solely of the form of equation ( [ eq : a-2reg1 ] ) .
but , as behavior in regions ( 1 ) , ( 2 ) , and ( 3 ) is governed by equation ( [ eq : a1 ] ) , the eikonal approximation must break down in region ( 2 ) in order for the solution in region ( 1 ) to convert to the solution given by equation ( [ eq : a-2reg3 ] ) in region ( 3 ) so that smooth matching may be accomplished with solution ( [ eq : a-2reg4 ] ) in region ( 4 ) .
the failure of the eikonal approximation around @xmath138 yields a reduction of equation ( [ eq : a1 ] ) to the hermite - type equation ( [ eq : b - ho ] ) .
the requirement that a solution to this equation exist imposes the eigenvalue condition , equation ( [ eq : eigcond ] ) .
this single condition can also be used to smoothly match solution ( [ eq : a-2reg3 ] ) in region ( 3 ) to solution ( [ eq : a-2reg4 ] ) in region ( 4 ) .
finally , region ( 4 ) is governed by equation ( [ eq : a2 ] ) .
in the limit @xmath154 , this equation yields a matching solution in the overlap with region ( 3 ) ; and , in the limit @xmath130 , it provides two solutions which both satisfy the boundary conditions as @xmath303 .
therefore , the single condition necessary to find a smooth solution which satisfies the boundary conditions is the eigenvalue condition , equation ( [ eq : eigcond ] ) .
to lowest order , this condition can be written in a more recognizable form in the limit @xmath115 as @xmath304 thus , in region ( ii ) of figure [ fig1 ] , where @xmath305 , the behavior is clearly demonstrated that stabilization occurs as the alfvn mach number approaches one .
adam , j. a. 1978 , j. plasma phys .
, 19 , 77 .
balbus , s. a. and hawley , j. f. 1991 , , 376 , 214 .
barnes , g. , macgregor , k. b. , and charbonneau , p. 1998
, , 498 , l169 .
bender , c. m. and orszag , s. a. 1978 , advanced mathematical methods for scientists and engineers , new york : mcgraw - hill .
bernstein , i. b. , frieman , e. a. , kruskal , m. d. , and kulsrud , r. m. 1958 , proc .
a244 , 17 .
blaes , o. m. and balbus , s. a. 1994 , , 421 , 163 .
caligari , p. , moreno - insertis , f. and schussler , m. 1995 , , 441 , 886 .
cally , p. s. 1983 , geophys .
fluid dynamics , 23 , 43 .
cally , p. s. and adam , j. a. 1983 , geophys .
fluid dynamics , 23 , 57 .
cattaneo , f. and hughes , d.w .
1988 , j. fluid mech . , 196 , 323 .
cattaneo , f. , chiueh , t. , and hughes , d.w .
1990 , j. fluid mech .
, 219 , 1 .
centrella , j. and wilson , j. r. 1984 , , 54 , 229 .
chandrasekhar , s. 1960 , proc .
, 46 , 253 .
chandrasekhar , s. 1961 , hydrodynamic and hydromagnetic stability , oxford : oxford univ . press .
charbonneau , p. and macgregor , k. b. 1997 , , 473 , l59 .
cowley , s. c. , kulsrud , r.m . and sudan , r. 1991 , phys .
fluids b , 3 , 2767 .
gilman , p. a. and fox p. a. 1997 , , 484 , 439 .
goldreich , p. and lynden - bell , d. 1965 , , 130 , 125 .
hassam , a. b. 1991 , comments plasma phys . and controlled fusion , 14 , 275 .
hassam , a. b. 1996 , nuclear fusion , 36 , 707 .
hassam , a. b. 1999 , phys .
plasmas , 6 , 3772 .
hayashi , y .- y . and young , w. r. 1987 , j. fluid mech . , 184 , 477 . hughes , d.w . and
cattaneo , f. 1987 , geophys . astrophys .
fluid dynamics , 39 , 65 .
ince , e. l. 1926 , ordinary differential equations , new york : dover .
kitchatinov , l. and rdiger , g. 1997 , , 286 , 757 .
miller , r. l. , waelbroeck , f. l. , hassam , a. b. , and waltz , r. e. 1995 , phys .
plasmas , 2 , 3676 .
newcomb , w. a. 1961 , phys .
fluids , 4 , 391 .
parker , e. n. 1979 , cosmical magnetic fields : their origin and thier activity , oxford : clarendon press .
parker , e. n. 1993 , , 408 , 707 .
roberts , k. v. and taylor , j. b. 1965 , phys .
fluids , 8 , 315 .
schou , j. et .
1998 , , 505 , 390 .
schussler , m. , caligari , p. , ferriz - mas , a. and moreno - insertis , f. 1994 , astron .
astrophys . , 281 , l69 .
schwarzschild , k. 1906 , nachr .
kgl . ges .
gttingen , p. 41 .
spiegel , e. a. and zahn , j .-
1992 , astron .
astrophys . , 265 , 106 .
thompson , m. j. et .
1996 , science , 272 , 1300 .
urpin , v. 1996 , , 280 , 149 .
velikhov , e. p. 1959
jetp , 36 , 995 .
waelbroeck , f. l. , and chen , l. 1991 , phys .
fluids b , 3 , 601 .
zweibel , e. g. , and heiles , c. 1997 , nature , 385 , 131 . | the effect of magnetic shear and shear flow on local gravitationally induced instabilities is investigated .
a simple model is constructed allowing for an arbitrary entropy gradient and a shear plasma flow in the boussinesq approximation . a transformation to shearing magnetic
coordinates achieves a model with plasma flow along the magnetic field lines where the coordinate lines are coincident with the field lines .
the solution for the normal modes of the system depends on two parameters : the alfvn mach number of the plasma flow and the entropy gradient .
the behavior of the unstable normal modes of this system is summarized by a stability diagram .
important characteristics of this stability diagram are the following : magnetic shear is stabilizing and the entropy gradient must exceed a threshold value for unstable mode growth to occur ; flow acts to suppress mode growth in a substantially unstable regime as expected , yet near marginal stability it can lessen the stabilizing effect of magnetic shear and enhance the growth rates of the instability ; and , as the alfvn mach number approaches one , the instability is completely stabilized .
analytical work is presented supporting the characteristics of the stability diagram and illuminating the physical mechanisms controlling the behavior of the model .
a derivation of the stability criterion for the case without shear flow , asymptotic solutions in the limit that the alfvn mach number approaches one and in the limit of zero growth rate , a complete wkb solution for large growth rates , an exactly soluble bounded straight field case , and energy conservation relations are all presented .
the implications of this work for astrophysical and fusion applications and the potential for future research extending the results to include compressibility are discussed . |
ground state spectroscopy on the lattice is by now a well understood physical problem with impressive agreement with experiment .
the lattice study of excited states is not as advanced .
the euclidean correlation function of an interpolating operator contains contributions from all states with the correct quantum numbers and the masses of the excited states show up only in the sub - leading exponentials
. a direct fit to a sum of exponentials is cumbersome since the signal is strongly dominated by the ground state . also with methods such as constrained fits , the maximum entropy method @xcite and a recently proposed method @xcite one still needs high statistics for reliable results .
an alternative method is the variational method @xcite where one diagonalizes a matrix containing all cross - correlations of a set of several basis operators @xmath0 with the correct quantum numbers , @xmath1 ( written here for infinite temporal extent of the lattice and operators projected to vanishing spatial momentum ) . for a large enough and
properly chosen set of basis operators each eigenmode is then dominated by a different physical state . finding these states
is equivalent to solving the generalized eigenvalue problem @xcite with eigenvalues behaving as @xmath2\;.\ ] ] each eigenvalue corresponds to a different energy level @xmath3 dominating its exponential decay ; @xmath4 is the distance to nearby energy levels .
the optimal operators which have maximal overlap with the physical states are linear combinations of the basis operators .
after normalization the largest eigenvalue gives the correlator of the ground state , the second - largest eigenvalue corresponds to the first excited state , and so on .
although the set of basis operators can be made arbitrarily large , in realistic calculations the intrinsic statistical errors make the diagonalization increasingly unstable .
the challenge is then to provide a set of operators large enough to span the physically relevant space but still small enough to allow significant results for the given statistics . in earlier work
@xcite we used three different interpolating fields for the nucleon sector , but could not resolve a signal for the roper excitation . here
we concentrate on the first two of those , @xmath5 u_c(x ) \;,\quad \chi_2(x ) = \epsilon_{abc } \left[u_a^t(x)\,c\ , d_b(x)\right ] \gamma_5 \,u_c(x ) \;.\ ] ] we concluded that in order to get a reliable roper signal it is also important to optimize the spatial properties of the interpolating operators .
it can be argued that a node in the radial wave function is necessary to capture reliably the roper state or other radially excited hadrons . in @xcite
we demonstrated that an elegant solution is to combine jacobi smeared quark sources with different widths to build the hadron operators and compute the cross - correlations in the variational method .
we find good effective mass plateaus for the first and partly the second and even the third radially excited states .
the eigenvalues @xmath6 are then fitted using standard techniques .
already earlier @xcite jacobi smeared sources were combined with point sources and cross - correlations studied in similar spirit ( see also @xcite ) .
the technique of jacobi smearing is well known @xcite . the smeared source lives in the timeslice @xmath7 and
is constructed by iterated multiplication with a smearing operator @xmath8 on a point - like source .
the operator @xmath8 is the spatial hopping part of the wilson term at timeslice 0 ; it is trivial in dirac space and acts only on the color indices .
this construction has two free parameters : the number of smearing steps @xmath9 and the hopping parameter @xmath10 .
these can be used to adjust the profile of the source .
we worked with two different sources , a narrow source @xmath11 and a wide source @xmath12 with parameters chosen such that the profiles approximate gaussian distributions @xcite .
the distribution of the two source - types correspond to half - widths of @xmath13 fm and @xmath14 fm .
the two sources allow the system to build up radial wave functions with and without a node .
the parameters were chosen such that simple linear combinations @xmath15 of the narrow and wide profile can approximate the first and second radial wave functions of the spherical harmonic oscillator : the coefficients @xmath16 approximate a gaussian with a half - width of @xmath17 fm , while @xmath18 approximate the corresponding excited wave function with one node . note , however , that the final contribution of the corresponding source terms is not pre - determined but results from the diagonalization . following the variational method
@xcite we compute the complete correlation matrix of operators .
the hadron sources we use for the correlation matrix are constructed from the narrow and wide quark sources .
respecting symmetries we then have for both types of nucleon operators ( [ chiops ] ) six combinations of wide or narrow quark sources , in total 12 interpolating field operators and a @xmath19 correlation matrix .
the final form of the wave function is determined through the eigenvectors resulting from the variational method .
we use the chirally improved dirac operator @xcite and work in the quenched limit , ie . , without dynamical fermions .
this operator is an approximate ginsparg - wilson operator and was well tested in quenched ground state spectroscopy @xcite where pion masses down to 250 mev have been reached at a considerably smaller numerical cost than needed for exact ginsparg - wilson fermions .
the gauge configurations were generated on a @xmath20 lattice with the lscher - weisz action @xcite ( further studies for larger lattices are in progress ) .
the inverse gauge coupling is @xmath21 , giving rise to a lattice spacing of @xmath22 fm as determined from the sommer parameter @xcite .
the statistics of our ensemble is 100 configurations . for the @xmath23
, @xmath24 quarks we use degenerate quark masses @xmath25 ranging from @xmath26 to @xmath27 .
more details are discussed in @xcite . and
@xmath28 baryons in units of the lattice spacing @xmath29 fm ( = 1/1.33 gev ) ( triangles : ground state , circles : negative parity ground state , squares : first positive parity excited state , full symbol : experimental value of ground state ) ; the mass of the strange quark has been adjusted to give physical @xmath30-meson mass .
[ fig1],title="fig:",width=257 ] and @xmath28 baryons in units of the lattice spacing @xmath29 fm ( = 1/1.33 gev ) ( triangles : ground state , circles : negative parity ground state , squares : first positive parity excited state , full symbol : experimental value of ground state ) ; the mass of the strange quark has been adjusted to give physical @xmath30-meson mass .
[ fig1],title="fig:",width=257 ] although we determine the full correlation matrix as discussed , it turns out to be numerically advantageous to analyze only a subset of operators .
the exponential decay of the three leading eigenvalues is clearly identified .
we identify these signals with the nucleon , the roper state and the next positive parity resonance @xmath31 .
the larger eigenvalues have too large statistical errors for a reliable interpretation ( see also @xcite concerning the problem of nucleon-@xmath32 ghost contributions ) . in principle
both interpolators @xmath33 and @xmath34 should couple to all states with the corresponding quantum numbers ; however , the coupling amplitude depends on the internal structure of the physical states ( with possibly different overlap ) .
since we can continuously increase the quark mass and thus make contact with the heavy quark region we also may argue in such a framework .
all roper states form an excited 56-plet of su(6 ) and in this multiplet the parity of any two - quark subsystem is positive .
the @xmath31 belongs to another multiplet containing both positive and negative parity two - quark subsystems .
the two - quark subsystem in the brackets in @xmath33 has positive parity , while the subsystem in the brackets of @xmath34 has negative parity .
hence in @xmath34 we should see only the signal from the @xmath31 state , and no signals from the nucleon and the roper .
this is exactly what we observe in @xcite , where the relevant figures for the nucleon channel are shown .
we may also combine quarks with different masses , allowing for a strange quark .
we fix the strange quark mass such that the @xmath30-meson mass has its experimental value in the chiral limit of the @xmath35-masses .
we then can determine also the @xmath36- and @xmath28-baryons using interpolator @xmath33 and again different width quarks sources .
first results are exhibited in fig .
we have to stress , however , that the results given here are for relatively small lattices ( spatial lattice size 1.8 fm ) .
a crucial test of our method is to check whether indeed the ground state is built from a nodeless combination of our sources and the excited states do show nodes .
this question can be addressed by analyzing the eigenvectors of the correlation matrix .
this has been done in @xcite and indeed confirms the expectation .
ground state masses approach , in naive chiral extrapolation , their experimental values well .
the excited state masses are above their experimental values .
since our results are for small spatial lattice size a volume dependence study still has to confirm this behavior .
excited states may be more affected by finite volume effects , but also be more sensitive to the quenched approximation .
finite volume studies , including also scaling properties , are under way .
n. mathur _ et al .
. lett .
b * 605 * ( 2005 ) 137 ; hep - lat/0405001 .
s. sasaki , prog .
* 151 * ( 2003 ) 143 .
d. guadagnoli , m. papinutto and s. simula , phys .
b * 604 * ( 2004 ) 74 , and these proceedings . c. michael , nucl .
b * 259 * ( 1985 ) 58 ; m. lscher and u. wolff , nucl .
b * 339 * ( 1990 ) 222 .
d. brmmel _ et al _ , phys .
d * 69 * ( 2004 ) 094513 ; nucl .
* 129 - 130 * ( 2004 ) 251 .
et al . _ , phys .
d * 70 * ( 2004 ) 054502 ; see also hep - lat/0409014 .
allton _ et al . _
[ ukqcd collaboration ] , phys .
d * 47 * ( 1993 ) 5128 .
t. burch , c. gattringer and a. schfer , hep - lat/0408038 .
c. best _ et al .
_ , phys .
d * 56 * ( 1997 ) 2743 ; c. alexandrou _ et al .
_ , nucl .
b * 414 * ( 1994 ) 815 . c. gattringer , phys .
d * 63 * ( 2001 ) 114501 ; c. gattringer , i. hip , c. b. lang , nucl . phys . b * 597 * ( 2001 ) 451 . c. gattringer _
[ bgr collaboration ] , nucl .
b * 677 * ( 2004 ) 3 ; nucl .
. suppl .
* 119 * ( 2003 ) 796 ; c. gattringer , nucl . phys .
* 119 * ( 2003 ) 122 .
m. lscher and p. weisz , commun .
* 97 * ( 1985 ) 59 ; g. curci , p. menotti and g. paffuti , phys .
b * 130 * ( 1983 ) 205 . c. gattringer , r. hoffmann and s. schaefer , phys .
d * 65 * ( 2002 ) 094503 . | whereas ground state spectroscopy for quenched qcd is well understood , it is still a challenge to obtain results for excited hadron states . in our study
we present results from a new approach for determining spatially optimized operators for lattice spectroscopy of excited hadrons . in order to be able to approach physical quark masses we work with the chirally improved dirac operator , i.e. , approximate ginsparg - wilson fermions .
since these are computationally expensive we restrict ourselves to a few quark sources .
we use jacobi smeared quark sources with different widths and combine them to construct hadron operators with different spatial wave functions .
this allows us to identify the roper state and other excited baryons , also in the strange sector . |
the first theoretical idea of neutron stars comes from baade and zwicky @xcite , being based on the analysis of supernova explosions . according to
, supernova explosion could be a transition from a star to a neutron star , consisting of closely packed neutrons in a compact size object .
after a few years , the derivation of the full general relativistic equation of hydrostatic equilibrium for spherically symmetric objects , now known as _ tolman - oppenheiemer - volkoff _ ( tov ) equation @xcite came into figure .
these equations were solved by oppenheimer and volkoff @xcite assuming that the matter consisted of non - interacting neutrons , and found that the maximal allowed mass for this to be 0.71@xmath3 .
later it was noticed that the inclusion of nuclear energy from the interacting neutrons increases this value substantially @xcite .
observational proof for the existence of neutron star was succeeded by jocelyn bell and and antony hewish in 1967 by observing the pulsating radio beams from objects named pulsars , the rotating neutron stars with a strong magnetic field @xcite .
generally , the behavior of matter in the interior of a neutron star is governed by the internal pressure , where all the neutrons obey the @xmath2-equilibrium condition with leptons and protons .
an extended review of theoretical and observational aspects of neutron star , from the surface to the core , with the emphasis on their structure and equations of state can be found in ref . @xcite .
further it is well known that the integral parameter for a foolproof structure of neutron star depends on their equation of state from an ideal theory , which can explain the @xmath2-equilibrium condition at high density . at present , a wide spectrum of different eos for neutron star matters has been designed from different interactions such as skyrme @xcite , the akmal - pandharipande - ravenhall @xcite and the relativistic mean field theory ( rmf ) @xcite ( see , e.g. , @xcite for reviews and the references therein for details ) .
an extensive investigation on the eos from these models show that they have almost similar properties at the nuclear saturation density ( matter where the density of protons is equal to that of the neutrons , @xmath4 ) @xmath5 0.16 @xmath6 but differ substantially at high density .
in other words , the properties of nuclear matter predicted from different models have analogous behavior around the saturation density but extremely different from each other at high density under @xmath2-equilibrium .
the recently observed massive neutron stars , j1614 - 2230 ( with mass of 1.97@xmath7 0.04@xmath3 ) @xcite and j0348 + 0432 ( with mass of 2.01@xmath7 0.04@xmath3 ) @xcite , have provided a reliable information on the existence of massive compact stars . at present , the uncertainties are quite large in the radius of compact stars @xcite . for example , using the pulse phase - resolved x - ray spectroscopy for pulsar j0437 - 4715 gives the radius of the compact star @xmath1 @xmath8 11 km with 3@xmath9 uncertainty @xcite .
furthermore , at high densities it is well known that there might be substantial population of heavy baryons ( i.e. hyperons ) , because these become energetically favorable once the fermi energy of neutrons reaches the order of their rest mass .
hence , it becomes quite necessary to include the hyperon matter in the study of highly dense compact objects .
in the last few decades , a large number of systematic studies including hyperon in the nuclear matter ( i.e. , hyper - nuclear matter or hyperon matter ) were carried out _ before _ @xmath10 _
after _ the observation of the pulsars and their identification with the neutron stars @xcite . most of these works demonstrate that the mass of the compact stars is reduced by 0.4@xmath3 when contribution of the hyperon matter is included in the eos , challenging the knowledge of the perceived modern mass @xcite .
this reduction in the mass of the compact stars due to hyperon matter in the eos under @xmath2-equilibrium is known as _
hyperon puzzle _ @xcite .
besides the calculation performed by including the @xmath11-@xmath12 cross - coupling and quartic terms in the rmf softens the eos to good agreement with the recent observed mass of compact stars @xcite .
in addition to that the generalized model having the low lying octet of baryons makes it difficult to obtain a neutron star mass greater than 2.0@xmath3 @xcite . in continuation of our earlier work @xcite , we here analyze the rotational attributes of neutron star with various hyperon - meson couplings within the framework of effective field theory motivated relativistic mean field model ( e - rmf ) @xcite . in brief , the motivation behind the present work is to investigate the effects of rotational profiles in the hyperon matter under @xmath2-equilibrium condition at high density . within the present scenario of the neutron star formation , the first millisecond pulsar b1937 + 21 rotating at frequency 641 hz @xcite , remained the most rapidly one for more than two decades . in 2006 , the discovery of pulsar j1748 - 2446ad with a faster rotating of frequency 716 hz was announced @xcite .
later observations of x - ray burst xte j1739 - 285 noticed a formation of neutron star with an initial rotation frequency @xmath13 1122 hz @xcite . on the other hand
the young pulsars has been known to have rotational frequency @xmath14 10 hz and in this category ( according to the observations so far ) psr j0537 - 6910 is known to be the fastest with @xmath15 62 hz @xcite . in fact , the angular velocity distribution of a neutron star evolves to a uniform rotation within a very short time period in any supernova event @xcite .
moreover , the sub - khz frequencies are still too low to significantly affects the structure of massive neutron stars @xcite .
however , for rapid rotation i.e. the sub - millisecond pulsar with super - khz frequencies , the rotation affects the massive neutron stars . reviewing the works concerning the rapidly rotating stars in general relativity
, there are few pioneering predictions , which have been observed by considering some of the sophisticated aspects of the neutron star : the nuclear equation of state , magnetic fields , magnetic field breaking and meridional flows etc .
here we give some of these references for such studies : butterworth & ipser ( 1976 ) @xcite ; komatsu _ et al . _
( 1989 ) @xcite ; bonazzola _ et al . _
( 1993 ) @xcite ; stergioulas & friedman ( 1995 ) @xcite ; laarakkers & poisson ( 1999 ) @xcite ; baumgarte _ et al . _
( 2000 ) @xcite ; ansorg _ et al . _
( 2002 ) @xcite ; birkl _ et al . _
2010 @xcite . for more extensive global collection of literature
, see friedman & stergioulas ( 2013 ) @xcite .
this present work will provide a relativistic mean field descriptions of the static and rotating ( sub - millisecond to super - millisecond ) compact star properties for various hyperon - meson couplings .
it is to be noted that the model used for the study of the eos of the hyper - nuclear is a chiral effective lagrangian ( e - rmf ) @xcite , the extension of the standard relativistic mean field ( rmf ) theory @xcite .
this paper is organized as follows : in section ii we discuss the theoretical setup for the relativistic mean field theory .
the parametrization and values of their coupling constant are also included in this section .
section iii@xmath16vi are assigned for the discussion of the results obtained from our calculation for the static and rotating compact stars . a short mathematical formulation for the equilibrium condition of static and rapidly rotating compact stars are also presented in the section iv & v.
finally , the summary and a brief conclusion are given in section vii .
the elementary theory for strong interaction to represent the complete description of nuclear equation of state is quantum chromodyanmics ( qcd ) . at present , it is not conceivable to describe the complete picture of the hadronic matter due to its non - peturbative properties . hence , one need to endorse the perception of effective field theory ( eft ) at low energy , known as quantum hadrodynamics ( qhd ) @xcite .
now - a - days , the mean field treatment of qhd has been used widely to describe the properties of infinite nuclear matter @xcite and finite nuclei @xcite . in this theory ,
the nucleons are considered as dirac particle , interact through the exchange of various mesons ( i.e. iso - scalar scalar @xmath9- , iso - scalar vector @xmath11- , iso - vector - vector @xmath12- and iso - vector - scalar @xmath17-mesons ) . the chiral effective lagrangian ( e - rmf ) is proposed by furnstahl , serot and tang @xcite , the extension of the standard relativistic mean field model @xcite . in e - rmf ,
the nonlinear lagrangian is expanded with the increasing order of the fields along with their derivative up to @xmath18 order of interaction under naive dimensional analysis @xcite and the concept of naturalness @xcite .
finally , the truncated lagrangian is given by @xmath19 the subscript @xmath20 ( i.e. @xmath20 = n , p for neutrinos and protons , respectively ) should be interpreted as sum over the states of nucleons .
the covariant derivative @xmath21 is defined as @xmath22 where , the term @xmath23 , and @xmath24 are the field tensors , @xmath25 @xmath26 here , @xmath27 , @xmath28 , @xmath29 and @xmath30 are the masses denoted for the baryon , @xmath9- , @xmath11- and @xmath12- meson , respectively . from this lagrangian , we derive the equation of motion and solve within mean field approximation self consistently .
the obtained field equations for @xmath9 , @xmath11 and @xmath12-meson are given by @xmath31 @xmath32 @xmath33 for a baryon species , the scalar density , @xmath34 , and baryon density @xmath35 are given as , @xmath36 @xmath37 where @xmath38 is the effective energy and @xmath39 and @xmath40 are the spin and isospin projection of baryon , respectively .
the quantity @xmath41 is the fermi momentum of the baryon , and @xmath42 is the effective mass , which can solve self - consistently .
now , the pressure density @xmath43 and the energy density @xmath44 for a given baryon density are expressed as follow , @xmath45 @xmath46 here , the constant @xmath47 is known as spin degeneracy parameter . the pressure and energy density contribution from leptons
are represented by @xmath48 and @xmath49 , respectively . [ cols="^,^,^,^,^,^,^,^,^,^,^,^,^,^,^,^,^,^ " , ] [ tab2 ]
in the interior of neutron stars , the neutron chemical potential goes beyond the total mass of the proton and electron . in other words , a probabilistic neutron star composition of neutron star
is asymmetric matter with an admixture of electrons rather than pure neutron matter .
here we have considered the @xmath2-equilibrium @xmath50 and the charge neutrality condition to monitor the concentrations of neutrons , protons and electrons inside the star .
it is worth mentioning that the contributions of neutrinos are not taken into account for simplicity of our calculation .
again , @xmath51 , @xmath52 , where @xmath53 and @xmath54 .
the term @xmath55 = @xmath56 , @xmath57 = @xmath58 and @xmath59 , @xmath60 are known as the chemical potentials and the fermi momentum for neutron and proton respectively .
up to this point , we have considered a neutron star comprised only of neutrons , protons and electrons .
as we know , the core of the neutron stars has a very high density ( @xmath61 78@xmath62 ) , therefore other low lying octet of baryons ( @xmath63 , @xmath64 , @xmath65 , @xmath66 , @xmath67 , and @xmath68 ) apart from the neutron and proton may become important in the eos of neutron star @xcite .
furthermore , at higher densities , the other strange mesons such as @xmath69 and @xmath70 are also crucial for the prediction of eos @xcite . in the present study
, we have not contemplated their effects , because of the limitation of the force parameter @xcite .
the electron _
e _ and muon @xmath71 are included for maintaining the charge neutrality and the @xmath2-equilibrium condition for the hyperons under the weak interaction @xcite , @xmath72 and @xmath73 . here ,
@xmath74 , @xmath75 , @xmath76 , @xmath77 and @xmath78 are the baryons , leptons , neutrino and anti - neutrino respectively . on the basis of the quark model
, one can assume that the hyperons interact with the mesons in two distinct modes by permitting the mixing and breaking of su(6 ) symmetry , keeping nuclear coupling constant intact @xcite .
we have used these two methods of parametrisations : ( i ) same coupling ratios as assumed by the quark model @xcite and ( ii ) different couplings strength for different baryons @xcite to deals with the octet of baryons in the eos of compact star .
the used hyperon - meson coupling ratios and their values for the present studies are listed in table [ tab2 ] .
( color online ) the equation of states obtained for nuclear and hyper - nuclear matter under charge neutrality as well as the @xmath79equilibrium condition from g2 force are compared with the empirical data ( shaded area in the graph ) by steiner et al for @xmath80 = @xmath1 with the uncertainty of 2@xmath9 @xcite . ] by imposing the above conditions on the eqs .
( [ eqnp],[eqne ] ) , we have self consistently calculated the energy density @xmath81 and pressure density @xmath82 of hyper - nuclear star matter as a function of baryon density .
the results obtained for nuclear ( i.e. , proton and neutron ) and the octet system ( i.e. proton , neutron , @xmath63 , @xmath64 , @xmath65 , @xmath66 , @xmath67 , and @xmath68 ) are shown in fig .
[ fig:2 ] . in those curves , there are five eoss for compact star , one from nucleonic matter and four from hyper - nuclear matter for various hyperon - meson coupling ratios . from the figure , it is clearly noticed that the eos from the nucleonic matter is a little stiffer compared to the hyperon matter at high density . in other words ,
the inclusion of hyperons into the compact star , gives a softer eos as compared to the nucleonic matter with respect to the baryon density .
comparison with the empirical data for @xmath80 = @xmath1 with the uncertainty 2@xmath9 of steiner _ et .
@xcite ( shaded region ) is also depicted . here
@xmath1 and @xmath80 are for the neutron radius and the photospheric radius , respectively .
more care inspection to the figure shows that the eos for nuclear system agrees well with the empirical data throughout the densities , but deviates a little for hyperon matter at high density .
for example , the eos obtained from the baryon octet family coincides with the empirical values up to the density @xmath61 6@xmath62 , then it becomes a little softer compared to empirical values .
hence , one can interpret that the inclusion of hyperon matter to the nucleon system makes the neutron star eos softer as shown in the figure , which is consistent with the other theoretical predictions @xcite .
the structure of a spherically symmetric and static compact star , consisting of relativistic matter modeled as perfect fluid , can be studied in term of energy density @xmath81 and the pressure density @xmath82 as a function of baryon density using tolman - oppenheimer - volkoff equations @xcite .
the general form of the tov equation is given by : @xmath83 \left[m+4\pi r^3 \cal p \right ] } { ( r-2 gm)},\ ] ] @xmath84 where @xmath85 and @xmath86 are the gravitational constant and the enclosed gravitational mass of radius @xmath87 , respectively .
for a given value of @xmath82 and @xmath81 , these equations can be integrated from the origin as an initial value problem for a given choice of central energy density .
the value of @xmath88 , where the pressure vanishes defines the surface of the star .
( color online ) the obtained static ( solid line ) and rotational ( dashed line ) compact star mass for nuclear and hyper - nuclear matter as a function of baryon density are given along with the recent mass observations @xcite .
see text for details . ]
( color online ) the obtained mass and radius trajectories nuclear and hyper - nuclear matter of static and rotational compact stars are compared with the recent mass observational datas @xcite ( the region between two solid horizontal blue line )
. the dashed orange vertical and horizontal lines are stands for the shift of mass and radius respectively due to the inclusion of rotational profile to the eoss of compact star .
a gravitational redshift of 0.35 @xcite observed from exo 0748 - 676 is also display in this figure .
the mass - radius constraints from thermal radiation of isolated neutron star rx j1856 @xcite ( orange hatched area ) and from qpos in the lmxbs 4u 0614 + 09 @xcite ( blue hatched area ) are given for constraining to the eoss of compact stars .
see text for more details . ] as we know , the fast rotating relativistic compact stars are sensitive to the stellar mass and to the equations of state at high density under @xmath2-equilibrium .
further , the rotational instabilities can produce gravitational waves , the detection of which would initiate a new field of observational asteroseismology of relativistic stars @xcite .
many interesting phenomena are studied from several independent numerical codes for obtaining accurate picture of rotating neutron stars in full general relativistic framework .
the metric of the space time for a realistic configurations of an axis - symmetric and stationary rotating compact star in spherical polar coordinate is given by @xcite : @xmath89 here , ( @xmath87 , @xmath90 , and @xmath91 ) are the spherical polar coordinates and the metric potentials @xmath77 , @xmath92 , @xmath2 , @xmath11 are functions of @xmath87 and @xmath90 only @xcite . in the limit of perfect fluid , the energy momentum tensor can be given as , @xmath93 where @xmath94 , @xmath43 , and @xmath95 are the total energy density , the pressure density and the four velocity , respectively @xcite . here , one should follow the three basic equations : ( i ) the einstein s equation for the metric potentials @xmath96 ( ii ) the rest mass conservation @xmath97 and ( iii ) the stress energy conservation @xmath98 see ref .
@xcite , for detailed mathematical expressions .
furthermore the einstein s equation is split into four component for the potentials and the four velocity as , @xmath99 here , @xmath100 is the spatial linear velocity with respect to an observer with zero angular momentum , @xmath101 now , we can use the limit on the maximum rotation , by the onset of mass shedding from the eos of the compact star . here , we have used the geometrized units , @xmath102 = @xmath85 = @xmath103 . for calculation of the rotational compact star properties like mass , radius and rotational frequency etc .
, we have used the well known rotational neutron star ( rns ) code , which is written by stergioulas @xmath10 friedman @xcite .
the calculated maximum mass @xmath0 and the radius @xmath1 for the static nuclear and hyperon stars are obtained from the well - known tov equations using the eoss of e - rmf .
the estimated results for the maximum mass as a function of density are compared with the observational data from pulsars j1614 - 2230 and j0348 + 0432 @xcite , shown in fig .
[ fig:3 ] .
we have shown the maximum mass and radius trajectory for the static and rotating compact star in fig .
[ fig:4 ] .
the dashed orange vertical and horizontal lines in the figure stands for the shift of mass and radius respectively due to the inclusion of rotational profile to the eoss of compact star ( will discuss in the preceding paragraph ) .
moreover , a gravitational redshift of 0.35 @xcite observed from exo 0748 - 676 is also shown in this figure .
the mass - radius constraints from thermal radiation of isolated neutron star rx j1856 @xcite ( orange hatched area ) and from qpos in the lmxbs 4u 0614 + 09 @xcite ( blue hatched area ) are given for constraining the eoss of compact stars .
ensuing these recent observations @xcite , it is clearly illustrated that the maximum mass predicted by any theoretical models should reach or near the limit @xmath61 2.0@xmath104 , which is consistent with our present prediction from the eos of nucleonic matter compact star .
but , the mass reduced somewhat by inclusion of hyperon matter to the eoss under @xmath2-equilibrium conditions . in other words , the maximum mass obtained from nuclear matter eos is reduced by @xmath61 0.4@xmath104 due to presence of hyperon matter at high density . for example , the mass predicted from the nuclear and hyper - nuclear matter are @xmath61 2.1@xmath104 and @xmath61 1.5@xmath104 , respectively from the tov equation ( see figs .
[ fig:3 ] @xmath10 [ fig:4 ] ) .
hence , one can easily contend that the predicted maximum mass from the hyperon matter eoss is underestimated to the the recent mass observations @xcite , which is the well known _
hyperon puzzle_. furthermore , there is no significant precise information regarding the constraint on the radius @xmath1 of the compact star .
hence in all cases , the radius of the neutron star corresponding to the maximum mass is within the range of @xmath61 12 km , which satisfy the standard value as predicted in the ref .
@xcite for @xmath105 skryme - hartree - fock forces .
( color online ) the gravitational mass @xmath106 radius configuration for eos of nuclear and hyperon matter compact stars with constant rotational frequency .
the rotational frequency is labeled for each line in the curve .
the horizontal line in each panel represent the recent mass observation @xcite are given for a ideal reference for the mass and radius .
see text for more details . ]
the stationary configurations of a compact star rotating at a given rotational frequency @xmath11 forms a one - parameter family characterized by central density @xmath107 , with limitation of two instabilities in the mass @xmath106 equatorial radius plane . for the high central density ,
the axis - symmetric perturbation makes it unstable , as a result the star collapse into a kerr black hole @xcite . in this section
we present results for compact stars rotating at sub - millisecond periods for a broad range of frequencies , [email protected] khz .
the calculations were performed for five eoss with nuclear and hyperon matter using rns equations .
mass @xmath106 radius relations for very fast rotating ( the keplerian frequency @xmath108 ) compact stars are given along with the non - rotating compact star properties in fig .
[ fig:3 ] @xmath10 [ fig:4 ] . from the figures
, it is clearly noticed that the rotational profile for the eos at @xmath108 brings a mass increase by @xmath109 and radius by @xmath61 2 @xmath104 ( see figs .
[ fig:3 ] @xmath10 [ fig:4 ] ) .
in other words , the maximum gravitational mass and radius for a given eos is enhanced by @xmath61 0.5@xmath3 and @xmath110 km , respectively for a compact star with keplerian frequency .
quantitatively , the mass predicted using eos of hyperon matter of certain coupling ratio are @xmath111 ( for static compact star using tov equations ) and @xmath112 for stationary rotation ( using rns equations ) .
the crucial question from the observational point of view is which the instability with respect to oscillations determine the bound state of the compact stars . in order to answer this question ,
we have calculated the gravitational mass - radius trajectories for several constant frequency sequences of uniformly rotating compact stars .
the results are displayed in fig .
[ fig:5a ] along with the recent mass observation @xcite .
the left and right panel of the figure are corresponding to the eos of nuclear and hyperon matter compact star , respectively . here , we give the results of mass - shedding for a given rotational frequency for only one eos of hyperon matter @xcite , as a example .
we have also found similar results for all other hyperon matter eoss obtained by using the various hyperon - meson couplings listed in table [ tab2 ] . in the figure , we have only shown the mass - shedding points for each frequency of millisecond compact stars with respect to axis - symmetric stationary rotation . from the figure ,
it is clear that the mass and the radius increases monotonically with the rotational frequency .
instability with respect to the mass - shedding from the equator implies that for a specific rotational frequency the gravitational mass and radius should be smaller than the maximum mass at keplerian limit . nevertheless , both features i.e. hypeon matter in the compact star and the rotation profile are of great practical importance and interest . in principle , they are very useful for selecting a perfect eos to constrain the mass measurement and is simultaneously able to work for a wide range of rotational frequencies .
( color online ) the particle fractions in hyperon matter as a function of density .
left panel , for constant coupling ratios @xcite and right panel , for variable coupling ratios @xcite .
the vertical black lines ( solid and dashed ) in each panel for the central density of the neutron star at the maximum mass . see text for more details . ]
in the above , we have mentioned that there are two different ways to incorporate the hyperon - meson coupling ratio into account : ( i ) constant coupling ratios as assumed by the quark model @xcite and ( ii ) variable couplings for different baryons @xcite .
the results obtained for the yield as a function of density using these coupling ratios are shown in fig .
[ fig:5 ] .
the left and right panel of the figure represents the prediction from the constant and variable couplings , respectively .
furthermore , the solid and dashed lines in each panel stand for the two sets of coupling for a specific approach in the hyperon - meson interactions . from the figure
, it is easily noticed that the @xmath66 is generated at @xmath113 2.0@xmath62 and its fraction increases rapidly to a saturation at @xmath114 @xmath61 3.1@xmath62 .
similarly , the @xmath115-hyperon generated at @xmath612.1@xmath62 and the yield becomes constant at @xmath615.1@xmath62 . from the figure , one can clearly notice that the yields for the two coupling ratios for each ( constant & variable ) approach have almost similar predictions . a careful investigation between two methodologies are found to differ a little from each other for the yields with respect to density . in other words ,
the procreation of hyperons as a function of density are altered a little for two approaches .
hence , one can conclude that the yields are marginally dependent on coupling ratios .
in the present study , we applied an effective field theory motivated relativistic mean field approach to the nuclear matter , where we investigated the influence of hyperon matter and the rotational profile on the properties of the compact star .
we have generated four eoss for the hyper - nuclear matter under @xmath2-equilibrium conditions at high density using various hyperon - meson coupling ratios .
all the predicted eoss for the hyperon stars have produced a maximum mass @xmath61 1.5@xmath3 , which is lower compared to the recent star observation data @xcite .
the maximum radii corresponding to the masses are located in the range of 12@xmath1614 km .
the influence of hyperon in the present study are consistent with the predictions of other models , both early and the recent ones in their mass spectrum under charge neutrality and @xmath2-equilibrium conditions . in continuation to the hyperon problem , we have included the rotational profile under axis - symmetric constant rotation with keplerian frequency into account using rotational star equation , which increase the mass of the compact star by @xmath109 solar mass in magnitude as compared to the static case . in other words ,
the eos of hyperon matter along with the rotational profile predicts the extreme mass @xmath61 2.2@xmath104 , which is in contrast with the late observation @xcite at keplerian limit .
furthermore , we have also calculated the mass - shedding for a given eos for rotational frequency sequences of uniformly rotating compact stars .
we found that the mass and the radius are increased monotonically with the rotational frequency for millisecond pulsar .
instability with respect to the mass - shedding from the equator implies that for a specific rotational frequency the gravitational mass and radius should be smaller than the maximum mass at keplerian limit .
hence , it is important to consider both the feature i.e. hypeon matter and the rotation profile in the compact stars . in principle
, they are very useful for selecting a perfect eos to constrain the mass measurement and is simultaneously able to work for a wide range of rotational frequencies . in summary ,
the present study is very useful in the admiration of the expectation of static and stationary rotating hyper - nuclear compact stars at high density .
this work has been supported by the 973 program of china ( grant no .
2013cb834400 ) , the nsf of china ( grants no . 11120101005 , no . 11275248 , and no .
11525524 ) , and the chinese academy of sciences ( grant no .
kjcx2-ew - n01 ) .
the computation of this work was supported by the hpc cluster of skltp / itp - cas and the supercomputing center , cnic of cas .
the authors are also thankful to g. narain , cheng - jun xia , bing wang and xu meng for discussions and also for careful reading of the manuscript .
99 w. baade and f. zwicky , phys .
rev . * 46 * , 76 ( 1934 ) .
r. tolman , phys . rev . * 55 * , 364 ( 1939 ) .
j. oppenheimer and g. volkoff , phys . rev . * 55 * , 374 ( 1939 ) .
s. chandrasekhar , ast .
j. * 74 * , 81 ( 1931 ) .
p. ghosh , _ rotation and accretion powered pulsars _ ,
world scientific , singapore , ( 2007 ) .
p. hansel , a. y. potekhin and d. g. yakovlev , * vol-326 * of astrophysics and space science library , ( 2007 ) ( http://www.astroscu.unam.mx/neutrones/home.html ) .
j. r. stone , j. c. miller , r. koncewicz , p. d. stevenson and m. r. strayer , phys .
c * 68 * , 034324 ( 2003 ) .
a. akmal , v. r. pandharipande and d. g. ravenhall , phys .
c * 58 * , 1804 ( 1998 ) . j. d. walecka , _ theoretical nuclear and subnuclear physics _ , imperial college and world scientific , london and singapore , ( 2004 ) , isbn 981 - 238 - 898 - 2 .
s. l. shapiro and s. a. teukolsky , _ black holes , white dwarfs and neutron stars : the physics of compact objects _ , john wiley and sons , new york , ( 1983 ) .
f. weber , _ pulsars as astrophysical laboratories for nuclear and particle physics _ , inst .
publishing , bristol , ( 1999 ) .
b. demorest , t. pennucci , s. m. ransom , m. s. e. roberts , and j. w. t. hessels , nature ( london ) * 467 * , 1081 ( 2010 ) .
j. antoniadis et al . , science * 340 * , 6131 ( 2013 ) . s. guillot , r. e. rutledge , and e. f. brown , astrophys .
j. * 732 * , 88 ( 2011 ) .
t. gver , p. wroblewski , l. camarota , and f. ozel , astrophys .
j. * 719 * , 1807 ( 2010 ) .
d. k. galloway and n. lampe , astrophys .
j. * 747 * , 75 ( 2012 ) . s. bogdanov , astrophys .
j. * 762 * , 96 ( 2013 ) . v. a. ambartsumyan and g. s. saakyan , astron .
zh . * 37 * , 193 ( 1960 ) .
s. a. moszkowski , phys .
d * 9 * , 1613 ( 1974 ) .
n. k. glendenning , phys .
b * 114 * , 392 ( 1982 ) .
n. k. glendenning , astrophys .
j. * 293 * , 470 ( 1985 ) .
n. k. glendenning , zeitschr .
phys . a * 326 * , 57 ( 1987 ) .
f. weber and m. k. weigel , nucl .
a * 493 * , 549 ( 1989 ) .
j. i. kapusta and k. a. olive , phys .
lett . * 64 * , 13 ( 1990 ) .
j. ellis , j. i. kapusta , and k. a. olive , nucl .
b * 348 * , 345 ( 1991 ) .
p. j. ellis , r. knorren , and m. prakash , phys .
b * 349 * , 11 ( 1995 ) .
j. schaffner and i. n. mishustin , phys .
c * 53 * , 1416 ( 1996 ) .
a. sedrakian , prog .
* 58 * , 168 ( 2007 )
. s. balberg , i. lichtenstadt , and g. b. cook , astrophys .
* 121 * , 515 ( 1999 ) .
h. -j . schulze and t. rijken ,
c * 84 * , 035801 ( 2011 ) .
d. logoteta , i. vida na , c. providencia , a. polls , and i. bombaci , j. phys .
* 342 * , 012006 ( 2012 ) .
xian - feng zhao , and huan - yu jia , phys .
c * 85 * , 065806 ( 2012 ) .
i. bombaci , arxiv:1601.05339 , ( 2016 ) .
p. arumugam , b. k. sharma , p. k. sahu , s. k. patra , t. sil , m. centelles and x. vinas , phys .
b * 601 * , 51 ( 2004 ) . t. k. jha , p. k. raina , p. k. panda and s. k. patra , phys .
c * 74 * , 055803 ( 2006 ) . c. j. horowitz , j. pielarewicz , phys .
lett . * 86 * , 5647 ( 2001 ) .
s. k. singh , m. bhuyan , p. k. panda and s. k. patra , j. phys .
g : nucl . part .
* 40 * , 085104 ( 2013 ) .
r. j. furnstahl , b. d. serot and h. b. tang , nucl .
a * 598 * , 539 ( 1996 ) .
d. c. backer , s. r. kulkarni , c. heiles _ et al .
_ , nature * 300 * , 615 ( 1982 ) . j. w. t. hessels , s. m. ransom , i. h. satirs , p. c. c. freire , v. m. kaspi , and f. camilo , science * 311 * , 1901 ( 2006 ) .
p. kaaret , z. prieskorn _ et al .
_ , astrophys . lett .
* 657 * , 97 ( 2007 ) .
f. e. marshall _ et al .
_ , astrophys .
j. * 603 * , 682 ( 2004 ) .
n. andersson , astrophys .
j. * 502 * , 708 ( 1998 ) .
n. andersson , k. d. kokkotas , int .
d * 10 * , 381 ( 2001 ) .
s. l. shapiro , s. a. teukolsky , i. wasserman , astrophys .
j. * 272 * , 702 ( 1983 ) .
e. m. butterworth , and j. r. ipser , astrophys .
j. * 204 * , 200 ( 1976 ) .
h. komatsu , y. eriguchi , and i. hachisu , mon . not .
r. astron .
237 * , 355 ( 1989 ) .
s. bonazzola , e. gourgoulhon , phys .
d * 48 * , 2635 ( 1993 ) .
n. stergioulas and j. l. friedman , astrophys .
j. * 444 * , 306 ( 1995 ) .
w. g. laarakkars , and e. poisson , astrophys . j. * 512 * , 282 ( 1999 ) . t. w. baumgarte , s. l. shapiro , and m. shibata , astrophys . j. * 528 * , l29 ( 2000 ) .
m. ansorg , a. kleinwchter , and r. meinel , astron .
astrophys . * 381 * , l49 ( 2002 ) .
r. birkl , n. stergioulas , and e. mller , phys .
d * 84 * , 023003 ( 2011 ) .
j. l. friedman and n. stergioulas , _ rotating relativistic stars _ , cambridge university press , united kingdom ( 2013 ) .
h. mller and b. d. serot , nucl .
a * 606 * , 508 ( 1996 ) .
r. j. furnstahl and b. d. serot , nucl .
a * 671 * , 447 ( 2000 ) .
i. bednarek and r. manka , int .
d * 10 * , 607 ( 2001 ) .
i. bednarek and r. manka , phys .
c * 73 * , 045804 ( 2006 ) .
b. k. sharma , p. k. panda and s. k. patra , phy .
c * 75 * , 035808 ( 2007 ) .
r. j. furnstahl , b. d. serot and h. b. tang , nucl .
a * 615 * , 441 ( 1997 ) .
s. k. singh , s. k. biswal , m. bhuyan , and s. k. patra , phys .
c * 89 * , 044001 ( 2014 ) s. k. singh , s. k. biswal , m. bhuyan , and s. k. patra , j. phys .
. phys . * 41 * , 055201 ( 2014 ) .
b. d. serot and j. d. walecka , in _ advances in nuclear physics _ , edited by j. w. negele and erich vogt _ plenum press , new york
* 16 * , p. 1
w. pannert , p. ring , and j. boguta , phy .
lett . , * 59 * , 2420 , ( 1986 ) .
j. boguta and a. r. bodmer , nucl .
a * 292 * , 413 ( 1977 ) .
s. k. patra , m. bhuyan , m. s. mehta and raj k. gupta , phys .
c * 80 * , 034312 ( 2009 ) .
m. bhuyan , s. k. patra , and raj k. gupta , phys .
c * 84 * , 014317 ( 2011 ) .
reinhard , rep .
. phys . * 52 * , 439 ( 1989 ) .
p. ring , prog .
* 37 * , 193 ( 1996 ) .
j. meng , h. toki , s. g. zhou , s. q. zhang , w. h. long and l. s. geng , prog .
. phys . * 57 * , 470 ( 2006 ) .
m. bhuyan , phys .
c * 92 * , 034323 ( 2015 ) .
h. liang , j. meng , and s. g. zhou , phys .
rep . * 570 * , 1 ( 2015 ) .
j. meng and s. g. zhou , j. phys .
g : nucl . part .
phys . * 42 * , 093101 ( 2015 ) .
h. mller and b. d. serot , nucl .
a * 606 * , 508 ( 1996 ) .
b. d. serot and j. d. walecka , int
e * 6 * , 515 ( 1997 ) .
j. m. lattimer , and a. w. steiner , eur .
j. a * 50 * , 40 ( 2014 ) .
h. nakada , phys .
c * 78 * , 054301 ( 2008 ) .
d. alonso and f. sammarruca , phys .
c * 67 * , 054301 ( 2003 ) .
e. n. e. van dalen , c. fuchs and a. faessler , nucl .
a * 744 * , 227 ( 2004 ) .
t. klahn et _ al .
c * 74 * , 035802 ( 2006 ) .
p. danielewicz , r. lacey and w. g. lynch , science * 298 * , 1592 ( 2000 ) .
w. g. lynch , m. b. tsang , y. zhang , p. danielewicz , m. famiano , z. li and a. w. steiner , prog .
. phys . * 62 * , 427 ( 2009 ) .
n. k. glendenning , _ compact stars , nuclear physics , particle physics and general relativity _ , new york : springer ( 1997 ) .
j. schaffner - bielich and a. gal , phys .
c * 62 * , 034311 ( 1999 ) .
s. weissenborn , d. chatterjee and j. schaffner - bielich , nucl .
a * 881 * , 62 ( 2012 ) .
a. r. bodmer , nucl . phys .
a * 526 * , 703 ( 1991 ) .
h. f. lu , j. meng , s. q. zhang , and s. -g .
zhou , euro .
j. a * 17 * , 19 ( 2003 ) .
h. f. lu , x. s. wang , and y. y. liu , comm .
phys . * 62 * , 272 ( 2014 ) .
j. schaffner , c. b. dover , a. gal , d. j. milliner , c. greiner and h. stocker , ann .
* 235 * , 35 ( 1994 ) .
j. boguta and s. bohrmann , phys .
b * 102 * , 93 ( 1981 ) . n. k. glendenning , s. a. moszkowski , phys .
lett . * 67 * , 2414 ( 1991 ) .
m. chiapparini , m. e. bracco , a. delfino , m. malheiro , d. p. menezes and c. providencia , nucl .
a * 826 * , 178 ( 2009 ) .
a. w. steiner , s. reddy and m. prakash , phys .
d * 66 * , 094007 ( 2002 ) .
a. w. steiner , j. m. lattimer and e. f. brown , astrophys .
j. * 722 * , 33 ( 2010 ) .
b. s. sathyaprakash and b. f. schutz , liv .
* 12 * , 2 ( 2009 ) .
m. anderson et al .
lett . * 100 * , 191101 ( 2008 ) .
j. cottam , f. paerels and m. mendez nature ( london ) * 420 * , 51 ( 2002 ) .
s. van straaten , e. c. ford , m. van der klis , m. mendez and p. kaaret astrophys . j. * 540 * , 1049 ( 2000 ) .
f. sammarruca , phys .
c * 79 * , 034301 ( 2009 ) .
et al . _ ,
c * 85 * , 035201 ( 2012 ) .
roy p. kerr , phys .
lett * 11 * , 237 ( 1963 ) . | in this theoretical study , we report an investigation of the equations of state ( eoss ) of hyper - nuclear matter and its composition as a function of density within the framework of effective field theory motivated relativistic mean field model .
we have used g2 force parameter along with various hyperon - meson coupling ratios by allowing the mixing and the breaking of su(6 ) symmetry to predict the eoss , keeping the nucleonic coupling constant intact .
we have estimated the properties of non - rotating and rapidly rotating configuration of compact stars by employing a representative set of equations of state .
the obtained results of the mass @xmath0 and radius @xmath1 for the compact stars are compared with the recent mass observations .
we noticed that the inclusion of hyperon matter to the eoss of nuclear matter under @xmath2-equilibrium condition decreases the maximum mass of the compact star by 0.4@xmath3 in magnitude in comparison to the nuclear matter prediction .
further , we have studied the stability and sensitivity of rotational frequency ( at sub - millisecond period ) on the configuration of the compact stars , because the angular frequency is significantly smaller than the mass - shedding ( keplerian ) frequency in slow rotation regime .
moreover , the yield of hyperon as a function of density for various hypeon - meson couplings are also estimated . |
in the presence of a perpendicular magnetic field , the quasi - free charge carriers of a two - dimensional electron system ( 2des ) condense in equidistant and numerously degenerated landau levels @xcite .
variations of the electrostatic potential locally shift the degeneracy and , therefore , cause the formation of landau bands .
the bending of the landau dispersion at the edge of a 2des is discussed thoroughly in the literature @xcite as it is the key for the understanding of the quantum hall effect @xcite . if the lateral extension of a 2des is limited electrostatically or by the physical edge of the underlying ( lithographically defined ) structure , the landau bands rise on a length scale which substantially exceeds the magnetic length . however
, if the confining barrier is fabricated by means of epitaxy , namely , by the technique of cleaved - edge overgrowth @xcite , the realization of a sharp edge potential becomes possible .
the corresponding edge channels are then located at distances on the order of the magnetic length @xcite . in this paper , we report on the magnetotunneling spectroscopy between two quantum hall systems which are separated by an atomically precise potential barrier within a gaas / algaas heterostructure
. the width of the barrier amounts to , i.e. it falls always below the magnetic length . in the vicinity of the barrier , a complex landau band structure exists which can be described in the case of weak coupling as a superposition of the mirror - inverted dispersions of both subsystems .
the degeneracy at the crossings is lifted by small landau band gaps @xcite .
the tunneling current through the structure becomes maximum when the fermi level coincides with one of these anticrossings @xcite .
the conductance is additionally determined by random tunneling centers in the barrier which are responsible for quantum interferences between opposite edge channels @xcite .
while the electron systems of our structure are field - induced by means of a gate electrode , cf .
[ fig : intro]a , a former experiment by kang . is based on modulation - doped electron films @xcite .
a fixed electron density implies that a certain landau band gap coincides with the fermi level only for a small range of the magnetic field .
in contrast , a sample design with a gate electrode allows to investigate a particular band gap at different magnetic fields provided that the fermi level is adjusted in a suitable way . in order to probe the same anticrossing for different barrier widths ,
it is , therefore , not necessary to prepare several samples as the effective shape of the barrier ( in units of the magnetic length @xmath0 and the cyclotron energy @xmath1 ) can simply be varied by the magnetic field .
carrying out measurements for different electron densities with the same sample is also indispensable if quantum interferences are investigated which depend on the random configuration of tunneling centers within the barrier .
another effect can be observed just with a single sample , namely , the small magnetic shift of the landau band gaps on the scale of the cyclotron energy which would otherwise be perturbed by the unavoidable fluctuations between successive growth processes .
all effects discussed in this paper depend basically on the shape of the landau band structure at the tunneling barrier ( fig .
[ fig : intro]b ) . for weakly coupled electron systems , ho has calculated the energy dispersion by using an approximation approach @xcite . though the result reflects the actual dispersion very well ,
the gap positions differ noticeably from the findings of a complete quantum mechanical calculation if a fine energy resolution is relevant as for our experiment @xcite . instead of combining the dispersions of both subsystems , takagaki and ploog
have developed a tight - binding model which is not only applicable for weakly , but also for strongly coupled electron systems @xcite .
the tunneling barrier is thereby represented by a reduced hopping amplitude between two simulation grid lines . since the effective shape of the barrier is not explicitly considered in this model ,
the latter is inappropriate to make accurate predictions on the gap position in dependence of the magnetic field .
the landau dispersion which is necessary for the interpretation of the obtained experimental data has been calculated exactly by solving the single - electron schrdinger equation . since the potential term in the hamiltonian is given by a superposition of the parabolic magnetic confinement potential and the piecewise constant conduction band offset of the intrinsic semiconductor heterostructure , the problem is generally closely related to the linear quantum harmonic oscillator .
the analytic solution becomes possible with an ansatz for the wave function which is composed of parabolic cylinder functions . for calculating the energy eigenstates ,
we have developed a dedicated algorithm which numerically solves the continuity conditions at the heterojunctions .
it is generally applicable to all systems which consist of up to three regions of constant potential .
hence , this method yields the landau dispersion for systems with either a rectangular barrier ( biased or not ) or a potential well ( quantum wire ) @xcite .
this article is organized as follows : the sample structure and the corresponding landau dispersion are introduced in the next two sections .
the subject of sec .
[ sec : tun ] consists of two different tunneling models which are the basis for the understanding of the landau oscillations and simultaneously occurring conductance fluctuations . in sec .
[ sec : basic ] we discuss a method for determining the electron density from magnetotunneling measurements .
the comparison of the conductance traces with the landau band structure is carried out in sec .
[ sec : ll - osci ] where the focus lies on the magnetic shift of the anticrossings .
section [ sec : ab ] deals with aharonov - bohm oscillations at the first conductance peak on the scale of the filling factor .
conductance fluctuations which appear at low magnetic fields are finally discussed in sec .
[ sec : fluct_low ] .
for a quantitative investigation of the energy dispersion in dependence of the magnetic field , it is necessary to have both a well - defined tunneling barrier and electron systems of variable density .
the molecular beam epitaxy ( mbe ) allows the fabrication of gaas / algaas interfaces with a roughness which is as low as one monolayer . for low - dimensional quantum systems like ours ,
one commonly needs a sharp potential modulation not just in one , but in two directions of space . for this purpose ,
pfeiffer . have developed the cleaved - edge overgrowth ( ceo ) method which allows to grow subsequently two layer sequences at right angles to another @xcite . with this technique ,
the sample structure of fig . [ fig : intro]a can be realized .
it contains two laterally adjacent electron systems which reside in the layers of the first growth step and are induced by means of a gate electrode which is fabricated during the second mbe step .
the source and drain contacts of the heterostructure consist of gaas layers which are silicon - doped to 217@xmath2 with thicknesses of 500 and 1100@xmath3 nm , respectively .
the quantum region is composed of a @xmath4 thick @xmath5 barrier and two embedding layers of intrinsic gaas ( each 2 m ) .
after the first growth step , the wafer is discharged from the mbe machine and chemically polished to a thickness of 80 to @xmath6 by means of a bromine methanol solution .
the wafer is then cleaved into small pieces which are scratched at a certain position and transferred into the growth chamber again . immediately after breaking the samples _ in - situ _ , the freshly exposed cleavage planes are overgrown with @xmath7 of @xmath8 .
the barrier region is followed by a @xmath9 thick layer of @xmath10-gaas with an electron density of 218@xmath2 .
it acts as the gate electrode . by etching a mesa ,
the buried source contact is exposed .
the samples are divided into stripes where each piece contains a @xmath11 long section of the cleavage plane .
ohmic contacts on the @xmath10-layers are realized with indium droplets .
these are deposited by a soldering iron and subsequently alloyed into the crystal at about 360@xmath12c .
if the active region of a ceo sample is accessed via @xmath10-layers of the ( 001 ) growth step , the occurrence of bulk leakage currents is generally possible @xcite . in the case of our structure ,
the leakage current flows apart from the coupled electron systems through the 4 m thick intrinsic region . in the ( 001 ) cross - section of the heterostructure , the induced 2dess account only for about 1/100,000 of the total area . in order to prevent that conductance oscillations in the quantum region can not be measured against the background of the bulk leakage current
, the aluminum content in the tunneling barrier has to be sufficiently high .
thus the sample design of fig .
[ fig : intro]a allows only landau band gaps which are considerably smaller than the cyclotron energy . since for such a dispersion the number of opposite edge states with significant overlap is small in comparison to the low - barrier case which is exemplarily shown in fig . [ fig : intro]b , this coupling regime is denoted here as weak independently from the absolute tunneling resistance .
the quality of the indium contacts and of the tunneling barrier is revealed by the two - point i - v - curves of fig .
[ fig : curves]a : both pairs of junctions on the source and drain contact layers show a perfect ohmic behavior .
the ratio of their reciprocal resistance coincides with the thickness ratio of the underlying @xmath10-contact layers . according to the @xmath13-@xmath14 curve d@xmath15s , which represents the leakage current between source and drain for a vanishing gate voltage , the bulk resistance amounts at least @xmath16 for applied voltages below @xmath17 .
therefore , magnetotransport measurements can be carried out with ac voltages which are small enough so that the bulk resistance always exceeds the resistance of the induced electron systems .
the magnetotransport properties of the quantum region have been investigated by means of lock - in technique . the corresponding circuit is shown in fig .
[ fig : intro]a . in spite of the four - point contact geometry ,
the active region is effectively accessible only via two leads , namely , the extended @xmath10-layers of the first growth step .
this fact prevents a direct determination of the charge carrier density by standard methods like shubnikov - de haas or ( quantum ) hall measurements .
the coupled electron systems have been studied by applying an ac voltage of @xmath18 via a high - impedance resistor @xmath19 .
this setup ensures a practically constant current of @xmath20 .
a quasistatic measurement is enabled by a low ac frequency of @xmath21 .
all measurements have been carried out within a @xmath22he system which provides temperatures down to @xmath23 .
the resistance traces shown in fig .
[ fig : curves]b in dependence of the magnetic field bear some resemblance to the shubnikov - de haas effect . in the following ,
the prominent long - period resistance oscillations are called _ landau oscillations _ as they originate from the periodic coincidence of the fermi level with the landau band gaps .
other magnetotransport features are two different types of conductance fluctuations which appear respectively at low field strengths ( @xmath24 ) and in the range of the rightmost resistance minimum ( in fig .
[ fig : curves]b best observable for @xmath25 ) . for the understanding of all phenomena ,
it is necessary to know the landau band structure at the tunneling barrier quantitatively .
provided that the electrostatic potential of a two - dimensional hall system is constant , the energy levels of the charge carriers are given by the solutions of the quantum harmonic oscillator @xcite .
close to potential variations , the degeneracy of the landau levels is lifted so that they form dispersive bands .
the particular landau bands increase monotonically for electrons approaching the confinement potential of a 2des @xcite . in the vicinity of a tunneling barrier
, the landau dispersion shows an oscillating structure which is calculated in the following . with regard to the conduction band shape of the gaas / algaas / gaas heterostructure ,
the treatment is restricted to electron systems which are unidirectionally modulated by three regions of constant potential .
the uniform potential within each region makes a representation of the energy eigenfunctions by parabolic cylinder functions possible .
besides , the constraint concerning the number of heterojunctions keeps the equation system which results from the continuity conditions solvable . for a system with a single potential discontinuity
, one has to determine only the energy eigenvalue which can be done by means of a standard solution method .
a sample structure with two heterojunctions already requires a dedicated solution algorithm for the simultaneous calculation of both the energy eigenvalue and a characteristic mixing parameter . for the analytic and numerical details we refer to refs .
@xcite and @xcite .
the schrdinger equation which describes the whole quantum region is solved in the scope of the effective mass approximation .
for the gaas layers and the algaas barrier , we use the same effective mass @xmath26 .
the aluminum content @xmath27 of the potential barrier corresponds to a conduction band offset of @xmath28 with respect to gaas @xcite . as the barrier height is a multiple of the cyclotron energy @xmath29 , the probability density for electrons at typical magnetic fields is very low inside the algaas material .
consequently , the assumption of a constant effective mass throughout the whole structure is a rather good approximation . for strongly coupled electron systems , however ,
the difference concerning both the cyclotron energy and the cyclotron radius becomes absolutely relevant @xcite .
if the conduction band offset is assumed to vary just along the @xmath30-axis , i.e. in the direction of the first growth step , the schrdinger equation for a single electron subjected to a magnetic field reads @xmath31 for electrons propagating parallel to the tunneling barrier and perpendicular to @xmath32 , it is appropriate to use the landau gauge @xmath33 .
this definition of the vector potential implies a wave function which represents a localized state in the @xmath30- and a plane wave in the @xmath34-direction : @xmath35 the integer @xmath36 , 1 , 2 , @xmath37 is the landau band index , and @xmath38 stands for the angular wavenumber which determines the wavelength along the @xmath34-direction .
the extension @xmath39 of the system is contained in the ansatz of eq . for the purpose of normalization .
the @xmath30-dependent component @xmath40 describes a bound state . by using eq . and the magnetic length @xmath41
, the schrdinger equation transforms to @xmath42+v(x)\right\}\varphi{_{nk}}(x)=e{_{nk}}\varphi{_{nk}}(x ) .
\label{eqn : sgl1dim}\ ] ] the elimination of the @xmath34-coordinate reduces the partial differential equation to an ordinary differential equation which now solely determines the localized state @xmath43 .
the @xmath30-quadratic term of the hamiltonian represents the magnetic confinement potential which is given by a parabola centered at @xmath44 .
only for bulk states , which are not influenced by potential variations , the quantity @xmath45 is always identical to the quantum mechanical expectation value of the electron location .
if @xmath45 resides , e.g. , within an infinitely high potential step , the probability density vanishes at @xmath45 and the center of mass deviates from this location . in the semiclassical picture
, the variable @xmath45 stands for the center of the skipping orbits of electrons propagating along a barrier .
therefore , the parameter @xmath45 is commonly called guiding center . while dispersion relations are generally plotted against the wavenumber @xmath38 , the close connection between the guiding center and the actual location of the electron often gives reasons to use @xmath44 as the abscissa in plots of the landau band structure .
this enables a direct comparison between the dispersion and the potential landscape even when @xmath45 substantially differs from the actual electron position .
electrons which occupy bulk states are described by wave functions @xmath43 which are already converged to zero at the next potential variation . for these electrons ,
the term @xmath46 may be neglected in eq .
which then becomes identical to the schrdinger equation for a quantum harmonic oscillator .
therefore , in regions where the conduction band is flat , the potential in the hamiltonian is except for a constant completely determined by the effect of the lorentz force :
@xmath47 the last term contains with @xmath48 a dimensionless variable which is used in the following in place of @xmath30 .
bulk electrons which are bound within the parabolic confinement potential @xmath49 occupy the equidistant energy levels @xmath50 .
the corresponding wave functions are given by the hermite functions : @xmath51 . from fig .
[ fig : hermite]a it follows that the extension of the wave functions is approximately limited by the parabola @xmath52 .
in particular , about 90 of all electrons reside within the interval @xmath53 $ ] .
hence , one can deduce a cyclotron radius of @xmath54 which is in fact the half extension of the wave function . for electrons with a guiding center which is located at a distance of about @xmath55 or less in respect of a potential step ,
the quantum harmonic oscillator is no longer a good approximation . for solving the schrdinger equation separately for each region of constant potential ,
it is necessary to admit for a moment the most general solution which also includes diverging eigenfunctions . by using the dimensionless variables @xmath56 and @xmath57 , where the latter is defined according to @xmath58
, the schrdinger equation transforms to the weber differential equation . for the @xmath59-th region of constant potential @xmath60
, we get @xmath61\right\}\varphi{_{nk}}(\xi)=0 .
\label{eqn : weberdeq}\ ] ] the corresponding two - dimensional solution space is spanned by the linearly independent parabolic cylinder functions @xmath62 @xcite .
if both outermost intervals are of the same material in our case gaas , it is appropriate to set @xmath63 which leads to the total wave function @xmath64 the parameters @xmath65 denote the positions of both heterojunctions which are assumed to be located symmetrically around the origin . as illustrated in fig .
[ fig : hermite]b , the function @xmath66 diverges on the negative @xmath30-axis if @xmath67 is not an integer . outside the barrier region , the wave functions are represented , therefore , just by one of both basis functions .
the essential parameters of the system are the energy eigenvalue @xmath57 and the scalar @xmath68 .
the latter represents the mixing ratio of both parabolic cylinder functions within the algaas region . the unknown variables @xmath57 and @xmath68 as well as the prefactors @xmath69 and @xmath70 are determined by the requirement of continuity for both @xmath71 and @xmath72 at the heterojunctions .
alternatively , one may also demand a continuous transition of @xmath71 and @xmath73 which allows to determine @xmath57 and @xmath68 independently from @xmath74 . the simultaneous solution of the two corresponding equations , which are highly non - linear ,
has been realized by a dedicated algorithm which is discussed in detail elsewhere @xcite . for a set of ( equidistant ) @xmath45-values , the method yields ( almost ) all solution tuples @xmath75 within a given energy range . the obtained eigenstates are , however , not yet assigned to the particular landau bands . for the determination of the band index @xmath76
, one may exploit the properties of the discrete energy spectrum of an one - dimensional schrdinger equation : the energy levels are non - degenerate and , if arranged according to ascending energy , the corresponding wave functions possess 0 , 1 , 2 , @xmath37 nodes @xcite . thus in order to obtain the actual landau dispersion , it is not only necessary to solve the continuity conditions , but also to count subsequently the zeros of the energy eigenfunctions . the band structure of our sample structure
is depicted in fig .
[ fig : weak]a for two different magnetic field amplitudes .
when the barrier height is considerably greater than the fermi level , the coupling of the adjacent electron systems is weak and the shape of the dispersion in units of @xmath0 and @xmath77 varies only slightly in dependence of the magnetic field .
a certain anticrossing is resolved in fig .
[ fig : weak]b .
an increase of the magnetic field by 1 t shifts the band gap by @xmath78 .
simultaneously , the gap is broadened from @xmath79 to @xmath80 .
generally , the size of the band gaps can not be resolved within our experiment , but their position is an accessible quantity . by changing the fermi level in an appropriate way ,
it is possible to trace the position of the anticrossings while the magnetic field is varied . on large length and energy scales , the approximation of ho @xcite reproduces the actual landau band structure very well ( fig . [ fig : weak]a ) .
however , in comparison to the complete quantum mechanical calculation , the position of the anticrossings is about @xmath81 higher and the increase of the gap positions as a function of @xmath82 is about twice as large as for systems where the coupling is switched on ( fig .
[ fig : weak]b ) .
all features of the magnetotransport curves shown in fig . [ fig : curves]b are explainable within the scope of two different models which have been developed recently .
while each describes an ideal situation , the actual sample structure possesses essential attributes of both .
the model of _ landau level mixing _
@xcite is based on a tunneling barrier which is invariant in the @xmath34-direction .
this property is also a prerequisite for our band structure calculation .
in contrast , the second model does not explicitly consider the dispersion of the landau levels .
it is rather based on the assumption that there exist _ tunneling centers _ within the potential barrier @xcite .
the model is especially valuable for the understanding of quantum interference patterns at certain regions of the conductance traces .
the formation of the landau oscillations is illustrated in figs .
[ fig : tun]a and b. the coupling of both electron systems is determined by the degree of mixing between the edge states on the left and right side of the tunneling barrier .
the coupling becomes maximum at the band extrema near the anticrossings .
figure [ fig : weak]a depicts the configuration of the edge channels for two different fermi levels . for @xmath83 , there exist two pairs of opposite channels containing electrons which counterpropagate along the barrier .
when the fermi level @xmath84 increases and exceeds an integer value , two additional edge channels emerge at all sample edges as well as parallel to the tunneling barrier .
however , if the fermi level increases further and enters one of the landau band gaps , two edge states disappear again at each anticrossing while the corresponding channels still exist at the other edges of the coupled quantum hall systems . from this reason and due to the fact that the chirality of the system suppresses backscattering into the contacts ,
the electrons in the remaining parts of the edge channels are supposed to tunnel immediately through the endings of the barrier @xcite . if the fermi level lies between the @xmath76-th and @xmath85-th landau band ( @xmath36 , 1 , 2 , @xmath37 ) and if the system is spin - degenerate , this effect gives rise to a conductance of @xmath86 .
the authors of ref .
@xcite show in the scope of a tight - binding model that the tunneling of the electrons is not strictly limited to the outermost regions of the barrier .
in fact , their simulation demonstrates that the electrons may propagate along the barrier for a certain distance until all of them have successively tunneled through the potential wall .
while both anticrossings around @xmath87 in fig .
[ fig : weak]b are located at the same energy level due to symmetry , this does generally not hold for energy gaps between higher order bands .
however , in weakly coupled electron systems , multiple band gaps of same order overlap with each other so that they are not distinguishable by magnetotransport measurements @xcite .
the existence of tunneling centers within the barrier is the origin for quantum interferences between counterpropagating edge channels [ 9 - 12 ] .
impurities within the barrier enhance the tunneling probability locally .
the same holds for a reduced aluminum content in the barrier which varies in the 18 monolayers thick algaas layer due to statistical fluctuations .
the illustration of fig .
[ fig : tun]c shows that for each pair of tunneling centers there exist two different sets of trajectories which both enclose the same magnetic flux @xmath88 .
the aharonov - bohm ( ab ) effect is possible as an electron at point a may interfere with itself at point b by taking the paths 1 or 2 which have the lengths @xmath89 and @xmath90 , respectively .
constructive interference takes place if the enclosed magnetic flux is a multiple of the dirac flux quantum @xmath91 .
therefore , the period of the conductance fluctuations is given by @xmath92 for the distance @xmath90 between the interfering edge channels we use in the following the difference of the corresponding centers of mass on the @xmath30-axis .
generally , the quantity @xmath90 can exceed the barrier width by a multiple . in conventional ab rings ,
the interfering trajectories are fixed within the current paths which are defined by means of lithography @xcite .
the course of the trajectories in our samples is , however , not only determined by the underlying structure , but also by the magnetic field so that variations of the magnetic length are relevant for the enclosed magnetic flux . for two interfering edge states at some energy and with guiding centers at @xmath93 , the length @xmath90 can be estimated to @xcite @xmath94 where the cyclotron radius is given by @xmath95 .
the last term of eq .
is valid for a flat density of states as discussed in sec .
[ sec : basic ]
. a more accurate specification of @xmath90 becomes possible by explicitly using the results of the band structure calculations .
for the evaluation of @xmath96 , there exists an analytic expression for the antiderivative appearing in the corresponding piecewise defined integral @xcite .
the interference of the electrons is perturbed by thermal fluctuations .
depending on temperature @xmath97 and the distance @xmath98 between the involved tunneling centers , one can distinguish between a coherent and incoherent regime @xcite .
an electron with the velocity @xmath99 takes the time @xmath100 to travel along pathway 1 outlined in fig .
[ fig : tun]c . according to the energy - time uncertainty principle ,
the average fluctuation of the particle energy is then @xmath101 . for an electron within the coherent regime
, @xmath102 has to be considerably greater than the average amplitude of the thermal fluctuations .
the corresponding relation @xmath103 leads to the definition of a critical temperature which specifies the transition between the coherent and incoherent regimes : @xmath104 the last term is based on the fact that the particle momentum is given by @xmath105 @xcite . as the landau dispersion is widely invariant in respect of the magnetic field ( if plotted against @xmath106 as in fig .
[ fig : weak ] ) , the quantity @xmath107 is approximately constant and for the coherent regime it holds @xmath108 because a real sample likely contains not just two , but several tunneling centers , there exist many competing interference paths .
therefore , one expects quasiperiodic conductance fluctuations instead of sinusoidal oscillations .
the coherence condition of eq .
depends on the temperature as well as on the magnetic field strength .
both parameters influence the contribution of the particular tunneling centers to the conductance of the whole system . for a decreasing temperature or a rising magnetic field
, an increasing number of different interference paths affects the conductance .
however , the @xmath82-dependence can not be investigated as the observed ab oscillations are restricted to just one conductance peak @xcite .
the comparison of the magnetotransport traces of fig .
[ fig : curves]b with the corresponding landau dispersions requires the knowledge of the exact charge carrier concentration in the coupled electron systems .
our sample structure does not , however , allow a direct measurement of this quantity as each electron system is effectively accessible just by one contact .
nevertheless , there exist two methods which offer the possibility to determine the electron density . on one hand , the electron concentration can be predicted by means of a capacitor model and on the other hand it is possible to infer it from magnetotransport measurements similar to the analysis of conventional shubnikov - de haas oscillations .
the gate element of the heterostructure in fig .
[ fig : intro]a can be modeled by a plate capacitor . the dielectric is characterized by both the relative permittivity @xmath109 of al@xmath110ga@xmath111as at low temperatures @xcite and the width @xmath112 of the potential barrier . from the corresponding capacitance @xmath113 it follows
@xmath114 the voltage @xmath115 is a sample - specific offset which is determinable just by experiment . because eq . only predicts the slope of @xmath116 , a measurement method is still indispensable for obtaining an absolute value for the electron density .
alternatively to the capacitor model , the electron concentration can be inferred from the resistance traces @xmath117 shown in fig .
[ fig : curves]b .
this method exploits the fact that the anticrossings of the landau dispersion in fig .
[ fig : weak]a are equidistant . figure [ fig : density]a confirms this feature by plotting the gap position versus the index of the particular lower landau band .
the following two equations reproduce the linear relation between the band index and the energy level of the anticrossings : [ eqn : shift - ac ] @xmath118 all energy gaps are shifted about @xmath119 with respect to the corresponding bulk landau levels .
the collective magnetic shift of @xmath120 reproduces the increase @xmath78 which is visible in fig .
[ fig : weak]b . the field dependence contained in the particular second terms of eq .
is one order of magnitude weaker than the collective shift , but it increases linearly with the band index . for the average field strength of 3 t of the conductance traces in fig .
[ fig : curves]b , the distance of the equally spaced band gaps is approximately @xmath121 .
the spacing of the anticrossings is thus almost identical to the landau splitting of bulk states .
the filling factor results from the electron density according to @xmath122 . in order to use @xmath67 as a natural comparative quantity for the dimensionless energy eigenvalue @xmath57 ,
it is still necessary to know the density of states which actually determines the fermi level .
the authors of refs .
@xcite and @xcite state an electron mobility of @xmath123 for their electron films which reside like ours along the cleavage plane of a ceo sample . for the analysis of magnetotunneling spectroscopy experiments
, they assume a strong broadening of the landau levels and employ a flat density of states . the comparison between the conductance traces of fig .
[ fig : curves]b and those of refs .
@xcite and @xcite makes it reasonable to use here the same asymptotic approximation @xmath124 .
the distance @xmath125 of the anticrossings then corresponds to a period of @xmath126 on the scale of the filling factor .
if @xmath127 denotes the mean spacing of the resistance minima on the axis of the reciprocal magnetic field , the two - dimensional electron density can be calculated according to @xmath128 a detailed error discussion in ref .
@xcite yields that the neglect of the small field dependence contained in @xmath125 does not considerably influence the results which are presented in the following on the basis of this relation . figure [ fig : density]b shows the electron density obtained from the minima and maxima of the resistance traces of fig .
[ fig : curves]b .
the data points for intermediate values of the gate voltage are gained from other measurements which are not shown here .
the electron density increases linearly with @xmath129 until the accumulation of electrons begins to saturate at v. for gate voltages below this value , the slope of the data points is @xmath130 with an uncertainty of 5 .
thus , the determination of the electron density according to eq .
corresponds very well to eq . , namely , to the capacitor model
this agreement also confirms that the ( enhanced ) zeeman splitting is not resolved with our samples . with an increasing gate voltage , at some point
a notable leakage current @xmath131 occurs through the algaas layer of the gate structure . for the experiment , it is not so much the leakage current which is relevant , but rather an accompanying bias voltage which builds up as a consequence of the current flow .
electrons are tunneling along the gradient of the gate voltage from the electron films into the @xmath10-layer of the gate electrode . while in the lower electron system the charge carriers are easily replaced by electrons from the source contact ( fig .
[ fig : intro]a ) , the upper system depletes at some degree because the electron supply from the common ground is restricted by the @xmath4 thick tunneling barrier .
therefore , an _
internal _ bias voltage emerges which is approximately proportional to the leakage current ( fig .
[ fig : density]b ) .
this bias shifts the electron systems against each other .
the consequences are the same as discussed in refs .
@xcite and @xcite for an applied _ external _ bias voltage .
the following quantitative comparison of the measurement results and the landau dispersion is restricted to gate voltages @xmath132 . on that condition
, a distortion of the conductance traces due to an internal bias can be excluded as the offset between both electron systems is then negligible small : @xmath133 .
the two - point conductance is plotted in fig .
[ fig : enhanced]a as a function of the filling factor . since the electron transport takes place in ( coupled ) one - dimensional edge channels
, it is advantageous to calibrate the ordinate in units of the conductance quantum @xmath134 . as stated before in sec .
[ sec : tun ] , one expects for the spin - degenerate system that the first conductance peak is limited to @xmath135 . however , the peaks at @xmath136 in fig .
[ fig : enhanced]a have an amplitude on the order of @xmath137 .
in contrast , kang .
report on conductance oscillations which reach a height of only about @xmath138 @xcite .
although the origin of the low current is not resolved , it is much more easier to comprehend a peak amplitude which falls below the twofold conductance quantum than if this threshold is exceeded . the tight - binding model of ref .
@xcite shows that the conductance stays below @xmath139 if the relative hopping amplitude is @xmath140 .
our result @xmath141 contradicts not only this conclusion , but also the theory of ref .
@xcite which predicts @xmath142 with @xmath143 for two spin - polarized quantum hall systems .
thus , the enhancement of the conductance in our ceo structure is probably based on features which are not considered in both models .
macroscopic defects of the cleavage plane are able to separate the @xmath11 long quantum region into several independent sections .
the extended @xmath10-layers connect these multiple active regions in parallel as illustrated in fig .
[ fig : enhanced]b .
the main reason for the fragmentation of the quantum region is the local injury of the upper @xmath144$]-ridge during the sample fabrication and preparation .
the most likely origin are steps and corrugations which emerge during the cleavage process .
the thin electron system can already be divided by steps which have a height on the order of the 2des thickness , namely , about @xmath145 .
another source for a partial damage of the electron systems consists in the manipulations during the preparation process .
the disruption of the quantum region hardly appears if the contacts are fabricated by means of photolithography on the cleavage plane as done by kang .
in contrast to extended contact layers as used here , small contacts access just one or very few independent sections of the active region .
the periodic coincidence of the fermi level with the landau band gaps is responsible for the sequence of conductance peaks shown in fig .
[ fig : enhanced]a . by varying the electron density , it is possible to measure the position of a certain anticrossing in dependence of the magnetic field . for this purpose ,
the conductance traces @xmath146 have been recorded for different gate voltages .
it is a consequence of the unresolved spin - splitting that the distance of the conductance peaks on the scale of the filling factor is about 2 instead of 1 . from the land factor @xmath147 of gaas
@xcite it follows for the ratio between the zeeman and landau splitting @xmath148 the high electron density of the coupled 2dess makes it necessary to consider additionally the influence of electron - electron interactions on the spin - splitting @xcite .
the corresponding exchange energy enhances the land factor which then has depending on the actual sample a value between @xmath149 and 7 @xcite . nevertheless , in similarity to refs .
@xcite and @xcite , the conductance traces of our sample reveal no clear spin - dependent features . according to the experimental results presented by kang . in fig . 1 of ref
. @xcite , it is by all means possible that each electron system shows shubnikov - de haas oscillations with a transition between spin - degenerate and -resolved landau levels while this feature is entirely absent in the magnetotunneling data @xcite .
.5 in the expected magnetic shift of the anticrossings is small if one considers the energy scale not absolutely , but in units of the cyclotron energy .
according to the band structure of fig .
[ fig : weak]b , the band gaps are shifted by @xmath150 .
likewise , the conductance traces of fig .
[ fig : enhanced]a have peaks with positions which depend on the magnetic field .
the peaks are shifted for an increasing electron density towards higher filling factors . for @xmath151
, they move to lower values again . in the following , a basis filling factor @xmath152 , 4 , 6 , @xmath37 is assigned to the particular peaks of each conductance trace .
the difference between the actual peak position and @xmath153 is plotted in fig .
[ fig : shift ] . for a constant basis filling factor
@xmath153 , the magnetic field increases in this graph from the left to the right .
in order to keep the fermi level within the same band gap , it is necessary to increase the electron density for compensating an enhanced degeneracy factor @xmath154 . the graphical comparison in fig .
[ fig : shift ] shows that the conductance peaks evolve in a very similar way for different basis filling factors .
the magnetic shift of the conductance peaks is compensated by another effect which emerges for gate voltages greater than v. the corresponding decrease of the peak positions is not correlated with the magnetic field , but is a consequence of the internal bias voltage which builds up due to the gate leakage current .
the potential offset between both electron films involves a lowering of the band gaps with respect to the electron system next to the cathode @xcite .
this effect shifts the conductance peaks towards lower filling factors and has been revealed in ref .
@xcite by applying an external bias voltage .
our sample structure exhibits a very similar behavior for the internal bias .
the pronounced boundary between rising and falling peak positions at @xmath155 is a consequence of the nearly exponential relation between the gate and bias voltages ( fig .
[ fig : density]c ) . because the charge carrier density of both electron systems changes simultaneously with @xmath156 , the spectroscopy of a certain band gap for different internal bias voltages requires an adjustment of the magnetic field .
hence , the displacement of the conductance peaks can not be investigated independently from the magnetic shift of the landau gaps .
nevertheless , a quantitative comparison with the results of kang . , who applied an external bias , reveals similar peak shifts . for a gate voltage which increases from 0.7 to v , the position of the @xmath157 peak diminishes in fig .
[ fig : shift ] by @xmath158 . at the same time , the internal bias rises according to fig . [ fig : density]b by @xmath159 . for this voltage offset , the first conductance peak in fig . 2 of ref .
@xcite decreases from @xmath160 to 1.04 .
thus , the shift @xmath161 obtained for an external bias is comparable to the displacement encountered for an internal bias of the same amplitude .
.5 in .53@cccccc ' '' '' @xmath162&@xmath163 & @xmath82&@xmath164&@xmath165[0pt]{$\scriptscriptstyle\bigtriangleup$}}$]&@xmath166 + @xmath167&@xmath168&@xmath169 & & & + 0.4&1.16&1.15&4.13&2.062&2.072 + 0.5&1.89&1.84&4.24&2.078&2.091 + 0.6&2.49&2.25&4.58&2.086&2.100 + 0.7&2.95&2.56&4.77&2.092&2.107 + 0.8&3.18&2.84&4.63&2.097&2.113 + all conductance peaks in fig .
[ fig : shift ] are shifted towards higher filling factors as the gate voltage increases up to v. due to the cropping of the conductance peak at @xmath152 ( this feature will be discussed in sec .
[ sec : ab ] ) , the following quantitative analysis is focused on the @xmath157 peak which can be localized much more accurately .
the band gap which is responsible for this conductance peak is twofold , more precisely , it consists of two anticrossings at opposed wave numbers ( fig .
[ fig : weak ] ) .
table [ tab : cmp ] compares the positions of the @xmath157 conductance peak for different magnetic fields with the result of corresponding band structure calculations .
the relevant band gap is characterized either by its extension @xmath170[0pt]{$\scriptscriptstyle\bigtriangleup$}}}};{\ensuremath{{\varepsilon}_\text{2}^{\scriptscriptstyle\bigtriangledown}}}]$ ] or its center @xmath171[0pt]{$\scriptscriptstyle\bigtriangleup$}}}}+{\ensuremath{{\varepsilon}_\text{2}^{\scriptscriptstyle\bigtriangledown}}})/2 $ ] and width @xmath172[0pt]{$\scriptscriptstyle\bigtriangleup$}}}})$ ] . for an increase of the gate voltage from 0.4 to v , the magnetic field
is raised from 1.15 to t in order to compensate for the enhanced electron density and to keep the fermi level between the landau bands @xmath173 and @xmath174 .
this shift corresponds to a displacement of the conductance peak by @xmath175 on the scale of the filling factor . according to the band structure calculation
, the higher magnetic field raises the energy gap by @xmath176 .
the conversion to the filling factor doubles this value : @xmath177 .
for another data set , which is not shown here , we get @xmath178 and @xmath179 , respectively .
the ratio between the measured and predicted shift of the conductance peak is @xmath180 .
the discrepancy between both values is not yet completely resolved , but it is probably a consequence of the coulomb interaction which is not considered in the calculated landau dispersion .
the influence of this many body effect on the band gap size is investigated in refs .
@xcite and @xcite .
both author groups arrive at the result that the electron - electron interaction enhances the energy gaps by a factor of about two .
the absolute position of the band gaps is , however , not explicitly determined there .
generally , the opposite edge channels are converging with an increasing magnetic field according to @xmath181 , with the consequence that the total coulomb energy rises .
this effect is most likely responsible for the encountered enhancement of the conductance peak displacement .
the landau dispersion as discussed so far is valid for a perfect tunneling barrier which is invariant in the @xmath34-direction . within a real barrier ,
however , there exist imperfections which lead to a local enhancement of the tunneling probability .
these tunneling centers , which are considered in the model of kim and fradkin @xcite , give rise to aharonov - bohm ( ab ) interferences between the opposite edge channels along the barrier @xcite .
in addition , the point contacts are responsible for conductance fluctuations observed at low magnetic fields @xcite . the latter effect which appears before the formation of distinct quantum hall edge channels will be discussed in sec .
[ sec : fluct_low ] . the rightmost resistance minimum of the curves in fig .
[ fig : curves]b shows some irregularities which are most pronounced for @xmath182 and disappear at higher gate voltages . figure [ fig : max]a gives a magnification of this feature .
the left conductance peak is substantially cropped and the peak at @xmath183 also exhibits slight irregularities . in general , there exist two different types of oscillations which appear simultaneously and exclusively at the first conductance peak on the scale of the filling factor : short - period and quasiperiodic ab oscillations as well as long - period conductance fluctuations .
a visual differentiation between both effects is possible by the black curves in figs .
[ fig : max]a and b. the two measurements differ only with respect to the sweep velocity of the magnetic field . a high value for @xmath184 in the right plot averages the ab oscillations out , whereas in fig .
[ fig : max]a both oscillation types are visible .
the two kinds of conductance fluctuations are not simply superimposed to the long - range landau oscillations , but they come along with a cropping and flattening of the respective conductance peak . while fig . [
fig : max ] reveals a clear distortion of the conductance peak at @xmath185 , the same maximum is in fig .
1(a ) of ref .
@xcite for @xmath186 entirely smooth .
however , in ref .
@xcite also a sample of kang
. shows a strong cropping of the conductance peak at a slightly lower temperature of @xmath187 .
the phase coherence of the interfering electrons is harmed by both thermal fluctuations and electrostatic potential gradients . the critical temperature given by eq .
separates the coherent and incoherent regime .
the calculation of the band structure for the mean magnetic field t of the first conductance peak yields @xmath188 for the two involved opposite edge channels ( inset of fig .
[ fig : fft]b ) .
in anticipation of the mean distance of the tunneling centers , @xmath189 , the critical temperature is then @xmath190 .
this result is insofar plausible as the ab oscillations of fig .
[ fig : max]a are almost suppressed at @xmath191 ( @xmath192 ) while in ref .
@xcite they exhibit at @xmath193 an amplitude which is more easily recognizable ( @xmath194 ) .
each phase - destroying effect which appears in addition to thermal fluctuations causes that some tunneling paths , first of all those with point contacts at a great distance , lose their coherence .
thereby , the shape of the conductance fluctuations changes since they are composed of all coherent ab oscillations .
the quantum interferences are influenced by the internal voltage in a similar way as by thermal fluctuations . due to the potential gradient across the barrier
, a tunneling electron possesses an excess energy with respect to the fermi level of the destination system . the bigger this difference , the more free states are available around the final wave number , and the more electron - electron scattering processes cause a loss of the phase information @xcite .
analogous to the thermal disorder , this effect is negligible as long as @xmath195 holds .
the energy equivalent to the critical temperature is @xmath196 .
consequently , the ab oscillations are expected to be essentially undisturbed until the internal bias voltage exceeds @xmath197 .
this threshold is reached in fig .
[ fig : density]b for a gate voltage of v. indeed , the conductance trace for @xmath198 shows in fig .
[ fig : max]a a significant indication of an initially emerging incoherence : in comparison to the data for @xmath199 , the right part of the conductance peak is already partially smoothed for @xmath200 .
the effect is further intensified at @xmath201 in that the cropping of the peak almost disappears while the ab signatures persist only at the curve shoulder around @xmath202 .
thus , the bias coherence condition is essentially consistent with the experimental data as well as this is the case for the temperature dependence .
although the amplitude of the ab oscillations shown in fig .
[ fig : fft]a is rather small , they are clearly revealed in the derivative of @xmath146 .
this is possible even for a high temperature of about @xmath203 which substantially exceeds the critical temperature of @xmath190 .
the conductance traces , which are composed of data points at a distance of @xmath204 , are smoothed before the differential quotient is calculated . for this purpose
, parabolas are fitted to the curve for each data point by considering altogether 12 points in the proximity of the central point .
the slope of the parabola at the abscissa of the middle sampling point is plotted in fig .
[ fig : fft]a .
the comparison of two successive field sweeps reveals that all fine structures of the conductance traces are reproducible to a very high degree .
it is possible to resolve half cycles with a minimum width of 5mt and an amplitude of at least @xmath205 .
the derivative @xmath206 exhibits a strong quasiperiodic character which gives evidence that there exist more than two tunneling centers within the barrier .
more precisely , this is an indication that several pairs of point contacts simultaneously fulfill the coherence conditions of eqs . and . the fourier spectrum of fig .
[ fig : fft]b is mainly concentrated within the two period length intervals 52@xmath207 and 106@xmath208 .
the comparison of the spectrum with the conductance trace @xmath146 in fig .
[ fig : fft]a suggests that the components of the second interval represent the second harmonic of the long - period conductance fluctuations .
the latter have a period of @xmath209 if the quantity is determined directly from @xmath146 .
the origin of these fluctuations will be discussed below . according to eq . , the distance of two tunneling centers which cause ab oscillations of a period @xmath210 is @xmath211 .
the distance @xmath90 of the involved edge channels can be determined either by the approximation formula of eq . or from a dedicated band structure calculation .
the wave functions shown in the inset of fig . [ fig : fft]b are located with their energy level next to the first band gap and have a distance of @xmath212 if the expectation value for the @xmath30-location is considered . for comparison , eq .
yields for the two states with @xmath213 an approximated distance of @xmath214 .
because the landau dispersion varies only slightly on the scales of the magnetic length and the cyclotron energy , the exact result @xmath215 is also a good approximation for system parameters which differ from those of fig .
[ fig : fft ] .
however , in units of the barrier width , the quantity @xmath90 may vary considerably : in our case ( @xmath216 , @xmath217 ) , the distance of the edge channels is @xmath218 while for the conditions of ref .
@xcite one gets @xmath219 .
generally , the distance of the opposite edge channels at a thin barrier is less determined by the barrier width , but rather by the extension @xmath220 of the wave functions .
the first interval of the fourier spectrum of fig .
[ fig : fft]b contains contributions of prominent ab oscillations with periods of @xmath221 , 63 , and @xmath222 .
if the distance between the interfering edge channels is @xmath223 , these values correspond to a distance of the tunneling centers of @xmath224 , 2.0 , and @xmath225 , respectively .
however , as the resolution of the fourier spectrum is limited to about 4mt due to the finite width of the conductance peak , reliable conclusions about the actual configuration of the tunneling centers are not reasonable , so that we simply state the average distance of the point contacts : @xmath226 .
if a pair of tunneling centers with a distance of @xmath98 belongs to the coherent regime according to eq . , the phase coherence length is at least @xmath227 which follows from fig .
[ fig : tun]c . thus , our experimental data yields @xmath228 .
the ab oscillations are superimposed by slowly varying conductance fluctuations which have according to fig .
[ fig : max]b a period length of about t. the authors of ref .
@xcite report a similar period of about t. the long - period conductance fluctuations occur in our samples for temperatures up to @xmath229 .
since this feature simultaneously emerges with the ab oscillations and disappears together with them when the internal bias increases ( fig .
[ fig : max]a ) , a connection to the quantum interferences at tunneling centers is obvious .
however , the theoretical work of ref .
@xcite gives no direct evidence for this effect .
nevertheless , from microstructured ab rings it is well - known that ab oscillations , which are periodic with respect to the flux quantum @xmath230 , simultaneously arise along with aperiodic fluctuations which have a higher amplitude as well as a considerably longer period @xcite .
the effect is attributed to the finite width of an ab ring @xcite .
when the electrons diffuse through the mesoscopic system , the corresponding trajectories lie from time to time nearer to the inner or outer radius of the ring - shaped electron path .
the corresponding fluctuations of the enclosed magnetic flux exist in addition to the continuous variation due to a rising or falling magnetic field .
the random shape of the electron trajectories within the conductor determines the phase difference @xmath231 between both pathways and is reflected by irregular fluctuations of the conductance .
the effect can be reduced or avoided if the area of the ab ring is very small compared to the area circumscribed by the whole ring path @xcite .
in contrast to conventional ab rings , the interfering electrons in our sample are not kept enclosed by a conductance path .
their location is rather determined by the magnetic confinement potential in combination with the shape of the conductance band . for a varying magnetic field ,
the enclosed flux @xmath88 changes by reason of two different effects : besides the proportionality with respect to the magnetic field , the latter also controls the flux indirectly via the distance @xmath90 of the interfering edge channels .
the field - dependence of this length comes from the landau dispersion and can be described according to eq . by the relation @xmath232 .
the parameter @xmath90 is a smooth function of the magnetic field only for an ideal system . in a real sample , the field - induced variation of the channel positions
is additionally influenced by the disorder potential .
the spatial shift of the edge channels is irregularly modified when the electrons at the fermi edge are forced by the disorder potential to propagate along a trajectory which deviates from that expected for a perfect rectangular potential wall . because of @xmath233 and @xmath234 , it is possible that small displacements of the edge channels cause a variation of the enclosed magnetic flux on the order of the flux quantum . however , due to the local character of disorder , the effect emerges in comparison to ab oscillations on considerably larger scales of the magnetic field .
the corresponding long - period and irregular conductance oscillations represent an abstract image of the disorder at the tunneling barrier .
the ab effect is restricted to the conductance peak which corresponds to the lowest landau band gap . already at the subsequent peak , the irregular oscillations are almost completely suppressed ( fig .
[ fig : max ] ) . though the ab signatures rapidly disappear from the conductance traces for a decreasing magnetic field , similar fluctuations emerge at low field strengths again ( fig .
[ fig : fluct]a ) .
the quasiperiodic oscillations start from zero field , get considerably stronger at about t and vanish at t when the regular landau oscillations arise . just like the ab oscillations
, the conductance feature in the low field regime is characterized by a quasiperiodic behavior and a high reproducibility .
furthermore , the fourier spectrum ( inset of fig [ fig : fluct]a ) extends over a similar range of period , namely , @xmath235@xmath236 . in spite of the resemblance between the conductance fluctuations at low magnetic fields and the quasiperiodic ab oscillations ,
there is evidence that the low field feature is not based upon the interference between counterpropagating edge channels .
the conductance traces of fig .
[ fig : fluct]b exhibit fluctuations which do not disappear for an increasing gate voltage as it is the case for the ab oscillations in fig .
[ fig : max]a .
this indicates that the quantum interferences do not take place immediately within the internal bias gradient , but they emerge apart from the tunneling barrier .
granted that the conductance feature at low magnetic fields was caused by the quantum interference of counterpropagating edge channels , one would expect according to eqs . and a period length of @xmath237 in contrast to the ab oscillations at the @xmath238 peak
, the field - dependence of the period length can not be neglected in the low field regime . though the conductance traces of fig .
[ fig : fluct]b fulfill @xmath239 at some degree , the absolute period of the conductance fluctuations does not conform to the theoretical prediction of eq .
: for the average distance @xmath240 of the tunneling centers and the field range 0.27t this formula yields a period of 8@xmath241 which substantially deviates from the fourier spectrum of the experimental data .
the conductance fluctuations at low magnetic fields originate most likely from the bulk of the 2dess .
recently , it has been shown with a cooled scanning probe microscope that electrons which are injected through a quantum point contact into a 2des flow at zero magnetic field along fan - shaped branches @xcite .
if just one transversal mode passes the contact , the electrons are restricted to a single branch . in our sample ,
the algaas barrier contains tunneling centers which are randomly distributed over the @xmath242 long potential wall . from their average distance of 2 m
, one may estimate the number of tunneling centers to be on the order of one hundred . in comparison ,
the 2dess with the electron density of @xmath243 at @xmath182 possess @xmath244 transversal modes @xcite . even if the momentum relaxation length @xmath245 fell below the phase coherence length @xmath246 by two orders of magnitude , the localization length @xmath247 still clearly exceeds the extension @xmath248 of the electron systems in the @xmath30-direction .
in addition , the short electron systems represent a single phase coherence unit , @xmath249 , and , therefore , belong to the regime of weak localization .
however , this effect does generally not cause conductance fluctuations in broad samples and would be anyway destroyed by low magnetic fields @xcite . as the imperfect potential barrier reduces the mode number of the total system by at least two orders of magnitude , strong localization ( @xmath250 ) becomes possible . in this regime , the electron branches which stem from the multiple slits in the tunneling barrier interfere and cause conductance fluctuations while the fermi level or the magnetic field are varied .
this effect vanishes at about t when all electrons which are injected by the tunneling centers are forced by the magnetic field to propagate within the same edge channels , cf .
@xcite , so that the fluctuating electron branches in the bulk of the electron systems do not exist any longer .
the conductance fluctuations in the low field regime allow
similar to the ab oscillations an estimation of the phase coherence length .
the starting point is the semiclassical condition @xmath251 which requires electrons to fulfill at least one complete cyclotron orbit .
the critical magnetic field @xmath252 is determined by the onset of significant oscillations which are defined here _ ad hoc _ as a ` dense sequence ' of fluctuations with an amplitude of minimally @xmath253 . for the conductance curve with @xmath182 in fig .
[ fig : fluct]b , this definition yields a phase coherence mobility of @xmath254 .
this quantity corresponds to a phase coherence length of @xmath255 which is of the same order as the value @xmath246 obtained from the ab oscillations . for an increasing gate voltage
, the momentum relaxation time rises due to a more efficient screening of remote ionized scatterers .
the corresponding increase of the phase coherence length @xmath256 @xcite is , indeed , confirmed by a decrease of @xmath252 in fig .
[ fig : fluct]b .
we have presented a gaas / algaas heterostructure which contains two laterally adjacent , field - induced quantum hall systems which are separated by a thin , epitaxially grown barrier . the structure includes a gate electrode in order to allow a detailed analysis of the landau band structure at the line junction for different magnetic fields . for the interpretation of the experimental data ,
the landau dispersion has been calculated exactly for non - interacting electrons residing in a 2des with a rectangular potential wall .
the conductance traces in dependence of the magnetic field are dominated by so - called landau oscillations which occur due to the periodic coincidence of the fermi level with the landau band gaps .
an observed enhancement of the conductance which exceeds the predictions according to the landauer - bttiker formalism is a consequence of macroscopic defects .
the latter divide the long and narrow quantum region into multiple independent sections which are connected in parallel by the extended @xmath10-contact layers .
our theory predicts for weakly coupled electron systems a magnetic shift of the anticrossings on the scale of the cyclotron energy of about @xmath257 .
the experiment confirms this effect semiquantitatively by a displacement of the conductance peaks on the scale of the filling factor for an increasing fermi level .
the conductance peak which corresponds to the gap between the first two landau bands is cropped and distorted by short- and long - period conductance fluctuations .
the short - period feature is a result of the ab interference between counterpropagating edge channels along the barrier .
the interference is made possible by tunneling centers which exist accidentally within the algaas barrier . for determining the distance of the involved tunneling centers from the fourier spectrum of the conductance traces ,
it is necessary to know the distance of the interfering edge channels . by using the result of a band structure calculation
, we obtain an average distance of the point contacts of about 2 m .
the observed ab oscillations basically fulfill two coherence conditions which describe the influence of the electron temperature and the internal bias voltage , respectively .
the bias results from gate leakage currents which lead to an offset between the two coupled electron systems .
long - period conductance fluctuations at the first conductance peak are a consequence of the disorder potential which distorts the edge channel positions in an irregular way for a varying magnetic field .
in addition to the interference effects at the cropped conductance peak , we observe conductance fluctuations which emerge at small magnetic fields and disappear at the onset of the regular landau oscillations . they occur in the regime of strong localization which is a consequence of an imperfect barrier .
the latter represents a multiple slit interferometer which essentially reduces the number of transversal modes in the whole system .
the phase coherence length determined from the conductance fluctuations , @xmath258 , is of the same order as @xmath246 which follows from the distance of the tunneling centers involved for the ab oscillations .
this work was supported financially by the _ deutsche forschungsgemeinschaft _ ( dfg ) via the priority program _
quanten - hall - systeme_. we would like to thank m. bichler and g. abstreiter as well as for sample growths .
m. h. is grateful to m. grayson for useful and motivating discussions throughout this project .
10 l. d. landau , z. phys . * 64 * , 629 ( 1930 ) .
b. i. halperin , phys .
b * 25 * , 2185 ( 1982 ) .
a. h. macdonald and p. steda , phys .
b * 29 * , 1616 ( 1984 ) .
k. v. klitzing , g. dorda , and m. pepper , phys .
* 45 * , 494 ( 1980 ) .
l. pfeiffer , k. w. west , h. l. stormer , j. p. eisenstein , k. w. baldwin , d. gershoni , and j. spector , appl .
. lett . * 56 * , 1697 ( 1990 ) .
m. grayson , m. huber , m. rother , w. biberacher , w. wegscheider , m. bichler , and g. abstreiter , physica e * 25 * , 212 ( 2004 ) .
+ m. huber , m. grayson , m. rother , w. biberacher , w. wegscheider , and g. abstreiter , phys .
94 * , 016805 ( 2005 ) .
l . ho , phys . rev .
b * 50 * , 4524 ( 1994 ) .
w. kang , h. l. stormer , l. n. pfeiffer , k. w. baldwin , and k. w. west , nature ( lond . )
* 403 * , 59 ( 2000 ) .
kim and e. fradkin , phys .
b * 67 * , 045317 ( 2003 ) ; phys . rev
. lett . * 91 * , 156801 ( 2003 ) . i. yang , w. kang , l. n. pfeiffer , k. w. baldwin , k. w. west , e .- a .
kim , and e. fradkin , phys .
b * 71 * , 113312 ( 2005 ) . m. habl , m. reinwald , w. wegscheider , m. bichler , and g. abstreiter , phys . rev .
b * 73 * , 205305 ( 2006 ) .
m. habl , ph.d .
thesis , universitt regensbug ( 2006 ) .
y. takagaki and k. h. ploog , phys .
b * 62 * , 3766 ( 2000 ) .
the complete algorithm , including the consideration of the effect of different effective masses in gaas and algaas , is to be published separately .
r. schuster , h. hajak , m. reinwald , w. wegscheider , g. schedelbeck , s. sedlmaier , m. stopa , s. birner , p. vogl , j. bauer , d. schuh , m. bichler , and g. abstreiter , phys .
( c ) * 1 * , 2028 ( 2004 ) .
t. feil , c. gerl , and w. wegscheider , phys .
b * 73 * , 125301 ( 2006 )
. + t. herrle , ph.d .
thesis , universitt regensburg ( 2007 ) .
i. barto and b. rosenstein , j. phys .
a : math . gen .
* 27 * , l53 ( 1994 ) ; phys .
( b ) * 193 * , 411 ( 1996 ) .
s. adachi , gaas and related materials ( world scientific , singapore , 1994 ) .
a. erdlyi ( ed . ) , higher transcendental functions ( mcgraw - hill , new york , 1953 ) .
+ j. spanier and k. b. oldham , an atlas of functions ( springer , berlin , 1987 ) .
+ w. magnus , f. oberhettinger , and r. p. soni , formulas and theorems for the special functions of mathematical physics ( springer , berlin , 1966 ) .
l. d. landau and e. m. lifshitz , quantum mechanics ( pergamon , oxford , 1965 ) .
r. a. webb , s. washburn , c. p. umbach , and r. b. laibowitz , phys .
lett . * 54 * , 2696 ( 1985 ) .
+ r. webb and s. washburn , phys . today
* 41 * ( 12 ) , 46 ( 1988 ) .
i. yang , w. kang , k. w. baldwin , l. n. pfeiffer , and k. w. west , phys .
92 * , 056802 ( 2004 ) . c. weisbuch and c. hermann , phys .
b * 15 * , 816 ( 1977 ) .
j. dempsey , b. y. gelfand , and b. i. halperin , phys .
lett . * 70 * , 3639 ( 1993 ) .
a. usher , r. j. nicholas , j. j. harris , and c. t. foxon , phys . rev .
b * 41 * , 1129 ( 1990 ) .
+ v. t. dolgopolov , a. a. shashkin , a. v. aristov , d. schmerek , w. hansen , j. p. kotthaus , and m. holland , phys .
lett . * 79 * , 729 ( 1997 ) . + t .- y .
huang , y .- m .
cheng , c .- t .
liang , g .- h .
kim , and j. y. leem , physica e * 12 * , 424 ( 2002 ) .
a. mitra and s. m. girvin , phys .
b * 64 * , 041309 ( 2001 ) .
m. kollar and s. sachdev , phys .
121304 ( 2002 ) . s. datta , electronic transport in mesoscopic systems ( cambridge university press , 1997 ) . c. p. umbach , s. washburn , r. b. laibowitz , and r. a. webb , phys .
b * 30 * , 4048 ( 1984 ) .
a. d. stone , phys .
* 54 * , 2692 ( 1985 ) .
m. a. topinka , b. j. leroy , s. e. j. shaw , e. j. heller , r. m. westervelt , k. d. maranowski , and a. c. gossard , science * 289 * , 2323 ( 2000 ) .
k. e. aidala , r. e. parrott , t. kramer , e. j. heller , r. m. westervelt , m. p. hanson , and a. c. gossard , nat . phys . * 3 * , 464 ( 2007 ) . | a quantum hall system which is divided into two laterally coupled subsystems by means of a tunneling barrier exhibits a complex landau level dispersion .
magnetotunneling spectroscopy is employed to investigate the small energy gaps which separate subsequent landau bands . the control on the fermi level permits to trace the anticrossings for varying magnetic fields .
the band structure calculation predicts a magnetic shift of the band gaps on the scale of the cyclotron energy .
this effect is confirmed experimentally by a displacement of the conductance peaks on the axis of the filling factor .
tunneling centers within the barrier are responsible for quantum interferences between opposite edge channels . due to the disorder potential ,
the corresponding aharonov - bohm interferometers generate additional long - period and irregular conductance features . in the regime of strong localization
, conductance fluctuations occur at small magnetic fields before the onset of the regular landau oscillations .
the landau dispersion is obtained by a dedicated algorithm which solves the schrdinger equation exactly for a single electron residing in a quantum hall system with an arbitrary unidirectional , threefold staircase potential . |
the statistical mechanics of dilute and semi - dilute solutions of interacting polymers may be greatly simplified by adopting a coarse - graining strategy for this complex many - body system , whereby the initial monomer or segment - level description is reduced to a `` soft - colloid '' representation involving only the centers of mass ( c.m . ) of the polymer coils .
this is formally achieved by tracing out the individual monomer degrees of freedom for fixed positions of the c.m.s of the polymers .
this reduction procedure results in effective interactions between the c.m.s which are state dependent , since they are in fact _ free _ energies of the `` fluid '' of monomers , for each configuration of the c.m.s of the @xmath4 interacting coils . in the low concentration limit
, this free energy reduces to a pairwise additive sum of effective potentials of mean force acting between the c.m.s of overlapping polymers , which depend only on temperature .
such pair potentials were first considered by flory and krigbaum @xcite , and later put on a firmer basis by grosberg _
@xcite , who showed that the effective pair potential between the c.m.s of linear self - avoiding walk ( saw ) polymers remains finite at full overlap in the scaling limit ( where the number of segments @xmath5 ) , and of the order of a few @xmath6 , reflecting the purely entropic origin of the effective interaction in the athermal saw model .
quantitative estimates of the effective c.m .
pair potential between two isolated saw polymers were obtained by monte carlo ( mc ) simulations of on- and off - lattice models and by renormalization group calculations @xcite .
these studies show that the zero concentration pair potential is purely repulsive and reasonably well represented by a gaussian of amplitude @xmath7 and width equal to the radius of gyration @xmath8 .
the saw model describes polymers in good solvent .
poor solvent conditions may be modeled by adding a finite attractive interaction between monomers and lowering the temperature .
mc simulations @xcite show that the resulting effective c.m .
pair potential becomes less repulsive as the temperature @xmath9 is lowered , and can become attractive in the vicinity of the @xmath1 temperature . when the polymer concentration is increased , three- and more - body effective interactions between the polymer c.m.s come into play .
these can also in principle be calculated by mc simulations of clusters of three or more polymers @xcite , but this systematic procedure becomes rapidly untractable , since the effective interactions depend on an increasing number of variables .
an alternative approach is to restrict the effective interactions between polymer c.m.s at finite concentrations to a sum of pair potentials , the latter being determined by an inversion procedure of the c.m .
pair distribution function based on integral equations for the pair structure of classical fluids @xcite .
the price to pay is that the resulting pair potentials now depend on polymer concentration , in addition to temperature .
the variation with concentration is moderate , but significant , in the case of the saw model for polymers in good solvent @xcite , and becomes very strong for polymers in poor solvent @xcite . since the effective pair potentials are obtained by inverting c.m .
pair correlation data from mc simulations , this strong variation with polymer concentration means that full monomer - level simulations must be carried out for each thermodynamic state of the polymer solution , thus partly defeating the purpose of the coarse - graining strategy .
it is hence important to explore numerically less demanding theoretical methods for the determination of the required polymer pair correlation functions .
one obvious candidate is the polymer reference interaction site model ( prism ) theory , which has proved very successful to determine the pair structure of interacting polymer solutions and melts @xcite .
prism grew out of the rism integral equation of chandler and andersen @xcite for the pair structure of simple molecular fluids .
the theory includes intramolecular connectivity via the macromolecular form factor , and provides accurate monomer - monomer pair distribution functions , @xmath10 .
the c.m .- c.m
. pair distribution function , @xmath11 , which needs to be inverted to extract the effective c.m .- c.m .
pair potential , @xmath3 , can be derived from @xmath10 by a simple , accurate relation which was established recently within the prism framework @xcite .
this strategy , which is highly economical on computational resources , is explicitly carried out in this paper , on the basis of the gaussian thread model for interacting polymers @xcite .
prism theory , in conjunction with an approximate closure relation , allows the monomer - monomer pair distribution function @xmath10 to be calculated analytically or numerically depending on the expression for the polymer form factor .
pair distribution function is then extracted from @xmath10 using the relation of ref .
@xcite together with appropriate form factors , and @xmath3 is finally obtained by hypernetted - chain ( hnc ) inversion @xcite .
the variation of the effective pair potential with temperature and polymer concentration is then examined in detail , and compared to the `` exact '' results extracted from mc simulations @xcite .
we consider solutions of @xmath4 linear polymers made up of @xmath12 freely jointed monomers or segments of diameter @xmath13 and spacing @xmath14 .
monomers belonging to the same or different chains are not allowed to overlap , i.e. they experience an infinite mutual repulsion for monomer - monomer distances @xmath15 , and , for @xmath16 , they attract each other via a yukawa interaction , @xmath17 where @xmath18 is the energy scale and @xmath19 the range of the monomer - monomer attraction @xcite .
the relevant thermodynamic variables are the temperature @xmath9 [ as usual , we define @xmath20 and the polymer concentration @xmath21 ( number of polymer coils per unit volume ) ; the corresponding monomer density is @xmath22 . as usual , the spatial extension of the chain will be characterized by its radius of gyration @xmath23 and the polymer overlap concentration , at which different chains start to overlap , is defined as @xmath24 . in this model of polymer solutions ,
the attractive interaction ( [ yukawa.eq ] ) represents solvent mediated monomer - monomer interactions in an effective way .
then , with respect to intermolecular interactions , large @xmath9 values correspond to good solvent conditions , while lower @xmath9 values mimic poor solvent conditions .
prism theory assumes all monomers to be equivalent regardless of their position along the polymer chain .
its aim is to calculate the site - averaged pair distribution function @xmath10 of monomers on different chains . to this end
one introduces the pair correlation function , @xmath25 , the direct correlation function @xmath26 , and their fourier transforms , respectively @xmath27 and @xmath28 , which are related by the following generalization of the ornstein - zernike ( oz ) equation @xcite , @xmath29,\ ] ] where @xmath30 is the form factor of the polymer chains . in the following
, it will be convenient to introduce the auxiliary quantities @xmath31 [ hence , @xmath32 and @xmath33 , such that eq .
( [ prismoz.eq ] ) may be rewritten as @xmath34.\ ] ] the oz equation , eq .
( [ prismozthread.eq ] ) , must be supplemented by a closure relation for @xmath35 , as well as by a prescription for @xmath36 .
while self - consistent schemes exist to compute the latter as a function of temperature and density , they are quite complex to implement @xcite . in the present work , we will make the simplifying assumption that the polymers obey gaussian statistics .
this approximation , which is in principle only valid in the melt or close to the @xmath1 temperature , is mainly needed for technical reasons , due to the lack of analytical results for non - gaussian polymers .
thus , the form factor @xmath36 is given by the well - known debye expression for ideal chains in the continuum limit , namely @xcite @xmath37,\ x = qr_g.\ ] ] replacing this exact form by its simple pad approximant , @xmath38 allows analytic results to be obtained within the thread limit @xcite , but we will see in the following that it can lead to significant problems in the present context .
in this work , we are interested in a mesoscopic level of description of polymer systems . for simplicity
, it is thus useful to get rid of the microscopic details of the polymer model .
a way of doing this in the prism framework is by taking the so - called thread limit . in the case of excluded volume interactions
only , it is defined by letting @xmath13 and @xmath14 go to zero and @xmath12 to infinity , keeping the ratio @xmath39 , @xmath8 and @xmath40 constant @xcite . in refs .
@xcite , the attractive part of the monomer - monomer interaction potential was left untouched when taking the thread limit .
however , for simple polymers , it turns out that the range of this attractive tail scales with the monomer size .
thus for this work it seemed physically appropriate to supplement the previous limiting process with the limit of a vanishing range of the attractive potential , namely @xmath41 , keeping the ratio @xmath42 constant . this can be interpreted as a special limit of the theory of refs .
@xcite , as shown in appendix [ thread.app ] . to escape the trivial case of non - interacting polymers , the thread limit requires a proper formulation of the closures of the prism - oz equation ( [ prismozthread.eq ] ) as well . in this work
, we have applied the so - called rmpy / hta closure , which has been found to be the most satisfactory in earlier work @xcite .
the details are given in appendix [ thread.app ] and we only quote here the equations relevant to our calculation .
the monomer - monomer correlations are thus given by @xmath43 with @xmath44 where @xmath45 is the @xmath1 temperature of the polymer model .
@xmath46 , which depends on @xmath47 only , is obtained by imposing that , in the presence of excluded volume interactions only ( i.e. for infinite temperature ) , @xmath48 obeys the hard core condition , @xmath49 .
it is plotted for both chain form factors , eq .
( [ debye.eq ] ) and ( [ omegaapp.eq ] ) , in fig . [ czero.fig ]
. density dependence of the direct correlation function @xmath46 , in the presence of excluded volume interactions only . for the exact form factor ( [ debye.eq ] )
, @xmath46 has to be numerically determined , while for the approximation ( [ omegaapp.eq ] ) , the analytic result ( [ czero.eq ] ) is plotted . ] with the above monomer pair correlation functions , effective pair potentials between polymer chains may now be determined .
we first consider in the next section the situation where the interaction site for the effective interaction between polymer coils is chosen to be a monomer ; subsequently effective interactions between polymer c.m.s are discussed in sec .
[ vcc.sec ] .
starting from the original polymer system , the choice of the effective coordinates used to specify the position of each polymer coil , and relative to which the coarse - graining procedure is performed , is largely arbitrary .
most of the time , the center of mass is considered , but it has sometimes been suggested to use specific monomers , like the midpoint or the endpoints , as in refs .
this is a very natural choice for star polymers as well , where the star center provides an obvious reference point @xcite . in this section
, we thus attempt to build such a monomer - monomer ( or site - site ) effective interaction . in the present prism - based approach , by construction
, no distinction is made between the different monomers on a chain .
however , one can build an effective monomer - monomer interaction based on the concept of an average monomer @xcite .
the idea goes as follows . on each chain
, a random monomer is tagged , and a coarse - grained polymer - polymer interaction is computed by tracing out , for fixed ( spatial ) positions of the tagged monomers , all the other monomer degrees of freedom . averaging
this free energy over the random locations of the tagged monomers on their chains then leads to a quantity which can be given the meaning of an effective interaction between average monomers . in practice
, an effective pair potential @xmath50 may be extracted by an inversion procedure , based on integral equations for the pair structure of simple fluids @xcite . for this purpose
, we consider a fluid of average monomers , of density @xmath21 ( each polymer chain is originally associated with a single tagged interaction site ) , the pair correlations of which are characterized by @xmath51 , where @xmath0 is the monomer - averaged correlation function determined in the previous section @xcite .
a direct correlation function @xmath52 may be extracted from @xmath53 via the oz relation for simple ( atomic ) fluids which in fourier space reads @xmath54.\ ] ] the effective interaction potential may finally be determined using one of the standard closures for simple fluids . since the effective interactions are expected to be relatively weak ( in particular , they contain no hard core repulsion ) , we follow earlier work @xcite , and adopt the very accurate hnc closure , which leads to the following expression for @xmath50 : @xmath55.\ ] ] with this closure , a fully analytic expression for @xmath56 can be obtained if the approximate expression ( [ omegaapp.eq ] ) for the chain form factor is used .
details are given in appendix [ anavaa ] .
we first consider the case of excluded volume interactions only , i.e. @xmath57 . the analytic @xmath56 is plotted in fig .
[ vaa.fig ] for various polymer densities .
it diverges at zero separation and , when the density increases , this divergence manifests itself at shorter distances and becomes steeper .
at the same time , the repulsive tail of the potential becomes longer - ranged , as can be clearly seen on the plot of @xmath58 , a quantity playing a crucial role for the calculation of the thermodynamic properties of fluid systems characterized by similar `` soft '' interaction potentials @xcite .
these results are in good qualitative agreement with the simulation data of bolhuis and louis for midpoint and endpoint effective pair potentials @xcite .
recently , much attention has been devoted to the development of analytic expressions for effective pair potentials between the centers of star polymers in solution @xcite . since a linear polymer , seen from one of its endpoints or from its midpoint , can be identified with a one- or two - arm star , respectively , it is interesting to compare those calculations with the present theory . since , according to eq .
( [ hzero.eq ] ) , @xmath59 vanishes linearly for @xmath60 , @xmath56 diverges like @xmath61 .
such a weak , logarithmic divergence is present in the star polymer models , where it has been derived using scaling arguments , but the prefactor , which should be given by the two - arm result in the limit of very long chains where end effects are expected to be negligible , appears overestimated by the present theory compared to scaling arguments ( which predict a prefactor equal to @xmath62 for a star with @xmath63 arms ) and renormalization group results @xcite .
this quantitative disagreement is not surprising , since the present calculation relies on the use of gaussian form factors in good solvent conditions .
however , a similar calculation done with an approximate form factor reproducing the correct asymptotic behavior of @xmath36 in good solvent still does not provide the correct value and this failure was in fact argued to be inherent to the thread prism approach @xcite . for large distances
, @xmath56 asymptotically takes a yukawa form @xcite , @xmath64 such a long range behavior was postulated for star polymers , on the basis of the blob picture of the star inner structure .
however , because the blob picture breaks down for small @xmath63 , it was advocated that this asymptotic behavior should not be valid for small functionalities , @xmath65 , and should be replaced by a gaussian tail , by analogy with the behavior of the effective pair potential between chain centers of mass @xcite .
the present calculation does not appear to support this choice .
finally , and maybe more importantly , the present calculation reveals that @xmath56 changes significantly with density .
for instance , the range of the asymptotic yukawa tail more than doubles between @xmath66 and @xmath67 .
if such a behavior also applies to star polymers , it could have a major impact on the calculation of the thermodynamic properties of these systems , in particular on their phase diagrams . in fig .
[ vaacomp.fig ] , the numerically determined @xmath68 for the exact debye form factor ( [ debye.eq ] ) is compared to the analytic result obtained with the pad approximant ( [ omegaapp.eq ] ) , at various polymer densities .
a thorough analysis shows that the differences between both are only quantitative .
both share the same @xmath61 divergence for small separations , and both display an asymptotic yukawa tail , which is of shorter range when ( [ debye.eq ] ) is used .
comparison of the effective pair potentials between average monomers [ multiplied by @xmath69 obtained with the exact and approximate polymer form factors , ( [ debye.eq ] ) and ( [ omegaapp.eq ] ) respectively , at various densities , in the presence of excluded volume interactions only . from left to right : @xmath70 . ]
we now turn to the effect of the attractive monomer - monomer interaction with changing temperature . in fig .
[ vaatemp.fig ] , the analytic @xmath56 calculated with the approximate chain form factor ( [ omegaapp.eq ] ) is plotted at density @xmath71 for various values of the temperature .
similar results are obtained at all densities , and with the exact debye form factor as well . the effective potential is seen to become less repulsive when the temperature is decreased from @xmath72 to @xmath45 , it vanishes identically at @xmath45 and grows more and more attractive if the temperature is further decreased .
this trend reverts when the `` spinodal '' line is approached , leading to non - monotonous , mostly repulsive effective potentials of increasing range . but a word of caution is needed here . indeed ,
as shown in appendix [ thread.app ] , at all finite temperatures , the hard core condition @xmath49 is no longer obeyed .
thus , the effective pair potential between average monomers does not diverge at full overlap as it should , if this condition was consistently enforced at all state points
. this might not be so important , since , as it has already been mentioned , the product @xmath58 is in general more relevant than the potential itself , and in this quantity the short range details of the potential are strongly reduced by the vanishing prefactor .
another source of concern is that it is not known to what extent the results obtained below @xmath45 when approaching the `` spinodal '' line are physically meaningful , since the corresponding remnant of a binodal line , which would properly identify the domain of existence of the low density phase , is not known in the present theory @xcite . for these reasons ,
the present results are only given for reference and would need to be validated , both by implementing alternative integral equation schemes or by investigating @xmath56 as a function of temperature by molecular simulation .
effective pair potentials between average monomers , obtained with the approximate polymer form factor ( [ omegaapp.eq ] ) , at various temperatures , for the reduced density @xmath71 . ] to summarize , we have shown in this section that integral equation theories for macromolecular systems , like the present thread prism theory , could be useful tools to compute monomer - based effective interaction potentials between polymers . at least in the case of excluded volume interactions only , they lead to very reasonable predictions for the effective interactions between average monomers on polymer chains , compared to the presently available simulation and theoretical results .
but , the concept of an average monomer , which is needed in the present theory because , at variance with star polymers , there is no obvious choice of a reference monomer , is rather artificial and in practice conceals significant technical difficulties .
indeed , this concept has a simple meaning only in the limit of very long chains , where all monomers can be considered as equivalent . in ref .
@xcite , the effective interactions between central monomers and between end monomers have been found to differ considerably , showing that strong finite length effects should be expected on effective interactions between average monomers .
thus , for a better control of these finite size effects , an alternative approach seems desirable , and it has been found that considering effective interactions between polymer c.m.s is appropriate in this respect .
as has been shown in recent monte carlo studies @xcite , an effective pair potential @xmath3 between the c.m.s of polymer chains may be extracted by an inversion procedure analogous to the one described in the previous section . to that purpose one
considers a simple fluid of density @xmath21 , the pair correlations of which are characterized by @xmath2 , the c.m .
correlation function of the polymer system .
the corresponding direct correlation function will be denoted @xmath73 .
correlation function @xmath2 is easily accessible in mc simulations for any polymer concentration , but can not be calculated directly from the original prism integral equation .
however an approximate relation between @xmath74 and @xmath75 was established in ref .
@xcite , which reads @xmath76 where @xmath36 is the previously discussed intramolecular monomer - monomer form factor , and @xmath77 is the c.m .- monomer form factor [ normalized such that @xmath78 .
this relation was shown to be quite accurate by comparing the resulting @xmath79 , or the corresponding c.m .
structure factor @xmath80 , to mc simulation data . to complete the results of ref .
@xcite , where only good solvent conditions were explored , and to illustrate the quality of the predictions of eq .
( [ magic.eq ] ) for @xmath81 , simulation results obtained for a lattice model slightly below its @xmath1 temperature are reported in fig .
[ magic.fig ] .
structure factor @xmath81 for @xmath82 saw polymers with monomer nearest - neighbor interaction @xmath83 on a simple cubic lattice , at @xmath84 and @xmath85 .
the `` exact '' structure factor from mc simulations is compared to the prediction of the approximation ( [ magic.eq ] ) using the `` exact '' monomer correlation function and form factors . ] however , to use eq .
( [ hnc.eq ] ) , we explicitly need the pair correlation function @xmath2 , i.e. we have to formulate the predictions of eq .
( [ magic.eq ] ) in real space , and not in reciprocal space . in fig .
[ magicinr.fig ] , we plot the simulation results of ref .
@xcite accordingly .
the agreement between the `` exact '' @xmath2 and its approximation computed from eq .
( [ magic.eq ] ) , using the `` exact '' monomer correlation function and form factors , is seen to be reasonably good , except at short distances where the results of the approximation lie well below the `` exact '' simulation data , and can even become smaller than @xmath86 at low densities , which is unphysical .
consequences of this underestimation of @xmath2 at full overlap , which is found under the conditions of fig .
[ magic.fig ] as well and appears to be systematic , are discussed below .
pair correlation functions @xmath2 for @xmath82 saw polymers on a simple cubic lattice . from bottom to top , @xmath87 .
the `` exact '' pair correlation functions are compared to the predictions of the approximation ( [ magic.eq ] ) using the `` exact '' monomer correlation functions and form factors . ] as mentioned in sec .
[ prism.sec ] , since there are no known analytic expressions for the two form factors appearing in eq .
( [ magic.eq ] ) for interacting polymers , we are led to use the expressions valid for gaussian polymer coils in the scaling limit @xcite , i.e. eq . ( [ debye.eq ] ) and @xmath88 where @xmath89 denotes the error function . in ref .
@xcite , it was shown that they yield qualitatively correct results for the calculation of @xmath79 , even in the good solvent regime .
the importance of using exact results is best seen when the high density limit of @xmath79 for @xmath90 is considered , which follows from eq .
( [ h.eq ] ) : @xmath91 indeed , while this ratio is always larger than @xmath86 if the exact form factors ( [ debye.eq ] ) and ( [ omegacm.eq ] ) are combined , it shows a region at small @xmath92 where it is smaller than @xmath86 if the approximation ( [ omegaapp.eq ] ) is used . in the latter case
, spurious singularities are thus encountered when eq .
( [ oz.eq ] ) is applied to compute @xmath93 at high densities . in this high density regime ,
an expression for the potential of mean force between the c.m.s of polymers in the melt has been derived by guenza , on the basis of the same thread prism theory as the one used here , in the case of excluded volume interactions only @xcite .
both approaches share similarities , like the fact that , to first order , the potential of mean force vanishes proportionally to @xmath94 .
however , since many aspects of these theories differ and since we are interested in the effective potential and not the potential of mean force , a detailed comparison was not attempted .
the outline of the present computation is thus the following .
first , the thread prism theory developed in sec . [ prism.sec ]
is used to obtain @xmath0
. then @xmath2 is derived by applying eq .
( [ magic.eq ] ) with form factors ( [ debye.eq ] ) and ( [ omegacm.eq ] ) , and is eventually inverted using eqs .
( [ oz.eq ] ) and ( [ hnc.eq ] ) ( with indices @xmath95 instead of @xmath19 ) to obtain @xmath96 .
we first consider the zero density limit .
when @xmath97 , the effective pair potential is simply the potential of mean force , i.e. @xmath98 $ ] . combining eqs .
( [ h.eq ] ) ( with @xmath99 ) and ( [ magic.eq ] ) , one finds @xmath100,\ ] ] where @xmath101 denotes a space convolution .
results for @xmath96 and @xmath102 are shown in fig .
[ vcctemp.fig ] for several temperatures . above @xmath45
, @xmath96 has a roughly gaussian shape , with an amplitude which decreases with temperature .
the potential vanishes identically for @xmath103 and becomes purely attractive below , retaining a roughly gaussian shape with a negative amplitude increasing in absolute value when @xmath9 is decreased .
these trends are globally in agreement with the behavior found in mc simulations @xcite , but quantitative differences are visible .
in the good solvent regime , for temperatures above approximately @xmath104 , it is impossible to compute @xmath96 .
this is due to the theory producing unphysical c.m .
pair distribution functions , which are negative in the vicinity of @xmath105 .
moreover , approaching this existence limit from below , the value of @xmath96 at full overlap is found to diverge , corresponding to the fact that @xmath106 goes to @xmath86 .
this behavior is at odds with all available simulation data and other theoretical approaches @xcite , showing that in good solvent the amplitude of the potential of mean force is always finite and has a maximum of about @xmath107 , a value reached for @xmath108 in the present theory .
this deficiency , which manifests itself throughout the high-@xmath9 , low-@xmath21 domain , is an obvious consequence of the underestimation of @xmath2 at full overlap , as illustrated in fig .
[ magicinr.fig ] : while the exact @xmath106 saturates at a value strictly greater than @xmath86 in the limit of infinite temperature , it is predicted by the present theory to go to @xmath109 in the same limit . at this point it is worth insisting on the fact that this problem appears to be associated with the use of eq .
( [ magic.eq ] ) , and is totally unrelated to the use of gaussian chain form factors to model good solvent conditions .
around and below @xmath45 , a more complex @xmath110 dependence of @xmath111 was found in the simulations of refs .
@xcite , the potential globally growing more and more negative , as found here , but retaining a repulsive component at short distances , while it is found strictly monotonous in the present calculation .
these two types of variations were also found in various theoretical approaches based on mean - field ideas @xcite , and in fact it is not clear what the generic behavior of @xmath96 is in poor solvents .
the available simulation data were indeed obtained for rather short chains , for which , on the scale of @xmath23 , the monomer hard cores are not of negligible size .
this could result in the observed short range repulsive contribution to @xmath96 , which would then be a finite size effect disappearing for longer chains @xcite . by comparison with mean - field theories ,
difficulties can in fact be anticipated with the present theory in the low temperature domain . indeed , in the zero density limit , following eq .
( [ meanfieldish.eq ] ) , @xmath96 is a simple function of the overlap between the monomer distributions around the centers of mass of two coils at distance @xmath110 .
however , mean - field theories built exclusively on this two polymer overlap are known to take into account only two - body effective monomer - monomer interactions .
this is for instance the case of the theory of flory and krigbaum @xcite , which can be recovered in a thread prism version by linearizing eq .
( [ meanfieldish.eq ] ) .
such approaches are valid in good solvent , since then pair interactions actually dominate , but they fail in @xmath45 and poor solvents , where inclusion of higher order interactions is required . by analogy
, this raises questions on the applicability of the present theory in the latter domain as well .
we now turn to the density dependence of the effective potentials derived within the present theory .
effective potentials obtained for various densities at @xmath112 are plotted in fig .
[ vcc.fig ] .
these results are typical of good solvent conditions .
starting from @xmath113 , where the potential is purely repulsive and nearly gaussian , @xmath3 first increases in amplitude and in range with @xmath47 , then , at a density which increases when the temperature approaches @xmath45 from above , the trend changes : the repulsion at full overlap decreases as @xmath47 is further increased , while the range of the effective potential still increases .
in addition , with increasing @xmath47 , a damped oscillatory tail , which is more clearly seen on plots of @xmath114 , progressively develops when the distance between the c.m.s of two coils greatly exceeds @xmath23 .
all these trends are consistent with the simulation data of refs .
the only qualitative difference is in the tail of the potential , which , within the resolution of the simulations , appears to be purely attractive .
the results of the present theory suggest that to unveil a possible oscillatory behavior , simulations of much larger bulk systems with very high statistical accuracy would be required to characterize the potential beyond its first minimum . despite the various concerns discussed above about the applicability of the present theory around and below the @xmath1 temperature ( lack of more - than - two - body effective monomer - monomer interactions , incomplete characterization of the domain of existence of the low density phase @xcite )
, we present the corresponding results for completeness .
when @xmath103 , the effective c.m .
potential is found to vanish identically at all densities .
typical results for the poor solvent domain , obtained for various densities at @xmath115 , are plotted in fig .
[ vcc2.fig ] . starting from @xmath113 , where the potential is negative and nearly gaussian ,
@xmath3 is first found to become more negative with increasing @xmath47 , while roughly retaining its shape .
then , upon approaching the `` spinodal '' instability line , at a density which increases when the temperature approaches @xmath45 from below , the value of the potential at full overlap starts to grow with @xmath47 and a repulsive tail progressively develops .
these tendencies are not corroborated by the presently available simulation data .
we have developed an integral equation approach to the computation of effective interactions between polymers in solution .
it is based on the thread limit of the prism theory , which is used directly for the calculation of effective interactions between average monomers , and in association with an approximate expression for the c.m .
pair distribution function , derived within the prism formalism as well , for the calculation of c.m .- c.m .
effective potentials . in the case of the effective interactions between average monomers , the predictions of the theory for polymers with excluded volume interactions only are in good qualitative agreement with the available simulation data .
moreover , we have shown that the theory lends itself to fully analytic calculations and compares rather favorably with other analytic theories like those developed for star polymers .
the effective c.m .- c.m .
interaction potentials derived from the present theory exhibit many of the trends observed in mc simulations .
the agreement with the latter is very satisfactory at intermediate temperature under good solvent conditions , i.e. @xmath90 . the theory indeed reproduces non - trivial features like the non - monotonic variation of the value of the potential at full overlap and the development of a long range tail when the density increases .
this agreement deteriorates at very high temperature , in particular for @xmath116 ( saw model ) , and around and below the @xmath1-temperature .
since the hnc closure used to compute the effective pair potential is exact for all practical purposes , when applied to `` soft - core '' systems which exhibit essentially `` mean - field fluid '' behavior @xcite , the observed discrepancies must be traced back to the other approximations made in the theory , namely the rmpy / hta closure used to compute @xmath0 in the thread limit , the use of form factors @xmath36 and @xmath77 valid for ideal rather than interacting chains , and eq .
( [ magic.eq ] ) relating @xmath79 to @xmath27 .
simulation data tend to show that the use of eq .
( [ magic.eq ] ) is the main cause of the breakdown of the theory at very high temperature , while there are indications that the present use of the rmpy / hta closure in the low temperature regime is not justified . in the present state of the theory ,
the use of ideal chain form factors is in fact not found to be a critical approximation . in conclusion ,
we have proposed a first step towards an integral equation theory of effective pair potentials between macromolecules in solution , which could bypass costly simulations for the determination of the c.m . pair correlation function .
the results appear very encouraging , since , while only the simplest ingredients have been combined in the present theory , and there is clearly room for improvement , it is already found to be able to capture various non - trivial behaviors observed in recent simulations .
it appears thus that integral equation theories could be in the future a valuable tool for the development of coarse - grained descriptions of complex polymeric systems .
the authors thank a.a .
louis for useful discussions and b.r . acknowledges financial support from the cole normale suprieure ( paris ) and schlumberger cambridge research .
in this appendix , we present our calculation of the monomer pair correlation function based on prism theory in the thread limit .
it is inspired by the work of chatterjee and schweizer @xcite ( see also ref .
@xcite for a related approach ) , a special asymptotic regime of which is obtained .
we start with eq .
( [ prismozthread.eq ] ) , involving quantities which all remain finite in the thread limit as defined in sec .
[ prism.sec ] .
the case of excluded volume interactions only is considered first , and then used as a reference system to include attractive interactions between the monomers in the closure relations .
considering first the excluded volume interaction only , a percus - yevick ( py)-like closure is used , in which @xmath117 ( @xmath118 is introduced here to make @xmath46 dimensionless ) , or equivalently in fourier space , @xmath119 this singular @xmath120 is required to account for the effect of excluded volume on intermolecular correlations in the limit of vanishing size of the hard core of the monomers in the thread limit .
the prism - oz relation ( [ prismozthread.eq ] ) then leads to eq .
( [ h.eq ] ) , with @xmath121 , and the value of @xmath46 is determined by enforcing the hard core condition @xmath49 . in general
, all the calculations have to be performed numerically .
this is the case in the present work when @xmath36 is chosen to be the debye function ( [ debye.eq ] ) .
however , fully analytic results can be obtained if the usual pad approximant of this function , eq .
( [ omegaapp.eq ] ) , is used .
then , one finds @xmath122 and @xmath123 where @xmath124 and @xmath125 .
we now introduce the attractive interaction between the monomers , which , for convenience , is taken of the yukawa form ( [ yukawa.eq ] ) . in ref .
@xcite , various closures of the prism - oz equation have been studied in the thread limit , for chains with monomers interacting through a hard core and an attractive tail . here
, we consider the reference molecular percus - yevick / high temperature approximation ( rmpy / hta ) @xcite , which has been found to be the most satisfactory in this earlier work . with the present notations ,
it reads @xmath126 where a renormalized attractive potential @xmath127 was introduced . @xmath128 and @xmath129 are respectively the renormalized monomer direct correlation function and the monomer pair distribution function of the reference system , the calculation of which is explained above . as mentioned in sec .
[ prism.sec ] , on physical grounds , we supplement the previously applied limiting process with the limit of a vanishing range of the attractive potential , namely @xmath41 , keeping the ratio @xmath42 constant . to take this supplementary limit in the rmpy / hta closure , we need to evaluate @xmath130 .
we assume that @xmath129 satisfies the hard - core condition , @xmath131 , and has a finite slope at the origin .
this hypothesis holds both for the analytic result , eq .
( [ hzero.eq ] ) , and when the exact debye form factor is used instead of its pad approximant ( see below ) .
then we can write the taylor expansion @xmath132 therefore @xmath133 where @xmath134 has been used .
now , taking the limit @xmath135 , we find @xmath136 since the contribution of the @xmath137 in ( [ betagv.eq ] ) goes to zero .
the rmpy / hta closure relation hence reduces to @xmath138 with the temperature @xmath45 given by @xmath139 [ remember that @xmath140 , as seen in eq . ( [ czero.eq ] ) ] .
the latter expression can be greatly simplified , using the fact that , for large @xmath92 , @xmath141 decays like @xmath142 ; this allows to show that @xmath143 which yields @xmath144 thus , we find that @xmath45 is the same for both choices of the polymer form factors , eq .
( [ debye.eq ] ) or ( [ omegaapp.eq ] ) , and is independent of the polymer density .
fourier transforming @xmath120 , defining @xmath145 as in eq .
( [ c.eq ] ) , and injecting the result in eq .
( [ prismozthread.eq ] ) , one obtains eq .
( [ h.eq ] ) .
as anticipated by its notation , @xmath45 is in fact the @xmath1 temperature of the present polymer model .
this can be first seen from eq .
( [ c2.eq ] ) , since one finds that @xmath120 , often interpreted as a renormalized two - body monomer - monomer interaction , vanishes for @xmath103 .
more evidently , this can be shown by considering the pressure of the polymer fluid .
following the `` compressibility route '' @xcite , the pressure is indeed easily calculated to be @xmath146d\rho ' \\ & = \rho - \left(1-\frac{\theta}{t}\right ) r_{g}^{3 } \int_{0}^{\rho } \rho ' \widetilde{c}_0(\rho ' ) d\rho ' , \end{split}\ ] ] and one sees that , for @xmath103 , the pressure is that of a perfect gas , meaning in particular that its second virial coefficient vanishes , which is the usual definition of the @xmath1 temperature . here
again , a fully analytic expression for @xmath0 can be obtained if the pad approximant of the chain form factor is used .
it retains the form ( [ hzero.eq ] ) , with @xmath147 unchanged and @xmath148 .
we show here that the present calculation is related to an asymptotic regime of the theory developed in refs .
@xcite , in which the range of the attractive tail was kept finite . to this purpose ,
the calculation of the pressure is considered . in refs .
@xcite , where the approximate chain form factor ( [ omegaapp.eq ] ) was used , the pressure was found to be given by ( see eq . ( 30 ) of ref .
@xcite ) @xmath149 } \right]\right),\end{gathered}\ ] ] where the notations of the present paper are used and @xmath14 and @xmath150 have been reexpressed as functions of @xmath12 , @xmath23 , and @xmath21 , using the relations @xmath151 and @xmath152 .
@xmath19 can be handled in a similar way , using the fact that @xmath153 , and all the variables which vanish in the thread limit as defined in sec .
[ prism.sec ] are then eliminated .
the only remaining singular quantity is @xmath12 , which should go to infinity . taking this limit in eq .
( [ pressure.eq ] ) , after a large @xmath12 expansion of the logarithm and denominators is performed , leads to @xmath154 which is precisely the result obtained by combining eqs .
( [ eost.eq ] ) and ( [ czero.eq ] ) .
this indicates that the present theory corresponds to an @xmath41 limit of the theory of chatterjee and schweizer .
the present polymer theory shows some limitations which are reviewed here .
first , in the presence of attractive interactions , the hard core condition @xmath49 is no longer enforced @xcite .
for instance , one finds for the analytic solution @xmath155 which is strictly larger than @xmath86 for all densities and finite temperatures .
such a feature is very common in this kind of thermodynamic perturbation type approximations , where the direct correlation function is completely specified for all values of @xmath110 .
we expect its impact to be modest on the outcome of the proposed coarse - graining procedure , since we will be interested in a description on a mesoscopic level , while this inaccuracy manifests itself on the microscopic , monomer level .
secondly , below the @xmath1 temperature , there is a domain where the integral equations exhibit a singularity , corresponding to the vanishing of the denominator of eq .
( [ h.eq ] ) .
this is in fact the remnant , when @xmath12 goes to infinity while taking the thread limit used in this work , of the finite @xmath12 spinodal instability found by chatterjee and schweizer @xcite , with a critical temperature equal to @xmath45 and an infinite critical value of @xmath156 .
the boundary of this low - temperature , high - density domain can be explicitly computed when the approximate expression of the form factor is used , leading to @xmath157,\ t<\theta.\ ] ] this means that , below the @xmath1 temperature , where polymer solutions are known to phase separate between polymer - rich and solvent - rich fluids , the proposed theory will only allow to investigate the latter phase .
a fully analytic expression for @xmath56 can be obtained if the approximate expression ( [ omegaapp.eq ] ) for the chain form factor is used .
indeed , the expression for @xmath158 is known ; solving eq .
( [ oz.eq ] ) for @xmath159 and identifying the result with eq .
( [ h.eq ] ) , one finds @xmath160 replacing @xmath36 by its expression ( [ omegaapp.eq ] ) , @xmath161 , and subsequently @xmath52 , are easily obtained .
two regimes are found , depending on the sign of @xmath145 . if @xmath162 , i.e. @xmath90 , one finds that @xmath52 is the difference between two yukawa terms , @xmath163 where @xmath164 if @xmath165 , i.e. @xmath166 , @xmath52 shows damped oscillations , @xmath167 where @xmath168 in this practical calculation , the order of the operations of monomer averaging and effective potential calculation has been inverted , compared to the description of the process given before . in the limit of very long chains where end effects are negligible , this should be harmless . however , there is no reason to think that in general these two operations commute and this could lead to subtle effects , for short chains for instance .
lang , a. ; likos , c.n . ; watzlawek , w. ; lwen , h. _ j. phys . : condens
. matter _ * 2000 * , _ 12 _ , 5087 ; louis , a.a . ;
bolhuis , p.g . ; hansen , j .-
p . _ phys .
e _ * 2000 * , _ 62 _ , 7961 ; louis , a.a .
* 2001 * , _ 359 _ , 939 ; likos , c.n . ;
hoffmann , n. ; lwen , h. ; louis , a.a .
_ j. phys . :
condens . matter _ * 2002 * , _ 14 _ , 7681 .
the impossibility to properly identify the domain of existence of the low density phase below @xmath45 is a consequence of the version of the thread prism theory considered in this work , which implements the limits @xmath171 , @xmath13 , @xmath14 , and @xmath41 in a strict sense .
as can be seen in appendix [ thread.app ] in the case of the spinodal line , this rejects the liquid branch to infinite polymer density and thus prevents a meaningful investigation of liquid - gas coexistence .
this was not the case in ref .
@xcite , where complete phase diagrams , including both spinodals and binodals , could be obtained . upon inversion of the low density simulation data of fig .
[ magic.fig ] for chains of length @xmath82 , the resulting effective c.m .- c.m .
pair potential does not show a repulsive short range component .
this tends to support the possibility that this repulsion is a short chain artifact , but a more systematic study in temperature and density would be necessary to reach a final conclusion . | we use the thread model for linear chains of interacting monomers , and the `` polymer reference interaction site model '' ( prism ) formalism to determine the monomer - monomer pair correlation function @xmath0 for dilute and semi - dilute polymer solutions , over a range of temperatures from very high ( where the chains behave as self - avoiding walks ) to below the @xmath1 temperature , where phase separation sets in .
an inversion procedure , based on the hnc integral equation , is used to extract the effective pair potential between `` average '' monomers on different chains . an accurate relation between @xmath0 , @xmath2 [ the pair correlation function between the polymer centers of mass ( c.m . ) ] , and the intramolecular form factors is then used to determine @xmath2 , and subsequently extract the effective c.m .- c.m .
pair potential @xmath3 by a similar inversion procedure .
@xmath3 depends on temperature and polymer concentration , and the predicted variations are in reasonable agreement with recent simulation data , except at very high temperatures , and below the @xmath1 temperature . |
we propose new connections between the pauli graphs @xcite , that encode the commutation relations of qudit observables , and prime number theory .
we already emphasized that the dedekind psi function @xmath8 ( with @xmath9 a prime ) is used to count the number of maximal commuting sets of the qudits @xcite and meets the riemann hypothesis ( rh ) at primorial numbers @xmath25 @xcite .
similarly , there exist striking connections between @xmath26 and the hardy - littlewood function @xmath27 for the goldbach distribution of prime pairs [ see sec . 3 for the definition of @xmath28 .
in particular , we observe that @xmath26 meets the so - called totally goldbach numbers at small @xmath3 s and that @xmath29 also meets rh at primorial numbers .
the euler constant @xmath30 , through the mertens formula @xmath31 , is an important ingredient of all the inequalities involved in this correspondence . in the rest of this section
, we report on the number theoretical _ coincidence _ between @xmath26 and the totally goldbach numbers at small @xmath3 s , as well on the already known theorem connecting @xmath26 and rh at primorial numbers @xmath18 . in section 2
, we explore in detail this coincidence by refering to the qudit pauli graphs .
then , in section 3 , we establish the connection between @xmath32 and rh at primorial numbers . in the discussion , the concept of a goldbach defect for encompassing the statements at low and high @xmath3 s is proposed .
let us start the exposition of our ideas with a few theorems and definitions .
goldbach s conjecture , formulated in 1742 , is that every even integer greater than @xmath33 is the sum of two primes .
to date it has been checked for @xmath3 up to @xmath34 @xcite . a pair ( @xmath35 ) of primes such that the even integer @xmath36 is called a golbach partition . * definition 1 : * a positive integer @xmath16 is _ totally goldbach _ if for all primes @xmath15 , with @xmath9 not dividing @xmath16 ( except when @xmath37 ) we have that @xmath17 is prime @xcite .
* theorem 1 : * the inequality @xmath10 is only satisfied at a totally goldbach number @xmath12 .
the only totally goldbach numbers not satisfying the inequality are @xmath38 and @xmath39 .
* proof : * the proof of theorem 1 follows from a combination of results in @xcite and @xcite ( for a more general setting , see @xcite )
. * definition 2 : * a positive integer @xmath16 is _ almost totally goldbach _ of index @xmath40 if for all primes @xmath15 , with @xmath9 not dividing @xmath16 ( except when @xmath37 ) we have that @xmath17 is prime with @xmath40 exceptions .
.almost totally goldbach numbers of index @xmath41 . [ cols=">,>,>",options="header " , ] [ tablecheck ] this proposed conjecture follows from numerical evidence as shown on table 4 . in theorem 5 , conjecture 1 is found to be equivalent to riemann hypothesis .
* proposition 2 : * for an even integer @xmath3 , the following inequality holds @xmath42 * proof : * one introduces the ratio @xmath43 that reads .
@xmath44 by splitting the product
as @xmath45 the ratio @xmath46 is rewritten as @xmath47 the right hand side product @xmath48 equals @xmath49 and the first one @xmath50 is lower than @xmath49 so that proposition 2 is satisfied . *
corollary 1 : * the primorial numbers @xmath18 are champion numbers , i.e. left to right maxima of the function @xmath51 .
one gets @xmath52 .
* proof : * according to @xcite , p. 90 , or @xcite ,
, the number of goldbach pairs is such that there exists a constant @xmath53 with @xmath54 .
using the same method than the one used in @xcite for the champions of @xmath55 one gets @xmath56 .
* theorem 5 : * conjecture 1 is equivalent to rh .
* proof : * if riemann hypothesis is not true , then according to theorem 4.2 of @xcite and the inequality ( [ 2ineqs ] ) there are infinitely many @xmath40 such @xmath57 , corresponding to as many violations of the hardy - littlewood inequality in conjecture 1 .
now we prove that conjecture 1 follows from riemann hypothesis .
it is known from theorem 3 ( b ) of @xcite that if rh is true then , for any @xmath58 , @xmath59 with @xmath60 the first tchebycheff function
. we introduce the new function @xmath61 and show below that , under rh , we have @xmath62 let us explicit the ratio @xmath63 where @xmath64 one immediately gets @xmath65 next , using the change of variables @xmath66 , one is interested with the range of values such that @xmath67 so that @xmath68 . replacing the sum by an integral at the right hand side of ( [ loggf ] ) , one obtains @xmath69_y^{\infty } \sim \frac{1}{y}\sim \frac{1}{x},\ ] ] at large @xmath70 . by (
* theorem 3 ( b ) ) this is dominated by @xmath71 hence , by ( [ loggf ] ) @xmath72 and the inequality ( [ gf ] ) follows .
then , by using the same reasoning that in @xcite , conjecture 1 follows from the inequality @xmath73 by observing that @xmath74 this completes the proof of theorem 5 .
it is useful to encapsulate some of the results obtained for goldbach pairs at small even dimensions and at primorial dimensions by introducing a measure that we denote the goldbach defect . *
definition 6 : * for a positive even integer @xmath3 , one defines a goldbach defect @xmath75 from the relation @xmath76 . * result 4 : * the champions of the function @xmath75 ( the numbers at which @xmath75 reaches a new record ) form the sequence @xmath77 the sequence @xmath78 is found to be a subset of the champions of the cototient function ( a051953 in sloane s encyclopedia ) . by its definition , sequence @xmath78 encompasses the even totally goldbach numbers , except for the number @xmath79 ( see definition 1 ) and the primorial numbers ( see conjecture 1 and corollary 1 )
. none of the ( small ) even dimensions in table 3 belongs to the sequence .
inumbers in @xmath78 satisfy @xmath80 in future work , the goldbach defect may turn to be useful for discussing the change in the symmetries of the pauli graphs .
the first not square free and not totally golbach dimension in the sequence @xmath78 is @xmath81 .
the cardinality of maximal commuting sets for this case decomposes as @xmath82 , i.e. @xmath83 . in conclusion , we found a cornerstone between the arithmetic of qudits , the goldbach pairs and riemann hypothesis .
it may be that the loss of coherence of a classical system , in comparison to a quantum system , crucially depends on its high dimensionality .
the inequalities of theorem 2 and 4 would serve to define an entropy .
mixtures of multiple qudit structures also have an attractive nested arithmetical and geometrical structure that is worthwhile to investigate in detail ( see table 1 in @xcite and @xcite ) . | let consider the pauli group @xmath0 with unitary quantum generators @xmath1 ( shift ) and @xmath2 ( clock ) acting on the vectors of the @xmath3-dimensional hilbert space via @xmath4 and @xmath5 , with @xmath6 .
it has been found that the number of maximal mutually commuting sets within @xmath7 is controlled by the dedekind psi function @xmath8 ( with @xmath9 a prime ) @xcite and that there exists a specific inequality @xmath10 , involving the euler constant @xmath11 , that is only satisfied at specific low dimensions @xmath12 . the set @xmath13 is closely related to the set @xmath14 of integers that are totally goldbach , i.e. that consist of all primes @xmath15 with @xmath9 not dividing @xmath16 and such that @xmath17 is prime @xcite . in the extreme high dimensional case , at primorial numbers @xmath18 ,
it is known that the inequality @xmath19 ( for every @xmath20 ) is equivalent to riemann hypothesis . introducing the hardy - littlewood function @xmath21 ( with @xmath22 the twin prime constant ) , that is used for estimating the number @xmath23 of goldbach pairs
, one shows that the new inequality @xmath24 is also equivalent to riemann hypothesis . in this paper
, these number theoretical properties are discusssed in the context of the qudit commutation structure . |
the canonical cold dark matter ( cdm ) model for the origin of structure in the universe provides a concise qualitative description of large - scale - structure data spanning several orders of magnitude in distance scales , but on closer inspection , there seem to be anomalies . for example , optical and infrared surveys seem to indicate more power on large scales relative to that on small scales ( peacock & dodds 1994 ) . a number of variations on the standard cdm model have been suggested which improve the agreement between some data and model predictions .
these include ( but are not limited to ) the introduction of a cosmological constant ( efstathiou , sutherland & maddox 1990 ; kofman , gnedin & bahcall 1993 ; krauss & turner 1995 ; ostriker & steinhardt 1995 ) or a lower hubble constant ( bartlett et al . 1994 ) .
however , a problem with discriminating between these various models from large - scale - structure measurements is that luminous matter may be ` biased ' relative to the mass distribution ( kaiser 1984 ) and this bias is likely to evolve with time and scale ( see , for example , matarrese et al .
further investigation of the bias of luminous objects is thus important in the recovery of the primordial power spectrum from large scale structure data . in this paper
, we investigate the implications of new data obtained from deep radio surveys : in particular , the vla _ first _ ( faint images of the radio sky at twenty centimeters ) survey ( becker , white & helfand 1995 ; white et al .
1996 ) .
the goal of _ first _ is to survey the 10,000 @xmath0 scheduled for inclusion in the sloan digital sky survey down to a flux - density threshold of 1 mjy . at present
, the survey covers almost 3,000 @xmath0 of sky where @xmath1 @xmath2 @xmath3 @xmath2 @xmath4 and @xmath5 @xmath2 @xmath6 @xmath2 @xmath7 .
this yields a catalog of @xmath8 sources , about one third of which are in double - lobed and multi - component sources .
the survey has been shown to be 95% complete down to 2 mjy and 80% complete to 1 mjy .
the mean redshift of the radio sources in the survey is at @xmath9 , as opposed to redshifts @xmath10 characteristic of sources in optical or infrared surveys , so the typical physical distances for fixed angular separations are larger .
therefore , clustering of _ first _ radio sources has the potential to probe the power spectrum on large scales and at earlier epochs . the first high - significance detection of an angular correlation function for radio sources was presented in cress et al .
( 1996 ) . using the cf - estimator proposed by landy and szalay ( 1993 ) on the first 1550 @xmath0 of the survey
, it was found that the correlation function between @xmath11 and @xmath12 is well - fit by a power law of the form @xmath13 where @xmath14 and @xmath15 .
results using a simpler estimator were also presented but when the survey was later expanded to include a new @xmath16 it became evident that the ls estimator was the more robust ( cress et al .
correlation - function measurements shown here are determined using the ls estimator in the expanded survey area .
the purpose of the work presented here is to develop the formalism required to interpret , within a cdm framework , angular - correlation - function measurements of deep surveys and to apply this to the observational results of cress et al .
we also discuss applications to measurements made from other radio surveys .
we restrict our investigation to spatially - flat models ( allowing for the possibility of a non - zero @xmath17 ) . given a power spectrum , calculation of the angular correlation function is straightforward and similar to that for optical or infrared surveys .
however , there are two important issues which we address here .
( 1 ) we investigate the uncertainties in the predictions which arise from imprecise knowledge of the survey redshift distribution .
( 2 ) since the typical redshifts of sources in the _ first _ survey are of order unity , the evolution of the power spectrum must be considered in the analysis or else there may be errors of order unity in the deduced primordial power spectrum . we include estimates of the non - linear evolution of the power spectrum .
the plan of the paper is as follows : in 2 , we discuss the redshift distribution inferred for the sample . in 3 , we discuss the time evolution of the power spectrum , and in 4 , we discuss the calculation of the angular correlation function .
results are presented and discussed in 5 .
we make some concluding remarks in 6 .
we use redshift - distribution estimates derived from two sources : ( i ) the 1.4-ghz radio luminosity function ( rlf ) presented by condon ( 1984 ) and ( ii ) a set of 2.7-ghz lf estimates presented by dunlop & peacock ( 1990 ) . in estimating the rlf , condon used faint radio sources with optical counterparts in the ugc catalog to determine a @xmath18 rlf for spiral galaxies .
he combined this with auriemma s ( 1977 ) estimate of the local rlf for elliptical galaxies and then found a model for the rlf at higher redshifts by allowing the local population to have evolved in density and luminosty .
the functions of redshift that describe the evolution were constrained using number counts ( down to sub - mjy thresholds ) at 1.4ghz and at other frequencies , spectral - index measurements and redshift measurements for a few very bright sources .
dunlop & peacock ( 1990 ) ( dp ) estimated the 2.7-ghz rlfs of steep- and flat - spectrum sources separately using similar constraints as those used by condon as well as some new data which included redshifts estimated from k - band photometry . to demonstrate the possible errors in the rlfs resulting from these imprecise redshift estimates , they analyze two distributions : one where assigned redshifts are the mean values for galaxies of a given k - band luminosity ( mean-@xmath19 distribution ) and one ( the high-@xmath19 distribution ) where redshifts are larger than the average ( they give reasons why one might expect this kind of bias ) .
seven models were presented and parameters for all seven models were determined for both mean- and high-@xmath19 distributions . here , we obtain the redshift distribution of _ first _
sources by integrating the rlf s over all luminosities that produce flux - densities greater than 1mjy .
the dp models were first shifted to 1.4ghz assuming a spectral index of @xmath20 for steep - spectrum sources and a spectral index of @xmath21 for flat - spectrum sources ( where the luminosity is related to the frequency by l@xmath22 ) .
figure 1 shows the redshift distribution @xmath23 for condon s lf and for a selection of the models given in dp for a flux - density limit of 1 mjy .
mean- and high-@xmath19 distributions produce similar results and the differences between dp s models 1 through 5 are smaller than the differences shown in the figure . in condon s model , the sharp increase in the number of sources near @xmath24 is due to starbursting galaxies . for larger flux - density limits ,
this low-@xmath19 peak is reduced considerably , and we have checked that our redshift distributions for larger flux - density limits agree with those shown in fig .
6 of loan , wall & lahav ( 1996 ) .
in cdm models , one starts with a scale - free primordial power spectrum , @xmath25 ( where @xmath26 is the comoving wavenumber ) .
this @xmath27 is then ` processed ' according a transfer function for cdm . at the epoch at which structure begins to grow , @xmath28 where @xmath29 for cdm is given by bond & efstathiou ( 1984 ) and @xmath30 is the normalisation .
we normalise to _ cobe _ 4-year data using liddle et al.s ( 1996 ) formula which is valid for spatially - flat models .
we will refer to this @xmath31 spectrum as the primordial power spectrum . in the linear regime
, the time evolution of the power spectrum can be calculated analytically and depends only on the expansion of the universe .
it is given by @xmath32/(1+z)^2 , \label{linearevolution}\ ] ] where the growth factor , @xmath33 , is given in carroll , press & turner ( 1992 ) , and @xmath34 .
in the highly non - linear regime , the cf obtained from a scale - free primordial spectrum obeys a simple scaling relation ( groth & peebles 1977 ) .
one can interpolate between the linear and highly non - linear regimes to produce semi - analytic models for clustering in the quasi - linear regime which can then be accurately fit to results from @xmath35-body simulations ( hamilton 1991 ; jain , mo , & white 1995 ; peacock & dodds 1996 ) .
pd use the dimensionless power spectrum @xmath36 . in our notation ,
the non - linear power spectrum is then given by @xmath37 , \label{pnlk}\ ] ] where the linear and nonlinear scales are related by @xmath38^{-1/3 } k_{nl}$ ] .
the form of the function , @xmath39 , is given in pd for various values of @xmath40 .
we take into account the time evolution of the power spectrum by evaluating pd s formulas using the redshift - dependent matter density @xmath41 and cosmological constant @xmath42 .
as is well known , the angular correlation function @xmath43 for a given population is related to the spatial correlation function @xmath44 by limber s equation ( rubin 1954 ; limber 1954 ; groth & peebles 1977 ; phillips et al . 1978 ; peebles 1980 ; baugh & efstathiou 1993 ) . if we assume that clustering is independent of luminosity ( or if we consider an averaged correlation function for the entire sample ) and that clustering is negligible on scales compared with the depth of the survey , then the angular correlation function in a flat universe is ( e.g. , baugh & efstathiou 1993 ) @xmath45 ^ 2 } , \label{limbereqn}\ ] ] where @xmath46 is the scale factor as a function of time @xmath47 and @xmath48 is the selection function ( the probability that a source at a distance @xmath49 is detected in the survey ) .
the physical ( not comoving ) separation between two sources separated by an angle @xmath50 is @xmath51 , \label{separation}\ ] ] where @xmath52 .
the total number of sources in a survey of solid angle @xmath53 is @xmath54 where @xmath23 is the redshift distribution of sources in the survey .
the spatial correlation function is the fourier transform of the spatial power spectrum @xmath55 : @xmath56 where @xmath26 is the comoving wavenumber . substituting ( [ separation ] ) , ( [ sourcenumber ] ) and ( [ powerspectrum ] ) into ( [ limbereqn ] ) , and integrating over @xmath57 , one obtains @xmath58 where @xmath59 to include non - linear effects , we use ( [ pnlk ] ) for the power spectrum .
the integral in ( [ wbande ] ) is then over @xmath60 and to determine the @xmath61 corresponding to a given @xmath60 , we need @xmath62 . before calculating the correlation function , we therefore set up a table of @xmath60 values for a range of @xmath61 s at various redshifts .
= 85truemm = 85truemm the form of the @xmath26 integrand in the cf estimate at @xmath63 is shown in figure 2 for @xmath64 and @xmath65 .
one sees that there are positive contributions to the cf from @xmath26 s in the range @xmath66 to @xmath67 with a large contribution coming from the larger @xmath26 s ( smaller scales ) . as one decreases the angle @xmath50 , the the peak contribution shifts towards even larger @xmath26 s . in figure 3
we have plotted @xmath68 , the redshift distribution of the clustering signal , for a range of angles ( @xmath69 universe ) .
one sees that the larger - angle cf estimates are dominated by contributions from nearby objects and even at @xmath70 one expects about half the signal to come from objects with @xmath71 .
this illustrates that the angular correlation function probes clustering at redshifts significantly smaller than those characteristic of the survey population .
more quantitatively , the models used here predict that the mean redshift probed by the correlation function at @xmath72 ( @xmath70,@xmath73 ) is expected to be @xmath74=0.4 ( 0.18 , 0.08 ) .
together , figures 2 and 3 highlight the fact that the cf measured for _ first _ is likely to be rather sensitive to smallish - scale correlations in fairly nearby sources .
figure 4 shows how the predictions for @xmath43 vary for different estimates of the redshift distribution and how they change when nonlinear evolution is considered .
there appears to be a fair amount of uncertainty in the predicted result , depending on which redshift distribution is chosen .
it should be noted , however , that condon s luminosity function is more carefully constructed to fit the low-@xmath19 population and it was pointed out above that these sources are expected to make large contributions to the clustering signal . in addition , peacock has indicated that dp s model 7 is the best model to use ( the difference between ` high-@xmath19 ' model 7 as opposed to ` medium-@xmath19 ' model 7 is not significant ) . if one limits oneself to dp7 and condon s model then uncertainties in the redshift distribution do not contribute very significantly to uncertainties in the correlation - function predictions . for the @xmath69 model shown in figure 4 ,
it is clear that the effect of nonlinear evolution is significant on scales less than @xmath75 . for a @xmath17-dominated model
the nonlinear contributions are significant out to scales of @xmath76
. the precise scale at which nonlinearities become important also depends on the redshift distribution .
note that in the standard - cdm model , @xmath77 for the linear - evolution spectrum is 1.22 while @xmath77 for the nonlinear - evolution spectrum is 1.17 .
= 85truemm = 85truemm figure 5 compares the measured cf in the _ first _ survey ( 1-mjy flux threshold ) with the cf s predicted by various cosmological models : the three solid lines are for @xmath64 and @xmath78 with @xmath79 taking on values of 0.5 ( lowest curve ) , 0.65 and 0.9 .
the dashed lines show results for @xmath80 and @xmath79 taking values 0.5 and 0.65 .
error bars in the plot are determined using a ` partition bootstrap method ' in which the cf is calculated for 10 subdivisions of the survey region and the standard deviation of these measurements at each angle is used as a measure of the error .
note that the spike in the cf at @xmath81 and possibly the dip at @xmath82 are associated with the sidelobe contamination discussed in cress et al .
( 1996 ) . at a first glance
it would appear that the data is best fit by an @xmath64 model with a high hubble constant , but this is misleading .
it is likely that at low fluxes , a significant fraction of the source counts come from fairly nearby `` star - bursting '' spiral galaxies as opposed to the agn - powered sources that one normally associates with radio sources ( condon 1984 ) . since spiral galaxies have been shown to have different clustering properties from ellipticals ( see , for example , hermit et al .
1996 ) and from agn in general ( for an estimate of the quasar correlation function see boyle et al .
1993 ) , it seems likely that the correlation function measured here will have contributions from populations with different clustering properties . in
3 we showed that on large angular scales ( @xmath83 ) , we expect the signal to be dominated by fairly local sources and so it is probably more reflective of the `` starburst '' clustering . on small angular scales we are probably more sensitive to the clustering of more distant agn . with this in mind
, we do not attach special significance to the curve which best matches the slope of the data .
instead , we calculate the bias @xmath84 , separately on small and large scales ( where @xmath85 ) and use this as an indicator of the relative bias of different populations in the survey . inferred biases for various cosmological models are shown in table 1 .
estimates of the bias calculated when the power spectrum is normalized to the observed cluster abundance are shown in brackets .
these were estimated by @xmath86 where @xmath87 is given by bunn & white ( 1997 ) and @xmath88 for @xmath89 ( viana & liddle 1996 ; white , efstathiou & frenk 1993 ) . [ cols="^,^,^,^,^,^,^",options="header " , ] table 1 also shows the value of @xmath90 calculated for various models assuming that the spatial correlation function of sources is approximated by a power law and that density fluctuations evolve linearly ; i.e. @xmath91 ( in comoving coordinates ) .
we used @xmath92 for the _ first _ measurement and @xmath93 for the other surveys but the @xmath90 results do not depend heavily on the chosen value of @xmath94 .
peacock & nicholson ( pn , 1991 ) measured the spatial correlation function for bright ( @xmath95 ) radio galaxies with redshifts between 0.01 and 0.1 and found it to be well fit by @xmath96 mpc and @xmath93 .
after correcting for redshift - space distortions , one would most likely obtain a real - space @xmath90 slightly less than this .
this result indicates that bright radio galaxies trace the matter distribution in a way similar to low - richness abell clusters . assuming linear evolution , estimates for @xmath90 in the _ first _ sample range from @xmath97 to @xmath98 depending on what redshift distribution is used .
note that in cress et al .
( 1996 ) , the @xmath90 obtained was based on a different estimate of the angular correlation function .
the value obtained here is consistent with the fainter sources in the sample being less clustered than the bright sources investigated by pn .
this picture is supported by ledlow & owen s ( 1995 ) observation that in a sensitive radio survey , field ellipticals and cluster ellipticals have a similar probability of being detected . one might thus expect sources to be clustered more like ellipticals than like clusters .
another consistency check is supplied by the measurement of the angular cross - correlation of abell clusters and _ first _ sources presented in cress et al .
( 1996 ) . using this result in a cross - correlation analog of limber s equation ( see seldner & peebles 1978 ) , we inferred the amplitude of the spatial cross - correlation of abell clusters and _ first _ sources .
dp s model 7 was used for the redshift distribution of _ first _ sources and the distribution of abell clusters was obtained by fitting a function of the form @xmath99 to data given in huchra et al .
( 1990 ) . by combining the autocorrelation and the cross - correlation we found the ratio of cluster bias to radio - source bias ( @xmath100 ) to be about 1.9 .
a similar estimate of @xmath100 was also obtained by combining the measured cluster - cluster spatial correlation and the radio - radio spatial correlation function inferred above .
mo , peacock & xia ( 1993 ) compared bright radio galaxies with abell clusters and obtained a value of 1.7 for this ratio .
this is also consistent with the idea that the faint sources in _ first _ are somewhat less clustered than the bright sources investigated by their work ( same as in pn ) .
= 85truemm angular correlation functions of radio sources have also been measured for the green bank and the parkes - mit - nrao ( pmn ) surveys done at 4.85 ghz ( loan , wall & lahav 1997 ; kooiman et al .
1995 ; sicotte 1995 ) and for wenss the 325-mhz westerbork northern sky survey ( rengelink et al .
we have investigated cdm predictions for these surveys using dp7 to estimate the redshift distribution of sources with @xmath101 mjy in the gb / pmn surveys and of sources with @xmath102 mjy in wenss . predicted correlation functions ( assuming no bias ) are shown in figure 6 .
taking @xmath103=0.0025 , @xmath104 and @xmath93 [ where @xmath105 , we have calculated the bias of sources in these surveys relative to cdm predictions and listed them in table 1 . the source composition in these two surveys is probably more similar to the sample investigated by pn , although wenss is likely to include fewer quasars .
the biases inferred for the pn sample are also shown in the table . while a large amount of variation in inferred clustering strength could be attributed to uncertainties in the correlation - function estimates and to the different clustering of different populations , data summarized in table 1 show some indication that as one probes deeper , the inferred bias increases . assuming that _ first _
( at small angles ) and the other surveys probe a similar population of objects , we have , with cluster normalization , @xmath106 at @xmath107 for the standard model and @xmath108 at @xmath109 .
one interpretation of this is that the clustering is evolving slower than cdm models predict
. however , matarrese et al . (
1996 ) , drawing on work by mo & white ( 1996 ) and by fry ( 1996 ) , have recently argued that in cdm models , the bias of a given population _ should _ decrease with time ; that is , the fluctuations in density inferred from luminous objects should tend towards the real fluctuations in the mass density as time goes by .
thus , the observations in table 1 could still be consistent with cdm predictions if bias evolution takes place .
interestingly enough , this apparent bias evolution is stronger than that seen in optical surveys ( matarrese et al .
this would be consistent with the general picture that the bias evolution for objects which tend to form at higher - density mass peaks is stronger .
if one writes @xmath110 , the value of @xmath111 which best fits the data is @xmath112 but this can vary significantly when uncertainties in the measurements are considered .
we have studied the implications of angular clustering of radio sources in deep surveys for standard - cdm like models for structure formation . we have examined the effect of uncertainties in the redshift distribution and considered nonlinear evolution of the power spectrum .
we have also studied what may be learned about the biasing of such sources and the evolution of such a bias .
we have shown that uncertainties in the redshift distribution of _ first _ sources can contribute a significant amount to the uncertainties in the cdm predictions of the angular correlation function even when bias is not considered . however , limiting oneself to condon s model and dp7 reduces the uncertainties significantly .
it should also be kept in mind that the uncertainty in the predicted angular correlation function which arise from uncertainties in the redshift distribution are still not much larger than the statistical uncertainty in the measured angular correlation function .
therefore , uncertainties in the redshift distribution do not currently affect strongly the implications of radio - source clustering for the origin of structure .
the effect of nonlinear evolution on the predictions has also been explored .
it appears that nonlinear contributions become important on scales of @xmath113 to @xmath76 depending on what model is used .
therefore , as data improve , the standard approximation of a power - law evolution of the clustering with redshift will break down .
we also found that although the mean redshift of sources in deep radio surveys may be of order unity , the clustering signal comes primarily from smaller redshifts .
that is , the angular correlation function is more sensitive to nearby clustering than one might expect from the @xmath23 distribution . furthermore , the redshift distribution of the clustering signal varies significantly with the angular scale probed .
therefore , the angular correlation function reflects the evolution of the three - dimensional power spectrum as well as the power spectrum at some fixed epoch . we have shown that spatially - flat cdm models can produce angular correlation functions similar to those observed in the _ first _ survey as long as some kind of biasing is invoked .
the needed biasing seems to be a bit larger than that needed for optical sources , but less than that needed for brighter radio sources .
this is probably due to the fact that _ first _ sources are a heterogeneous sample which contain numerous nearby starbursting galaxies as well as more distant brighter radio sources .
we have also begun to explore what may be learned about the evolution of bias from the _ first _ and other radio surveys . at this stage , it is difficult to make very quantitative statements about the evolution of clustering because the cf measurements that probe higher-@xmath19 ranges have large uncertainties associated with them .
in addition , bias evolution will have to be understood better if we are to learn about the evolution of mass clustering .
this point is particularly significant for a heterogeneous sample of objects such as _ first _ sources .
one way to improve our ability to make quantitative statements about clustering evolution is to make more precise measurements of the clustering of sources that probe the higher-@xmath19 range . as the _ first _ survey coverage increases
, we will be able to obtain significant clustering signals for samples with higher flux thresholds .
increasing the flux threshold should increase the average redshift probed by the correlation function and thus improves one s chances of probing larger scales .
another way of increasing the redshift probed by the cf might be to remove all sources that have optical counterparts in the apm survey .
this is currently being investigated .
improving our estimates of the redshift distribution of faint sources is also important and there is some hope that this will be done in the near future .
we thank john peacock , jim condon , ofer lahav , david helfand , alexandre refregier and jacqueline van gorkom for useful comments .
this work was supported by the u. s. department of energy under contract de - fg02 - 92er40699 , nasa grant nag5 - 3091 and the alfred p. sloan foundation .
the first project is supported by grants from the national geographic society , nsf grant ast94 - 19906 , nato , igpp , columbia university , and sun microsystems .
bartlett j. g. et al . , 1994 , astro - ph/9407061 baugh c. m. , efstathiou g. , 1993 , mnras , 265 , 145 becker r. h. , white r. l. , helfand , d. j. , 1995 , apj , 450 , 559 ( bwh ) bond j. r. , efstathiou g. , 1984 , apj , 285 , l45 boyle b.j .
et al . , 1994 , mnras , 271 , 639 bunn e.f , white m. , 1997 , apj , 480 , 6 carroll s. m. , press w. h. , turner e. l. , 1992 araa , 30 , 499 condon j. j. , 1984 , apj , 287 , 461 cress c. m. , helfand d. j. , becker r. h. , gregg m. d. , white r. l. , 1996 , apj , 473 , 7 cress c. m. et al .
, 1997 , to appear in jackson n. , bremer m. , eds , cosmology and the new radio surveys . cambridge university press , cambridge davis m. , peebles p. j. e. , apj suppl . , 34 , 425 davis m. , groth e. j. , peebles p. j. e. , 1977 , apj , 212 , l107 dunlop j. s. , peacock j. a. , 1990 mnras , 247 , 19 efstathiou g. , sutherland w. j. , maddox s. j. , 1990 , nat . , 348 , 705 fry j. n. , 1996 , apj , 461 , l65 groth e. j. , peebles p. j. e. , 1977 , apj , 217 , 385 jain b. , mo h. j. , white s. d. m. , 1995 , mnras , 276 , 25 hamilton a. j. s. , matthews a. , kumar , p. , lu e. , 1991 , apj , 374 , l1 hermit s. , santiago b. x. , lahav o. , strauss m. a. , davis m. , dressler a. , huchra j. p. , 1996
, astro - ph/9608001 huchra j.p .
et al . , 1990 ,
apj , 365 , 66 kaiser n. , 1984 , apj , 284 , l9 kofman l. a. , gnedin n. y. , bahcall n. a. , 1993 , apj , 413 , 1 kooiman b. l. , burns j. o. , klypin a. a. , 1995 , apj , 448 , 500 krauss l. m. , turner m. s. , 1995 , gen .
grav . , 27 , 1137 landy s. d. , szalay a. s. , 1993 , apj , 412 , 64 ledlow m. j. , owen f. n. , 1995 , aj , 109 , 853 liddle a. r. , lyth d. h. , roberts d. , viana p. t. p. , 1996 ,
mnras , 278 , 644 limber d. n. , 1954 , apj , 119 , 655 loan a. j. , wall j. v. , lahav o. , 1996 , astro - ph/9612190 matarrese s. , coles p. , lucchin f. , moscardini l. , 1996 , astro - ph/9608004 mo h. j. , peacock j. a. , xia , x. y. , 1993 , mnras , 260 , 121 mo h. j. , white s. d. m. , 1996 , mnras , 282 , 347 ostriker j. p. , steinhardt p. , 1995
, astro - ph/9505066 peacock j. a. , dodds s. j. , 1994 , mnras , 267 , 1020 peacock j. a. , dodds s. j. , 1996 , mnras , 280 , l19 ( pd ) peacock j. a. , nicholson d. , 1991 , mnras , 253 , 307 peebles p. j. e. , 1980 , the large scale structure of the universe ( princeton : pup ) phillips s. , fong r. , ellis r. s. , fall s. m. , macgillivray h. t. , 1978 , mnras , 182 , 673 rengelink et al . , 1997 , to appear in jackson n. , bremer m. , eds , cosmology and the new radio surveys .
cambridge university press , cambridge rubin v. c. , 1954 , proc .
nas , 40 , 541 seldner m. , peebles p. j. e. , 1978 , apj , 225,7 sicotte h. , 1995 , ph.d .
thesis , princeton university viana p.
t. p. , liddle a. r. , 1996 , mnras , 281 , 323 white r. l. , becker r. h. , helfand d. j. , gregg m. d. , 1996 , apj , 475 , 479 white s. d. m. , efstathiou g. , frenk c. s. , 1993 , mnras , 262 , 1023 | we develop the formalism required to interpret , within a cdm framework , the angular clustering of sources in a deep radio survey .
the effect of nonlinear evolution of density perturbations is discussed as is the effect of the assumed redshift distribution of sources .
we also investigate what redshift ranges contribute to the clustering signal at different angular scales .
application of the formalism is focussed on the clustering detected in the _ first _ survey but measurements made for other radio surveys are also investigated .
we comment on the implications for the evolution of clustering .
-0.5 in # 1#1 # 1#23.6pt galaxies : clustering cosmology : theory large - scale structure of universe . |
synchronization is observed in many natural , social , physical and biological systems , and has found applications in a variety of fields @xcite .
the large number of networks of coupled dynamical systems that exhibit synchronized states are subjects of great interest . as a theoretical paradigm
, the kuramoto model is usually used to investigate the phase synchronization among nonidentical oscillators @xcite . however
, some real physical systems that display synchronized phenomenon ( e.g. the coupled double planar pendulums , the two - state quantum system , etc . ) can not be properly described by the simply kuramoto model . actually , the kuramoto oscillator is equal to the left - invariant flow on the simplest nontrivial lie group @xmath1 ( see the refs .
@xcite for the fundamental knowledge of lie group ) . accordingly ,
in this paper , by extending the concept of kuramoto oscillator to left - invariant flow on general lie group , we investigate the generalized phase synchronization on networks .
in particular , the left - invariant flows corresponding to non - abelian lie groups are named non - abelian oscillators , which can be applied on some non - abelian dynamical systems like quantum systems .
this paper is organized as follow : in section 2 , the concept of synchronization of non - abelian oscillators is introduced . in section 3 ,
the analyses and simulations on some typical examples are given , including the @xmath0-dimensional torus @xmath2 , the identity component of 3-dimensional general linear group @xmath3 ( i.e. @xmath4 ) , the special unitary group @xmath5 and the special orthogonal group @xmath6 .
finally , we sum up this paper and discuss the relevance of phase synchronization of non - abelian oscillators to the real world in section 4 .
some mathematic remarks are given in the appendix .
first of all , we review the simple oscillator : @xmath7 it is obvious that @xmath8 is the solution of the equation @xmath9 which relates to the dynamical system @xmath10 on the circle @xmath11 . in terms of lie groups
, @xmath12 is an abelian lie group with its lie algebra @xmath13 and the map @xmath14 is the exponential map of @xmath15 .
the system ( 3 ) is also called the left - invariant flow , of which the tangent vector field is called left - invariant vector field on @xmath15 .
this fact motivates the idea that the left - invariant flows on the general lie groups can be considered as _ general oscillators_. in particular ,
the left - invariant flows on the non - abelian lie groups can be considered as _ non - abelian oscillators_. let @xmath16 be a lie group and @xmath17 its lie algebra .
if @xmath18 , then the left - invariant flow determined by @xmath19 on the @xmath16 is @xmath20 different from those defined on the linear euclidean spaces , the above dynamical system ( 5 ) is defined on the manifold @xmath16 .
next we consider the synchronization of network , of which each node is located a general oscillator on the lie group @xmath16 with its lie algebra @xmath17 .
the collective synchronization behavior of the coupled general oscillators starts by randomly choosing an element @xmath21 in @xmath17 and an initial phase @xmath22 for each node @xmath23 . in the case of the general oscillator
@xmath24 and @xmath25 correspond to the frequency @xmath26 and @xmath27 in eq .
( 3 ) , respectively .
it is worthwhile to note that , in general , the phase space of our general oscillator is both non - abelian group in algebra , and non - linear space , which is a manifold , in geometry .
thus one may expect that there will be more interesting phenomena in this model . [
0.8 ] ( a ) and @xmath28 ( b ) vs coupling strength @xmath29 for different values of the rewiring probability @xmath30 .
all the simulation results are the average over 100 independent runs corresponding to the case of @xmath31.,title="fig : " ] [ 0.8 ] ( a ) and @xmath28 ( b ) vs coupling strength @xmath29 for different values of the rewiring probability @xmath30 . all the simulation results
are the average over 100 independent runs corresponding to the case of @xmath31.,title="fig : " ] for a dynamical system on the manifold @xmath16 , its equations of motion should be written in a local coordinate system because , generally , there does not exist a global coordinate system on a manifold .
let @xmath32 be a local coordinate system about the identity @xmath33 of lie group @xmath16 , @xmath34 , and @xmath35 the coordinates of @xmath33 .
if @xmath36 , @xmath37 and @xmath38 are the coordinates of @xmath39 and @xmath40 in this coordinate system respectively , then the multiplication function @xmath41 can be defined as : @xmath42 suppose that @xmath18 , then the equations of the left - invariant flow determined by @xmath19 are @xmath43 where @xmath44 is the coordinates of @xmath19 with respect to the basis @xmath45 , and @xmath46 is the matrix valued function : @xmath47 the set of equations ( 7 ) is the intrinsic dynamic of a single general oscillator .
note that , the multiplication functions are the local representation of the multiplication in the lie group , and the matrix valued function ( 8) is the map induced by multiplying by @xmath48 at left .
this map translates every vector in the lie algebra to @xmath48 .
( 7 ) mean that the left - invariant flow determined by @xmath19 are made from the vector field whose value at @xmath48 is just the vector into which translating @xmath19 .
similar to the case of coupled simple oscillators @xcite , the coupled dynamics of general oscillators are : @xmath49 where @xmath50 is the coupling function assumed smooth , @xmath51 is the set of @xmath23 s neighboring nodes , and @xmath52 is the coordinates of initial phase of node @xmath23 . according to the existence and uniqueness theorem of ordinary differential equations , eq .
( 9 ) has an unique solution .
then , by the theory of lie groups , this solution can be extended to an one - parameter subgroup , which corresponds to a left - invariant flow .
in this section , we will show some analyses and simulations on several typical examples .
all the simulations are obtained based on the watts - strogatz ( ws ) network @xcite , which can be constructed by starting with one - dimensional lattice and randomly moving one endpoint of each edge with probability @xmath30 .
the network size @xmath53 and average degree @xmath54 are fixed . in our numerical simulation
below , the coordinates @xmath55 are chosen randomly and uniformly in the range @xmath56 , and the initial phases @xmath52 in @xmath57 .
all the numerical results are obtained by integrating the dynamical equations using the runge - kutta method with step size 0.01 . the order parameters ( see below ) are averaged over 2000 time steps , excluding the former 2000 time steps , to allow for relaxation to a steady state .
we firstly investigate the @xmath0-dimensional torus , @xmath58 , which is a connected compact lie group .
denote @xmath59 and its lie algebra @xmath60 with @xmath61 \equiv 0 $ ] ( i.e. abelian ) .
this is a direct extension of the simple kuramoto oscillators , and will degenerate to karumoto oscillators when @xmath62 .
it follows that an element in @xmath63 , that is an intrinsic frequency , is @xmath64 .
[ 0.8 ] vs the coupling strength @xmath29 .
all the simulation results are obtained by averaging 100 independent runs.,title="fig : " ] it is easy to see that the general oscillation determined by @xmath19 is periodic if @xmath65 are integer - linearly dependent , and quasi - periodic if @xmath66 are integer - linearly independent . for the equations of motion we have @xmath67 with the initial conditions @xmath68 , where @xmath29 is the coupling strength . to characterize the synchronized states of the @xmath69th dimension
, we use the order parameter @xmath70 where @xmath71 signifies the time averaging .
1 displays the relations between the phase order parameter and the coupling strength for various values of the rewiring probability @xmath30 , where @xmath31 .
since the dynamical behaviors in different dimensions are independent , and each one is equal to a simply kuramoto oscillator , the synchronization behavior in each dimension is the same as that of a kuramoto oscillator @xcite . in a word ,
the small - world effect ensures the global synchronization for sufficiently large coupling strength , and the higher disorder ( i.e. larger @xmath30 ) will enhance the network synchronizability .
this example can be used to describe the collective behavior of the coupled double planar pendulums in a network .
furthermore , according to the liouville - arnold theorem , the phase space of a integrable system , if it is compact , is a @xmath0-torus .
therefore , this example can be widely applied to describe the synchronization phenomenon of integrable systems .
@xmath4 , the identity component of @xmath3 , is a connected lie group of dimension 9 ( actually , @xmath4 is a open subset of @xmath72 ) .
denote @xmath73 , and @xmath74 its lie algebra containing all the @xmath75-real matrices .
the coordinates @xmath76 for all @xmath77 can be addressed on the basis @xmath78 of @xmath79 ( see the case ( a ) of appendix a ) , where @xmath80 in this circumstances the equations of motion are @xmath81 with the initial conditions @xmath82 , where @xmath83 and @xmath29 is the coupling strength .
[ 0.8 ] vs the coupling strength @xmath29 .
all the simulation results are obtained by averaging 100 independent runs.,title="fig : " ] since @xmath84 is a subset of the euclidean space @xmath72 , the coupling term of arbitrary pair of nodes @xmath23 and @xmath85 can be directly written as @xmath86 .
therefore , the order parameter is defined as : @xmath87 } } \right\rangle } \right],\ ] ] where @xmath88 and @xmath89 . in fig .
2 , we report the simulation results about the order parameter @xmath90 as a function of @xmath29 for different rewiring probabilities . similar to the situation of @xmath2 , this system can approach to the global synchronized state for sufficiently large @xmath29 when @xmath30 is large enough .
it is worthwhile to emphasize that the order parameter of different systems can not compare to each other directly , since their definitions are not the same .
@xmath91 , the special unitary group , is a connected compact lie group consisted of all anti - hermitian traceless matrices .
denote @xmath92 , and @xmath93 its lie algebra having the following basis @xmath94 give an arbitrary element @xmath95 , its local coordinates @xmath96 can be found from : @xmath97 and @xmath98 the corresponding equations of motion are @xmath99 with the initial condition @xmath100 , where @xmath29 is the coupling strength .
it is worthwhile to emphasize that the behaviors of this dynamical system is independent to the local coordinate systems .
it follows that the local coordinates ( 14 ) in @xmath5 can be replaced by @xmath101 which is convenient for the description of two - state quantum systems since the matrices @xmath102,@xmath103 , and @xmath104 are relative to pauli matrices as : @xmath105 note that , a one - parameter subgroup @xmath106 can be considered as the time - evolution in the quantum mechanics . since the subgroup @xmath107 , where @xmath108 is an arbitrary real number , represents the same time - evolution as @xmath109 does , we call two different nodes @xmath23 and @xmath85 are synchronous if they are only differ in a phase factor @xmath110 , that is @xmath111 therefore , if the equation @xmath112 is hold for every @xmath113 , the completely global synchronization is achieved .
consequently , for each pair of nodes @xmath23 and @xmath85 , define the dynamical distance " between them as @xmath114 where @xmath115 signifies the modulus of the complex number .
accordingly , the order parameter is @xmath116.\ ] ] in fig .
3 , we report the simulation results about the order parameter @xmath90 as a function of @xmath29 for different rewiring probabilities .
analogously , this system can approach to the global synchronized state for sufficiently large @xmath29 when @xmath30 is large enough .
note that , @xmath117 for @xmath118 . thus in the uncoupling case ( i.e. @xmath119 ) , the order parameter @xmath90 approaches to 0.37 . in terms of the hopf bundle @xmath120 ,
the base space @xmath121 is the set of states of a two - state quantum system ( see the ref .
@xcite and appendix b for details ) .
therefore , let a two - state quantum system be located at each node of a small - world network , the whole system can achieve the synchronized state if the coupling strength is large enough . note that , the special orthogonal group @xmath6 has an isomorphic lie algebra to @xmath93 for the existence of the double covering @xmath122 , which is a homomorphism .
therefore , the dynamical system @xmath6 has the same synchronized behavior as that of @xmath5 .
the case where each node is located with a top can be represented by the system @xmath6 @xcite .
in this paper , we extended the concept of kuramoto oscillator to the left - invariant flow on general lie group , and studied the generalized phase synchronization on small - world networks .
the left - invariant flow on lie group can be used to represent many significant physical systems , such as the integrable systems ( e.g. the double planar pendulums ) , the two - state quantum systems ( e.g. ising model ) , the motions of top , and so on . in particular , the dynamics of two - state quantum systems on complex networks are extensively studied recently @xcite , the present work can provide us a theoretical frame to investigate their collective synchronized behaviors . it is intuitive that the low - dimensional systems are more easily to synchronize .
however , one may note that in the regular networks ( i.e. @xmath123 ) , the system @xmath124 does not exhibit synchronized behavior while the obvious synchronization is observed in the higher dimensional systems @xmath4 and @xmath5 when the coupling strength gets large .
this is an interesting phenomenon that worths a further study in the future .
we thank dr .
hui - jie yang for helpful discussions .
this work was partially supported by the national natural science foundation of china under grant nos .
70471084 , 70471033 and 70571074 , the special research founds for theoretical physics frontier problems under grant no .
a0524701 , the specialized program under the presidential funds of the chinese academy of science , and the graduate student foundation of ustc under grant no . kd2005007 .
in order to investigate the collective synchronization behaviors of the coupled non - abelian oscillators , we have to integrate eq .
( 9 ) numerically .
because of the using of local coordinate system , a question rises : does the eq . ( 9 ) express the behavior of the system in the whole manifold @xmath16 ?
the answer is as follows .
manifold @xmath16 is a connected open subset of euclidean space @xmath125 . in this case
, we can directly use the natural coordinate system of @xmath125 as the local coordinate system of @xmath16 .
@xmath16 is a connected compact lie group . in this case
, we have @xmath126 , so we can choose the coordinate system of @xmath17 as the local coordinate system of @xmath16 .
according to the theory of fiber bundles , the condition ( 20 ) is equivalent to that @xmath127 and @xmath128 is all in the same fiber of the hopf bundle @xmath129 . in order to interpret the map
@xmath30 we consider the 2-sphere @xmath121 as the complex projective line @xmath130 in which a point is written as @xmath131 $ ] , where @xmath132 and @xmath133 are two complex numbers and not all zero . here , the point @xmath131 $ ] is identified with @xmath134 $ ] for any nonzero complex number @xmath135 .
then the map @xmath30 can be written as @xmath136.\ ] ] it is easy to show that the inverse image @xmath137 $ ] is the sphere @xmath15 @xcite .
this fact tells us that , in the sense of our definition , the collective synchronization behaviors of the coupled non - abelian oscillators @xmath138 can represent the two - state quantum system .
strogatz2003 s. h. strogatz , _ sync - how the emerges from chaos in the universe , nature , and daily life _ ( hyperion , new york , 2003 ) .
y. kuramoto , _ chemical oscillations , wave and turbulence _ ( springer - verlag , berlin , 1984 ) .
a. pikovsky , _ synchronization _ ( cambridge university press , cambridge , 2001 ) .
j. a. acebrn , l. l. bonilla , c. j. p. vicente , f. ritort , and r. spigler , rev .
mod . phys . * 77 * , 137 ( 2005 ) .
b. f. schutz , _ geometrical method of mathematical physics _ ( cambridge university press , 1980 ) .
t. brocker , and t. t. dieck , _ representations of compact lie groups _ ( springer - verlag , 1985 ) .
h. hong , m. y. choi , and b. j. kim , phys .
e * 65 * , 026139 ( 2002 ) .
d. j. watts , and s. h. strogatz , nature * 393 * , 440 ( 1998 ) .
t. frankel , _ the geometry of physics , an introduction _
( cambridge university press , cambridge , 2004 ) . v. i. arnold , _ mathematical methods of classical mechanics _ ( springer - verlag , new york , 1978 ) . c. p. herrero , phys .
e * 65 * , 066110 ( 2002 ) .
h. hong , b. j. kim , and m. y. choi , phys .
e * 66 * , 018101 ( 2002 ) . c. -p .
zhu , s. -j .
xiong , y. -j .
tian , n. li , and k. -s .
jiang , phys .
rev . lett . * 92 * , 218702 ( 2004 ) . | in this paper , by extending the concept of kuramoto oscillator to the left - invariant flow on general lie group , we investigate the generalized phase synchronization on networks .
the analyses and simulations of some typical dynamical systems on watts - strogatz networks are given , including the @xmath0-dimensional torus , the identity component of 3-dimensional general linear group , the special unitary group , and the special orthogonal group .
in all cases , the greater disorder of networks will predict better synchronizability , and the small - world effect ensures the global synchronization for sufficiently large coupling strength .
the collective synchronized behaviors of many dynamical systems , such as the integrable systems , the two - state quantum systems and the top systems , can be described by the present phase synchronization frame .
in addition , it is intuitive that the low - dimensional systems are more easily to synchronize , however , to our surprise , we found that the high - dimensional systems display obviously synchronized behaviors in regular networks , while these phenomena can not be observed in low - dimensional systems . |
shell - type supernova remnants ( snrs ) have a special significance among expected galactic @xmath0-ray sources , because only they have a sufficiently large _ hydrodynamical energy output _ to replenish the dominant nucleonic component of the cosmic rays ( crs ) in the galaxy .
the arguments are as follows : except possibly for @xmath0-ray bursts at cosmological distances , sn explosions are the most violent `` events '' in the universe .
they have the largest single release of mechanical energy @xmath2 erg , where a mass @xmath3 between about one and several @xmath4 is ejected with a velocity @xmath5 km / sec . within the galaxy , assuming a sn rate @xmath6 this also implies the largest mean hydrodynamical energy input into the interstellar medium ( ism ) @xmath7 ; the factors between each of these energy inputs may very well reach values of the order of ten .
nevertheless @xmath8 is still only marginally sufficient to replenish the crs against their escape from the galaxy @xmath9 for an energy - independent `` grammage '' @xmath10 g @xmath11 , as derived from the cr composition at gev energies . in this relation @xmath12 and @xmath13
denote the cr confinement volume and the interstellar gas mass , respectively .
the grammage @xmath14 decreases with increasing particle energy as suggested by the observed energy dependence of the cr secondary to primary ratio @xmath15 g @xmath11 ( e.g. swordy et al . ,
1990 ) , with @xmath16 denoting rigidity , as well as by cr propagation theory ( ptuskin et al .
therefore the cr _ source spectrum _ is significantly harder than the steady state particle spectrum observed near the solar system .
this increases the cr energy input requirement from the conventionally estimated few percent of @xmath8 ( e.g. berezinsky et al . ,
1990 ) to @xmath17 erg sec@xmath18 , i.e. to something between 10 and 20 percent of @xmath8 ( drury et al .
thus , energetically there are no plausible sources for the cr nucleon component in the galaxy other than snrs , and their particle acceleration ought to be very efficient .
the same argument does not hold for the cr electrons whose energy density is lower by 2 orders of magnitude at energies exceeding @xmath19 gev .
therefore the physical nature of the electron sources is energetically much less constrained than that of the cr nucleons .
apart from the snr energetics there are quite strong arguments from _ acceleration theory _ : the energization of nonthermal particles in snrs is usually assumed to be due to diffusive shock acceleration at the outer shock that sweeps up the circumstellar medium during the expansion of the remnant ( for reviews of diffusive shock acceleration theory , see e.g. drury , 1983 ; blandford and eichler , 1987 ; berezhko and krymsky , 1988 ; jones and ellison , 1991 ) .
all models agree that a hard , power law - type nucleon spectrum with maximum energies around 100 tev should be produced during the evolution of a typical snr in a diffuse interstellar medium ( ism ) of not too high density .
we shall evaluate the models , as they have been developped to date , in the next section .
theoretical estimates of the @xmath1 - decay @xmath0-ray luminosity of snrs ( drury et al .
, 1994 ; naito and takahara , 1994 ) have led to the conclusion that the resulting @xmath0-ray flux would be _ difficult to observe with present instruments_. we shall discuss the recent tev searches in the northern hemisphere and compare the model predictions with the upper limits .
_ electron acceleration _ in snrs is more difficult to estimate quantitatively since the injection efficiency and even the very process of acceleration for electrons are as yet much less clearly determined . on the other hand , it is empirically plausible that electrons are copiously accelerated in snrs .
for example , the well - known correlation between the integrated far infrared and radio continuum luminosities of late type galaxies is dominated by sn precursor stars .
the far inrared emission is due to dust reradiation of the stellar uv emission .
the radio continuum is mainly synchrotron emission by @xmath20 gev diffuse galactic cr electrons .
it is consistent with a hard source spectrum of cr electrons @xmath21 and a @xmath22 acceleration efficiency ratio of electrons to nucleons ( lisenfeld et al . , 1996 ) .
in addition , several individual snrs show x - ray continua which have been attributed to nonthermal synchrotron emission by multi - tev electrons ( e.g. koyama et al . , 1995 )
. the corresponding inverse compton ( ic ) emission might therefore be able to swamp the nucleonic @xmath0-ray emission from snrs .
altogether , real snrs are not the ideal spherically symmetric configurations as theoreticians tend to picture them for simplicity
. the circumstellar environment into which a supernova explodes can in fact be quite complicated and we shall at least qualitatively attempt to evaluate the effects of these uncertainties . at the durban icrc just before this workshop the cangaroo group has announced the detection of the southern hemisphere sn 1006 in tev @xmath0-rays .
this is a tremendously exciting result which suggests that snrs are indeed capable to accelerate multi - tev electrons , as expected from the asca observations in the hard x - ray region .
why should then not multi - tev nucleons be accelerated efficiently as well ?
we may again use the theoretical estimates for nucleon acceleration , given a set of assumed parameters for this object that have been derived from observations in different wavelength regions .
then it appears that the observed tev @xmath0-ray flux could be produced in hadronic interactions only with considerable difficulties ( ! ) .
this seemingly paradoxical result , when seen in the light of the non - detection of other nearby snrs in the northern hemisphere , can be rationalized if we assume that sn 1006 is an ic @xmath0-ray source .
this conclusion is of course highly provisional and needs further careful scrutiny .
tycho s snr , even though considerably younger , might have a similar @xmath1 - decay flux as sn 1006 .
if tycho s x - ray spectrum has only a very small nonthermal component as suggested by the dominance of emission lines , it might indeed be possible to detect its @xmath0-ray emission of hadronic origin in a deep observation with present instruments like the hegra stereoscopic system , or whipple and cat .
the results should be known relatively soon .
the simplest models treat crs as test particles in a prescribed , time - dependent substrate given by a hydrodynamic solution for the explosion dynamics , for example a sedov solution . in the first concrete estimates ( cesarsky and lagage , 1981 ; ginzburg and ptuskin , 1981 ; lagage and cesarsky , 1983 ) it was assumed that the outer snr shock was locally plane . with the pitch angle scattering
mean free path @xmath23 at the bohm limit ( which implies that a particle is turned around in its motion parallel to the mean magnetic field after each gyration ) , an upper limit to the achievable maximum momentum @xmath24 could be calculated by equating @xmath25 .
will be smaller than the above estimated value .
however , a very strong shock should lead to such strong excitation of scattering mhd waves by the accelerating particles themselves ( bell , 1978 ; mckenzie and vlk , 1982 ) , that the bohm limit may indeed approximately be reached in such a system . ] here @xmath26 , with @xmath27 denoting the diffusion coefficient in the upstream ( 1 ) and downstream ( 2 ) region , respectively ; the particle speed is given by @xmath28 ( e.g. drury , 1983 ) .
for typical values of the snr parameters one obtains in this manner @xmath29 . the actual _ nucleon _ spectrum in snrs has first been approximately calculated in the so - called `` onion shell '' models ( bogdan and vlk , 1983 ; moraal and axford , 1983 ) : the spectral index @xmath30 of the accelerated particles at the shock varies with time , following the time dependence of the shock compression ratio @xmath31 . to simplify the time dependence of the solution , time
is divided into n intervals and at each time @xmath32 an imaginary shell surface with radius @xmath33 is created .
the @xmath34 shell contains the material between the shock radii @xmath35 and @xmath33 . after being accelerated `` at time @xmath36 '' , the particles in a given radial shell k
remain confined _ inside _ the expanding snr and therefore loose energy adiabatically . in this kinematic , quasi - stationary model ,
the maximum particle energy depends on the strength of the shock at @xmath37 and the actually calculated wave spectrum at this time ( vlk et al . , 1988 ) . to obtain the spatially integrated spectrum inside the snr at time t
, all shells k with @xmath38 are summed up .
particles are finally assumed to be released from their snr source when the shock velocity @xmath39 decreases below the alfvn velocity @xmath40 km sec@xmath18 of the ambient medium .
the interesting result was that the final source spectrum corresponds approximately to a
_ single power law _ in momentum @xmath41 with @xmath42 over a large range in particle momenta above the injection momenta ( fig .
1 ) . at the end of the so - called sweep - up phase , when the remnant has swept up an amount of interstellar material comparable to the ejected mass from the explosion , the spectrum is very hard @xmath43 , with the maximum particle energies achieved . after the end of the sedov phase , the spectrum is softer , with the maximum energy @xmath44 only slightly exceeding @xmath45 times the proton rest energy @xmath46 for the particular parameters of a hot ism . in such a test particle picture
the fraction of injected particles @xmath47 has to be restricted to smaller values during early times t in order to not violate total energy flux conservation since no backreaction of the accelerated particles is assumed in such models . for _ electrons _ the determination of @xmath24 has to be modified , since it may be determined by _ energy loss processes _ rather than by the system age t alone , as for nucleons .
thus @xmath48 ( reynolds and chevalier , 1981 ) .
reynolds ( 1996 ) has applied such models to the x - ray synchrotron emission from sn 1006 .
he parametrized the scattering mean free path by a factor @xmath49 times the electron gyro radius , and introduced the speculative possibility of particle escape from the remnant at some maximum energy . with these additional parameters in the model he argued that the observed x - ray continuum emission from sn 1006 could be explained by the acceleration of electrons up to 100 tev in the remnant blast wave .
a drastic simplification of the onion shell model for electrons was recently made by mastichiadis ( 1996 ) : he assumed the accelerated particles at any given time to be distributed uniformly over the remnant interior , subject to subsequent adiabatic energy losses .
taking also the magnetic field strength to be uniform inside the remnant and restricting the injection rate by globally requiring an electron to proton ratio of @xmath50 at release , he could show that the ic yield of these electrons on the mbr at tev energies is comparable with the expected @xmath1 - decay emission from the nucleonic component .
even though these electron models simply ignore the difficulty of injecting electrons into the shock acceleration process , they justify themselves by the fact that multi - tev electrons seem to be required empirically to explain the x - ray synchrotron emission from a number of snrs ( section 5 ) .
it seems clear that we have to reckon with electron ic @xmath0-ray emission in all snrs .
the test particle models discussed above are not only of a kinematic and quasi - stationary character , but they also neglect the fact that the nucleonic component must have a strong dynamical influence on the snr shock ( `` backreaction '' ) modifying the acceleration process significantly , if this process is to be efficient in the first place . a simple hydrodynamic approximation for the time - dependent nonlinear equations for the _ coupled dynamics _ of the thermal gas ( plasma ) and the accelerating crs is the so - called 2-fluid model ( axford et al . , 1977 , 1982 ; drury and vlk , 1981 ) .
it consists in taking the kinetic energy moment of the cr transport equation and introducing the cr pressure gradient force into the momentum balance for the overall system of plasma and crs . extending this fluid approximation to a 3-fluid model ( mckenzie and vlk , 1982 ) ,
also the scattering wave energy can be selfconsistently included .
nonlinear shock modification then leads to the following effects ( fig .
2 ) : in the frame of the shock , the pressure @xmath51 of the accelerating particles has a gradient that decelerates the incoming flow in the precursor , before the gas pressure , the mass velocity and the mass density finally jump to their downstream values in a subshock .
acceleration is the more efficient the more the shock is smoothed out since in this case irreversible gas heating is reduced and a higher fraction of the free flow energy must go into cr energy .
the total compression ratio exceeds the canonical value of 4 and therefore at high energies where the diffusion length is large , the spectrum of accelerated particles is expected to be be harder than @xmath43 , as has been pointed out by eichler ( 1984 ) and ellison and eichler ( 1984 ) in their kinetic discussion of nonlinear shocks . in a snr this should be the case during the early evolutionary phases when shock modification is strong .
particle diffusion into the upstrem medium excites hydromagnetic waves there .
they propagate against the incoming flow with the alfvn speed and get further amplified at the ( sub)shock .
an important aspect is that only these selfexcited waves permit efficient acceleration since snr shocks have a short life - time , typically of the order of @xmath52 ; the much weaker average ism hydromagnetic turbulence alone would make this process basically ineffective .
the hydrodynamic approximation has been applied to snrs in various forms ( drury et al . , 1989 ; markiewicz et al . ,
1990 ; dorfi , 1990 , 1991 ; kang and jones , 1991 ) .
all assume the shock normal to be parallel to the external magnetic field ( parallel shock ) , in an otherwise spherically symmetric configuration .
this includes adiabatic energy losses in the remnant interior and diffusive `` losses '' of particles from the shock into the interior .
obviously none of these effects can be taken into account in an application of plane shock theory to such a configuration .
also heating of the thermal gas by the dissipation of the selfexcited alfvn waves has been included , even though the diffusion coefficient was maintained at the bohm limit .
this is no contradiction in the limit of strong wave excitation .
such model calculations show that for very reasonable injection rates 10 - 30 percent of the entire hydromagnetic sn energy @xmath53 can be readily converted into cr nuclei during the early evolutionary stages , when the snr shock is strong .
the ultimate release of crs into the ism is determined by escape and the details of adiabatic cooling during the late phases where the radiative cooling of the thermal gas and the external pressure compete in decelerating the remnant expansion .
however , simple estimates and the numerical results of dorfi ( 1991 ) indicate that an efficiency of cr release of about 10 percent is quite possible .
even though the hydrodynamid fluid models constitute a simple means to calculate the snr dynamics including cr production , they do not determine the energy spectra and their temporal evolution .
the resulting knowledge of the instantaneous shock compression ratio says little about the overall spectrum , integrated over the remnant interior , which is relevant for the resulting @xmath0-ray spectra . from a quite different angle ,
such spectra have been estimated by baring et al .
( 1997a ) using the monte carlo approach to particle scattering of ellison and eichler ( 1984 ) .
this is a plane parallel approximation for steady shocks and has a kinematic character , assuming the overall dynamics of the system to be determined in an independent fashion .
it is therefore no surprise that the estimated spectra can only give a qualitative indication of the true situation of a dynamical system of basically spherical symmetry .
the results were also presented at this workshop by baring et al .
( 1997b ) .
their emphasis is on @xmath0-ray cutoffs and the @xmath0-ray spectral curvature due to the hardening of the spectrum for successively higher energy particles .
the first nonlinear and fully time - dependent kinetic models for particle acceleration in snrs have been presented by berezhko et al .
these authors numerically solved the full combined dynamics of thermal plasma and crs , using a phenomenological injection model and the bohm limit for the scattering mean free path .
the only limitation of this calculation is the assumption of a parallel shock everywhere , despite the restriction to spherical symmetry . such an approximation clearly can not describe asymetries like that of the x - ray continuum emission from sn 1006 .
whether a calculation of the full angular dependence might even imply dynamical instabilities due to the variations in injection rate and acceleration efficiency with angle is unknown . in the following discussion we shall ignore this potential difficulty and assume spherical symmetry .
the resulting solutions show strong shock modifications with increased compression ratios , and highly efficient nucleon acceleration for moderate injection rates .
the cr spectra at the shock are hardening towards higher energies .
the _ spatially integrated overall snr spectrum at late times _ is again an approximate single power law @xmath54 , but harder than the corresponding test particle spectra , since now @xmath55 .
we shall apply this model to the @xmath0-ray emission of snrs in the next section .
the first detailed calculation of the @xmath0-ray emission from snrs is due to dorfi ( 1991 ) .
he applied the 2-fluid model , adopting the @xmath1 - decay @xmath0-ray emissivities appropriate for the observed nucleon spectrum in the ism from higdon and lingenfelter ( 1975 ) , for the 100 mev energy region .
the result was that the @xmath0-ray fluxes from nearby snrs at @xmath19 kpc distance were near the lower limit of detectability by the cos - b satellite . using simplified 2-fluid models for the particle acceleration , drury et al .
( 1994 , referred to as dav in the following ) also calculated the @xmath0-ray emission , concentrating on the variations with energy and discussing in particular the background problems for its detection ( see also naito and takahara , 1994 ) .
dav derived the emissivities for various spectral indices @xmath56 , appropriate for the spectra from cr sources , and pointed out that the diffuse galactic @xmath0-ray background emission at energies @xmath57 decreases with an integral spectral index @xmath58 , corresponding to the diffuse galactic cr energy spectrum , whereas the emission from the sources like snrs should decrease considerably slower , with @xmath59 between 1.1 and 1.3 , according to the test particle results .
they concluded that observations of snrs against the strong diffuse galactic @xmath0-ray background should best be done in the tev region of currently operating imaging atmospheric cherenkov telescopes ( iacts ) . at energies above 100 mev
, corresponding to satellite instruments like cos - b or even egret , the @xmath1- decay @xmath0-ray emission from snrs would as a rule be drowned in the diffuse galactic background which also would explain the unsuccessful search for snrs in @xmath60 mev @xmath0-rays so far ( lebrun et al . , 1985 ; bhat et al . , 1985 ) ;
unfortunately this is still true today despite repeated efforts in this direction ( e.g. osborne et al . , 1995 ) .
it was also pointed out that also in the tev range a detection would be difficult , even for the best present instruments .
their critical parameter is @xmath61 , where @xmath62 denotes the fraction of total sn energy that is in crs at any given time during snr evolution .
since typically @xmath63 , we have @xmath64 or possibly considerably smaller , even for close - by snrs . then the flux is indeed marginal for a typical source that extends over about 1 degree . for @xmath65 and @xmath66 ,
the flux of tev @xmath0-rays exceeds @xmath67 at @xmath68 .
the relatively large angular diameter of snrs makes them usually _ extended sources _ , making background rejection by an iact difficult . in calculating the @xmath0-ray emissivity
the dav model effectively assumes that the accelerating particles are distributed _ uniformly _ across the snr for all energies .
then the @xmath0-ray energy spectrum is that of a uniform nucleon energy spectrum and for @xmath69 it is proportional to the nucleon spectrum . for
the spatially integrated @xmath0-ray luminosity this concentrates all dynamical effects into the single efficiency function @xmath70 for all @xmath0-ray energies . in general of course , the particle diffusion length inside the remnant depends on both e and t. according to dav , @xmath71 is very small during the sweep - up phase , i.e. for very young snrs with @xmath72 , where @xmath73 denotes the external mass density , and @xmath74 a characteristic ejecta velocity ( fig .
the shock radius @xmath75 still increases linearly with time in this phase .
the @xmath0-ray flux reaches a very flat maximum during the subsequent sedov phase with a slow decrease therafter .
the shortcoming of the dav theory lies in its acceleration aspect .
therefore it is important to substitute the 2-fluid model by a kinetic acceleration model .
this has been done by berezhko and vlk ( 1997a ) , based on the theory of berezhko et al .
( 1994 ) described before .
in addition , the _ distribution of the sn ejecta velocities _ ( chevalier and liang , 1989 , and refs . therein ) was taken into account .
in fact , the fastest ejecta leading the expanding sn material have much higher velocities than those represented by the mean ejecta speed @xmath76 . the very high shock velocities which result from this differential sweep - up at early times @xmath77 increase the acceleration efficiency strongly , with a harder overall proton momentum spectrum at @xmath77 than at @xmath78 ( fig .
4a ) , and the @xmath0-ray flux rises much faster during the sweep - up phase than in the dav model at a given @xmath0-ray energy ( fig .
5 ) . at @xmath79
the flux exceeds that of dav by a factor of several for the same parameters
@xmath80 due to a harder particle spectrum , an enhanced cr energy @xmath81 , and a larger compression ratio relative to the hydrodynamic approximation . on the other hand
, after a few times @xmath82 the kinetic @xmath0-ray flux falls off roughly linearly with time , by a factor of about 10 over a time period 10 times larger than the time of maximum during the subsequent sedov phase .
[ cols="^,^ " , ] in an assumed steady state for a strong shock , the necessary existence of an upper cutoff in the momentum distribution implies the escape of particles at this upper cutoff . in a time
dependent situation that may not be the case .
nevertheless there might always be a loss of particles at the upper cutoff due to insufficient wave generation there .
another cause for particle losses may be given by purely geometrical conditions , for example in acceleration at a spatially limited shock , like a planetary bow shock in the solar wind .
finally , in a snr shock the ionization of the upstream circumstellar medium may be insufficient if the gas density is high , despite the ionisation of a radiative precursor by the hot snr gas . as a result wave damping due to ion - neutral friction
may inhibit wave excitation for very high energy particles ( e.g. vlk et al .
, 1981 ; draine and mckee , 1993 ; drury et al . , 1997 ) .
already in moderately dense clouds ( @xmath83 ) this implies a cutoff well below 1 tev .
gas clouds in the environment of a snr may be partially shielded if their magnetic field is swept back by the wind from the massive progenitor star .
the snr will then engulf the cloud but not necessarily produce a great intensity of energetic particles that could illuminate the cloud along the magnetic field lines ( fig .
the shock transmitted into the cloud will be rather slow and may therefore not produce vhe particles while crossing the cloud .
thus , even a physically interacting cloud may not necessarily lead to an enhanced @xmath84-decay @xmath0-ray luminosity to the extent calculated by aharonian et al .
( 1994 ) .
if the sn progenitor is a massive star of a mass @xmath85 , it will strongly modify its environment by its supersonic mass loss .
combined with the high rotation velocity , the wind will draw out the stellar magnetic field into an archimeden spiral geometry ( parker , 1958 ) . at a typical distance of @xmath86 1
pc the radial wind flow will adapt to the exterior pressure in a termination shock beyond which an expanding hot bubble of rarefied shocked stellar wind material creates a dense shell of swept up , cooling interstellar gas ( fig .
the same effect exists for lower mass sn progenitors , but quantitatively it becomes only important for massive stars .
when the star subsequently explodes as a sn , particle acceleration will be extremely rapid initially ( vlk and biermann , 1988 ) due to the high b - field @xmath86 1 gauss of stellar origin .
possibly this will be true even at later times ( biermann 1993a , 1993b ) .
since also the gas density in the wind is very high initially , @xmath87 , the @xmath84- decay @xmath0-ray emission may also be very large at early times as pointed out by kirk et al .
( 1995 ) . the snr evolution in this complex wind / bubble structure has been modelled in the kinetic approximation ( berezhko and vlk , 1997b ) , again assuming acceleration at a parallel shock while maintaining overall spherical symmetry .
roughly speaking , the @xmath0-ray luminosity is indeed very large in the wind during an early episode of a few weeks to months , but then it decreases to quite a low value in the shocked wind bubble until the dense shell is reached after @xmath88 years .
subsequently the high external gas density leads to a high @xmath0-ray luminosity which persits until the shock weakens and the accelerated particles leave the remnant .
the result is a complex time structure of the @xmath0-ray flux as opposed to the case of a snr in a uniform ism , and the observability of such an object will depend critically on the phase of its development .
thus non - detection of such objects in @xmath0-rays at the present time , even though they may look spectacular in other wavelengths , is not a sufficient argument against acceleration .
we may have looked at it simply at the wrong time !
however , this is not the whole story yet . in a wind bubble
the magnetic field is largely azimuthal except in the very polar regions , near the stellar axis of rotation .
therefore not only injection should be strongly inhibited but possibly also the acceleration itself .
the above model calculations should then only constitute an upper limit to the true @xmath0-ray luminosity at any given time .
we should perhaps conclude that massive stars ( and their demise as sn ) which are amongst the most impressive fireworks in the sky , are not necessarily the clearest fingerprints of nonthermal processes in snrs in general .
in addition these sn explosions are likely to leave a pulsar or a black hole behind .
objects like @xmath0-cygni or ic 443 may well be in this category : too complex and possibly too confused to serve as templates for cr acceleration .
after this complex digression to uncertainties in the theoretical picture and to astronomical difficulties we should now return to comparatively safe grounds . these are sne in a uniform ism .
sn ia without obvious interaction with the environment are the natural @xmath0-ray candidates : old stellar objects of low mass , disintegrating completely .
two nearby remnants are of special interest here .
sn 1006 has been anounced as a detected tev source by the cangaroo group during the durban conference just before this workshop .
tycho s snr may be a similar object yet with different characteristics , even though only an upper limit exits ( lessard et al . , 1997 ) .
let us first discuss the @xmath84- decay @xmath0-ray flux . in order to determine the a - parameter we follow mastichiadis and de jager ( 1996 , and references therein ) who take @xmath89 kpc ,
@xmath90 erg , and @xmath91 @xmath92 , which gives @xmath93 , see table 1<3.1 $ ] , a diameter of 30 arc min , and a radio spectral index of 0.6 .
the uncertainties in the product @xmath94 from x - ray data correspond at least to a factor of 2 due to the general difficulties of modelling the _ thermal _ x - ray emission of snrs . ] . with an integral spectral index @xmath95
this results in a theoretically predicted flux of @xmath96 .
translating this to 3 tev with a spectral index @xmath97 gives @xmath98 , ( fig .
the observed cangaroo flux is @xmath99 . since in fact
not only the nonthermal x - rays but also the tev @xmath0-rays appear to come from only parts of the snr , one might increase the ratio of observed to predicted flux by about 25 percent to a value of about 22 .
even though sn 1006 should have essentially reached the sedov phase , we may nevertheless take @xmath100 and @xmath101 , to obtain ( in this case ) an upper bound of @xmath102 , ( fig .
6 ) . reducing this flux by 25 percent , as before ,
gives @xmath103 .
such an upper bound still disregards the possibly higher gas compression ratio between the possibly remaining ejecta and the shock which might exceed that assumed by dav by a factor up to @xmath104 .
yet this additional factor is not certain and should in any case not explain the above minimum discrepancy by a factor of @xmath105 .
thus , it does not seem possible to explain the observed flux by nucleonic @xmath0-rays without significantly revising the parameters @xmath106 , which are independent of acceleration theory .
of course such a revision can not be excluded , given the uncertainties .
on the other hand , the strong nonthermal x - ray emission is suggestive of an ic explanation for sn 1006 .
this presupposes a fairly low magnetic field , fitted to the radio and x - ray spectra as @xmath107 by mastichiadis and de jager ( 1996 ) , for a gyro factor @xmath108 .
probably @xmath109 and then @xmath110\ll 16 $ ] . such low values in a snr
may be entirely possible if we remember that the x - ray emission most likely comes from those regions where the shock normal is roughly parallel to the external magnetic field of several @xmath111 g . using the fit to the radio and x - ray spectra , mastichiadis and de jager ( 1996 )
applied the simple electron acceleration model of mastichiadis ( 1996 ) to obtain an ic @xmath0-ray flux that roughly agrees with the observed cangaroo flux provided we choose @xmath112 , which is an acceptable value .
clearly , more detailed modelling is needed , but our preliminary conclusion is that sn 1006 is an electron ic source in tev @xmath0-rays .
this 425 old snr is probably approaching the end of the sweep - up phase ( tan and gull , 1985 ) .
for an assumed external density @xmath113 , sn energy @xmath114 ( smith et al . , 1988 ) , and distance @xmath115 kpc ( heavens , 1984 ) , one obtains @xmath116 , similar to sn 1006 . with @xmath95
this gives @xmath117 which for @xmath118 gev lies a factor @xmath119 below the upper limit of the whipple observation ( lessard et al .
, 1997 ) . for @xmath120 and @xmath121 , which is an entirely reasonable choice for the evolutionary phase of tycho
, this factor reduces to @xmath122 , and if we would assume a postshock compression ratio @xmath86 twice that of dav , whipple should just have missed a detection of tycho ! ( see fig .
4 ) with an angular diameter of 8 arc min , tycho s snr is essentially a point source for tev @xmath0-ray astronomy with consequent advantages for the @xmath0-hadron seperation . if the dominance of emission lines in the x - ray spectrum ( becker et al . , 1980 ; petre et al . , 1993 ) can be taken as an indication that the nonthermal electron component is very small , then a detection of tycho s snr in tev @xmath0-rays may indeed be the first detection of nuclear @xmath0-rays in a cr source .
and there may not be many more such sources in the small sample of nearby candidates available .
thus every effort to find tev @xmath0-rays in tycho should be made .
at the same time , it is clear that only @xmath0-ray _ spectroscopy _ together with detailed hard x - ray continuum observations will ultimately enable us to clearly distinguish a @xmath84-decay source from an ic @xmath0-ray source .
the conclusions from this discussion can be formulated quite concisely and we will simply do this here : * snrs fall into two major categories regarding tev @xmath0-ray emission * * snrs in a uniform ism and with progenitor masses @xmath123 are rapidly rising and very slowly decreasing tev @xmath0-ray sources , except at very high ism densities , where the upper cutoff energy may not reach present threshold energies of a few hundred gev , either because the snr becomes too soon too weak to generate particles of sufficiently high energies , or because the ionisation in the upstream gas is too low to allow strong wave excitation . * * snrs in wind bubbles of massive progenitor stars @xmath124 should have very low @xmath0-ray emission ( except at very early times ) until reaching the swept - up shell of interstellar gas . the average magnetic field geometry is unfavourable for efficient acceleration in the first place . * the theoretical estimates show that it is difficult but not impossible to detect nearby snrs in @xmath0-rays .
these conclusions have been dramatically confirmed by the recent detection of sn 1006 .
we interpret the present upper limits either as the result of too short observation times ( tycho ) or of an unfavourable evolutionary phase in a massive star snr . *
the ambiguity of an electronic vs. a nuclear origin of the @xmath0-ray emission from snrs can only be resolved by detailed spectrocopy , both in the tev and the hard x - ray regions .
this observational effort needs to be accompanied by detailed modelling of the synchrotron and ic emission characteristics of the sources . * as in all other field of astronomy , detailed physics conclusions are only possible in a _ multi - wavelength approach_. for snrs this implies especially a reliable determination of such basic parameters like distance , total explosion energy , and ambient density .
i would like to thank markus he , evgeny berezhko , michail malkov and felix aharonian for informative discussions on snr physics , @xmath0-ray emission , and acceleration theory .
allen , g.e . ,
geohane , j.w . , gotthelf , e.v .
, petre , r. , jahoda , k. , rothschild , r.e . ,
lingenfelter , r.e .
, heindl , w.a . , marsden , d. , gruber , d.e .
, pelling , m.r . , blanco , p.r .
1997 , to appear in apj letters | a number of nearby northern hemisphere shell - type supernova remnants ( snrs ) has been observed in tev @xmath0-rays , but none of them could be detected so far .
this failure calls for a critical reevaluation of the theoretical arguments for efficient particle acceleration and resulting @xmath0-ray emission of snrs which are presumed to be the sources of the galactic cosmic rays .
we first discuss diffusive shock acceleration in shell - type snrs .
observational upper limits are compared with theoretical predictions for the @xmath0-ray flux and found to be roughly consistent . as a next step
the empirical arguments from the observations of x - ray power law continua for inverse compton @xmath0-ray emission at tev energies due to electrons are contrasted to the nucleonic @xmath1 - decay emission from the same objects .
emphasis is given to the possible problems for vhe @xmath0-ray production due to the environmental conditions a sn progenitor finds itself in .
finally , a point is made for the simplest case of sne ia , expected to explode in a uniform circumstellar medium .
in particular the very recently detected southern hemisphere remnant of sn 1006 is compared with tycho s snr .
on the basis of the assumed parameters for the two remnants we argue that the tev emission from sn 1006 is dominated by inverse compton radiation , whereas tycho could very well be predominantly a @xmath1 - decay @xmath0-ray source .
* particle acceleration and gamma - ray production in shell remnants + * h.j.vlk + _ max - planck - institut fr kernphysik , heidelberg , germany + _ |
in this talk , we will discuss high - order corrections to heavy quark production in qcd . to get a better confidence in our ability to predict cross sections for various heavy quark reactions , including open flavor and quarkonia production ,
we need to gain control over a class of corrections associated with what is sometimes called partonic threshold . this talk will describe the nature of these corrections , and report on the progress of the past few years @xcite in their resummation to all orders in perturbation theory .
corrections due to resummation may turn out to be small , in which case our confidence in low - order perturbative cross sections should increase , or , they may turn out to be large , and may afford sensitivity to qcd dynamics in a regime where all orders of perturbation theory are relevant .
it is possible that both scenarios apply in different cross sections , and/or in different kinematic ranges of the same cross section .
inclusive heavy quark production is just one example of a qcd process at large momentum transfer
. we may consider the class of cross sections in which we sum over all final states that include any heavy system @xmath2 , which can only be produced by a short - distance process in partonic scattering .
suppose for simplicity that the total mass @xmath3 of the system @xmath4 is of order @xmath5 , the total ( hadronic ) center of mass energy , and that the rapidity @xmath6 of the produced system is not a large parameter .
such a cross section can be expressed to leading power by the factorized expression & = & _ ab _
q^2/s^1dz _ a / a(x_a,^2 ) _ b / b(x_b,^2 ) + & & ( z - q^2 x_ax_bs ) _ abfx(z , e^-2yx_ax_b , q , q , _
s(^2 ) ) , [ basicfact ] which is illustrated in fig .
[ cutcross ] . the @xmath7 s
are parton distributions ( in some factorization scheme , like dis or @xmath8 ) , while @xmath9 is a partonic hard - scattering function , which at lowest order is the born cross section for @xmath10 , = _
born+_s^(1 ) + . [ sigmaborn ] these relations
may be written down for both unpolarized and polarized cross sections .
corrections in either case are power - suppressed @xcite ; beginning at @xmath11 for unpolarized or longitudinally polarized cross sections , and at @xmath12 for transversely polarized cross sections . in eq .
( [ basicfact ] ) , the limit @xmath13 is associated with corrections that tend to enhance the cross section .
these enhancements are the object of threshold resummation .
we will discuss below the origin of this terminology , why such corrections can sometimes be large , and how they can be resummed to all orders .
much of the following discussion follows ref .
@xcite .
the kinematics of the partonic process require that @xmath14 , so that @xmath15 in eq . ( [ basicfact ] ) . at @xmath16 ,
the partons have just enough energy to produce the observed final state , with no extra hadronic radiation .
this is what we shall refer to as the elastic limit @xcite , or partonic threshold .
it is important to distinguish partonic threshold from the usual concept of a threshold . in particular , in heavy quark production , we shall assume that the heavy quarks of mass @xmath17 are produced with nonzero velocity @xmath18 , and hence with a total invariant mass @xmath19 .
thus , only for @xmath20 does partonic threshold coincide with true threshold . for the drell - yan production of highly relativistic lepton pairs with @xmath21 ,
partonic threshold still refers to @xmath16 , and is a source of potentially large corrections .
typical hard - scattering cross sections @xmath22 in eq .
( [ basicfact ] ) are distributions in the variable @xmath23 rather than simply functions of @xmath23 , because they include contributions from virtual as well as real gluons .
we are interested in a class of potentially large , positive corrections due to such distributions , which occur in all @xmath24 .
we next explain in what sense they are
large " , and why they are positive to all orders . at order @xmath25 , the leading logarithmic distributions in eq.([basicfact ] ) are of the form ( suppressing color factors , which depend on parton type ) -_s^nn ! _ + , [ lln ] whose integral with a smooth function @xmath26 ( such as the convolution of parton distributions in eq .
( [ basicfact ] ) ) is defined as -_s^nn!_0 ^ 1dz ^2n-1((1-z)^-1 ) & & + & & = _
s^nn!_0 ^ 1dz(1 ) ^2n-1((1-z)^-1)+ ~_s^nn!f(1)(2n-1)!+ , [ factorial ] where we have kept only the first term in the expansion of @xmath26 about @xmath16 .
it is evident that such terms give , at least formally , contributions that grow even faster than @xmath27 at @xmath28th order .
if they had alternating signs , these contributions might add up to a finite number somehow , but they are all of the same sign . the signs in eq .
( [ factorial ] ) come directly from the manner in which hard - scatter - ing cross sections are computed .
the fully inclusive drell - yan cross section @xmath29 illustrates the situation .
the convolution in eq .
( [ basicfact ] ) factors into simple products of functions for each pair of initial - state partons @xmath30 and @xmath31 , under moments with respect to @xmath32 , ^dy_ab(n , q ) & = & _ 0 ^ 1 d^n-1 + & = & _ ab _ a / a(n , ) _
ab^dy(n , q , ) _ b / b(n , ) . identifying @xmath33 as the contribution to the cross section from the parton combination @xmath34 ,
ab^dy(n , q , ) = _ i = a , b 1 _ 0 ^ 1 d ^n-1d^ab_ab()dq^2 . neglecting parton labels ,
this is the ratio of the moment of a physical cross section to a product of moments of parton distributions . because @xmath9 is , by construction , dependent only on short - distance behavior ,
the ratio may be computed in perturbation theory , as illustrated schematically in fig .
[ mtratio ] ( for the dis scheme ) .
the numerator is a moment of the perturbative partonic drell - yan cross section , while the denominator is the product of moments of two perturbative parton distributions . for quark - antiquark processes ,
the parton distributions are the same , so the denominator is the square of squared partonic amplitudes , summed over final states . at each order , both the numerator and denominator in fig .
[ mtratio ] have double - logarithmic terms like eq .
( [ lln ] ) . before moments ,
the perturbative drell - yan cross sections include distributions in @xmath35 , like eq .
( [ lln ] ) above , while in the deeply inelastic scattering cross section the same sort of terms depend on bjorken @xmath36 through @xmath37 .
after moments , both give double - logarithmic @xmath38 at @xmath28th order , with @xmath39 the moment variable .
these leading logarithms are the finite remainders of corrections from @xmath28 pairs of real and virtual gluons that attach to the scattered quarks in dis and the annihilating pair in dy . in the denominator ,
each dis parton distribution , which is itself of the form of a cross section , has both incoming _ and _ outgoing quarks , while in the numerator , dy involves incoming quarks only . on physical grounds ,
the numerator must vanish as @xmath40 and the denominator must vanish as @xmath41 .
the reason for this may be seen by recalling the relations of @xmath23 and @xmath36 to the invariant mass @xmath42 of hadrons in the final state for the two cases : & : & w^2q^2(1-z)^2 , + dis & : & w^2~q^2(1-x)/x .
[ wsquared ] the limits @xmath40 and @xmath41 thus both correspond to nearly elastic scattering @xcite : for drell - yan , the annihilation of a quark pair into an electroweak vector boson , for dis , the scattering of a quark into a nearly massless jet of particles .
it is not difficult to verify that in these limits , the partonic cross section is suppressed @xcite . indeed , in these limits , we expect the coherent scattering of hadronic bound states , whose contributions are normally suppressed by a power of @xmath3 compared to incoherent partonic scattering , to dominate @xcite .
the outgoing quarks give an extra suppression in dis when the hard - scattering function @xmath9 , fig .
[ mtratio ] , is computed in perturbation theory .
that is , the product of dis denominators is suppressed even more than the single dy numerator .
then the ratio grows with moment @xmath39 , from the elastic limit in @xmath23 space .
this is the source of the terms shown in eq .
( [ lln ] ) in @xmath9 , and is the reason why they all have the same sign .
the drell - yan cross section is the benchmark for the resummation of singular distributions .
as above , singular distributions at @xmath16 translate into logarithms of the moment variable @xmath39 .
logarithms of @xmath39 ( to all logarithmic order , not just leading or next - to - leading logarithm ) in the moments of the inclusive drell - yan cross section exponentiate @xcite , _
q|q^dy(n , q,)=_born(q ) e^c(_s)+e(n , q,,_s ) , [ dyexp ] where @xmath43 stands for @xmath44 . in the exponent ,
the function @xmath45 is known to two loops , while the function @xmath46 , which organizes all logs of @xmath39 , has the following form in the @xmath8 scheme , e(n , q,,_s ) & = & -_0 ^ 1dx .
[ dyexponent ] the functions @xmath47 and @xmath48 are finite series in @xmath43 @xcite , given for @xmath8 by g_1(_s)= 2c_f ( _ s + 12k ( _ s)^2)+ , g_2(_s ) = o(_s^2 ) , [ g12def ] where @xcite k = c_a(6718-^26 ) -59n_f .
( [ dyexp ] ) and ( [ dyexponent ] ) resum all logarithms of @xmath39 in the sense of an order - by - order expansion , by reexpanding the running couplings in terms of @xmath43 .
the resummed integrals , however , are ill - defined for @xmath41 , no matter how large @xmath49 is , since the perturbative running coupling diverges at @xmath50 .
such a divergence is sometimes called an infrared renormalon , and may give hints on the structure of power corrections to eq .
( [ basicfact ] ) .
most importantly , it is possible to regulate the singularity in the running coupling without affecting the leading power behavior of the resummed cross section @xcite .
different approaches to this problem can give somewhat different predictions @xcite , but in any reasonable prescription the threshold corrections remain relatively modest , although nonnegligible in at least some physically relevant regions . it has been noted in several phenomenological applications that threshold resummation , and even fixed - order expansions based upon it , significantly reduces sensitivity to the factorization scale @xcite .
the reason for this reduction is easy to see in the dependence on the factorization scale , @xmath51 in eq .
( [ dyexponent ] ) .
indeed , for the drell - yan , and all related cross sections , the dependence on @xmath51 in the hard - scattering function matches that part of the scale dependence of the @xmath8 distributions due to the singular part of the splitting functions @xmath52 .
that is , @xmath47 in eq . ( [ g12def ] )
is exactly the coefficient of @xmath53 in @xmath54 .
because resummation matches phase space to all orders at partonic threshold @xcite , all associated factorization scale dependence is also matched . to be specific , we may write the moments of the drell - yan cross section in a form that resums the singular threshold distributions , ^dy_ab(n , q ) & = & _ q _ q / a(n , ) _ q|q^dy(n , q , ) _
|q / b(n , ) + & = & _ q _ q / a(n , ) m_q(,q ) _
q|q^dy(n , q , q ) _
|q / b(n , ) m_|q(,q ) + & & + o(1/ n ) , where in the second line the factors @xmath55 absorb the @xmath51-dependence in the exponentials of eq . (
[ dyexponent ] ) .
these factors compensate for the singular part of the evolution of the parton distributions , and the @xmath51-dependence of the resummed expression is suppressed by a power of the moment variable , = o(1/ n ) .
of course , the importance of the remaining sensitivity to @xmath51 depends on the kinematics .
beyond leading logarithms , there are important differences between the electro - weak - induced drell yan cross sections and the qcd - induced top or jet cross sections .
these are due primarily to the presence of final - state radiation from scattered quarks in the latter case , which is absent in the former , and to the interplay of color exchange in the hard scattering with the soft radiation . as in the case of eq .
( [ dyexp ] ) , resummation is based first of all on the factorization properties of the cross section in the neighborhood of the elastic limit @xcite .
the situation is illustrated in fig .
[ hs ] . near the elastic limit ,
all gluons emitted into the final state have energies limited by @xmath56 .
correspondingly , gluons with energies of order @xmath3 can appear only in virtual diagrams .
standard factorization methods may then be used to separate the relatively soft but still perturbative gluons from the underlying hard scattering .
this may be done order - by - order in perturbation theory , keeping both the hard and soft factors of @xmath9 free of soft and collinear divergences .
the process of factorization may be thought of as the construction of an effective field theory for soft gluons in the presence of the hard scattering @xcite . in the relevant effective field theory , the incoming partons that annihilate into the heavy quarks and the outgoing heavy quarks themselves
are represented by ordered exponentials ( wilson lines ) .
the wilson lines are tied together in the amplitude and its complex conjugate at local vertices , @xmath57 and @xmath58 in fig .
[ hs ] , which describe the flow of color between the initial and final states .
indices @xmath59 and @xmath60 label matrices in color space .
we will consider here the annihilation of light quarks ( color indices @xmath61 and @xmath62 ) into heavy quarks ( indices @xmath63 and @xmath64 ) , q_a_1(p_a)+|q_a_2(p_b ) q_a_4(p_1 ) + |q_a_3(p_2 ) , [ qqqq ] with kinematic invariants , t_1=(p_a - p_2)^2-m^2 , u_1=(p_b - p_2)^2-m^2 , s=(p_a+p_b)^2 . in this case
a convenient basis for the @xmath65 s is one that represents color singlet and octet exchange in the @xmath66-channel , ( t_1 ) _ \{a_i } = _ a_1a_2_a_3a_4 , ( t_2 ) _ \{a_i } = _ c ( t_c^(f))_a_2a_1(t^(f)_c)_a_4a_3 .
[ tdefs ] other bases , particularly singlet exchange in the @xmath66- and @xmath67- channels , are also interesting @xcite . as in fig .
[ hs ] , each choice of effective vertices leads to a separate soft function @xmath68 , which depends on @xmath69 only , rather than @xmath3 itself .
the ( virtual ) hard - scattering functions , @xmath70 and @xmath71 , are labelled by the same color exchange indices . as in most factorizations and constructions of effective field theories ,
the new vertices require renormalization .
thus we renormalize the soft functions @xcite , s_ji^(un)((1-z)q)= z^_jj s_ji^(ren)((1-z)q)z_ii , [ sren ] and the hard functions h_i^(un)(q)=z^-1_ijh_j^(ren)(q ) , [ hren ] where the @xmath72 are cutoff - dependent renormalization constants .
solutions to the renormalization group equation for @xmath68 that follows from ( [ sren ] ) are generally ordered exponentials @xcite , but at leading logarithm in @xmath68 , which is next - to - leading logarithm in the overall cross section , we can diagonalize the anomalous dimension , and separate the evolution of particular linear combinations of composite color vertices @xcite .
we can apply these results to the production of a pair of heavy quarks with total invariant mass @xmath73 at fixed rapidity @xmath6 .
the cross section is , as usual , a convolution of hard - scattering functions @xmath74 with parton distributions @xmath75 and @xmath76 , as in eq .
( [ basicfact ] ) , with @xmath77 .
corresponding to the drell - yan result , ( [ dyexp ] ) , we now have , to next - to - leading logarithmic accuracy , choosing @xmath78 , _
q|qq|q(n ) = _
ijs^(0)_ji h_i(q)h^*_j(q ) e^c_ji(_s)+e_ji(n,_s ) , [ qqqqexp ] where again @xmath43 stands for @xmath79 . the function @xmath80 is known to one loop only at this time . to next - to - leading log ,
we need only the lowest - order soft functions , @xmath81 .
the function @xmath82 , which contains the logs of the moment variable @xmath39 has a form very similar to the drell - yan case , but now with a dependence on the effective color vertices , through a third function , @xmath83 , e_ji^(ab)(n,_s ) & = & -_0 ^ 1dxx^n-1 - 11-x .
[ qqqqexponent ] as before , the @xmath84 , @xmath85 are finite functions of their arguments .
in the @xmath8 scheme , @xmath47 and @xmath48 are given for incoming light quarks by ( [ g12def ] ) above .
dependence on color exchange in the hard scattering is contained entirely in the new functions @xmath86 . to determine @xmath86 , we go to a color basis that diagonalizes the renormalization matrix @xmath87 in eqs .
( [ hren ] ) and ( [ sren ] ) . in this basis ,
@xmath88=-\lambda_i[\alpha_s]\ , , \ ] ] where the eigenvalues @xmath89 of the corresponding anomalous dimension matrix are complex in general , and may depend on the directions of the incoming and outgoing partons .
the anomalous dimension matrix of the effective vertices @xmath57 in fig . [ hs ] for light to heavy quark annihilation in the singlet - octet basis of eq
. ( [ tdefs ] ) @xcite is : _
s= ( cc - c_f(l_+1+i ) & ( ) + & + 2 ( ) & c_f + & + ) .
[ gammaoneloop ] here @xmath90 is the vertex function in the eikonal approximation for the production of a pair of heavy quarks with center of mass velocity @xmath18 , l_= ( + i ) , = . solving for the eigenvalues , substituting them in eq .
( [ qqqqexp ] ) , and expanding the result to first order in @xmath43 , we can derive an explicit one - loop expression for the cross section for heavy quark production through light quark annihilation @xcite .
this result is consistent @xcite with the explicit one - loop formulas given in @xcite for the single - particle inclusive cross section .
much the same considerations apply to electroproduction of heavy quarks .
it is interesting to note that for the process @xcite ^*(p_a)+g_i(p_b ) q_a_4(p_1 ) + |q_a_3(p_2 ) , as for direct photon production @xcite there is only a single color tensor , coupling the produced pair to the gluon in an octet state . as a result , there is only one eigenvalue , which in this case is given ( in the conventions of ref .
@xcite ) by _ 8 = ( c_a2 + ( 12c_a - c_f ) [ l_+ 1 ] ) . resolved processes in photoproduction , however , involve nontrivial color mixing @xcite . in the limit @xmath91 , we can make contact with quarkonia production in the nonrelativistic qcd ( nrqcd ) formalism @xcite .
the use of resummation as a tool for heavy - quark and other hard - scattering cross sections may help to put the successes and shortcomings of fixed - order perturbative qcd into a larger context .
the reduction of factorization scale dependence is an encouraging observation in this direction
. it will be interesting to study as well renormalization scale independence @xcite .
other directions for investigation include the inversion of the moments and the closely - related issue of power - suppressed corrections .
the development of resummed polarized cross sections , for which next - to - leading order calculations are becoming available @xcite , is sure to be a further source of insight .
the treatment of scale - independence in this talk grew out of discussions with werner vogelsang .
we thank eric laenen and jeff owens for their collaboration in the development of some of this work , and stan brodsky and paolo nason for helpful conversations . this work was supported in part by the national science foundation , under grant phy9722101 , and by the u.s .
department of energy .
99 j .- w . qiu and g. sterman , in proceedings of _ polarized collider workshop _ , university park , pa , nov 15 - 17 , 1990 , ed .
j. collins , s.f .
heppelman and r.w .
robinett , american inst.phys . , 1991 ( aip conference proceedings , 223 ) p.249 ; ; _ ibid .
_ * 378 * , 52 ( 1992 ) . | we review some techniques of resummation applied to heavy quark production in hadronic scattering and electroproduction near partonic threshold , and discuss the reduction of factorization scale dependence in resummed cross sections .
= cmr8 # 1#2#3#4#1 * # 2 * , # 3 ( # 4 ) itp - sb-99 - 27 * resummation for heavy quark production + near partonic threshold * gianluca oderda@xmath0 , nikolaos kidonakis@xmath1 , george sterman@xmath0 _ @xmath0institute for theoretical physics + state university of new york at stony brook + stony brook , ny 11794 - 3840 _ _ @xmath1department of physics + florida state university , tallahassee , fl 32306 - 4350 _ |
recently the atlas and cms collaborations have announced the discovery of a new boson with the approximate mass of 125 gev , and suggested that this might be the long sought higgs boson @xcite .
if this is confirmed by future data , it will be extremely important to accurately measure the higgs coupling to the particles of the standard model ( sm ) .
the @xmath4 and @xmath5 couplings are of particular importance , because of the role they play in the unitarization of longitudinal weak boson scattering amplitudes . recall that for a given @xmath6 scattering amplitude @xmath7 the @xmath8-th partial wave projection is defined by @xmath9 where @xmath10 is the cosine of the scattering angle , and @xmath11 is the @xmath8-th legendre polynomial . for any value of @xmath8 , unitarity demands @xmath12 to lie on a unit circle centered on @xmath13 , the argand circle .
in particular this implies the bounds @xmath14 consider for instance the @xmath15 scattering . at large values of the center - of - mass ( cm ) energy
the amplitude grows like @xmath16 where @xmath17 is the electroweak vev , @xmath18 gev , and @xmath19 if @xmath20 is well below one , for instance , the @xmath21 projection undergoes unitarity loss at a few tevs @xcite .
if this is the case , new particles , not heavier than a few tevs , must come into play and unitarize the amplitude .
it is often assumed that a light higgs boson is associated with perturbative dynamics .
this implies that , at high energy , the partial wave projections of the longitudinal weak boson scattering amplitudes have a real part approaching a small and constant value , far from the unitarity bounds of eq .
( [ eq : unitbound ] ) .
the reason for this is that the imaginary part of an amplitude is due to loop corrections , which , in a perturbative regime , are small .
thus in order for the @xmath12 projections to lie on the argand circle , also the real part has to be small .
therefore , the terms growing like @xmath22 must vanish . if @xmath23 , this implies that new states must supply the right - hand side of eq .
( [ eq : wwh ] ) , @xmath24 in such a way that the coefficient of the term growing like @xmath22 vanishes : @xmath25 this , for example , is the scenario conjectured in @xcite and @xcite . despite the common lore
, we believe it possible for a light higgs to coexist with strong dynamics at the unitarization scale .
this occurs if the higgs boson is a composite scalar of new strong dynamics , technicolor ( tc ) , which is responsible for breaking the electroweak symmetry .
naively the natural mass scale for the lightest tc scalar singlet is of the order of 1 tev , as obtained by scaling up the lightest scalar isosinglet of quantum chromo - dynamics ( qcd ) , the @xmath26 meson .
however , in viable tc theories naive qcd scaling - up does not hold , as there are additional mechanisms which lead to a large suppression of the mass of the lightest scalar , relative to the naive estimate .
first , viable tc theories are expected to be near - conformal in a range of energies above the chiral symmetry breaking scale . to see this , recall that the sm fermions acquire mass by interacting with the tc vacuum through a spontaneously broken gauge interaction , extended technicolor ( etc ) @xcite .
this , however , introduces unwanted flavor - changing neutral currents , which can only be eliminated by raising the mass of the etc gauge bosons .
doing so would result in unacceptably small sm fermion masses , unless the tc vacuum is enhanced , from the tc to the etc scale , by near - conformal dynamics and a large anomalous dimension of the tc vacuum @xcite .
if tc dynamics feel the presence of a nearby fixed point , the mass of the lightest scalar is expected to be reduced , relative to a qcd - like spectrum .
the reason for this is that tc theories with a spontaneously broken scale invariance are expected to feature a massless scalar singlet in the spectrum , the technidilaton @xcite . in asymptotically - free tc theories
scale invariance is only approximate , and the technidilatonic higgs boson is expected to acquire a small dynamical mass @xcite . by small we mean smaller than the natural mass scale of the light technihadrons , which is of the order of 1 tev .
second , the higgs mass in tc , as well as in the sm , is modified by radiative corrections due to gauge interactions and etc - induced yukawa interactions . as in the sm , in tc
the radiative corrections to the higgs mass are quadratic in the cutoff . however , unlike the sm , a theory of technihadrons has a physical cutoff , the scale of compositeness .
the latter is estimated to be of the order of @xmath27 @xcite , where @xmath28 is the `` technipion '' decay constant , analogous to the pion decay constant of qcd .
if the fermionic sector of tc consists of @xmath29 weak `` technidoublets '' , each contributing an equal amount to @xmath30 , then @xmath31 .
since the cutoff is physical , the cutoff - dependence of the radiative corrections to the higgs mass is also physical , and , being quadratic , large .
the dominant contribution to @xmath32 is negative : it arises from the top loop , and is of the order of @xmath33 @xcite .
finally , the tc higgs can become lighter by `` geometric scaling '' , as the number of technicolors and/or weak technidoublets is increased .
the tc states become lighter as @xmath29 increases because , as mentioned in the last paragraph , the square of the electroweak vev receives contribution from @xmath29 weak pairs .
if @xmath17 is to be kept fixed at 246 gev , this forces the whole tc spectrum to scale like @xmath34 .
the scaling as a function of the number of technicolors , @xmath35 , is much more subtle .
the large-@xmath35 limit of @xmath36 was first obtained by t - hooft with technifermions in the fundamental representation @xcite .
this allows rescaling from qcd , which corresponds to @xmath37 .
corrigan and ramond noted that scaling - up qcd is also possible with technifermions in the two - index antisymmetric representation , as the latter coincides with the anti - fundamental representation for @xmath37 @xcite .
the t hooft and corrigan - ramond large-@xmath35 limits lead to qualitatively very different spectra .
in particular , only @xmath38 scalars , in the t hoof limit , with @xmath17 fixed at 246 gev , become lighter at large values of @xmath35 , whereas multiquark states do not exist at @xmath39 . on the other hand , in the corrigan - ramond limit scalar masses are always expected to scale down like @xmath40 .
large @xmath35 scaling can be obtained for other higher - dimensional representations , even though a comparison with qcd is no longer possible .
for instance , in the two - index symmetric representation scalars masses are expected to scale like @xmath41 , in the large @xmath35 limit @xcite . for strongly coupled dynamics at the unitarization scale ,
we neither expect sum rules like the one of eq .
( [ eq : sum ] ) to hold , nor the partial wave projections of longitudinal weak boson scattering amplitudes to have a small real part .
instead , we expect the latter to quickly grow ( with either an overall plus or minus sign ) and oscillate between the unitarity bounds -1/2 and 1/2 .
this is indeed the scenario of pion - pion scattering in qcd @xcite . because of this , in models with strong unitarization we expect enhancement of @xmath42 processes at the lhc , where each @xmath1 denotes either a @xmath2 or a @xmath3 boson @xcite .
here we analyze this in the context of near - conformal tc , with a light composite higgs in the spectrum , and vector meson dominance ( vmd ) .
the interesting aspect of this analysis is that it allows to probe tc for masses and coupling of the spin - one resonances lying outside the lhc reach for significant production through drell - yan ( dy ) processes .
in fact in models of vmd the coupling of the light sm fermions to the spin - one technimesons is suppressed by @xmath43 , where @xmath44 is the weak coupling and @xmath45 is the spin - one technimeson coupling . thus increasing @xmath45 reduces sensitivity to dy production @xcite .
on the other hand , increasing @xmath45 does not lead to a reduction of the cross section for @xmath42 , as the partial wave projections of longitudinal weak boson scattering amplitudes are always expected to oscillate between -1/2 and 1/2 .
furthermore , dy production of spin - one technimesons becomes increasingly kinematically forbidden as the latter become heavier . on the other hand , in @xmath42 processes the strength of the amplitude does not decrease as the technimeson masses become larger .
in fact the technimesons are expected to unitarize the weak boson scattering amplitudes , no matter what their mass is .
our goal is therefore to show that the discovery of tc through @xmath42 processes is complementary to dy production of spin - one technimesons .
the remainder of this note is organized as follows . in sec .
[ sec : unit ] we review aspects of unitarity of longitudinal weak boson scattering in near - conformal tc , with a light higgs and a technirho isospin triplet unitarizing the amplitudes . in sec.[sec : vector ] we analyze the @xmath42 processes at the lhc , and the potential for discovery of near - conformal tc . finally , in sec .
[ sec : conclusions ] we offer our conclusions .
the model we use as a template for our analysis is near - conformal tc with an @xmath46 chiral symmetry breaking pattern . by gauging the electroweak subgroup of @xmath47
this becomes a model of dynamical electroweak symmetry breaking .
the near conformal behavior of the underlying gauge theory is reflected , in the effective theory , by the presence of a light higgs , which we choose to take as the chiral partner of the technipions eaten by the @xmath2 and @xmath3 bosons , and by a near degenerate set of spin - one vector and axial - vector isotriplets .
such a model has been termed `` vanilla tc '' , and its phenomenology analyzed in @xcite .
unitarity of longitudinal weak boson scattering amplitudes in tc has instead been studied in @xcite .
we denote the normalized @xmath48 generators with @xmath49 , @xmath50 , where @xmath51 are the pauli matrices . employing a linear representation for the chiral group leads to the effective lagrangian , @xmath52 , and @xmath53
are termed @xmath54 , @xmath55 , and @xmath56 , respectively . ]
@xmath57 + \frac{1}{2}\
{ \rm tr}\left[d_\mu m d^\mu m^\dagger\right ] + m^2\ { \rm tr } \left[c_{l\mu}^2+c_{r\mu}^2\right ] \nonumber \\ & -&\tilde{g}^2\ \alpha_1\ { \rm tr}\left[c_{l\mu } m c_r^\mu m^\dagger\right ] -i\ \frac{\tilde{g}\ \alpha_2}{4}\ { \rm tr}\left[c_{l\mu}\left(m d^\mu m^\dagger - d^\mu m m^\dagger\right ) + c_{r\mu}\left(m^\dagger d^\mu m - d^\mu m^\dagger m\right)\right ] \nonumber \\ & + & \frac{\tilde{g}^2\ \beta}{4}\ { \rm tr } \left[c_{l\mu}^2+c_{r\mu}^2\right ] { \rm tr}\left[m m^\dagger\right ] + \frac{\mu^2}{2}\ { \rm tr}\left[m m^\dagger\right ] -\frac{\lambda}{4}\ { \rm tr}\left[m m^\dagger\right]^2 + { \cal l}_{\overline{\psi}\psi-{\rm resonance } } + { \cdots } \ , \label{eq : l}\end{aligned}\ ] ] where the ellipses denote higher dimensional terms , suppressed by inverse powers of the cutoff @xmath58 . here @xmath59 is the sm lagrangian without the higgs sector and the yukawa interactions , whereas @xmath60 contains the etc - induced interactions of the sm fermions with the tc vacuum ( mass terms ) and the tc resonances . the field - strength tensors @xmath61 and @xmath62 are associated to the spin - one resonance triplets @xmath63 and @xmath64 , respectively , with coupling @xmath45 . the fields @xmath65 and @xmath66 are defined by @xmath67 where @xmath68 and @xmath69 are the electroweak boson gauge eigenstates .
these definitions follow from the fact that the fields @xmath70 and @xmath71 are taken to transform like @xmath72 and @xmath69 , respectively , under the electroweak gauge group . in the limit of zero electroweak gauge couplings , and because of chiral symmetry breaking , the technimeson mass eigenstates are not @xmath70 and @xmath71 , but the vector and axial linear combinations , @xmath73 when the electroweak gauge couplings are switched on , @xmath74 and @xmath75 further mix with @xmath76 and @xmath69 to form mass eigenstates .
the spin - zero matrix @xmath77 transforms like the bifoundamental of @xmath47 , and is defined by @xmath78 \ , \ ] ] where @xmath79 are the technipions `` eaten '' by the sm weak bosons , and @xmath80 is the chiral symmetry breaking vev .
the latter is not quite equal to @xmath81 gev .
in fact the mixing term @xmath82 , which arises from the term proportional to @xmath52 in the effective lagrangian , must be removed prior to quantization .
this requires a shift of the @xmath83 fields and a tree - level renormalization of the technipions .
thus @xmath80 approaches @xmath81 gev only for large values of the axial technimeson mass , and/or small values of the coupling @xmath45 .
the precise relation between @xmath17 and @xmath80 is @xmath84 \ .
\label{eq : fp}\ ] ] upon higgsing , the spin - zero field @xmath77 contributes to the mass of the spin - one vector and axial triplets . for zero electroweak gauge couplings
this gives @xmath85 in addition to the sm parameters , the lagrangian of eq .
( [ eq : l ] ) contains five additional parameters : @xmath86 , @xmath45 , @xmath87 , @xmath52 , and @xmath53 . in a qcd - like tc theory
we could impose both weinberg sum rules ( wsrs ) @xcite on the vector and axial resonances to eliminate two of these parameters .
however in near - conformal tc we only expect the first wsr to approximately hold @xcite , which leaves us with four input parameters . in @xcite
these were taken to be the mass of the axial technimeson , the coupling @xmath45 , the lagrangian parameter @xmath53 , and the contribution to the peskin - takeuchi @xmath88 parameter @xcite from the axial and vector technimesons , @xmath89 .
the latter is given by @xcite @xmath90 \ .
\label{eq : sva}\ ] ] at high energies the longitudinal weak boson scattering amplitudes approach the scattering amplitudes of the corresponding eaten goldstone bosons @xcite .
we adopt the principle of local unitarization : at a given cm energy @xmath91 , the scattering amplitudes are unitarized by leading - order interactions , and exchanges of states with mass not heavier than @xmath91 . in our tc model
the isospin invariant amplitude is unitarized by higgs and spin - one vector exchanges , @xmath92s -\frac{g_{h\pi\pi}^2}{v^2}\frac{s^2}{s - m_h^2 } -g_{v\pi\pi}^2\left[\frac{s - u}{t - m_v^2}+\frac{s - t}{u - m_v^2}\right ] \ .
\label{eq : invariant}\end{aligned}\ ] ] here @xmath93 is the coupling of the vector technimeson to the goldstone bosons , and @xmath94 is the relevant linear combination of the higgs couplings to the eaten goldstone bosons in units of @xmath95 @xcite
. the parameters of eq .
( [ eq : invariant ] ) , @xmath96 , @xmath93 , and @xmath94 , can be computed in terms of the input parameters @xmath97 , @xmath45 , @xmath89 , and @xmath53 @xcite .
this gives @xmath98 \
, \label{eq : gvpph}\ ] ] and @xmath99 the isospin projections of the invariant amplitude are @xcite @xmath100 the goldstone boson scattering amplitudes can in turn be expressed as functions of the isospin projections @xmath101\ , \nonumber \\ & & { \cal m}(w_l^+ w_l^-\to w_l^+ w_l^- ) = \frac{1}{6}\left[2i_0 + 3i_1+i_2\right]\ , \nonumber \\ & & { \cal m}(w_l^\pm z_l\to w_l^\pm z_l ) = \frac{1}{2}\left[i_0+i_2\right]\ , \nonumber \\ & & { \cal m}(w_l^\pm w_l^\pm \to w_l^\pm w_l^\pm ) = i_2\ , \nonumber \\ & & { \cal m}(z_l z_l\to z_l z_l ) = \frac{1}{3}\left[i_0 + 2i_2\right ] \
.\end{aligned}\ ] ] for the @xmath15 scattering , and taking into account corrections due to nonzero electroweak couplings , this gives @xmath102\frac{-2(1+x)e^2}{v^2 } \
, \label{eq : wwhv}\ ] ] whence , comparing with eq ( [ eq : wwhrho ] ) , @xmath103 in @xmath15 scattering ( the quantity between parenthesis in eq . ( [ eq : wwhv ] ) ) , for @xmath104 , @xmath105 , and up to small electroweak corrections .
the blue lower curve is for @xmath106 tev , whereas the red upper curve is for @xmath107 tev .
right : the corresponding behavior of @xmath108 ( solid curves ) and @xmath109 ( dotted curves).,title="fig:",width=288 ] in @xmath15 scattering ( the quantity between parenthesis in eq .
( [ eq : wwhv ] ) ) , for @xmath104 , @xmath105 , and up to small electroweak corrections .
the blue lower curve is for @xmath106 tev , whereas the red upper curve is for @xmath107 tev .
right : the corresponding behavior of @xmath108 ( solid curves ) and @xmath109 ( dotted curves).,title="fig:",width=288 ] in general we expect the sum rule of eq .
( [ eq : sum ] ) to be violated , @xmath110 . the behavior of @xmath111 is shown in fig . [ fig : sumrule ] ( left ) , up to small electroweak corrections , for @xmath104 and @xmath105 .
the blue lower curve is for @xmath106 tev , whereas the red upper curve is for @xmath107 tev .
( [ eq : sva ] ) implies an upper bound on @xmath45 , @xmath112 which for @xmath104 is approximately @xmath113 . in fig .
[ fig : sumrule ] ( right ) we plot the corresponding behavior of @xmath108 ( solid curves ) and @xmath109 ( dotted curves ) .
it might seem unreasonable that @xmath20 , which is the @xmath4 coupling in sm units , decreases with increasing @xmath97 , since this appears to be the decoupling limit .
however this is not so , because , in addition to @xmath17 , we are also keeping @xmath89 fixed . then inspection of eqs .
( [ eq : fp ] ) and ( [ eq : sva ] ) shows that @xmath52 and @xmath80 must grow , when @xmath97 grows , and @xmath94 , from eq .
( [ eq : gvpph ] ) , decreases . since larger values of @xmath45
give a larger coefficient of the term growing like @xmath22 , we also expect the longitudinal weak boson scattering amplitudes to grow with @xmath45 .
this is shown in fig .
[ fig : unitlong ] , where the @xmath114 projections of @xmath15 ( left ) , @xmath115 ( center ) , and @xmath116 ( right ) scattering amplitudes are plotted as a function of the cm energy , up to small electroweak corrections , for @xmath104 and @xmath105 . increasing dashing size corresponds to increasing values of @xmath45 , with @xmath117 .
the blue curves are for @xmath106 tev , whereas the red curves are for @xmath107 tev .
the amplitudes increase with @xmath45 and @xmath97 , in accordance with the behavior of @xmath111 in fig .
[ fig : sumrule ] .
we therefore expect signals , in the lhc processes @xmath0 , to grow accordingly .
partial wave projections of @xmath15 ( left ) , @xmath115 ( center ) , and @xmath116 ( right ) scattering amplitudes , as a function of the cm energy , for @xmath104 , @xmath105 , and up to small electroweak corrections . increasing dashing size
corresponds to increasing values of @xmath45 , with @xmath117 .
the blue curves are for @xmath106 tev , whereas the red curves are for @xmath107 tev .
the amplitudes increase with @xmath45 , in accordance with the behavior of @xmath111 in fig .
[ fig : sumrule ] ( left).,title="fig:",width=220 ] partial wave projections of @xmath15 ( left ) , @xmath115 ( center ) , and @xmath116 ( right ) scattering amplitudes , as a function of the cm energy , for @xmath104 , @xmath105 , and up to small electroweak corrections .
increasing dashing size corresponds to increasing values of @xmath45 , with @xmath117 .
the blue curves are for @xmath106 tev , whereas the red curves are for @xmath107 tev .
the amplitudes increase with @xmath45 , in accordance with the behavior of @xmath111 in fig .
[ fig : sumrule ] ( left).,title="fig:",width=220 ] partial wave projections of @xmath15 ( left ) , @xmath115 ( center ) , and @xmath116 ( right ) scattering amplitudes , as a function of the cm energy , for @xmath104 , @xmath105 , and up to small electroweak corrections . increasing dashing size corresponds to increasing values of @xmath45 , with @xmath117 .
the blue curves are for @xmath106 tev , whereas the red curves are for @xmath107 tev .
the amplitudes increase with @xmath45 , in accordance with the behavior of @xmath111 in fig .
[ fig : sumrule ] ( left).,title="fig:",width=220 ]
since the work of @xcite , many detailed studies have been dedicated to weak boson scattering in a strongly coupled regime @xcite . in the absence of a light higgs , the lhc , running at 14 tev ,
would very probably be able to determine whether the weak bosons interact strongly at high energies @xcite .
see also @xcite for the analysis at 7 , 8 , and 10 tev . in the presence of a light tc higgs the prospects for discovering strong dynamics trough weak boson scattering are substantially unchanged as unitarization still occurs in a strong regime but may require larger amounts of luminosity .
weak boson scattering , in a strongly coupled regime and a generic composite higgs , has been investigated in @xcite . as motivated in sec .
[ sec : intro ] , a light composite higgs is expected to show up in models of near - conformal tc , like the `` vanilla tc '' model introduced in sec . [
sec : unit ] .
this also features a near degenerate set of spin - one vector and axial isotriplets . the higgs and the vector technimesons unitarize the longitudinal weak boson scattering amplitudes , leaving a footprint of strong dynamics in @xmath118 processes . in this section
we inquire on the possibility of uncovering such a footprint in the high energy region . for this purpose
we consider the most relevant totally leptonic channels , in which both weak bosons decay into leptons : @xmath119 , @xmath120 , @xmath121 , and @xmath122 .
these channels give a cleaner signature than semi - leptonic channels to four leptons due to its low cross section . ] .
two perturbative orders contribute to these channels at the lhc : the purely electroweak @xmath123 , which includes signal and electroweak background , and @xmath124 , which corresponds to the qcd background .
moreover , in the @xmath119 channel @xmath125 + 2 light jets production gives relevant contribution to the background .
this occurs when the @xmath126-quarks are lost in the beam pipe or have too low energy , and has been taken into account .
our study is based on the complete simulation of these processes at the parton level , using monte carlo techniques .
samples for the @xmath123 tc signal have been generated using the madgraph / madevent @xcite framework .
the tc model has been implemented in feynrules @xcite and used for phenomenological studies in @xcite .
the processes have been generated using the decay chain technique , in which the final state fermions are decay products of the weak bosons .
the @xmath124 qcd background and the @xmath123 sm samples have been produced using the phantom program @xcite , with the complete @xmath127 set of diagrams .
for @xmath128 we have used madgraph , relying also on the decay of the top quarks . tab .
[ tab : cutsbasic ] features a set of basic cuts to be employed for this kind of analysis .
it includes basic detector coverage ( e.g. : @xmath129 ) , minimum transverse momenta for all observed partons , and a cut against low energy jet activity @xmath130 , in order to avoid contamination from the underlying event .
we require a minimum di - lepton invariant mass to avoid photon singularity , and , in addition , we consider two different ranges for the mass of the lepton pair in the channel with two neutrinos , @xmath131 .
we first select same - flavor charged leptons with a mass in the interval @xmath132 gev @xmath133 gev , corresponding to a lepton pair produced by the decay of a @xmath3 boson . with the additional requirement of a large missing transverse momentum , @xmath134
, we cut out @xmath135 jets production , and isolate the @xmath136 channel . when the mass of the lepton pair is outside the quoted range , or the two oppositely charged leptons belong to different families , we consider the event to be a candidate for the @xmath137 channel .
since we are interested in high energy scattering , we require @xmath138 for this kind of events .
furthermore , we employ additional cuts especially tailored to eliminate certain types of background . these include suppression of top production : @xmath139 in @xmath140 jets channel , and @xmath141 in @xmath142 , in which the top momenta can not be reconstructed .
.basic cuts imposed to isolate highly energetic @xmath143 events .
see text for details . [
cols="<",options="header " , ]
it is possible that the recently observed boson at 125 gev is a composite scalar singlet of a near - conformal tc theory : a tc higgs .
the presence of a nearby fixed point reduces the mass of the tc higgs , relative to qcd - like estimates , and quadratically divergent radiative corrections from the top quark lead to a further large reduction .
if the near - conformal tc scenario is correct , then we expect also a near - degenerate set of spin - one vector and axial technimeson triplets to be in the spectrum . in @xcite the dy production of such resonances , at the lhc , was analyzed .
the results are summarized in fig . 1 of @xcite : the lhc reach , at 13 tev and 100 @xmath144 ,
is drastically reduced as the technimeson coupling @xmath45 is increased .
the reason for this is that in vmd ( which is typically assumed to hold ) the sm fermion coupling with the spin - one tc resonances is reduced by a factor of @xmath43 , where @xmath44 is the weak coupling .
dy production also drops with increasing mass of the spin - one resonances , as obviously expected from kinematical reasons . in this note
we investigated signatures of near - conformal tc by analyzing highly energetic weak boson scattering amplitudes .
the reason for this is that in a strongly coupled regime , and at high energies , the longitudinal weak boson scattering amplitudes are expected to saturate the unitarity bounds of eq .
( [ eq : unitbound ] ) . in a perturbative regime , in contrast
, scattering amplitudes are expected to drop to negligible values at high energies . in near - conformal tc , at a few tevs , the longitudinal weak boson scattering amplitudes are unitarized by the light tc higgs and by the spin - one vector triplet . since amplitudes are unitary but large , we expect enhanced signals in @xmath0 processes , which can therefore be used to test the tc hypothesis .
in contrast to dy production , the enhancement of weak boson scattering drops neither with increasing @xmath45 nor with increasing mass of the spin - one resonances , as the partial wave projections of the scattering amplitudes are always expected to saturate the unitarity bound .
we actually find the signal for the @xmath145 , @xmath146 , @xmath147 , and @xmath148 channels to increase with @xmath45 and @xmath97 ( @xmath97 is the mass of the spin - one axial triplet , which , being near degenerate with the mass of the spin - one vector triplet @xmath96 , can be conveniently used to parametrize the overall mass of the lightest states in the tc spin - one sector ) .
this shows that tc searches in @xmath0 channels are complementary to dy searches , which are most effective at small values of @xmath45 and @xmath97 .
our results are summarized in tab .
[ tab : pbsm ] , in which we show the probability of excluding the sm at the 95% c.l .
, for different near - conformal tc scenarios ( different values of @xmath45 and @xmath97 ) , and different integrated luminosities .
for instance , if the tc scenario with @xmath107 tev and @xmath149 is realized in nature , then the lhc will be able to exclude the sm , with a 77% probability , at the 95% c.l .
, already for 50 @xmath144 of integrated luminosity .
our analysis includes leptonic decays only , and can therefore be improved by including additional decay modes for the weak bosons .
the analysis can also be refined by investigating a larger set of points in the parameter space .
the reach of this analysis goes to regions of the parameter space inaccessible to the usual dy production of resonances and is therefore an important tool for probing this type of theory .
we would like to thank m.t .
frandsen for reviewing the manuscript .
the work of r.f . is supported by the marie curie iif grant .
99 g. aad _ et al . _ [ atlas collaboration ] , phys .
b * 710 * , 49 ( 2012 ) [ arxiv:1202.1408 [ hep - ex ] ] .
s. chatrchyan _ et al . _
[ cms collaboration ] , phys .
b * 710 * , 26 ( 2012 ) [ arxiv:1202.1488 [ hep - ex ] ] .
_ acta phys.polon._ * b8 * ( 1977 ) 475 .
a. falkowski , s. rychkov and a. urbano , jhep * 1204 * , 073 ( 2012 ) [ arxiv:1202.1532 [ hep - ph ] ] .
b. bellazzini , c. csaki , j. hubisz , j. serra and j. terning , arxiv:1205.4032 [ hep - ph ] . s. dimopoulos and l. susskind , nucl .
b * 155 * , 237 ( 1979 ) .
e. eichten and k. d. lane , phys .
b * 90 * , 125 ( 1980 ) .
b. holdom , phys .
d * 24 * , 1441 ( 1981 ) .
k. yamawaki , m. bando and k. -i .
matumoto , phys .
rev . lett . * 56 * , 1335 ( 1986 ) .
m. bando , k. -i .
matumoto and k. yamawaki , phys . lett .
b * 178 * , 308 ( 1986 ) .
d. d. dietrich , f. sannino and k. tuominen , phys .
d * 72 * , 055001 ( 2005 ) [ hep - ph/0505059 ] .
t. appelquist and y. bai , phys .
d * 82 * , 071701 ( 2010 ) [ arxiv:1006.4375 [ hep - ph ] ] .
a. manohar and h. georgi , nucl .
b * 234 * , 189 ( 1984 ) .
r. foadi , m.t .
frandsen , and f. sannino , in preparation g. t hooft , nucl .
b * 72 * , 461 ( 1974 ) .
e. corrigan and p. ramond , phys .
b * 87 * , 73 ( 1979 ) .
f. sannino , acta phys .
b * 40 * , 3533 ( 2009 ) [ arxiv:0911.0931 [ hep - ph ] ] .
m. harada , f. sannino and j. schechter , phys .
d * 54 * , 1991 ( 1996 ) [ hep - ph/9511335 ] .
j. bagger , v. d. barger , k. -m .
cheung , j. f. gunion , t. han , g. a. ladinsky , r. rosenfeld and c. p. yuan , phys .
d * 49 * , 1246 ( 1994 ) [ hep - ph/9306256 ] .
j. bagger , v. d. barger , k. -m .
cheung , j. f. gunion , t. han , g. a. ladinsky , r. rosenfeld and c. -p .
yuan , phys . rev .
d * 52 * , 3878 ( 1995 ) [ hep - ph/9504426 ] .
a. belyaev , r. foadi , m. t. frandsen , m. jarvinen , f. sannino and a. pukhov , phys .
d * 79 * , 035006 ( 2009 ) [ arxiv:0809.0793 [ hep - ph ] ]
. j. r. andersen , t. hapola and f. sannino , phys .
d * 85 * , 055017 ( 2012 ) [ arxiv:1105.1433 [ hep - ph ] ] .
m. t. frandsen , m. jarvinen and f. sannino , phys .
d * 84 * , 035025 ( 2011 ) [ arxiv:1103.3271 [ hep - ph ] ] .
r. foadi , m. jarvinen and f. sannino , phys .
d * 79 * , 035010 ( 2009 ) [ arxiv:0811.3719 [ hep - ph ] ] .
s. weinberg , phys .
* 18 * , 507 ( 1967 ) .
t. appelquist and f. sannino , phys .
d * 59 * , 067702 ( 1999 ) [ hep - ph/9806409 ] .
m. e. peskin and t. takeuchi , phys .
d * 46 * , 381 ( 1992 ) .
r. foadi , m. t. frandsen , t. a. ryttov and f. sannino , phys .
d * 76 * , 055005 ( 2007 ) [ arxiv:0706.1696 [ hep - ph ] ] .
m. s. chanowitz and m. k. gaillard , _ nucl .
* b261 * ( 1985 ) 379 .
hikasa and k. igi , _ phys .
_ * b261 * ( 1991 ) 285288 . c. englert , b. jager , m. worek , and d. zeppenfeld , _ phys .
* d80 * ( 2009 ) 035027 , [ http://xxx.lanl.gov/abs/0810.4861[arxiv:0810.4861 ] ] .
a. ballestrero , g. bevilacqua and e. maina jhep 05 ( 2009 ) 015 , arxiv:0812.5084 [ hep - ph ] .
a. ballestrero , g. bevilacqua , d. buarque franzosi and e. maina jhep 11 ( 2009 ) 126 , arxiv:0909.3838 [ hep - ph ] .
a. ballestrero , d. b. franzosi and e. maina , jhep * 1106 * , 013 ( 2011 ) [ arxiv:1011.1514 [ hep - ph ] ] .
a. ballestrero , d. buarque franzosi , l. oggero and e. maina , jhep * 1203 * , 031 ( 2012 ) [ arxiv:1112.1171 [ hep - ph ] ] .
a. ballestrero , d. b. franzosi and e. maina , jhep * 1205 * , 083 ( 2012 ) [ arxiv:1203.2771 [ hep - ph ] ] .
k. cheung , c .- w .
chiang , and t .- c .
yuan , _ phys .
_ * d78 * ( 2008 ) 051701 , [ http://xxx.lanl.gov/abs/0803.2661[arxiv:0803.2661 ] ] .
f. maltoni , t. stelzer , jhep 0302 ( 2003 ) 027 ; t. stelzer and w. f. long , comput .
. commun . * 81 * ( 1994 ) 357 ; + j. alwall _ et al .
_ , jhep 0709:028,2007 , arxiv:0706.2334 ; + h. murayama , i. watanabe and k. hagiwara , kek-91 - 11 .
a. ballestrero , a. belhouari , g. bevilacqua , v. kashkan and e. maina , comput .
commun . * 180 * , 401 ( 2009 ) [ arxiv:0801.3359 [ hep - ph ] ] .
b. jager , c. oleari , and d. zeppenfeld , _ jhep _ * 07 * ( 2006 ) 015 , [ http://xxx.lanl.gov/abs/hep-ph/0603177[hep-ph/0603177 ] ] . | the recently observed boson at 125 gev could be a light composite scalar from near - conformal technicolor dynamics : a technicolor higgs .
if this is the case , unitarization of longitudinal weak boson scattering amplitudes , which is due to exchanges of the higgs and spin - one vector technimesons , is expected to occur in a strong regime , with saturation of the unitarity bounds .
this implies that @xmath0 processes , where @xmath1 is either a @xmath2 or a @xmath3 boson , are enhanced , relative to the standard model .
we show that this allows probing near - conformal technicolor for couplings and masses of the spin - one resonances which are not directly accessible for direct drell - yan production . |
elastic hadron scattering constitutes a hard challenge for qcd .
the problem concerns the large distances involved ( confinement ) , which renders difficult the development of a formal nonperturbative calculational scheme for scattering states , able to describe soft diffractive processes . at this stage analyticity , unitarity , crossing and their consequences , still represent a fundamental framework for the development of theoretical ideas , aimed to reach efficient descriptions of the experimental data involved . in this context ,
dispersion relations , connecting real and imaginary parts of the scattering amplitude , play an important role as a useful mathematical tool , in the simultaneous investigation of particle - particle and antiparticle - particle scattering .
dispersion relations in integral form , for hadronic amplitudes , were introduced in the sixties , as consequences of the cauchy s theorem and the analytic properties of the scattering amplitude , dictated by unitarity @xcite .
however , two kinds of limitations characterize this _ integral _ approach : ( 1 ) its nonlocal character ( in order to evaluate the real part , the imaginary part must be known in all the integration space ) ; ( 2 ) the restricted class of functions that allows analytical integration . later on it
was shown that , for hadronic forward elastic scattering _ in the region of high and asymptotic energies _ , these integral relations can be replaced by derivative forms @xcite . since then
, the formal replacement of integral by derivative relations and their practical use have been widely discussed in the literature @xcite , mainly in the seminal papers by kol and fischer @xcite .
@xcite for a recent critical review on the subject . despite the results that have been obtained with the _ derivative _ approach , the high - energy condition ( specifically , center - of - mass energies above 10 - 20 gev ) turns out difficult any attempt to perform global fits to the experimental data connecting information from low and high energy regions .
a first step in this direction appears in ref .
@xcite , where new representations for the derivative relations , extended to low energies , have been introduced by cudell , martynov and selyugin and to which we shall refer in what follows . however , a rigorous formal extension of the derivative dispersion relations down to the physical threshold , providing a complete analytical equivalence between integral and differential approaches , is still missing and that is the point we are interested in . in this work , we first demonstrate that , for a class of functions of physical interest as forward elastic scattering amplitudes , the integral relations can be analytically replaced by derivative forms without the high - energy approximation .
therefore , in principle , for this class of functions , derivative relations hold for any energy above the physical threshold .
we then check the consistences of the results obtained with the integral relations and the extended derivative dispersion relations by means of a simple analytical parametrization for the total cross sections from proton - proton ( @xmath0 ) and antiproton - proton ( @xmath1 ) scattering ( highest energy interval with available data ) .
in addition , we compare the results with those obtained through the standard derivative relations ( high - energy condition ) and the derivative representation by cudell - matynov - selyugin .
we shall show that , above the physical threshold , only the extended relations lead to exactly the same results as those obtained with integral forms .
we proceed with a critical discussion on the limitations of our analysis from both formal and practical points of view .
the manuscript is organized as follows . in sec .
[ sec : dr ] we recall the main formulas and some conditions involving the _ integral dispersion relations _ ( idr ) , the _ standard derivative dispersion relations _ ( sddr ) and the _ cudell - martynov - selyugin representations _ ( cmsr ) ; we also present , in certain detail , the replacement of idr by the _ extended derivative dispersion relations _ ( eddr ) . in sec .
[ sec : pra ] we check the consistences and exemplify the applicability of all these results in simultaneous fits to the total cross section and the ratio @xmath2 of the real to imaginary parts of the forward amplitude , from @xmath0 and @xmath1 scattering . in sec .
[ sec : critical ] we present a critical discussion on all the obtained results .
the conclusions and some final remarks are the contents of sec .
[ sec : conclusions ] .
first , it is important to recall that analyticity , unitarity and crossing lead to idr for the scattering amplitudes in terms of a _ crossing symmetric variable_. for an elastic process , @xmath3 , in the forward direction , this variable corresponds to the energy of the incident particle in the laboratory system , @xmath4 @xcite . in this context and taking into account polynomial boundedness , the one subtracted idr for crossing even ( @xmath5 ) and odd ( @xmath6 ) amplitudes , in the physical region ( @xmath7 ) , read @xcite @xmath8 @xmath9 where @xmath10 is the subtraction constant .
the connections with the hadronic amplitudes for crossed channels , such as @xmath0 and @xmath1 elastic scattering , are given by the usual definitions : @xmath11 the main practical use of the idr concerns simultaneous investigations on the total cross section ( optical theorem ) and the ratio @xmath2 of the real to imaginary parts of the forward amplitude , which is also our interest here . in terms of the crossing symmetric variable @xmath4
these physical quantities are given , respectively , by @xcite @xmath12 @xmath13 where @xmath14 is the scattering angle in the laboratory system .
basically , at high energies , the replacement of idr by sddr is analytically performed by considering the limit @xmath15 in eqs .
( [ eq : idre ] ) and ( [ eq : idro ] ) @xcite . it should be recalled that an additional high - energy approximation is considered in these integral equations , when they are expressed in terms of the center - of - mass energy squared @xmath16 and not @xmath4 @xcite .
however , based on a rigorous replacement ( discussed in sec .
[ sec : eddr ] ) , we consider the derivative relations in terms of the crossing symmetric variable @xmath4 . in this case
the sddr read @xcite @xmath17 \frac{{\mathrm{im\:}}f_{+}(e)}{e } , \label{eq : sddre}\ ] ] @xmath18 { \mathrm{im\:}}f_{-}(e ) .
\label{eq : sddro}\ ] ] necessary and sufficient conditions for the convergence of the above tangent series have been established by kol and fischer , in particular through the following theorem @xcite : let @xmath19 .
the series @xmath20 f(x ) \nonumber\end{aligned}\ ] ] converges at a point @xmath21 if and only if the series @xmath22 is convergent .
for example , in the case of @xmath23 , @xmath24 a real constant , the ratio test demands @xmath25 for the series to be absolutely convergent ( which will be our interest in sec .
[ sec : pra ] ) .
recently , the following representations have been introduced for the derivative dispersion relations @xcite @xmath26 \frac{{\mathrm{im\:}}f_{+}(e)}{e } \nonumber \\ & & - \frac{2}{\pi } \sum_{p=0}^{\infty } \frac{c_+(p)}{2p+1}\left(\frac{m}{e}\right)^{2p } , \label{eq : cmsre}\end{aligned}\ ] ] where , @xmath27.\ ] ] and @xmath28 \frac{{\mathrm{im\:}}f_{-}(e)}{e } \nonumber \\ & - & \frac{2}{\pi } \sum_{p=0}^{\infty } \frac{c_-(p)}{2p+1}\left(\frac{m}{e}\right)^{2p+1 } , \label{eq : cmsro}\end{aligned}\ ] ] where @xmath29,\ ] ] and @xmath30 and @xmath31 .
we note the presence of correction terms in the form of infinity series , which go to zero as the energy increases , leading to the sddr , eqs .
( [ eq : sddre]-[eq : sddro ] ) .
we shall use this representation in sec .
[ sec : pra ] , where their applicability is discussed in detail . in this section
we present our analytical replacement of the idr by derivative forms without the high - energy approximation .
we also specify the class of functions for which this replacement can be formally performed .
let us consider the even amplitude , eq .
( [ eq : idre ] ) . integrating by parts we obtain @xmath32 following ref .
@xcite , we define @xmath33 and @xmath34 , so that the integral term in the above formula is expressed by @xmath35 where @xmath36 . expanding the logarithm in the integrand in powers of @xmath37 , @xmath38 and assuming that @xmath39 is an analytic function of its argument
, we perform the expansion @xmath40 substituting the above formulas in eq .
( [ eq:11 ] ) and integrating term by term , _ under the assumption of uniform convergence of the @xmath41 series _ , we obtain @xmath42 where @xmath43 \nonumber\end{aligned}\ ] ] and @xmath44 is the incomplete gamma function @xmath45 . with this procedure and from @xmath46 ,
( [ eq:10 ] ) is expressed by @xmath47 which can be put in the final form @xmath48 where the correction term @xmath49 is given by @xmath50 with analogous procedure for the odd relation we obtain @xmath51 where @xmath52 equations ( [ eq : eddre ] ) and ( [ eq : eddro ] ) are the novel eddr , which are valid , in principle , for any energy _ above _ the physical threshold , @xmath53 .
we note that the correction terms @xmath54 as @xmath55 , leading , in this case , to the sddr , eqs .
( [ eq : sddre ] ) and ( [ eq : sddro ] ) .
we also note that the structure of the cmsr , eqs .
( [ eq : cmsre ] ) and ( [ eq : cmsro ] ) , are similar to the above results , but without the logarithm terms .
these terms come from the evaluation of the primitive at the lower limit in the integration by parts . since theorem [ theo:1 ] insures the uniform convergence of the series expansion associated with @xmath56 the condition imposed by this theorem defines the class of functions for which the eddr hold .
for example , that is the case for @xmath23 , @xmath57 , referred to in sec .
[ sec : sddr ] .
other conditions are discussed by kol and fischer @xcite .
in this section we verify and discuss the consistences between the analytical structures of the idr and the eddr in a specific example : the connections of the total cross section with the @xmath2 parameter from @xmath0 and @xmath1 scattering .
firstly , it is important to note that the efficiency of both integral and derivative approaches in the description of the experimental data , depends , of course , on the theory available , namely the input for the imaginary part of the amplitude . in the absence of a complete model , valid for any energy above the physical threshold
, we shall consider only as a _ framework _
, a pomeron - reggeon parametrization for the scattering amplitude @xcite . for @xmath0 and @xmath1 scattering this analytical model assumes nondegenerate contributions from the even ( @xmath5 ) and odd ( @xmath6 ) secondary reggeons ( @xmath58/@xmath59 and @xmath2/@xmath60 , respectively ) , together with a simple pole pomeron contribution : @xmath61 where @xmath62 for @xmath0 and @xmath63 for @xmath1 . as usual , the pomeron and the even / odd reggeon intercepts are expressed by @xmath64 we stress that the pomeron - reggeon phenomenology is intended for the high - energy limit ( rigorously , @xmath4 or @xmath65 ) .
its use here , including the region of low energies , has only a framework character .
however , as we shall show , this model is sufficient for a comparative analysis of the consistences .
we shall return to this aspect in sec . [
sec : critical ] . in what follows ,
the point is to treat simultaneous fits to the total cross section and the @xmath2 parameter from @xmath0 and @xmath1 scattering and compare the results obtained with both idr and eddr .
schematically , with parametrization ( [ eq:15]-[eq:16 ] ) we determine @xmath66 through eq .
( [ eq:3 ] ) and then @xmath67 either by means of the idr , eqs .
( [ eq : idre]-[eq : idro ] ) or the eddr , eqs .
( [ eq : eddre]-[eq : eddro ] ) . returning to eq .
( [ eq:3 ] ) we obtain @xmath68 and @xmath69 and , at last , eqs .
( [ eq : tcs ] ) and ( [ eq : rho ] ) lead to the analytical connections between @xmath70 and @xmath71 for both reactions .
moreover , through the same procedure , we shall also compare the above results with those obtained by means of both the sddr , eqs .
( [ eq : sddre]-[eq : sddro ] ) and the cmsr , eqs .
( [ eq : cmsre]-[eq : cmsro ] ) .
we first present the fit procedure and then discuss all the obtained results .
for the experimental data on @xmath72 and @xmath73 , we made use of the particle data group archives @xcite , to which we added the values of @xmath2 and @xmath74 from @xmath1 scattering at 1.8 tev , obtained by the e811 collaboration @xcite .
the statistical and systematic errors were added in quadrature .
the fits were performed through the cern - minuit code , with the estimated errors in the free parameters corresponding to an increase of the @xmath75 by one unit . to fit the data as function of the center - of - mass energy
, we express the lab energy in the corresponding formulas in terms of @xmath76 , namely @xmath77 .
we included all the data above the physical threshold , @xmath78 gev , that is , we did not perform any kind of data selection .
since the ensemble has a relatively large number of experimental points just above the threshold , the statistical quality of the fit is limited by the model used here as framework .
in fact , with the these choices and procedures we obtained reasonable statistical results ( in terms of the @xmath75 per degree of freedom ) only for an energy cutoff of the fits at @xmath79 4 gev . however , we stress that our focus here is in tests on the consistences among the different relations and representations and not , strictly , on the statistical quality of the fits ( we shall return to this point in sec . [
sec : critical ] ) . in each of the four cases ( idr , sddr , cmsr and eddr ) ,
we consider two variants of the fits , one neglecting the subtraction constant ( that is , taking @xmath80 ) and the other considering the subtraction constant as a free fit parameter .
the numerical results and statistical information on the fits are displayed in table [ tab:1 ] ( @xmath81 ) and table [ tab:2 ] ( @xmath10 free ) .
the corresponding curves together with the experimental data are shown in fig .
[ fig:1 ] ( @xmath81 ) and fig .
[ fig:2 ] ( @xmath10 free ) . [ cols="^,^,^,^,^",options="header " , ] and @xmath2 from @xmath0 and @xmath1 scattering , by means of integral dispersion relations ( idr ) , standard derivative dispersion relations ( sddr ) , cudell - martynov - selyugin representation ( cmsr ) and the extended derivative dispersion relations ( eddr ) and considering the subtraction constant @xmath80 ( table [ tab:1 ] ) .
the curves corresponding to idr ( solid ) and eddr ( dot - dashed ) coincide .
, title="fig : " ] and @xmath2 from @xmath0 and @xmath1 scattering , by means of integral dispersion relations ( idr ) , standard derivative dispersion relations ( sddr ) , cudell - martynov - selyugin representation ( cmsr ) and the extended derivative dispersion relations ( eddr ) and considering the subtraction constant @xmath80 ( table [ tab:1 ] ) .
the curves corresponding to idr ( solid ) and eddr ( dot - dashed ) coincide .
, title="fig : " ] but considering the subtraction constant @xmath10 as a free fit parameter ( table [ tab:2]).,title="fig : " ] but considering the subtraction constant @xmath10 as a free fit parameter ( table [ tab:2]).,title="fig : " ] the main goal of this section is to discuss the consistences among the results obtained by means of distinct analytical connections between the real and imaginary parts of the amplitude .
however , some phenomenological consequences can also be inferred from this study , as discussed in what follows . from tables [ tab:1 ] and [ tab:2 ]
we see that , as expected , the best statistical results are obtained with the subtraction constant @xmath10 as a free fit parameter .
however , as we shall show , taking @xmath81 gives suitable information not only on the practical equivalence between the idr and the differential forms ( sddr , cmsr and eddr ) , but also on the important role played by the subtraction constant .
for that reason we shall treat separately the cases @xmath81 and @xmath10 as a fit parameter .
from table [ tab:1 ] we see that , for @xmath81 , the numerical results obtained with the idr and the eddr are exactly the same , up to four figures and that this does not occur in the case with the sddr and the cmsr neither .
that is an important result since it demonstrates the accuracy of our analytical results for the extended derivative relations .
we note that the high values of @xmath82 , in all the cases , are consequences of the specific analytical model considered ( intended for the high - energy region ) and the energy cutoff used .
we add the fact that we did not performed any data selection , but used all the available data from the pdg archives .
however , as already commented this disadvantage has no influence in our main goal , namely tests of consistences .
the effects of the equivalences ( idr and eddr ) and differences ( idr and sddr or cmsr ) in the description of the experimental data are shown in fig .
[ fig:1 ] .
the curves corresponding to idr ( solid ) and eddr ( dot - dashed ) coincide at all the energies above the threshold and we see that even with the fit cutoff at @xmath83 gev , the description of the experimental data below this point is reasonably good in both cases . on the other hand , the differences between the exact results ( idr and eddr ) and the sddr or cmsr are remarkable for @xmath72 at the highest energies and for @xmath73 in the region of low energies ( below @xmath84 gev ) . in the case of the total cross section ,
the results with sddr and cmsr indicate a faster increase with the energy then those with the idr and eddr .
we stress the importance of this point , since it gives different solutions for the well known puzzle between the cdf data @xcite and the e710/e811 data @xcite at @xmath85 tev ; in this respect , we see that the exact results ( idr and eddr ) favor the e811/e710 results .
in particular the values for the pomeron intercept read ( table [ tab:1 ] ) : @xmath86 ( idr and eddr ) , @xmath87 ( sddr ) and @xmath88 ( cmsr ) . with @xmath10 as a free fit parameter our results demonstrate , once more , an effect that we have already noted before @xcite , namely the high - energy approximation can be absorbed by the subtraction constant .
in fact , from fig . [ fig:2 ] we see that in this case , the differences between the sddr / cmsr and the exact results idr / eddr , practically disappear . from table
[ tab:2 ] we can identify the subtraction constant as the responsible for this complementary effect : the numerical values of the fit parameters and errors are practically the same in all the four cases , except for the values of @xmath10 , that is , in practice , the differences are absorbed by this parameter .
we conclude that the subtraction constant affects the fit results even in the region of the highest energies ; this effect is due to the correlations among the free parameter in the fit procedure , as previously observed @xcite . of course ,
also in this case the numerical values obtained with the idr and eddr are exactly the same , including the value of the subtraction constant up to four figures ( table [ tab:2 ] ) .
in particular , we note that all the four variants indicate the same result for the intercept of the pomeron , namely @xmath89 .
the corresponding result for the total cross section lies nearly between the cdf and e811/e710 results , barely favoring the last ones ( figure [ fig:2 ] ) .
we have demonstrated that for the class of functions defined by theorem [ theo:1 ] , idr can be formally replaced by differential operators without any high - energy approximation ; we have also verified the equivalence between the integral and extended derivative results , in the particular case of a simple phenomenological parametrization for @xmath0 and @xmath1 scattering . despite the encouraging results reached the whole analysis has limitations from both formal and practical points of view . in what follows
we summarize the main critical points , giving references where more details can be found and providing also suggestions for further investigations .
since the eddr involve two contributions , the tangent operators ( sddr ) and the correction terms @xmath90 , we shall consider the two cases separately .
we also present some critical remarks on the pomeron - reggeon model used in sect .
[ sec : pra ] as a practical framework .
first , let us discuss some aspects related to the dispersion approach as it has been treated and widely used in the literature till now , namely the sddr , eqs .
( [ eq : sddre ] ) and ( [ eq : sddro ] ) . as commented in sect .
[ sec : sddr ] , these equations are obtained by considering the limit @xmath15 in the idr , eqs .
( [ eq : idre ] ) and ( [ eq : idro ] ) .
this condition is a critical one , which puts serious practical and formal limitations in any use of the sddr , because that means to go to lower energies by passing through different thresholds , resonances , poles , up to @xmath91 ! in this sense ,
the expression of the tangent operator as an integral from @xmath91 to @xmath92 does not guarantee any local character for the differential approach ( or the corresponding integral ) , even in the case of convergence of the series .
in other words , this representation of the non - local operator ( integral ) in terms of local operators ( tangent series ) , does not guarantee the non - locality of the result .
moreover , the representation does not apply near the resonances @xcite and the convergence of the series has been discussed in several works leading some authors to argue that , in a general sense , the mathematical condition for the convergence excludes all cases of physical interest " @xcite .
these and other aspects were extensively discussed in the seventies and eighties @xcite and some points have been recently reviewed in @xcite .
however , there is a fundamental point developed by some authors that enlarger the practical applicability of the sddr under some special conditions @xcite .
as stated by kol and fischer @xcite , in discussing the replacement of idr by sddr we must distinguish two formulations : ( 1 ) to consider the case of asymptotic energies and a finite number of terms in the tangent series ; ( 2 ) to consider finite energies and an infinity number of terms in the series .
the former case applies for smooth behaviors of the amplitude ( as it is the case at sufficiently high energies , specifically @xmath93 10 - 20 gev ) . that includes a wide class of functions of physical interest , mainly if only the first term can be considered @xcite ( see @xcite for a recent analysis even beyond the forward direction ) .
the later case , however , is critical for at least two reasons .
first , because the condition of convergence of the series ( theorem [ theo:1 ] ) limits the class of functions of practical applicability . secondly and more importantly , since the high - energy approximation is enclosed , all the strong limitations referred to before applies equally well to this case . in conclusion , the class of functions for which the sddr have a practical applicability depends strongly on the formalism considered and is narrower in the case of finite energies , namely entire functions in the logarithm of the energy .
let us now discuss the eddr , with focus on the role of the correction terms @xmath90 in eqs .
( [ eq : eddre ] ) and ( [ eq : eddro ] ) .
first we note that these infinity series are analytically associated with the fixed lower limit @xmath94 in the integral representation and they correspond to the contributions that are neglected in the high - energy approximation ( @xmath15 ) .
for example , for the even case we have the formal identity @xmath95 which means that all the physical situation concerns the region _ above _ the physical threshold @xmath53 .
therefore , from a formal point of view , the critical points raised above on the sddr ( tangent operator only ) , concerning the infinity series in the region @xmath96 , do not apply in this case and the critical point here concerns only the convergence of the correction series and their practical applicability . on the one hand , from a formal point of view ( as already discussed at the end of sect .
[ sec : eddr ] ) , the convergence of the correction series is ensured by theorem [ theo:1 ] and that means a narrower class of functions than that associated only with entire functions in the logarithm of the energy .
this restriction is due to the infinity number of derivatives in @xmath97
. we shall give and discuss some examples in what follows .
from a practical point of view , it is obvious that the efficiency and/or real applicability of not only the derivative approach ( eddr ) , but also the integral one ( idr ) , depends on the specific physical problem involved . in this scenario ( the physical problem )
we expect to find some specific limitations that are independent of the formal aspects referred to above and these aspects demands also some comments . in principle and in a _ general sense _ ,
if we attempt to apply dispersion techniques directly to the experimental data ( related to the imaginary part of a function ) , we are faced with the problem of error propagation from the experimental uncertainties .
even if we can reproduce " the experimental behavior by means of suitable analytical parameterizations , with statistical errors inferred for the free parameters , these errors should , in principle , be propagated too . in this case , the infinity series in both sddr and eddr have certainly limited usefulness ( see for example @xcite for the sddr case ) . however , if error propagation from the fit results is not of interest or can be neglected , and , most importantly , one has a correct " or acceptable model for the imaginary part of the amplitudes , then we are restricted only to the the formal conditions discussed above and the derivative approach becomes reliable , including the correction series ( theorem [ theo:1 ] ) .
let us now discuss the specific physical problem that motivated the present analysis . as commented in our introduction and in sect .
[ sec : dr ] , we focused the dispersion techniques in the context of hadron scattering , in special in the elastic case , for which a complet theory is still absent .
the main goal concerns the connections between total cross section and the @xmath2 parameter for energies above @xmath98 = 5 - 10 gev . in terms of dispersion techniques
the usual way to treat the subject is by means of idr , sddr and the analyticity prescriptions for even and odd amplitudes ( and recently the cmsr ) . in this specific case , besides the absence of a pure qcd treatment , the subject is characterized by three kinds of problems : ( 1 ) formal justification of the usual phenomenology ; ( 2 ) approximated descriptions of the experimental data by phenomenological models ; ( 3 ) experimental data available ( problems ( 2 ) and ( 3 ) are certainly connected ) .
since these problems affect the practical applicability and efficiency of the dispersion techniques , let us shortly discuss some aspects involved .
\(1 ) as it is well know , the usual phenomenology for the total cross sections is based on the reggeon concepts and involves distinct contributions from pomerons and secondaries reggeons .
in this context , analytical parameterizations for the total cross sections are characterized by power and logarithm functions of the energy ( reggeons , simple , double and triple pole pomerons ) and the fits are performed not below @xmath98 = 5 gev .
we note that all these contributions belong to the class of functions defined by theorem [ theo:1 ] ( the tangent series can be summed leading to closed analytical results ) and they have been used and investigated in several works @xcite . however , as we have already pointed out @xcite , the central problem here concerns the fact that these contributions are formally justified only for asymptotic energies ( @xmath4 or @xmath76 @xmath99 ) , which certainly is not the case for the energies considered .
the applicability of these models seems to be justified only under the hypothesis that the accelerators have already reached the energies that can be considered asymptotic in the mathematical context , which seems to us a dangerous assumption .
( 2 - 3 ) a close look at the bulk of experimental data available shows that these ensembles present several discrepancies due to spurious data , normalization problems and other effects . in this respect ,
recent analysis have pointed out the necessity of some screening criterion in order to select the
correct " experimental information .
we shall not discuss this question here because it seems to us an open problem .
but the point is that this fact puts serious limitations in any interpretation of statistical tests of the fits , as the popular @xmath75 per degree of freedom and , consequently , not only in the efficiency of the phenomenological descriptions , but also in the possible selection of the best phenomenological model . at last
let us return to the applicability of the eddr , now in this context .
despite of all the above problems , the known and usual phenomenological approach is characterized by analytical parameterizations for the imaginary parts of the amplitude and statistical tests on the quality of the fit . in this case , with specific analytical representations for the total cross sections , without error propagation from the fit parameters and in the regge context we understand that the correction terms , we have introduced , can have a suitable applicability in the context of the dispersion techniques .
the point is that the class of functions for which they hold includes all the usual regge parameterizations and since the high - energy approximations is absent the fits can be formally extend to lower energies .
however to reach a good statistical description of the data , specifically near the threshold , demands a correct " model for the imaginary part of the amplitude , which , to our knowledge is still lacking .
we shall return to this point in what follows .
based on all the above limitations in the phenomenological context , we have chosen one of the possible ( and popular ) models in order to check the equivalences ( and differences ) among the different dispersion representations analyzed in this work .
although , as demonstrated in sect .
[ sec : pra ] , this choice is sufficient for our aim , some drawnbacks involved demand additional comments . in the mathematical context ,
as demonstrated by kol and fischer @xcite , some formal results , theorems and representations for the derivative relations were obtained under the assumption of the froissart - martin bound , @xmath100 , but other forms of sddr do not require this bound .
therefore , since the simple pole pomeron contribution , that dominates at the asymptotic energies , violates this bound , the model assumed is not an example in full agreement with the totality of the formal results .
however , as already exemplified ( sect .
[ sec : sddr ] ) , the model belongs to the class of functions defined by theorem [ theo:1 ] and therefore , in this restrictive sense it seems to us to be an acceptable choice . in the formal phenomenological context ,
when applied below asymptotic ( infinity ) energies , the model suffers from all the drawbacks already discussed . despite of this ,
its use above , let us say , @xmath98 = 10 gev , could be explained ( not justified ) by the fact that the regge approach is the only known formalism , able to describe some global characteristics of the soft scattering .
what is presently expected is the development of a microscopic theory able to justify its efficiency . now let us focus in the low energy region , above the physical threshold , @xmath101 1.88 gev @xmath102 10 gev and discuss the usefullness and practical applicability of the eddr . to our knowledge
, there is no model proposed for this interval and that could explain the fact that fit procedures , even through idr , make use of energy cutoffs at @xmath103 5 gev ( @xcite is a typical example ) . in this sense , the usefulness of the correction terms @xmath90 could be questioned .
however , we understand that the lack of a phenomenological approach for that region may also be a mirror of the present stage , characterized by a focus ( probably excessive ) on the highest and asymptotic energies ( the great expectations from the tevatron , rhic , lhc ) . in our oppinion , independently of the fact that the asymptopia " might be resolved or not in a short term , the connection between resonance region ( above the physical threshold ) and the high - energy region ( above 10 gev ) still remains a fundamental problem demanding solution .
in this respect we understand that the eddr can play an important role in further investigations . concerning the practical applicability of the extended relations in this region ,
it is obviously limited , due to the lack of a correct " or accepted analytical model for the imaginary part of the amplitude .
one way to circumvent this problem could be the introduction of a different parametrization for this particular region .
that was the procedure used in ref .
@xcite ; although without justification or explicit reference to the analytical form used , the authors obtained reasonable fit results .
however , beyond the lack of any physical meaning , this procedure puts limitations on the equivalence between integral and derivative representations .
based on the above facts and aimed only to check and compare the results obtained through different dispersion representations , we considered the pomeron - reggeon parametrization extended up to the low energy region , with a fit cutoff at @xmath98 = 4 gev . certainly the statistical results displayed in tables [ tab:1 ] and [ tab:2 ] indicate that the confidence level is very low and even a look at figs .
[ fig:1 ] and [ fig:2 ] shows that the data near the resonance are not adequately described . as a consequence the numerical results in tables
[ tab:1 ] and [ tab:2 ] may be questionable on physical grounds .
however , we insist that all the figures in these tables are fundamental for a definite check of all the analytical representations investigated ( which is the only aim of sect . [
sec : pra ] ) .
at last we note that one may think that it might be possible to find a suitable function , in agreement with the convergence condition and able to fit all the experimental data of interest on secure statistical grounds ; that would be enough for our tests of consistences .
we are not sure about this possibility , but the point is that the use of a known and popular parametrization , even with limited efficiency , can bring new insights for further developments mainly because it gives information on what should be improved .
we have obtained novel analytical expressions for the derivative dispersion relations , without high - energy approximations .
the mathematical results are valid for the class of functions specified by theorem [ theo:1 ] . in principle
, their applicability can be extended to any area that makes use of dispersion techniques , with possible additional constraints , dictated by the analytical and experimental conditions involved . in special , under aadequate circumstances ,
the local character of the derivative operators may be a great advantage . for scattering amplitudes belonging to the class of functions defined by theorem [ theo:1 ] , the eddr are valid for any energy above the physical threshold .
since the experimental data on the total cross sections indicate a smooth variation with the energy ( without oscillations just above the physical threshold and a smooth systematic increase above @xmath104 gev ) , this class includes the majority of functions of physical interest . using as framework a popular pomeron - reggeon parametrization for the total cross sections , we have checked the numerical equivalence between the results obtained with the idr ( finite lower limit @xmath94 ) and the eddr , as well as the differences associated with the sddr and the cmsr .
we have also presented a critical discussion on the limitations of the whole analysis from both formal and practical points of view .
we stress that , as in the case of idr , the practical efficiency of the eddr in the reproduction of the experimental data on @xmath74 and @xmath2 depends on the model considered . here , in order only to check the consistences among the different analytical forms , we made use of a particular pomeron - reggeon parametrization , for which a cutoff at @xmath105 4 gev was necessary .
for example , by considering the full nondegenerated case ( four contributions , each one from each meson trajectory , @xmath106 ) , this cutoff can be reduced @xcite , or the @xmath82 can be reduced for the same cutoff . despite the limitations of our practical example ( sec . [
sec : critical ] ) , some interesting phenomenological aspects could be inferred . in particular , although already noted @xcite , we have called the attention to the role of the subtraction constant as a practical regulator " , in the replacement of idr by derivative forms , a fact that is clearly identified in table [ tab:2 ] : the high - energy approximation is absorbed by the constant . in this respect , we have demonstrated that this artifice , which lack physical meaning , can be avoided by the direct use of the eddr .
however , this observation does not depreciate the important role of the subtraction constant as a free fit parameter , since the best statistical results are obtained in this context ( tables [ tab:1 ] and [ tab:2 ] ) .
in particular , we note that the effect of this parameter is to provide a slight higher value for the pomeron intercept , @xmath107 ( @xmath81 ) and @xmath108 ( @xmath10 free ) . to our knowledge ,
a well established theoretical approach for total cross sections just above the physical threshold and in the region connecting low and high energies is still absent . in this sense , despite all the limitations discussed , we hope that the local analytical operators , developed here for these regions , can contribute , as a formal mathematical tool , for further developments on the subject . n.v .
gribov and a.a .
migdal , yad . fiz .
* 8 * , 1002 ( 1968 ) [ sov
* 8 * , 583 ( 1969 ) ] ; j.b .
bronzan , in : argonne symposium on the pomeron , anl / hep-7327 ( 1973 ) p. 33 ; j.d .
jackson , in : 1973 scottish summer school , lbl-2079 ( 1973 ) p. 39 .
i. vrko , czech .
math . j. * 35 * , 59 ( 1985 ) ; m.j .
menon , a.e .
motter , and b.m .
pimentel , phys .
b * 451 * 207 ( 1999 ) ; yu .
s. vernov , m.n .
mnatsakanova , physics of particles and nuclei * 32 * , 589 ( 2001 ) .
j. fischer and p. kol , phys .
b * 64 * , 45 ( 1976 ) ; phys . rev .
d * 17 * , 2168 ( 1978 ) ; p. kol and j. fischer , j. math . phys . * 25 * , 2538 ( 1984 ) ; j. fischer and p. kol , czech . j. phys .
b * 37 * , 297 ( 1987 ) .
vila and m.j .
menon , in _ sense of beauty in physics a volume in honour of adriano di giacomo _ , edited by m. delia , k. konishi , e. meggiolaro and p. rossi ( edizioni plus , pisa university press , pisa , 2006 ) , p. 153
; hep - ph/0601194 . | we discuss some formal and practical aspects related to the replacement of integral dispersion relations ( idr ) by derivative forms , _ without high - energy approximations_. we first demonstrate that , for a class of functions with physical interest as forward scattering amplitudes , this replacement can be analytically performed , leading to novel extended derivative dispersion relations ( eddr ) , which , in principle , are valid for any energy above the physical threshold .
we then verify the equivalence between the idr and eddr by means of a popular parametrization for total cross sections from proton - proton and antiproton - proton scattering and compare the results with those obtained through other representations for the derivative relations .
critical aspects on the limitations of the whole analysis , from both formal and practical points of view , are also discussed in some detail
. published in _
brazilian journal of physics _ * 37 * , 358 ( 2007 ) |
in a recent work , one of the authors ( @xcite ) concluded that his explanation in terms of ground - level atomic polarization of the `` enigmatic '' linear polarization peaks of the d - lines , observed by @xcite in `` quiet '' regions close to the solar limb , implies that the magnetic field in the lower solar chromosphere must be either isotropically distributed and extremely weak ( with @xmath5 ) or , alternatively , practically radially oriented .
that investigation was based on a formulation of line scattering polarization that is valid in the absence of magnetic fields .
the suggestion that the magnetic field of the lower solar chromosphere can not be stronger than about 0.01 g unless it is oriented preferentially along the radial direction was based on the sizeable amount of ground - level polarization required to fit the @xmath6 observations of @xcite , and on the assumption that the atomic polarization of the ground - level of must be sensitive to much weaker magnetic fields than the atomic polarization of the upper levels of the d@xmath1 and d@xmath4 lines . on the whole , landi deglinnocenti s ( @xcite )
argument that the observed linear polarization peaks in the cores of the d - lines are due to the presence of ground - level atomic polarization seems very convincing .
however , for a rigorous interpretation of spectropolarimetric observations ( e.g. , @xcite ) it is of fundamental importance to clarify the physical origin of this polarization by carefully investigating how it is actually produced , and modified by the action of a magnetic field of given strength and inclination .
in this letter we shall focus on the `` solar prominence case '' , in which a slab of solar chromospheric plasma at 6000 k , situated at 10 arcsec ( @xmath7 ) above the visible solar limb , and permeated by a magnetic field of given strength and orientation , is illuminated from below ( hence , anisotropically ) by the photospheric radiation field , which is assumed to be unpolarized and with rotational symmetry around the solar radial direction through the scattering point .
the degree of anisotropy of the incident radiation field is calculated as in landolfi & landi deglinnocenti ( 1985 , hereafter ll85 ) , using the limb - darkening data for the na i d - lines from pierce & slaughter ( 1982 ) , and a gaussian absorption profile with @xmath8 ( corresponding to @xmath9 ) .
the resulting anisotropy factors for the two lines are @xmath10 and @xmath11 , where @xmath12 ( with @xmath13 the radiation field tensors ; see , for example , trujillo bueno 2001 and note that @xmath14 ) . in order to investigate this problem ,
we have applied the quantum theory of spectral - line polarization in the limit of complete frequency redistribution , and in the collisionless regime , as developed by @xcite .
the excitation of the atomic system is described by a set of @xmath15 elements , which are the irreducible spherical tensors of the atomic density matrix ( e.g. , the review by @xcite ) .
we adopt a three - level model of consisting of the ground level ( @xmath16 ) , the upper level of the d@xmath1 line ( @xmath17 ) , and the upper level of the d@xmath4 line ( @xmath18 ) .
we also take into account the hyperfine structure ( hfs ) of sodium , due to its nuclear spin with @xmath19 .
we describe the zeeman splittings of the levels most generally in terms of the incomplete paschen - back effect .
in fact , as shown , e.g. , in fig . 5 of ll85 , in the 1050 g range
numerous level crossings occur among the magnetic sublevels of the hfs levels with @xmath20 of p@xmath21 , whereas fields with strengths in the kilogauss range are necessary to produce level crossing within the hfs levels with @xmath22 of s@xmath0 . therefore , besides population imbalances and quantum interferences ( or coherences ) between the magnetic sublevels of each @xmath23-level , we must also take into account coherences between magnetic sublevels pertaining to different @xmath23-levels within a given @xmath24-level
. instead , we neglect coherences between the two @xmath24-levels p@xmath0 and p@xmath21 , since they are presumably of secondary importance for the generation of the line - core polarization peaks , given the sizeable energy separation between the upper levels of the d@xmath1 and d@xmath4 lines .
our model atom then implies 384 @xmath25 elements ( with @xmath26 and @xmath27 ) , which are the unknowns of the linear system representing the statistical equilibrium problem for . in this letter , these quantities are calculated in a reference frame with the @xmath28-axis ( i.e. , the quantization axis ) along the solar radial direction through the scattering point . in summary ,
our approach is similar to that of ll85 , but with the following fundamental improvement . in the expressions of the stokes components of the emission vector ( @xmath29 ) we now take fully into account the energy separation of the various hfs components of the d@xmath1 and d@xmath4 lines , along with their zeeman splittings in the presence of the external magnetic field
this is crucial in order to obtain non - zero linear polarization for the d@xmath1 line .
we have solved numerically the linear system of 384 equations in the unknowns @xmath30 mentioned before , for magnetic strengths between 0 and 1000 g , and for various inclinations ( @xmath31 ) of the magnetic field vector from the solar vertical .
the eight @xmath32 elements quantify the populations of the various @xmath23-levels , and they produce the dominant contribution to the emergent stokes-@xmath33 parameter .
the @xmath34 elements ( the _ alignment _ components ) contribute to the _ linear _ polarization signals , which we quantify by the stokes parameters @xmath35 and @xmath36 . the @xmath37 elements ( the _ orientation _ components ) contribute to the _ circular _ polarization of the scattered radiation .
( we recall that in an aligned atomic system , states of different @xmath38 are unequally populated , while the populations in @xmath39 and @xmath40 are the same .
in contrast , an oriented system is characterized by different populations in the @xmath39 and @xmath40 states .
we are dealing here with an atomic system which is both aligned and oriented ) . the contributions from the longitudinal and transverse zeeman effects are also accounted for , although they become dominant only for relatively strong fields .
given that we are interested in understanding the generation of _ linear _ polarization signals in the presence of weak magnetic fields , we focus here on the _ alignment _ components . in fig .
[ fig : alignment ] , for each @xmath23-level , we show @xmath41 , which quantifies the fractional _ population imbalance _ of the level .
since the spectral dependence of the incident radiation field is practically negligible over the frequency intervals encompassing the zeeman components of each of the two spectral lines ( _ flat - spectrum approximation _ ) , a necessary condition for inducing atomic alignment by means of an unpolarized radiation field is that the illumination of the atomic system be anisotropic .
moreover , atomic orientation can only be originated through the alignment - to - orientation conversion mechanism discussed by @xcite .
figure [ fig : alignment ] shows the sensitivity of @xmath42 to the magnetic field strength and inclination .
first of all , we note that the largest values are obtained for the level p@xmath21 , which can carry atomic alignment even neglecting hfs . on the contrary ,
both the lower and upper levels of the d@xmath1 line , with electronic angular momentum @xmath43 , can carry atomic alignment only because of hfs , as each of these levels splits into two _
polarizable _ hfs levels with @xmath44 and @xmath45 .
however , it is found that only the level @xmath46 can be polarized directly via the anisotropic illumination .
the levels of the d@xmath1 line , instead , are directly sensitive only to radiation intensity , but they nonetheless become polarized when the atomic polarization of the level p@xmath21 is transferred to the level s@xmath0 via spontaneous emission in the d@xmath4 line , and then from the level s@xmath0 to the level p@xmath0 via radiative absorption in the d@xmath1 line , in a process known as _ repopulation pumping _
( e.g. , @xcite ) .
in fact , this explains one of the various remarkable features of fig .
[ fig : alignment ] , i.e. , the fact that the atomic alignment in the lower and upper levels of the d@xmath1 line are equally sensitive to the magnetic strength , independently of the magnetic field inclination . for instance , we see that a _ non - vertical _ magnetic field of the order of 0.01 g is sufficient to produce a serious reduction of the atomic alignment of both the lower and upper levels of the d@xmath1 line .
this is due to hanle depolarization of the s@xmath0 ground level , which occurs when the larmor frequency corresponding to the magnetic field becomes comparable to the inverse lifetime for radiative absorption of that level . nonetheless , the alignment of the level @xmath45 of s@xmath0 is still significant for fields up to 10 g , except for field inclinations close to the van vleck angle ( @xmath47 ) . for non - vertical fields the alignment of the level p@xmath21 is also sensitive to magnetic strengths between 0 and 10 g , but
the depolarization takes place rather smoothly .
an interesting point to note here is the sizeable feedback of the ground - level polarization on the alignment of the @xmath23-levels of p@xmath21 , with the exception of the level @xmath44 .
( this behavior can be understood analytically via inspection of the corresponding transfer rates ) . as previously indicated , such a feedback takes place because the upper level of the d@xmath4 line can be repopulated as a result of absorptions from the _ polarized _ ground level .
the most remarkable feature of fig . [ fig : alignment ] is that , independently of the magnetic field inclination ( e.g. , even for a purely vertical magnetic field ) , the atomic alignment of each of the two levels involved in the d@xmath1 line transition is suddenly reduced for magnetic strengths larger than 10 g , and practically vanishes for strengths larger than 100 g. we stress the fact that this depolarization is _ not _ due to the hanle effect , as it occurs also for vertical fields .
a thorough investigation of this phenomenon shows that the vanishing of atomic alignment in the levels with @xmath43 sets in when the electronic and nuclear angular momenta , @xmath48 and @xmath49 , are decoupled , for the atom in the excited state p@xmath21 . in the case of
, this decoupling is reached in the limit of the complete paschen - back effect of the level p@xmath21 , i.e. , for magnetic strengths @xmath3 . in such regime , it is found that the transfer of atomic alignment from the level @xmath46 to the ground level is inhibited . at the same time
, the alignment of the level @xmath45 of p@xmath21 must vanish as well .
the analytical proof of these properties will be given in a forthcoming paper ( casini et al .
2001 , in preparation ) .
it is also of interest to note that the repopulation pumping process works efficiently in thanks to the fact that the hfs of the level @xmath46 is of the same order of magnitude of its natural width .
if the frequency intervals between the hfs levels of p@xmath21 were instead substantially smaller than the natural width of this level , then we would have a negligible hfs interaction during the lifetime of the level @xmath46 , with the result of a drastic reduction in the efficiency of the repopulation pumping process that polarizes the ground level of sodium , regardless of the magnetic field strength .
the theory of the hanle effect for a two - level atom devoid of hfs predicts no modication of the emergent linear polarization with increasing strength of a magnetic field oriented parallel to the symmetry axis of the incident radiation ( e.g. , @xcite ) . for this reason , and given the conclusions of the previous section , it is of great interest to investigate the emergent polarization of the d - lines for 90@xmath50 scattering events as a function of the strength of a vertical magnetic field . for this case , and choosing the reference direction for positive stokes @xmath35 parallel to the limb , the only non - zero stokes parameter is @xmath35 .
figure 2 shows profiles of @xmath51 , where @xmath52 indicates the peak intensity of the emission line , for magnetic strengths between 0 and 100 g. the top panels refer to the solution of the statistical equilibrium problem outlined in [ sec : polar ] .
the results shown in the two bottom panels , instead , are obtained assuming that the ground level of is totally unpolarized ( e.g. , by the presence of depolarizing collisions ) .
the first interesting feature of the d@xmath4 line polarization is the _ increase _ of the linear polarization degree as the strength of the vertical magnetic field increases beyond 10 g. this is caused by the interferences , @xmath53 ( not shown in fig .
[ fig : alignment ] ) , of the hfs levels in the p@xmath21 level .
our calculations for _ inclined _ fields with @xmath54 ( not given here ) show first a decrease and then an increase of the linear polarization , but the maximum polarization amplitude of the d@xmath4 line corresponds to the @xmath55 case .
in general , the polarization of both d@xmath1 and d@xmath4 fluctuates significantly with strength and inclination .
we note that the polarization of the d@xmath4 line is practically unaffected by ground - level polarization for fields larger than 10 g. as shown in fig .
[ fig : alignment ] , for weaker fields the existing ground - level polarization has a significant feedback on the alignment of the @xmath46 level , which in turn produces a significant but small enhancement of the emergent linear polarization in the d@xmath4-line core , with respect to the case of totally unpolarized ground level ( see fig . [
fig : profiles ] ) . as shown in fig .
[ fig : profiles ] , for magnetic strengths @xmath56 , the emergent linear polarization of the d@xmath1 line owes its very existence to the presence of ground - level polarization .
note that its stokes @xmath35 profile is _
antisymmetric_. of @xcite .
see also @xcite .
] in fig .
[ fig : profiles ] we can see that a 1 g _ vertical _ field has practically no effect on the emergent linear polarization , while a 10 g _ vertical _ field reduces the signal by a factor of two , as a result of the significant reduction of the atomic alignment of the hfs levels of the upper level p@xmath0 ( see fig .
[ fig : alignment ] ) .
what we see for stronger fields in the corresponding top panel of fig .
[ fig : profiles ] is the result of the combined action of the atomic alignment of the upper level of the d@xmath1 line and of the transverse zeeman effect .
since the zeeman splittings for the magnetic strengths of interest are determined by the incomplete paschen - back effect , and the d - lines are the combination of transitions between @xmath23-levels weighted by different land @xmath57-factors , the stokes @xmath35 profiles can be asymmetric , even if the ground level is forced to be totally unpolarized ( see bottom panels of fig .
[ fig : profiles ] ) .
nonetheless , it can be analitically proven that the wavelength integrated stokes @xmath35 parameter of d@xmath1 is always zero ( see ll85 ) .
it is also of interest to point out that , if the field is horizontal with randomly distributed azimuth , then the stokes @xmath35 amplitude of the d@xmath1 line is reduced by a factor of four for a magnetic strength of 1 g , with respect to the non - magnetic case . in general ,
the linear polarization of the d@xmath1 line is dominated by the transverse zeeman effect for magnetic strengths @xmath58 ( the alignment of the d@xmath1 levels is practically zero for such field intensities ) . for example , from fig .
[ fig : profiles ] we see that the amplitude of the stokes @xmath35 signal in the presence of a _ vertical _ field with @xmath59 is comparable to the amplitude produced by scattering processes in the absence of magnetic fields , but the linear polarization signature has a profile characteristic of the transverse zeeman effect , with a polarization peak at the line core . on the other hand , the stokes @xmath35 signal in the d@xmath4 line for @xmath59 is still `` scattering like , '' being determined essentially by the contribution of the alignment of the level @xmath46 .
in fact , for this line , the signature of the transverse zeeman effect begins to appear only for fields @xmath60 .
one of the most interesting results of this investigation is that the atomic polarization of the hfs levels of the @xmath61 and @xmath62 states of is practically negligible for @xmath63 , and virtually vanishes for @xmath64 , even for a purely vertical field .
consequently , the characteristic antisymmetric scattering polarization signature of the d@xmath1 line is practically suppressed in the presence of fields larger than 10 g , regardless of their inclination .
concerning the observable effects , we find that the scattering polarization amplitude of the d@xmath4 line increases steadily with the magnetic strength , in the case of vertical fields larger than 10 g , whereas the contribution of atomic alignment to the linear polarization of the d@xmath1 line rapidly decreases . on the contrary , for vertical fields such that @xmath56 ( or , alternatively , for turbulent or canopy - like fields with a predominance of much weaker fields ) it is possible to have a non - negligible scattering polarization signal for the d@xmath1 line , but then the maximum d@xmath4 core amplitude corresponds to the @xmath65 case . from this
we tentatively conclude that spectropolarimetric observations of the `` quiet '' solar chromosphere showing significant scattering polarization peaks in both the d@xmath1 and d@xmath4 line cores can not be interpreted in terms of one - component magnetic field models , suggesting that the magnetic structuring of the solar chromosphere could be substantially more complex than previously thought .
for instance , in the presence of a topologically complex distribution of `` weak '' solar magnetic fields , the d@xmath4 line core would respond mainly to the strongest and preferentially radially oriented fields , while the d@xmath1 line to the weakest and more randomly oriented fields .
it remains to be seen whether or not this conclusion is validated after we take fully into account radiative transfer effects and the role of _ dichroism _ ( @xcite ) on the emergent polarization of the `` enigmatic '' na i d - lines .
one of the authors ( j.t.b . ) is grateful to the high altitude observatory ( u.s.a . ) and to the university of florence ( italy ) for short - term visitor grants , which have facilitated this collaboration .
this work has been partly funded by the spanish pnaya through project aya2001 - 1649 . | we demonstrate that the atomic alignment of the hyperfine - structure components of the ground level s@xmath0 of and of the upper level p@xmath0 of the d@xmath1 line are practically negligible for magnetic strengths @xmath2 , and virtually zero for @xmath3 .
this occurs independently of the magnetic - field inclination on the stellar surface ( in particular , also for vertical fields ) .
consequently , the characteristic antisymmetric linear - polarization signature of the scattered light in the d@xmath1 line is practically suppressed in the presence of magnetic fields larger than 10 g , regardless of their inclination .
remarkably , we find that the scattering polarization amplitude of the d@xmath4 line increases steadily with the magnetic strength , for vertical fields above 10 g , while the contribution of alignment to the polarization of the d@xmath1 line rapidly decreases .
therefore , we suggest that spectropolarimetric observations of the `` quiet '' solar chromosphere showing significant linear polarization peaks in both d@xmath1 and d@xmath4 can not be interpreted in terms of one - component magnetic field models , implying that the magnetic structuring of the solar chromosphere could be substantially more complex than previously thought . |
the laser interferometer space antenna ( lisa ) , a nasa and esa space mission to detect gravitational waves ( gws ) in the @xmath0@xmath1 hz range @xcite , will produce time series consisting of the superposition of the signals from millions of sources , from our galaxy to the edge of the observable universe .
some of the signals ( such as those from extreme mass - ratio inspirals , or emris ) are very complex functions of the physical parameters of the sources ; others ( such as those from galactic white - dwarf binaries ) are simpler , but their resolution will be confused by the presence of many other similar signals overlapping in frequency space .
thus , data analysis is integral to the lisa measurement concept , because no source can be identified without first carefully teasing out its individual voice from the noisy party of each data set .
understanding data analysis is therefore important to demonstrate that lisa can meet its science requirements , and to translate them into decisions about instrument design .
the idea of the mock lisa data challenges ( mldcs ) arose in late 2005 from this realization .
the mldcs have the purpose of encouraging and tracking progress in lisa data - analysis development , and ( as a useful byproduct ) of prototyping the lisa computational infrastructure : common data formats , standard models of the lisa orbits , noises and measurements , software to generate waveforms and to simulate the lisa response , and more .
the mldcs are a coordinated ( but voluntary ) effort in the gw community , whereby a task force chartered by the lisa international science team periodically issues a number of data sets containing synthetic noise and gw signals from sources of undisclosed parameters ; challenge participants return detection candidates and parameter estimates , together with descriptions of their search methods .
these results are then compiled and compared to the previously secret challenge `` key '' .
challenge 1 , issued in jun 2006 with results due in dec 2006 ( see @xcite ) , tackled the detection and parameter characterization of _ verification binaries _
( galactic binaries of known frequency and position ) ; of loud unknown galactic binaries , either alone or in small , moderately interfering groups ; and of relatively loud inspirals of supermassive black - hole ( mbh ) binaries .
all sources were represented by somewhat idealized waveforms , and they were staged on instrument noise alone .
altogether , challenge 1 successfully demonstrated the detection of all three source classes .
ten research collaborations submitted entries , adopting a variety of methods ( template - bank , stochastic and genetic matched filtering ; time frequency ; tomography ; hilbert transform ) . despite the short timescale , all challenges were `` solved '' by at least one group , although some searches locked on strong secondary maxima of the source parameter probabilities .
more important , challenge 1 helped set the playing field and assemble the computational tools for the more realistic challenge 2 .
challenge 2 , issued in jan 2007 with results due at the end of jun 2007 , raised the bar by proposing three complex subchallenges .
data set 2.1 contained a full population of galactic binary systems ( about 26 million sources ) .
data set 2.2 contained a different realization of the galaxy , plus 46 mbh binary inspirals with single - interferometer signal - to - noise ratios ( snrs ) between 10 and 2000 and a variety of coalescence times , and five emris with snrs between 30 and 100 .
last , five more data sets ( denoted 1.3.15 , since they were actually released at the time of challenge 1 ) contained a single emri signal over instrument noise alone .
see @xcite for more details about the signal models and the exact source content of the data sets .
thirteen collaborations ( comprising all the researchers listed as participants in the byline of this article , and most task force members ) submitted a total of 22 entries , including a proof - of - principle analysis for stochastic backgrounds performed on data set 2.1 .
altogether , challenge 2 successfully demonstrated the identification of @xmath2 galactic binaries , the accurate estimation of mbh inspiral parameters , and the positive detection of emris . in the rest of this paper
, we describe some of its highlights .
all the solutions submitted by participating groups , together with technical write - ups of their methods and findings , can be found at the url www.tapir.caltech.edu/~mldc/results2 .
a few groups are also contributing descriptions of their work to the proceedings of this conference .
five groups submitted galactic - binary catalogs for data sets 2.1 and 2.2 : glig : : a collaboration of research groups at institutions in the uk , united states and new zealand developed a reversible - jump markov chain monte carlo ( rj mcmc ) code that can sample models with different numbers of sources ; for lack of time , however , they only submitted parameter sets for the verification binaries .
impan : : krlak and blaut developed an @xmath3-statistic , template - bank based matched - filtering search @xcite , and submitted parameters for 404 sources for data set 2.1 .
mtjpl : : the montana state
jpl collaboration used a metropolis
hastings monte carlo ( mhmc ) code that ran separately for overlapping frequency bands and for different hypotesized numbers of sources ; model comparison was then used to determine the most probable number of sources in each band .
the collaboration submitted parameter sets for 19,324 sources for data set 2.1 , and 18,461 sources for data set 2.2 .
prixwhelanaei : : prix and whelan developed an @xmath3-statistic , template - bank based matched - filtering search using a hierarchical scheme that enforced trigger coincidence between tdi observables , followed by a coherent follow - up using noise - orthogonal tdi combinations @xcite .
they submitted parameter sets for 1777 sources for data set 2.1 , and 1737 sources for data set 2.2 .
utb : : nayak , jimenez and mohanty used a tomographic reconstruction technique and submitted parameters for 3862 sources in data set 2.1 .
evaluating the performance of these searches brings up several problems of principle : while we know that many of the @xmath4 million galactic sources that were injected into the data sets can not be recovered because they are ( relatively ) too weak , we do not have a precise estimate of how many sources should be recoverable .
thus , the notion of false dismissal is not well defined . to make matters worse ,
the notion of false positive is also ill defined , because a single recovered source can provide a good fit to the blended signal from several injected sources , which may well be the `` right '' answer with the knowledge we have , since it is the best fit to the data with the smallest numbers of parameters .
the task force devoted considerable time to the analysis of galactic - binary searches , and we do not have space here to describe all the treatments that we applied to the data .
instead we will limit our report to techniques that pair up individual recovered sources with individual sources from the challenge key , with the understanding that this will overestimate the number of false positives , and say nothing about false dismissals . ]
one way to proceed is to associate the reported and injected sources that have the strongest signal correlation , limiting the search to the bright ( @xmath5 ) injected sources that could in principle have been found : in the left panel of figure [ fig : correlation ] we show the distribution of correlations generated with this procedure for data set 2.1 .
detections with the highest correlations can be considered `` safest , '' while those with the lowest correlations probably represent spurious associations .
another procedure is to associate the reported and injected sources that minimize the _ doppler metric _ that spans the frequency sky - location subspace of the full parameter space , and automatically maximizes correlation over the _ extrinsic parameters _ ( amplitude , polarization , inclination , initial phase ) : the right panel of figure [ fig : correlation ] shows the resulting distribution of correlations .
the utb entry , which includes frequency and sky position but not the extrinsic parameters , can only be plotted this way .
generally , this is a softer criterion , and all searches do better by it ( especially the prixwhelanaei entry , whose long - wavelength approximation for the lisa response is prone to extrinsic - parameter errors ) . ]
figure [ fig : paramerrors ] shows the snrs of the recovered sources and the errors for the intrinsic and extrinsic parameters , computed after associating sources by correlation ( and by doppler metric only for the utb entry ) , again for data set 2.1 .
the error in frequency is in most cases within a small fraction of a fourier bin , and the errors in sky position are within a few degrees ; by contrast , the errors in the amplitude and in the ( extrinsic ) orientation angles are larger .
the @xmath6 graph for the prixwhelanaei suggests a systematic error in the definition of initial phase . altogether , these challenges demonstrated a solid capability in analyzing signals from the galaxy and resolving a large number of binaries .
as we mentioned , deciding how well they were recovered is not an easy question to answer , because of the difficulty of defining ( at least operationally ) a notion of _ identity _ for recovered sources .
these problems deserve careful attention in the future .
four groups reported parameter sets for the mbh binaries in data set 2.2 : aeise : : babak and porter used an @xmath3-statistic , template - bank based matched - filtering search , followed by an mcmc stage .
mtaei : : cornish and porter used an mhmc matched - filtering search with a _ frequency - annealed _ scheme where shorter , lower - frequency templates are used in the initial phases of the search and then progressively extended .
jplct : : the jpl cit collaboration used a three - stage pipeline consisting of a track search in the time frequency ( tf ) plane , followed by template - bank based matched filtering , and by an mcmc refinement @xcite .
lisafrance : : the french collaboration used a tf track search alone , and therefore reported only mass and time - of - coalescence parameters .
all the four mbh binaries in the data set ( mbh-1 , 2 , 4 and 5 , with total snrs @xmath7 2583 , 25 , 174 and 117 ) were positively detected by aeise , mtaei , and jplct ; the tf method used by lisafrance identified mbh-1 and 4 , but not mbh-2 , and could report only a time of coalescence for mbh-5 .
@l|r@r@r@r@r@r@r@r@r@r & snr & @xmath8 & @xmath9 & @xmath10 & @xmath11 & @xmath12 & @xmath13 & @xmath14 & @xmath15 & @xmath16 + & & @xmath17 & @xmath17 & @xmath18 & @xmath19 & @xmath20 & @xmath20 & @xmath20 & @xmath19 & @xmath19 + aeise & 2247.60 & 147.0 & 3386.2 & 19.0 & @xmath21 & @xmath22 & 77.8 & @xmath23 & 13.8 & @xmath24 + mtaei & 2583.34 & 7.9 & 9.1 & 3.6 & 1.65 & @xmath25 & 3.5 & 4.94 & 1.2 & @xmath26 + jplct & 2582.42 & 27.5 & 28.7 & 16.0 & 4.81 & 12.2 & 12.3 & @xmath27 & @xmath28 & 7.45 + lisafrance & & 2944.6 & 72.7 & 67.8 & & & & & & + @l|r@r@r@r@r@r@r@r@r@r & snr & @xmath8 & @xmath9 & @xmath10 & @xmath11 & @xmath12 & @xmath13 & @xmath14 & @xmath15 & @xmath16 + & & @xmath18 & @xmath29 & @xmath18 & @xmath20 & @xmath20 & @xmath30 & @xmath30 & @xmath30 & @xmath19 + aeise & 81.38 & 1396.3 & 149.9 & 3.4 & @xmath31 & @xmath32 & 574.0 & @xmath33 & @xmath34 & @xmath35 + mtaei & 174.13 & 148.8 & 21.3 & 2.1 & @xmath36 & @xmath37 & 15.1 & @xmath38 & @xmath39 & @xmath40 + & 174.11 & 17.1 & 20.5 & 33.3 & @xmath41 & @xmath42 & 16.7 & @xmath43 & @xmath44 & @xmath45 + jplct & 174.11 & 4.2 & 9.5 & 2.1 & @xmath46 & @xmath47 & 1.2 & @xmath48 & @xmath49 & @xmath50 + & 174.12 & 124.7 & 9.0 & 35.4 & @xmath51 & @xmath52 & 6.3 & @xmath53 & @xmath54 & @xmath53 + lisafrance & & 34394.1 & 1804.1 & 280.8 & & & & & & + tables [ tab : mbh1 ] and [ tab : mbh4 ] show fractional parameter errors for mbh-1 and 4 , together with the snr recovered by the best - fit candidates , computed as @xmath55 with @xmath56 the usual noise - weighted inner product , and @xmath57 and @xmath58 two noise - orthogonal tdi observables ( see , e.g. , @xcite ) . in table
[ tab : mbh1 ] , we see that the jplct search for mbh-1 locked onto a secondary probability maximum with @xmath59 only slightly lower than the optimal value , but with sky positions off by several degrees , which also led to errors in the other parameters . the jplct authors hypothesize that this was caused by first subtracting a rough mbh-1 model from the data , then subtracting the resolvable galactic binaries , and finally refining the mbh search .
mbh-4 ( table [ tab : mbh1 ] ) is an interesting example of a `` true '' bimodal probability distribution for the source parameters .
mtaei and jplct each submitted two candidates , placed at rather different sky locations , quoting relative probability ratios of 1:1 and 1.18:1 . in this case , it was probably the sky position and orientation of this source that conspired to degrade lisa s positional sensitivity , since they resulted in a very weak signal in one of the noise - orthogonal observables .
altogether , this challenge demonstrated a solid capability in the detection and parameter estimation of mbh inspirals with moderate snrs , even in the presence of a strong galactic background , at least if the inspirals can be considered close to our idealized model : circular and adiabatic with negligible spin effects .
these restrictions are being relaxed for the upcoming challenge 3 .
three groups reported parameter sets for the emris in data sets 1.3.11.3.4 .
no group tackled the problem of detecting these systems in data set 2.2 ( on top of the galactic background ) .
bbgp : : babak and colleagues used an mcmc matched - filtering search that modeled the signal with a sequence of progressively longer templates ( a _ time - annealed _ scheme ) .
etfag : : gair , mandel and wen used a tf track search that ( for now ) targeted only the intrinsic parameters and sky position @xcite .
mt : : cornish used an mhmc matched - filtering search , running it in parallel on individual month - long segments , which were subsequently strung together for full detections .
table [ tab : emri ] shows typical recovered snrs and errors . while it is clear that the matched - filtering searches locked on several secondary probability maxima with comparable probabilities , the recovered snrs correspond to solid detections with exceedingly low false - alarm probabilities .
the errors are quoted as fractions of the allowed parameter ranges , and they are quite large . intriguingly , the tf search was the most accurate in determining the sky position . altogether , these challenges demonstrated a positive capability of detecting emris , at least if their signals are similar in complexity to the _ kludge _ waveforms used in this challenge @xcite ; however , the prospects for accurate parameter estimation are still uncertain , and a good focus for further challenges .
@l@|@l@@l@l@l@l@l@l@l@l@l@l@l@ & snr & & & & & & & & & & + bbgp & 74.86 & @xmath60 & @xmath61 & @xmath62 & @xmath63 & @xmath64 & @xmath65 & @xmath66 & @xmath67 & @xmath68 & @xmath69 + & 72.96 & @xmath70 & @xmath71 & @xmath72 & @xmath73 & @xmath74 & @xmath65 & @xmath66 & @xmath67 & @xmath68 & @xmath75 + & 72.52 & @xmath76 & @xmath77 & @xmath78 & @xmath79 & @xmath80 & @xmath81 & @xmath82 & @xmath83 & @xmath84 & @xmath85 + & 72.49 & @xmath76 & @xmath77 & @xmath78 & @xmath86 & @xmath80 & @xmath81 & @xmath87 & @xmath83 & @xmath84 & @xmath88 + & 70.59 & @xmath89 & @xmath90 & @xmath91 & @xmath92 & @xmath93 & @xmath94 & @xmath94 & @xmath95 & @xmath64 & @xmath96 + etfag & & @xmath97 & @xmath98 & & & @xmath99 & @xmath100 & @xmath81 & @xmath101 & @xmath102 & + mt & 74.85 & @xmath103 & @xmath104 & @xmath105 & @xmath72 & @xmath106 & @xmath107 & @xmath108 & @xmath109 & @xmath110 & @xmath111 + & 76.52 & @xmath112 & @xmath113 & @xmath60 & @xmath114 & @xmath115 & @xmath116 & @xmath117 & @xmath118 & @xmath114 & @xmath119 +
we are very excited about the outcome of the first two mldcs , which have given a convincing demonstration that a significant portion of the lisa science objectives could already be achieved with techniques that are currently in hand .
most of the research groups that participated in challenge 1 have successfully made the transition to the greater complexity of challenge 2 .
challenge 3 will continue to move in the direction of more realistic signals , featuring chirping galactic binaries and precessing binaries of spinning mbhs .
it will also include two new classes of signals : an isotropic primordial gw background and bursts from the cusps of cosmic strings . between july and
dec 2007 we are also running challenge 1b , a repeat of challenge 1 conceived to provide a softer entry point for research groups new to the mldcs .
the mldc conventions , file formats , and software tools ( see lisatools.googlecode.com ) have matured to the point where interested parties can use them to generate a variety of data sets .
this enables a wealth of interesting side investigations , such as the studies of the lisa science reach that are now being undertaken by the lisa science team . to obtain more information and to participate in the mlcds ,
see the official mldc website ( astrogravs.nasa.gov/docs/mldc ) and the task force wiki ( www.tapir.caltech.edu/listwg1b ) .
sb , ekp , and jtw s work was supported by the german aerospace center ( dlr ) and the max - planck society .
mb acknowledges funding from nasa grant nng04gd52 g and support from the nasa center for gravitational wave astronomy at university of texas at brownsville ( nag5 - 13396 ) . nc was supported by nasa grants nng05gi69 g and nnx07aj61 g .
db and sf acknowledge funding from nsf grant phy-0601459 and the ligo laboratory .
jc s , cc s and mv s work was carried out at the jet propulsion laboratory , california institute of technology , under contract with the national aeronautics and space administration .
jg acknowledges support from st catharine s college , cambridge .
i m would like to thank the brinson foundation , nasa grant nng04gk98 g and nsf grant phy-0601459 for financial support .
rp s work was supported by the max - planck society .
mv was supported by lisa mission science office and by jpl s human resources development fund .
lw s work was supported by the alexander von humboldt foundation s sofja kovalevskaja programme , funded by the german federal ministry of education and research .
99 bender p , danzmann p and the lisa study team 1998 `` laser interferometer space antenna for the detection of gravitational waves , pre - phase a report '' * mpq 233 * ( garching : max - planck - institt fr quantenoptik ) arnaud k a et al .
( the mock lisa data challenge task force ) 2006 _ laser interferometer space antenna : 6th international lisa symp .
( greenbelt , md , 1923 jun 2006 ) _ ed merkowitz s m and livas j c ( melville , ny : aip ) p 619 ; _ ibid . _
p 625 arnaud k a et al .
( the mock lisa data challenge task force and challenge 1 participants ) 2007 _ class .
* 24 * s529 arnaud k a et al .
( the mock lisa data challenge task force ) 2007 _ class .
* 24 * s551 p. jaranowski , a. krlak , and b. f. schutz , phys .
d * * 5**8 , 063001 ( 1998 ) .
a. krlak , m. tinto , and m. vallisneri , _ phys .
d _ , * 70 * , 022003 ( 2004 ) .
various authors 2008 proc .
7th amaldi conf . on gravitational waves ( sydney , 814 july 2007 )
_ j. phys .
* xx * brown d a , crowder j , cutler c , mandel i and vallisneri m 2007 _ class .
* 24 * s595 vallisneri m , crowder j and tinto m 2007 `` sensitivity and parameter - estimation precision for alternate lisa configurations '' _ preprint _
arxiv.org/0710.4369 gair j r , mandel i and wen l 2007 proc .
7th amaldi conf . on gravitational waves ( sydney , 814 july 2007 ) ,
_ preprint _
prix r and whelan j t 2007 _ class .
grav . _ * 24 * s565 ; _ poster _ www.ligo.caltech.edu/docs/g/g070462-00.pdf | the mock lisa data challenges are a program to demonstrate lisa data - analysis capabilities and to encourage their development .
each round of challenges consists of several data sets containing simulated instrument noise and gravitational - wave sources of undisclosed parameters .
participants are asked to analyze the data sets and report the maximum information about source parameters .
the challenges are being released in rounds of increasing complexity and realism : in this proceeding we present the results of challenge 2 , issued in january 2007 , which successfully demonstrated the recovery of signals from supermassive black - hole binaries , from 20,000 overlapping galactic white - dwarf binaries , and from the extreme mass - ratio inspirals of compact objects into central galactic black holes . |
i am grateful to the organizers of lattice-01 for the opportunity to present this talk .
i thank claude bernard , joachim hein , takashi kaneko , francesco knechtli , kostas orginos , and howard trottier for providing results used in this talk .
this work was supported by the u.s .
s. naik , nucl . phys . *
b316 * , 238 ( 1989 ) ; c. bernard ( milc ) , phys . rev .
d * 58 * , ( 1998 ) 014503 ; c. bernard ( milc ) , phys .
d * 55 * , 1133 ( 1997 ) ; g.p .
lepage , nucl .
( proc . suppl . )
* 60a * 267 ( 1998 ) ; j.f .
lage and d.k .
sinclair , phys .
d * 59 * , ( 1999 ) 014511 ; nucl .
( proc . suppl . )
* 63 * , 892 ; k. orginos and d. toussaint , phys .
d * 59*:014501 ( 1999 ) ; nucl . phys .
b ( proc . suppl . ) * 73 * , 909 ( 1999 ) ; k. orginos , r.l
. sugar and d. toussaint , phys .
d * 60 * ( 1999 ) 054503 , nucl .
b(proc . suppl . )
* 83 * , 878 ( 2000 ) . ;
luo , phys .
d * 57 * , 265 ( 1998 ) ; g.p .
lepage , phys .
d * 59 * , ( 1999 ) 074502 ; c. bernard ( milc ) , phys .
d * 61 * , 111502 ( 2000 ) ; u.m .
heller , f. karsch and b. sturm , phys .
d * 60 * , ( 1999 ) 114502 ; f. karsch , nucl . phys .
b ( proc . suppl . )
* 60a * , 169 ( 1998 ) ; b. beinlich , a. bicker , f. karsch , e. laermann , a. peikert , nucl .
( proc . suppl . )
* 63 * ( 1998 ) 895 ; o. kaczmarek , f. karsch , e. laermann , nucl . phys .
( proc . suppl . )
* 73 * ( 1999 ) 441 ; f. karsch , e. laermann , a. peikert , ch .
schmidt , s. stickan , nucl .
( proc . suppl . )
* 94 * ( 2001 ) 411 .
s. kim and s. ohta , nucl .
( proc . suppl . )
* 47 * ( 1996 ) 350 ; nucl .
( proc . suppl . )
* 63 * ( 1998 ) 185 ; nucl .
( proc . suppl . )
* 73 * ( 1999 ) 195 ; phys .
d * 61 * ( 2000 ) 074506 .
j. gasser and h. leutwyler , nucl .
b250 * ( 1985 ) 465 . | i summarize recent developments in spectrum calculations using kogut - susskind quarks .
theoretical developments include one - loop computations with improved actions .
i present some recent simulation results , mostly from a milc collaboration project using three flavors .
effects of dynamical quarks are clearly seen in the isovector @xmath0 meson propagator and in the mass ratio `` j '' .
the kogut - susskind formulation of lattice fermions , often called `` staggered fermions '' , has long been a popular formulation for qcd simulations including dynamical quarks .
the basic reason for this is the single remaining exact chiral symmetry , which is sufficient to protect the quark mass from additive renormalization . from a practical viewpoint ,
an important consequence is a lower limit on the eigenvalues of @xmath1 , which insures that the simulation will be well behaved and not encounter `` exceptional configurations . '' also , but less importantly , there are simply fewer fermionic degrees of freedom to handle .
the combined effect of these advantages is that we can use our limited computer power to push to smaller dynamical quark masses , or larger physical sizes , or perhaps to larger statistics , by using kogut - susskind quarks .
the price that we pay is that it can be a painful exercise to figure out the lattice implementation of your desired operator and , more importantly , flavor symmetry is broken by effects of order @xmath2 .
although they are of order @xmath2 , in practice the effects of flavor symmetry breaking are unpleasantly large at accessible lattice spacings .
the situation can be greatly improved by the use of an improved action which suppresses the coupling of the quarks to high momentum gluons , as well as fixing up the quarks dispersion relation .
see ref .
@xcite for some relevant references .
the most important point is that exchange of a gluon with momentum near @xmath3 can scatter a low momentum quark from one corner of the brillioun zone into another , resulting in a mixing of different flavors .
`` fattening '' the links , by averaging paths in the parallel transport , effectively puts a form factor into the quark - gluon vertex which suppresses such exchanges . ' '' '' + = 3.00 in ' '' '' + ' '' '' + = 3.30 in ' '' '' + ' '' '' + = 2.90 in ' '' '' + ' '' '' + = 3.00 in ' '' '' + ' '' '' + = 3.00 in ' '' '' + ' '' '' + = 3.10 in ' '' '' + at this conference we have seen the first results of one - loop calculations done with improved kogut - susskind actions . in ref .
@xcite the quark mass renormalization is calculated , and in ref .
@xcite renormalization constants for several operators are shown .
the results are encouraging , with none of the coefficients coming out surprisingly large .
in contrast , the conventional kogut - susskind action often has large one - loop corrections .
this can again be understood in terms of a form factor suppressing coupling to high momentum gluons@xcite , so that a diagram like fig .
2b of ref .
@xcite will not get large contributions when the quark is in another corner of the brillioun zone .
in other words , we suppress the unwanted doublers in the loop diagrams .
figure [ scaling_fig ] shows the result of one scaling test of the improved kogut - susskind action . in this figure
we plot the @xmath4 mass in units of the static quark potential as a function of lattice spacing . for the length scale we use @xmath5 , the distance where @xmath6 .
all these points are from quenched simulations , and they are all interpolated to the quark mass where @xmath7 , or @xmath8 mev .
next one is motivated to ask whether it is practical to further improve the kogut - susskind action . at this conference dipierro and
mackenzie reported on experiments with empirically tuning the coefficients of the paths in the action@xcite .
the gains were limited , which they interpret as evidence of the need for four quark operators . in ref .
@xcite a one - loop improved action was presented , and methods for handling these four quark operators by introducing auxiliary bosonic fields were sketched . taking a different approach , a. hasenfratz and knechtli combined fattening of the links with a projection back onto unitary su(3 ) matrices in the `` hyp '' , for `` hypercubic blocking '' action@xcite .
this action produces smaller mass splittings among the pions than the @xmath2 tadpole action .
these authors also describe an algorithm for dynamical simulation of this action , which can not be expressed in a simple way as a sum over paths . in another theoretical development presented at this conference ,
levkova and manke have worked out an action for unimproved dynamical kogut - susskind quarks on an anisotropic lattice@xcite .
while their primary motivation is high temperature qcd , this approach could be useful for spectroscopy of glueballs , hybrids and excited states .
the milc collaboration has used the @xmath2 tadpole improved action for a set of hadron spectrum calculations@xcite .
these calculations used three flavors of dynamical quarks , as well as a quenched run and one two - flavor run , on lattices tuned to match the lattice spacing at about 0.13 fermi . for masses larger than the strange quark mass @xmath9 three degenerate flavors
were used , while for smaller masses one quark mass was held fixed at @xmath9 .
quark masses down to @xmath10 , or @xmath11 , were used .
the lattice size was @xmath12 , or @xmath13 fm .
in addition , some preliminary results at a finer lattice spacing of @xmath14 fm are available .
the big advantage of using lattices matched in lattice spacing and physical size for different numbers and masses of quarks is that the effects of dynamical quarks can be convincingly exposed .
the simplest quantity that shows this is the static quark potential .
figure [ potmatch_fig ] shows the potential in quenched qcd and in three - flavor qcd with all the quarks at the strange quark mass .
the slopes of these two potentials agree at the point chosen to define the length scale , but their overall shape is different .
notice that we do not see , and do not expect to see , string breaking in the distance range shown here . to quantify the change in shape we look at dimensionless quantities such as @xmath15 or @xmath16 .
figure [ r0_fig ] shows @xmath15 for the three flavor runs .
it also contains two flavor results with wilson quarks from cppacs and sesam , and a two - flavor improved kogut - susskind point from milc .
one aspect of lattice spectroscopy where sea quarks are expected to have important effects is in the chiral behavior of hadron masses . in the case of the goldstone pion , at lowest order
the squared pion mass is proportional to the quark mass , so corrections to this behavior can be displayed by plotting the squared pion mass divided by the quark mass .
figure [ mpisq_over_fig ] is such a plot for the milc runs in both full and quenched qcd .
the increase at the right side is understood as the transition from chiral behavior to heavy quark behavior , while the much sharper upturn at low quark mass is the expected chiral logarithm .
although the quenched and three flavor curves look similar , the theoretical expectation is that the three flavor line has a finite limit as @xmath17 , while the quenched line diverges logarithmically .
the quenched pion masses in fig .
[ mpisq_over_fig ] are similar to earlier quenched results of kim and ohta using the conventional kogut - susskind action at a lattice spacing of about 0.46 fm@xcite .
a crude comparison is in fig .
[ mpisq_ko_fig ] , where i have arbitrarily rescaled both the vertical and horizontal axes to match the results at the strange quark mass . while it is tempting to simply fit fig .
[ mpisq_over_fig ] to the chiral behavior predicted in ref .
@xcite , it is probably necessary to take account of the remaining flavor symmetry violations .
claude bernard has been fitting this data both directly to the continuum form , and to a form that takes into account flavor symmetry violations using the lagrangian of lee and sharpe@xcite with empirically determined non - goldstone pion masses in each loop , to calculate corrections to the pion mass@xcite . in the quenched case , where the coefficient of the chiral logarithm can be a parameter
, reasonable fits can be obtained both with and without the flavor symmetry breaking corrections .
however , the coefficient of the logarithm comes out about 0.06 with the simple fitting , and about 0.14 when the flavor symmetry breaking is included .
when the three flavor data is fit , with the coefficients of the logarithms determined by the chiral theory , good fits can only be obtained when the flavor symmetry breaking is included .
an application of these chiral corrections to the pion mass was presented at this conference by the ohio state group@xcite . here a particular combination of chiral lagrangian parameters was computed and found to be inconsistent with @xmath18 .
the boulder `` hyp '' action@xcite was used for the valence quarks in this project . in view of bernard
s observations on the effects of flavor symmetry breaking , it is appropriate that they used a valence quark action which makes these effects as small as possible .
the milc spectrum calculations on matched quenched and three flavor lattices@xcite found some interesting differences between the quenched and full qcd spectra ( and at least one interesting lack of differences ) .
the most striking difference is in the coupling of hadrons to two particle intermediate states , which is presumably present in full qcd calculations but is represented only by `` hairpin '' diagrams in quenched calculations .
figure [ a0_decay_fig ] shows masses obtained for the isovector @xmath0 propagator on matched quenched and full qcd lattices .
we interpret the low quark mass behavior of the full qcd results as an avoided level crossing between a @xmath19 state and a two meson @xmath20 state . in simple language
, we are seeing a meson decay in the lattice simulations . ' '' '' + = 3.00 in ' '' ''
+ ' '' '' + = 3.00 in ' '' '' + the ukqcd collaboration pointed out that the ratio `` j '' , defined as j = m_k^ * is a case where the quenched approximation produces an answer different from the real world@xcite .
figure [ j_fig ] shows this ratio on the matched quenched and full qcd milc lattices , with a clear increase in the value when the sea quarks are included .
this result can be compared with two flavor wilson quark simulations by the cppacs , ukqcd and jlqcd collaborations@xcite , which also show a larger value in two flavor qcd .
the nucleon to rho mass ratio is a much studied quantity where lattice simulations typically disagree with the real world number .
many groups have looked at this , and it has become clear that this quantity is especially sensitive to effects of the lattice spacing .
thus it is interesting to use the matched lattices to look for effects of sea quarks .
figure [ ape_both_fig ] shows this ratio in the milc calculations , with no discernable difference between the quenched and full qcd curves .
this is in contrast to ukqcd calculations on matched lattices with two flavors of wilson quarks , where the two flavor numbers are larger than the quenched ( see fig .
9 of ref .
@xcite ) . the two plusses in fig .
[ ape_both_fig ] are quenched points at @xmath14 fm .
preliminary three - flavor results at @xmath21 and @xmath22 fm show a similar trend .
although this lattice spacing dependence is much less than seen with the conventional kogut - susskind action , it is clear that a careful continuum extrapolation will still be required for this quantity . ' '' '' |
in standard cosmology , the universe is thought to have experienced big - bang nucleosynthesis ( bbn ) at a very early stage .
the light nuclides d , t+@xmath3he , @xmath0he and @xmath7li+@xmath7be are produced in the standard bbn ( sbbn ) at observable levels , while this model does not make appreciable quantities of @xmath6li .
the wilkinson microwave anisotropy probe ( wmap ) satellite has measured the temperature fluctuations of the cosmic microwave background ( cmb ) radiation , and parameters characterizing the standard big bang cosmology have been deduced @xcite from these data . for the baryon - to - photon ratio @xmath12 deduced from fits to the cmb , the bbn model predicts abundances of the light elements except for @xmath6li and @xmath7li which are more - or - less consistent with those inferred from astronomical observations .
spectroscopic lithium abundances have been detected in the atmospheres of metal poor stars .
nearly constant abundances of @xmath6li and @xmath7li in metal - poor population ii ( pop ii ) stars have been inferred .
there is about a factor of three under - abundance of @xmath7li in metal - poor halo stars ( mphss ) with respect to the sbbn prediction when using the baryon - to - photon ratio @xmath12 .
this is called the @xmath7li problem @xcite . in addition ,
spectroscopic measurements obtained with high resolution indicate that mphss have a very large abundance of @xmath6li , i.e. at a level of about 3 orders of magnitude larger than the sbbn prediction of the @xmath6li abundance , which is called the @xmath6li problem @xcite .
cayrel et al .
@xcite studied line asymmetries to be generated by convective doppler shifts in stellar atmospheres , and found that the convective asymmetry might mimic the presence of @xmath6li and an error of @xmath6li/@xmath7li amounts to a few percent that is roughly comparable to the values estimated from mphss .
the @xmath6li problem , therefore , may not exist in fact , since the convective asymmetry could give a possible solution to the @xmath6li problem within the framework of sbbn .
the possibility of @xmath6li production in non - standard bbn triggered by the decay of unstable relic neutral massive particles @xmath13 has been studied @xcite .
several critical constraints on the properties of @xmath13 particles were derived from the studies of radiative decay @xcite , hadronic decay or annihilation @xcite of @xmath13 particles along with the bbn constraints on the light elements .
these particle decay induces electromagnetic and/or hadronic showers triggering the destruction of preexisting nuclei and the production of different nuclear species .
a recent detailed study @xcite of the radiative decay and its influence on the @xmath6li production has found a parameter region of lifetime @xmath14 s and abundance parameter @xmath15 gev where the non - thermal nucleosynthesis of @xmath6li can explain the observed abundance level in mphss .
this parameter region satisfies the two observational constraints on the cmb energy spectrum and the primordial light element abundances .
three important characteristics were found for the interesting parameter region .
first , @xmath3he and @xmath16 are the seeds for @xmath6li in the processes @xmath0he(@xmath3he,@xmath2)@xmath6li and @xmath0he(@xmath16,@xmath4)@xmath6li .
second , the excess of @xmath6li abundance is therefore regulated by the amounts of @xmath3he and @xmath16 which are produced by the non - thermal photodisintegration of @xmath0he , i.e. @xmath0he(@xmath1,@xmath2)@xmath3h and @xmath0he(@xmath1,@xmath4)@xmath3he . hence , the radiative decay model which results in @xmath6li - production above the mphs abundance level is also reflected by an enhancement of the @xmath3he abundance with respect to the sbbn value .
third , the radiative decay does not resolve the @xmath7li problem @xcite .
it is therefore concluded that other mechanisms such as the stellar depletion of the lithium isotopes in the atmosphere of mphss @xcite or new burst of late - time bbn on the exotic @xmath13-bound nuclei in the case of negatively - charged leptonic particles @xmath17 @xcite must operate to lower the @xmath7li abundance .
a recent measurement of @xmath0he photodisintegration reactions , @xmath0he(@xmath1,@xmath2)@xmath3h and @xmath0he(@xmath1,@xmath4)@xmath3he with laser - compton photons @xcite shows much smaller cross sections than those estimated from the other previous experiments @xcite and those summarized in ref .
@xcite at the photon energies 20 mev @xmath18 30 mev .
if these non - thermal photon energies dominate the destruction of @xmath0he , the production of @xmath3he and @xmath16 and also the subsequent production of @xmath6li via @xmath0he(@xmath3he,@xmath2)@xmath6li and @xmath0he(@xmath16,@xmath4)@xmath6li , this would change the parameter region of @xmath19 and @xmath20 of massive relic particles @xmath13 so that the resultant non - thermal nucleosynthesis of @xmath6li can explain the abundance level observed in mphss .
the first purpose of this article is to study the sensitivity of non - thermal bbn of all light elements d , t , @xmath3he , @xmath0he , @xmath6li , @xmath7li and @xmath7be to the photodisintegration cross sections of @xmath0he .
the second purpose is to infer the uncertainties of the two parameters @xmath19 and @xmath20 of massive relic particles @xmath13 which would arise from the uncertainties of the measured reaction cross sections . in sec .
[ sec2 ] we present a result of a new measurement of @xmath0he photodisintegration cross sections . in sec .
[ sec3 ] we briefly explain the model of non - thermal nucleosynthesis and the calculated result of the effect of the considered change of the photodisintegration cross sections . in sec .
[ sec4 ] we summarize our conclusion and offer an outlook for measurements of @xmath0he photodisintegration .
so far the cross section data of the photodisintegration of @xmath21he have been obtained from the direct photodisintegration experiments as well as the inverse radiative capture experiments .
indirect probes such as the ( @xmath2,@xmath22 ) reaction @xcite and the ( @xmath23li,@xmath23be ) reaction @xcite have also been applied to investigate the property of the dipole excitations of @xmath21he .
+ the direct experiments have been performed by detecting either charged fragments ( @xmath2 , @xmath24h , @xmath24he ) or neutrons from the photodisintegration reactions . as incident real photon beams , either continuous bremsstrahlung photons or quasi - monochromatic ones generated with various methods such as the photon tagging , the positron annihilation in flight , and the laser compton - backscattering were used .
since the energies of the emitted particles are small due to high threshold energies of the photodisintegrations of @xmath21he , many efforts have been devoted to detect those particles clearly from the backgrounds caused by the incident high - energy @xmath1-rays . in the inverse experiments ,
the reaction cross sections are much smaller than the photodisintegrations because of the difference of the phase space factors , and therefore the measurements have been performed with great care to the influence of the background @xmath1-rays as well as the determinations of the experimental parameters such as the detector efficiency for high - energy capture @xmath1-rays , the effective target thicknesses , the incident beam intensity , and so on .
+ in spite of the above experimental efforts , there have been large discrepancies in the previous data as shown in fig .
[ fig1 ] . especially in the energy region below @xmath25 30 mev , the data show either a pronounced peak at around 25@xmath2626 mev or a rather smooth curve as a function of the excitation energy . in order to determine the cross sections more accurately , shima et al .
recently performed measurements using quasi - monochromatic laser - compton photons and a time projection chamber ( tpc ) containing a gas mixture of he and cd@xmath27 @xcite .
the method had the following features : + @xmath28 thanks to the quasi - monochromatic and well - collimated @xmath1-ray beam , the background due to low - energy @xmath1-rays is very small .
+ @xmath28 since the tpc gas serves as an active target , low - energy charged fragments from the photodisintegrations can be detected simultaneously for @xmath0he(@xmath1,@xmath2 ) and @xmath0he(@xmath1,@xmath4 ) with efficiencies of nearly 100% and a solid angle of 4@xmath29 .
+ @xmath28 the absolute sensitivity of the measurement can be accurately checked with the d(@xmath1,@xmath4)@xmath2 reaction , whose cross section is well known in the energy region of the present interest . + previously shima et al . measured the cross sections in the @xmath1-ray energy region from 20 to 30 mev using the laser - compton @xmath1-ray source at the national institute of advanced industrial science and technology ( aist ) @xcite .
he(@xmath1,@xmath2)@xmath24h ( upper panel : a ) and @xmath21he(@xmath1,@xmath4)@xmath24he ( lower panel : b ) reactions .
the open circles stand for the published data @xcite from experiments performed using quasi - monochromatic laser - compton photon beams at aist .
the dotted circles are the latest data of the ( @xmath1,@xmath4 ) reaction measured by means of a tagged photon beam @xcite .
the other symbols indicate the previous data ( see the references in ref .
the error bars show @xmath30 uncertainties in the cross section data .
the solid curves are the most probable excitation functions determined from the experimental data at @xmath31 30 mev @xcite and the previous ones at higher energies 30 mev @xmath32 116 mev @xcite .
the dash - dotted curves are the fitting functions to some of the available data summarized in ref .
@xcite.[fig1],width=302 ] as shown in fig .
[ fig1 ] , the cross sections from the experiments are found to monotonically increase as a function of the @xmath1-ray energy up to @xmath25 30 mev , being quite different from the standard fitting functions previously evaluated by cyburt et al . @xcite . with the above mentioned situation in mind ,
the sensitivity of the bbn to the photodisintegration cross sections of @xmath21he was studied in the present paper by using the recent experimental results as well as the previously adopted standard fitting functions in ref .
@xcite .
the present excitation functions were determined as follows .
since no sharp resonance has been observed in the excitation functions of the photodisintegration of @xmath21he in the energy region up to @xmath25 100 mev , we assumed the excitation functions can be approximated by a fitting function in ref .
@xcite , i.e. , @xmath33 three parameters , @xmath34 , @xmath35 and @xmath36 are determined by fitting this functions to the data measured at aist @xcite and the data measured at high energies of @xmath11 30 mev @xcite by means of the @xmath37minimization method .
this function is suitable for use in numerical calculation because it has no discontinuity and no divergence at high energies . in this study
we take published values for errors of cross sections at all fitting procedures .
the most probable fitting functions for the cross sections of the @xmath0he(@xmath1,@xmath2)@xmath3h and @xmath0he(@xmath1,@xmath4)@xmath3he reactions turn out to be @xmath38 and @xmath39 respectively , and they are plotted in fig .
reaction @xmath40-values are taken from experiments as @xmath41 mev and @xmath42 mev , respectively .
we take these two cross sections as our recommended cross sections in this paper .
the dash - dotted curves are standard expressions from ref .
parameter values are given as @xmath43 mb , @xmath44 and @xmath45 for @xmath0he(@xmath1,@xmath2)@xmath3h reaction , and @xmath46 mb , @xmath44 and @xmath45 for @xmath0he(@xmath1,@xmath4)@xmath3he reaction .
as another implementation , we fit all data adopted in this study including the data measured at aist @xcite at low energies of @xmath47 mev and previous ones obtained in other experiment .
the fitting parameters turn out to be @xmath48 mb , @xmath49 , and @xmath50 for @xmath0he(@xmath1,@xmath2)@xmath3h reaction , and @xmath51 mb , @xmath52 , and @xmath53 for @xmath0he(@xmath1,@xmath4)@xmath3he reaction .
the derived cross sections are displayed in fig .
[ fig2 ] . .
the solid curves and the short - dashed curves are the most probable excitation functions determined from all data adopted in this study including the experimental data at @xmath5 mev @xcite and the previous ones at whole energy range of @xmath54 116 mev @xcite .
the dash - dotted curves are the same as in fig .
[ fig1].[fig2],width=302 ] one can see in fig .
[ fig1 ] and fig .
[ fig2 ] that there is a large dispersion in experimental data at low photon energies of 20 mev @xmath55 mev .
this dispersion is much larger than @xmath56 deviation of any single fitting function .
we therefore adopt the above three functions in bbn calculations and compare the calculated results with one another in order to study the sensitivity of bbn with the radiative particle decay to the @xmath0he photodisintegration reaction cross sections .
we assume the creation of high energy photons from the radiative decay of a massive particle @xmath13 with a mass @xmath57 and a mean life of @xmath58 s. we represent the emitted photon energy by @xmath59 .
we assume that the decaying dark particle is non - relativistic , and almost at rest in the expanding universe . a decay - produced energetic photon interacts with the cosmic background and induces an electromagnetic cascade shower .
the faster processes , pair production through background photons @xmath60 ( @xmath61 ) and inverse compton scattering of produced electrons and positrons through background photons ( @xmath62 ) , produce electromagnetic showers and the non - thermal photon spectrum realizes a quasi - static equilibrium @xcite .
the attained zeroth generation photon spectrum is @xcite , @xmath63 k_0(e_x / e_{\gamma})^{2.0 } & \mbox{for $ e_x\leq e_{\gamma}<e_c$ } , \\ % [ 0.2 in ] 0 & \mbox{for $ e_c\leq e_{\gamma}$ } \\
% [ 0.2 in ] \end{array } \right .
\label{spectrum}\ ] ] where @xmath64\right)$ ] is a normalization constant fixed by energy conservation of the injected photon energy .
this spectrum has a break in the power law at @xmath65 and an upper cutoff at @xmath66 .
we take the same energy scaling with the temperature @xmath67 of the background photons as in @xcite , i.e. @xmath68 and @xmath69 at the temperature @xmath67 , where the cascade spectrum was calculated by numerically solving a set of boltzmann equations . here
@xmath70 is the rest mass of an electron .
because the rates of electromagnetic interactions are faster than the cosmic expansion rate , the photon spectrum @xmath71 is modified into a new quasi - static equilibrium ( qse ) .
this distribution is given by @xmath72 where @xmath73 is the number density of the decaying particles at a redshift @xmath74 .
the quantity @xmath75 is the energy degradation rate through three slower processes ; compton scattering ( @xmath76 ) , bethe - heitler ordinary pair creation in nuclei ( @xmath77 ) , and double photon scattering ( @xmath78 ) for the zeroth - generation photons .
we use this steady state approximation for the cosmic non - thermal constituent of photons .
the equation for the production and destruction of nuclei by non - thermal photons is given by @xmath79_a y_t - \sum_p \left[a\gamma\right]_p y_a , \label{dydt2}\ ] ] where we have defined the reaction rate @xmath80_a & \hspace{-5pt } = & \hspace{-5pt } \frac{n_\gamma^0 \zeta_x}{\tau_x } \left(\frac{1}{2h_r t}\right)^{3/2 } \exp(-t/\tau_x ) \nonumber \\ & & \hspace{-5pt } \times \int_{0}^\infty de_\gamma s_\gamma^{\rm qse}(e_\gamma)\ , \sigma_{\gamma + t \rightarrow a } ( e_\gamma ) , \nonumber \\ \label{agamma}\end{aligned}\ ] ] and @xmath81 @xmath82 is the mole fraction of a particular nuclear species @xmath83 , and @xmath84 and @xmath85 are number densities of nuclei @xmath83 and baryons .
the first and second term on the right - hand side are the source ( @xmath86 ) and sink ( @xmath87 ) terms for nucleus @xmath88 .
@xmath89 is a non - thermal photon energy .
the cross section of the process @xmath90 is denoted by @xmath91 .
further we use @xmath92 to represent the number of identical species of nuclei in a production or destruction process ; @xmath93 when particles @xmath88 and @xmath94 are identical and @xmath95 when they are not .
for example , in the process @xmath0he(@xmath1,@xmath96)d ,
we defined @xmath98{\mathstrut 8\pi g \rho_{\rm rad}^0/3}$ ] , where the superscript 0 denotes present values ( @xmath99 ) , therefore @xmath100 and @xmath101 are the present photon number density and present radiation energy density of the cosmic background radiation ( cbr ) , respectively .
we defined @xmath102 .
the equation describing the secondary production and destruction is obtained by taking account of the energy loss of nuclear species while propagating through the background . in general , because of the high energy loss rate , the primary particles establish a quasi - static equilibrium .
the abundance evolution is then represented by @xmath103_s -({\rm sink~term } ) , \label{dysdt}\ ] ] where @xmath104_s$ ] is the reaction rate for a secondary process @xmath105 with any combination of particles @xmath106 , @xmath88 , and @xmath107 .
for example , @xmath108he@xmath109@xmath110 for a secondary process @xmath0he(@xmath1,n)@xmath3he(@xmath111,@xmath2)@xmath6li is given by @xmath112_{^6{\rm li } } \hspace{-5pt } & = & \hspace{-5pt } \frac{\eta ( n_\gamma^0)^2 \zeta_x}{\tau_x } \left(\frac{1}{2h_r t}\right)^3 \exp(-t/\tau_x ) \nonumber \\ & & \hspace{-5pt } \times \int_{e_{p,\rm th}}^{{\mathcal e}_{^3{\rm he}}(e_c)}\!\!\!\!\ ! de_{^3{\rm he } } \frac{\sigma_{^3{\rm he}(\alpha , p)^6{\rm li}}(e_{^3{\rm he}})\beta_{^3{\rm he}}}{b_{^3{\rm he}}(e_{^3{\rm he } } ) } \nonumber \\ & & \hspace{-5pt } \times \int_{{\mathcal e}_{^3{\rm he}}^{-1}(e_{^3{\rm he}})}^{e_c } \!\!\!\!\ !
de_\gamma s_\gamma^{\rm qse}(e_\gamma)\sigma_{^3{\rm he}(\alpha , p)^6{\rm li } } ( e_\gamma ) , \nonumber \\ \label{tt'}\end{aligned}\ ] ] where @xmath113 is the baryon - to photon ratio : @xmath114 , and @xmath115 is the velocity of the primary particle @xmath88 .
@xmath116 is the energy loss rate of the primary particle by coulomb scattering .
@xmath117 is the energy of the nuclide @xmath88 produced by the photodisintegration process @xmath118 , and @xmath119 is the energy of the non - thermal photons which produce the primary species @xmath88 with energy @xmath120 .
@xmath117 and @xmath119 are derivable in the limit of low energy scattering , where the relevant nuclei are non - relativistic , i.e. @xmath121 and @xmath122 , where @xmath123 is the threshold energy of primary photodisintegration reaction .
we focus on the non - thermal production of mass 3 nuclides , @xmath3h and @xmath3he , and secondary production of @xmath6li . in fig .
[ fig5 ] , photodisintegration cross sections of @xmath0he(@xmath1,@xmath4)@xmath3he and @xmath0he(@xmath1,@xmath2)@xmath3h are shown as a function of non - thermal photon energy @xmath89 ( above threshold energy @xmath124 mev ) .
these cross sections are fitted with data from the laser - compton photon experiment ( solid curves ) and those from ref .
@xcite ( dotted curves ) .
dashed line is the energy spectrum of non - thermal photon produced by the radiative decay when the cosmic temperature is 10 ev corresponding to the decay life @xmath125 s @xcite .
it is normalized with respect to intensity at @xmath126 mev .
since the yield of non - thermally produced mass 3 nuclides is proportional to the integration of the non - thermal photon spectrum times photodisintegration cross sections , the change of photodisintegration cross sections at @xmath5 mev has relatively large influence on the resulting @xmath3he abundance .
he(@xmath1,@xmath4)@xmath3he and @xmath0he(@xmath1,@xmath2)@xmath3h with data from the laser - compton photon experiment at low energies of @xmath5 mev are shown by solid lines above @xmath127 mev .
the standard cross sections @xcite are also shown as dotted lines .
cross sections of secondary @xmath6li production reactions @xmath3h(@xmath111,@xmath4)@xmath6li and @xmath3he(@xmath111,@xmath2)@xmath6li are superimposed .
the energy spectrum of non - thermal photon produced by the radiative decay when the cosmic temperature is 10 ev corresponding to the decay life @xmath128 s is also shown ( dashed line ) , which is normalized with respect to intensity at @xmath126 mev.[fig5],width=302 ] on the other hand , non - thermal secondary production of @xmath6li is not affected by @xmath0he photodisintegration cross sections at low energies .
when we consider the non - relativistic limit of nuclear reactions , the center of mass energy @xmath129 in @xmath3he(@xmath111,@xmath2)@xmath6li and @xmath3h(@xmath111,@xmath4)@xmath6li reactions is given by @xmath130 where @xmath131 , @xmath132 and @xmath133 are the reduced mass , the mass of a nucleus @xmath134 ( @xmath16 or @xmath3he ) and that of @xmath0he , @xmath135 and @xmath136 are the relative velocity and the velocity of @xmath134 , @xmath137 is the kinetic energy of @xmath134 , respectively .
the energy of @xmath134 generated in primary @xmath0he(@xmath1,@xmath4)@xmath3he and @xmath0he(@xmath1,@xmath2)@xmath3h reactions , i.e. , @xmath138 ( @xmath139 ) is given by @xmath140 where @xmath141 is the mass of proton and @xmath123 is the threshold energy of @xmath0he photodisintegration reaction .
the center of mass energy @xmath129 in the secondary reactions thus relates to the non - thermal photon energy in primary reactions as @xmath142 the inequality means that non - thermal @xmath3he and @xmath3h nuclides lose energies from a time of their production to a time of nuclear reactions to produce @xmath6li .
we plot the cross sections of @xmath3he(@xmath111,@xmath2)@xmath6li and @xmath3h(@xmath111,@xmath4)@xmath6li reactions as a function of @xmath89 neglecting the energy losses of @xmath3he and @xmath3h in fig .
since secondary @xmath6li production needs photon energy of at least @xmath8 mev to produce energetic @xmath3he and @xmath3h , the yield of @xmath6li is insensitive to the change of the @xmath0he photodisintegration cross section at low energy of @xmath5 mev .
( see sec .
[ sec3]b1 . ) on the other hand , a change of the cross section at high energies of @xmath11 50 mev causes a change of resulting yield of @xmath6li ( see sec .
[ sec3]b2 for estimation of uncertainty on the @xmath6li yield associated with a possible uncertainties in cross sections at high energies . )
figure [ fig3 ] shows contours corresponding to the constraints for the primordial abundance @xcite for the present result with our recommended smaller cross sections of @xmath0he photodisintegration eqs .
( [ cross11 ] ) and ( [ cross12 ] ) ( thick lines ) .
for example , a contour of the @xmath0he mass fraction @xmath143 is shown in the ( @xmath19,@xmath20 ) plane .
the region above this contour should be excluded for @xmath144 .
the contours of @xmath7li / h lower limit , @xmath7li / h@xmath145 , d / h upper and lower limits , d / h@xmath146 and d / h@xmath147 ( dashed line ) , respectively , and @xmath3he / h upper limit , @xmath3he / h@xmath148 are also drawn .
dotted lines show contours of the same constraints for the result with larger standard cross sections of @xmath0he photodisintegration @xcite .
the thick solid line shows a constraint from the consistency requirement of the cmb with a blackbody @xcite .
the region above the line is excluded .
the contour for the mphss level of @xmath6li / h=@xmath149 is plotted .
the gray region above the contour and below the nucleosynthesis plus cmb constraints is allowed and abundant in @xmath6li .
plane corresponding to the adopted constraints for the primordial abundances in the present calculation with smaller cross sections of @xmath0he photodisintegration eqs .
( [ cross11 ] ) and ( [ cross12 ] ) ( thick lines ) .
contours for the mass fraction of @xmath0he @xmath150 and the number ratios of @xmath3he / h=@xmath151 , d / h=@xmath152 , d / h=@xmath153 ( dashed line ) , and @xmath7li / h=@xmath154 are shown . the notation `` over '' and `` low '' identifies overproduced and underproduced regions , respectively . the same constraints on the result calculated with larger standard cross sections of @xmath0he photodisintegration @xcite is shown as dotted lines .
the region above the thick solid line is excluded by the consistency requirement of the cmb with a blackbody .
the contour of @xmath6li / h=@xmath149 is also drawn .
the gray region above the contour and below the nucleosynthesis plus cmb constraints is the allowed region where the abundant @xmath6li is produced.[fig3],width=302 ] for this figure , we use the two sigma upper limit on the observed @xmath3he / h abundance ratio from galactic hii region @xcite as a conservative constraint on primordial abundance . alternatively a constraint from @xmath3he / d ratio is often used . since deuterium is fragile and
more easily burned in stars than @xmath3he is , the @xmath3he / d ratio is thought to increase monotonically as a function of time by stellar processing from the formation of first stars to the solar system formation . the @xmath3he / d ratio for primordial abundances can , therefore , be constrained by the solar @xmath3he / d ratio . since the solar @xmath3he / d ratio is ( @xmath3he / d)@xmath155=0.82 @xcite
, one can obtain a constraint of @xmath3he / d @xmath156 for primordial abundances .
although this constraint is slightly stronger than the constraint for the @xmath3he / h ratio , the contours for both constraints are not distinguishable from each other in fig .
[ fig3 ] .
one can see that @xmath7li abundance does not change drastically by the new @xmath0he photodisintegration cross sections .
the contour for the @xmath0he over - destruction by the non - thermal photons shifts upward by @xmath25 300 % at most for @xmath157 s with respect to that with the larger cross sections ( dotted line ) .
since a destruction rate of @xmath0he is lowered at low energies when its cross sections are taken to be small , larger energy is necessary to destruct @xmath0he considerably .
the @xmath0he destruction leads to non - thermal production of @xmath3he and d. then , d over - destruction regions , which are above the region of this figure , shift upward reflecting the change of relic abundance of seed nuclide @xmath0he .
the @xmath3he overproduction region also shifts upward by @xmath25 30 % .
since the cross sections of @xmath0he(@xmath1,@xmath2)@xmath3h and @xmath0he(@xmath1,@xmath4)@xmath3he are lowered , the amount of produced mass 3 nuclides are lowered .
one finds an upward shift of d overproduction region .
we confirmed that this change is caused since d is produced by @xmath2(@xmath4,@xmath1)d using non - thermally produced neutrons by @xmath0he(@xmath1,@xmath4)@xmath3he , whose cross section changed .
figure [ fig3 ] shows that the contour of @xmath6li abundance also changes slightly by @xmath158 % at most for @xmath159 s , but this change does not result from the change of photodisintegration cross sections of @xmath0he at low energy @xmath31 30 mev .
it is due to the difference of excitation functions between our recommended fit and the standard fit @xcite at higher energy , @xmath11 30 mev .
( see fig .
[ fig5 ] . )
the reason has already been explained at the beginning of this section sec .
[ sec3]b .
we pick up the most interesting parameter region , where @xmath6li is produced at the level higher than that of mphss .
we take a look at the result of nucleosynthesis of the parameters @xmath160=(10@xmath161 s , @xmath162 gev ) @xcite for this reason . in fig .
[ fig4 ] , the abundance of @xmath13 particle @xmath163 is shown as a function of the cosmic photon temperature . it is normalized as @xmath163/@xmath164 , where @xmath164 is the initial abundance before the @xmath13-decay decreases its abundance . as the indicator of energy amount injected by the @xmath13 decay at an epoch , we also plot @xmath165 .
most of the energy content is injected in the time scale of @xmath19 .
these lines are for only the case of @xmath166 s. particles @xmath163/@xmath164 , where @xmath164 is the initial abundance before @xmath13-decays . also shown
is @xmath167 , which characterizes the amount of energy from the particle decay at time @xmath16 .
the decay life is assumed to be @xmath166 s.[fig4],width=302 ] figure [ fig6 ] shows the @xmath3he and @xmath3h abundances as a function of temperature @xmath67 [ @xmath168 when parameters are @xmath169=(10@xmath161 s , @xmath162 gev ) ( upper panel ) .
the results with our recommended and the standard cross sections of @xmath0he photodisintegration correspond to the solid and dashed lines , respectively .
an apparent reduction of non - thermally produced @xmath3he abundance is found in our recommended case .
the resulting difference of non - thermal yield of @xmath3he is about 10 % . he and @xmath3h abundances when parameters are @xmath169=(10@xmath161 s , @xmath162 gev ) .
the solid and dashed lines are the results for our recommended and the standard cross sections of @xmath0he photodisintegration .
( lower panel ) ratios of @xmath3he and @xmath3h abundances calculated with our recommended cross sections to those with the standard cross sections.[fig6],width=302 ] the measured @xmath0he photodisintegration cross sections are still subject to large uncertainties which depend on the different experimental setup and methods over the wide energy range . the energy dependence of the cross sections is not well understood theoretically .
if either cross section of @xmath0he(@xmath1,@xmath2)@xmath3h or @xmath0he(@xmath1,@xmath4)@xmath3he in the energy region of @xmath8 mev is different from the adopted cross section , non - thermal bbn abundance of @xmath6li would change significantly .
it should be noted here that the data below 30 mev of shima et al .
@xcite are significantly smaller than previous ones , and therefore one may expect enhancement of the cross sections in the higher energy range from the well - known thomas - reiche - kuhn ( trk ) sum rule , which relates the energy - integrated cross section @xmath170 to the ground state property of a target nucleus .
the e1 sum rule is expressed by @xmath171 where @xmath172 with @xmath111 , @xmath173 , @xmath174 , @xmath175 and @xmath88 the fine structure constant , the nucleon mass , the neutron number , the proton number and the nucleon number , respectively .
@xmath176 is the so - called trk enhancement factor defined as @xmath177 ] |0\rangle , \label{kappa}\ ] ] where @xmath178 is the nuclear ground state wave function , @xmath179 is the dipole operator @xmath180 with @xmath181 and @xmath182 the third component of the isospin operator and the @xmath74-coordinate of the @xmath83th particle in the center - of - mass frame , respectively , @xmath183 is the nuclear potential .
there is another sum rule , i.e. bremsstrahlung sum rule , which is expressed by @xmath184 where @xmath185 fm @xcite and @xmath186 fm @xcite are the root mean square charge radii for proton and @xmath0he , respectively .
energy - weighted integrals @xmath187 with @xmath188 for eq .
( [ e1sum ] ) and @xmath189 for eq .
( [ bremsum ] ) are listed in table [ tab1 ] for the four different models of photodisintegration cross sections of @xmath0he(@xmath1,@xmath2)@xmath3h and @xmath0he(@xmath1,@xmath4)@xmath3he .
the three models of our recommended fit , i.e. , eqs .
( [ cross11 ] ) and ( [ cross12 ] ) , the fit with all data adopted in this study , and the standard fit @xcite lead to values smaller than the sum rules , i.e. @xmath190 mev mb with @xmath191 for the trk sum rule and @xmath192 mb for the bremsstrahlung sum rule .
we expect that the model using eqs .
( [ cross11 ] ) and ( [ cross12 ] ) is closer to the lower limit of true cross sections in consideration of two sum rules . [
cols="^,^,^,^,^",options="header " , ] in addition to the three models in the second , third and forth columns in table [ tab1 ] above mentioned , we make another new fitting function of the cross sections so that the energy weighted integrals satisfy the two sum rules : as for low energy cross sections at @xmath31 30 mev , we fit the experimental data @xcite measured by using quasi - monochromatic laser - compton photon beams at aist . at higher energy @xmath11 30 mev
, we assume a simple energy dependence @xmath193 .
resultant parameters are @xmath194 mb and @xmath195 . in this new model where we require the sum rules ,
the constructed cross sections are typically @xmath196 times larger than our recommended ones at 50 mev @xmath197 135 mev , where the upper limit is the meson mass .
the both trk sum rule and bremsstrahlung sum rule do not apply to such high energies @xmath11 135 mev because new degrees of freedom of mesons as well as nucleons play an important role and the above two sum rules break down .
we can expect that the constructed cross sections which satisfy the sum rules are close to the upper limits to the realistic cross sections .
we compare the primordial abundance calculated with new cross sections which satisfy the sum rules and that of our recommended cross sections . in fig .
[ fig7 ] thick solid and dashed lines correspond to the results with our recommended and new cross sections of @xmath0he photodisintegration , respectively .
the dark and light gray regions above the solid and dashed contour lines and below the nucleosynthesis plus cmb constraints are the allowed region where the abundant @xmath6li is produced . the @xmath6li production occurs more efficiently in the case with larger cross sections than with smaller ones .
this is because the @xmath0he destruction is more effective for larger cross sections at higher energies so that the more abundant energetic mass 3 nuclides which synthesize @xmath6li are produced .
for this reason the @xmath6li abundance produced by using the large @xmath0he - photodisintegration cross sections which satisfy the sum rules are presumed to be a maximum yield from the viewpoint of nuclear structure physics .
the allowed region of the properties of relic @xmath13 particle in the @xmath169 plane , therefore , should not move even below the light gray region in fig [ fig7 ] , and we expect that the reality is located between the dark and light gray regions . plane corresponding to the adopted constraints for the primordial abundances in the calculation with our recommended cross sections of @xmath0he photodisintegration ( thick lines ) and those satisfying sum rules ( dashed lines ) .
the adopted abundance constraints on light elements are the same as those in fig [ fig3 ] .
the dark ( for our recommended cross sections ) and light ( for larger ones ) gray regions above the solid and dashed contour lines , respectively , and below the nucleosynthesis plus cmb constraints are the allowed region where abundant @xmath6li is produced.[fig7],width=302 ] however , it has been known that @xmath170 highly depends on various effects of residual interactions among nucleons such as the meson - exchange currents @xcite , the tensor correlations @xcite , the short - range interactions @xcite , and so on . in the case of @xmath0he ,
the calculated values of @xmath170 , which were integrated out up to @xmath198 mev , have uncertainty of more than @xmath199 % , depending on the nucleon - nucleon potentials and the nuclear models for the ground state of @xmath0he @xcite . in summary , both the existing experimental data and the theoretical calculations for the @xmath0he photodisintegration cross section in the energy range up to @xmath200 mev contain large uncertainties , and therefore a precise measurement of the @xmath0he photodisintegration cross section at @xmath201 mev is highly desirable .
comprehensive theoretical study of the nuclear structure and reactions of @xmath0he are also necessary in order to clarify many unresolved nuclear effects and also refine the applicability of an empirical formula such as eq.([cross21 ] ) .
these nuclear physics studies would be important to constrain the life time and the abundance of long - lived relic @xmath13 particle more precisely .
a recent measurement of @xmath0he photodisintegration reactions , @xmath0he(@xmath1,@xmath2)@xmath3h and @xmath0he(@xmath1,@xmath4)@xmath3he with laser - compton photons shows lower cross sections at low energies than those estimated by other previous experiments .
we studied the sensitivity of non - thermal bbn of all light elements d , t , @xmath3he , @xmath0he , @xmath6li , @xmath7li and @xmath7be to the photodisintegration cross section of @xmath0he .
the change of cross sections of @xmath0he photodisintegration has an influence on the non - thermal yields of light elements , d , @xmath3he and @xmath0he , which are related to the photodisintegration cross sections at low energy ( @xmath202 mev ) .
the upper limit of allowed regions of @xmath13-abundance parameter @xmath20 for these light nuclei shifts upward by @xmath203 % for @xmath159 s for this change of the cross sections .
this arises from the upshift of @xmath3he abundance contour ( see fig . [ fig3 ] ) . on the other hand ,
the non - thermal @xmath6li production is not very sensitive to the change of cross sections at low energy , since the non - thermal secondary synthesis of @xmath6li needs energetic photons of @xmath8 mev .
the non - thermal nucleosynthesis triggered by the radiative particle decay is one of candidates of the production mechanism of @xmath6li observed in mphss . in the interesting parameter region of @xmath204
s@xmath205 s and @xmath206 gev @xmath207 gev which satisfies the @xmath6li production above the abundance level observed in mphss , the lowering of the photodisintegration cross sections at low energy @xmath5 mev as measured in the recent experiment using laser - compton photons leads to @xmath208 % reduction of resulting @xmath3he abundance , whereas the @xmath6li abundance does not change for the change of the cross sections of @xmath0he(@xmath1,@xmath2)@xmath3h and @xmath0he(@xmath1,@xmath4)@xmath3he .
let us briefly discuss other impacts of such a precise cross section measurement .
clarifying the effects of photodisintegrations of @xmath0he will affect more strongly the @xmath209-process in core - collapse supernova ( sn ) explosions through the neutrino - nucleus interactions specifically of @xmath209+@xmath0he .
the weak transition rates for @xmath0he(@xmath210 ) , @xmath0he(@xmath211,@xmath212 ) , and @xmath0he(@xmath213,@xmath214 ) are determined similarly to the giant electric dipole resonance observed in the photodisintegrations with the help of theoretical calculation @xcite .
in fact , several experiments of measuring the @xmath0he photodisintegration cross sections @xcite were carried out for this purpose .
the precise knowledge of the @xmath0he(@xmath209,@xmath209@xmath2 ) , ( @xmath209,@xmath209@xmath4 ) , ( @xmath211,@xmath215 ) , and ( @xmath213,@xmath216 ) cross sections is required to determine the unknown parameters for neutrino oscillations through the msw effect on the @xmath7li and @xmath217b production triggered by the @xmath209+@xmath0he reactions @xcite .
the energy range @xmath218 mev is very important for the @xmath209-process nucleosynthesis in sne .
the mean neutrino energy of sn neutrinos is presumed to be about 10 - 25 mev in numerical simulations of the neutrino transfer in core - collapse sne , and the threshold energies for all neutrino - induced spallation reactions of @xmath0he are @xmath2520 mev .
therefore , the difference between the newly measured @xcite and previous @xmath0he photodisintegration cross sections at 20 mev @xmath18 30 mev could be critical . as a result
, the absolute yields of @xmath7li and @xmath217b produced in the @xmath209-process in core - collapse sne would be different from one another , depending on the assumed @xmath209-process reaction rates as demonstrated theoretically @xcite although the ratio of @xmath7li/@xmath217b does not change largely .
another recent focus of photodisintegration of @xmath0he is on the mechanism of the core - collapse sne .
most sn simulations still do not succeed in the sn explosion in spite of detailed numerical studies of the neutrino transfer calculations inside the core .
haxton @xcite proposed that the neutrino - induced excitations of @xmath0he and heavier nuclei could deposit extra - energy to the ejected materials and revive the shock wave , which motivated a recent theoretical study on the role of @xmath0he spallation reactions in the core - collapse sne @xcite .
his theoretical suggestion also motivated recent experimental studies of photodisintegrations of @xmath0he @xcite in order to estimate the neutrino - induced reaction cross sections for @xmath0he(@xmath210 ) , @xmath0he(@xmath211,@xmath212 ) , and @xmath0he(@xmath213,@xmath214 ) .
as such , it is important and even critical to study the @xmath0he(@xmath1,@xmath2 ) and @xmath0he(@xmath1,@xmath4 ) reactions precisely for the discussions of the problem of sn - neutrino oscillation and sn explosion as well as the cosmological discussion concerning the bbn with a radiative decay of long - lived relic particles . this work has been supported in part by the mitsubishi foundation , the grant - in - aid for scientific research ( 17540275 , 20244035 ) of the ministry of education , science , sports and culture of japan , and the jsps core - to - core program , international research network for exotic femto systems ( efes ) .
mk acknowledges the support by grant - in - aid for jsps fellows ( 18.11384 ) .
99 d. n. spergel _ et al . _
[ wmap collaboration ] , astrophys .
j. suppl . * 148 * , 175 ( 2003 ) .
d. n. spergel _ et al . _
[ wmap collaboration ] , astrophys .
j. suppl .
* 170 * , 377 ( 2007 ) .
s. g. ryan , t. c. beers , k. a. olive , b. d. fields and j. e. norris , astrophys .
j. * 530 * , l57 ( 2000 ) . j. melendez and i. ramirez , astrophys . j. * 615 * , l33 ( 2004 ) .
m. asplund , d. l. lambert , p. e. nissen , f. primas and v. v. smith , astrophys . j. * 644 * , 229 ( 2006 ) .
s. inoue , w. aoki , t. k. suzuki , s. kawanomoto , a. e. garca - prez , s. g. ryan and m. chiba , iau symposium * 228 * , 59 ( 2005 ) .
r. cayrel _ et al .
_ , astron .
astrophys . * 473 * , l37 ( 2007 ) .
k. jedamzik , phys .
d * 74 * , 103509 ( 2006 ) .
k. jedamzik , phys .
* 84 * , 3248 ( 2000 ) .
m. kawasaki , k. kohri and t. moroi , phys .
d * 63 * , 103502 ( 2001 ) .
r. h. cyburt , j. ellis , b. d. fields and k. a. olive , phys .
d * 67 * , 103521 ( 2003 ) .
m. kusakabe , t. kajino and g. j. mathews , phys .
d * 74 * , 023526 ( 2006 ) .
k. jedamzik , phys .
d * 70 * , 063524 ( 2004 ) .
k. jedamzik , phys .
d * 70 * , 083510 ( 2004 ) .
m. kawasaki , k. kohri and t. moroi , phys .
d * 71 * , 083502 ( 2005 ) .
d. cumberbatch , k. ichikawa , m. kawasaki , k. kohri , j. silk and g. d. starkman , phys . rev .
d * 76 * , 123005 ( 2007 ) . in a parameter region of @xmath219
s and @xmath220 gev , @xmath6li and @xmath7li problems seem to be fixed @xcite , where the @xmath7li abundance decreases because of @xmath7be destruction by a primary @xmath7be(@xmath1,@xmath3he)@xmath0he reaction .
it is , however , found that in this parameter region where @xmath7li abundance decreases by photodissociation to the level of the @xmath7li plateau , either the d abundance was too low or the number ratio @xmath3he / d was too large .
this is absolutely excluded in light of standard stellar evolution and chemical evolution @xcite .
j. r. ellis , k. a. olive and e. vangioni , phys .
b * 619 * , 30 ( 2005 ) .
o. richard , g. michaud and j. richer , astrophys .
j. * 619 * , 538 ( 2005 ) .
d. l. lambert , aip conf .
proc . * 743 * , 206 ( 2005 ) .
m. pospelov , phys .
lett . * 98 * , 231301 ( 2007 ) .
k. hamaguchi _ et al .
_ , phys .
b * 650 * , 268 ( 2007 ) .
r. h. cyburt _ et al .
_ , jcap * 0611 * , 014 ( 2006 )
. k. kohri and f. takayama , phys .
d * 76 * , 063507 ( 2007 ) . c. bird , k. koopmans and m. pospelov , arxiv : hep - ph/0703096 .
m. kusakabe , t. kajino , r. n. boyd , t. yoshida and g. j. mathews , phys .
d * 76 * , 121302(r ) ( 2007 ) .
m. kusakabe , t. kajino , r. n. boyd , t. yoshida and g. j. mathews , astrophys . j. * 680 * , 846 ( 2008 ) .
m. kawasaki , k. kohri and t. moroi , phys .
b * 649 * , 436 ( 2007 ) .
m. kawasaki , k. kohri , t. moroi and a. yotsuyanagi , arxiv:0804.3745 [ hep - ph ] . t. jittoh , k. kohri , m. koike , j. sato , t. shimomura and m. yamanaka , phys .
d * 76 * , 125023 ( 2007 ) .
t. jittoh , k. kohri , m. koike , j. sato , t. shimomura and m. yamanaka , arxiv:0805.3389 [ hep - ph ] .
k. jedamzik , phys . rev .
d * 77 * , 063524 ( 2008 ) .
k. jedamzik , jcap * 0803 * , 008 ( 2008 ) .
m. pospelov , arxiv:0712.0647 [ hep - ph ] .
m. pospelov , j. pradler and f. d. steffen , arxiv:0807.4287 [ hep - ph ] .
t. shima _ et al .
_ , phys . rev .
c * 72 * , 044004 ( 2005 ) .
s. nakayama _ et al .
_ , phys .
c * 76 * , 021305(r ) ( 2007 ) .
m. kawasaki and t. moroi , astrophys .
j. * 452 * 506 ( 1995 ) .
r. j. protheroe , t. stanev and v. s. berezinsky , phys .
d * 51 * , 4134 ( 1995 ) .
v. s. berezinskii , s. v. bulanov , v. a. dogiel , v. l. ginzburg and v. s. ptuskin 1990 , astrophysics of cosmic rays , ed .
v. s. verezinskii and v. l. ginzburg ( new york : north - holland ) .
t. suzuki , s. chiba , t. yoshida , t. kajino and t. otsuka , phys .
c * 74 * , 034307 ( 2006 ) .
t. yoshida , t. kajino , h. yokomakura , k. kimura , a. takamura and d. h. hartmann , phys .
lett . * 96 * , 091101 ( 2006 ) .
t. yoshida , t. kajino , h. yokomakura , k. kimura , a. takamura and d. h. hartmann , astrophys .
j. * 649 * , 319 ( 2006 ) .
t. yoshida , t. suzuki , s. chiba , t. kajino , h. yokomakura , k. kimura , a. takamura and d. h. hartmann , submitted to astrophys .
j. w. c. haxton , phys .
lett . * 60 * , 1999 ( 1988 ) .
n. ohnishi , k. kotake and s. yamada , astrophys .
j. * 641 * , 1018 ( 2006 ) . | a recent measurement of @xmath0he photodisintegration reactions , @xmath0he(@xmath1,@xmath2)@xmath3h and @xmath0he(@xmath1,@xmath4)@xmath3he with laser - compton photons shows smaller cross sections than those estimated by other previous experiments at @xmath5 mev .
we study big - bang nucleosynthesis with the radiative particle decay using the new photodisintegration cross sections of @xmath0he as well as previous data .
the sensitivity of the yields of all light elements d , t , @xmath3he , @xmath0he , @xmath6li , @xmath7li and @xmath7be to the cross sections is investigated .
the change of the cross sections has an influence on the non - thermal yields of d , @xmath3he and @xmath0he . on the other hand ,
the non - thermal @xmath6li production is not sensitive to the change of the cross sections at this low energy , since the non - thermal secondary synthesis of @xmath6li needs energetic photons of @xmath8 mev .
the non - thermal nucleosynthesis triggered by the radiative particle decay is one of candidates of the production mechanism of @xmath6li observed in metal - poor halo stars ( mphss ) . in the parameter region of the radiative particle lifetime and the emitted photon energy which satisfies the @xmath6li production above the abundance level observed in mphss
, the change of the photodisintegration cross sections at @xmath9 mev as measured in the recent experiment leads to @xmath10 % reduction of resulting @xmath3he abundance , whereas the @xmath6li abundance does not change for this change of the cross sections of @xmath0he(@xmath1,@xmath2)@xmath3h and @xmath0he(@xmath1,@xmath4)@xmath3he .
the @xmath6li abundance , however , could show a sizable change and therefore the future precise measurement of the cross sections at high energy @xmath11 50 mev is highly required . |
in 1964 , hohenberg and kohn @xcite proved a theorem asserting that the ground- state properties of a stationary , non - relativistic system can be calculated from a variational principle involving only the one - body density @xmath4 of the system , without recourse to the schrdinger equation and its wavefunction . the theorem was soon generalized to finite-@xmath0 @xcite , and to relativistic systems @xcite .
thus it became clear that all thermodynamic properties of @xmath1-particle quantum systems , entangled , interacting , or not , could be calculated without recourse to the @xmath1-body wavefunction .
the method , re - written in the form of the kohn - sham theory has now become the preferred method in computational quantum mechanics ( qm ) .
furthermore , since linear transport properties depend only on equilibrium correlation functions ( c.f . , the kubo relations ) , a large class of quantum phenomena depends only on the density @xmath5 , a measurable quantity having a familiar ` ontology ' in quantum as well as classical mechanics . unlike the many - electron wavefunction of conventional qm
, the one - body density is an observable .
for instance , the x - ray scattering from a specimen of matter gives direct information regarding @xmath5 .
the development of ` time - dependent ' density functional theory @xcite , if substantiated , implies that density functional theory may enable a very different discussion of quantum phenomena than has been the custom in foundational studies . in this article , for the sake of brevity and simplicity , we limit ourselves mainly to non - relativistic equilibrium systems . in the following presentation
we look at dft and compare it with relevant aspects of bohmian mechanics , treatment of hybrid systems within dft , as well as classical representations inspired by dft ideas .
furthermore , finite-@xmath0 physics leads us to conclude that there is no need for _ ad hoc _ decoherence theories to eliminate macroscopic cat states from qm .
to be specific , we consider a system of electrons subject to an external potential @xmath6 , and interacting with one another via the coulomb potential . in many cases this external potential
is produced by a fixed set of nuclei , as in molecular physics or quantum chemistry calculations at @xmath7 .
however , the ions can not be treated as fixed in dealing with liquid metals and hot plasmas .
hence , in discussing true hybrid systems , we allow the nuclei ( or ions ) to be a second interacting subsystem , interacting with the electron subsystem , and the finite-@xmath0 discussion becomes not only appropriate , but also necessary .
however , in simplified models , e.g. , the jellium model , the ions provide a fixed uniform positive distribution of charge that exactly neutralizes the negative charge of the electron distribution .
the @xmath8 singularities in the coulomb potential @xmath9 for electron - electron or ion - ion interactions are exactly canceled by the corresponding term in the ion - electron interactions . however , while keeping such specific models in mind , we can simply work with an unspecified external potential @xmath10 and a corresponding one - body electron distribution @xmath5 .
then according to hohenberg and kohn @xcite , the exact ground state density @xmath5 and the ground - state energy @xmath11 $ ] corresponding to that external potential @xmath10 are obtained by the minimization of the following energy functional .
@xmath12=t[n]+u_{cou}[n]+\int d\vec{r}v_{ext}(\vec r)n(\vec r)\ ] ] the kinetic energy contribution , viz .
, @xmath13 $ ] , and the coulomb electrostatic energy @xmath14 $ ] are said to be ` universal functionals ' of @xmath5 , in the sense that they do not depend on the external potential @xmath15 specific to the system .
the original hohenberg - kohn theorem was established for non - degenerate ground states at @xmath7 .
this has since been restated within the language of functional analysis @xcite , while degenerate ground states , spin - polarized systems , finite-@xmath0 systems , relativistic systems , multi - component hybrid systems , superconductors , liquids and plasmas etc .
, have been studied within the first three decades of dft @xcite .
it is clear that a direct minimization of @xmath11 $ ] via @xmath16 would require a knowledge of the kinetic energy functional @xmath13 $ ] , and the electrostatic potential @xmath17 $ ] .
the latter is a known functional of @xmath5 , but @xmath13 $ ] is unknown .
although this is a universal functional of @xmath4 , it turns out to be highly non - local .
hence gradient expansions merely lead to thomas - fermi like theories .
these , sometimes known as ` orbital - free dft ' @xcite , presently lack the full accuracy of methods that use a wavefunction .
since dft is presented via a one - body density , the kohn - sham formulation of dft uses a one - body ` kohn - sham wavefunction ' to construct a non - local kinetic energy , where a single kohn - sham electron moves in the electrostatic potential of the other electrons and an additional exchange - correlation potential @xmath18 $ ] which is a functional of the one - body density .
thus , the many - electron problem is replaced by an effective one electron problem in kohn - sham theory . here the kohn - sham electron moves in a non - local kohn - sham potential that is made up of the electrostatic potential @xmath17 $ ] , and the non - local @xmath18 $ ] . however , unlike @xmath13 $ ] , approximate models of @xmath18 $ ] have proved to be quite successful , when used in the kohn - sham equation where @xmath13 $ ] is constructed for each case , instead of relying on a knowledge of the universal functional @xmath13 $ ] .
it is of interest at this stage to examine bohmian mechanics , where the schrdinger many - particle wavefunction @xmath19 is used to construct a non - linear equation of motion containing the external potential as well as a quantum potential @xmath20 .
let us limit ourselves to a single particle problem in 1-d for the sake of brevity .
bohm @xcite writes the wavefunction in the form : @xmath21 so that @xmath22 ( positive ) , and @xmath23 are real functions . substituting this form of the wavefunction into the schrdinger equation ,
one obtains bohm s form of the equations as : @xmath24&=&0 \\ \label{bo2 } \frac{\partial s}{\partial t}+v_{ext}+\frac{(\nabla s)^2}{2m}+q(x , t)&=&0\\ \label{qpot - eq } q(x , t)&=&\frac{-\hbar^2}{2m}\frac{\nabla^2r}{r } \end{aligned}\ ] ] eq .
[ bo1 ] is simply the equation of continuity .
the term @xmath25 in eq .
[ bo2 ] is interpreted as the kinetic energy of a point - like particles with momentum @xmath26 = @xmath27 . in effect , eq .
[ bo2 ] has the form : @xmath28 this is exactly like a classical hamilton - jacobi equation with the extra potential @xmath29 , called the quantum potential , or the bohm potential .
since we are interested in the simplest form of dft ( systems in equilibrium ) in this presentation , let us consider an electron confined within an infinite potential well of width @xmath30 .
the eigenfunctions are of the form @xmath31 .
it is easily shown that the particle current @xmath32 is zero , and the bohmian particle is _ at rest _ in some unknown location @xmath33 in the well , with the probability @xmath34 of _ being _ there .
einstein felt that the particle being at rest in the well was not the correct physical picture . a more detailed discussion of the bohmian particle in a well , and a possible resolution to einstein s objections @xcite are given in chapter 6 of ref .
@xcite .
does the bohmian particle in the well have no kinetic energy ? it is easy to show that the kinetic energy is completely resident in bohm s quantum potential while the particle current is zero .
in fact , the kinetic energy functional @xmath35 $ ] per electron confined in the quantum well , given by hohenberg - kohn theory ( or conventional qm ) is essentially the quantum potential of bohm .
although some advocates of bohmian mechanics have moved away from the quantum potential ( mainly due to philosophical reasons ) , we find it to be of a sort similar to the effective potentials based on density - functionals used in condensed - matter many - body physics . for instance , one of the objections adduced against the quantum potential is that it has the ` unusual effect ' of making the particle density @xmath1 influence the dynamics of individual particles .
however , that is precisely the message of dft .
it is typical of many - particle physics that the individual dynamics is influence by the over - all particle distribution .
for instance , a low - order ( mean - field ) many - body theory like the hartree theory of many electrons requires that an electron moves in the many - body field of the other electrons produced by the density distribution @xmath36 .
hartree - fock theory involves a highly non - local hartree - fock potential dependent on the density @xmath5 .
interestingly , drr , goldstein and zhangi @xcite summarize their attitude to the quantum potential with the statement , ` we believe that the most serious flaw in the quantum - potential formulation of bohmian mechanics is that it gives a completely wrong impression of the lengths to which we must go in order to convert orthodox quantum theory into something more rational .
the quantum potential suggests
we must incorporate into the theory a quantum potential of a _ grossly _ non - local character ' .
we emphasized the word ` grossly ' in the above to high - light where the attitudinal squeamishness resides .
in contrast , most physicists who use the highly non - local hartree - fock potential as a basic first step in their calculations have given up any such squeamishness .
regarding ` the lengths ' needed to reach the classical hamilton - jacobi form , as seen from above , we need only a few lines .
the insights from dft enable us to consider the bohm potential as a functional of the density , without invoking a wavefunction .
furthermore , eq . [ qpot - eq ] may be used as a differential equation for the direct determination of @xmath37 given in terms of @xmath38 , subject to satisfying the equation of continuity , and minimizing the energy . once @xmath22 is determined by iteratively solving eq .
[ qpot - eq ] , @xmath23 can be obtained from the continuity equation . in effect ,
a time - dependent system is described by the four vector consisting of the current @xmath32 and the density @xmath39 .
these are precisely the quantities treated in dft , and in bohmian mechanics .
the above considerations suggest that bohmian mechanics can be formulated entirely in a language free of the wavefunction , using only density functionals , exactly as in dft .
the hohenberg - kohn theorem reduces the @xmath1-electron problem to that of a functional of an effective one - body problem and an exchange - correlation potential @xmath18 $ ] .
given that the basic electron - electron interactions ( that lead to quantum correlations including those of the bell type ) are pair - interactions and pauli exclusion effects , the reduction to a one - body problem achieved in dft appears a bit mysterious , and one would expect that fundamental two - particle correlations are embedded in the @xmath40 .
in fact , the latter is most transparently evaluated as a coupling constant integration over the pair - distribution function ( pdf ) , i.e. , @xmath41 of the system .
this is the probability of finding an electron at @xmath42 , given an electron at @xmath43 .
the pdf is well - defined both classically and quantum mechanically . in qm ,
the pdf is calculated from the many - particle wavefunction by integrating over all except two space variables @xmath44 of two electrons . in a classical system
, we can consider a uniform electron fluid ( as in jellium ) , with a mean density @xmath1 and wigner - seitz radius @xmath45 .
we position the origin on the first electron , so that @xmath46 .
then , invoking radial symmetry and uniformity , we can use @xmath47 rather than @xmath48 for the pdf , where @xmath49 is the radial distance of the ( second ) electron from the origin .
when classical mechanics holds ( i.e. , for @xmath50 and for very small de broglie wavelengths ) , this @xmath47 for an interacting coulomb fluid can be calculated using the modified - hyper - netted - chain ( mhnc ) equation @xcite , or via molecular dynamics ( md ) . in qm
, even the non - interacting electron system is of interest if all the particles have the same spin
. then the pauli exclusion principle controls the entanglement of the @xmath1-electron system .
this non - interacting set of @xmath1-entangled electrons is described by a slater determinantal wavefunction .
the fully spin - polarized system ( parallel spins @xmath51 ) , and the fully unpolarized system ( anti - parallel spins @xmath52 ) are of interest .
they can be analytically treated at @xmath7 as in ref .
@xcite , and extended to finite-@xmath0 as in ref .
furthermore , the pauli exchange effect ( which leads to entanglement since the wavefunction is a slater determinant ) can be re - written as a non - local quantum potential which is a universal function of @xmath53 .
this ` pauli - exclusion ' potential @xmath54 can be constructed such that it reproduces the quantum @xmath47 when used in the classical mhnc equation .
this provides an interesting and useful classical map of the quantum problem that has been very fruitful in applications to interacting quantum problems ranging from 2-d electron layers @xcite , hydrogen plasmas , electrons in graphene , and quantum dots @xcite .
that is , correlated , interacting quantum problems can be replaced by equivalent classical problems where pair - potentials involve additional potentials containing quantum effects which are usually non - local .
bohmian mechanics achieves this directly from the @xmath55 part of the wavefunction .
dft achieves this via functionals of the one - body density @xmath5
. naturally , non - local contributions are usually quite important in both dft and bohmian mechanics .
gives the probability of finding a @xmath51-spin electron at @xmath49 if there is already a @xmath51-spin electron at the origin ( coulomb repulsion neglected ) .
we plot @xmath53 on the @xmath33 axis , and the result is valid for any density , @xmath56 being the sphere - radius containing an average of one electron .
thus the ` entanglement ' due to pauli exclusion holds even if @xmath56 is a lunar distance .
the pauli exclusion is exactly mimicked by a scale - independent repulsive potential @xmath57 when used in a classical mhnc equation .
, height=6 ] the pdf of non - interacting fully spin - polarized entangled electrons , denoted by @xmath58 is shown in fig .
[ pdf - fig ] .
it has the form : @xmath59 ( the large-@xmath49 behavior is normalized to unity , whereas if @xmath60 , it is natural to normalize the spin - resolved functions to 1/2 ) . here
@xmath61 is the spherical bessel function , while its argument @xmath33 contains @xmath62 , where @xmath63 is the fermi momentum .
the probability of finding an electron at @xmath49 , given that there is an electron at the origin of coordinates is affected by their mutual coulomb interaction .
this is not included in the above calculation , and hence the superfix ` zero ' in @xmath64 .
a treatment including the coulomb potential is given in ref .
@xcite , where it is shown that remarkably good agreement with monte - carlo quantum simulations is obtained for these classical maps .
since the pdf scales as @xmath53 , the correlations remain undiminished , whatever be the mean - separation @xmath65 that is imposed on any two electrons in the entangled system .
that is , _ local realism _ as required by the einstein - podolsky - rosen gedanken experiment does not hold .
thus the correlations unearthed by the bell inequalities are very transparently and simply revealed by an examination of the pdf of correlated electron systems .
this topic is discussed in greater detail in chapters 7 and 8 of ref .
formal discussions of hybrid systems ( i.e. , systems containing quantum particles as well as classical particles ) @xcite rarely bother to discuss the role of temperature , partly because some of these discussions deal with model problems dealing with just one or two particles . in reality , all particles in any system obey quantum mechanics . however , if some particles have de broglie wavelengths @xmath66 smaller than some assigned length scales associated with the system , then such particles among themselves can be treated within classical mechanics .
the interactions between the particle deemed to be classical , and those deemed to be quantum occur via a hamiltonian @xmath67 whose characteristic length scales and energy scales have to be investigated separately . in the following
we look at systems at finite temperature @xmath0 .
if a system is coupled to a heat bath and held at some finite temperature @xmath0 , the average kinetic energy of the particles is a function of @xmath0 . in the classical limit
the average kinetic energy of an atom is @xmath68 , using energy units for @xmath0 where the boltzmann constant becomes unity . at @xmath7 , for our model system of fermions ( jellium ) , the kinetic energy becomes @xmath69 , where @xmath70 is the fermi energy .
the latter can be thought of as a temperature @xmath71 .
hence we see that we can ascribe a de broglie wavelength @xmath66 and a thermal momentum @xmath72 to any particle in a system at an effective temperature @xmath0 , with @xmath73 and @xmath66 given by : @xmath74 it is clear that @xmath66 is infinite for systems at @xmath7 , or for sufficiently dilute quantum systems ( @xmath75 as @xmath76 ) .
thus all systems become quantum systems as @xmath77 since their de broglie wavelengths become infinite . however , if at some finite ( even if low ) temperature , e.g. , 1 kelvin , if the mass of a particle is sufficiently large , then if @xmath78 atomic unit , we may assume that superposition and entanglement become unimportant .
such heavy particles are classical particles .
hence , since all systems are quantum systems at @xmath7 , and heavy particles become classical at some finite-@xmath0 , any quantum theory of hybrid systems @xcite has to ultimately incorporate a quantum treatment of thermal fields . the conventional quantum approach to finite-@xmath0 problems
is difficult , and involves a typical doubling of the hilbert space as in umezawa s thermofield dynamics @xcite , or in the martin - schwinger - kedysh contour technique @xcite .
these methods are useful for tackling a class of non - equilibrium physics .
however , even static finite-@xmath0 quantum calculations ( e.g. , via the matsubara method @xcite ) are restricted to perturbation theory . on the other hand ,
finite-@xmath0 density functional theory @xcite presents a computationally and conceptually much simpler approach to finite-@xmath0 problems , as well as two - temperature quasi - equilibria . here
the finite @xmath0 one - body density functionals @xmath79 determine all the thermodynamics and linear transport properties of the system _ without _ the need for a wavefunction . the explicit construction of a dft and the calculation of such finite-@xmath0 hybrid systems for a system of protons ( classical ) and electrons ( quantum ) at finite-@xmath0 has been given in ref .
the method is applicable to any electron - ion system , where the ions are treated classically .
the usual born - oppenheimer approximation is not needed .
if the ion distribution is denoted by @xmath80 , and the electron distribution is @xmath4 , the total free energy @xmath81 , [ n])$ ] is a functional of both distributions .
hence we have two coupled hohenberg - kohn - mermin type equations to determine the finite-@xmath0 distributions of the hybrid system .
that is , the functional derivatives with respect to variations @xmath82 lead to two coupled euler - lagrange equations .
@xmath83}{\delta n}=0,\;\ ; \frac{\delta f[\rho , n]}{\delta \rho}=0.\ ] ] the first of these equation can be reduced to a kohn - sham equation for electrons at finite-@xmath0 , or simply retained as a hohenberg - kohn variational form .
the second equation leads to a classical hohenberg - kohn equation for the ion distribution that is identified as a form of the mhnc equation . in effect
, it is a boltzmann distribution for the density @xmath84 in terms of the ` potential of mean force ' ( pmf ) used in the statistical mechanics of liquids @xcite .
the pmf is the appropriate kohn - sham potential for this classical system .
the full practical implementation of this approach to a typical hybrid system ( an interacting gas of electrons and protons at finite @xmath0 inclusive of the formation of hydrogenic bound states ) is given in ref .
@xcite .
the equation [ debrog - eq ] is of great importance in showing us that there is no need for _ ad hoc _ theories of decoherence to eliminate macroscopic cat states from qm .
let us consider a 1 kg cat at room temperature ( 300 k , i.e. , @xmath85 hartees ) .
the atoms in the cat have internal energies with approximately @xmath68 arising from translational motion .
the center of mass will also make random oscillatory motions about an equilibrium position consistent with its temperature and any motion of the cat .
the corresponding center - of - mass de broglie wavelength @xmath86 turns out to be @xmath87 the ` radius ' of a proton is about 0.88 femtometers ( 1 fm=10@xmath88 meters ) .
the de borglie wavelength of the cat is about one hundred - millionth of the size of a proton !
even when cooled to one kelvin , the de broglie wavelength remains totally negligible .
even a free electron close to a cat s body would not get entangled with the cat s quantum energy states .
matter has to be squashed to densities where nuclear reactions begin before superpositions are possible at such small values of @xmath89 .
clearly , schrdinger s cat states of macroscopic bodies ( entanglements or superpositions ) do not exist in nature except at @xmath0=0 , or possibly in white dwarfs .
one may argue that the mass @xmath90 of the center of mass is not the proper quantity to use .
one may claim that qm should be applied to some molecules of the cat that are specific to it being dead or alive in schrdinger s cat paradox . in that case
we are no longer applying qm to a macroscopic system and then there is no difficulty in having superpositions or entangled states .
chemical reactions ( e.g. , cyanide reacting with the cat ) are simply processes where reactive atomic groups enter into entanglements .
such processes pose no difficulty and are not ` paradoxical ' .
one may also question the use of a momentum calculated from @xmath68 .
this is the classical kinetic energy of an ideal gas .
we know from dulong and petit s law ( or from the more modern einstein - debye theory of solids ) that classically , a system held together by chemical bonds , e.g. , a cat , has three degrees of vibration per atom , and three degrees of translation , each having an energy @xmath91 . for solids in the low - temperature limit , we need the fermi energy of the system . such detailed treatments merely introduce numerical factors of little consequence .
the basic conclusion that @xmath89 is very very small remains firm .
these considerations imply that explicit decoherence theories like those of penrose ( ` objective reduction ' due to quantum gravity ) @xcite , or the ` spontaneous localization ' model of ghirardi , rimini and weber @xcite are unnecessary additions to quantum mechanics , if we grant that the relevant de broglie lengths are much smaller than even nuclear radii .
the tradition of studying just one or two quantum particles , usually in a two - state model has a long history in foundational studies . however ,
if the many - body problem is examined , qm can be approached using one - body density functionals , without recourse to wavefunctions , both at @xmath7 and at finite-@xmath0 .
an examination of bohmian mechanics within the perspective of dft is conceptually rewarding and clarifies the nature of the bohmian quantum potential . finite-@xmath0
quantum mechanics leads us to consider the thermal de broglie wavelength which , being negligible for macroscopic systems , ensures that there are no macroscopic schrdinger cat states .
90 hohenberg p and kohn w 1964 _ phys . rev . _
* 136b * 864 mermin n d 1965 _ phys . rev . _ * 137 * a1441 rajagopal a k and callaway j 1973 _ phys .
rev b _ * 7 * 1912 ; macdonald a h and vosko s h 1979 _ j phys c _ * 12 * 2977 runge e and gross e k u 1984 _ phys
lett . _ * 52 * 997 lieb e 1985 _ density functional methods in physics _
nato series * b123 * , ed . dreizler r m and providencia j ( plenum new york ) karasiev v v and trickey s b 2012 preprint ( arxiv:1109.6602 ) gross e k u and
dreizler r m 1993 _ density functional theory _
nato asi series * 337 * ( plenum new york ) bohm d 1952 _ phys
_ * 85 * 166 and 1953 _ phys .
rev . _ * 89 * 1040 einstein a 1953 _ scientific papers presented to max born _
( _ festschrift _ ) ( hafner new york ) p 33 and bohm d in the same volume , p 13 - 19 dharma - wardana m w c 2013 _ a physicist s view of matter and mind _
( world scientific singapore ) drr d goldstein s and zhangi n 1995 _ bohmian mechanics and quantum theory : an appraisal _ , eds . cushing j t fine a and goldstein s ( kluwer academic press new york ) ( _ preprint _ quant - ph/9511016 )
mahan g d 1981 _ many - particle physics _ ( plenum new york ) dharma - wardana m w c and perrot f 2000 _ phys .
lett . _ * 84 * 959 perrot f and dharma - wardana m w c 2001 _ phys .
lett . _ * 87 * 206404 dharma - wardana m w c 2012 _ int .
_ * 112 * 53 heslot a 1985 _ phys .
d _ * 31 * 1341 umezawa h 1993 _ advanced field theory : micro , macro and thermal physics . _
( aip new york ) rammer j and smith h 1986 _ rev .
* 58 * 323 dharma - wardana m w c and perrot f 1982 _ phys .
a _ * 26 * 2096 hansen
j - p and macdonald i r 2006 _ theory of simple liquids _
( academic london ) penrose r 1996 _ gen . relativ . grav .
_ * 28 * 581 ghirardi c rimini a and weber t 1986 _ phys .
d _ * 34 * , 470 | the advent of the hohenberg - kohn theorem in 1964 , its extension to finite-@xmath0 , kohn - sham theory , and relativistic extensions provide the well - established formalism of density - functional theory ( dft ) .
this theory enables the calculation of all static properties of quantum systems _ without _ the need for an @xmath1-body wavefunction @xmath2 .
dft uses the one - body density distribution instead of @xmath2 .
the more recent time - dependent formulations of dft attempt to describe the time evolution of quantum systems without using the time - dependent wavefunction .
although dft has become the standard tool of condensed - matter computational quantum mechanics , its foundational implications have remained largely unexplored .
while all systems require quantum mechanics ( qm ) at @xmath0=0 , the pair - distribution functions ( pdfs ) of such quantum systems have been accurately mapped into classical models at effective finite-@xmath0 , and using suitable non - local quantum potentials ( e.g. , to mimic pauli exclusion effects ) .
these approaches shed light on the quantum @xmath3 hybrid @xmath3 classical models , and provide a new way of looking at the existence of non- local correlations without appealing to bell s theorem .
they also provide insights regarding bohmian mechanics .
furthermore , macroscopic systems even at 1 kelvin have de broglie wavelengths in the micro - femtometer range , thereby eliminating macroscopic cat states , and avoiding the need for _ ad hoc _ decoherence models . |
the idea behind the approach is based on the fact that the probabilities @xmath2 depend on the system s size @xmath102 and therefore are awkward to use in mimicking a macroscopic experiment .
thus , according to normalization condition , @xmath103 , and in equilibrium @xmath104 or @xmath105 provided the corresponding mean exists . on the other hand ,
@xmath21 and @xmath23 are the `` intensive '' variables , e.g. @xmath106 , and do not depend on the system s size provided the mean particle concentration @xmath107 is kept constant . to see that @xmath108 introduced in the main text are indeed activities one can proceed as follows .
adding a weak external potential ( e.g. electric field giving rise additional potential energy @xmath109 at site @xmath7 ) changes the equilibrium concentrations and hence the corresponding equation to @xmath110 $ ] . here
@xmath111 can indeed be interpreted as activities , since @xmath112 have a form of chemical potentials on the sites .
the master equation can be interpreted as a combination of the local continuity equation @xmath113 and the local linear response equation @xmath114 with @xmath115 .
close to equilibrium and for very small potentials @xmath116 the values of @xmath117 are close to unity , so that the deviations of chemical potentials from their equilibrium values of zero are small . therefore ,
close to equilibrium , @xmath118 and @xmath119 the external potential @xmath109 is set to zero in the main text .
it is interesting to note that not the chemical potentials but the activities at the sites are reinterpreted as electric potentials , which leads to differences in the behavior of the random resistor - capacitor model and the random potential model far from equilibrium , i.e. for large concentration gradients .
in a stationary situation corresponding to quasiequilibrium we have @xmath120 where the sum runs over the nearest neighbors of the site @xmath3 .
moreover , in the thermodynamical limit @xmath40 the permeability of the membrane tends to zero , and all local currents in it as well : @xmath121 now we show that in quasi - equilibrium the local concentrations and the local activities ( `` potentials '' ) tend to be independent from each other ( although the local transition rates are correlated with the boltzmann factors and thus local concentrations ) . to see this we note that according to eq.([means ] ) the activity ( potential ) at a site @xmath3 is the weighted arithmetic mean of the activities of the neighbors it is connected to , and eq.([curr ] )
states that the difference between the activities of the connected sites under quasi - equilibrium tends arbitrarily small .
therefore our system can be considered as composed of large , practically equipotential regions whose potential hardly fluctuates around its mean @xmath122 depending on the region s position . in these regions the concentrations @xmath123 can be averaged over the physically small volume still containing the large number of sites , so that @xmath124 .
we note that this last property does not rely on homogenization or on the isotropy and will hold even if our system is built of independent parallel or interwoven wires !
let us now assume that at large scales the corresponding electric system homogenizes . in this case the total voltage profile ( which in our case is mapped on the profile of activities ) obtained by the coarse graining ( moving average ) of the voltages ( activities ) over macroscopic domains of the system which are large enough compared to the lattice spacing but small compared to @xmath31 follows a linear behavior ( similar to those of the effective concentration in fig.1 ) .
then the coarse - grained activity is a linear function of the coordinate , and thus also the coarse grained concentration is linear in coordinate , with the proportionality factor @xmath125 between the both .
we start this section by showing the discretization procedure linking the fokker - planck equation to the master equation .
the equation in a continuous @xmath8-dimensional space is discretized using a regular hypercubic lattice with lattice constant @xmath4 considered sufficiently small .
the sites are identified by the vectors @xmath126 , the site energies @xmath127 are given by corresponding values of the random potential @xmath64 at the position of site @xmath126 , @xmath128 where @xmath129 is an integer number and @xmath16 runs from 1 to @xmath8 .
correspondingly , the derivatives @xmath130 along a certain direction @xmath16 are linked to the forward differences @xmath131 in the following way : @xmath132 \nonumber \\ \partial_k^2 p(\mathbf{x } ) & \rightarrow & \frac{1}{a^2 } \delta^2_k p_{\mathbf{i } } = \frac{1}{a^2 } [ p_{\mathbf{i } + a\hat{k } } - 2 p_{\mathbf{i } } + p_{\mathbf{i } - a\hat{k}}]\end{aligned}\ ] ] where @xmath133 is the unit vector in the @xmath16-th direction .
hence , from @xmath134 = \\ & = & d
\sum_{k=1}^d \big [ \partial^2_k p(\mathbf{x } ) + \beta \partial_k [ \partial_k u(\mathbf{x } ) \cdot p(\mathbf{x})]\big ] \end{aligned}\ ] ] follows @xmath135 + \nonumber \\ & & + \frac{\beta}{a}\partial_k \big ( ( e_{\mathbf{i } + a\hat{k } } - e_{\mathbf{i } } ) p_{\mathbf{i}}\big)\big ] \end{aligned}\ ] ] we consider the potential to be a slowly varying function of the position ; this allows to substitute the remaining partial derivative with half of a double step forward difference , the consequent error being of the smaller order of magnitude @xmath136 .
@xmath137 + \\ & & + \frac{\beta}{2 } \big ( ( e_{\mathbf{i } + 2a\hat{k } } - e_{\mathbf{i } + a\hat{k } } ) p_{\mathbf{i } + a\hat{k } } - ( e_{\mathbf{i } } - e_{\mathbf{i}-a\hat{k } } ) p_{\mathbf{i}-a\hat{k}}\big)\big ] \nonumber\end{aligned}\ ] ] neglecting again second order terms we replace the quantity @xmath138 with @xmath139 , and add and subtract the term @xmath140 @xmath141 + \nonumber \\ & & + \frac{\beta}{2 } \big ( ( e_{\mathbf{i } + a\hat{k } } - e_{\mathbf{i } } ) p_{\mathbf{i}+a\hat{k } } - ( e_{\mathbf{i } } - e_{\mathbf{i } -a\hat{k } } ) p_{\mathbf{i } -a\hat{k } } + \nonumber \\ & & + ( e_{\mathbf{i } + a\hat{k } } - e_{\mathbf{i } } ) p_{\mathbf{i } } - ( e_{\mathbf{i } + a\hat{k } } - e_{\mathbf{i } } ) p_{\mathbf{i}}\big)\big ] = \\ & = & \frac{d}{a^2 } \sum_{k=1}^d \big [ [ p_{\mathbf{i } + a\hat{k } } - 2 p_{\mathbf{i } } + p_{\mathbf{i } -a\hat{k } } ] + \nonumber \\ & & + \frac{\beta}{2 } \big ( ( e_{\mathbf{i } + a\hat{k } } - e_{\mathbf{i } } ) p_{\mathbf{i}+a\hat{k } } - ( e_{\mathbf{i } } - e_{\mathbf{i } -a\hat{k } } ) p_{\mathbf{i } -a\hat{k } } + \nonumber \\ & & + ( e_{\mathbf{i } + a\hat{k } } - e_{\mathbf{i } } ) p_{\mathbf{i } } + ( e_{\mathbf{i } -a\hat{k } } - e_{\mathbf{i } } ) p_{\mathbf{i}}\big)\big ] = \\ & = & \frac{d}{a^2 } \sum_{k=1}^d \big [ [ p_{\mathbf{i } + a\hat{k}}\big(1 - \frac{\beta}{2 } ( e_{\mathbf{i } } - e_{\mathbf{i } + a\hat{k}})\big ) + \nonumber \\ & & + p_{\mathbf{i } -a\hat{k}}\big(1 - \frac{\beta}{2 } ( e_{\mathbf{i } } - e_{\mathbf{i } -a\hat{k}})\big ) + \\ & & - p_{\mathbf{i}}\big(2 - \frac{\beta}{2 } ( e_{\mathbf{i } + a\hat{k } } - e_{\mathbf{i } } ) - \frac{\beta}{2 } ( e_{\mathbf{i } -a\hat{k } } - e_{\mathbf{i}})\big)\big ] \nonumber\end{aligned}\ ] ] since the difference between the energies at neighbouring sites is small , in the lowest nonvanishing order we can put @xmath142 \end{aligned}\ ] ] at this point , the discretized fokker - planck equation is shown to be equivalent to the master equation @xmath143\end{aligned}\ ] ] by setting @xmath144 implying that , up to constant factors @xmath145 in what follows @xmath146 can be changed for @xmath147
. thus , returning to the continuous limit when calculating the macroscopic ( effective medium ) conductance of the continuous disordered medium we can take its local conductivity to be @xmath148 with @xmath149 .
we now proceed to compute the hashin - shtrikman bounds for the effective diffusion constant by considering two different distributions of the energy variable @xmath127 . in the binary case @xmath150 with @xmath151 .
the local conductance follows again a binary distribution @xmath152 or , equivalently , @xmath153 here we introduce the following notation for different averages which will repeatedly appear in what follows ( weighted arithmetic , arithmetic and geometric mean ) : @xmath154 the notation @xmath155 will be continuously used for the mean of any physical quantity @xmath156 over the energy distribution .
the effective medium approximation then gives the effective conductance @xmath157 as the solution for the self - consistency condition @xmath158 resulting in @xmath159 ^ 2 } } \ ; \right]\ ] ] the percolating case is easily recovered in the limit @xmath160 , @xmath161 , giving the wellknown result ( [ 14 ] ) @xmath162 while simple calculations can show that in the two - dimensional symmetric case , @xmath163 , @xmath164 , @xmath54 equals the geometric mean @xmath165 , as stated in the main text .
eq.([bingeff ] ) is finally used to calculate the effective diffusion constant @xmath166 ^ 2 } } \ ; \right ] \nonumber\end{aligned}\ ] ] following the strategy outlined in [ 11 ] , we now introduce a free parameter @xmath167 and define the quantity @xmath168 according to [ 11 ] , the following inequalities hold : @xmath169 denoting @xmath170 we get the following inequalities @xmath171 \\ g^ * & < & g_2^ * = g_2\big [ 1 - \frac{dp(g_2 - g_1)}{dg_2 -(1- p)(g_2 - g_1)}\big ] \end{aligned}\ ] ] and define @xmath172 \\ d_2^ * & = & \frac{a^2 g_2^*}{\langle e^{-\beta e_{\mathbf{i } } } \rangle } = d \frac { g_2}{\langle g \rangle}\big [ 1 - \frac{dp(g_2 - g_1)}{dg_2 -(1- p)(g_2 - g_1)}\big ] \nonumber\end{aligned}\ ] ] giving the bounds @xmath173 through a simple rescaling of the units we can fix @xmath174 and express @xmath175 via the contrast @xmath176 and @xmath54 via its rescaled value @xmath177 , to get a better graphical representation .
the corresponding boundaries and the ema result for @xmath178 then read @xmath179 - ( x+1)}{2(d-1)[p + ( 1-p)x ] } \times \\ & \times & \left[\ ; 1 + \sqrt{1 + \frac{4(d-1)x } { [ d[p + ( 1-p)x ] - ( x+1)]^2 } } \ ; \right ] \nonumber \end{aligned}\ ] ] @xmath180 \\ \tilde{d}_2^ * & = & \frac{x}{p + ( 1-p)x } \big [ 1 - \frac{d p(x - 1)}{dx - ( 1- p)(x - 1)}\big ] \end{aligned}\ ] ] with @xmath181 let us now consider the case in which the possible values of the site energies @xmath127 are exponentially distributed in the interval @xmath182 $ ] @xmath183 performing the change of variables it is easy to show that the bond conductance is a random variable uniformly distributed in the interval @xmath184 $ ] .
eq.([scc ] ) then gives the following condition to be satisfied by the effective conductance @xmath185 e^{-g_2/dg^ * } = \big [ ( d-1)g^ * + g_1\big ] e^{-g_1/dg^*}\ ] ] or by the effective diffusivity @xmath60 @xmath186 e^{-g_2 d/(d\langle g \rangle d^ * ) } = & & \nonumber \\
= \big [ ( d-1)\langle g \rangle d^ * + g_1 d\big ] e^{-g_1d/(d\langle g \rangle d^ * ) } & & \end{aligned}\ ] ] which can be rewritten through the previous rescaling as @xmath187 e^{- 2x/(d(x+1)\tilde{d}^ * ) } = & & \nonumber
\\ = \big [ ( d-1)(x+1)\tilde{d}^ * + 2\big ] e^{-2/(d(x+1 ) \tilde{d}^ * ) } & & \end{aligned}\ ] ] the solution @xmath178 of this equation is shown graphically in fig .
3 of the main text together with the two bounds we get ready to calculate .
we consider again @xmath188 and the inequalities @xmath189 in this case we have @xmath190\ ] ] and setting respectively @xmath191 and @xmath192 , we can write @xmath193 \nonumber \\ & = & 1 - \frac{1}{y_1 } \ln\big(1 + y_1\big ) \\
b(g_2 ) & = & 1 - \frac{dg_2}{g_2 - g_1}\ln\big [ \frac{dg_2}{dg_2 - g_2 + g_1}\big ] \nonumber \\ & = & 1 + \frac{1}{y_2}\ln\big(1 - y_2\big)\end{aligned}\ ] ] where the couple of parameters @xmath194 has been introduced . at the end
the two bounds for the conductivity are given by @xmath195 \\
g^ * & < & g_2^ * = g_2\big [ 1 - d \big(\frac{y_2}{\ln(1 - y_2 ) } + 1\big)\big ] \end{aligned}\ ] ] from which we obtain @xmath196 \\ d_2^ * & = & \frac{d}{\langle g \rangle}g_2\big [ 1 - d \big(\frac{y_2}{\ln(1 - y_2 ) } + 1\big)\big ] \end{aligned}\ ] ] and the following bounds @xmath197 then , fixing again @xmath198 , @xmath199 , @xmath200 , we write @xmath201 \\ \tilde{d}_2^ * & = & \frac{2x}{x+1 } \big [ 1 - d - \frac{x - 1}{x \ln(1 - \frac{x - 1}{dx } ) } \big ] \end{aligned}\ ] ] with @xmath181 | a relation between the effective diffusion coefficient in a lattice with random site energies and random trasition rates and the macroscopic conductivity in a random resistor network allows for elucidating possible sources of anomalous diffusion in random potential models .
we show that subdiffusion is only possible either if the mean boltzmann factor in the corresponding potential diverges or if the percolation concentration in the system is equal to unity ( or both ) , and that superdiffusion is impossible in our system under any condition .
we show also other useful applications of this relation .
a classical particle s diffusion in a random potential or hopping on a lattice with disordered site energies @xmath0 is a versatile theoretical model with a wide range of applications @xcite .
the particle s motion typically corresponds to normal diffusion , but can get subdiffusive in the presence of deep traps or in the case of infinte contrast when approaching percolation transition .
a question arises , whether there can be other cases leading to subdiffusion except for these two ( or combinations thereof ) . in what follows
we show that the diffusion coefficient in a discrete disordered lattice is always finite ( i.e. that no superdiffusion can be observed ) but may vanish ( possibly giving rise to subdiffusion ) . independently of the particular distribution of _ nonzero _
transition rates , this is possible either if the percolation threshold in the corresponding network is unity ( e.g. in one dimension , on finitely ramified fractals , or when we are _ already at percolation threshold _ ) or if the mean boltzmann factor @xmath1 diverges ( or both ) .
according to the arrhenius law , the last situation corresponds to the divergence of the mean sojourn time at a site , pertinent to trapping .
we start from the master equation for the probabilities @xmath2 to find a particle at a site @xmath3 of a lattice ( with lattice spacing @xmath4 ) @xmath5 where @xmath6 are transition rates from site @xmath7 to site @xmath3 different from zero only for nearest neighbors .
eq .
( [ meq0 ] ) can either follow from some microscopic scheme or be obtained by a discretization of the fokker - planck equation for the overdamped motion in a continuous potential .
the system is taken to be homogeneous and isotropic in statistical sense .
this requirement excludes underdamped cases for which the velocities and coordinates enter differently , thus leading to anisotropy of the state ( phase ) space . in what follows we consider a @xmath8-dimensional lattice with total of @xmath9 sites assigned energies @xmath0 being identically distributed random variables .
we assume that the system is isothermic and possesses true thermodynamical equilibrium under appropriate boundary conditions , i.e. that the transition rates fulfill the detailed balance condition @xmath10 at equilibrium ( the superscript 0 will denote the corresponding value at equilibrium throughout the work ) .
the transition rates @xmath11 are not necessarily bounded from above , and some of them may be put to zero to mimic percolation situations .
the values of @xmath12 are given by @xmath13 , where @xmath14 denotes the boltzmann factor , @xmath15 is the temperature and @xmath16 is the boltzmann constant .
our discussion maps the initial problem onto the one for random resistor - capacitor networks .
let @xmath17 be the corresponding conductivities of the bonds , and @xmath18 be the effective conductivity of a corresponding network in the static regime .
then the effective diffusion coefficient in a network follows as @xmath19 the statements done in the first paragraph are then demonstrated by using the results from the theory of electric circuits and from the percolation theory .
some other useful applications of eq.([maineq ] ) are shown .
eq.([maineq ] ) by itself is not new , but we give here its physical derivation which stresses its general applicability and its connection with thermodynamics .
thus , the discussion for the case of a barrier model ( all @xmath0 are the same , but the transition rates fluctuate ) is contained in ref .
@xcite , eq.(2.15 ) . moreover , eq.([maineq ] ) naturally appears when applying effective medium approximation ( ema ) , like the one of @xcite .
a derivation for a continuous case ( langevin description in ito interpretation ) is given in @xcite .
note that the ito prescription may correspond to the trap model in the discrete case @xcite , i.e. to a situation different from the one of ref .
@xcite . in our work
we confine ourselves to a discrete setup which allows for the application of the theory of electric circuits for the analysis of the results ( although we make a limiting transition to continuum to illustrate some outcomes of the approach ) .
we start by rewriting eq.([meq0 ] ) as an equation for mean numbers ( `` concentrations '' ) of non - interacting particles at the corresponding sites , @xmath20 , connected with probabilities via @xmath21 with @xmath22 being the total number of particles . in equilibrium
all @xmath23 are proportional to the boltzmann factors , @xmath24 with prefactor @xmath25 depending on the number of particles , on the system s size and on distribution of @xmath14 . putting the detailed balance condition into the form @xmath26 we denote @xmath27 where @xmath28 is now a property of the bond . using this notation we rewrite eq.([meq0 ] ) as an equation for the temporal evolution of _ activities _ @xmath29 ( see the appendix ) : @xmath30 eq.([fuga ] ) is formally equivalent to the evolution equation of node potentials in a random resistor - capacitor model @xcite , with conductivities @xmath17 and capacitances @xmath23 .
let us now calculate the effective diffusion coefficient provided it exists ( i.e. the system homogenizes at large scales ) .
for random resistor networks the homogenization of conductivity is mathematically proved for local conductivities bounded from above and from below , see @xcite and references therein .
the boundness from below excludes the conductor - isolator percolation model , but homogenization still holds _ provided _ the system percolates @xcite .
physically , it is known that the conductor - superconductor system homogenizes below the percolation threshold for superconductor @xcite .
we mimic a stationary experiment on measuring the diffusion coefficient via the first fick s law : the system , in a form of a membrane of thickness @xmath31 and of transversal dimension @xmath32 separates two reservoirs , the left one with a well - stirred solution of particles at concentration @xmath33 , the other one with a slightly lower concentration @xmath34 .
the membrane is considered as impermeable for the solvent , and both concentrations are kept constant .
the constant particles current @xmath35 through the membrane is measured and connected with the mean diffusion coefficient inside it .
since in general a jump of the ( free ) energy per particle can form on a contact between the membrane and the solution ( e.g. when the fluid is a good solvent for diffusing particles and the membrane is , on the average , a bad one , or other way around ) the effective diffusion coefficient _ inside _ the membrane has to be defined through @xmath36 where @xmath37 and @xmath38 are the mean particle concentrations in the layers of the membrane in immediate contact with the solution , see fig.1 . in a stationary state @xmath39 . moreover , in the thermodynamical limit @xmath40 the permeability of the membrane and thus the current tend to zero .
we will call this situation `` quasi - equilibrium '' in what follows .
the contact with solution is modelled by additional arrays of sites to the left and to the right from the membrane , with constant particles concentrations and constant energies @xmath41 which can be chosen arbitrarily ( @xmath41 defines the quality of the solvent ) .
these additional sites are connected to the ones on the membrane s sides via extremely high transition rates fulfilling the detailed - balance condition . in this case a local equilibrium between the surface sites and the solutions persists independently on the particles distribution inside the bulk . due to this the activities of the surface sites
are all equal to @xmath42 and @xmath43 at the left resp .
right boundary of the membrane , where the prefactor @xmath44 depends on @xmath41 .
thus , @xmath45 in the leftmost layer are proportional to @xmath46 and in the rightmost layer to @xmath47 so that the mean concentrations in the layers are @xmath48 and @xmath49 , where we assume that the distribution of the site energies in the surface layers is the same as in the bulk .
we then calculate the corresponding total current @xmath35 through the system noting that the equations for the currents and activities in a stationary state are the same as the ones given by the kirchhoff s laws for an electric circuit .
making such a reinterpretation we see that @xmath50 where @xmath51 is the effective conductance ( conductivity of a bond in the _ effective ordered medium _ with the same total conductivity as our heterogeneous one ) , where the subscript em denotes the effective medium mean .
therefore @xmath52 .
the prefactor @xmath53 is introduced to restore the dimension as follows from eq .
( [ exp ] ) , when passing from distances @xmath31 measured in lattice units to distances measured in centimeters .
we note that since @xmath23 are proportional to the boltzmann factors , and since rescaling of all @xmath17 by a constant factor leads to changing @xmath54 by the same factor , the proportionality factor @xmath25 cancels out ; this gives eq.([maineq ] ) .
the result is rather transparent .
if we are able to measure the effective conductivity of the system , we can connect it with the effective mobility @xmath55 ( and thus with the diffusion coefficient ) via nernts - einstein equation @xmath56 , where @xmath57 is the equilibrium concentration of particles with charge @xmath58 . reverting this expression
we get @xmath59 , which is essentially eq.([maineq ] ) .
one may argue that the correct way is to define @xmath60 through the gradient of the coarse - grained concentration , and not via the total concentration difference . as we show in the appendix
, this definition leads to the same result since local concentrations and local activities decouple under quasi - equilibrium ( but _ only _ under this condition ! ) .
let us first discuss some applications of eq.([maineq ] ) other than discussed in the preface .
eq.([maineq ] ) gives the possibility to obtain the universal bounds on the effective diffusion coefficient based on those for the effective conductance , i.e. the universal wiener bounds @xcite and the tighter hashin - shtrikman bounds for isotropic systems @xcite , as well as to generalize some exact results for two - dimensional systems based on duality @xcite . in this presentation
we concentrate on continuum models where eq.([meq0 ] ) arises from discretization of a fokker - planck equation for @xmath61 : @xmath62 with constant diffusion coefficient @xmath63 and disordered potential @xmath64 .
the details of calculations are given in the appendix .
the universal wiener bounds for the conductance are given by @xmath65 , in our cases this corresponds to @xmath66 note that the lower bound reproduces the exact result for the one - dimensional system with random potential and constant diffusion coefficient . in figs .
[ fig : binhs ] and [ fig : homohs ] we plot the hashin - shtrikman bounds for two cases : the case of the binary disorder @xmath67 with probability @xmath68 and @xmath69 with probability @xmath70 , and the case of @xmath0 possessing an exponential distribution with cutoffs , the one with density @xmath71^{-1}$ ] for @xmath72 and vanishing elsewhere ( apart from cutoffs this distribution is reminiscent of the exponential energy distributions leading to ctrws ) . as a comparison ,
the results of the effective medium approximation ( ema ) , see refs .
@xcite , are shown .
the results are plotted as the function of a contrast @xmath73 $ ] being the ratio of the maximal and the minimal value of @xmath17 .
[ cols="^ " , ] in the limit of very strong disorder @xmath60 may vanish or diverge . in the first case @xmath74 , the system either does not show any transport ( does not percolate ) or shows anomalous transport slower than diffusion ( i.e. shows _ subdiffusiom _ ) . in the second case
it might show superdiffusion .
if @xmath60 vanishes , it can do so either because the numerator @xmath75 vanishes or because the denominator @xmath76 diverges , as well as in the cases when both possibilities are realized simultaneously ( which may give rise to subdiffusion of mixed origins @xcite ) .
we discuss the conditions under which the corresponding behavior may take place . if @xmath60 diverges , it can do so because the numerator diverges , or because the denominator vanishes , or both .
as we proceed to show , none of these possibilities can be realized .
this excludes not only superdiffusion , but also the `` compensated '' cases of normal diffusion when both numerator and denominator vanish or diverge simultaneously . for further discussion
we first recapitulate the following properties of percolation systems : ( i ) the mixture of resistors with given finite conductivity ( at concentration @xmath68 ) with insulating bonds ( of zero conductivity ) at concentration @xmath70 possesses zero conductance below the percolation threshold @xmath77 and finite conductance above it .
the corresponding system homogenizes at scales above the correlation length @xcite .
this homogenization also takes place for arbitrary distribution of the conductivities of the resistors @xcite .
similarly , ( ii ) the mixture of resistors with given finite conductivity ( at concentration @xmath68 ) with superconducting bonds ( at concentration @xmath70 ) possesses finite conductance below the percolation threshold @xmath78 for superconducting bonds , with @xmath79 , and infinite conductance above it .
these properties do hold not only for the bernoulli percolation model but also in the case when the short - range correlations in the occupation probabilities of the bond by the corresponding resistors / insulators / superconductors are present .
this statement is a ( silently assumed ) basis of all renormalization group approaches in percolation . using the results of the theory of electric circuits ,
see e.g. @xcite , we show that the total conductivity of a resistor network is a non - decaying function of the conductivity of each particular bond .
let us consider the system as placed between two `` superconducting '' bars considered as a terminal 1 of the system .
let us consider the poles @xmath3 and @xmath7 between which @xmath17 is switched as terminal 2 .
using the theory of two - terminal circuits we calculate the input impedance ( total conductivity ) @xmath80 as a function @xmath17 : @xmath81 , where @xmath82 are the elements of the impedance matrix of the system . for a system of reciprocal passive elements ( no batteries , no diodes )
this matrix is non - negatively definite and symmetric as a consequence of non - negative heat production and of reciprocity theorem . in the case of pure resistor network
the matrix is real .
thus , @xmath83 and @xmath84 , so that @xmath80 is a non - decaying function of @xmath17 .
now we show that the numerator never diverges .
we fix some @xmath85 and declare the fraction @xmath58 of bonds ( starting from the ones with largest @xmath86 ) to be superconductive .
the lowest conductivity of a changed bond is @xmath87 .
the superconducting bonds are non - percolating by construction , and the conductance of the remaining system is finite , being smaller than a conductance of the resistor - superconductor mixture where all conductivities are put to @xmath87 .
thus the numerator can only diverge if @xmath88 , i.e. never in finite dimension .
the numerator does not vanish for a system with percolation concentration @xmath89 .
let us remove a portion @xmath90 of bonds with smallest @xmath86 without destroying percolation and denote the largest removed conductivity by @xmath91 .
the rest of the system percolates and has a conductance which is larger then the conductance of a two - phase system constructed of resistors with @xmath92 and @xmath93 , which is nonzero since we are above percolation threshold .
thus the numerator can only vanish if @xmath94 , and no bonds can be removed . the denominator in eq.([maineq ] ) can diverge if the corresponding mean value of the boltzmann factor diverges .
since @xmath95 is proportional to the sojourn time at a site @xmath3 in equilibrium , this corresponds to diverging mean sojourn time at a site , i.e. to a trap model , which in high dimensions is equivalent to ctrw with a broad distribution of waiting times .
the denominator can not vanish .
let @xmath96 be the probability density of @xmath0 , and @xmath97 its median , @xmath98 .
since @xmath99 where the integrand is non - negative , we have @xmath100 since @xmath101 is monotonically decaying . summarizing our findings we state that there exists an exact correspondence between the effective diffusion coefficient in a random potential and macroscopic conductivity in a random resistor model .
this simple relation allows us to obtain exact bounds on the effective diffusion coefficient .
it also allows for elucidating possible sources of anomalous diffusion in such model .
thus , the subdiffusion is possible either if the mean boltzmann factor of the corresponding potential diverges ( energetic disorder ) or if the percolation concentration in a system is equal to unity , i.e. if the system is already at the percolation threshold , in one dimension , or on finitely ramified fractals ( structural disorder ) , and that superdiffusion is impossible in our system under any condition . *
acknowledgement .
* the work was supported by bmu within the project `` zuverlssigkeit von pv modulen ii '' . |
as it is well known , magnetic vortex states are experimentally observed in ferromagnetic disk - shaped nanostructures .
these topological objects exhibit a planar - like configuration of spins outside the core , where a perpendicular magnetization ( up or down polarization ) is observed @xcite . as long as one could manipulate these vortices
, several possibilities would emerge .
therefore , studies taking into account the application of external potentials such as static and sinusoidal magnetic fields are nowadays very common in the literature @xcite .
for instance , the simplest effect induced by an external field is the gyrotropic mode , which is the lowest excitation of the vortex structure .
this mode is simply the elliptical vortex core motion ( with the resonance frequency ) around the disk center .
the sense of gyration in an elliptic trajectory ( clockwise or counterclockwise ) is determined only by the vortex core polarization .
external potentials such as magnetic fields or ( d.c . ) spin - polarized currents can also stimulate dramatic effects such as the switching behavior @xcite .
however , to the authors knowledge , no internal mechanisms able to induce the core switching were reported neither experimentally nor theoretically .
such a chance would only occur if the vortex could interact with possible inhomogeneities present in the nanostructure .
for instance , a possibility of triggering this process could be obtained by removing some small portions of the magnetic nanodisk , in such a way that the cavities ( antidots ) so created work by attracting and eventually affecting the vortex structure @xcite .
recently , rahm and coworkers @xcite have experimentally studied the cases of one , two , three and four antidots artificially inserted in a disk with diameter @xmath0 and separated by around @xmath1 .
their results not only confirmed the previous statement about vortex pinning around the hole defects @xcite as well as have put forward the possibility of using these stable states as serious candidates for magnetic memory and logical applications @xcite .
although experimental results are provided for such systems , it is worthy noticing that a broad theoretical analysis is still lacking . in this paper
, we argue that defects present in the dots may induce interesting possibilities for the magnetization dynamics , including the switching behavior of vortex structures .
this is a very interesting topic and open new possibilities for technologies and experiments .
indeed , a very recent work @xcite has shown that magnetization dynamics and evolution is deeply modified when cavities are introduced .
their samples consist of co - made elongated ring - shaped nanomagnets , which like permalloy has negligible magnetostatic anisotropy .
what is clearly observed in these experiments is that the reversal pathway of the so - called diagonal onion state is drastically affected by the inserted holes once they attract and sometimes pin the transient vortex - like domain - wall configurations . actually , not only the presence , but also the sizes and locations of the cavities ( two or four in their hollowed samples ) remarkably change magnetization evolution and eventually lead to its reversal .
we will follow a different line in which the magnetization reversal takes place due to the effects produced by the hole on the vortex core motion .
no further agent is necessary in the switching process , like occur to other usual methods where external fields are one of the main ingredients . before starting our analysis
, it would be useful to describe how a magnetic vortex can arise in nanodisks .
indeed , a small revision will help to justify the magnetic model used here .
the vortex state in magnetic nanodot materials is the result of the competition between exchange and magnetostatic interactions .
particularly , in finite systems like magnetic dots , densities of magnetic charges " are induced in their volumes and surfaces .
considering the magnetization @xmath2 , the volumetric and superficial densities are defined as @xmath3 and @xmath4 respectively , where @xmath5 is the unit vectors normal to the surfaces of the dot at each point .
therefore , in order to minimize the magnetostatic energy , the magnetization tends to point parallel to the dot surfaces and it is one of the properties responsible for the formation of the vortex ground state .
however , as long as we go towards the vortex center , exchange energy density gradually increases like @xmath6 .
then , in order to regularize exchange contribution , the magnetic moments around the center tend to revolve , developing out - of - plane projection , so that exactly at the vortex center it is perpendicular to the disk face , whose direction defines what is called vortex polarization ( @xmath7 , if the centered moment points _ up _ , or @xmath8 , if it is pointing _ down _ ) . here
, we study very thin ferromagnetic nanodisks with thickness @xmath9 and radius @xmath10 so that the aspect ratio @xmath11 . in this case
, one can assume @xmath12 and hence , the only source of magnetostatic interactions is the superficial magnetic charges . to write the vortex state in these dots
, it is convenient to parametrize the magnetization @xmath13 by two scalar fields , the polar @xmath14 and azimuthal @xmath15 angles , @xmath16 . using @xmath17 ,
the vortex state configuration can be written as @xmath18 where the magnetization unit vector deviates out of the plane at the vortex core @xmath19 and outside this region , @xmath20 .
the vortex core size is approximately the exchange length @xmath21 ( which can be taken to be of the order of the lattice spacing @xmath22 in our simulations on a discrete lattice ) . in this notation ,
the values of @xmath23 at the vortex center defines the vortex polarization . on the other hand
, @xmath24 defines the vortex chirality .
the flux closed in - plane vortex ( @xmath24 ) and the perpendicular component in the center ( @xmath23 ) are independent of each other so that four different magnetization states of one disk are possible .
+ our paper is organized as follows : in section ii , the model and associated methods of numerical analysis are presented and discussed .
section iii is devoted to the study of sample with a single hole .
there , we emphasize the vortex core reversal process induced by its interaction with the defect and the non - ordinary circumstances favoring it .
the case of doubly hollowed disk is studied in section iv , where differently from the former case , no holes configuration yields the core switching : indeed , it is always captured by one of the defects before any reversal be observed in our simulations .
we finally close our work by point out our conclusions and prospects for forthcoming investigation .
+ trajectory of the vortex core ( beginning at the disk center @xmath25 ) , in the @xmath26-plane , after turning off the external field . here
, no hole is inserted and the arrows indicate the direction of the motion . ]
in order to describe a thin magnetic nanodisk we pursue an alternative strategy substituting the disk by a two - dimensional film with magnetization vectors distributed in a regular square lattice inside a circumference of radius @xmath10 .
in addition , the magnetostatic interactions due to the presence of the magnetic charges in the lateral and top surfaces of the dot are replaced by local potentials .
since we are going to study systems with one and two holes and also with no hole , the model used is summarized by the following general hamiltonian : @xmath27 + where @xmath28 for sites inside the holes and @xmath29 for the remaining sites of the film . here , @xmath30 is the unit spin vector at position @xmath31 ( @xmath32 is the saturation magnetization ) , the sum @xmath33 is over nearest - neighbor spins , the terms with positive constants @xmath34 mimic the magnetostatic energies ( see below ) in the top face of the disk , @xmath35 , while @xmath36 accounts for the lateral edge . at the holes borders the analogous constants taking into account the surface charges at the edges are @xmath37 and @xmath38 , for holes @xmath39 and @xmath40 respectively . for the case without defects ,
@xmath41 while for the situation with a single defect @xmath42 . in turn
, @xmath43 is an external magnetic field .
the hole defects are introduced by removing a number of neighbor spins around sites @xmath44 and @xmath45 from the system ( @xmath46 and @xmath47 ) .
of course , the number of neighbor spins removed around a particular position ( @xmath48 or @xmath49 ) defines the hole sizes @xmath50 and @xmath51 .
now , the magnetostatic energy favors the spins to be parallel to the film surfaces .
therefore , the local unit vectors @xmath5 defined on the points characterizing the surfaces of the material are always perpendicular to these surfaces ( or boundaries ) : @xmath52 is perpendicular to the disk face , in the xy - plane ; @xmath53 is radially perpendicular to the circumference envelop , with radius @xmath10 , at each point ; @xmath54 and @xmath55 are radially perpendicular to the border points of holes @xmath39 ( pointing to @xmath48 ) and @xmath40 ( directed to @xmath49 ) , respectively .
as we have already mentioned , in principle , our model , hamiltonian ( [ exchange - magnetostatic - holes ] ) , could describe a very thin disk with aspect ratio @xmath11 .
in fact , several known experimental results with nanodisks can be qualitatively reproduced using it . for instance , the vortex core precession mode ( which has a rather low eigenfrequency @xcite ) can be easily obtained , at least qualitatively , with this simple model ( see below ) .
however , it is not well known how defects can change this picture .
defects must be naturally present @xcite or artificially inserted @xcite in the film samples .
it has been predicted theoretically @xcite and also observed experimentally @xcite that small holes incorporated into the magnetic structures can attract and capture magnetic vortices . on the other hand
, we argue that another effect can occur : the vortex core switching due to a involved dynamical state of the magnet , which should include the coherent magnetization oscillations ( near the defect ) matched to the vortex motion .
actually , a somewhat similar effect has been observed in quite recent experiments @xcite .
there , the pathway developed to reverse the diagonal onion state in elongated nanorings is quite sensitive to intentionally inserted cavities , as well as their sizes and locations . here
, the picture comprises a vortex - like configuration with polarization , say , up , which after interacting with a hole , of a given size at a specified position ( these values clearly depend on other parameters like the whole nanodisk size , applied field amplitude for resonantly exciting the gyrotropic mode , etc ) , turns it to be down , eventually reversing the sense of its gyrotropic motion .
what should be stressed is that such a remarkable phenomenon show up in our simulations only for very special situations involving the physical parameters , for instance , the hole must be relatively small ( around 5%-10% of the sample size ) .
in addition , the insertion of a second similar hole in any position deeply jeopardize the magnetization evolution and the vortex core is always captured by one of the holes before any core reversal process .
magnetization in the @xmath56-direction as a function of time and its fourier transform . ]
our results are obtained by using spin dynamics simulations for a square lattice occupying all possible points inside a circumference of radius @xmath10 ( in most simulations we have used @xmath57 ) .
it is solved by employing the fourth - order predictor - corrector method .
the calculations presented here consider only @xmath58 , @xmath59 and @xmath60 .
the choice of other values of @xmath35 does not alter the essential physics if it is not large enough ( in all investigations we have used the range @xmath61 . for @xmath62 , the vortex becomes essentially planar and does not develop the out - of - plane components at the core @xcite .
on the other hand , we have verified only a limited range of @xmath36 .
indeed , the interval @xmath63 leads to the same quantitative and qualitative results . in order to excite the vortex core precession mode ,
an external perturbation was applied . here
, we have chosen a sinusoidal external magnetic field of the form @xmath64 , with @xmath65 and @xmath66 .
these are the best conditions we have identified in our simulations favoring the core reversal .
for instance , large changes in these field parameters will not produce the main phenomena we intend to describe . in all cases ,
the duration of the sinusoidal field was @xmath67 and the total time of the simulations was @xmath68 with time increment equal to @xmath69 .
firstly , in the absence of the defects and after turning the field off , we observe the precession mode with a circular trajectory , as expected ( see , fig .
[ orbita1 ] ) . for a better analysis of this mode
, we have plotted the behavior of the magnetization in the @xmath56-direction @xmath70 as a function of time and its fourier transform ( see fig . [ mxsb ] ; the same behavior can be seen for @xmath71 ) .
the fourier transform presents a peak at a well defined frequency @xmath72 , which is the frequency of the vortex core precession ( the resonance frequency , see fig .
[ cxcysb ] ) . of course , it depends on the size @xmath10 of the disk .
more realistically , the resonance frequency depends on the aspect ratio @xmath73 .
however , since our model assumes a very thin disk , we can , in principle , obtain @xmath74 as a function of @xmath73 only for @xmath75 .
we also studied the behavior of @xmath74 as a function of @xmath76 , which appears to be essentially linear , at least for @xmath77 ( see fig .
[ fxr ] ) , in qualitatively accordance with experimental findings @xcite .
all above comparisons ( and others not presented here ) show that this model works very well .
it gives us a perspective that hamiltonian ( [ exchange - magnetostatic - holes ] ) is also able to predict new facts .
position components ( @xmath78 ) of the vortex center as function of time .
note that the fourier transform leads to a peak at the same frequency observed in fig.[mxsb ] .
this is the resonance frequency of the gyrotropic motion ] the behavior of the gyrotropic frequency , @xmath79 , as function of the ` aspect ratio ' , @xmath80 , obtained from the present model .
points are simulational results while the line is the linear fitting . ] before studying the problem of interest , we have also plotted the behavior of the @xmath81-component of the magnetization as function of time ( and its fourier transform ) in the absence of a defect ( see fig .
[ mzsb ] ) . in this case , the average magnetization over the whole sample oscillates around a small positive value ( because the out - of - plane spins forming the vortex core are pointing along the positive @xmath81-direction ) all the time without abrupt changes , indicating that there is not any magnetization reversal during the vortex core precession , as expected .
in addition , the main peak in the fourier transform occurs at a frequency @xmath82 , which is different from @xmath74 .
@xmath83 versus @xmath84 and its fourier transform .
note that @xmath83 oscillates around a positive value . ]
this model involves several parameters and , of course , it is very hard to treat in details all aspects .
therefore we have exposed here the general behavior of the vortex core dynamics and the main particular phenomena found for specific conditions .
we notice that a hole , independently of its size and position , in general , attracts the vortex core to its center , i.e. , the core almost always falls into the hole ( sometimes , scattering - like events may occur ) .
some theoretical @xcite and experimental works have already reported the capture process @xcite .
hence , it would be interesting to consider other possibilities for the vortex dynamics in the presence of such type of defects .
two situations are analyzed in details .
the first one , which is more remarkable , describes the vortex core switching due to the vortex - defect interaction .
it happens only for very special circumstances and our work indicates the most suitable configurations , in order to aid possible experimental probing .
the another one relies on the fate of the core to be trapped by one of the two holes ( two configurations that yield this capture are presented ) .
such a situation also serves to illustrate how a relatively small perturbation eliminates the magnetization reversal and how the center of the vortex oscillates after a capture process .
now we consider some effects that a hole may induce on the vortex core dynamics .
when a small hole is inserted in the circular film , the vortex core suffers new influences so that the magnetization dynamics should become much richer .
indeed , holes induce an attractive effective potential for the vortex @xcite and therefore , such force may compete with the effects of applied fields , magnetostatic , anisotropies , etc .
+ typical trajectory of the vortex core in a disk containing a hole ( at position @xmath85 ) .
the vortex core motion starts in the disk center @xmath25 and the precession mode is excited by a field pulse .
the comparison of this motion with that shown in fig .
[ orbita1 ] clearly shows the effect of the hole on vortex core dynamics . ]
@xmath70 versus @xmath84 and its fourier transform for a film with a hole . ] magnetization in the @xmath81-direction and its fourier transform .
the vortex - hole interaction causes the switching . ] in this section we consider the presence of a single hole of radius @xmath86 . for the external field
, we use the parameters of the previous section . in a nanodisk with a cavity
, one expects that magnetic charges are also induced in the cavity walls analogously as they appear in the usual surfaces of the disk .
therefore , taking into account only the presence of defect @xmath39 at position @xmath48 , in principle we should use @xmath87 , which corresponds to the influences of the magnetostatic energy coming from the edge of the hole cavity . here
we will show only influences of very small defects as compared with the disk size and consequently , @xmath37 could be neglected in a first approximation since the hole wall can support only a tiny amount of magnetic charges .
more realistic analysis should then take this parameter to be small , for example , as compared with @xmath36 .
however , it is illustrative to study the system for a large range of @xmath37 values , say , @xmath88 , for elucidating its effects on the whole properties of the magnetization . generally , for the smallest hole considered ( four neighbor spins removed , @xmath89 ) , the vortex core is almost always captured ( or sometimes scattered ) by the defect for @xmath90
hence , large enough contributions of magnetostatic energy coming from the cavity wall ( i.e. , large @xmath37 ) leads to the common phenomena of capture , or even the more seldom scattering - like effect , observed for larger holes . indeed , larger holes always capture the vortex core , independently of @xmath37 ( recall that besides magnetostatic , exchange gets lower as the hole size increases ) , what agrees with experimental@xcite and theoretical@xcite results , which consider defects occupying an appreciable fraction of the disk .
therefore , if one wishes looking for different possibilities , perhaps small defects should be the focus . with this strategy
, we first analyse several parameter possibilities and later we shall consider @xmath91 and use @xmath92 ( or , for simplicity , @xmath93 ) for all subsequent calculations . indeed , the most important phenomenon we would like to describe here happens only for @xmath94 .
physically , it is reasonable to think that a small cavity wall has only a tiny contribution to the magnetostatic energy .
the size and position of the hole can be controlled and so we have studied several cases . for example , for a central hole with size @xmath22
, the resonance frequency of the gyrotropic mode is given by @xmath95 , which is smaller than the one we have observed for the usual nanodisk @xmath72 ( a fact qualitatively predicted by analytical treatment @xcite ) .
the problem with this configuration is that the ground state is a vortex pinned to the hole ( of course , without an out - of - plane core ) and , therefore , the vortex needs to be previously released from the cavity to exhibit the gyrotropic mode .
changes in the defect size and position lead to the following general situations : depending on the initial conditions ( external sinusoidal fields , essentially ) , then , as the vortex core moves sufficiently close the defect border , it can be either transmitted or captured by the cavity .
these are interesting effects due to core - hole interaction but , for very special features of the film , a much more rare dynamical event may be triggered .
actually , choosing the coordinate system in which the disk center is put at @xmath25 and if and only if we use @xmath96 ( or @xmath97 ) for the center position ( along the @xmath98-axis at @xmath99 ) of a hole with size @xmath100 ( four neighbor spins removed ) , then a fabulous phenomenon occurs : the magnetic vortex core is reversed .
for all disk sizes studied , only one configuration was found to be able of reversing the core magnetization .
always , the hole must be relatively small and placed near enough the disk border , perpendicularly to the field . particularly , for @xmath58 and @xmath101 ,
the asymmetric conditions with @xmath102 seems to be the unique defect position possible to trigger the switching process .
it should be emphasized that the switching of magnetic domains generally depends on a myriad of detailed features of the magnetic particles , and topological effects ultimately limit this possibility .
therefore , the occurrence of such a delicate dynamical effect should not be expected for any ordinary configuration of the magnet with a hole .
usually , what is observed is the capture of the core when the defect is large enough ; conversely , only small changes in its dynamics are observed for very small defects .
the main lesson is : the core reversal seems to demand relatively small hole ( in our case around @xmath103 of the nanodisk dimension ) combined with other favorable events , like its location and suitable applied oscillating field .
another important point is that , assuming @xmath104 and for special conditions and parameters ( like @xmath105 and @xmath106 used here ) , the reversal of the magnetization of the vortex core ( along the @xmath81-direction ) can be always reproduced . in fig.[orbita2 ] we show the trajectory of the vortex core . note that this mode starts with the same frequency of the case without a hole ( fig .
[ orbita1 ] ) in a trajectory outer the hole .
however , after some revolutions , when the vortex core moves again very near the hole border ( in a region in between the hole and disk walls ) , it quickly changes the direction of the motion simultaneously with the magnetization reversal along the @xmath81-direction ( see video of the vortex core motion and the switching phenomenon as auxiliary material @xcite ) .
really , a change in the sense of gyration unambiguously indicates a change in the vortex core polarization .
then the vortex core again develops a new precession mode with a smaller frequency @xmath107 as shown in fig .
[ mx ] ( an orbit with radius smaller than @xmath108 ) .
an appreciable part of the kinetic energy of the vortex was transformed in spin waves . to see more clearly the switching behavior ,
the z - component of the average magnetization @xmath83 is plotted in fig .
[ mz ] , which can be compared with fig .
note the rapid change of the mean value of @xmath83 from positive ( when the core direction was up ) to negative ( when the core is down ) . in the fourier transform
, the main peak happens in almost the same frequency for the case without a hole .
( color online ) the instant immediately before the switching process .
the out - of - plane magnetization fluctuations can be perceived in the irregularities present around the green surface .
note the deep protuberance with negative values of @xmath109 near and below the hole and the core .
such a protuberance will pull the vortex core , leading to the inversion of its polarity . ]
thus , the lacking of experimental observation of core switching induced by vortex - hole interaction may be credited to the large holes generally inserted into the samples , around @xmath110 of its size . in these cases
the common phenomenon is the capture of the core by the defect @xcite ) .
basic physics concepts may help to understand this even better : as it is known the effective potential ( exchange + magnetostatic ) experienced by the vortex is globally minimized inside the hole @xcite .
conversely , if the hole is very small , it only slightly affects vortex dynamics and no appreciable effects are expected to occur .
actually , the phenomenon of core reversal predicted here takes place for a hole of intermediary size , situation in which the minima provided by the exchange and magnetostatic contributions are comparable . in summary , our simulations as a whole agree with experimental observations upon vortex core dynamics in the presence of intentionally inserted cavities : generally , the core is captured by one of the defects , mainly when they are sufficiently large , like in the experiments of refs.@xcite . only for special circumstances the core reversal is expected to occur , demanding among other a non - centered relatively small hole .
another peculiarity of core reversal induced by vortex - hole interaction is the fact that the polarization switching occur only from internal dynamical events : the vortex must be only under motion ( the gyrotropic is the more appropriate for it is easily reproducible ) and pass suitably close to the defect , so that the attraction from the hole ultimately dominates its inertia , reversing its motion sense
. this in turn , demands the switching in its polarization , so that the sign of the gyrotropic factor in the landau - lifshitz - gilbert equation , @xmath111 is kept unaltered .
another dynamical explanation for this phenomenon is : firstly , the vortex motion excites spin - wave modes .
differently from the case without defects , the presence of the hole makes these modes with larger amplitudes due to reflections in its border , mainly when the vortex core moves near it .
secondly , for a magnet of finite size , one expects that the radiation of spin - waves , their reflection off the disk ( and hole ) boundaries and their effect back on the vortex core will result in the establishment of a dynamical state of the magnet which includes both the confined moving vortex and the coherent magnetization oscillations matched to the vortex motion .
then , under this special situation , strong out - of - plane spin fluctuations arise near the hole border , changing coherently the direction of the vortex core magnetization when it passes close to the defect . figure [ ricardo3d ] shows an instant of the core motion immediately before the magnetization reversal .
our aim in this section is to consider the effect of two holes on the vortex core dynamics in circular thin films .
also , this study will be useful for showing that the reversal magnetization phenomenon is completely suppressed if a second defect , even small and distant from the first , is introduced into the system .
in addition , the pinned vortex dynamical behavior is detailed . like the former case
, the disk radius is selected to be @xmath58 .
( color online ) three - dimensional view of fig .
( [ bidimensional1 ] ) .
the vortex polarization is up ( yellow - red peak ) .
note the out - of - plane fluctuations generated by the core motion .
these spin waves propagate through the system and are reflected in the disk and holes borders.,width=340 ] mean magnetization @xmath112 as a function of time and its fourier transform .
the resonance frequency is approximately @xmath113 .
this figure shows clearly that the vortex core is captured by one hole at @xmath114.,width=264 ] mean magnetization @xmath115 as a function of time and its fourier transform.,width=264 ] the results are obtained by choosing an external sinusoidal field to excite the gyrotropic mode .
again , the field is applied for a short time ( @xmath116 ) and its parameters are @xmath65 and @xmath117 . of course
, the vortex motion will depend on the holes size and position as well as the external field . in all observed cases ,
the vortex finishes trapped in one of the antidots just it finds the defect in its way , even for smaller holes .
since this is the common phenomenon for all possible positions of the defects , we intend to know the dynamical process after the trapping . to illustrate this , we choose the following geometry : hole @xmath39 at position @xmath118 and hole @xmath40 at @xmath119 as shown in figs .
( [ bidimensional1 ] ) and ( [ tridimensional2 ] ) .
larger defects will capture the core even faster .
what should be stressed is that the presence of the second defect , even if smaller and distant of the first , will change the vortex dynamics drastically so that no core switching have been observed before its capture by one of the holes .
figure ( [ magx ] ) shows the mean magnetization @xmath112 along the @xmath56-direction as a function of time . as the vortex
core moves , this mean magnetization oscillates almost harmonically from @xmath120 to @xmath121 , indicating that the motion is approximately circular ( or elliptical ) around the disc center . the orbit is outer the holes .
however , at about @xmath122 , the core decreases considerably its velocity near the hole @xmath39 , moves slowly in spiral ( outer the antidot @xmath40 and inner antidot @xmath39 ) , until hitting antidot @xmath40 again , where it is captured at about @xmath123 .
now , the sense of gyration is not changed and consequently , there is no switching process .
the @xmath112 magnetization oscillations stop abruptly becoming almost constant around @xmath124 ( actually , @xmath112 oscillates very rapidly with a very small amplitude around @xmath124 ) .
we also have calculated the fourier transform of @xmath112 ( see fig.([magx ] ) ) .
it has two main peaks : the first one refers to the oscillating external field ( along the @xmath56-direction ) and the second is related to the resonance frequency . in our unit system ,
such resonance frequency is given by @xmath125 , which can be compared with the analogous problem of a film without defects @xmath126 .
similar graphics are obtained for @xmath115 as shown in fig.([magy ] ) .
the difference is that , after being pinned , the mean @xmath98-magnetization oscillates around zero ( with a relatively large amplitude as compared to its counterpart in @xmath56 ) and not around a finite value as happens to @xmath112 . in addition , there is no the first large peak exhibited in the @xmath112 .
the reasons are simple : first , the holes were displayed in the @xmath98-axis ( at @xmath127 and @xmath128 ) and second , the external sinusoidal field ( responsible for the first peak in the fourier transform of @xmath112 ) is applied along the @xmath56-direction and hence , it can not appear in @xmath115 .
the trajectory of the vortex core is shown in fig.([orbitarod ] ) . based on the above description ,
an interesting fact to report is that the vortex center oscillates around the hole center with small amplitude and large frequency when the vortex is pinned to the defect .
this phenomenon looks like the one theoretically predicted in layered ferromagnetic materials @xcite .
nevertheless , this effect must be very richer in nanodisks because the surface magnetostatic energy has an effective contribution , which is absent in the layered systems .
analytical estimate @xcite predicts that such oscillations are related to the vortex mass and take place around @xmath129 ghz , which is far beyond current possibilities of observation in these systems @xcite , @xmath130ghz .
however , a study considering all microscopic details of the pinned vortex oscillation will require further investigation .
+ in our present analysis , the oscillation frequency of this mode can be seen in the fourier transform of @xmath112 or @xmath115 and is given by @xmath131 .
it is much higher than the resonance frequency , and much lower than mass - like vibrations , as above .
such an intermediary frequency could be somewhat related to the _ fractional gyrotropic _ mode , which is predicted to occurs around @xmath132 and is provided by a remanent ( fractional ) gyrotropic vector of the vortex core , partially captured by the hole@xcite .
another interesting feature can be easily seen when one compares the fourier transforms of @xmath112 and @xmath115 .
note that , after being captured , the magnetization amplitude oscillations are larger in the @xmath98-direction than in the @xmath56-direction .
it means that the pinned vortex center oscillates much more along the @xmath56-axis than along the @xmath98-axis .
therefore , it describes an ellipse with larger semi - axis along the @xmath56-direction , of the order of the hole size . for a trapped vortex
, the difference in the amplitudes of oscillation of @xmath115 and @xmath112 increases as the hole responsible for the capture is placed nearer the circumference border of the film .
hence , this ellipse turns out to be an almost straight line if the hole is located very near the disk boundary , while it takes a circular shape if the cavity is centered in the disk .
therefre , several procedures of production and propagation of spin waves with different phases could be built only by changing the defect position , which is easily manipulated by lithographic techniques .
thus , our results indicate that the vortex - hole system could also works as a mechanism to create and control spin waves in confined magnetic thin films . the complete orbit of the vortex core until capture by hole @xmath40 .
the vortex starts its motion in the disk center at @xmath25 .
the symbols @xmath133 and @xmath134 indicate the positions of antidots @xmath39 and @xmath40 respectively . in the studied process
, the core follows an almost perfect gyrotropic mode , completing five laps around the film before feeling a strong effect of hole @xmath39 , which changes its motion direction.,width=283,height=283 ] mean magnetization @xmath135 as a function of time and its fourier transform.,width=283 ] we have also plotted the mean magnetization along the @xmath81-direction , @xmath135 , and its fourier transform , fig.([magz ] ) .
there , we clearly realize that , before the capture process , @xmath135 oscillates around a small positive value , once the vortex is _ up_-polarized ( negative if the polarity were _ down _ ) .
after some vortex core laps , at about @xmath122 , the amplitude of oscillation increases considerably .
this is just the moment the vortex core motion suffers a strong disturb ( near hole @xmath39 ) and starts to slow down , changing the orbit until @xmath136 , when the core is captured by hole @xmath40 .
therefore , the kinetic energy of the core decreases considerably and a large amount of spin waves is released during the instant of the capture process .
the pinned vortex oscillations also generate more and more spin waves due to the rapid rotations of vortex center .
the amplitude of @xmath135 remains large after trapping but now its oscillations are quite small , meaning that the pinned vortex polarization practically vanishes .
( color online ) a picture of the instant that the vortex core was captured by hole @xmath40 .
note the large amount of spin waves generated in this process .
such waves can be seen by the out - of - plane fluctuations given by irregularity in the spin pattern ( green surface ) on the film.,width=302 ] + for the sake of completeness , we now discuss a situation where the holes are not aligned along the same line crossing the disk center .
let us consider a disk with radius @xmath137 centered at position @xmath25 with two small holes ( @xmath138 ) displayed at @xmath85 and @xmath139 .
figure ( [ orbita2holes ] ) shows the orbit of the vortex core until its capture by one of the holes .
even in this situation ( many others , with the holes closer and apart from each other were also tested ) the main facts remain the same : after being excited by the applied sinusoidal field the vortex core starts its gyrotropic motion around the disk center . whenever passing near a hole its dynamics changes ( similarly to the former case studied , say , both holes centered along @xmath98-axis , as depicted in figures [ tridimensional2 ] , [ magx ] , and [ magy ] ) , and after some revolutions it is eventually captured by one of the defects .
analogously to the previous configuration with two holes , no core reversal was observed in our simulations , neither for smaller ( only @xmath39 spin removed ) or larger ( @xmath140 spins removed ) defects .
however , we can not strictly state that no other sizes could trigger core reversal ( with two or more holes ) , once their sizes in our simulations can be only largely varied ( keeping its circular shape ) , say , while the smallest hole is obtained by removing a unique spin , the next possible size demands we take 4 spin away , and so forth . the main point must be recalled : if core reversal could somewhat occur with two or more holes , the simplest system where this phenomenon would certainly be easiest and clearest observed is that comprising a nanodisk with a single non - centered relatively small hole . the orbit of the vortex core in a nanodisk with radius @xmath58 centered at @xmath25 containing two holes at positions @xmath141 and @xmath142 .
as observed for several geometries with two defects , the fate of the core is to be captured by one of the holes . ]
in this paper we have considered an alternative model for circular magnetic thin films containing artificial defects and we have also predicted an interesting way of reversing the magnetization of the vortex core which is completely dependent on its dynamics and interaction with a relatively small hole , inserted in the sample .
therefore , it should be stressed that the present switching process is not stimulated by external agents , like usually occur by means of the application of magnetic or spin - polarized dc current fields .
the conditions necessary are summarized by the following properties : first , only one hole must be present and localized at a position near the lateral border of the disk ; we suggest that the center of the hole must be placed at a distance on the order of the vortex core size away from the disk border .
second , the hole must be small and third , the field required to excite the gyrotropic mode should be applied perpendicularly to the line joining the disk and hole centers.[although the field is an external agent , we emphasize once more that its role is restrict for resonantly excite the gyrotropic mode , putting the vortex core in motion .
such a field has no direct influence on the reversal mechanism , which is completely triggered by vortex - hole interaction ] .
clearly , the suitable parameters of the applied magnetic field , relative hole size , etc , should be probed in actual experiments . + as we observed in our simulations , those special conditions force the core to move in a circular trajectory passing exactly in between the hole and disk borders generating a large amount of spin waves , which will also develop important role in the switching process ( for example , see fig.([ricardo3d ] ) .
nowadays , such defects were already intentionally introduced in nanostructures @xcite but they occupy a relatively large fraction of the material ( about @xmath143-@xmath144 percent ) . hence , the switching mechanism occurring during the vortex motion can be met only by further miniaturization of defects in nanostructures . also in this line
, we should mention the very recent experiment@xcite where hole defects induce magnetization reversal in elongated @xmath145 rings .
the results are aimed at the same direction of our work and are an experimental indication that our proposal may have further and broad relevance . an important difference between our prediction and
the phenomenon experimentally observed in ref.@xcite is that the last needs a permanent presence of an external magnetic field ( while in our case the mechanism of the vortex switching is based only on internal interactions during the vortex motion ) . in this aspect ,
the dynamical characteristic of the phenomenon predicted here is another differential with relation to the usual core reversals induced by external potentials . to achieve more details about the vortex - hole interactions ,
we have also considered a system with two holes . in this case
no switching process was observed for several configurations analyzed . since one of the holes
will trap the vortex , we have then addressed our attention to the magnetization dynamics after the capture .
when the vortex center falls into a hole , some of its energy is radiated away in the form of spin waves .
the vortex center oscillates inside the hole with a frequency much larger than the resonance frequency .
the bound state vortex - hole releases a large quantity of spin waves forming coherent magnetization oscillations matched to the vortex structure ( part of the kinetic energy of the core lost in the capture is transformed in spin waves ) .
these results demonstrate a significant coupling between the vortex ( with the core or even captured , without the core ) and spin waves in a disk .
+ to conclude we would like to suggest that the switching mechanism reported here can be used in technological applications . actually , this phenomenon naturally lends itself to applications in binary data storage .
therefore , it seems that our investigations may provide fundamentally new ways of using magnetic nanostructures in technology .
+ the authors thank cnpq , fapemig and capes ( brazilian agencies ) for financial support .
t. shinjo , t. okuno , r. hassdorf , k. shigeto , and t. ono , science * 289 * , 930 ( 2000 ) .
j. miltat and a. thiaville , science * 298 * 555 ( 2002 ) .
a. wachowiak , j. wiebe , m. bode , o. pietzsch , m. morgenstern , and r. wiesendanger , science * 298 * , 577 ( 2002 ) .
r. pulwey , m. rahm , j. biberger , and d. weiss , ieee trans .
* 37 * , 2076 ( 2001 ) .
kravchuk , d.d .
sheka , y. gaididei , and f.g .
mertens , j. appl .
phys . * 102 * , 043908 ( 2007 ) .
xiao , j. rudge , e. girgis , j. kolthammer , b.c .
choi , y.k .
hong , and g.w .
donohoe , j. appl .
* 102 * , 103904 ( 2007 ) .
k. yamada , s. kasai , y. nakatan , k. kobayashi , h. kohno , a. thiaville , and t. ono , nat .
* 6 * , 270 ( 2007 ) .
v. novosad , f.y .
fradin , p.e .
roy , k.s .
buchanan , k. yu .
guslienko , and s.d .
bader , phys .
b * 72 * , 024455 ( 2005 ) .
k. yu guslienko , b.a .
ivanov , v. novosad v , y. otani , h. shima , and k. fukamichi , j. appl .
phys . * 91 * , 8037 , ( 2002 ) .
t. uhlig , m. rahm , c. dietrich , r. hllinger , m. heumann , d. weiss , and j. zweck , phys .
lett . * 95 * , 237205 ( 2005 ) .
paula , a.r .
pereira , and l.a.s .
ml , phys . lett .
* a329 * , 155 ( 2004 ) .
m. rahm , j. stahl , w. wegscheider , and d. weiss , app .
* 85 * , 1553 ( 2004 ) .
m. rahm , j. stahl , and d. weiss , app .
lett . * 87 * , 182107 ( 2005 ) .
gao , a.o .
adeyeye , and c.a .
ross , j. appl .
phys.*103 * , 063906 ( 2008 ) .
guslienko , x.f .
han , d.j .
keavney , r. divan , asn s.d .
barder , phys .
lett . * 96 * , 067205 ( 2006 ) .
compton and p.a .
crowell , phys .
* 97 * 137202 ( 2006 ) .
m. rahm , j. biberger , v. umansky , and d. weiss , j. appl . phys . * 93 * , 7429 ( 2003 ) .
pereira , s.a .
leonel , p.z .
coura , and b.v .
costa , phys .
b * 71 * , 014403 ( 2005 ) . g.m .
wysin , phys .
b * 49 * 8780 ( 1994 ) .
pereira , a.r .
moura , w.a .
moura - melo , d.f .
carneiro , s.a .
leonel , and p.z .
coura , j. appl . phys .
* 101 * , 034310 ( 2007 ) .
wysin , phys .
b * 71 * , 094423 ( 2005 ) .
k. kuepper , l. bischoff , ch .
akhmadaliev , j. fassbender , h. stoll , k.w .
chou , a. puzic , k. fauth , d. dolgos , g. schtz , b. van waeyenberge , t. tyliszczak , i. neudecker , g. woltersdorf , and c.h .
back , appl .
lett . * 90 * , 062506 ( 2007 )
. w. a. moura - melo , a. r. pereira , r. l. silva , and n.m .
oliveira - neto , j. app .
( 2008 ) , in press .
see epaps document for video of the vortex core motion around the hole and subsequent switching process .
k. kpper , private communication . | we investigate the influence of artificial defects ( small holes ) inserted into magnetic nanodisks on the vortex core dynamics .
one and two holes ( antidots ) are considered . in general
, the core falls into the hole but , in particular , we would like to remark an interesting phenomenon not yet observed , which is the vortex core switching induced by the vortex - hole interactions .
it occurs for the case with only one hole and for very special conditions involving the hole size and position as well as the disk size .
any small deformation in the disk geometry such as the presence of a second antidot changes completely the vortex dynamics and the vortex core eventually falls into one of the defects . after trapped , the vortex center still oscillates with a very high frequency and small amplitude around the defect center .
= 3.4 in |
stars experience mass loss in the form of winds during their lives and , as the star evolves , its wind changes characteristics . for some stars , at certain evolutionary phases ,
the amount of mass lost through stellar winds can amount to a significant portion of its own mass .
this is the case , for example , of supergiant stars , which present massive winds that are believed to be mainly driven by radiation pressure on dust grains ( for the red supergiants ) or on spectral lines ( for the blue supergiants ) .
on the other hand , for cool stars at the main sequence phase , we believe their winds are significantly less massive , carrying only a small amount of mass compared to the total mass of the star .
although these winds are quite rarefied , they carry a significant amount of angular momentum , which affects the stellar rotational evolution .
p - cygni profiles , the traditional mass - loss signatures observed in the denser winds of giant and supergiant stars , are not formed in the rarefied winds of cool dwarf stars . to detect these winds ,
other indirect methods have been proposed ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
so far , the method developed by @xcite has been the most successful one , enabling estimates of mass - loss rates for about a dozen cool dwarf stars . due to the lack of observational constraints on winds of cool low - mass stars , many works assume these winds to be identical to ( or a scaled version of ) the solar wind . in the next section , i highlight a few observed properties of cool dwarf stars that suggest that their winds can actually be significantly different from the solar wind .
we believe the solar coronal heating is ultimately caused by the sun s magnetism , although it is still debatable which mechanism heats the solar corona . in the stellar case ,
coronal heating is much less understood .
stellar coronae can be much hotter than the solar corona .
the temperature of the x - ray emitting coronae of solar analogs can exceed @xmath1 mk , more than one order of magnitude larger than the solar coronal temperature of @xmath2 mk . if the temperature of the x - ray emitting ( closed ) corona is related to the temperature of the stellar wind ( flowing along open field lines ) , as one would naively expect , then we may expect stellar winds of cool dwarf stars to have temperatures that could be much larger than the solar wind temperature . because the terminal velocities of thermally - driven winds are very sensitive to the choice of temperature of the stellar wind @xcite , the winds of cool dwarf stars might have a variety of terminal velocities . thanks to our privileged position immersed in the solar wind , we have access to a great quantity of data that allow a detailed understanding of the physics that is operating in the sun . measurements of mass flows and magnetic field of the solar wind have revealed an asymmetric solar wind . in particular , ulysses measurements showed that the solar wind structure depends on the characteristics of the solar magnetic field @xcite .
the solar wind characteristics change along the solar cycle , presenting a simple bimodal structure of fast and slow flows during solar minimum , when the geometry of the solar magnetic field is closest to that of an aligned dipole . at solar maximum ,
when the the axis of the large - scale dipole becomes nearly perpendicular to the solar axis of rotation , the solar wind shows a more complex structure .
although the richness of details of the magnetic field configuration is only known for our closest star , modern techniques have made it possible to reconstruct the large - scale surface magnetic fields of other stars .
the zeeman - doppler imaging ( zdi ) technique is a tomographic imaging technique that allows us to reconstruct the large - scale magnetic field ( intensity and orientation ) at the surface of the star from a series of circular polarisation spectra .
this method has now been used to investigate the magnetic topology of stars of different spectral types and has revealed fascinating differences between the magnetic fields of different stars .
for example , solar - type stars that rotate about two times faster than our sun show the presence of substantial toroidal component of magnetic field , a component that is almost non - existent in the solar magnetic field @xcite .
the magnetic topology of low - mass ( @xmath3 ) very active stars seem to be dictated by interior structure changes : while partly convective stars possess a weak non - axisymmetric field with a significant toroidal component , fully convective ones exhibit strong poloidal axisymmetric dipole - like topologies @xcite .
in addition to magnetic field characteristics , the symmetry of a stellar wind also depends on the rotation rate of the star . in the presence of fast rotation
, a magnetised wind can lose spherical symmetry , as centrifugal forces become more important with the increase of rotation rate @xcite .
the distribution of rotation periods in cool main - sequence stars is very broad , ranging from stars rotating faster than once per day to indefinitely long periods
. the variety of observed rotation rates , intensities and topologies of the magnetic fields of cool , dwarf stars indicate that their winds might come in different flavours and might be significantly different from the solar one .
one might also bear in mind that , similarly to the sun , low - mass stars are also believed to host magnetic and activity cycles ( see fares contribution , this volume ) , which imply that the characteristics of their winds can also vary in a time scale of the cycle periods .
there are two most commonly used ways to model stellar winds , which i summarise below . 1
. one approach consists of computing the detailed energetics of the wind , starting from the photosphere , passing through the chromosphere , until it reaches the stellar corona . depending on the physics that is included in such models , this approach can become computationally expensive . as a result , it has been limited to analytical , one- and two - dimensional solutions .
in addition , it is also usually focused in the inner most part of the corona and usually only adopts simplified magnetic field topologies .
the second approach commonly used to model winds of cool stars consists of adopting a simplified energy equation , usually assuming the wind to be isothermal or described by a polytropic equation of state ( in which thermal pressure @xmath4 is related to density @xmath5 as @xmath6 ) . in this case
, one is allowed to perform multi - dimensional numerical simulations of stellar winds and can incorporate more complex magnetic field topologies because of the simplified energetics that are considered in the second approach , its domain can extend considerably farther out than the first one . as a result , polytropic winds can be useful in the characterisation of the interplanetary medium and also to characterise interactions between exoplanets and the winds of their host - stars . in the next section ,
i present how to make more realistic stellar wind simulations by incorporating recent insights acquired on the magnetic topology of different stars into stellar wind models . to illustrate how observationally reconstructed surface maps ( see section [ sec.magnetism ] )
can be incorporated in the simulations of stellar winds , i will present the work done in @xcite , where we performed numerical simulations of the stellar wind of the planet - hosting star @xmath7 boo ( spectral type f7v ) .
@xmath7 boo is a remarkable object , not only because it hosts a giant planet orbiting very close to the star ( located at 0.046 au from the star ) , but also because it is the only star other than the sun for which a full magnetic cycle has been reported in the literature @xcite .
these observations suggest that @xmath7 boo undergoes magnetic cycles similar to the sun , but with a cycle period that is about one order of magnitude smaller than the solar one ( about 2 years as opposed to 22 years for the solar magnetic cycle ) .
the surface magnetic maps of @xmath7 boo reconstructed by @xcite , @xcite and @xcite were used as boundary conditions for the stellar wind simulations .
a potential field extrapolation is adopted at the initial state of the simulation ( fig .
[ fig.sim_conditions]a ) . as the simulation evolves in time ,
stellar wind particles and magnetic field lines are allowed to interact with each other .
figure [ fig.sim_conditions]b shows the self - consistent solution found for the magnetic field lines , after the wind solution relaxed in the grid .
note that the magnetic field lines become stressed , wraping around the rotational axis of the star ( pointing towards positive @xmath8 ) .
colour - coded are the reconstructed large - scale surface field of @xmath7 boo at jun/2006 ( fig . [ fig.sim_conditions]a ) and the wind velocity at the equatorial plane of the star ( fig . [ fig.sim_conditions]b ) .
@xcite found that variations of the stellar magnetic field during the cycle directly influence the outflowing wind .
therefore , the rapid variation of the large - scale magnetic field of @xmath7 boo implies that the environment surrounding the close - in planet should be varying quite rapidly .
in addition , @xcite estimated the mass - loss rate ( @xmath9 ) of @xmath7 boo , showing that this star seems to have a denser wind than that of the sun , with @xmath9 that are 2 orders of magnitude larger than the solar value @xmath10 ( @xmath11 ) .
when the wind outflows from the star , it permeates the entire extrasolar system , interacting with any body that it encounters on its way .
the interaction between stellar winds and exoplanets can lead to observable signatures , some of which are absent in our own solar system . in the solar system , the giant planets and
the earth emit at radio wavelengths .
such planetary radio emission is due to the interaction between the magnetic planets and the solar wind , and its power is proportional to the stellar wind energy dissipated in the wind - planet interaction . by analogy to what is observed in the solar system ,
it is expected that exoplanets also interact with the winds of their host stars and should , therefore , also generate radio emission . because the energy dissipated by the stellar wind
is larger at closer distances to the star ( because the wind density and magnetic fields are significantly larger than further out from the star ) , radio emission of close - in planets , such as @xmath7 boo b , is expected to be several orders of magnitude larger than the emission from the planets in the solar system .
in addition , the relatively dense wind of @xmath7 boo estimated in @xcite implies that the energy dissipated in the stellar wind - planet interaction can be significantly higher than the values derived in the solar system .
combined with the close proximity of the system ( @xmath12 pc ) , the @xmath7 boo system has been one of the strongest candidates to verify exoplanetary radio emission predictions .
using the detailed stellar wind model developed for its host - star , @xcite estimated radio emission from @xmath7 boo b , exploring different values for the _ assumed _ planetary magnetic field .
for example , they showed that , for a planet with a magnetic field similar to jupiter s ( @xmath13 g ) , the radio flux is estimated to be @xmath14 mjy , occurring at an emission frequency of @xmath15 mhz .
although small , this emission frequency lies in the observable range of current instruments , such as lofar . to observe such a small flux , an instrument with a sensitivity lying at a mjy level
is required . the same estimate was done considering the planet has a magnetic field similar to the earth ( @xmath16 g ) . although the radio flux does not present a significant difference to what was found for the previous case , the emission frequency ( @xmath17 mhz ) falls at a range below the ionospheric cut - off , preventing its possible detection from the ground . in fact , due to the ionospheric cutoff at @xmath18mhz , radio detection with ground - based observations from planets with magnetic field intensities @xmath19 g should not be possible @xcite . despite many attempts ,
radio emission from exoplanets has not been detected so far .
such a detection would not only constrain local characteristics of the stellar wind , but would also demonstrate that exoplanets are magnetised . fortunately
, there may be other ways to probe exoplanetary magnetic fields , in particular for transiting systems , through signatures of bow shocks during transit observations . based on hubble space telescope / cosmic origins spectrograph ( hst / cos ) observations using the narrow - band near - uv spectroscopy
, @xcite showed that the transit lightcurve of the close - in giant planet wasp-12b presents both an early ingress when compared to its optical transit , as well as excess absorption during the transit , indicating the presence of an asymmetric distribution of material surrounding the planet .
motivated by these transit observations , @xcite suggested that the presence of bow shocks surrounding close - in planets might lead to transit asymmetries at certain wavelengths , such as the one observed in wasp-12b at near - uv wavelengths . the main difference between bow shocks formed around exoplanets and the ones formed around planets in the solar system is the shock orientation , determined by the net velocity of the particles impacting on the planet s magnetosphere . in the case of the earth , the solar wind has essentially only a radial component , which is much larger than the orbital velocity of the earth . because of that , the bow shock surrounding the earth s magnetosphere forms facing the sun .
however , for close - in exoplanets that possess high orbital velocities and are frequently located at regions where the host star s wind velocity is comparatively much smaller , a shock may develop ahead of the planet . in general , we expect that shocks are formed at intermediate angles ( see also * ? ? ?
* ; * ? ? ?
due to their high orbital velocities , close - in planets offer the best conditions for transit observations of bow shocks .
if the compressed shocked material is able to absorb stellar radiation , then the signature of bow shocks may be observed through both a deeper transit and an early - ingress in some spectral lines with respect to the broadband optical ingress @xcite .
the sketch shown in figure [ fig.sketch2 ] illustrates this idea . , which can be estimated by measuring the difference between the phases ( @xmath20 ) at which the near - uv and the optical transits begin .
, scaledwidth=90.0% ] in the case of wasp-12b , @xcite suggested that the shocked material , which is able to absorb enough stellar radiation in the near - uv , can cause an asymmetry in the lightcurve observed ( see lightcurve sketches in figure [ fig.sketch2 ] ) , where the presence of compressed material ahead of the planetary orbit causes an early ingress , while the lack of compressed material behind the planetary orbit causes simultaneous egresses both in the near - uv transit as well as in the optical one .
this suggestion was verified by @xcite , who performed monte carlo radiation transfer simulations of the near - uv transit of wasp-12b .
they confirmed that the presence of a bow shock indeed breaks the symmetry of the transit lightcurve an interesting outcome of the observations of bow shocks around exoplanets is that it permits one to infer the magnetic field intensity of the transiting planet . by measuring the difference between the phases
at which the near - uv and the optical transits begin ( @xmath20 illustrated in figure [ fig.sketch2 ] ) , one can derive the stand - off distance from the shock to the centre of the planet , which is assumed to trace the extent of the planetary magnetosphere @xmath21 .
at the magnetopause , pressure balance between the coronal total pressure and the planet total pressure requires that @xmath22 ^ 2}{8\pi } + p_c= \frac{[b_{p}(r_m)]^2}{8\pi } + p_{p } , \ ] ] where @xmath23 , @xmath24 and @xmath25 are the local coronal mass density , thermal pressure , and magnetic field intensity at orbital radius @xmath26 , and @xmath27 and @xmath28 are the planet thermal pressure and magnetic field intensity at @xmath21 . the relative velocity between the material surrounding the planet and the planet itself is @xmath29 . in the case of a magnetised planet with a magnetosphere of a few planetary radii , the planet total pressure is usually dominated by the contribution from the planetary magnetic pressure ( i.e. , @xmath30 ) .
@xcite showed that , in the case of wasp-12b , eq . ( [ eq.equilibrium ] ) reduces to @xmath31 . further assuming that stellar and planetary magnetic fields are dipolar , we have @xmath32 where @xmath33 and @xmath34 are the magnetic field intensities at the stellar and planetary surfaces , respectively .
( [ eq.bplanet ] ) shows that the planetary magnetic field can be derived directly from _ observed _ quantities . for wasp-12 , using the upper limit of @xmath35 g @xcite and the stand - off distance obtained from the near - uv transit observation @xmath36 @xcite
, we predicted an upper limit for wasp-12b s planetary magnetic field of @xmath37 g. in theory , the suggestion that through transit observations one can probe the planetary magnetic field is quite straightforward - all it requires is a measurement of the transit ingress phase in the near - uv . in practice , however , acquisition of near - uv transit data requires the use of space - borne facilities , making follow - ups and new target detections rather difficult . in order to optimise target selection , @xcite presented a classification of the known transiting systems according to their potential for producing shocks that could cause observable light curve asymmetries .
the main considered assumption was that , once the conditions for shock formation are met , planetary shocks absorb in certain near - uv lines , in a similar way as wasp-12b .
in addition , for it to be detected , the shock must compress the local plasma to a density sufficiently high to cause an observable level of optical depth .
this last hypothesis requires the knowledge of the local ambient medium that surrounds the planet . by adopting simplified hypotheses ,
namely that up to the planetary orbit the stellar corona can be treated as in hydrostatic equilibrium and isothermal , @xcite predicted the characteristics of the ambient medium that surrounds the planet for a sample of @xmath38 transiting systems , and discussed whether such characteristics present favourable conditions for the presence and detection of a bow shock .
excluding systems that are quite far ( @xmath39 pc ) , the planets that were top ranked are : wasp-19b , wasp-4b , wasp-18b , corot-7b , hat - p-7b , corot-1b , tres-3 , and wasp-5b .
as the wind outflows from the star , it permeates the interplanetary medium , interacting with any planet encountered on its way .
the proper characterisation of stellar winds is therefore crucial to constrain interactions between exoplanets and their surrounding environments and also essential for the study of space weather events on exoplanets .
stellar winds are affect by the stellar rotation , magnetism and coronal temperature , properties that can vary significantly from star to star . as a consequence , stellar winds of cool stars
can actually be significantly different from the solar wind . in this talk ,
i illustrated how one can take an extra step towards more realistic models of stellar winds of low - mass stars by incorporating observationally reconstructed surface magnetic maps into simulations of stellar winds .
i also showed that dramatic differences in stellar magnetism and orbital radius can make the interplanetary medium of exoplanetary systems remarkably distinct from the one present in the solar system .
in addition , i showed that the interaction of the stellar winds with exoplanets can lead to observable signatures that are absent in our own solar system .
aav acknowledges support from a royal astronomical society fellowship and thanks the iau for a travel support grant to attend the symposium . , j. , morin , j. , petit , p. , delfosse , x. , forveille , t. , aurire , m. , cabanac , r. , dintrans , b. , fares , r. , gastine , t. , jardine , m. m. , lignires , f. , paletou , f. , velez , j. c. r. , and thado , s. : 2008a , , 545 , l. , haswell , c. a. , froning , c. s. , hebb , l. , holmes , s. , kolb , u. , helling , c. , carter , a. , wheatley , p. , cameron , a. c. , loeillet , b. , pollacco , d. , street , r. , stempels , h. c. , simpson , e. , udry , s. , joshi , y. c. , west , r. g. , skillen , i. , and wilson , d. : 2010b , , l222 , j. , donati , j. , petit , p. , delfosse , x. , forveille , t. , albert , l. , aurire , m. , cabanac , r. , dintrans , b. , fares , r. , gastine , t. , jardine , m. m. , lignires , f. , paletou , f. , ramirez velez , j. c. , and thado , s. : 2008 , , 567 | the great majority of exoplanets discovered so far are orbiting cool , low - mass stars whose properties are relatively similar to the sun .
however , the stellar magnetism of these stars can be significantly different from the solar one , both in topology and intensity .
in addition , due to the present - day technology used in exoplanetary searches , most of the currently known exoplanets are found orbiting at extremely close distances to their host stars ( @xmath0 au ) .
the dramatic differences in stellar magnetism and orbital radius can make the interplanetary medium of exoplanetary systems remarkably distinct from that of the solar system . to constrain interactions between exoplanets and their host - star s magnetised winds and to characterise the interplanetary medium that surrounds exoplanets , more realistic stellar wind models , which account for factors such as stellar rotation and the complex stellar magnetic field configurations of cool stars , must be employed .
here , i briefly review the latest progress made in data - driven modelling of magnetised stellar winds .
i also show that the interaction of the stellar winds with exoplanets can lead to several observable signatures , some of which that are absent in our own solar system . |
many situations in the recent research in quantum games @xcite appear to be based on a general idea that is quite interesting as well .
it is to take a classical game exhibiting certain features , generalize it to quantum domain , and see how the situation changes in the course of this generalization .
in this course noncooperative games have attracted an earlier attention with the ruling solution concept of a nash equilibrium ( ne ) .
this development looks reasonable because in classical game theory as well the earlier research was focused on noncooperative games and interest in coalition formation was revived later .
players in noncooperative games are not able to form binding agreements even if they may communicate . on the other hand
the distinguishing feature of cooperative games is a strong incentive to work together to receive the largest total payoff .
these games allow players to form coalitions , binding agreements , pay compensations , make side payments etc .
in fact , von neumann and morgenstern @xcite in their pioneering work in the theory of games offered models of coalition formation where the strategy of each player consists of choosing the coalition he wishes to join . in coalition games , that are part of cooperative game theory , the possibilities of the players are described by the available resources of different groups ( coalitions ) of players . joining a group or remaining outside is part of strategy of a player affecting his / her payoff .
recent work in quantum games @xcite gives rise to a natural and interesting question : what is the possible quantum mechanical role in cooperative games that are an important part of the classical game theory ? in our opinion it may be quite interesting , and fruitful as well , to investigate coalitions in quantum versions of cooperative games .
our motivation in present paper is to investigate what might happen to the advantage of forming a coalition in a quantum game compared to its classical analogue .
we rely on the concepts and ideas of von neumann s cooperative game theory @xcite and consider a three - player coalition game in a quantum form .
we then compare it to the classical version of the game and see how the advantage of forming a coalition can be affected . in usual classical analysis of the coalition
games the notion of a strategy disappears ; the main features are those of a coalition and the value or worth of the coalition .
the underlying assumption is that each coalition can guarantee its members a certain amount called the `` value of a coalition '' @xcite__. _ _ _ the value of coalition measures the worth the coalition possesses and is characterized as the payoff which the coalition can assure for itself by selecting an appropriate strategy , whereas the ` odd man ' can prevent the coalition from getting more than this amount_. using this idea we study cooperative games in quantum settings to see how advantages of making coalitions can be influenced in the new settings .
the preferable scheme to us to play a quantum game has been recently proposed by marinatto and weber @xcite . in this scheme
an initial quantum state is prepared by an arbiter and forwarded to the players .
each player possesses two quantum unitary and hermitian operators i.e. the identity @xmath0 and the inversion or pauli spin - flip operator @xmath1 .
players apply the operators with classical probabilities on the initial quantum state and send the quantum state to the ` measuring agent ' who decides the payoffs the players should get .
interesting feature in this scheme is that the classical game is reproduced when the initial quantum state becomes unentangled @xcite .
classical game is therefore embedded in the quantum version of the game . in this paper using marinatto and weber s scheme @xcite we find a quantum form of a symmetric cooperative game played by three players . in classical form of this game
any two players out of three get an advantage when they successfully form a coalition and play the same strategy .
we find a quantum form of this game where the advantage for coalition forming is lost and players are left with no motivation to cooperate .
a classical three person normal form game @xcite is given by three non - empty sets @xmath2 , @xmath3 , and @xmath4 , the strategy sets of the players @xmath5,@xmath6 , and @xmath7 and three real valued functions @xmath8
, @xmath9 , and @xmath10 defined on @xmath11 .
the product space @xmath11 is the set of all tuples @xmath12 with @xmath13 , @xmath14 and @xmath15 .
a strategy is understood as such a tuple @xmath12 and @xmath8 , @xmath9 , @xmath10 are payoff functions of the three players .
the game is usually denoted as @xmath16 .
let @xmath17 be the set of players and @xmath18 be an arbitrary subset of @xmath19 .
the players in @xmath18 may form a coalition so that , for all practical purposes , the coalition @xmath18 appears as a single player .
it is expected that players in @xmath20 will form an opposing coalition and the game has two opposing `` coalition players '' i.e. @xmath18 and @xmath20 .
we study quantum version of an example of a classical three player cooperative game discussed in ref .
each of three players @xmath21 and @xmath7 chooses one of the two strategies @xmath22 .
if the three players choose the same strategy there is no payoff ; otherwise , the two players who have chosen the same strategy receive one unit of money each from the ` odd man . '
payoff functions @xmath8 , @xmath9 and @xmath10 for players @xmath23 and @xmath7 respectively are given as @xcite @xmath24 with similar expressions for @xmath9 and @xmath10 .
suppose @xmath25 ; hence @xmath26 .
the coalition game represented by @xmath27 is given by the following payoff matrix @xcite @xmath28{lllll}% & & $ \left [ 1\right ] $ & & $ \left [ 2\right ] $ \\ $ \left [ 11\right ] $ & & $ 0 $ & & $ 2$\\ $ \left [ 12\right ] $ & & $ -1 $ & & $ -1$\\ $ \left [ 21\right ] $ & & $ -1 $ & & $ -1$\\ $ \left [ 22\right ] $ & & $ 2 $ & & $ 0$% \end{tabular}\ ] ] here the strategies @xmath29 $ ] and @xmath30 $ ] are dominated by @xmath31 $ ] and @xmath32 $ ] .
after eliminating these dominated strategies the payoff matrix becomes @xmath28{lll}% & $ \left [ 1\right ] $ & $ \left [ 2\right ] $ \\ $ \left [ 11\right ] $ & $ 0 $ & $ 2$\\ $ \left [ 22\right ] $ & $ 2 $ & $ 0$% \end{tabular}\ ] ] it is seen that the mixed strategies @xmath33 + \frac{1}{2}\left [ 22\right ] \label{cltc}\\ & \frac{1}{2}\left [ 1\right ] + \frac{1}{2}\left [ 2\right ] \label{lftc}%\end{aligned}\ ] ] are optimal for @xmath18 and @xmath20 respectively . with these strategies a payoff @xmath34 for players @xmath18
is assured for all strategies of the opponent ; hence , the value of the coalition @xmath35 is @xmath34 i.e. @xmath36 . since @xmath37 is a zero - sum game @xmath35 can also be used to find @xmath38 as @xmath39 .
the game is also symmetric and one can write @xmath40 in quantum form of this three player game the players implement their strategies by applying the identity operators in their possession with probabilities @xmath41 and @xmath42 respectively on the initial quantum state . in marinatto and weber s
scheme @xcite the pauli spin - flip or simply the inversion operator @xmath1 is then applied with probabilities @xmath43 and @xmath44 by players @xmath5 , @xmath6 and @xmath7 respectively .
if @xmath45 is the density matrix corresponding to initial quantum state the final state after players have played their strategies corresponds to @xcite @xmath46 where the unitary and hermitian operator @xmath47 can be either @xmath0 or @xmath1 .
@xmath48 , @xmath49 and @xmath50 are the probabilities with which players @xmath5 , @xmath6 and @xmath7 apply the operator @xmath47 on the initial state respectively .
@xmath51 corresponds to a convex combination of all possible quantum operations .
let the arbiter prepares the following three qubit pure initial quantum state @xmath52 where the eight basis vectors of this quantum state are @xmath53 for @xmath54 .
the initial state ( [ instate ] ) can be imagined as a global state ( in a @xmath55 dimensional hilbert space ) of three two - state quantum systems or ` qubits ' .
a player applies the unitary operators @xmath0 and @xmath1 with classical probabilities on @xmath45 during his ` move ' or ` strategy ' operation . fig .
1 shows the scheme to play this three player quantum game where players @xmath6 and @xmath7 form a coalition and player @xmath5 is ` leftout ' .
[ ptb ] let the matrix of three player game be given by @xmath56 constants @xmath57 with @xmath58 @xcite .
we write the payoff operators for players @xmath59 and @xmath7 as @xcite @xmath60 payoffs to players @xmath59 and @xmath7 are then obtained as mean values of these operators @xcite @xmath61\ ] ] where , for convenience , we identify the players moves only by the numbers @xmath62 and @xmath42 . the cooperative game of eq .
( [ payoffs ] ) with the classical payoff functions @xmath8 , @xmath9 and @xmath10 for players @xmath5 @xmath6 and @xmath7 respectively , together with the definition of payoff operators for these players in eq .
( [ payoper ] ) , imply that @xmath63 with these constants the payoff to player @xmath5 , for example , can be found as @xmath64{c}% -4rq-2p+2pr+2pq+r+q\\ -4rq+2p-2pr-2pq+3r+3q-2\\ 4rq+2pr-2pq-3r - q+1\\ 4rq-2pr+2pq - r-3q+1 \end{array } \right ] \left [ \begin{array } [ c]{c}% \left| c_{111}\right| ^{2}+\left| c_{222}\right| ^{2}\\ \left| c_{211}\right| ^{2}+\left| c_{122}\right| ^{2}\\ \left| c_{121}\right| ^{2}+\left| c_{212}\right| ^{2}\\ \left| c_{112}\right| ^{2}+\left| c_{221}\right| ^{2}% \end{array } \right ] \label{poff}%\ ] ] similarly payoffs to players @xmath6 and @xmath7 can be obtained .
classical mixed strategy payoffs can be recovered from the eq .
( [ poff ] ) by taking @xmath65 .
the classical game is therefore imbedded in its quantum form .
the classical form of this game is symmetric in the sense that payoff to a player depends on his / her strategy and not on his / her identity .
these requirements making symmetric the three - player game are written as @xmath66 now in this quantum form of the game @xmath67 becomes same as @xmath68 when @xmath69 and then payoff to a @xmath70 player remains same when other two players interchange their strategies .
the symmetry conditions ( [ rqmnts ] ) hold if , together with eqs .
( [ rqmnts1 ] ) , following relations are also true @xmath71{cc}% \alpha_{1}=\beta_{1}=\gamma_{1 } , & \alpha_{5}=\beta_{6}=\gamma_{7}\\ \alpha_{2}=\beta_{3}=\gamma_{4 } , & \alpha_{6}=\beta_{5}=\gamma_{6}\\ \alpha_{3}=\beta_{2}=\gamma_{3 } , & \alpha_{7}=\beta_{7}=\gamma_{5}\\ \alpha_{4}=\beta_{4}=\gamma_{2 } , & \alpha_{8}=\beta_{8}=\gamma_{8}% \end{array}\ ] ] these form the extra restrictions on the constants of payoff matrix and , together with the conditions ( [ rqmnts1 ] ) , give a three player symmetric game in a quantum form .
no subscript in a payoff expression is then needed and @xmath72 represents the payoff to a @xmath70 player against two other players playing @xmath73 and @xmath42 .
the payoff @xmath72 is found as @xmath74 the term ` mixed strategy ' in the quantum form of this game is defined as being a convex combination of quantum strategies with classical probabilities . for this assume that the pure strategies @xmath75
$ ] and @xmath76 $ ] correspond to @xmath77 and @xmath78 respectively .
the mixed strategy @xmath79 + ( 1-n)\left [ 2\right ] $ ] , where @xmath80 , means that the strategy @xmath81 $ ] is played with probability @xmath82 and @xmath83 $ ] with probability @xmath84 .
now suppose the coalition @xmath18 plays the following mixed strategy @xmath85+(1-l)[22 ] \label{cltq}%\ ] ] where the strategy @xmath86 $ ] means that both players in the coalition @xmath18 apply the identity operator @xmath0 with zero probability
. similarly the strategy @xmath87 $ ] can be defined .
the strategy of the coalition in eq .
( [ cltq ] ) means that the coalition @xmath18 plays @xmath86 $ ] with probability @xmath88 and @xmath87 $ ] with probability @xmath89 .
similarly we suppose the player in @xmath20 plays following mixed strategy @xmath90+(1-m)[2 ] \label{lftq}%\ ] ] in this case the payoff to the coalition @xmath18 is obtained as @xmath91}+l(1-m)p_{\wp\lbrack112]}+\nonumber\\ & ( 1-l)mp_{\wp\lbrack221]}+(1-l)(1-m)p_{\wp\lbrack222 ] } \label{cltpq}%\end{aligned}\ ] ] where @xmath92}$ ] is the payoff to @xmath18 when all three players play @xmath77 i.e. the strategy @xmath75 $ ] .
similarly @xmath93}$ ] is coalition payoff when coalition players play @xmath78 and the player in @xmath20 plays @xmath77 .
now from eq .
( [ qpayoff ] ) we get @xmath94 } & = 2p(0,0,0)=-4(\left| c_{211}\right| ^{2}+\left| c_{122}\right| ^{2})\nonumber\\ p_{\wp\lbrack112 ] } & = 2p(0,0,1)=2(\left| c_{111}\right| ^{2}+\left| c_{222}\right| ^{2}+\left| c_{211}\right| ^{2}+\left| c_{122}\right| ^{2})\nonumber\\ p_{\wp\lbrack221 ] } & = 2p(1,1,0)=2(\left| c_{111}\right| ^{2}+\left| c_{222}\right| ^{2}+\left| c_{211}\right| ^{2}+\left| c_{122}\right| ^{2})\nonumber\\ p_{\wp\lbrack222 ] } & = 2p(1,1,1)=-4(\left| c_{211}\right| ^{2}+\left| c_{122}\right| ^{2})\end{aligned}\ ] ] therefore from eq .
( [ cltpq ] ) @xmath95 to find the value of coalition @xmath35 in the quantum game we find @xmath96 and equate it to zero i.e. @xmath97 is such a payoff to @xmath18 that the player in @xmath20 can not change it by changing his / her strategy given in eq .
( [ lftq ] ) .
it gives , interestingly , @xmath98 and the classical optimal strategy of the coalition @xmath99 + \frac{1}{2}\left [ 22\right ] $ ] becomes optimal in the quantum game as well . in the quantum game
the coalition then secures following payoff , also termed as the value of the coalition @xmath100 similarly we get the value of coalition for @xmath20 as @xmath101 note that these values reduce to their classical counterparts of eq .
( [ vcltc ] ) when the initial quantum state becomes unentangled and is given by @xmath102 .
classical form of the coalition game is , therefore , a subset of its quantum version .
suppose the arbiter now has at his disposal a quantum state @xmath103 such that @xmath104
. in this case @xmath35 becomes a negative quantity and @xmath105 because of the normalization given in eq .
( [ instate ] ) .
a more interesting case is when the arbiter has the state @xmath106 at his disposal . because now both @xmath35 and @xmath38 are @xmath107 and the players are left with no motivation to form a coalition . a quantum version of this cooperative game ,
therefore , exists in which players are deprived of motivation to form a coalition .
the payoff to a @xmath70 player against @xmath73 , @xmath42 players in classical mixed strategy game can be obtained from eq .
( [ qpayoff ] ) by taking @xmath65 .
it gives @xmath108 note that @xmath109 and classical mixed strategy game is zero - sum and @xmath110 .
the quantum version of this game is not zero - sum always because from eq .
( [ qpayoff ] ) we have @xmath111 and the quantum game becomes zero - sum only when @xmath112 .
there may appear several guises in which the players can cooperate in a game .
one possibility is that they are able to communicate and , hence , able to correlate their strategies . in certain situations
players can make binding commitments before or during the play of a game . even in the post - play behavior
the commitments can make players to redistribute their final payoffs .
the two - player games are different from multi - player games in an important aspect . in two - player games the question before
the players is whether to cooperate or not . in multi - player case
the players are faced with a more difficult task .
each player has to decide which coalition to join .
there is also certain uncertainty that the player faces about the extent to which players outside his coalition may coordinate their actions .
analysis of cooperative games isolating coalition considerations instead of studying elaborate strategic structures has drawn more attention .
recent exciting developments in quantum game theory provide a motivation to see how forming a coalition and its associated advantages can be influenced in already proposed quantum versions of these cooperative games . to study this we selected an interesting but simple cooperative game as well as a recently proposed scheme telling how to play a quantum game .
we allowed the players in the quantum version of the game to form a coalition similar to the classical game .
the underlying assumption in this approach is that because the arbiter , responsible for providing three qubit pure quantum initial states to be later unitarily manipulated by the players , can forward a quantum state that correspond to the classical game , therefore , other games corresponding to different initial pure quantum states are quantum forms of the classical game .
this assumption , for example , reduces the problem of finding a quantum version of the classical coalition game we considered , with an interesting property that the advantage of making a coalition is lost , to finding some pure initial quantum states .
we showed that such quantum states can be found and , therefore , there are quantum versions of the three - player coalition game where the motivation for coalition formation is lost . in conclusion , we considered a symmetric cooperative game played by three players in classical and quantum forms . in classical form of this game , which is also embedded in the quantum form , forming a coalition
gives advantage to players and players are motivated to do so . in quantum form of the game ,
however , an initial quantum state can be prepared by the arbiter such that coalition forming is of no advantage . the interesting function in these situations
i.e. ` value of coalition ' is greater for coalition then for player outside ; when the game is played classically .
these values become same in a quantum form of the game and motivation to form a coalition is lost .
there is , nevertheless , an essential difference between the two forms of the game i.e. classical game is zero - sum but its quantum version is not . | we study two forms of a symmetric cooperative game played by three players , one classical and other quantum . in its classical form
making a coalition gives advantage to players and they are motivated to do so . however in its quantum form the advantage is lost and players are left with no motivation to make a coalition . |
the spin-@xmath9 anisotropic heisenberg model , or the xxz model , is one of the most important quantum spin models , which can be used to explain the quantum essence of many magnetic materials , such as la@xmath10cuo@xmath11 and thin films of @xmath12he@xcite .
recently , for this model , many works have been done on translational symmetric latticespaper2,paper3,paper4,paper5,squre , paper6,paper7 .
for example , f c alcaraz and a l malvezzi studied the xxz chain in the presence of an external field and obtained the phase diagram of this system accurately@xcite .
continentino and sousa , respectively , have investigated the critical properties of the two - dimensional ( 2d ) anisotropic heisenberg modelpaper4,paper5,squre , and found that , in the isotropic heisenberg limit , this system does not exhibit finite temperature phase transition ( i.e. , the critical temperature @xmath13 ) , which is in accordance with the theorem of mermin and wagner@xcite . however , as to this model on the simple cubic lattice @xmath14 no longer equals zero in the isotropic heisenberg limit , which means the existence of finite temperature phase transitionpaper6,paper7,paper9,paper10,paper11 . on the other hand
, there has been some interest in the critical phenomena of xxz model on fractal lattices , and particularly on the diamond - type hierarchical ( dh ) lattices .
because of the special geometrical and topological property , dh lattices are good candidates to investigate the spin systems in non - integer dimensions , and since they have a much lower symmetry than other fractals , so they may provide insights into other low - symmetry problems such as random magnets , surfaces , and the likekaufman1,paper18 . in these aspects , some theoretical works have been done . in 1983 , using the renormalization group ( rg ) method , caride _
et al _ have investigated the critical behavior of the heisenberg model on the wheatstone - bridge - basis hierarchical lattice .
their results show that in the isotropic heisenberg limit @xmath14 approaches zero as a continuous function of the anisotropic parameter@xcite .
latterly , the phase diagram of this model was obtained by souza on another kind of dh lattice with fractal dimension @xmath15@xcite , and the results agree well with those on the square lattice .
recently , using a real - space rg method , the anisotropic heisenberg spin - glass model on a three - dimensional dh lattice has been studied@xcite . in the study of the anisotropic heisenberg model ,
many effective approximate methods have been applied . such as the mean - field approximation@xcite ,
series expansion@xcite , rg method @xcite and monte carlo simulation@xcite , etc . among these methods ,
the rg theory has been proved to be very powerful and it has been widely used to investigate the critical behavior of different spin systemspaper2,paper6,paper14,paper15 . in this paper , using the rg method , we investigate the quantum spin-@xmath9 xxz ferromagnetic model on two kinds of dh lattices with fractal dimensions @xmath16 and @xmath8 , respectively .
our results show that the systems exhibit finite temperature phase transitions in the isotropic heisenberg limit @xmath17 . besides
, we also investigate the above systems in the range of @xmath5 and find that they exhibit xy - like fixed points .
the outline of the remainder of this paper is as follows . in the next section ,
the model and the calculation method are presented ; sec .
iii gives the results ; sec .
iv is a discussion about some interesting quantum effects and sec .
v gives a brief conclusion .
some of the more tedious of the formulations in sec .
ii are illustrated in the appendix .
the effective hamiltonian of the xxz model can be written as @xmath18 , \label{eq1}\]]where @xmath19 , in which @xmath20 is the exchange coupling parameter ( @xmath21 and @xmath22 correspond to the ferromagnetic and the antiferromagnetic model , respectively ) , @xmath23 is the boltzmann constant , and @xmath24 is the absolute temperature .
@xmath25 @xmath26 are spin pauli operators on the site @xmath27 .
the sum is over all the nearest - neighbor spin pairs @xmath28 , and @xmath29 , @xmath30 $ ] represents the anisotropic parameter .
note that the hamiltonian contains , as particular cases , the ising model @xmath31for @xmath32 , the isotropic heisenberg model @xmath31for @xmath33 and the xy model ( for @xmath34 ) .
the dh lattices are all constructed by an iterative manner .
1@xmath35 illustrates the first three construction stages of dh lattice with fractal dimension @xmath2 ( lattice a , for simplicity ) . as can be seen , the initiator is a two - point lattice joined by a single bond ( construction stage @xmath36 ) . then the initiator is replaced with the generator ( the cluster of @xmath37 stage ) . replacing every single bond on the generator itself , we get the second stage of the lattice .
if this procedure is repeated infinite times , we then construct a dh lattice with self - similar structurepaper16,paper162,paper17,paper18 . using the same procedure as that of lattice a , we can construct another kind of dh lattice ( lattice b , for simplicity ) with fractal dimension @xmath8 ( see fig .
1@xmath38 ) . in this section ,
using the rg method proposed by caride@xcite , we give the main calculation procedure of the xxz ferromagnetic model on lattice a. the case of lattice b can be solved in the same way . in the next rg procedure , the hamiltonians , except for the ising limit @xmath39 and the high temperature limit @xmath40 , among the neighboring generators do not commute with each other , which leads to the impossible of decoupling the genorator from the whole lattice . in order to achieve this goal ,
the noncommutativity among the neighboring generators were neglected .
this method had been used in refs.paper12,paper13,paper19,paper20 and the quantum effects and the approximation of this method had been detailedly discussed in ref.paper20 . based on the above approximation , we take out the generator from lattice a to perform the rg transformation , which is shown in fig .
2 . as can be seen , after summation of the internal spins @xmath41 , @xmath42 and @xmath43 ( decimation ) , the generator ( fig .
2@xmath35 ) is transformed into a new structure , which contains two spins ( i.e. , @xmath44 and @xmath45 ) joined by a single bond ( fig .
2@xmath46 ) .
this procedure can be described as@xmath47where @xmath48 and @xmath49 are , respectively , the hamiltonians associated with the clusters @xmath35 and @xmath38 in fig .
2 , tr@xmath50 denotes the partial trace over states of the internal spins @xmath51 and @xmath52 . according to eq .
( [ eq1 ] ) , the hamiltonians @xmath48 and @xmath49 are@xmath53 \notag \\ & & + k\left ( \sigma _ { 1}^{z}\sigma _ { 3}^{z}+\sigma _ { 1}^{z}\sigma _ { 4}^{z}+\sigma _ { 1}^{z}\sigma _ { 5}^{z}+\sigma _ { 3}^{z}\sigma _ { 2}^{z}+\sigma _ { 4}^{z}\sigma _ { 2}^{z}+\sigma _ { 5}^{z}\sigma _ { 2}^{z}\right ) \label{eq3}\end{aligned}\]]and@xmath54+k_{0 } , \label{eq4}\]]respectively , where @xmath55 is a constant included to make eq .
( [ eq2 ] ) possible . by calculating the partial trace in eq .
( [ eq2 ] ) , we can obtain the rg recurrent relations between the new parameters @xmath56 and the original parameters @xmath57 , @xmath58 .
firstly , we expand @xmath59 as@xmath60note that , in the expansion of @xmath59 , the anticommutation among the spin pauli operators are considered .
further , in the basis which simultaneously diagonalize @xmath61 and @xmath62 , we express both sides of eq .
( [ eq5 ] ) in the form of matrix . the left - hand side of eq .
( [ eq5 ] ) can be expressed as @xmath63 in which @xmath64@xmath65and@xmath66are eigenvalues of @xmath49 , where@xmath67following the same steps as above , the right - hand side of eq .
( [ eq5 ] ) can also be expressed in the form @xmath68by combining eqs .
( [ matrix6])-([matrix11 ] ) , we get@xmath69@xmath70and@xmath71obviously , the coefficients @xmath72 , @xmath73 and @xmath74 are all functions of @xmath75 and @xmath76 .
so from eqs .
( [ eq12])-([eq14 ] ) , we obtain the relations between the new parameters @xmath77 and the expansion coefficients @xmath78 as follows @xmath79@xmath80analogously , @xmath81 can also be written as@xmath82 \notag \\ & & + \sum_{\left\langle i , j\left ( > i\right ) \right\rangle \neq \left\langle k , l\left ( > k\right ) \right\rangle } \left [ d_{ij , kl}\left ( \sigma _ { i}^{x}\sigma _ { j}^{x}+\sigma _ { i}^{y}\sigma _ { j}^{y}\right ) \sigma _ { k}^{z}\sigma _ { l}^{z}+e_{ij , kl}\left ( \sigma _ { i}^{x}\sigma _ { j}^{x}+\sigma _ { i}^{y}\sigma _ { j}^{y}\right ) \left ( \sigma _ { k}^{x}\sigma _ { l}^{x}+\sigma _ { k}^{y}\sigma _ { l}^{y}\right ) \right ] \notag \\ & & + \sum_{\left\langle i , j\left ( > i\right ) \right\rangle
\neq \left\langle k , l\left ( > k\right ) \right\rangle } f_{ij , kl}\sigma _
{ i}^{z}\sigma _ { j}^{z}\sigma _ { k}^{z}\sigma _ { l}^{z } , \label{eq18}\end{aligned}\]]where @xmath83 , @xmath84 , @xmath85 , @xmath86 , @xmath87 and @xmath88 are all functions of @xmath89 and @xmath90 , which can be determined by diagonalizing both sides of eq .
( [ eq18 ] ) in the same basis of @xmath91 , @xmath92 , @xmath93 , @xmath94 and @xmath95 . because the matrices of the two sides are too lengthy
, we just give the diagonal matrix form of @xmath48 in the appendix . from eqs .
( [ eq2 ] ) , ( [ eq5 ] ) and ( [ eq18 ] ) , we can obtain the relations among the expansion coefficients@xmath96@xmath97and@xmath98these expressions , together with eqs .
( [ eq15 ] ) and ( [ eq16 ] ) , we finally get @xmath99@xmath100since @xmath83 , @xmath101 and @xmath102 are all functions of @xmath89 and @xmath103 , the rg recurrent relations between the new parameters @xmath104 and the original parameters @xmath105 are determined by eqs .
( [ eq22 ] ) and ( [ eq23 ] ) . however , the analytical expression about their right - hand sides are difficult to obtain . in order to obtain the phase diagram and the fixed points
, we will numerically solve the above two equations and the results will be presented in the next section .
by numerically iteration , the rg recurrent equations ( [ eq22 ] ) and ( eq23 ) , the phase diagram and the fixed points for the system with fractal dimension @xmath2 can be obtained , which are shown in fig .
3 . the curve @xmath35 in fig .
3 gives the critical line of the system .
as can be seen , in the range of @xmath106 $ ] , the phase space is divided into the paramagnetic phase ( p ) and the ferromagnetic phase ( f ) by the critical line . in this range
, we can also find that the system exhibits two unstable fixed points , i.e. , the ising unstable fixed point ( ip for short ) @xmath107 @xmath108 and the isotropic heisenberg unstable fixed point ( hp ) @xmath109 .
in fact , the value of the ip is in accordance with that of the classical modelpaper16,paper162 . at the ip
, the correlation length critical exponent can be calculated by @xmath110 in which , @xmath111 is the scaling factor and @xmath112 , thus , @xmath113 .
we note that this value is the same as that of the classical ising model@xcite .
it is worth mentioning that , at the hp , the system exhibits a finite temperature phase transition with critical temperature @xmath114 .
however , this is disagree with the result of the wheatstone - bridge - basis hierarchical latticepaper12 , where only zero temperature phase transition exists , i.e. , @xmath115 when @xmath116 ( curve @xmath38 in fig .
3 ) . in addition
, the isotropic heisenberg critical exponent is obtained as @xmath117 which differs greatly from the results @xmath118 of the 2d regular lattices@xcite and other hierarchical lattices with lower fractal dimensions @xmath119paper12,paper13 .
we also investigate the above system in the range of @xmath5 .
our results show that the system exhibits an unstable fixed point , which is not given in refs .
@xcite and we call it xy - like fixed point ( xyp )
. the critical line between the hp and the xyp divides the phase space into paramagnetic phase ( p ) and ferromagnetic phase ( f ) as well . using the same calculation procedure
as the lattice a , we can also investigate the critical behavior of the xxz model on the lattice b , i.e. , the dh lattice with @xmath8 . the phase diagram and the fixed points
are presented in fig .
4 . we can see that the phase space is divided into paramagnetic phase ( p ) and ferromagnetic phase ( f ) by the critical line . in the range of @xmath120 $ ] ,
the system also exhibits two unstable fixed points , i.e. , the ising fixed point ( ip ) and the isotropic heisenberg fixed point ( hp ) . at the ip @xmath121 @xmath122
we obtain the critical temperature @xmath123 and the critical exponent @xmath124 . compared with the results of three dimentional heisenberg model calculated by other rg methodspaper20,paper21 , our results are more consistent with that from series expansion ( @xmath125)@xcite . at the hp @xmath126 @xmath127 , the critical temperature @xmath128 is very close to that of three - dimensional heisenberg system ( @xmath129)@xcite . _ _ _ _ in addition , the critical exponent is @xmath130 which agrees well with that obtained by sousa _
et al _ on the simple cubic lattice where @xmath131@xcite . in summary
, we can see that the lattice b can be regarded as an approximation for the simple cubic lattice .
besides , just as the case of the lattice a , the system on lattice b also exhibits a xy - like fixed point ( xyp ) @xmath132 .
in this section , we discuss the possible effects of quantum fluctuation .
we assume the generator of lattice a as a whole . in this case
, there will be no noncommutativity and the result should be exact in our calculation . with this method , the exact calculation of the dh lattice
will be noted @xmath133 $ ] . on the other hand , for the rg transformation in fig .
2 , we apply a modified migdal - kadanoff method , which is an approximate one . in this method , the original cell ( fig .
2@xmath35 ) can be considered as a combination of 3 arrays in parallel , each of which is made up of two interactions in series .
the renormalized interaction @xmath89 and anisotropy @xmath103 can be firstly calculated for each combination in series and then combined in parallel .
the approximate calculation provides @xmath134 $ ] .
we use the convenient ratios introduced in ref .
@xcite@xmath135and@xmath136the @xmath24-dependences of @xmath137 and @xmath138 for typical values of @xmath139 are indicated in fig .
5 . in the high temperature limit ,
both @xmath137 and @xmath138 tend to unity for all values of @xmath103 ( @xmath140 and @xmath141 ) . in the range of low temperature , where the quantum effect tends to drive the system to a disordered phase ,
both @xmath137 and @xmath138 show oscillational behaviours .
the fact is usually due to quantum fluctuation which , at low temperatures ( @xmath24 ) and smaller anistropy ( @xmath139 ) are important@xcite . as we see in fig . 5 ,
the quantum fluctuation of xy - like model ( @xmath141 ) is stronger than the anistropic heisenberg model ( @xmath142 ) at the low - temperature region .
so , there will be considerable error when we calculate the xy - like model at lower temperatures .
however , in the range of @xmath142 , it is a very good approximation and when the anistropy @xmath103 is negative , as we can see in fig .
5 , it follows the same tendency as the positive one .
so , we can obtain approximate result at bigger temperatures and give qualitatively correct picture of the phase diagram .
when it comes to the lattice b , both @xmath143 and @xmath138 have the same type of behaviour and we can obtain the same conclusion .
in conclusion , using the rg method , we have studied the quantum spin-@xmath9 anisotropic xxz model on two kinds of dh lattices with fractal dimensions @xmath16 and @xmath3 , respectively .
phase diagrams , fixed points and critical exponents of the systems are obtained .
the results show that , in the ising limit @xmath144 , the values of the fixed points and the critical exponents agree well with those in refs . paper11,paper162,paper17 . in the isotropic heisenberg limit @xmath145
, there are finite temperature phase transitions on the above two lattices .
furthermore , the systems exhibit xy - like fixed points in the range of @xmath5 .
the quantum effects of this system show that , at low temperatures , the xy - like model has stronger fluctuation than the anistropic heisenberg model .
so , we conclude that there will be a considerable error when we calculate the xy - like model at low temperature . besides
, our results also indicate that the dh lattice with @xmath8 can be regarded as an approximation for the simple cubic lattice . as a comment
, this method can be extended to investigate spin systems with @xmath146 on other lattices and we are presently working along these lines .
this work is supported by the national science foundation for post - doctoral scientists of china ( 2005037442 ) and the science foundation of qufu normal university .
we thank prof .
shu chen for useful discussions .
one of the authors ( zhang ) would like to acknowledges many fruitful discussions with zhong - qiang liu , guang - hou sun , xian - ming li , xun - chang yin , xin zhang and zhong - yang gao .
fig . 3 phase diagram of the dh lattice with @xmath2 . the critical line @xmath156 separates the phase space into paramagnetic phase ( p ) and ferromagnetic phase ( f ) .
the ip , hp and xyp , respectively , denote the ising , isotropic heisenberg and xy fixed points .
@xmath46 is the critical line of the wheatstone - bridge - basis hierarchical lattice for comparison .
fig . 4 phase diagram of the dh lattice with @xmath8 .
the ip , hp and xyp , respectively , denote the ising , isotropic heisenberg and xy fixed points .
p and f correspond to the paramagnetic phase and ferromagnetic phase , respectively . | in this paper , the phase diagrams and the critical behavior of the spin-@xmath0 anisotropic xxz ferromagnetic model ( the anisotropic parameter @xmath1 $ ] ) on two kinds of diamond - type hierarchical ( dh ) lattices with fractal dimensions @xmath2 and @xmath3 , respectively , are studied via the real - space renormalization group method .
it is found that in the isotropic heisenberg limit @xmath4 , there exist finite temperature phase transitions for the two kinds of dh lattices above .
the systems are also investigated in the range of @xmath5 and it is found that they exhibit xy - like fixed points
. meanwhile , the critical exponents of the above two systems are also calculated .
the results show that for the lattice with @xmath2 , the value of the ising critical exponent @xmath6 is the same as that of classical ising model and the isotropic heisenberg critical exponent @xmath7 is a finite value , and for the lattice with @xmath8 , the values of @xmath6 and @xmath7 agree well with those obtained on the simple cubic lattice .
we also discuss the quantum fluctuation at all temperatures and find the fluctuation of xy - like model is stronger than the anistropic heisenberg model at the low - temperature region . by analyzing the fluctuation , we conclude that there will be remarkable effect of neglecting terms on the final results of the xy - like model .
however , we can obtain approximate result at bigger temperatures and give qualitatively correct picture of the phase diagram . |
let @xmath0 be a flow on @xmath1 with countably many periodic orbits @xmath2 .
we can look to each closed orbit as a knot in @xmath1 .
it was r. f. williams , in 1976 , who first conjectured that non trivial knotting occur in the lorenz system ( @xcite ) . in 1983 ,
birmann and williams introduced the notion of template , in order to study the knots and links ( i.e. finite collections of knots , taking into account the knotting between them ) contained in the geometric lorenz attractor ( @xcite ) . a template , or knot holder , consists of a branched two manifold with charts of two specific types , joining and splitting ,
together with an expanding semiflow defined on it , see figure [ charts ] .
the relationship between templates and links of closed orbits in three dimensional flows is expressed in the following result , known as template theorem , due to birman and williams in @xcite .
[ charts ] + given a flow @xmath0 on a three - manifold @xmath3 , having a hyperbolic chain - recurrent set , the link of closed orbits @xmath4 is in bijective correspondence with the link of closed orbits @xmath5 on a particular embedded template @xmath6 . on any finite sublink , this correspondence is via ambient isotopy .
we define a _ lorenz flow _ as a semi - flow that has a singularity of saddle type with a one - dimensional unstable manifold and an infinite set of hyperbolic periodic orbits , whose closure contains the saddle point ( see @xcite ) . a lorenz flow , together with an extra geometric assumption ( see @xcite )
is called a _ geometric lorenz flow_. the dynamics of this type of flows can be described by the iteration of one - dimensional first - return maps @xmath7\setminus \{c\}\rightarrow [ a , b]$ ] with one discontinuity at @xmath8a , b[$ ] , increasing in the continuity intervals @xmath9 and @xmath10c , b]$ ] and boundary anchored ( i.e. @xmath11 and @xmath12 ) , see @xcite . these maps are called lorenz maps and sometimes we denote them by @xmath13 , where @xmath14 and @xmath15 correspond , respectively , to the left and right branches . in @xcite , holmes studied families of iterated horseshoe knots which arise naturally associated to sequences of period - doubling bifurcations of unimodal maps .
it is well known , see for example @xcite , that period doubling bifurcations in the unimodal family are directly related with the creation of a 2-renormalization interval , i.e. a subinterval @xmath16 containing the critical point , such that @xmath17 is unimodal .
basically there are two types of bifurcations in lorenz maps ( see @xcite ) : one is the usual saddle - node or tangent bifurcations , when the graph of @xmath18 is tangent to the diagonal @xmath19 , and one attractive and one repulsive @xmath20-periodic orbits are created or destroyed ; the others are homoclinic bifurcations , when @xmath21 and one attractive @xmath20-periodic orbit is created or destroyed in this way , these bifurcations are directly related with homoclinic bifurcations of flows modelled by this kind of maps ( see @xcite ) . considering a monotone family of lorenz maps ( see @xcite ) , the homoclinic bifurcations are realized in some lines in the parameters space , called _ hom - lines _ or _ bifurcation bones_. it is known that ( see @xcite and @xcite ) , in the context of lorenz maps , renormalization intervals are created in each intersection of two hom - lines .
these points are called _ homoclinic points _ and are responsible for the self - similar structure of the bifurcation skeleton of monotone families of lorenz maps . so it is reasonable to say that _ homoclinic points are the lorenz version of period - doubling bifurcation points_. in @xcite , williams considered lorenz maps with a double saddle connection , i.e. , such that @xmath21 , and introduced a new determinant - like invariant to classify templates generated by them , in @xcite these templates are called _ sub - lorenz templates _
( note that , in a monotone family of lorenz maps , maps corresponding to homoclinic points have a double saddle connection ) .
this invariant is a zeta - function that counts periodic orbits taking into account their knot type and the linking between them . in @xcite ,
michael sullivan introduced one other zeta function that counts periodic orbits in positive templates , taking into account the number of twists in their stable manifolds . in this paper
we will study the effect of renormalization over these two invariants , in the context of sub - lorenz templates .
symbolic dynamics is a very useful combinatoric tool to study the dynamics of one - dimensional maps .
let @xmath22 , @xmath23 , be the @xmath24-th iterate of the map @xmath25 .
we define the _ itinerary _ of a point @xmath26 under a lorenz map @xmath25 as @xmath27 , where @xmath28 it is obvious that the itinerary of a point @xmath26 will be a finite sequence in the symbols @xmath29 and @xmath30 with @xmath31 as its last symbol , if and only if @xmath26 is a pre - image of @xmath31 and otherwise it is one infinite sequence in the symbols @xmath29 and @xmath30 .
so we consider the symbolic space @xmath32 of sequences @xmath33 on the symbols @xmath34 , such that @xmath35 for all @xmath36 and : @xmath37 or @xmath38 , with the lexicographic order relation induced by @xmath39 .
it is straightforward to verify that , for all @xmath40 $ ] , we have the following : 1 . if @xmath41 then @xmath42 , and 2 . if @xmath43 then @xmath41 .
we define the _ kneading invariant _ associated to a lorenz map @xmath13 , as @xmath44 we say that a pair @xmath45 is _ admissible _ if @xmath46 for some lorenz map @xmath25 .
denote by @xmath47 , the set of all admissible pairs .
consider the _ shift map _
@xmath48 , @xmath49 .
the set of admissible pairs is characterized , combinatorially , in the following way ( see @xcite ) .
[ t3 ] let @xmath50 , then @xmath51 if and only if @xmath52 , @xmath53 and , for @xmath54 we have : ( 1 ) if @xmath55 then @xmath56 ; ( 2 ) if @xmath57 then @xmath58 ; with inequality ( 1 ) ( resp . ( 2 ) ) strict if @xmath59 ( resp .
@xmath60 ) is finite . on the other hand
, the kneading invariant of a map characterizes completely its combinatorics , more precisely , considering a pair @xmath51 , denote by @xmath61 the set of sequences @xmath62 that satisfy conditions ( 1 ) and ( 2 ) from the previous proposition , then we have the following proposition whose proof can be found in @xcite .
let @xmath63 and @xmath25 be a lorenz map , then there exists @xmath64 such that @xmath65 if and only if @xmath66 . in the context of lorenz maps ,
we define renormalizability on the following way , see for example @xcite : let @xmath25 be a lorenz map , then we say that @xmath25 is renormalizable if there exist @xmath68 with @xmath69 and points @xmath70 such that @xmath71 is a lorenz map .
the map @xmath72}$ ] is called the @xmath73-renormalization of @xmath25 .
let @xmath74 be the length of a finite sequence @xmath75 , it is reasonable to identify each finite sequence @xmath76 with the corresponding infinite periodic sequence @xmath77 , this is the case , for example , when we talk about the knot associated to a finite sequence .
denote @xmath78 .
it is easy to prove that a pair of finite sequences @xmath79 is admissible if and only if the pair of infinite periodic sequences @xmath80 is admissible .
we define the @xmath81-product between a pair of finite sequences @xmath50 , and a sequence @xmath82 as @xmath83 where @xmath84 now we define the @xmath81-product between two pairs of sequences , @xmath85 , @xmath59 and @xmath60 finite , as @xmath86 the next theorem states that the reducibility relative to the @xmath81-product is equivalent to the renormalizability of the map
. the proof can be found , for example , in @xcite .
let @xmath25 be a lorenz map , then @xmath25 is renormalizable with renormalization @xmath87 iff there exist two admissible pairs @xmath88 and @xmath89 such that @xmath90 , @xmath91 , @xmath92 and @xmath93 . we know from @xcite that @xmath94 if and only if both @xmath51 and @xmath95 , so for each finite admissible pair @xmath88 , the subspace @xmath96 is isomorphic to the all space @xmath97 , this provides a self - similar structure in the symbolic space of kneading invariants .
it is straightforward to verify that the @xmath67-product of kneading invariants is associative , consequently this self - similar structure is nested .
now we will state a series of properties concerning the @xmath67-product , that are in the basis of our results .
the following lemma states that the order structure is reproduced at each level of renormalization . [ l0 ] let @xmath88 be one admissible pair of finite sequences , and @xmath98 , then @xmath99 .
the proof is straightforward . for any sequence @xmath59 and @xmath100 ,
we denote @xmath101}=x_p\ldots x_q.\ ] ] since we have an order structure in @xmath32 , we will denote @xmath102=\left\lbrace x\in \sigma : a\leq x \leq b \right\rbrace .\ ] ] for any finite sequence @xmath103 , consider also the numbers @xmath104 and @xmath105 [ l1 ] let @xmath88 and @xmath106 be admissible pairs and @xmath107 and @xmath108 be any two sequences in @xmath109 such that @xmath110 .
consider @xmath111 , then a sequence @xmath112}0 \right\rbrace $ ] belongs to @xmath113}(x , y)*a , z_{[p,|z|-1]}(x , y)*b]\cap \sigma^+ ( ( x , y)*(s , w))$ ] if and only if @xmath114}(x , y)*c$ ] , with @xmath115\cap \sigma^+ ( ( s , w))$ ] . * proof * obviously @xmath116}=z_{[p,|z|-1]}$ ] .
if @xmath117 , then , since @xmath118}(x , y)*a$ ] , we have that @xmath119 and @xmath120}(x , y)*a = z_{[p,|z|-1]}x_{[0,|x|-1]}(x , y)*\sigma(a)$ ] , so @xmath121-p-1 ] } \geq x_{[0,|x|-1]}$ ] . on the other hand , since @xmath122 , then @xmath117 implies that @xmath121-p-1 ] } \leq x_{[0,|x|-1]}$ ] .
analogously we see that , if @xmath123 then @xmath124+|y|-p-1 ] } = y_{[0,|y|-1]}$ ] . repeating these arguments ad infinitum
we prove that @xmath114}(x , y)*w$ ] .
we will now prove that @xmath125 .
let us suppose by contradiction that @xmath126 .
then it happens one of the following situations : 1 .
there exists @xmath127 such that @xmath128 and @xmath129 .
there exists @xmath127 such that @xmath130 and @xmath131 .
if it happens situation ( i ) , then there exists @xmath132 such that @xmath133}=s_{[0,r-1]}$ ] and @xmath134 .
but then , with @xmath135})+|y|n_r(c_{[0,l-1]})$ ] , we have that @xmath136 and @xmath137 and this implies that @xmath138 .
if it happens situation ( ii ) , we obtain the contradiction analogously @xmath139 [ l10 ] let @xmath88 be one admissible pair of finite sequences , @xmath140 and @xmath141 , then @xmath142}(x , y)*z \geq ( \overline{y})^{\infty}$ ] , for any sequence @xmath103 .
analogously , if @xmath143 and @xmath144 , then @xmath145}(x , y)*z \leq ( \overline{x})^{\infty}$ ] , for any sequence @xmath103 .
* proof * since @xmath88 is admissible , then @xmath142}(\overline{y})^{\infty}>(\overline{y})^{\infty}$ ] , so there exists @xmath127 such that @xmath146}=y_{[0,l-1]}$ ] and @xmath147 .
if @xmath148 the result follows immediately .
if @xmath149 , then necessarily @xmath150 , because otherwise we would have @xmath151 and @xmath152 , and this violates admissibility .
but then , @xmath153}(\overline{y})^{\infty}\leq ( \overline{x } ) \leq ( x , y)*z\ ] ] and this gives the result .
the proof of the second part is analogous .
@xmath139 [ l2 ] let @xmath88 be one admissible pair of finite sequences and @xmath154 .
if @xmath155 and @xmath156 } \neq w'_{[q,|w'|-1]}$ ] then @xmath157}(x , y)*z \leq w'_{[q,|w'|-1]}(x , y)*z'\ ] ] for any sequences @xmath158 .
* proof * the proof is divided in four cases : @xmath159 and @xmath160 ; @xmath161 and @xmath162 ; @xmath163 and @xmath164
. we will only demonstrate specifically the first case , since the others follow with analogous arguments .. following the hypotheses , there exists @xmath127 such that @xmath165}=y_{[q,(q+l-1)\mod |y|]}$ ] and @xmath166 . if @xmath167 , then the result follows immediately . if @xmath168 and @xmath169}=y_{[q , q+|x|-p-1]}$ ] , then @xmath170 , because otherwise we would have @xmath171 and @xmath172}y^{\infty } > x^{\infty}$ ] , and this violates admissibility of @xmath88 .
so @xmath170 and , from lemmas [ l0 ] and [ l10 ] , @xmath173}(x , y)*z',\ ] ] and the result follows . if @xmath174 and @xmath175}=y_{[q,|y|-1]}$ ] , then @xmath176 , because otherwise we would have @xmath177 and @xmath178 , which contradicts admissibility of @xmath88
. so @xmath176 and @xmath179}(x , y)*z\leq x^{\infty } \leq ( x , y)*z'\ ] ] and the result follows .
@xmath139 let us now introduce some more notations : for @xmath180 , @xmath181}=z_l\cdots z_p.\ ] ] @xmath182 @xmath183 and @xmath184 is the map that associates to each @xmath185 , the position occupied by @xmath59 in the lexicographic ordenation of @xmath186
. for each @xmath187 , denote @xmath188 \text { if } m(x , y)=0 \\ \\ \left [ x_{[|x|-m(x , y),|x|-1]}(x , y)*\phi _ { ( s , w)}^{-1}(k ) , x_{[|x|-m(x , y),|x|-1]}(x , y)*\phi _ { ( s , w)}^{-1}(k+1)\right ] \\
\text { if } m(x , y)\neq 0 \end{array } \right.\ ] ] [ main ] from the previous three lemmas we can take the following conclusions : 1 . if @xmath189 , denote by @xmath190 the set @xmath191}(x , y)*\sigma ^k(z )
: z\in \left\lbrace s , w \right\rbrace \ : and \
: z_{k-1}=l\right\rbrace $ ] , from lemmas [ l0 ] and [ l1 ] , @xmath192}(x , y)*\sigma ^2(w),x_{[p,|x|-1]}(x , y)*\sigma ( s)\right ] \cap \sigma((x , y)*(s , w))$ ] , analogously denoting @xmath193}(x , y)*\sigma ^k(z ) : z\in \left\lbrace s , w \right\rbrace \ : and \ : z_{k-1}=r\right\rbrace $ ] the setwe have that @xmath194}(x , y)*\sigma ( w),y_{[p,|y|-1]}(x , y)*\sigma ^2(s)\right ] \cap \sigma((x , y)*(s , w))$ ] on the other hand , if @xmath195 , then @xmath169}=y_{[p,|y|-1]}$ ] and @xmath196}(x , y)*\sigma ( w ) , x_{[p,|x|-1]}(x , y)*\sigma ( s)\right ] \cap \sigma ( ( x , y)*(s , w))=\left\lbrace x_{[p,|x|-1]}(x , y)\sigma ^k(z ) : z\in \left\lbrace s , w \right\rbrace\right\rbrace $ ] . without risk of confusion
, we will denote these sets by @xmath190 .
the ordenation of the elements of the sets @xmath190 and @xmath197 is induced by the ordenation of the sequences @xmath198 such that @xmath199 .
this follows immediately from lemma [ l0 ] .
3 . for each @xmath200 , if @xmath201 then @xmath202 . on the other hand , @xmath203 .
this follows immediately from the definitions .
let @xmath204 $ ] and @xmath205 $ ] , it follows from lemma [ l2 ] that @xmath206 iff @xmath207 .
moreover , from ( 3 ) and lemma [ l0 ] , if @xmath208 , then @xmath209 and @xmath210 5 . from ( 2 ) and ( 3 ) , performing some straightforward computations with the lengths of @xmath211 and @xmath212 we see that , for @xmath213 , then @xmath214 if and only if both @xmath215 and @xmath216 are contained in @xmath217 and @xmath218 , where @xmath219 $ ] and @xmath220 and @xmath221 are such that @xmath222}\left ( \phi_{(s , w)}^{-1}(k)\right ) + |y|n_{r[0,p-1]}\left ( \phi_{(s , w)}^{-1}(k)\right)$ ] and @xmath223}\left ( \phi_{(s , w)}^{-1}(k')\right ) + |y|n_{r[0,q-1]}\left ( \phi_{(s , w)}^{-1}(k')\right)$ ] .
this remark is the kernel of the proofs of all our main results , so we will illustrate it with one example
. we will consider @xmath224 and @xmath225 . denoting by @xmath226 , for @xmath227 and setting that @xmath228 if @xmath107 is located at the left of @xmath108 , we have the following ordenations of the members of the pairs @xmath88 and @xmath106 ,
moreover , the relative position of each member gives the maps @xmath229 and @xmath230 .
@xmath231 @xmath232 we will now consider the @xmath67-product @xmath233 we used the underscripts to indicate the corresponding iterate of the shift map and the upperscripts to indicate the element of @xmath106 that generated each subword in the @xmath67-product . denoting by @xmath234 we will now order the elements of the @xmath67-product pair : @xmath235 as we can see , the ordered disposition of the members of @xmath236 is obtained from the ordered disposition of @xmath88 , inflating each @xmath237 , @xmath238 and substituting it by the elements of the corresponding @xmath239 , ordered according with the ordenation of the @xmath240 , @xmath241 .
now we will follow @xcite to introduce sub - lorenz templates and williams zeta - functions .
we say that a lorenz map @xmath25 haves a double saddle connection if @xmath242 for some @xmath243 . in this case
the points @xmath244 we can define a finite markov partition for the semiflow . the restriction of the semiflow to this partition is called a sub - lorenz template . for @xmath245 , we construct the sub - lorenz template @xmath246 following the procedure of figure [ figtempl ] [ figtempl ] we identify each sub - lorenz template with the corresponding kneading invariant @xmath88 and denote it with @xmath247 .
we associate to the template the transition matrix @xmath248 $ ] where @xmath249 now we associate to @xmath247 a labeled transition matrix @xmath250 $ ] , where @xmath251 in the previous example we have @xmath252\ ] ] and @xmath253\ ] ] given a group @xmath254 and a matrix @xmath107 with entries in @xmath254 , we define the _ link - ring _ @xmath255 as follows : 1 . for each sequence ( called a cycle below )
@xmath256 such that the product @xmath257 let @xmath258 be the equivalence class under cyclic permutations of this product .
these equivalence classes are called * _ free - knot symbols _ * and the indices @xmath259 are called _ nodes _ of the free knot symbol .
2 . define @xmath255 to be the free abelian group generated by the free knot symbols defined in ( 1 ) .
the nodes in the matrix correspond to the cells in the markov partition .
the cycles in the matrix correspond to periodic orbits on the markov partition , and thus to periodic orbits in the flow .
since there is no natural way to choose a specific point for a periodic orbit , the natural invariant is the cyclic permutation class .
* we do not permit products of letters to commute * , since this usually corresponds to different orbits and consequently to different knot types , for example @xmath260 and @xmath261 correspond , in the first case to the unknot and in the second to the trefoil knot . however , when we consider a product of words , it represents the union of two knots ( i.e. * link * ) and in this case we do nt care which happens first , so * we permit products of free - knot symbols ( words ) to commute*. by a * _ free - link symbol _ * in the ring @xmath255is meant a product @xmath262 of free - knot symbols , no two of which have a node in common .
( just as a product of two cycles can occur in a determinant , only if they have no node in common ) in @xcite , williams defined the following determinant - like function and proved the following two results about it : let @xmath263 be a labeled transition matrix of a sub - lorenz template , then @xmath264 for lorenz attractors @xmath265 with a double saddle connection , @xmath266 where @xmath267 is the collection of all links @xmath29 in the attractor which have at most one point in each partition set , @xmath268 is the number of components in @xmath29 , and @xmath269 means the free - link symbol of @xmath29 . for lorenz attractors
@xmath265 with a double saddle connection , we have that @xmath270 from this result , we name @xmath271 , the * _ williams zeta function _ * of the template @xmath265 .
finally we state our factorization result about williams zeta - functions : let @xmath272 be a sub - lorenz template generated by a renormalizable lorenz map @xmath25 , with a double saddle connection and kneading invariant kneading invariant @xmath273 , then @xmath274 \times \left[(x , y)*\zeta_w(t_{(s , w ) } ) + 1\right],\ ] ] where @xmath275 such that the sum is taken over all free - link symbols of @xmath276 , @xmath277 .
* proof * denote @xmath278 . 1 .
let @xmath279 be the markov partition associated to @xmath236 , and @xmath280 be a cell of @xmath279 , then @xmath281 for some @xmath282 or @xmath283 for some @xmath284 such that @xmath285 .
moreover , from 3 of remark [ main ] we have that @xmath286 for all @xmath20 and , from 4 , the free - knot symbols associated to nodes in @xmath287 are exactly the same free - knot symbols from @xmath247 and consequently the free - link symbols are also the same .
2 . from ( 5 ) of remark [ main ] , the free - link symbols associated to knots in @xmath288 are exactly those that can be written in the form @xmath289 , where @xmath290 are free - link symbols of @xmath276 .
3 . finally , from ( 3 ) of remark
[ main ] there are no cycles , simultaneously with nodes in @xmath291 and in @xmath292 , so all free - link symbols of type @xmath293 where @xmath294 is a free link symbol associated to nodes in @xmath291 and @xmath295 is a free link symbol associated to nodes in @xmath296 will be terms of the zeta - function of @xmath272 .
finally , from ( i ) , ( ii ) and ( iii ) we have that @xmath297 @xmath139
in this section we will present a factorization formula for the twist - zeta function presented by sullivan in @xcite .
a ribbon is an embedded annulus or mbius band in @xmath1 . like knots and templates ,
ribbons can be braided . a ribbon which has a braid presentation such that each crossing of one strand over another is positive and each twist in each strand is positive ,
will be called a positive ribbon .
the core and boundary of positive ribbons are positive braids .
if @xmath30 is a ribbon and @xmath298 is a braid presentation of @xmath30 , we define the computed twist @xmath299 where @xmath300 is the sum of the half twists in the strands of @xmath298 and @xmath20 is the number of strands of the core . for a given template @xmath302 , let @xmath305 be the number of closed orbits with computed twist @xmath306 .
let @xmath307 .
define the sullivan zeta function of the template to be the exponential of a formal power series : @xmath308 we now define a twist matrix @xmath309 $ ] whose entries are nonnegative powers of @xmath300 and @xmath31 s , by considering the contribution to @xmath301 as an orbit goes from one element of a markov partition to another .
let @xmath310 if there is no branch going from the @xmath311-th to the @xmath24-th partition element and @xmath312 if there is such a branch , where @xmath313 is the amount of computed twist an orbit picks up as it travels from the @xmath311-th to the @xmath24-th partition element . in the case of a sub - lorenz template @xmath247 ,
since we have one curl in each ribbon from one partition element to other , see figure [ figtempl ] then we obtain @xmath314 simply substituting each @xmath315 element of the transition matrix by @xmath316 , i.e. @xmath317 * proof * from remark [ main ] , the markov partition @xmath323 associated to @xmath272 can be splited as @xmath324 in such a way that all the iterates of each periodic orbit are exclusively contained in @xmath291 or in @xmath325 .
so we can split the sum @xmath326 where @xmath327 ( respectively @xmath328 ) means simply that we are counting orbits contained in @xmath291 ( respectively in @xmath325 ) . from ( 4 ) of remark
[ main ] , @xmath329 .
now , from ( 5 ) of remark [ main ] , each ribbon leaving a cell @xmath330 makes @xmath74 curls without splitting if @xmath330 is on the left of @xmath31 and makes @xmath331 curls without splitting if @xmath330 is on the left of @xmath31 , before reenter in @xmath332 .
moreover @xmath333 ( respectively @xmath334 ) is not empty if and only if @xmath335 , so @xmath336 and the result follows .
r. williams , _ the structure of lorenz attractors_. in a. chorin , j. marsden and s. smale , editors , _ turbulence seminar , berkeley 1976/77 _ , lecture notes in mathematics , volume 615 , springer ( 1977 ) , 94 - 116 . | we describe the williams zeta functions and the twist zeta functions of sub - lorenz templates generated by renormalizable lorenz maps , in terms of the corresponding zeta - functions of the sub - lorenz templates generated by the renormalized map and by the map that determines the renormalization type .
nuno franco and lus silva |
one of the long - standing problems in astrophysics is the process of star formation in galaxies .
the standard scenario assumes stars to form from gas that falls in the potential wells of dark matter halos .
subsequent interacting or merging stages among galaxies might trigger additional bursts of star formation .
the complex nature of star formation makes this problem rather an untractable one from an analytical point of view , so that the best approach towards understanding the distribution of stellar populations in galaxies requires a heavy use of rough approximations and all - too - often dangerous generalizations .
it is the purpose of current phenomenological models describing the formation and evolution of the stellar component in galaxies to reveal the mechanisms which describe the wide range of galaxy colors and luminosities as well as their connection to morphology .
the current status of the determination of the ages of the stellar populations in galaxies is rather controversial due to the degeneracy between age and metallicity ( worthey 1994 ) .
observations of early - type galaxies by two different groups using similar techniques targeting narrow spectral indices to infer a luminosity - weighted age give contradictory results . while trager et al .
( 2000 ) find a large age spread in the sample of field and group early - type systems of gonzlez ( 1993 ) , kuntschner ( 2000 ) reports a large metallicity spread in fornax cluster ellipticals .
so far , any observational measurement of age is plagued by many degeneracies which render a direct estimate uncertain .
an alternative approach modelling the formation and chemical enrichment of the stellar component of galaxies is needed in order to reveal the actual scenario of galaxy formation .
= 2.6 in -0.3truein
the model presented here describes the process of star formation in early - type galaxies in terms of four parameters : star formation efficiency ( @xmath0 ) , ejected gas fraction in outflows ( @xmath1 ) , formation redshift ( @xmath2 ) and infall timescale ( @xmath3 ) .
the latter two parameters refer to the epoch ( @xmath4 ) at maximum and spread of a gaussian profile for the infalling gas , i.e. : @xmath5 this gas will be turned into stars according to a linear schmidt - type law , where the proportionality constant is the star formation efficiency parameter .
the model is described in more detail in ferreras & silk ( 2000a , b ) .
this generic description allows us to include multi - burst scenarios in galaxies undergoing several merging stages with enough gas to fuel star formation at each merging event .
figure 1 shows a comparison between two star formation histories : one with three equally strong starburst events and a second one as a simplification to the former using our approach .
the top panel shows the evolution of the gas mass which is proportional to the star formation rate .
the bottom panel shows the evolution of the metallicity and the top inset traces the evolution of rest frame @xmath6 color for both scenarios .
one can see that at times after the last bursting episode , the evolution in both cases is roughly undistinguishable .
hence , we conclude that our four - parameter model can account not only for a standard `` monolithic '' scenario but also for multi - burst formation histories .
furthermore , a gaussian profile for infall avoids the overproduction of low metallicity stars .
in fact , a suitable choice of infall parameters ( @xmath3,@xmath2 ) can reproduce the local metallicity distribution of stars ( rocha - pinto & maciel 1996 ) .
= 3.5 in -0.3truein for a given set of four parameters ( @xmath1,@xmath0,@xmath2,@xmath3 ) we can trace a star formation history and convolve it in age and metallicity with the simple stellar populations of bruzual & charlot ( in preparation ) .
hereafter a closed cosmology ( @xmath7 , @xmath8 , @xmath9 km s@xmath10 mpc@xmath10 ) , and a hybrid initial mass function between scalo and salpeter ( ferreras & silk 2000b ) are used . out of these four parameters , we find only @xmath1 and @xmath0 can generate the range of observed @xmath6 colors in nearby early - type cluster galaxies ( bower , lucey & ellis 1992 ) .
hence , the luminosity sequence of these systems could be explained either by a range of outflows ( @xmath1-sequence ) , by a range of star formation efficiencies ( @xmath0-sequence ) , or by some combination thereof .
figure 2 shows the predicted color - magnitude relation ( cmr ) of coma galaxies at the redshift of cluster cl0016 + 16 ( @xmath11 , ellis et al .
1997 ) assuming a @xmath1-sequence ( _ left _ ) or a @xmath0-sequence ( _ right _ ) , and a range of infall parameters ( @xmath2,@xmath3 ) .
a sequence driven by @xmath0 results in an age spread for the stellar populations .
this causes the remarkable departure of the predictions from the observed cmr ( shaded area ) for extended star formation histories ( @xmath12 , @xmath13 gyr ) .
however , because of the age - metallicity degeneracy , we find that quite a large range of the parameters agree with the observations within error bars .
hence , we can not use photometric measurements of moderate redshift clusters in order to determine whether age ( @xmath0 ) or metallicity ( @xmath1 ) drive the cmr .
one of the most age - sensitive observables is the mass - to - light ratio .
hence , the predicted evolution of @xmath14 with lookback time should be different for sequences driven by age or by metallicity .
we consider the evolution with redshift of the slope of the correlation between @xmath14 in rest frame @xmath15-band and _ stellar _ mass .
this slope change is parametrized by @xmath16 defined as follows : @xmath17 for a @xmath1-sequence ( driven by outflows ) , the range of luminosities is related by a spread in metallicities . as we evolve the cluster to higher redshifts , the mass - to - light ratio will decrease uniformly across the luminosity sequence because of lookback time , and there will also be a relative change of @xmath14 among early - type galaxies caused by its very weak metallicity dependence , which makes the decrease in mass - to - light ratio slighly larger in galaxies with a higher metallicity , thereby flattening the slope of @xmath14 vs @xmath18 ( i.e. @xmath19 ) .
on the other hand , a @xmath0-sequence ( driven by efficiency ) will introduce a significant age spread which varies with galaxy mass , so that @xmath14 at the fainter end ( which has a lower efficiency and thus a larger age spread ) will decrease more than the bright end , steepening the slope ( i.e. @xmath20 , see figure 5 in ferreras & silk 2000b ) .
this behavior makes the study of the evolution of @xmath14 with redshift a suitable candidate to infer the star formation history of early - type cluster galaxies .
unfortunately , this observable still poses a long string of uncertainties which prevent it from establishing a clearcut way of breaking the degeneracy between age and metallicity : mass - to - light ratios require time - consuming spectral observations in order to measure velocity dispersions , and can only be achieved with 10 m class telescopes for clusters at moderate and high redshifts .
furthermore , the measured @xmath14 ratios ( inferred from observations of velocity dispersions , surface brightnesses and galaxy sizes ) rely on a set of assumptions about the structure of the galaxy .
any correlation between galaxy structure and mass or luminosity will add systematic errors which are hard to estimate .
however , alternative age - dependent observables such as balmer spectral indices are also plagued by model - dependent uncertainties . despite all these caveats , the study of the evolution of @xmath14 with lookback time is still one of the best methods to determine the stellar demography in galaxies .
bower , r. g. , lucey , j. r. & ellis , r. s. 1992 , , 254 , 601 ferreras , i. & silk , j. 2000a , , march 20 , astro - ph/9910385 ferreras , i. & silk , j. 2000b , , in press gonzlez , j. j. 1993 , ph.d .
thesis , university of california kuntschner , h. 2000 , , in press ( astro - ph/0001210 ) rocha - pinto , h. j. & maciel , w. j. 1996 , , 279 , 447 trager , s. c. , et al .
2000 , , in press ( astro - ph/0001072 ) van dokkum , p. g. , et al . 1998 , , 504 , l17 worthey , g. 1994 , , 95 , 107 | a phenomenological model for chemical enrichment in early - type galaxies is presented , in which the process of star formation is reduced to a set of four parameters : star formation efficiency ( @xmath0 ) , fraction of ejected gas in outflows ( @xmath1 ) , formation redshift ( @xmath2 ) and infall timescale ( @xmath3 ) . out of these four parameters , only variations of @xmath1 or @xmath0
can account for the color - magnitude relation . a range of outflows results in a metallicity sequence , whereas a range of star formation efficiencies will yield a mixed age + metallicity sequence .
the age - metallicity degeneracy complicates the issue of determining which mechanism contributes the most ( i.e. outflows versus efficiency ) .
however , the determination of the slope of the correlation between mass - to - light ratio and mass in clusters at moderate or high redshift will allow us to disentangle age and metallicity .
# 1_#1 _ # 1_#1 _ = # 1 1.25 in .125 in .25 in |
studies of distant galaxies and active galactic nuclei ( agn ) require redshift measurements to determine distances and associated lookback times . for optically - bright objects
, redshifts can be measured efficiently with multi - slit spectrographs on large telescopes @xcite .
however , spectroscopic redshift measurements are very difficult for galaxies and agn that are at @xmath12 , obscured , or intrinsically faint . as a result
, we had to rely almost exclusively on photometric redshifts determined from broad - band photometry for galaxies with @xmath13 ( see , e.g. , * ? ? ? * ; * ? ? ? * , and many other studies ) .
the combo-17 survey @xcite pioneered the use of medium - bandwidth filters as a compromise between imaging and spectroscopy .
these filters sample the spectral energy distributions ( seds ) of galaxies at a resolution of @xmath14 and provide a redshift quality of 12% , intermediate between spectroscopy and broad - band imaging .
this improvement enables measurements of rest - frame colors and the environment of galaxies , and improves the accuracy of determinations of the physical properties of galaxies .
it also opens up the possibility to directly detect strong emission lines , which is particularly relevant for the identification of agn . with the development of medium - band filter sets in the near - ir @xcite and on 8 m class telescopes @xcite
, accurate redshifts are now becoming available for objects that are well beyond the limits of spectroscopy , and the full potential of this technique is being realized . in this paper , we present deep 18-band optical medium - band photometry from the subaru telescope in the @xmath0 `` extended '' chandra deep field - south , as part of the multiwavelength survey by yale - chile ( musyc ; @xcite ) .
the ecdf - s field has an extensive set of ancillary data ( see * ? ? ?
* ) , recently augmented by very deep _ spitzer _ irac data in the simple survey ( damen et al .
2010 , submitted ) .
furthermore , the field contains the goods - south field and the hubble ultra deep field .
our medium band survey is similar to that of ilbert et al .
( 2009 ) in the larger cosmos field , although we include 18 medium bands while there are only 12 available for the larger cosmos field .
we discuss the observations in section [ sec : obs ] and the details of data reduction in section [ sec : data ] .
we include ancillary data in our catalog from the musyc survey , which is described in section [ sec : adddata ] .
the details of our photometry are discussed and the public catalog is presented in section [ sec : phot ] .
we describe the photometric redshift determinations and evaluate the improvements provided by medium - band filters in section [ sec : redshifts ] .
finally in section [ sec : redseq ] we present the color - magnitude diagrams of the ecdf - s out to redshift 1.2 and investigate the occurrence of dusty galaxies on the red sequence . throughout this paper
we assume @xmath15 , @xmath16 and @xmath17 .
we carried out a medium - band imaging campaign on the subaru telescope using the wide field - of - view camera suprime - cam @xcite .
observations were collected in 6 runs over 2 years , in january , february and december 2006 , and january , march and december 2007 ; the observations are summarized in table [ tab : obs ] .
we show the response curves of the medium - band filters used in this study in figure [ response ] .
these filters were designed to maximize the accuracy of photometric redshifts by evenly sampling the wavelength range from 400 - 900 nm with an effective resolution of @xmath18 @xcite .
suprime - cam consists of ten 2k @xmath19 4k ccd chips that cover a large field of view ( 24@xmath20 27 ) . a single raw exposure in filter ia651
is shown in figure [ raw ] ; each of the 10 individual chips show large gradients in background ( from sensitivity variations and vignetting ) , as well as regions of bad pixels , and there are small gaps between the individual ccds .
therefore , our observing strategy consisted of taking multiple suprime - cam exposures , dithering the pointing slightly to cover the small chip gaps .
the median number of individual exposures for a given filter was 9 , but the number of exposures and the exposure time were varied depending on conditions at the telescope and the available time each night .
the long exposure on @xmath21 was designed to be used as a detection image for relevant science projects , but for our purposes here we prefer the significantly deeper musyc @xmath2 image as a detection image .
table [ tab : obs ] lists each filter ( 1 ) , the full - width at half maximum in nanometers ( 2 ) , its observation date ( 3 ) , the number of dithered exposures ( 4 ) , and the total exposure time in seconds ( 5 ) .
.medium band observations [ cols="^,^,^ , > , > " , ] comparing non - x - ray sources with high quality spectroscopic redshifts , we find a median accuracy @xmath22=@xmath23 out to z@xmath35 ( figure [ eazycomp ] , _ left _ ) . limiting ourselves to the subsample of sources with photometric redshifts @xmath24 , where the balmer break falls within our medium bands , we find @xmath25 ( figure [ eazycomp ] , _ right _ ) .
we report the nmad ] as in @xcite .
the normalization factor of 1.48 ensures the nmad of a gaussian distribution is equal to its standard deviation and the subtraction of median(@xmath26 ) corrects any offset from zero of the gaussian .
the nmad is a useful measure of dispersion because it is less sensitive to outliers that the standard deviation @xcite ] for comparison with the works of others ( e.g. , @xcite ) , and refer to it with the symbol @xmath27 .
we define outliers as 10@xmath28 , and find that @xmath35% of the overall sample are labeled as outliers . in table
[ tab : photzvmag ] , we show the photometric redshift quality as a function of galaxy magnitude ; listing the magnitude bin , number of objects found within the given magnitude bin , the percent of the galaxies in the bin with @xmath29 , the median 68% confidence interval for the galaxies in the bin as determined by eazy , the number of galaxies in the bin with high quality spectroscopic redshifts , the median @xmath30 , @xmath31 and the percent of galaxies in the bin with spectroscopically confirmed redshifts that are outliers . the number of high quality comparison spectroscopic redshifts is a strong function of magnitude , but overall we maintain median @xmath32 ( @xmath33 ) to an r - band magnitude of 25 .
the scatter in the photometric vs. spectroscopic redshifts almost certainly underestimates the actual uncertainty in the redshifts @xcite , and therefore we use the median 68% confidence intervals computed by eazy to compare the quality of the photometric redshifts for all galaxies .
this confidence interval is the difference between the 68% confidence upper and lower bounds on the photometric redshift computed from the probability distribution @xmath34 @xcite and is a strong function of galaxy apparent magnitude .
we find good quality photometric redshifts ( @xmath35 @xmath36 1 ; @xcite ) down to r - band magnitude of 25 for 70% of the sample ( 90% of the sample has @xmath37 ) , but miss a large fraction of the galaxy population at fainter magnitudes .
for many of the remaining sources poor fits are obtained by eazy because the photometry is too uncertain , due to faintness of the sources , or intrinsic variability in the source over the time period of which the photometry was taken @xcite .
there are also cases where the intrinsic sed may not be matched by the templates and/or degeneracies in color - z space result in multiple peaks in the redshift - probability distribution .
we note that when considering x - ray counterparts ( figure [ eazycomp ] , purple diamonds ) , our accuracy is maintained @xmath38 , but the outlier fraction increases to 18% .
we include the x - ray counterparts in the last line of table [ tab : photzvmag ] for comparison .
additionally , we compare the photometric redshift accuracy as a function of redshift in table [ tab : photzvsz ] . because the spectroscopic sample contains a variety of selection effects ( [ sec : specz ] ) , table [ tab : photzvsz ] includes the median 68% confidence intervals computed by eazy for all objects with @xmath35 @xmath36 1 .
we use photometric redshift to select the sources in each redshift bin so that we are comparing comparable sources in our measurements of @xmath39 and @xmath31 . comparing to the galaxies with spectroscopically
determined redshifts , the completeness of the photometric redshift determinations ( @xmath35 @xmath36 1 ) varies as a function of redshift .
we fit high quality photometric redshifts to over 90% of galaxies , but we note that the completeness falls to 80% in the interval between @xmath40 .
overall , with medium - band photometry we have achieved highly accurate photometric redshifts for the majority of the sources in the ecdf - s .
the photometric redshifts are available online through the musyc webpage .
we quantify the effect of the medium - band filters , comparing the photometric redshifts with and without the additional information these filters provide .
when excluding the medium - band filters , we retain the broad band optical filters , as well as the @xmath41 band and irac photometry .
overall the medium - band filters provide a factor of three improvement over photometric redshifts using broad - band filters alone , decreasing the median @xmath30 from 0.026 to 0.008 ( comparing sources with @xmath35 @xmath36 1 ) .
we note that with the addition of the medium band filters , high quality ( @xmath35 @xmath36 1 ) photometric redshifts are obtained for 20% more of the spectroscopic redshift sample .
the improvement is most noticeable at @xmath42 where the median@xmath43 falls from 0.022 to 0.006 and at @xmath44 where the median @xmath30 falls from 0.024 to 0.006 .
these redshift intervals are where the balmer break ( 3700 ) and lyman limit ( 912 ) fall inside the region of the spectrum covered by the medium - band filters .
however , we still see improvement at @xmath45 , where the median @xmath30 falls from 0.032 to 0.016 . in figure
[ eazycompnomed ] , we compare the spectroscopic sources with good fits to the seds ( @xmath35 @xmath36 1 ) with ( left ) and without ( right ) the medium - band filters .
the medium - band filters not only tighten the accuracy around the z_phot = z_spec line , but also can help to rule out false redshift solutions ( so called catastrophic failures ) for sources with z_spec @xmath46 1.2 . in each of these cases at @xmath47 ,
the broad band photometry alone is fit with by a balmer break feature in the sed at low redshift ( @xmath48 ) but when the medium band photometry is added , the same optical region of the sed is fit by a lyman alpha decrement caused by absorption by the igm and the photometric redshift increases to a value more consistent with the spectroscopic determination .
an example is shown in figure [ eazycompnomedsed ] , for a source with @xmath49 and @xmath50 .
we use two methods to identify the stars in our catalog , the first using a @xmath51 color selection @xcite and the second fitting stellar sed templates @xcite .
the @xmath51 diagram is traditionally used to select moderate redshift ( @xmath52 ) galaxies @xcite , but is also an efficient discriminating between stars and galaxies ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . because our filter set is slightly different than that used by @xcite , we apply the offsets determined by @xcite of -0.04 mag in ( z - k ) and 0.56 mag in ( b - z ) to our colors before plotting them in the bzk diagram . in figure
[ fig : bzkstars ] , we show the @xmath53 diagram for at all sources in our catalog with @xmath54 ( i.e. , @xmath55 ) .
the stellar sequence is clearly separated from the galaxies by the colors @xmath56 , shown as a solid black line in figure [ fig : bzkstars ] . for comparison
, we include the handful of spectroscopically identified stars ( large blue + in figure [ fig : bzkstars ] ) . only 2 of the 108 spectroscopically identified ( @xmath54 ) stars are missed by the @xmath53 star color selection . although this method of color selection is effective at identifying stars , we are limited to the sources bright enough to be detected in our relatively shallower k - band coverage . to identify stars in all of our objects with photometric redshifts , we evaluate a @xmath57 value for @xmath58 stellar sed template @xcite fits for each object . since the @xcite templates do not go past 2.5@xmath59 m , we do not include the _ spitzer _ irac bands in the stellar fits .
we flag as potential stars objects with reduced @xmath60 .
these sources are included in the @xmath53 diagram as blue crosses .
the stars identified through template fitting agrees well with those selected in our @xmath53 color - selection , @xmath390% of stellar - template stars with @xmath54 have @xmath53 colors of stars ( figure [ fig : bzkstars ] , dark blue crosses ) .
additionally , 80% of @xmath53-color - selected stars are also identified as stars by the stellar template fit . finally , we look at the subset of sources detected in the gems imaging @xcite . selecting sources with stellar fwhm from the gems imaging , @xmath380% of them
are selected by our template fitting method as stars , similar to the 84% of point like sources that the cosmos team identifies as stars through sed fitting @xcite .
furthermore , looking at bright ( @xmath61 ) extended sources in the gems imaging , we find fewer that 2% are misidentified as stars by this template fit classification .
( turquoise + ) and those which fit stellar seds ( dark blue x ) , mostly lie along the stellar sequence in the bottom region on the figure .
also indicated are high redshift bzk galaxies including star forming sbzk galaxies ( top left ) and passive pbzk galaxies ( top right ) .
overall , stars selected by template fitting to stellar seds are consistent with those indicated here on the bzk diagram.,scaledwidth=47.0% ] combo-17 produced the first highly accurate photometric redshifts of the ecdf - s region , and we wish to compare our own photometric redshifts to theirs @xcite .
we restrict ourselves to the brighter photometry ( r @xmath36 24 ) where their published photometric redshifts are more reliable ( 10% accuracy ) , restricting ourseleves to r @xmath3622 , we have fewer sources for comparison , but the accuracy of combo-17 ( 2% ) approaches our own @xcite .
to examine our consistency , we compare our photometric redshifts to those determined in combo-17 for all sources contained in both catalogs , finding @xmath62 at r @xmath36 22 and @xmath63 at r @xmath3624 .
therefore , our photometric redshift determinations are consistent with those determined by combo-17 to within their errors .
to compare our respective accuracies , we use spectroscopic sources ( figure [ fig : c17 ] ) , and compare z_spec with photometric redshifts for combo-17 and our musyc survey . for r @xmath36 22 , we find @xmath62 for combo-17 and @xmath64 for musyc , and considering all sources r@xmath6524 we find @xmath66 for combo-17 and @xmath67 for our musyc redshifts .
therefore we find that although our photometric redshifts agree with those from combo-17 survey to within their respective uncertainties , the larger number of deeper medium - band filter observations in the musyc survey improves the photometric redshift accuracy and extend this accuracy to fainter source populations .
we note that these differences come not only from the deeper medium - band data , but also by including jhk and irac photometry .
are more accurate for many sources that were outliers in combo-17 .
, scaledwidth=47.0% ]
here we demonstrate the quality of our data by showing the color - magnitude relation as a function of redshift in the e - cdfs .
the colors and morphologies of galaxies form a bimodal population @xcite .
this bimodality represents a fundamental relationship between mass and star formation history in galaxies , because the color of a galaxy is a proxy for the stellar age , where blue galaxies are younger and red galaxies are older .
the red sequence contains the brightest galaxies , which typically have early - type morphologies while the blue cloud galaxies typically have late - type morphologies .
relatively few galaxies are observed in the green valley , suggesting that galaxies spend only a short time here .
_ how _ the bimodal galaxy population was created is one of the outstanding questions in galaxy studies today .
rest - frame fluxes were computed following the procedure of @xcite , which measures the rest - frame fluxes from the best - fitting template @xcite .
this procedure is very different from a @xmath69-correction as it uses the observed medium bands that are closest in observed wavelength to the redshifted rest - frame band of interest .
the brammer et al .
( 2009 ) methodology is embedded in the eazy photometric redshift code ( brammer et al .
2008 ) . in figure
[ fig : colormag ] , we show the distribution of galaxies in the rest - frame @xmath70 versus @xmath71 plane .
this plot most directly relates observables to the bi - modal color sequence as a function of redshift . the restframe @xmath70 color covers the balmer break and
therefore is a good measure of recent star formation and the luminosity in the @xmath71 band is a proxy for stellar mass . in figure
[ fig : colormag ] , we also include the red sequence cut defined by @xcite in the combo-17 survey ( green dashed line ) and our completeness limits in each redshift bin ( blue solid line ) .
we see a sequence of red points separated from the main blue clump of galaxies at all redshifts in our sample , highlighting the quality of our data .
the location of this red sequence is consistent with that found by @xcite . in color space
both passive galaxies and dusty galaxies can appear red in @xmath70 colors . to investigate the nature of the red sequence ( figure [ fig : colormag ] ) , we use the irac photometry to separate out red passively evolving galaxies from red , dusty , star - forming galaxies .
the rest - frame @xmath72 color can distinguish between these two populations because dust - free galaxies are blue in v - j color , while the dust - obscured galaxies are still red @xcite .
following @xcite , we use a color - color diagram , rest - frame u - v vs v - j , to investigate the nature of the bimodal galaxy color sequence for all galaxies with 0.2 @xmath36 z @xmath36 1.2 ( figure [ uvvj ] ) .
we color - code the galaxies from the red sequence on the color - magnitude diagram red using @xcite s red sequence cut .
as expected , we find that passive and star forming galaxies separate cleanly in color - color space ( see * ? ? ?
we find that just over 20% of the galaxies identified as red sequence members in figure [ fig : colormag ] have rest - frame @xmath72 colors which place them in the star - forming sequence , consistent with previous work at these redshifts using morphological information ( e.g. , * ? ? ?
* ) and with the results in @xcite and whitaker et al.(2010 , in prep ) .
splitting our redshift interval in two ( @xmath73 and @xmath74 ) the fraction of red sequence galaxies with dusty colors is consistent between the two bins ( 21% and 23% respectively ) .
in this paper we present new deep 18-medium - band photometry in the well - studied e - cdfs field .
we reduced the raw data for the medium - band subaru suprime - cam observations using a combination of standard routines and custom tasks .
the ( public ) catalog includes photometry from 10 ground - based broad - band images ( @xmath75 , @xmath76 , @xmath77 , @xmath71 , @xmath78 , @xmath79 , @xmath80 , @xmath81 , @xmath82 , @xmath69 ) , 4 irac images ( @xmath83 @xmath59 m , @xmath84 @xmath59 m , @xmath85 @xmath59 m , @xmath86 @xmath59 m ) , and 18 medium - band images ( @xmath87 , @xmath88 , @xmath89 , @xmath90 , @xmath91 , @xmath92 , @xmath93 , @xmath94 , @xmath95 , @xmath96 , @xmath97 , @xmath98 , @xmath99 , @xmath100 , @xmath101 , @xmath102 , @xmath103 ) .
the full catalog provides multiwavelength seds for @xmath380,000 galaxies in the ecdf - s down to @xmath104}\sim27 $ ] ( 40,000 at [email protected] , the median depth of the medium band imaging ) , although the accuracy and completeness of the photometric redshifts declines at @xmath105 .
we computed accurate photometric redshifts using eazy , a public photometric redshift code .
the addition of the medium band proves a factor of four improvement in the photometric redshift accuracy over the use of broad band filters alone . comparing to spectroscopic redshifts
, we find a scatter in @xmath30 of 0.008 for the full sample , @xmath106 at @xmath107 where the balmer break is covered by the medium band filters and 0.01 at @xmath108 where the lyman limit ( 912 ) is covered by the medium band filters .
we find that the additional filters in the optical region even improve the photometric redshifts for sources at @xmath109 , decreasing the median @xmath30 from 0.03 to 0.02 .
we demonstrate that these photometric redshifts are sufficiently accurate to determine precise rest - frame colors of the galaxies .
we detect the bimodal galaxy distribution out to @xmath110 and find that 20% of the galaxies on the red - sequence have longer wavelength colors consistent with being dusty . in keeping with the spirit of the musyc collaboration , we provide public access to the images , photometry , and photometric redshifts . in future papers we will combine these accurate redshifts with the extensive ancillary data in the ecdf - s .
we thank the referee for helpful comments which improved this paper . support from nsf grants
ast-0407295 , ast-0449678 , ast-0807570 , nasa jpl grants 1277255 and 1282692 , and yale university is gratefully acknowledged .
support for the work of ks was provided by nasa through einstein postdoctoral fellowship grant number pf9 - 00069 issued by the chandra x - ray observatory center , which is operated by the smithsonian astrophysical observatory for and on behalf of nasa under contract nas8 - 03060 .
we also thank shunju sasaki for help with the imaging and data reductions . , n. 1999 , in astronomical society of the pacific conference series , vol .
191 , photometric redshifts and the detection of high redshift galaxies , ed .
r. weymann , l. storrie - lombardi , m. sawicki , & r. brunner , 31+ , g. a. , lira , p. , barrientos , l. f. , aguirre , p. , francke , h. , taylor , e. n. , quadri , r. , marchesini , d. , infante , l. , gawiser , e. , hall , p. b. , willis , j. p. , herrera , d. , & maza , j. 2008 , , 681 , 1099 , m. r. , hogg , d. w. , bahcall , n. a. , baldry , i. k. , brinkmann , j. , csabai , i. , eisenstein , d. , fukugita , m. , gunn , j. e. , ivezi , . , lamb , d. q. , lupton , r. h. , loveday , j. , munn , j. a. , nichol , r. c. , okamura , s. , schlegel , d. j. , shimasaku , k. , strauss , m. a. , vogeley , m. s. , & weinberg , d. h. 2003 , , 594 , 186 , a. , mignoli , m. , daddi , e. , pozzetti , l. , fontana , a. , saracco , p. , poli , f. , renzini , a. , zamorani , g. , broadhurst , t. , cristiani , s. , dodorico , s. , giallongo , e. , gilmozzi , r. , & menci , n. 2002 , , 392 , 395 , m. , faber , s. m. , newman , j. , phillips , a. c. , ellis , r. s. , steidel , c. c. , conselice , c. , coil , a. l. , finkbeiner , d. p. , koo , d. c. , guhathakurta , p. , weiner , b. , schiavon , r. , willmer , c. , kaiser , n. , luppino , g. a. , wirth , g. , connolly , a. , eisenhardt , p. , cooper , m. , & gerke , b. 2003 , in presented at the society of photo - optical instrumentation engineers ( spie ) conference , vol .
4834 , society of photo - optical instrumentation engineers ( spie ) conference series , ed .
p. guhathakurta , 161172 , e. , van dokkum , p. g. , gronwall , c. , ciardullo , r. , blanc , g. a. , castander , f. j. , feldmeier , j. , francke , h. , franx , m. , haberzettl , l. , herrera , d. , hickey , t. , infante , l. , lira , p. , maza , j. , quadri , r. , richardson , a. , schawinski , k. , schirmer , m. , taylor , e. n. , treister , e. , urry , c. m. , & virani , s. n. 2006 , , 642 , l13 , e. , van dokkum , p. g. , herrera , d. , maza , j. , castander , f. j. , infante , l. , lira , p. , quadri , r. , toner , r. , treister , e. , urry , c. m. , altmann , m. , assef , r. , christlein , d. , coppi , p. s. , durn , m. f. , franx , m. , galaz , g. , huerta , l. , liu , c. , lpez , s. , mndez , r. , moore , d. c. , rubio , m. , ruiz , m. t. , toft , s. , & yi , s. k. 2006 , , 162 , 1 , b. , mcintosh , d. h. , barden , m. , bell , e. f. , rix , h .- w .
, borch , a. , beckwith , s. v. w. , caldwell , j. a. r. , heymans , c. , jahnke , k. , jogee , s. , koposov , s. e. , meisenheimer , k. , snchez , s. f. , somerville , r. s. , wisotzki , l. , & wolf , c. 2007 , , 172 , 615 , o. , arnouts , s. , mccracken , h. j. , bolzonella , m. , bertin , e. , le fvre , o. , mellier , y. , zamorani , g. , pell , r. , iovino , a. , tresse , l. , le brun , v. , bottini , d. , garilli , b. , maccagni , d. , picat , j. p. , scaramella , r. , scodeggio , m. , vettolani , g. , zanichelli , a. , adami , c. , bardelli , s. , cappi , a. , charlot , s. , ciliegi , p. , contini , t. , cucciati , o. , foucaud , s. , franzetti , p. , gavignaud , i. , guzzo , l. , marano , b. , marinoni , c. , mazure , a. , meneux , b. , merighi , r. , paltani , s. , pollo , a. , pozzetti , l. , radovich , m. , zucca , e. , bondi , m. , bongiorno , a. , busarello , g. , de la torre , s. , gregorini , l. , lamareille , f. , mathez , g. , merluzzi , p. , ripepi , v. , rizzo , d. , & vergani , d. 2006 , , 457 , 841 , o. , capak , p. , salvato , m. , aussel , h. , mccracken , h. j. , sanders , d. b. , scoville , n. , kartaltepe , j. , arnouts , s. , floch , e. l. , mobasher , b. , taniguchi , y. , lamareille , f. , leauthaud , a. , sasaki , s. , thompson , d. , zamojski , m. , zamorani , g. , bardelli , s. , bolzonella , m. , bongiorno , a. , brusa , m. , caputi , k. i. , carollo , c. m. , contini , t. , cook , r. , coppa , g. , cucciati , o. , de la torre , s. , de ravel , l. , franzetti , p. , garilli , b. , hasinger , g. , iovino , a. , kampczyk , p. , kneib , j .-
p . , knobel , c. , kovac , k. , leborgne , j. f. , lebrun , v. , fvre , o. l. , lilly , s. , looper , d. , maier , c. , mainieri , v. , mellier , y. , mignoli , m. , murayama , t. , pell , r. , peng , y. , prez - montero , e. , renzini , a. , ricciardelli , e. , schiminovich , d. , scodeggio , m. , shioya , y. , silverman , j. , surace , j. , tanaka , m. , tasca , l. , tresse , l. , vergani , d. , & zucca , e. 2009 , , 690 , 1236 , i. , franx , m. , rudnick , g. , schreiber , n. m. f. , rix , h .- w . ,
moorwood , a. , van dokkum , p. g. , van der werf , p. , rttgering , h. , van starkenburg , l. , van de wel , a. , kuijken , k. , & daddi , e. 2003 , , 125 , 1107 , i. , huang , j. , franx , m. , rudnick , g. , barmby , p. , daddi , e. , van dokkum , p. g. , fazio , g. g. , schreiber , n. m. f. , moorwood , a. f. m. , rix , h .- w . , rttgering , h. , trujillo , i. , & van der werf , p. 2005
, , 624 , l81 , o. , vettolani , g. , garilli , b. , tresse , l. , bottini , d. , le brun , v. , maccagni , d. , picat , j. p. , scaramella , r. , scodeggio , m. , zanichelli , a. , adami , c. , arnaboldi , m. , arnouts , s. , bardelli , s. , bolzonella , m. , cappi , a. , charlot , s. , ciliegi , p. , contini , t. , foucaud , s. , franzetti , p. , gavignaud , i. , guzzo , l. , ilbert , o. , iovino , a. , mccracken , h. j. , marano , b. , marinoni , c. , mathez , g. , mazure , a. , meneux , b. , merighi , r. , paltani , s. , pell , r. , pollo , a. , pozzetti , l. , radovich , m. , zamorani , g. , zucca , e. , bondi , m. , bongiorno , a. , busarello , g. , lamareille , f. , mellier , y. , merluzzi , p. , ripepi , v. , & rizzo , d. 2005 , , 439 , 845 , g. p. , reed , r. , armandroff , t. , boroson , t. a. , & jacoby , g. h. 1998 , in society of photo - optical instrumentation engineers ( spie ) conference series , vol . 3355 , society of photo - optical instrumentation engineers ( spie ) conference series , ed .
s. dodorico , 577585 , m. , shimasaku , k. , okamura , s. , furusawa , h. , kashikawa , n. , ota , k. , doi , m. , hamabe , m. , kimura , m. , komiyama , y. , miyazaki , m. , miyazaki , s. , nakata , f. , sekiguchi , m. , yagi , m. , & yasuda , n. 2004 , , 611 , 685 , g. h. , young , e. t. , cadien , j. , engelbracht , c. w. , gordon , k. d. , kelly , d. m. , low , f. j. , misselt , k. a. , & et al .
2004 , in society of photo - optical instrumentation engineers ( spie ) conference series , vol .
5487 , society of photo - optical instrumentation engineers ( spie ) conference series , ed .
j. c. mather , 5061 , m. , hasinger , g. , ilbert , o. , zamorani , g. , brusa , m. , scoville , n. z. , rau , a. , capak , p. , arnouts , s. , aussel , h. , bolzonella , m. , buongiorno , a. , cappelluti , n. , caputi , k. , civano , f. , cook , r. , elvis , m. , gilli , r. , jahnke , k. , kartaltepe , j. s. , impey , c. d. , lamareille , f. , lefloch , e. , lilly , s. , mainieri , v. , mccarthy , p. , mccracken , h. , mignoli , m. , mobasher , b. , murayama , t. , sasaki , s. , sanders , d. b. , schiminovich , d. , shioya , y. , shopbell , p. , silverman , j. , smoli , v. , surace , j. , taniguchi , y. , thompson , d. , trump , j. r. , urry , m. , & zamojski , m. 2009 , , 690 , 1250 , l .- g . , riess , a. g. , dahlen , t. , livio , m. , panagia , n. , challis , p. , tonry , j. l. , filippenko , a. v. , chornock , r. , ferguson , h. , koekemoer , a. , mobasher , b. , dickinson , m. , giavalisco , m. , casertano , s. , hook , r. , blondin , s. , leibundgut , b. , nonino , m. , rosati , p. , spinrad , h. , steidel , c. c. , stern , d. , garnavich , p. m. , matheson , t. , grogin , n. , hornschemeier , a. , kretchmer , c. , laidler , v. g. , lee , k. , lucas , r. , de mello , d. , moustakas , l. a. , ravindranath , s. , richardson , m. , & taylor , e. 2004 , , 613 , 200 , e. n. , franx , m. , van dokkum , p. g. , quadri , r. f. , gawiser , e. , bell , e. f. , barrientos , l. f. , blanc , g. a. , castander , f. j. , damen , m. , gonzalez - perez , v. , hall , p. b. , herrera , d. , hildebrandt , h. , kriek , m. , labb , i. , lira , p. , maza , j. , rudnick , g. , treister , e. , urry , c. m. , willis , j. p. , & wuyts , s. 2009 , , 183 , 295 , e. , cristiani , s. , dickinson , m. , giavalisco , m. , kuntschner , h. , haase , j. , nonino , m. , rosati , p. , cesarsky , c. , ferguson , h. c. , fosbury , r. a. e. , grazian , a. , moustakas , l. a. , rettura , a. , popesso , p. , renzini , a. , stern , d. , & the goods team .
2008 , , 478 , 83 | we present deep optical 18-medium - band photometry from the subaru telescope over the @xmath0 extended chandra deep field - south ( ecdf - s ) , as part of the multiwavelength survey by yale - chile ( musyc ) .
this field has a wealth of ground- and space - based ancillary data , and contains the goods - south field and the hubble ultra deep field .
we combine the subaru imaging with existing @xmath1 and _ spitzer _ irac images to create a uniform catalog . detecting sources in the musyc `` @xmath2 '' image we find @xmath3 40,000 galaxies with @xmath4 , the median 5@xmath5 limit of the 18 medium bands .
photometric redshifts are determined using the eazy code and compared to @xmath6 spectroscopic redshifts in this field .
the medium band filters provide very accurate redshifts for the ( bright ) subset of galaxies with spectroscopic redshifts , particularly at @xmath7 and at @xmath8 . for @xmath7 ,
we find a @xmath9 scatter in @xmath10 of @xmath11 , similar to results obtained with a similar filter set in the cosmos field . as a demonstration of the data quality ,
we show that the red sequence and blue cloud can be cleanly identified in rest - frame color - magnitude diagrams at @xmath7 .
we find that @xmath320% of the red - sequence - galaxies show evidence of dust - emission at longer rest - frame wavelengths .
the reduced images , photometric catalog , and photometric redshifts are provided through the public musyc website . |
the basic key predistribution scheme of eschenauer and gligor @xcite has been recognized as a typical solution to secure communication in wireless sensor networks ( wsns ) and studied extensively in the literature over the last decade [ 1][11 ] .
the idea is that cryptographic keys are assigned before deployment to ensure secure sensor - to - sensor communications . the @xmath0-composite key predistribution scheme proposed by chan _
@xcite as an extension of the basic eschenauer
gligor scheme @xcite ( the @xmath0-composite scheme in the case of @xmath1 ) has received much interest @xcite since its introduction .
the @xmath0-composite scheme works as follows . for a wsn with @xmath3 sensors , prior to deployment , each sensor is independently assigned @xmath4 different keys which are selected _ uniformly at random _ from a pool of @xmath5 keys . then
two sensors establish a _ secure _ link in between after deployment _ if and only if _ they share at least @xmath0 key(s ) _ and _ the physical link constraint between them is satisfied . @xmath5 and @xmath4 are both functions of @xmath3 for generality , with the natural condition @xmath6 . examples of physical link constraints include the reliability of the transmission channel and the distance between two sensors close enough for communication
. the @xmath0-composite scheme with @xmath7 outperforms the basic eschenauer gligor scheme with @xmath1 in terms of the strength against small - scale network capture attacks while trading off increased vulnerability in the face of large - scale attacks @xcite . in this paper
, we investigate topological properties related to node degree in wsns employing the @xmath0-composite key predistribution scheme with general @xmath0 under the _ on_/_off _ channel model as the physical link constraint comprising independent channels which are either _ on _ or _
off_. the degree of a node @xmath8 is the number of nodes having secure links with @xmath8 ; and the minimum ( node ) degree of a network is the least among the degrees of all nodes .
specifically , we demonstrate that the number of nodes with certain degree is asymptotically equivalent in distribution to a poisson random variable , establish the asymptotic probability distribution for the minimum degree of the network , and derive the asymptotically exact probability for the property that the minimum degree is no less than an arbitrary value .
yaan @xcite and we @xcite consider the wsns with @xmath1 and show results for several topological properties , yet results about node degree in these prior work are weaker than our analytical findings even when the general @xmath0 is set as @xmath2 . our approach to the analysis is to explore the induced random graph models of the wsns .
as will be clear in section [ sec : systemmodel ] , the graph modeling a wsn under the @xmath0-composite scheme and the on / off channel model is an intersection of two distinct types of random graphs .
it is the intertwining @xcite of these two graphs that makes our analysis challenging .
we organize the rest of the paper as follows .
section [ sec : systemmodel ] describes the system model in detail .
afterwards , we present and discuss the results in section [ sec : res ] .
subsequently , we provide numerical experiments in section [ sec : expe ] to confirm our analytical results , whereas section [ related ] is devoted to relevant results in the literature .
next , we conclude the paper and identify future research directions in section [ sec : conclusion ] .
we elaborate the graph modeling of a wsn with @xmath3 sensors , which employs the @xmath0-composite key predistribution scheme and works under the on / off channel model .
we use a node set @xmath9 to represent the @xmath3 sensors . for each node @xmath10 ,
the set of its @xmath4 different keys is denoted by @xmath11 , which is uniformly distributed among all @xmath4-size subsets of a key pool of @xmath5 keys .
the @xmath0-composite key predistribution scheme is modeled by a uniform @xmath0-intersection graph denoted by instead of @xmath0 so we have uniform @xmath12-intersection graph @xmath13 .
this work uses @xmath0 following the @xmath0-composite key predistribution scheme @xcite . ]
@xmath14 , which is defined on the node set @xmath15 such that any two distinct nodes @xmath16 and @xmath17 sharing at least @xmath0 key(s ) ( an event denoted by @xmath18 ) have an edge in between . clearly , @xmath18 equals event @xmath19 $ ] , where @xmath20 with @xmath21 as a set means the cardinality of @xmath21 . under the on / off channel model ,
each node - to - node channel is independently _ on _ with probability @xmath22 and _ off _ with probability @xmath23 , where @xmath24 is a function of @xmath3 with @xmath25 .
denoting by @xmath26 the event that the channel between distinct nodes @xmath16 and @xmath17 is _ on _ , we have @xmath27 } = p_n$ ] , where @xmath28 $ ] denotes the probability that event @xmath29 happens , throughout the paper .
the on / off channel model is represented by an erds - rnyi graph @xmath30 @xcite defined on the node set @xmath15 such that @xmath16 and @xmath17 have an edge in between if event @xmath31 occurs .
finally , we denote by @xmath32 the underlying graph of the @xmath3-node wsn operating under the @xmath0-composite key predistribution scheme and the on / off channel model . we often write @xmath33 rather than @xmath34 for notation brevity .
graph @xmath33 is defined on the node set @xmath15 such that there exists an edge between nodes @xmath16 and @xmath17 if and only if events @xmath18 and @xmath31 happen at the same time .
we set event @xmath35 .
then @xmath33 can be seen as the intersection of @xmath36 and @xmath30 , i.e. , @xmath37 we define @xmath38 as the probability that two different nodes share at least @xmath0 key(s ) and @xmath39 as the probability that two distinct nodes have a secure link in @xmath33 .
clearly , @xmath38 and @xmath40 are the edge probabilities in graphs @xmath36 and @xmath33 , respectively . @xmath38 and
@xmath40 both depend on @xmath41 and @xmath0 , while @xmath40 also depends on @xmath24 . by definition
, @xmath42 is determined through @xmath43 = \sum_{u = q}^{k_n } \mathbb{p}[|s_{i } \cap s_{j}| = u ] , \label{psq1}\end{aligned}\ ] ] where it is shown @xcite that @xmath44 \nonumber \\ \quad & = \begin{cases } \frac{\binom{k_n}{u}\binom{p_n - k_n}{k_n - u}}{\binom{p_n}{k_n } } , & \textrm{for } \max\{0,2k_n - p_n\ } \leq u \leq k_n , \\ 0 , & \textrm{otherwise}. \end{cases } \label{psq2}\end{aligned}\ ] ] from @xmath45 and the independence of @xmath46 and @xmath47 , we obtain @xmath48 = \mathbb{p } [ { c}_{i j } ] \cdot \mathbb{p } [ \gamma_{i j } ] = p_n \cdot p_{s , q}. \label{eq_pre}\end{aligned}\ ] ]
we present and discuss our results in this section . throughout the paper
, @xmath0 is an arbitrary positive integer and does not scale with @xmath3 ; @xmath49 is the base of the natural logarithm function , @xmath50 .
all limits are understood with @xmath51 .
we use the standard asymptotic notation @xmath52 .
in particular , for two sequences @xmath53 and @xmath54 , @xmath55 signifies @xmath56 ; namely , @xmath53 and @xmath54 are asymptotically equivalent .
we detail the results of graph @xmath33 in theorem [ thm : exact_qcomposite2 ] and corollary [ cor : exact_qcomposite ] below .
the detailed proofs of theorem [ thm : exact_qcomposite2 ] and corollary [ cor : exact_qcomposite ] can be found in the full version @xcite and are omitted here owing to the space limitation .
the basic idea is to use the method of moments @xcite .
[ thm : exact_qcomposite2 ] for graph @xmath33 under @xmath57 and @xmath58 , the following properties ( a ) and ( b ) hold . * ( a ) * if @xmath59 then for @xmath60 , it follows that @xmath61 denoting the number of nodes with degree @xmath62 , is asymptotically equivalent in distribution to a poisson random variable with mean @xmath63 . in other words , with @xmath64 denoting a poisson random variable with mean @xmath65 , it holds that @xmath66 & \sim \mathbb{p } [ \textrm{po } ( \lambda_{n , h } ) = i ] , \quad\textrm{for } i=0,1,2,\ldots .
\nonumber\end{aligned}\ ] ] * ( b ) * if for some integer @xmath67 and some sequence @xmath68 satisfying @xmath69 it holds that @xmath70 then defining @xmath71 as the minimum degree of @xmath33 , we obtain : + for @xmath72 , it follows that as @xmath73 , @xmath74 and for @xmath75 , properties ( b1)(b4 ) below hold .
* @xmath76 0 as @xmath73 ; * if @xmath77 , then as @xmath78 , @xmath79 * if @xmath80 , then as @xmath78 , @xmath81 * if @xmath82 , then as @xmath78 , @xmath83 property ( b ) of theorem [ thm : exact_qcomposite2 ] presents the asymptotic probability distribution for the minimum degree of the network .
we present a corollary of theorem [ thm : exact_qcomposite2 ] below .
the corollary is established with the help of a graph coupling argument @xcite ( see the full version @xcite for the detailed proof ) .
[ cor : exact_qcomposite ] for graph @xmath33 under @xmath57 and @xmath58 , with some sequence @xmath84 defined by @xmath85 for some positive integer @xmath86 , and with @xmath71 denoting the minimum degree of @xmath33 , it holds that @xmath87 & = \begin{cases } e^{- \frac{e^{-\beta ^*}}{(k-1 ) ! } } , & \textrm{if } \lim\limits_{n \to \infty } \beta_n = \beta ^ * \in ( -\infty , \infty ) , \\ 1 , & \textrm{if } \lim\limits_{n \to \infty } \beta_n = \infty , \\ 0 , & \textrm{if } \lim\limits_{n \to \infty } \beta_n = -\infty .
\end{cases } \nonumber \end{aligned}\ ] ] corollary [ cor : exact_qcomposite ] presents the asymptotically exact probability and a zero one law @xcite ( a kind of phase transition @xcite ) for the event that graph @xmath33 has a minimum node degree no less than @xmath86 . setting @xmath24 to @xmath2 in theorem [ thm : exact_qcomposite2 ] and corollary [ cor : exact_qcomposite ]
, we obtain corresponding results @xcite for topological properties in uniform @xmath0-intersection graph @xmath36 . in theorem [ thm : exact_qcomposite2 ] and corollary [ cor : exact_qcomposite ] , given @xmath88 , we have @xmath89 for all @xmath3 sufficiently large since @xmath0 does not scale with @xmath3 . from @xmath58 ,
it is clear that @xmath90 for all @xmath3 sufficiently large .
we check the practicality of the conditions in theorem [ thm : exact_qcomposite2 ] and corollary [ cor : exact_qcomposite ] : @xmath88 , @xmath91 , and ( [ peq1])([peq2 ] ) .
first , the condition @xmath88 follows in wireless sensor network applications @xcite since @xmath92 is often logarithmic @xcite with @xmath3 , the number of sensor nodes in the network .
second , the condition @xmath91 also holds in practice since the key pool size @xmath5 is expected to be several orders of magnitude larger than @xmath4 @xcite . finally , ( [ peq1])([peq2 ] ) present the range of @xmath93 that is of interest .
analogous results to those of theorem [ thm : exact_qcomposite2 ] and corollary [ cor : exact_qcomposite ] can be given with @xmath94 at all places substituted by @xmath95 , due to @xmath96 from ( [ eq_pre ] ) and the replacement of @xmath42 by @xmath97 given lemma [ lem_eval_psq ] below .
however , _
extra _ conditions have to be added for some results .
the details as well as the proof of lemma [ lem_eval_psq ] are provided in the full version @xcite .
[ lem_eval_psq ] if @xmath88 and @xmath58 , then it follow that @xmath98 . for @xmath99 in graph @xmath100 with @xmath101 , @xmath102 , @xmath103 , @xmath104 and @xmath105 .
, scaledwidth=27.0% ] has a minimum node degree at least @xmath86 as a function of @xmath106 for @xmath107 and @xmath108 with @xmath109 , @xmath101 , @xmath103 , and @xmath110 .
, scaledwidth=27.0% ]
to confirm the results in theorem [ thm : exact_qcomposite2 ] , we now provide numerical experiments in the non - asymptotic regime . as we will see from the simulation results , the experimental observations are in agreement with our theoretical findings . in all experiments ,
we fix the number of nodes at @xmath101 and the key pool size at @xmath103 . in figure
[ f4 ] , we plot the probability distribution for the number of nodes with degree @xmath62 in graph @xmath100 for @xmath111 from both the simulation and the analysis , with @xmath102 , @xmath112 and @xmath105 . on the one hand , for the simulation , we generate @xmath113 independent samples of @xmath100 and record the count ( out of a possible @xmath113 ) that the number of nodes with degree @xmath62 for each @xmath62 equals a particular non - negative number @xmath114 ( m is the horizontal axis in figure [ f4 ] )
. then the empirical probabilities are obtained by dividing the counts by @xmath113 .
on the other hand , we approximate the analytical curves by the asymptotic results as explained below .
property ( a ) of theorem [ thm : exact_qcomposite2 ] notes that with the parameter conditions therein , the number of nodes in @xmath115 with degree @xmath62 is asymptotically equivalent in distribution to a poisson random variable with mean @xmath116 .
we derive @xmath117 by computing the corresponding probability of @xmath40 in @xmath100 through @xmath118 \label{eq : thresholdsab}\end{aligned}\ ] ] given ( [ psq1][eq_pre ] ) and @xmath119 . then for each @xmath62
, we plot a poisson distribution with mean @xmath117 as the curve corresponding to the analysis .
we observe that the curves generated from the simulation and those obtained by the analysis are close to each other , confirming the result on asymptotic poisson distribution in property ( a ) of theorem [ thm : exact_qcomposite2 ] . in figure [ f5 ] ,
we depict the probability that graph @xmath100 has a minimum node degree at least @xmath86 from both the simulation and the analysis , for @xmath107 and @xmath108 with @xmath120 and @xmath110 and @xmath106 varying from 29 to 36 ( we still set @xmath101 and @xmath103 ) .
similar to the experiments for figure [ f4 ] above , we also generate @xmath113 independent samples of graph @xmath100 and record the count that the minimum degree of graph @xmath100 is no less than @xmath86 ; and the empirical probability of @xmath100 having a minimum degree at least @xmath86 is derived by averaging over the @xmath113 experiments .
the analytical curves in figure [ f5 ] are also approximated by the asymptotical results as follows .
first , we compute the corresponding probability of @xmath40 in @xmath100 through ( [ eq : thresholdsab ] ) . then based on ( [ peq2 ] ) , we determine @xmath121 through @xmath122 . then with an approximation to the asymptotical results in corollary [ cor : exact_qcomposite ] , we plot the analytical curves by considering that the minimum degree of @xmath100 is at least @xmath123 with probability @xmath124 .
the observation that the curves generated from the simulation and the analytical curves are close to each other is in accordance with corollary [ cor : exact_qcomposite ] .
erds and rnyi @xcite propose the random graph model @xmath125 defined on a node set with size @xmath3 such that an edge between any two nodes exists with probability @xmath24 _ independently _ of all other edges .
for graph @xmath125 , erds and rnyi @xcite derive the asymptotically exact probabilities for connectivity the property that the minimum degree is at least @xmath2 , by proving first that the number of isolated nodes converges to a poisson distribution as @xmath73 .
later , they extend the results to general @xmath86 in @xcite , obtaining the asymptotic poisson distribution for the number of nodes with certain degree and the asymptotically exact probabilities for @xmath86-connectivity and the event that the minimum degree is at least @xmath86 , where @xmath86-connectivity is defined as the property that the network remains connected in spite of the removal of any @xmath126 nodes - connectivity throughout this paper means @xmath86-vertex - connectivity @xcite . ] . since its introduction , graph @xmath125 has been widely investigated @xcite . for graph @xmath36 ,
bloznelis _ et al .
_ @xcite demonstrate that a connected component with at at least a constant fraction of @xmath3 emerges asymptotically when the edge probability @xmath40 exceeds @xmath127 .
bloznelis and uczak @xcite have recently considered connectivity and perfect matching . still in @xmath36 ,
bloznelis _ et al . _
@xcite investigate assortativity and clustering , while for asymptotic node degree distribution , bloznelis @xcite analyzes clustering coefficient and the degree distribution of a typical node .
we @xcite compute the probability distribution for the minimum node degree .
recently , bloznelis and rybarczyk @xcite and we @xcite have derived the asymptotically exact probability of @xmath86-connectivity .
several variants or generalizations of graph @xmath128 are also considered in the literature @xcite .
when @xmath1 , for graph @xmath129 ( also referred to as a random key graph @xcite or a uniform random intersection graph @xcite ) and some of its variants , a number of properties have been extensively studied in the literature including component evolution @xcite , connectivity @xcite , @xmath86-connectivity @xcite , node degree distribution @xcite and independent sets @xcite . in graph
@xmath130 , yaan @xcite presents zero one laws for connectivity and for the property that the minimum degree is at least @xmath2 .
we extend yaan s results to general @xmath86 for @xmath130 in @xcite .
krishnan _ et al .
_ @xcite and krzywdziski and rybarczyk @xcite describe results for the probability of connectivity asymptotically converging to 1 in wsns employing the @xmath0-composite key predistribution scheme with @xmath1 ( i.e. , the basic eschenauer gligor key predistribution scheme ) , not under the on / off channel model but under the well - known disk model @xcite , where nodes are distributed over a bounded region of a euclidean plane , and two nodes have to be within a certain distance for communication .
simulation results in our work @xcite indicate that for wsns under the key predistribution scheme with @xmath1 , when the on - off channel model is replaced by the disk model , the performances for @xmath86-connectivity and for the property that the minimum degree is at least @xmath86 do not change significantly .
in this paper , we analyze several topological properties related to node degree in a wireless sensor network operating under the @xmath0-composite key predistribution scheme with on / off channels .
the network is modeled by the superposition of an erds - rnyi graph on a uniform @xmath0-intersection graph .
numerical simulation is shown to be in agreement with our theoretical findings .
two future research directions are as follows . to begin with ,
we can consider physical link constraints different with the on / off channel model , where one candidate is the aforementioned disk model .
another extension is to derive the asymptotically exact probability and thus a zero one law for @xmath86-connectivity in graph @xmath33 .
note that a zero law for @xmath86-connectivity follows immediately from corollary [ cor : exact_qcomposite ] since @xmath86-connectivity implies the property of minimum node degree being at least @xmath86 .
the one law and the asymptotically exact probability result will follow if we show that under certain conditions , the probability that @xmath33 has a minimum node degree no less than @xmath86 but is not @xmath86-connected converges to @xmath131 as @xmath78 .
j. zhao , o. yaan , and v. gligor .
@xmath86-connectivity in secure wireless sensor networks with physical link constraints the on / off channel model . ,
available online at + http://arxiv.org/abs/1206.1531 [ http://arxiv.org/abs/1206.1531 ] j. zhao , o. yaan , and v. gligor .
topological properties of wireless sensor networks under the @xmath0-composite key predistribution scheme with unreliable links .
technical report cmu - cylab-14 - 002 , cylab , carnegie mellon university , january 2014 .
available online at + http://www.cylab.cmu.edu/files/pdfs/tech_reports/cmucylab14002.pdf [ http://www.cylab.cmu.edu/files/pdfs/tech_reports/cmucylab14002.pdf ] j. zhao , o. yaan , and v. gligor . on @xmath86-connectivity and minimum vertex degree in random @xmath12-intersection graphs .
available online at http://www.andrew.cmu.edu/user/junzhao/papers/s-int-k-con.pdf [ http://www.andrew.cmu.edu/user/junzhao/papers/s-int-k-con.pdf ] j. zhao , o. yaan , and v. gligor .
results on vertex degree and @xmath86-connectivity in uniform @xmath12-intersection graphs .
technical report cmu - cylab-14 - 004 , cylab , carnegie mellon university , january 2014 .
available online at + http://www.cylab.cmu.edu/files/pdfs/tech_reports/cmucylab14004.pdf [ http://www.cylab.cmu.edu/files/pdfs/tech_reports/cmucylab14004.pdf ] m. bradonji , a. hagberg , n. w. hengartner , n. lemons , and a. g. percus .
the phase transition in inhomogeneous random intersection graphs . , 2013 .
available online at + { http://arxiv.org/abs/1301.7320 } [ \{http://arxiv.org / abs/1301.7320 } ] . | the @xmath0-composite key predistribution scheme @xcite is used prevalently for secure communications in large - scale wireless sensor networks ( wsns ) .
prior work @xcite explores topological properties of wsns employing the @xmath0-composite scheme for @xmath1 with unreliable communication links modeled as independent on / off channels . in this paper
, we investigate topological properties related to the node degree in wsns operating under the @xmath0-composite scheme and the on / off channel model .
our results apply to general @xmath0 and are stronger than those reported for the node degree in prior work even for the case of @xmath0 being @xmath2 .
specifically , we show that the number of nodes with certain degree is asymptotically equivalent in distribution to a poisson random variable , present the asymptotic probability distribution for the minimum degree of the network , and establish the asymptotically exact probability for the property that the minimum degree is at least an arbitrary value .
numerical experiments confirm the validity of our analytical findings .
@ps@headingsps .
@headings @ps@ieeetitlepagestyleps .
@ieeetitlepagestyle # 1 ps .
@headings @ps@headings oddhead # 1 evenhead # 1 ps .
@ieeetitlepagestyle @ps@headings oddhead # 1 evenhead # 1 ps .
@headings random intersection graph , random key graph , s - intersection graph , connectivity , node degree , key predistribution , wireless sensor network . |
the recent advances in the manipulation of ultracold gases have made it possible to employ these systems to accurately simulate several non - trivial problems in condensed matter physics .
as examples , we may mention the exploration of the bcs to bec crossover @xcite and the superfluid to mott - insulator transition @xcite .
disorder plays an important role in the theory of solid state , affecting in a substantial way the transport properties of various systems .
special attention has in the past been dedicated to studies of light propagation in strongly - scattering powders and electron transport in the presence of impurities @xcite .
it looks therefore interesting to introduce a controlled disorder in the experiments with ultracold atoms , in order to provide a closer modeling of realistic systems of condensed matter physics .
it was indeed predicted that atomic gases stored in optical lattices would be good candidates to experimentally observe the effect of a disordered or quasi - periodic potential on an interacting bose gas or on interacting fermi - bose mixture @xcite .
first experimental results along this research line have been recently reported , i.e. the observation of a bose glass in a quasi - periodic potential @xcite and the study of spatial coherence properties of an interacting bose gas trapped in a lattice in presence of a disordered ensemble of fermionic atoms @xcite . in this paper
, we consider a variant of this line of research , that is the possibility of looking for genuine localized states of a non - interacting matter wave exposed to static disorder in continuous space .
localized states are stationary states with a square integrable wave function at an energy where the classical motion is not bounded spatially . in a paper that dates back to the early years of quantum mechanics ,
von neumann and wigner @xcite showed that the schrdinger equation can admit square integrable eigenstates embedded in the continuum of states with energy higher than the maximum of the potential .
after the work of anderson @xcite , it is expected that disordered potentials can generically lead in 3d to a quantum phase transition , a macroscopic number of localized states being present at low energy .
such a phase transition in 3d is not straightforward to observe , as it is sensitive to decoherence and wave absorption effects , and requires a mean free path of the wave @xmath0 smaller than its wave length @xmath1 , as stated by the ioffe - regel criterion @xmath2 the study of localization of light is a well developed experimental subject : strong localization of light has been reported in semi - conductor powders @xcite , and weak localization effects of light in a gas of cold atoms are the subject of an intensive experimental study @xcite . on the contrary , for matter waves , no direct evidence of localization was obtained in 3d .
matter waves made of ultracold atoms are good candidates in this respect , due to their weak coupling to the environment and to the possibility of tuning their interactions with a feshbach resonance @xcite .
an open problem is however to know if strong enough disorder can be introduced in these gases to lead to reasonably short localization lengths in 3d .
a natural way to produce a disordered potential in atomic gases is to use the speckle pattern of a laser beam @xcite .
many experiments on bose - einstein condensates in 1d random optical potentials have very recently been reported @xcite , and they provide evidence for disorder - related effects such as fragmentation of the condensate , suppression of diffusion , frequency shifts and damping of collective oscillations .
theoretically , these effects were discussed in @xcite . genuine strong localization in 1d , in the non - interacting regime ,
has not been reported yet in these experiments , and the implementation of the disordered optical potential in 3d remains to be done .
also the theoretical analysis of matter wave localization in a speckle pattern , recently performed in 2d @xcite , has not been done in a detailed way in the 3d strong localization regime .
( color online ) sketch of the model of disorder considered here : a matter wave ( a , blue ) scatters on randomly distributed b particles ( red ) , each occupying the vibrational ground state of a node in a 3d optical lattice ( here the average occupancy @xmath3 is 0.1 ) .
the lattice does not act on the matter wave a. ] an alternative method to realize a disordered potential was proposed in @xcite : a matter wave , made of atoms of a species @xmath4 , scatters off a set of atoms of another species @xmath5 that are trapped in randomly chosen sites of a deep optical lattice with no multiply occupied sites ( see fig .
[ cap : sketch ] ) . as we will choose the lattice to be very far detuned for the species @xmath4 , the matter wave moves unaffected through the optical lattice , and interacts only with the @xmath5 atoms : this excludes classical localization effects in local potential minima .
the disorder can be made very strong , since ( i ) the correlation length of the disorder can be as small as @xmath6 m ( the spatial period of the optical lattice ) , and ( ii ) the scattering cross section of the matter wave off a single @xmath5 atom can be made as high as allowed by quantum mechanics ( the so - called unitary limit ) by use of a @xmath7 interspecies feshbach resonance , making it possible to dramatically reduce the mean free path of the matter wave .
furthermore , as we shall take advantage of , this model allows a straightforward exact numerical analysis even in 3d , when @xmath5 atoms are modeled as fixed point - like scatterers , as is known for light waves @xcite .
it is the 3d version of this scheme that we analyze in this paper .
after the detailed presentation of our model and its practical implementation given in sec .
[ sec : our - proposal ] , we show numerically in sec .
[ sec : localized - states ] that it leads to the appearance of a large number of localized states for a range of parameters accessible in present experiments , provided that the effective coupling between the matter wave and a single scatterer is tuned to a value of the order of the mean scatterer separation .
section [ sec : effective ] is dedicated to the quantitative description of the scattering between the matter wave and a single trapped scatterer , and to the discussion of the confinement - induced resonances thereby arising : we show indeed that large enough effective coupling constants can be obtained .
our conclusions are presented in sec .
[ sec : conclusions ] , including a discussion of a possible strategy of observation of strong localization .
the scatterers are a set of identical particles , whose chemical species and quantum numbers will be indexed by the letter @xmath5 , randomly occupying ( with filling factor @xmath8 ) the sites of a 3d cubic optical lattice . the potential seen by the @xmath5 atoms
is produced by a superposition of three laser standing waves of common intensity and mutually orthogonal linear polarizations along the @xmath9 , @xmath10 and @xmath11 axes,@xmath12,\label{eq : opticallattice}\ ] ] where @xmath13 is the modulation depth of the lattice and @xmath14 is the laser wavenumber .
we shall denote the lattice spatial period along each axis by @xmath15 .
multiple occupation of a lattice well is assumed to be absent , by use of polarized fermions , or by creation of vacancies in a unit occupancy mott phase state @xcite , or simply by choosing @xmath3 sufficiently small to make it statistically irrelevant .
we choose the lattice depth @xmath16 to be much larger than the recoil energy @xmath17 of the @xmath5 atoms so that the tunneling time of @xmath5 atoms from one lattice site to another ( @xmath18 for @xmath19 ) is negligible over the duration of the experiment and the disordered spatial pattern of @xmath5 atoms is static @xcite .
we also require that each @xmath5 atom is prepared in the vibrational ground state of the local lattice micro - trap , which can be achieved in practice by raising adiabatically the optical lattice on top of a condensate cloud of atoms @xmath5 @xcite , or by applying raman laser cooling sideband techniques @xcite to an optical molasses .
this condition is crucial to ensure that each @xmath7 scattering event is elastic when the @xmath4 atoms have sufficiently low energy : indeed , energy conservation guarantees that the @xmath5 atom is left in the vibrational ground state after scattering with a @xmath4 atom of momentum @xmath20 if@xmath21 where @xmath22 is the oscillation frequency of a @xmath5 atom in a micro - trap .
a last point is to ensure that spontaneous emission processes are negligible for the @xmath5 atoms . in order to achieve large values of @xmath16 with negligible heating of the trapped scatterers ,
we require the lattice to be blue - detuned with respect to the strongest transition of the @xmath5 atoms ( in blue - detuned lattices , particles are trapped in the minima of intensity of the stationary light field ) . including the lamb - dicke type reduction factor coming from the trapping of @xmath5 atoms close to the nodes of the laser field , one gets the fluorescence rate @xmath23 where @xmath24 is the atom - laser detuning and @xmath25 is the spontaneous emission rate of @xmath5 atoms . for the bosonic @xmath26rb isotope of rubidium ( @xmath27 nm , @xmath28 nm ) and an optical lattice tuned at @xmath29 nm , only at @xmath30 nm to the blue of the strongest rubidium transition at 780 nm , at the required lattice intensity @xmath19 the tunneling time is @xmath31s @xcite and the fluorescence rate is @xmath32 , allowing experimental times up to @xmath33ms . the same calculation , taking for @xmath5 the fermionic isotope @xmath34k of potassium ( @xmath35 nm , @xmath36 nm ) and an optical wavelength @xmath37 nm leads to @xmath38s and @xmath39 . the matter wave to be strongly - localized is made of atoms of another species , that we will label by @xmath4 .
we shall ignore interaction effects among these @xmath4 atoms .
one way to fulfill this condition in a real experiment would be to take spin polarized fermionic atoms : @xmath40-wave interactions are prohibited by the exclusion principle and @xmath3-wave interactions are very weak at low energies in the absence of a @xmath3-wave resonance .
the @xmath4 atoms experience interactions with the trapped species @xmath5 . at low incoming kinetic energy of a @xmath4 particle
, we model these interactions by static contact potentials , corresponding to infinitely - massive point - like scatterers , each located at the center of a micro - well occupied by a @xmath5 atom : @xmath41 where the sum is taken over the @xmath42 scatterers .
the effective scattering length @xmath43 of a @xmath4 atom on a trapped @xmath5 atom , when expressed in units of the harmonic - oscillator length @xmath44 , is a function of the dimensionless ratios @xmath45 and @xmath46 , @xmath47 being the @xmath7 scattering length in free space . the value of @xmath48 and the validity condition of our model potential , eq .
( [ eq : veff ] ) , will be given in sec .
[ sec : effective ] .
the @xmath4 atoms also experience the optical lattice potential , with the same spatial dependence as in eq .
( [ eq : opticallattice ] ) but with a different modulation amplitude @xmath49 .
we require the optical lattice to be much closer to resonance with @xmath5 atoms than with @xmath4 atoms , @xmath50 , such that @xmath49 will be much smaller than @xmath16 .
in particular , we impose that @xmath51 so that , in the absence of @xmath5 atoms , the @xmath4 atoms can be safely considered as free . in this respect , a particularly promising combination is given by fermionic @xmath52li for the species @xmath4 ( @xmath53 nm ) and @xmath26rb for the species @xmath5 : taking @xmath29 nm and a laser intensity such that @xmath19 , one finds @xmath54 . if one takes for @xmath5 the fermionic @xmath34k with @xmath55 nm , one finds @xmath56 @xcite .
we shall therefore neglect the effect of the optical lattice on the @xmath4 atoms and take as a model hamiltonian for the matter wave : @xmath57 we note in passing that in the original anderson model the @xmath4 particles were instead assumed to be in the tight - binding regime , so that the strong localization reported in this paper is not _
stricto sensu _
anderson localization . to end this section , we briefly discuss two effects not included in our model hamiltonian that may impose limitations in a real experiment .
as we shall see , the production of localized states with a short localization length ( of the order of the lattice spacing @xmath58 ) requires the use of a large and positive value of @xmath59 , obtained by a @xmath7 feshbach resonance . as a consequence ,
the matter wave @xmath4 has a weakly bound state with a trapped atom @xmath5 , of spatial extension @xmath60 .
a first undesired effect is therefore the formation of such @xmath7 dimers .
a first stage that may lead to a dimer production is during the feshbach ramp of @xmath43 from @xmath61 to @xmath62 .
this may be avoided by using a ramping time longer than the inverse of the dimer binding frequency , @xmath63s for our previous example with lithium and rubidium . once @xmath43
is set to @xmath62 , one may fear that three - body collisions @xmath64 lead to the formation of a dimer . for our model hamiltonian eq .
( [ eq : model ] ) the @xmath4 particles are an ideal gas and the dimer formation does not happen : the trapped @xmath5 particle , being replaced by a fixed scatterer with no degree of freedom , can not mediate a @xmath65 interaction . in the opposite limit where the @xmath5 scatterer is supposed to move freely , the rate of dimer formation per @xmath5 atom is @xmath66 , where we used eqs .
( 11,12 ) of @xcite , taking a dimer binding energy @xmath67 with @xmath68 the @xmath7 reduced mass , @xmath69 , and assuming that the @xmath4 atoms are degenerate fermions of fermi momentum @xmath70 with @xmath71 ; the constant @xmath72 is @xmath73 for the @xmath74li , @xmath75rb case , resulting for @xmath76 in a dimer formation rate much smaller than e.g. the @xmath5 fluorescence rate @xmath77 due to the lattice @xcite .
the calculation of the actual dimer formation rate in our model , taking into consideration the trapping of the @xmath5 atoms , requires the solution of a three - body problem with no center - of - mass separability , which is beyond the scope of the present work .
a second undesired effect is the one of gravity .
if the lattice is arranged to be stationary in a free - falling frame , this frame having initially an upward velocity component @xmath78 in the lab frame and finally a downward velocity component @xmath78 , the overall vertical motion of the lattice in such a fountain - like experiment is less than 3 cm for a total time of 150 ms .
longer times may be obtained if one compensates gravity , e.g. by a using the inflexion point of the optical potential produced by a far - detuned gaussian laser beam , or by using electro - optical potentials @xcite .
a drastic solution is of course to perform the experiment in a micro - gravity environment @xcite .
the ioffe - regel criterion eq .
( [ eq : ircrit ] ) is considered as a necessary condition to achieve strong localization @xcite .
it is simple to check that the setup that we consider can satisfy this criterion for experimentally reasonable parameters . as we will see in sec .
[ sec : effective ] , the effective scattering length presents confinement induced resonances that allow one to reach the unitary regime for the interaction of the matter wave with a trapped @xmath5 atom , with the maximal cross - section @xmath79 . in this case
, the ioffe - regel criterion reads : @xmath80 which for a filling factor @xmath81 yields @xmath82 , achievable with sub - recoil laser cooling techniques , or with a low density degenerate fermi gas with a fermi wavevector @xmath83 .
since the ioffe - regel criterion is not proved to be sufficient , we numerically investigate in this section the possibility for the disordered model hamiltonian eq .
( [ eq : model ] ) to lead to matter wave localization .
in particular , we shall find that the unitary regime @xmath84 is not the most favorable one .
a criterion of strong localization presented by kramer and mackinnon @xcite for electrons in a solid consists in showing that , at the fermi energy @xmath85 , off - diagonal elements of the resolvent @xmath86 in real space decrease exponentially with the distance between the two considered points . for our model hamiltonian
, the calculation of the matrix elements of the resolvent is straightforward to implement numerically , using a technique well known for scalar light waves in a gas of scatterers @xcite .
these matrix elements are indeed given in presence of @xmath42 point - like scatterers by @xmath87_{jl}g_{0}\left(\mathbf{r}_{l}-\mathbf{r}^{\prime}\right).\label{eq : r - g - rprime}\end{aligned}\ ] ] here @xmath88 is the propagator in free space of a particle of positive energy @xmath89 : @xmath90 and we have introduced the @xmath91 matrix @xmath92 : @xmath93 where @xmath94 is the identity matrix and @xmath95 a complex symmetric ( not hermitian ) matrix with elements defined by @xmath96 the _ exact _ calculation of the resolvent in coordinates space is in this way reduced to the inversion of the @xmath91 matrix @xmath92 @xcite .
we have implemented the criterion by kramer and mackinnon for a variable energy @xmath97 , and indeed we found an exponential decay of @xmath98 for sufficiently low energies . however , as shown in the appendix [ appen : kinnon ] , this rapid decay is not a proof of localization but may be due to the fact that @xmath97 is in a spectral gap of the system @xcite . the most direct way to prove
localization is to exhibit stationary states that are ` localized ' inside the disordered potential , that is with a wave function strongly peaked inside the scattering medium , decreasing exponentially towards the borders of the scattering medium . to this end
, we use the fact that the wave function @xmath99 when not identically zero , is an exact eigenstate of @xmath100 with energy @xmath97 , whatever the arbitrary location of its center @xmath101 @xcite . here
we take @xmath102 so that @xmath103 belongs to the continuum of the energy spectrum of @xmath100 , like the scattering states .
a useful expression of @xmath104 is then @xmath105 \label{eq : phi_jolie}\ ] ] where @xmath106 is a constant factor and @xmath107 . in practice
, we choose @xmath108 inside the scattering medium ; to see if @xmath104 is localized or not , one just has to compare the values of @xmath104 inside and outside the scattering medium ; one can even watch how the modulus of @xmath104 decays inside the medium .
we test the proposed technique in 1d , where analytical results can be obtained .
we assume that a quantum particle of mass @xmath109 interacts in 1d with @xmath42 point - like scatterers that are randomly distributed in the interval @xmath110 $ ] . using the transfer matrix formalism , as detailed in the appendix [ appen:1d ]
, we can obtain exact expressions for the matrix elements of the resolvent @xmath111 as a function of @xmath9 .
assuming that there is no scatterer in @xmath112 , we obtain @xmath113 where @xmath114 is such that @xmath115 , @xmath116 is the complex reflection coefficient for the set of scatterers in the half - space @xmath117 oriented from @xmath112 to @xmath118 , and @xmath119 is the complex reflection coefficient for the set of scatterers in the half - space @xmath120 oriented from @xmath112 to @xmath121 . in the half - space @xmath122
we also have a simple expression : @xmath123 where @xmath124 is the transmission coefficient of the set of scatterers in the half - space @xmath120 oriented from @xmath125 to @xmath126 .
a similar expression holds for @xmath127 , see appendix [ appen:1d ] . using furstenberg theorem @xcite , we know that in the thermodynamic limit @xmath128 , with a fixed density of scatterers , the modulus of the transmission coefficients tends exponentially to zero , @xmath129 where @xmath130 is the localization length , so that the modulus of @xmath119 and @xmath116 tends to one .
we then find that the wavefunction @xmath131 outside the scattering medium is of the order of @xmath132 . to calculate @xmath104 inside the medium , we first take , to zeroth order in @xmath132 , @xmath133 : eq .
( [ eq : psi1d_in ] ) then leads to a purely real @xmath134 ( of order unity ) , that is to a vanishing @xmath135 .
going to next order , @xmath136 leads to @xmath137 . for a generic value of @xmath20
, the wavefunction @xmath138 has therefore the behavior depicted in fig.[cap : schema_1d]a .
however , for specific values @xmath139 of @xmath20 , the above reasoning is incorrect .
assume that @xmath20 is such that @xmath140 is a real and positive number .
then the denominator @xmath141 of eqs.([eq : psi1d_in],[eq : psi1d_out ] ) is extremely small , of order @xmath142 .
this leads to @xmath143 decreasing exponentially from @xmath112 outwards , from a value @xmath144 to a value @xmath145 @xcite , as depicted in fig.[cap : schema_1d]b . in this case
, @xmath138 corresponds to a ` localized ' state inside the medium . strictly speaking ,
since we consider disorder over a finite region , @xmath138 is not localized : the exponential decrease of the envelope stops outside the scattering medium , so that @xmath138 is not square integrable .
the state @xmath138 rather corresponds to a resonance , whose lifetime is of the order of @xmath146 over the energy width of the resonance . in practice ,
@xmath104 can be considered as localized if the resonance lifetime is much longer than the duration of the experiment .
assuming that the phase of @xmath140 varies linearly with @xmath20 close to @xmath139 , one finds that this ` localized ' state is present on a narrow interval in @xmath20 of width @xmath147 , so that its energy width , or equivalently its inverse lifetime , scales as @xmath148 this illustrates how the proposed technique , defined in eq .
( [ eq : the_def ] ) , gives access to localized wavefunctions . in presence of a 1d scattering medium of length @xmath149 much larger than the localization length @xmath130 , so that the modulus of the transmission coefficient for each half of the medium @xmath150 . ( a ) for a generic positive energy : the wavefunction decreases exponentially inside the medium , being of modulus @xmath151 out of the medium and @xmath152 in the center @xmath112 of the medium .
( b ) for specific values of the energy , the wavefunction is ` localized ' inside the medium : its modulus decreases from @xmath144 in @xmath112 to @xmath145 for @xmath153 .
[ cap : schema_1d ] ] we now present numerical results obtained for a single realization of a random potential obtained by a monte carlo generation of the positions of a finite number @xmath42 of scatterers at the nodes of a cubic lattice with a given filling factor @xmath3 .
( color online ) ratio of the values of @xmath154 in the center and outside the scattering medium ( see text ) as a function of @xmath155 , with @xmath156 , for a given realization of the disorder with @xmath157 scatterers that occupy the nodes of a cubic lattice with 21 sites per side , with a filling factor @xmath81 . the value of @xmath158 is calculated as explained in @xcite ( red solid curve ) or with the extrapolation from the far field behavior , eq .
( [ ang_average ] ) with @xmath159 ( green dashed curve ) .
the energy intervals where the matter wave significantly penetrates the scattering medium correspond to the narrow peaks in this figure .
we have checked that the wave function is actually spatially localized in such an energy interval .
the vertical lines mark the locations of the resonances obtained by the spectral method : dashed lines for the very long lived resonances ( @xmath160 ) and dotted lines for the broader resonances . the temporal decay rate @xmath161 is obtained from eq .
( [ eq : width ] ) . ] in fig .
[ cap : scan ] we plot the ratio of the square of the amplitudes of @xmath104 inside and outside the scattering medium , for @xmath162 scatterers and a filling factor @xmath81 . in order to avoid a choice that might pick a node of @xmath163 , we plot @xmath164 , where both numerator and denominator are averaged over a few points @xcite .
the graph reveals that the phenomenology is similar to the 1d case : one has generically @xmath165 , except for narrow energy intervals , corresponding to the peaks in the figure , where the matter wave can significantly penetrate the scattering medium .
we have verified for a large number of peaks that the wavefunction @xmath104 is indeed ` localized ' inside the medium .
we illustrate this phenomenology for a generic value of @xmath20 , and for one that corresponds to a peak in the function @xmath166 taking now a larger number of scatterers @xmath167 . as can be seen in fig .
[ cap : etats3d ] , in the generic case @xmath163 decays essentially exponentially when entering the medium , whereas for the specific value of @xmath20 the wavefunction decays essentially exponentially from the medium center towards the outside .
what is the associated localization length @xmath130 ?
if one takes @xmath168 , one obtains from a fit the estimate @xmath169 @xcite .
plot of @xmath170 along the straight line passing through @xmath108 and parallel to @xmath11 axis , for two values of @xmath20 , ( a ) a generic value @xmath171 and ( b ) a value @xmath172 corresponding to a peak for @xmath166 as a function of @xmath20 ( in the spirit of fig.[cap : scan ] ) . note the similarity with the 1d case sketched in fig.[cap : schema_1d ] . in ( a )
the position @xmath108 is close to the center of the scattering medium : @xmath173 ; in ( b ) it is close to the ` center ' of the localized state : @xmath174 .
the effective scattering length is given by @xmath175 .
to get a clear evidence of the exponential decay of @xmath104 , we used a larger scattering medium than in fig .
[ cap : scan ] : @xmath176 atoms on the lattice within a sphere of radius @xmath177 , with an occupation probability @xmath81 .
, title="fig : " ] + plot of @xmath170 along the straight line passing through @xmath108 and parallel to @xmath11 axis , for two values of @xmath20 , ( a ) a generic value @xmath171 and ( b ) a value @xmath172 corresponding to a peak for @xmath166 as a function of @xmath20 ( in the spirit of fig.[cap : scan ] ) . note the similarity with the 1d case sketched in fig.[cap : schema_1d ] . in ( a ) the position @xmath108 is close to the center of the scattering medium : @xmath173 ; in ( b ) it is close to the ` center ' of the localized state : @xmath174 .
the effective scattering length is given by @xmath175 .
to get a clear evidence of the exponential decay of @xmath104 , we used a larger scattering medium than in fig .
[ cap : scan ] : @xmath176 atoms on the lattice within a sphere of radius @xmath177 , with an occupation probability @xmath81 .
, title="fig : " ] in the complex plane @xmath11 .
the resolvent has poles on the negative side of the real axis , corresponding to bound states , and a branch cut on the positive side of the real axis , corresponding to the continuum .
the analytic continuation of the resolvent across the branch cut from @xmath178 to @xmath179 may present a discrete set of poles in the fourth quadrant : the associated states are the resonances of the system .
[ cap : poles - of - the - resolvent ] ] in this subsection , we adopt a different point of view on localization @xcite , based on the properties of the analytic continuation of the resolvent @xmath180 from the half - plane @xmath181 to the half - plane @xmath182 , across the branch cut @xmath183 corresponding to the continuous spectrum of the hamiltonian .
this analytic continuation can have poles in @xmath184 , where @xmath185 and @xmath161 are positive ( see fig .
[ cap : poles - of - the - resolvent ] ) : these poles correspond to resonances .
the lifetime of the state associated to the resonance is given by @xmath186 .
the key property of localized states is that they correspond to resonances with a decay rate @xmath161 that tends to zero in the limit of an infinite extension of the disorder , so that they become in this limit square integrable stationary states of the hamiltonian .
we expect that the poles associated to these narrow resonances will leave signatures on the real axis , in the form of eigenvalues of the matrix @xmath92 with vanishing real part and tiny imaginary part , for an energy @xmath187 close to the real part @xmath185 of the poles .
as opposed to the poles of @xmath188 , the eigenvalues of @xmath92 can be calculated in a straightforward manner , and in fig .
[ cap : scan ] we show with vertical lines the values of @xmath20 for which the matrix @xmath92 has a purely imaginary eigenvalue : the ones with the smallest imaginary parts ( dashed vertical lines ) are in good correspondence with the narrow peaks in @xmath189 .
now , for a value of @xmath20 within the width of a peak in fig.[cap : scan ] , we give an analytical argument relating the spatial decay of @xmath163 to the presence of a tiny eigenvalue @xmath190 of @xmath92 of modulus much smaller than the other eigenvalues . in the large @xmath191 limit ,
( [ eq : phi_jolie ] ) reduces to the far field expression @xmath192 , \label{eq : far_field}\ ] ] with @xmath193 .
as shown in the appendix [ appen : magique ] , the sum over @xmath194 in the right hand side is typically @xmath195 times larger than the first term so that the angular average of @xmath196 is given , apart from oscillating terms @xmath197 , by @xmath198 this allows to estimate @xmath199 by extrapolating this far field expression down to @xmath200 , where @xmath201 is a distance of order the size of the scattering medium @xcite , see the green dashed curve in fig.[cap : scan ] .
we now need to estimate the wavefunction inside the medium .
close to the scattering center located in @xmath202 , one finds @xmath203 .
this suggests that the @xmath204 are also localized , in the sense that they decrease rapidly from a center in the medium outwards : this we have checked numerically .
then , averaging spatially over a small volume @xmath205 , @xmath206 being smaller than the mean distance between scatterers , one finds that @xmath207 is at most of the order of @xmath208 , a value reached when @xmath108 is close to the ` center ' of the localized wavefunction .
we reach the estimate @xmath209 in the right hand side , the first factor has a geometrical origin whereas the second one is sensitive to matter wave interference effects due to multiple scattering on the @xmath5 atoms : since the matrix elements of @xmath210 are of the order of unity , only interference effects can indeed lead to a very small expectation value of this matrix . using arguments detailed in the appendix [ appen : magique ] ,
we ultimately arrive at @xmath211 , \label{eq : decay_to_m0}\ ] ] which links the spatial decay of the wavefunction to the smallness of an eigenvalue of the matrix @xmath92 .
a useful application of this formula is to give the shape of the resonances in fig.[cap : scan ] .
we linearize the @xmath20 dependence of @xmath190 around the value @xmath139 such that @xmath190 is purely imaginary : @xmath212 , where @xmath213 . anticipating fig .
[ cap : eigenvalues - of - m ] , it appears that the imaginary part of an eigenvalue of @xmath214 , when tiny for @xmath215 , remains tiny for even lower values of @xmath20 , so that the derivative of @xmath216 is also tiny ; since the real part varies on the contrary over an interval of width @xmath217 , its derivative is not extremely small : @xmath218 is essentially real @xcite .
( [ eq : decay_to_m0 ] ) then leads to a lorentzian shape of the peaks in fig.[cap : scan ] , in agreement with the numerics .
since the peaks have a width much narrower than @xmath139 , this also leads to a lorentzian dependence with the energy @xmath219 , with a half - width at half maximum @xcite : @xmath220 when combined to eq .
( [ eq : decay_to_m0 ] ) , this leads to a formula nicely relating the inverse lifetime @xmath161 of the localized state to its spatial decay : @xmath221 if one then assumes a state with a wavefunction localized close to the center of the medium and decaying exponentially as @xmath222 away from the center , eq .
( [ eq : decay_to_m0 ] ) leads to @xmath223 , where @xmath149 is the diameter of the scattering medium , and eq .
( [ eq : width_3d ] ) leads to a resonance energy width @xmath224 , thus obeying the same scaling as in 1d , see eq .
( [ eq : width_1d ] ) . an important practical consequence of the present spectral approach is to give at once the range of values of @xmath43 for which one can hope to have localized states . for a given value of @xmath20
, @xmath92 will have a tiny purely imaginary eigenvalue if @xmath225 is opposite to the real part of an eigenvalue of @xmath95 with a very small imaginary part , see eq .
( [ eq : m ] ) . in fig .
[ cap : eigenvalues - of - m ] the eigenvalues of @xmath95 are shown as points in the complex plane , for various values of the incoming wave number @xmath20 and for two realizations of the disorder with widely different densities of scatterers .
we observe that , as the incoming wave number @xmath20 decreases below the inverse of the mean distance between the scatterers @xmath226 , many eigenvalues acquire an extremely small imaginary part and accumulate in the region @xmath227 ( see the green dashed lines in fig .
[ cap : eigenvalues - of - m ] ) , similarly to earlier calculations for light waves @xcite . in a matter wave experiment
, this suggests to tune @xmath228 to a value close to @xmath229 ( as we have done in fig.[cap : scan ] ) : this might be achieved in practice by using a feshbach resonance , as shown in the next section . in this way ,
as @xmath20 decreases below @xmath229 , one obtains a very large sequence of values of @xmath20 such that @xmath92 has a tiny purely imaginary eigenvalue , that is one may have access to a large number of localized states .
this is illustrated in fig.[cap : fonc_aeff ] .
finally , the representation in fig.[cap : eigenvalues - of - m ] is also useful to understand what happens in the low energy limit , @xmath230 it shows that , for each eigenvalue @xmath231 of the matrix @xmath92 , @xmath232 and @xmath233 have a finite limit @xcite ; a numerical inspection reveals that some of the eigenstates of @xmath234 , having a tiny value of @xmath235 , are localized .
can these localized states in the zero energy limit be accessed in a real experiment ? for a given realization of disorder
, this would require that @xmath43 be tuned exactly to one of the corresponding values of @xmath236 , which is unrealistic . in the same way
, it would be very difficult to adjust @xmath43 to hit , at a _ given _ value of @xmath20 , one of the peaks in fig.[cap : fonc_aeff ] .
fortunately , in a real experiment , @xmath43 is fixed and a broad interval of @xmath20 can be populated by the atomic wavepackets ; then , fig.[cap : eigenvalues - of - m ] shows that the real parts of the eigenvalues of @xmath214 with tiny imaginary parts are increasing functions of @xmath20 , so that they have a chance to cross the value @xmath237 for some value of @xmath20 within the experimentally populated interval , and to lead to a peak in fig.[cap : scan ] and to an observable localized state in the experiment .
of @xmath95 for different values of the wave number @xmath20 of the matter wave .
the eigenvalues of @xmath92 are then simply deduced from these eigenvalues by a shift of @xmath225 along the real axis . left column : cubic lattice with 21 sites / side , @xmath81 and 872 scatterers ;
the inverse mean distance between scatterers is @xmath238 ( same realization of disorder as in fig .
[ cap : scan ] ) .
right column : cubic lattice with 209 sites / side , @xmath239 and 878 scatterers ; the inverse mean distance between scatterers is @xmath240 .
the green dashed lines mark the values @xmath241 .
the real axis is in units of @xmath226 , and common values of @xmath242 are taken in both columns , so as to reveal a possible universality in the low @xmath3 limit .
the imaginary axis is in units of @xmath20 , as justified by eq .
( [ eq : decay_to_m0 ] ) .
[ cap : eigenvalues - of - m ] ] , ratio of the values of @xmath154 in the center and outside the scattering medium as a function of the effective scattering length @xmath43 @xcite .
the dashed line gives the result for @xmath243 .
the same realization of disorder is used as in fig.[cap : scan ] .
the inset is a magnification of the region @xmath59 .
[ cap : fonc_aeff ] ] to conclude this section , it is useful to discuss the number of localized states that can be supported by a finite size scattering medium , to have an idea of the number of atoms that may populate these localized states in a real experiment , and to have an estimate of the number of states that can be considered as localized for a given duration of the experiment . to this end , we present in fig.[cap : histo ] an histogram giving the number of purely imaginary eigenvalues of @xmath92 per class of inverse lifetimes @xmath161 .
it is apparent on this figure how an increase of the volume of the scattering medium ( here by a factor @xmath244 ) leads to both an increase in the total number of localized states ( for a given lifetime ) and to the appearance of a tail of states of significantly longer lifetimes ( here by about two orders of magnitude ) . per class of inverse lifetimes @xmath186 .
the filling factor of the lattice is @xmath81 and the effective scattering length is @xmath69 .
black dashed histogram : @xmath245 scatterers within a cube with 21 sites per side .
red solid line histogram : @xmath246 scatterers within a cube with 31 sites per side .
these histograms were constructed by a dichotomy search of the values @xmath139 of the momentum @xmath247 $ ] such that the matrix @xmath92 has a purely imaginary eigenvalue ; the associated lifetime was then calculated with eq .
( [ eq : width ] ) .
eigenvalues with negative values of @xmath161 are of course not included ( see @xcite ) .
the decay rate @xmath161 is given in units of the recoil angular frequency @xmath248 of the species @xmath4 . in practice , only the eigenvalues with small enough values of @xmath161 are expected to produce an observable resonance in @xmath166 as a function of @xmath20 , see fig.[cap : scan ] .
if the matter wave is made of atoms of @xmath52li , a duration of the experiment of 0.3s corresponds to a minimal observable inverse lifetime of @xmath249 for an optical lattice laser wavelength of 779 nm .
[ cap : histo ] ]
we will solve here the two - body problem of a free atom @xmath4 scattering on a single trapped particle @xmath5 prepared in the ground state of a harmonic oscillator potential . in the limit of vanishing incoming kinetic energy of the @xmath4 particle ,
this gives access to the effective scattering length introduced in eq .
( [ eq : veff ] ) , which is the key parameter of our model hamiltonian with disorder .
the hamiltonian for a free @xmath4 particle and a harmonically trapped @xmath5 one is , in the absence of @xmath4-@xmath5 interaction , @xmath250 it admits the @xmath40-wave solution @xcite @xmath251 which represents a spherical matter wave ( @xmath4 ) of wave vector @xmath20 and a particle ( @xmath5 ) in the ground state of the harmonic potential , @xmath252 , with the harmonic oscillator length @xmath253 .
the state @xmath254 has an energy @xmath255 we model the @xmath4-@xmath5 interaction with a regularized contact potential , which leads to the full hamiltonian : @xmath256 here , the derivative is taken for a fixed value of the center of mass position @xmath257 of the two particles , the coupling constant @xmath258 is expressed in terms of the reduced mass @xmath259 and the free - space scattering length @xmath47 ( relative to the @xmath4-@xmath5 interaction in the absence of the trapping potential ) by @xmath260 the schrdinger equation @xmath261 can be reformulated equivalently in the integral form@xmath262 in terms of the two - particle retarded green s function for the non - interacting hamiltonian@xmath263 and the regularized part of the two - particle wave function,@xmath264\right|_{\mathbf{r}_{a}=\mathbf{r}_{b}}\ ] ] where @xmath265 and @xmath266 since we consider @xmath40-wave scattering only . inserting eq .
( [ eq : scatteringsolution ] ) into the definition of @xmath267 , we find that the regularized part of the wave function satisfies an equation of the form@xmath268 where @xmath269 , @xmath270 , and @xmath271 is an integral operator independent of the scattering length @xmath47 .
the detailed derivation of this equation and the explicit form of @xmath271 are presented in appendix [ appen : inte ] .
as we now show , the knowledge of @xmath272 directly leads to the value of the scattering amplitude @xmath273 for the @xmath4 wave .
expanding @xmath274 on the basis of eigenstates @xmath275 of @xmath276 , and projecting in position space , one is able to calculate the integral over the wavevector @xmath277 so that one is left with a sum over the vibrational states of the @xmath5 particle : @xmath278 .
\label{eq : greeninposspace}\end{gathered}\ ] ] the low - energy assumption ( [ eq : elasticscattering ] ) ensures that the only open exit channel corresponds to a @xmath5 particle in the ground vibrational state , and therefore that @xmath279 as a consequence , the terms of eq .
( [ eq : greeninposspace ] ) involving excited states of the harmonic oscillator give exponentially vanishing contributions and can be neglected when @xmath4 is at a distance @xmath280 from the center of the harmonic well .
expanding @xmath281 and substituting into eq .
( [ eq : scatteringsolution ] ) , we get @xmath282\phi_{0}(r_{b}),\label{eq : incoming+outgoing}\ ] ] where we have introduced the scattering amplitude @xmath283 in the limit of zero - energy , the scattering amplitude defines the _ effective scattering length _ @xmath284 through @xmath285 we have solved numerically eq .
( [ eq : integralequation ] ) , and in figs .
[ cap : effective0.15]-[cap : effective6.67 ] we plot @xmath48 as a function of @xmath286 for different values of @xmath287 .
we choose @xmath288 , corresponding to the physical cases of a mixture of @xmath289k and @xmath290li , a mixture of two different internal states of atoms of the same species , and a mixture of @xmath291li and @xmath292k , respectively .
it is apparent that @xmath43 presents a series of intriguing resonances , which can be used to tune @xmath43 to a high value and whose physical original are discussed in
[ subsec : reson ] . in figs .
[ cap : effectiveegales ] and [ cap : r_e ] we also plot , in the case @xmath293 , the behavior of the effective range @xmath294 .
it is defined as a coefficient in the low @xmath20 expansion of the inverse scattering amplitude : @xmath295.\ ] ] the replacement of a trapped @xmath5 particle by a fixed point - like scatterer , as done in eq .
( [ eq : veff ] ) , is allowed when the @xmath296 term in the above expansion is negligible , i.e. when @xmath297 in the ideal regime for matter wave localization , @xmath298 . since @xmath299 is generally of order @xmath300 , this condition eq .
( [ eq : cond_contact ] ) is satisfied .
note that @xmath294 diverges when @xmath301 , a generic phenomenon : expanding @xmath273 rather than @xmath302 in powers of @xmath20 , @xmath303 one generically expects that @xmath304 has a finite limit when @xmath305 .
this was demonstrated analytically both for a square well @xcite and a van der waals interaction potential @xcite . in this limit , eq .
( [ eq : cond_contact ] ) is thus violated for any finite @xmath20 ; however , this is not an ideal regime to obtain matter wave localization .
we now show that the effective coupling constant experiences an infinite set of resonances due to the presence of the external confinement . in the limit @xmath230 , the resolvent has real matrix elements in position space , see eq .
( [ eq : greeninposspace ] ) ; so does the symmetric operator @xmath271 , which then admits real eigenvalues @xmath306 and orthonormal eigenvectors @xmath307 ( see eqs.([eq : uf1],[eq : uf2 ] ) ) .
we then rewrite eq .
( [ eq : a_eff ] ) as @xmath308 where the scalar product of functions of the single variable @xmath201 is defined as @xmath309 .
this means that a singularity in @xmath48 is expected whenever the denominator of eq .
( [ eq : f_k__rewritten ] ) vanishes , i.e. whenever the inverse free space scattering length @xmath286 equals @xmath310 , provided that the non - interacting wave function has non - zero overlap with the corresponding eigenstate @xmath311 .
confinement - induced resonances have been been analyzed theoretically in different contexts , such as 1d wave guides @xcite , 3d optical lattices @xcite , and quasi-2d condensates @xcite . in 1d
wave guides the effect is particularly remarkable : due to the presence of the transverse confinement , a contact potential acquires a bound state for any value of the 3d scattering length ( while in free space the contact potential has a bound state only for @xmath312 ) .
confinement - induced modifications of two - body scattering properties have very recently been observed experimentally by the zurich group in 1d waveguides @xcite and in 3d optical lattices @xcite . in most of the cited papers , the underlying translational symmetry and the harmonic nature of the confinement
permit the factorization of the center - of - mass motion : this in turn implies that a single confinement - induced resonance can exist , since only one state in the closed channel is coupled to the open channel @xcite .
when this factorization is not possible , as in our setup or in the case of a 1d wave guide with anharmonic transverse confinement @xcite , an infinite set of states in the closed channel has non - zero coupling to the open one , and an infinite number of resonances appears . in practice
however only a few of them may be resolved and relevant in an experiment since they become increasingly sharper as @xmath313 becomes larger ( i.e. @xmath314 for large @xmath315 ) .
the position of these resonances can be predicted analytically in various limits , depending on the sign of @xmath47 , as we now discuss .
when @xmath312 , the pseudopotential admits a bound state in which the two particles can `` sit '' for a variable time , forming a molecule that oscillates in the harmonic well . to understand this point ,
we rewrite the two - body hamiltonian ( [ eq : twobodyh ] ) as@xmath316,\label{eq : h_in_com_and_rel_coord}\end{gathered}\ ] ] and treat the terms in the square parenthesis as a perturbation .
the unperturbed part admits the factorized eigenstates @xmath317 that describe a bound molecule with internal wavefunction @xmath318 and center - of - mass in an eigenstate of the harmonic oscillator of angular momentum @xmath319 and radial quantum number @xmath320 .
since both the initial state and the hamiltonian are spherically symmetric , conservation of angular momentum allows only @xmath319 intermediate molecular states . within this unperturbed approximation
, @xmath48 diverges each time the energy of the oscillating molecule corresponds to the ground state energy of the pair of atoms , i.e. at the values of @xmath321 that satisfy @xmath322 as can be seen in the upper part of fig .
[ cap : posres ] , this formula describes the position of the resonances with @xmath312 in a wide region of the graph , since corrections to it are only @xmath323 @xcite . for @xmath324 and for a large enough @xmath325 ratio ,
the presence of at least one resonance can be demonstrated by a very simple variational argument performed at the unitary limit @xmath326 .
a contact potential characterized by the scattering length @xmath47 can be replaced by the simple boundary condition @xmath327 , where @xmath328 is an arbitrary function of the center of mass coordinate @xmath257 .
the @xmath40-wave ansatz @xmath329 with @xmath1 and @xmath330 variational parameters , satisfies the boundary condition imposed by a unitarity limited contact potential , i.e. characterized by @xmath326 .
its energy can be calculated from the hamiltonian ( [ eq : h_in_com_and_rel_coord ] ) : the term @xmath331 has a vanishing contribution when averaged over this state , and the variational energy assumes the minimum value @xmath332 we might imagine continuously tuning @xmath47 from @xmath333 , where no bound state can exist , towards @xmath334 .
if @xmath335 ( i.e. if @xmath336 ) a bound state for @xmath326 is guaranteed by the former ansatz , and at least one resonance for @xmath48 must exist in the @xmath324 region .
the following more elaborate calculation allows to prove the existence of a resonance for @xmath324 , for an arbitrary mass ratio @xmath325 .
one performs a variational calculation directly on the integral operator @xmath337 , taking as a variational function @xmath338 . using eqs.([eq : uf1],[eq : uf2 ] ) and performing gaussian integrals we obtain in harmonic oscillator units : @xmath339^{-3/2 } \right.\nonumber \\ & & \left.-\left[\tau+\alpha-\alpha \exp(-\tau)\right]^{-3/2}\right\}.\end{aligned}\ ] ] from the inequality @xmath340 valid for any @xmath341 , we conclude that the integrand is a negative function , so that @xmath342 . since the trial wavefunction @xmath343 is proportional to @xmath344 ,
we conclude that @xmath337 admits at least one eigenvector @xmath345 with a non - zero overlap with @xmath346 and with a negative eigenvalue @xmath347 .
the identity eq .
( [ eq : f_k__rewritten ] ) then implies the existence of a resonance in @xmath43 for @xmath324 .
when @xmath324 and @xmath348 , the position of the resonances can be found with the aid of the born - oppenheimer approximation . for a fixed position of the massive particle @xmath4 , one calculates the energy of the @xmath5 particle , which then constitutes an effective potential for the @xmath4 particle . restricting for simplicity to the mean field regime @xmath349 , one can assume that the particle @xmath5 remains in the ground state of the well , thereby creating an effective gaussian attractive well for the @xmath4 particle .
hence the effective hamiltonian for @xmath4 : @xmath350 as argued above , the characteristic range of the potential is of order @xmath351 .
this hamiltonian can be easily solved numerically , and predicts a divergence of @xmath48 whenever the combination @xmath352 equals the critical value for the appearance of a new bound state ( see fig . [
cap : posres ] , dashed lines in lower graph ) . in the opposite limit @xmath353 ,
there is one resonance left on the @xmath324 side .
it is intuitive that its position @xmath354 tends to @xmath334 in this limit , the @xmath5 particle being then perceived by @xmath4 as a fixed scatterer of scattering length @xmath47 , for which the resonance is obtained for @xmath355 . at @xmath356 ,
one then expects that @xmath357 . to formalize this intuition ,
we expand the integral operator @xmath271 in powers of the mass ratio @xmath358 : from eqs.([eq : uf1],[eq : uf2 ] ) , we get @xmath359 it is then possible to check analytically , by calculation of gaussian integrals , that one has exactly @xmath360 , with @xmath361 .
using perturbation theory , we obtain the series expansion of the lowest eigenvalue of @xmath337 : @xmath362 in harmonic oscillator units .
the resulting lowest order expression for the resonance position is @xmath363 , \label{eq : pos_asympt}\ ] ] and is shown as a dashed line in fig.[cap : posres ] @xcite .
in this paper , we have shown numerically that the way of producing a disordered potential for matter waves proposed in @xcite , i.e. the use of atoms randomly trapped at the nodes of an optical lattice , indeed leads to the appearance of exponentially localized states in a three - dimensional geometry .
our numerical method allows to directly compute the wavefunction of the localized states ; it is based on the fact that the matrix elements of the resolvent of the hamiltonian can be calculated extremely efficiently for the interaction of the matter wave with point scatterers , a fact already used with success in the context of light localization @xcite .
the method also allows to obtain analytical results in a straightforward manner , such as a check of the existence of localized states in 1d from furstenberg theorem or the derivation of a link between the inverse lifetime of a localized state and its spatial decay in 3d , for a finite spatial extension of the scattering medium .
the main physical result is that numerous long - lived localized states appear for a wave number of the matter wave smaller than the inverse of the mean separation between scatterers , when the effective scattering length @xmath43 of the matter wave on a trapped atom is positive and of the order of the mean separation of the scatterers .
for the numerical examples of this paper , with a 10% occupancy of the lattice sites and for a wavenumber of the matter wave as large as @xmath364 , there are localized states with localization lengths that can be as small as the lattice period @xmath58 , usually a sub - micron quantity .
this extremely strong localization allows to have very long - lived localized states ( lifetime larger that @xmath365 inverse recoil angular frequency of the matter wave ) even in disordered samples with a radius as small as 10 lattice periods . by a full solution of the scattering problem of a free atom with a harmonically trapped one , we have shown that the required large values of @xmath59 can be obtained by using an inter - species feshbach resonance , and we have characterized intriguing confinement - induced resonances that appear in this two - body scattering process . how to proceed in a real experiment to get evidence of these localized states ?
a possibility is to extend to matter waves what was proposed for light in @xcite : one introduces the matter wave wavepackets inside the scattering medium at a low value of @xmath43 , then one tunes @xmath43 to the desired high value and one lets the matter wave evolve in presence of the scattering medium ( but in the absence of an external trapping potential ) . after an adjustable time @xmath366 ,
one measures the number of remaining matter wave atoms @xmath367 in the scattering medium .
since the component of the matter wave wavefunction in localized states decays exponentially in time with very weak rates @xmath161 , the function @xmath367 should have a long tail , as compared to the case of a purely ballistic or even diffusive expansion @xcite . a further check that this long tail has
a decay rate varying exponentially with the size of the scattering medium would be a very convincing evidence of strong localization @xcite .
we are grateful to uri gavish , dominique delande , martin weitz , leonardo fallani , chris .
j. pethick , henrik smith , anders s. srensen , fabrice gerbier , jean dalibard and christophe salomon for stimulating discussions .
laboratoire kastler brossel is a research unit of cole normale suprieure and of universit pierre et marie curie , associated to cnrs .
our research group in paris is a member of the ifraf institute .
as illustrated in fig .
[ cap : decayingresolvent ] , in our model the off - diagonal matrix elements of the resolvent @xmath368 decay exponentially in @xmath369 at low energy @xmath97 .
is this a signature of localization ?
no , because this decay may be due to the fact that the energy @xmath97 is in a spectral gap of the system .
and this occurs as well in the forbidden bands of a periodic system . to illustrate this statement for our system ,
we have enclosed our lattice of scatterers in a box of side @xmath149 , imposing periodic boundary conditions on the walls of the box .
this amounts to replacing eq .
( [ eq : g0free ] ) by the particle propagator satisfying the correct boundary conditions:@xmath370 with @xmath371 and @xmath372 ( a triplet of integers ) .
indeed we found that , for the ensemble of scatterers used in fig .
[ cap : decayingresolvent ] , the ground state of the system , once enclosed in a box of side @xmath373 ( slightly larger than the scattering medium ) , is characterized by a wave number @xmath374 . the exponential decay shown at @xmath375 by @xmath98
is therefore simply indicating that at such low energy no state can exist deep inside the medium . in a scattering experiment
, we might imagine a plane wave coming from infinity that scatters on the trapped @xmath5 atoms : if @xmath376 the incoming wave undergoes total reflection , and inside the random medium only penetrates an evanescent wave , that decays exponentially from the boundary of the medium towards its interior .
this example clearly points out that an exponential decay of @xmath98 is not a sufficient criterion to prove localization in our system , since it does not guarantee the existence of states deep inside the random potential . in the criterion of kramer and mackinnon , introduced for a solid ,
the energy is taken equal to the fermi energy , with the assumption that the density of states does not vanish at the fermi energy .
this criterion then turns out not to be practical in our case , since it requires a diagonalization of the hamiltonian in order to calculate the density of states .
we consider a 1d free space geometry with @xmath42 dirac scatterers located in distinct positions @xmath377 .
a quantum particle of mass @xmath109 interacts with the @xmath42 scatterers , with a coupling constant @xmath258 . using the transfer matrix formalism
we calculate the resolvent @xmath378 , for @xmath379 , @xmath97 real and @xmath380 , which will give access to stationary wavefunctions in the limit @xmath381 . taking for simplicity @xmath382 different from the positions @xmath383 , we use the fact that @xmath384 solves schrdinger s equation with a source term : @xmath385\psi(x)= \delta(x ) .
\label{eq : schr}\ ] ] over a position interval containing neither one of the @xmath383 nor @xmath382 , @xmath386 is an eigenstate of @xmath387 so that , introducing the unique @xmath388 such that @xmath389 with @xmath390 , one has : @xmath391 where we assumed that the first @xmath40 scatterer positions @xmath392 are @xmath393 and the @xmath394 other ones @xmath395 are @xmath396 , and where @xmath194 runs from 1 to @xmath397 and @xmath398 from @xmath399 to @xmath400 .
all the unknown coefficients @xmath401 and @xmath402 shall now be determined from boundary conditions obeyed by @xmath386 .
a first boundary condition is that @xmath386 should not diverge exponentially in @xmath403 .
this imposes @xmath404 and @xmath405 .
the other boundary conditions originate from the fact that each dirac distribution in eq .
( [ eq : schr ] ) introduces a discontinuity of the first order derivative of @xmath386 , whereas @xmath386 remains continuous . integrating eq .
( [ eq : schr ] ) over an infinitesimal interval around @xmath406 leads to @xmath407 with @xmath408 .
these two equations allow to connect the unknown coefficients on the right of @xmath406 to the ones on the left by a two - by - two transfer matrix . for @xmath409
we thus have @xmath410 and for @xmath411 : @xmath412 a simple calculation leads to the explicit expression @xmath413 with @xmath414 .
we then introduce the two following matrices , one associated to the first @xmath40 scatterers , @xmath415 and the other one to the @xmath394 last scatterers : @xmath416 they allow to express the coefficients @xmath417 in terms of @xmath418 and the coefficients @xmath419 in terms of @xmath420 : @xmath421 the last two unknowns , @xmath418 and @xmath420 , are obtained from the boundary conditions imposed by @xmath422 in eq .
( [ eq : schr ] ) , @xmath423 and @xmath424 which imposes two equations on the coefficients @xmath425 , @xmath426 combined with eq .
( [ eq : gros1 ] ) and eq .
( [ eq : gros2 ] ) , this leads to the following system : @xmath427 which can be solved explicitly : @xmath428 we then proceed with the limit @xmath11 tending to a real and positive energy @xmath97 : @xmath381 and @xmath429 .
the @xmath430 then become physically meaningful transfer matrices in the @xmath431 group . in the main text we introduced the reflection and transmission coefficients @xmath119
, @xmath124 of the last @xmath394 first scatterers on the axis oriented from @xmath118 to @xmath121 .
this means that there exists a stationary solution of the usual schrdinger equation equal to @xmath432 for @xmath433 and equal to @xmath434 for @xmath435 . by definition of the transfer matrices : @xmath436 similarly , introducing the reflection and transmission coefficients @xmath437 of the first @xmath40 scatterers on the axis oriented this time from @xmath121 to @xmath118
, we imply the existence of a solution of schrdinger s equation equal to @xmath438 for @xmath439 and equal to @xmath440 for @xmath441 , which imposes @xmath442 this allows a physical interpretation of the coefficients of @xmath443 and @xmath444 : @xmath445 this leads to very simple expressions for the green s function @xmath386 out of the scattering medium and over the interval between two scatterers containing @xmath382 : @xmath446
this appendix is useful to derive eq .
( [ eq : decay_to_m0 ] ) , an equation which relates the spatial decay of @xmath448 to the smallness of an eigenvalue of @xmath92 .
let us assume that we are at a positive energy @xmath97 such that the matrix @xmath92 has an eigenvalue @xmath190 extremely close to zero , much closer to zero anyway than all the other eigenvalues .
let @xmath449 be the associated eigenvector of @xmath92 .
the corresponding adjoint vector is an eigenvector of @xmath450 with the eigenvalue @xmath451 . since the matrix @xmath92 is ( complex ) symmetric
, @xmath450 is simply @xmath452 , the complex conjugate of @xmath92 , so that one may take as adjoint vector the complex conjugate @xmath453 of @xmath449 .
the imposed normalization condition is then @xmath454 for compactness we use here dirac s notation , even if @xmath92 does not act in a hilbert space .
one has to calculate the vector @xmath455 to fully determine @xmath448 , see eq .
( [ eq : phi_jolie ] ) .
this vector solves the linear system @xmath456 with the source term @xmath457 .
as the eigenvalue @xmath190 of @xmath92 is the only one to be extremely close to zero , we take the approximate expression @xmath458 this leads to @xmath459 to have access to an estimate of @xmath460 , we have to calculate the expectation value of the imaginary part of the matrix @xmath92 on the vector @xmath455 .
since @xmath92 is symmetric , both the real part and the imaginary part of @xmath92 are hermitian matrices , with real expectation values , so that @xmath461 then using the approximation eq .
( [ eq : approx_d ] ) we obtain @xmath462 an immediate application of this result is that the quantity @xmath463 , where @xmath464 is a unity vector , is typically much larger than unity .
the calculation of the average of @xmath465 over the unit sphere indeed leads to @xmath466 since @xmath467 in the low @xmath190 limit @xcite , it is correct to neglect the term @xmath468 in eq .
( [ eq : far_field ] ) as was done in the main text .
to have access to an estimate of @xmath469 , we have to calculate the maximal value of all the @xmath470 .
let us call @xmath471 the index such that @xmath472 is the biggest one . according to the approximation eq .
( [ eq : approx_d ] ) , we then have @xmath473 \label{eq : approx_in}\ ] ] where @xmath474 denotes the component @xmath471 of the vector @xmath449 . the expressions eqs.([eq : approx_out],[eq : approx_in ] )
greatly simplify when one chooses a position @xmath108 that tends towards @xmath475 .
the fact that the results shall not depend on this specific choice of @xmath108 is established in the appendix [ appen : math ] . in this limit ,
all the components @xmath476 of the source term @xmath477 are negligible as compared to @xmath478 so that @xmath479 and so on .
then eq .
( [ eq : est_int ] ) reduces to @xmath480 .
\label{eq : c9}\ ] ] as shown in the appendix [ appen : math ] , the components of the vector @xmath449 are real in the limit of a vanishing @xmath481 so that @xmath482 may be pulled out of the imaginary part . since the @xmath483
decrease roughly exponentially at large distances @xmath484 over a length scale @xmath485 of the order of the mean scatterer separation , as was known from studies of light localization @xcite , the normalization condition eq .
( [ eq : norm ] ) leads to @xmath486 , where @xmath487 is the mean scatterers density .
we then get eq .
( [ eq : decay_to_m0 ] ) .
we consider here a value of @xmath20 such that the matrix @xmath92 has one ( and only one ) eigenvalue @xmath190 of extremely small modulus .
we then show that the corresponding eigenvector @xmath449 of @xmath92 , normalized as in eq .
( [ eq : norm ] ) , is close to a vector with real components , a property used in the appendix [ appen : magique ] and necessary to obtain the equation eq .
( [ eq : decay_to_m0 ] ) .
first we give a physical argument .
starting from the approximation eq .
( [ eq : approx_m ] ) , and keeping terms only to leading order in @xmath488 , we obtain from eq .
( [ eq : phi_jolie ] ) : @xmath489 \label{eq : useful_phi}\ ] ] where @xmath490 now let us assume that the phase of @xmath491 ( modulo @xmath492 ) changes significantly ( that is in a way not tending to zero with @xmath190 ) when @xmath108 is varied .
this means that the @xmath493 dependent wavefunction @xmath448 spans a subspace of the hilbert space of dimension 2 when @xmath108 is varied , a subspace generated by the wavefunctions @xmath494 and @xmath495 , or equivalently by @xmath496 and @xmath497 .
this leads to the physically suspect situation that _ two _ independent localized states can be associated to a given resonance . to avoid this suspect situation , one
is led to the assumption that the function @xmath496 has a constant phase ( modulo @xmath492 ) . taking the limit @xmath498
, this implies that all the @xmath499 have the same phase ( modulo @xmath492 ) . from the normalization condition
( [ eq : norm ] ) we conclude that all the @xmath499 are real , apart from small terms tending to zero with @xmath190 .
furthermore , this leads to the conclusion that the function @xmath496 has to be real everywhere , apart from terms that tend to zero with @xmath190 @xcite .
a consequence that is important for the appendix [ appen : magique ] is that @xmath500 are approximately real , so that the result eq .
( [ eq : c9 ] ) does not depend indeed on the specific choice of @xmath108 .
now we give a mathematical argument to show that @xmath449 is almost real .
we start from @xmath501 and we split the matrix @xmath92 in a real part and an imaginary part , @xmath502 . since @xmath92 is symmetric , both @xmath503 and @xmath504 are real symmetric . taking the squared norm of the identity @xmath505 , where @xmath94 is the @xmath91 identity matrix , leads to @xmath506\vec{v_0},\ ] ] where we used the fact that @xmath507 that can be proved as in eq .
( [ eq : pour_appen_suiv ] ) .
next , we expand @xmath449 in the orthonormal eigenbasis of @xmath504 , @xmath508 so that @xmath509 it remains to use the following property of @xmath504 , valid for an arbitrary vector @xmath510 : @xmath511 to show that @xmath512 @xcite . as a consequence , @xmath513 . using the expansion eq .
( [ eq : expansion ] ) and eq .
( [ eq : sum_rule ] ) , one gets an upper bound on @xmath514 so that @xmath515 ||\vec{v_0}||^2.\ ] ] in the large scattering medium diameter @xmath128 , an exponential decrease of the imaginary part of @xmath190 is expected , @xmath516 . close to the center of a resonance peak , @xmath517 ,
so that @xmath518 $ ] . on the contrary , the density of states for the spectrum of @xmath503 is not expected to be exponentially peaked : the spacing between successive eigenvalues is expected to scale as @xmath519 at most .
this is apparent in fig.[cap : poles - of - the - resolvent ] , and we have checked it numerically for increasing numbers of scattering centers .
this implies that @xmath449 , when expanded on the eigenvectors of @xmath503 , populates essentially one eigenvector of @xmath503 , the one with the eigenvalue nearest to zero . since this ( unit norm )
eigenvector of @xmath503 is proportional to a real eigenvector , we deduce from eq .
( [ eq : norm ] ) that the components of @xmath449 are almost real , at the @xmath520 accuracy level .
to calculate the non - interacting two - particle green s function numerically , we adapt a technique used in @xcite .
the feynman propagator @xmath521 associated with the non - interacting hamiltonian @xmath276 @xcite,@xmath522 can be factorized as the product of the two factors , @xmath523 , the first term describing a free particle,@xmath524 and the second one a particle in a 3d harmonic oscillator,@xmath525\right)\label{eq : hoprop}\end{gathered}\ ] ] where @xmath526 for @xmath527 . setting @xmath528 for @xmath529
, @xmath274 is obtained as the fourier transform of @xmath530,@xmath531 for simplicity , we introduce dimensionless variables by expressing quantities in harmonic oscillator units and , even though the derivation has been carried out for a generic mass ratio , we restrict ourselves here to the special case @xmath532 ( see end of appendix for the case @xmath230 with an arbitrary mass ratio @xmath325 ) . in order to find the equation satisfied by @xmath267 , we rewrite eq .
( [ eq : scatteringsolution ] ) as@xmath533\label{eq : scatteringsolutionrewritten}\end{gathered}\ ] ] and let the two particles approach each other .
we introduce here the center - of - mass and relative coordinates @xmath534 and @xmath535 .
let us turn our attention to the first integral appearing in eq .
( [ eq : scatteringsolutionrewritten]):@xmath536 performing the gaussian integration over @xmath537 @xcite one finds @xmath538 where @xmath539 we have here introduced the shorthand notation @xmath540/2 $ ] and the phase factor @xmath541 , which equals @xmath542 for @xmath543 , where @xmath544 and @xmath545 are the consecutive solutions of @xmath546 . in the latter expression , eq .
( [ eq : u ] ) , the contribution of the neighborhood of @xmath547 diverges as @xmath548 for @xmath549 : @xmath550 we finally find@xmath551 that separates out the expected @xmath548 divergent contribution and defines @xmath552 with @xmath553 .
the term @xmath554 regularizes @xmath555 in the neighborhood of @xmath547 , and we have taken the limit @xmath556 .
let us now consider the remaining term appearing in eq .
( [ eq : scatteringsolutionrewritten ] ) : when @xmath557 and @xmath558 , the green s function @xmath274 diverges as @xmath559 , or equivalently as @xmath560 , but the second integral in eq .
( [ eq : scatteringsolutionrewritten ] ) is convergent in @xmath561 : @xmath562\\ \stackrel{\mathbf{r}\rightarrow0}{=}\int\frac{d\mathbf{u}}{\left|\mathbf{u}\right|^{4}}\left[\mathbf{u}\left.\frac{\partial\psi_{reg}}{\partial\mathbf{u}}\right|_{\mathbf{u=0}}+o\left(\left|\mathbf{u}\right|^{2}\right)\right]\end{gathered}\ ] ] ( the first order term vanishes due to spherical symmetry ) . in this term
therefore no divergence arises and we may set @xmath563 .
the angular integrations can be performed analytically and we obtain : @xmath564\\ = \frac{1}{r}\int_{0}^{\infty}d\rho\,\left[\tilde{\psi}_{\mathrm{reg}}(\rho)-\frac{\rho}{r}\tilde{\psi}_{\mathrm{reg}}(r)\right]f_{2}(r,\rho)\end{gathered}\ ] ] with@xmath565}}{\left(t|\sin(t)|\right)^{3/2}}\frac{\sin(2r\rho y)}{\left(2\pi\right)^{2}y},\ ] ] which is symmetric under the exchange of @xmath487 and @xmath201 . we have here introduced the radial wave function @xmath269 and the function @xmath566/2 $ ] .
writing @xmath567 we can cancel the divergent contribution on both sides of eq .
( [ eq : scatteringsolutionrewritten ] ) , and the remaining terms constitute the implicit integral equation @xmath568\end{gathered}\ ] ] that needs to be solved numerically in order to determine @xmath267 .
the latter equation can be written in a symbolic , more compact form as@xmath569 where @xmath271 is a symmetric integral operator , which is real in the limit @xmath570 . if one only aims at calculating strictly zero energy properties ( i.e. @xmath48 , and not @xmath571 or @xmath299 ) , the treatment presented in this appendix
can be drastically simplified setting @xmath572 in eqs .
( [ eq : freeprop],[eq : hoprop],[eq : gasfourierofk ] ) and @xmath573 in eq .
( [ eq : gasfourierofk ] ) : the derivation of @xmath271 proceeds in an analogous way , but the resulting integral equation is much easier to solve since the integrand in both @xmath555 and @xmath574 become real damped functions with no finite time singularities . for an arbitrary value of the mass ratio @xmath575 , they are given by : @xmath576^{3/2}}+\right.\\ \left.\frac { \exp(3\tau/2)}{\left [ \mathrm { sh } ( \tau ) + \frac \tau \alpha \mathrm { ch}(\tau)\right]^{3/2 } } \exp\left(-r^2 \frac { \mathrm { ch}(\tau ) + \frac \tau { 2\alpha}\mathrm { sh } ( \tau)-1}{\mathrm { sh } ( \tau ) + \frac \tau \alpha \mathrm { ch}(\tau ) } \right)\right\ } , \label{eq : uf1}\end{gathered}\ ] ] and @xmath577}{[\tau\mathrm{sh}(\tau)\exp(-\tau)]^{3/2}}\frac{\mathrm{sh}(2r\rho y)}{\left(2\pi\right)^{2}y } , \label{eq : uf2}\end{gathered}\ ] ] with @xmath578 and @xmath579 .
10 c. a. regal , m. greiner and d. s. jin , phys .
lett . * 92 * , 040403 ( 2004 ) ; m.w .
zwierlein , c.a .
stan , c.h .
schunck , s. m. f. raupach , a. j. kerman , w. ketterle , phys .
lett . * 92 * , 120403 ( 2004 ) ; m. bartenstein , a. altmeyer , s. riedl , s. jochim , c. chin , j. h. denschlag , r. grimm , phys . rev .
lett . * 92 * , 120401 ( 2004 ) ; t. bourdel , l. khaykovich , j cubizolles , j. zhang , f. chevy , m. teichmann , l. tarruell , s.j.j.m.f .
kokkelmans , c. salomon , phys .
. lett . * 93 * , 050401 ( 2004 ) ; c. chin , m. bartenstein , a. altmeyer , s. riedl , s. jochim , j. hecker denschlag , r. grimm , science * 305 * , 1128 ( 2004 ) ; j. kinast , a. turlapov , j. e. thomas , q. chen , j. stajic , k. levin , science * 307 * , 1296 ( 2005 ) .
m. greiner , o. mandel , t. esslinger , t. w. hnsch , and i. bloch , nature * 415 * , 39 ( 2002 ) .
g. montambaux and e. akkermans , _ physique msoscopique des lectrons et des photons _
( edp sciences , 2004 ) .
p. a. lee and t. v. ramakrishnan , rev .
57 * , 287 ( 1985 ) .
b. kramer and a. mackinnon , rep .
. phys . * 56 * , 1469 ( 1993 ) .
b. damski , j. zakrzewski , l. santos , p. zoller , and m. lewenstein phys . rev .
* 91 * , 080403 ( 2003 ) j. von neumann and e. wigner , phys . z.
* 30 * , 465 ( 1929 ) .
this paper contains a superficial algebraic error , that has been corrected in : f. h. stillinger and d. r. herrick , phys .
a * 11 * , 446 ( 1975 ) .
p. w. anderson , phys . rev . * 109 * , 1492 ( 1958 ) .
d. s. wiersma , p. bartolini , a. lagendijk , and r. righini , nature * 390 * , 671 ( 1997 ) .
t. jonckheere , c. a. mueller , r. kaiser , c. miniatura , d. delande , phys .
* 85 * , 4269 ( 2000 ) ; g. labeyrie , d. delande , c. a. mueller , c. miniatura , r. kaiser , europhys . lett . * 61 * , 327 ( 2003 ) ; g. labeyrie , d. delande , r. kaiser , c. miniatura , cond - mat/0603153 .
j. m. vogels , c. c. tsai , r. s. freeland , s. j. j. m. f. kokkelmans , b. j. verhaar , and d. j. heinzen , phys .
a * 56 * , r1067 ( 1997 ) ; s. inouye , m.r .
andrews , j. stenger , h .- j .
miesner , d.m .
stamper - kurn , and w. ketterle , nature * 392 * , 151 - 154 ( 1998 ) .
d. boiron , c. mennerat - robilliard , j .-
fournier , l. guidoni , c. salomon , and g. grynberg , eur .
j. d * 7 * , 373 ( 1999 ) .
j. e. lye , l. fallani , m. modugno , d.s .
wiersma , c. fort , and m. inguscio , phys .
* 95 * , 070401 ( 2005 ) .
d. clment , a. f. varon , m. hugbart , j. a. retter , p. bouyer , l. sanchez - palencia , d. m. gangardt , g. v. shlyapnikov , and a. aspect _ _ , _ _ phys
lett . * 95 * , 170409 ( 2005 ) . c. fort , l. fallani , v. guarrera , j. e. lye , m. modugno , d. s. wiersma , and m. inguscio , phys .
lett . * 95 * , 170410 ( 2005 ) .
t. schulte , s. drenkelforth , j. kruse , w. ertmer , j. arlt , k. sacha , j. zakrzewski , and m. lewenstein , phys .
lett . * 95 * , 170411 ( 2005 ) . m. modugno , phys .
a * 73 * , 013606 ( 2006 ) .
r. c. kuhn , c. miniatura , d. delande , o. sigwarth , c. a. mller , phys .
lett . * 95 * , 250403 ( 2005 ) .
u. gavish and y. castin , phys .
* 95 * , 020401 ( 2005 ) .
m. rusek , a. orlowski , phys . rev .
e * 51 * , r2763 ( 1995 ) .
emmanuel mandonnet , phd thesis of universit paris 6 ( 2000 ) , online at http://tel.ccsd.cnrs.fr/tel-00011872 .
s. e. hamann , d. l. haycock , g. klose , p. h. pax , i. h. deutsch , and p. s. jessen , phys .
. lett . * 80 * , 4149 ( 1998 ) .
h. perrin , a. kuhn , i. bouchoule , and c. salomon , europhys .
lett . * 42 * , 395 ( 1998 ) .
we used the asymptotic formula for the half - width of the ground energy band in a 1d standing wave of modulation depth @xmath580 : @xmath581 $ ] , a formula taken from r. campbell , _ thorie gnrale de lquation de mathieu et de quelques autres quations diffrentielles de la mcanique _ ( masson , 1955 ) . with this definition of the tunneling time
, the probability for the @xmath5 particle not to have tunneled during @xmath582 behaves as @xmath583 at short times , where @xmath584 is the spatial dimension of the lattice . if @xmath4 is a gas of fermionic @xmath34k atoms ( @xmath585 nm ) and @xmath5 a gas of @xmath26rb , the calculation for @xmath586 nm yields @xmath587 . due to the smaller detuning and the larger mass ratio @xmath325
, the optical potential would induce noticeable effects on the potassium matter wave .
d. petrov , phys .
a * 67 * , 010703(r ) ( 2003 ) .
( 11,12 ) of @xcite are derived under the condition that the kinetic energy of the three particles _ in their center - of - mass frame _ is less than the dimer binding energy @xmath588 : @xmath589 where @xmath590 , @xmath591 , @xmath592 are the wavevectors of the two @xmath4 atoms and @xmath593 is the wavevector of the @xmath5 atom .
whereas low enough values of @xmath594 may be chosen by adjusting the density of @xmath4 atoms , the value of @xmath595 is fixed by the zero - energy oscillation of the trapped @xmath5 particle , @xmath596 .
the @xmath597 limit of eq .
( [ eq : ekin ] ) therefore results in a condition that looks difficult to satisfy in general for @xmath59 : @xmath598 in the limit @xmath599 , however , as it is the case for the li - rb mixture , the vibrational energy in eq .
( [ eq : tough ] ) is multiplied by a small number @xmath600 whereas the binding energy @xmath588 increases as @xmath601 : for @xmath602 and @xmath603 , we thus find that the condition is satisfied for @xmath74li , @xmath75rb , so that eqs .
( 11,12 ) of @xcite may be used .
the @xmath604 matrix of the matter wave is related to @xmath92 by @xmath605_{jl } |\mathbf{r}_j\rangle \langle \mathbf{r}_l|$ ] .
j. m. ziman , _ models of disorder _ ( cambridge university press , cambridge , england , 1979 ) , chapter 8 .
this results from the fact that the hamiltonian @xmath606 is real in position space .
let @xmath607 by definition of @xmath188 , @xmath608 where @xmath609 is a real differential operator .
taking the limit @xmath610 of the imaginary part of the latter equation leads to @xmath611 alternatively , using the spectral expansion of the hamiltonian and of the resolvent with real eigenwavefunctions , one finds @xmath612 , where @xmath613 is the projector on the eigenspace of @xmath606 of energy @xmath97 .
this shows that , when one takes all possible values of @xmath614 , the wavefunctions @xmath104 span the entire eigenspace of energy @xmath102 . for @xmath615 ,
the density of states vanishes in 3d , and so does @xmath104 .
h. furstenberg , trans .
* 108 * , 377 ( 1963 ) ; for the application of this theorem to a product of transfer matrices , see k. ishii , supplement of the progress of theoretical physics , * 53 * , 77 ( 1973 ) , and v. baluni , j. willemsen , phys .
a * 31 * , 3358 ( 1985 ) .
this does not hold , however , when @xmath119 or @xmath116 is a real negative number , in which case one recovers the generic behavior of @xmath138 depicted in fig.[cap : schema_1d]a .
this means that @xmath382 is a ` bad ' choice , corresponding to the location of a node of the localized state .
we define @xmath616 as the average of the 6 values @xmath617 , with @xmath618 chosen between the 6 values @xmath619 .
similarly , we define @xmath620 as the average of the 36 values @xmath621 , with @xmath622 chosen between the 6 values @xmath623 , and @xmath618 as before .
requiring that the modulus square of the wavefunction decreases by four orders of magnitude over the radius of the medium ( which is a minimal requirement to pretend to be localized ) yields for the parameters of fig .
[ cap : etats3d ] a maximal localization length of the order of @xmath624 .
the numerical observation of larger values of @xmath130 would require a much larger scattering medium .
a. shalgi and y. imry , les houches session lxi , mesoscopic quantum physics ( e. akkermans , g. montambaux , j .- l . pichard and j. zinn - justin eds . ) , 1995 .
a strong reformulation of this far field extrapolation technique is to note that , out of a sphere containing the scattering medium , the function @xmath625 so that it admits the multipolar expansion on spherical harmonics , @xmath626 , where @xmath627 solves the spherical bessel differential equation with the asymptotic behavior at large @xmath191 , @xmath628 .
the complex coefficients @xmath629 obey the exact relation : @xmath630 where @xmath199 is obtained from the far field extrapolation .
so showing that this @xmath199 is small amounts to showing that _ all _ the amplitudes @xmath629 in the various output channels @xmath631 are small .
we also note that the intuitive fact that @xmath632 is minimal in @xmath215 leads to a real value for @xmath218 .
note that @xmath633 has to be positive for consistency of the present low-@xmath481 theory : with the reasoning leading to eq .
( [ eq : useful_phi ] ) , we get the approximation for the resolvent @xmath634 .
the linearization procedure in @xmath20 around @xmath139 leads to @xmath635 $ ] , where @xmath219 , @xmath636 and @xmath161 is given by eq .
( [ eq : width ] ) .
all the complex poles of the analytic continuation of the resolvent have a negative imaginary part , so that one should have @xmath637 and therefore @xmath638 .
this is compatible with the ` motion ' of the low - imaginary part eigenvalues of @xmath92 in the complex plane for increasing @xmath20 , see fig.[cap : eigenvalues - of - m ] .
we restrict to @xmath40-wave scattering for the @xmath4 waves , since we expect the @xmath3-wave scattering amplitude to vanish generically as @xmath640 at low @xmath20 . there may be values of @xmath641 leading to a @xmath3-wave resonance , but we assume that they differ from the locations of the @xmath40-wave resonances .
r. a. duine and h. t. c. stoof , phys . rep . *
396 * , 115 ( 2004 ) .
v. v. flambaum , g. f. gribakin , and c. harabati , phys .
a * 59 * , 1998 ( 1999 ) .
m. olshanii , phys .
81 * , 938 ( 1998 ) .
t. bergeman , m. g. moore , and m. olshanii , phys .
lett . * 91 * , 163201 ( 2003 ) .
v. peano , m. thorwart , c. mora , and r. egger , new j. phys .
* 7 * , 192 ( 2005 ) .
p. o. fedichev , m. j. bijlsma , and p. zoller , phys .
lett . * 92 * , 080401 ( 2004 ) .
d. s. petrov , m. holzmann , and g.v .
shlyapnikov , phys .
lett . * 84 * , 2551 ( 2000 ) .
h. moritz , t. stferle , k. gnter , m. khl , and t. esslinger , phys .
lett . * 94 * , 210401 ( 2005 ) .
t. stferle , h. moritz , k. gnter , m. khl , and t. esslinger , phys .
* 96 * , 030401 ( 2006 ) .
in fact , to first order in the perturbation series the term @xmath642 gives an energy shift proportional to @xmath643 while the term @xmath644 gives zero to first order , and a complex shift proportional to @xmath645 to second order .
one may wonder if one can , from this formalism suited to the case @xmath599 , propose an approximate formula giving the value of @xmath646 close to the resonance with a negative @xmath647 and leading to the desired value of @xmath43 .
we find that the answer is yes if one imposes the constraint @xmath648 , in which case one has the expansion @xmath649 the first term in the right - hand side actually corresponds to the born approximation . because of the imposed constraint on @xmath650 , this formula may be not very useful in practice . some reader may fear that the matter wave @xmath4 spends a very long time ` inside ' a given scatterer because scattering is resonant .
however the wigner time delay @xcite , @xmath651 , where @xmath652 is the @xmath40-wave phase shift , has here a negative value @xmath653 ( for @xmath654 and using the low-@xmath20 approximation @xmath655 ) .
in particular , this would allow to discriminate long dwell times due to localization effects from a parasitic matter wave trapping effect in the scattering medium , due e.g. to the formation of a weakly bound state between a matter wave atom and a trapped atom as discussed in
[ subsec : model_ham ] .
note that this particular spurious contribution to @xmath367 could also be eliminated by switching back @xmath43 to a low value just before measuring @xmath367 , so as to transform the weakly bound states into deeply bound states that can be discriminated from free atoms .
micro - gravity can presently be obtained e.g. at zarm ( a 120 meters high drop - tower facility in bremen , germany , where an experiment with ultracold atoms is planned @xcite ) , on airplanes performing parabolic flights ( experiments with atomic clocks on parabolic flights have been reported in @xcite ) , or on an orbiting space station ( experiments with ultracold atoms on the international space station have been planned by the european space agency ) .
e. rasel , j. reichel , private communication .
p. laurent , p. lemonde , e. simon , g. santarelli , a. clairon , n. dimarcq , p. petit , c. audoin , and c. salomon , eur .
j. d * 3 * , 201 ( 1998 ) .
a careful reader may fear that the fraction in the right hand side of eq .
( [ eq : approx_out ] ) is indeed of large modulus but almost real , so that its imaginary part is actually not large . as shown in the appendix [ appen : math ] ,
the vector @xmath449 is almost real so that the numerator of the considered fraction is almost real . in the limit @xmath656 ,
@xmath496 involves a factor @xmath657 so it can not be real for all @xmath191 .
however , using the calculation of the appendix [ appen : magique ] , one finds that the angular average of @xmath658 is of order @xmath659 , so that the imaginary part of @xmath496 at large @xmath191 is @xmath520 and tends to zero with @xmath190 . from schwartz inequality latexmath:[$|\sum_j x_j \exp(i
k\mathbf{n}\cdot\mathbf{r}_j)|^2 \leq of @xmath504 @xcite gives a value @xmath661 in the large @xmath662 limit , where @xmath149 is the diameter of the scattering medium , that is a value @xmath663 times smaller than the present upper bound .
g. bruun , y. castin , r. dum , k. burnett , eur .
j. d * 7 * , 433 - 439 ( 1999 ) .
r. p. feynman and a. r. hibbs ,
" quantum mechanics and path integrals " , mc graw - hill ( 1965 ) , chap . 4 .
@xmath664 for @xmath120 .
the result for @xmath117 is the complex conjugate of the latter expression . | the possibility of using ultracold atoms to observe strong localization of matter waves is now the subject of a great interest , as undesirable decoherence and interactions can be made negligible in these systems .
it was proposed that a static disordered potential can be realized by trapping atoms of a given species in randomly chosen sites of a deep 3d optical lattice with no multiple occupation .
we analyze in detail the prospects of this scheme for observing localized states in 3d for a matter wave of a different atomic species that interacts with the trapped particles and that is sufficiently far detuned from the optical lattice to be insensitive to it .
we demonstrate that at low energy a large number of 3d strongly localized states can be produced for the matter wave , if the effective scattering length describing the interaction of the matter wave with a trapped atom is of the order of the mean distance between the trapped particles .
such high values of the effective scattering length can be obtained by using a feshbach resonance to adjust the free space inter - species scattering length and by taking advantage of confinement - induced resonances induced by the trapping of the scatterers in the lattice . |
extremely high energy ( ehe ) cosmic rays beyond @xmath1 have been observed in a couple of decades @xcite , but their origin still remains enigmatic . in regard to the generation of the ehe particles ,
there are two alternative schemes : the `` top - down '' scenario that hypothesizes topological defects , z - bursts , and so on , and the traditional `` bottom - up '' ( see , e.g. , * ? ? ?
* for a review ) . in the latter approach ,
we explore the candidate celestial objects operating as a cosmic - ray `` zevatron '' @xcite , namely , an accelerator boosting particle kinetic energy to zev ( @xmath3 ) ranges . by simply relating the celestial size to the gyroradius for the typical magnetic field strength
, one finds that the candidates are restricted to only a few objects ; these include pulsars , active galactic nuclei ( agns ) , radio galaxy lobes , and clusters of galaxies @xcite .
in addition , gamma - ray bursters ( grbs ) are known as possible sources @xcite . as for the transport of ehe particles from the extragalactic sources , within the gzk horizon
@xcite the trajectory of the particles ( particularly protons ) ought to suffer no significant deflection due to the cosmological magnetic field , presuming its strength of the order of @xmath4 ( see , e.g. , * ? ? ? * for a review ) .
according to a cross - correlation study @xcite , some super - gzk events seem to be well aligned with compact , radio - loud quasars .
complementarily , self - correlation study is in progress , showing small - scale anisotropy in the distribution of the arrival direction of ehe primaries @xcite .
more recently , the strong clustering has been confirmed , as is consistent with the null hypothesis of isotropically distributed arrival directions @xcite . at the moment
, the interpretation of these results is under active debate . in the bottom - up scenario , the most promising mechanism for achieving ehe
is considered to be that of diffusive shock acceleration , which has been substantially studied for solving the problems of particle acceleration in heliosphere and supernova remnant ( snr ) shocks ( see , e.g. , * ? ? ? * for a review ) . in general , it calls for the shock to be accompanied by some kinds of turbulence that serve as the particle scatterers @xcite . concerning the theoretical modeling and its application to extragalactic sources such as agn jets , grbs , and so forth , it is still very important to know the actual magnetic field strength , configuration , and turbulent state around the shock front .
at this juncture , modern polarization measurements by using very long baseline interferometry began to reveal the detailed configuration of magnetic fields in extragalactic jets , for example , the quite smooth fields transverse to the jet axis ( 1803 + 784 : * ? ? ? * ) .
another noticeable result is that within the current resolution , a jet is envisaged as a bundle of _ at least _ a few filaments ( e.g. , 3c84 : @xcite ; 3c273 : @xcite ) , as were previously confirmed in the radio arcs near the galactic center ( gc ; * ? ? ?
* ; * ? ? ?
* ) , as well as in the well - known extragalactic jets ( e.g. , cyga : @xcite , @xcite ; m87 : @xcite ) .
the morphology of filaments can be self - organized via the nonlinear development of the electromagnetic current filamentation instability ( cfi ; * ? ? ?
* and references therein ) that breaks up a uniform beam into many filaments , each carrying about net one unit current @xcite . fully kinetic simulations indicated that this subsequently led to the coalescence of the filaments , self - generating significant toroidal ( transverse ) components of magnetic fields @xcite .
as could be accommodated with this result , large - scale toroidal magnetic fields have recently been discovered in the gc region @xcite .
accordingly , we conjecture that a similar configuration appears in extragalactic objects , particularly agn jets @xcite . it is also pointed out that the toroidal fields could play a remarkable role in collimating plasma flows @xcite . relating to this point ,
the agn jets have narrow opening angles of @xmath5 in the long scales , although in close proximity to the central engine the angles tend to spread ( e.g. , @xmath6 for the m87 jet ; * ? ? ?
* ) . moreover , there is observational evidence that the internal pressures are higher than the pressures in the external medium ( e.g. , 4c32.69 : @xcite ; cyga : @xcite ; m87 : @xcite ) .
these imply that the jets must be _ self_-collimating and stably propagating , as could be explained by the kinetic theory @xcite . in the nonlinear stage of the cfi
, the magnetized filaments can often be regarded as strong turbulence that more strongly deflects the charged particles .
when the shock propagation is allowed , hence , the particles are expected to be quite efficiently accelerated for the dsa scenario @xcite .
indeed , such a favorable environment seems to be well established in the agn jets .
for example , in the filamentary m87 jet , some knots moving toward the radio lobe exhibit the characteristics of shock discontinuity @xcite , involving circumstantial evidence of in situ electron acceleration @xcite .
as long as the shock accelerator operates for electrons , arbitrary ions will be co - accelerated , providing that the ion abundance in the jet is finite ( e.g. , * ? ? ?
* ; * ? ? ?
it is , therefore , quite significant to study the feasibility of ehe particle production in the filamentary jets with shocks : this is just the original motivation for the current work .
this paper has been prepared to show a full derivation of the diffusion coefficient for cosmic - ray particles scattered by the magnetized filaments .
the present theory relies on a consensus that the kinetic energy density ( ram pressure ) of the bulk plasma carrying currents is larger than the energy density of the magnetic fields self - generated via the cfi , likely comparable to the thermal pressure of the bulk .
that is , the flowing plasma as a reservoir of free energy is considered to be in a high-@xmath0 state . in a new regime
in which the cosmic - ray particles interact off - resonantly with the magnetic turbulence having no regular field , the quasi - linear approximation of the kinetic transport equation is found to be consistent with the condition that the accelerated particles must be rather free from magnetic traps ; namely , the particles experience meandering motion .
it follows that the diffusion anisotropy becomes small .
apparently , these are in contrast with the conventional quasi - linear theory ( qlt ) for small - angle resonant scattering , according to which one sets the resonance of the gyrating particles bound to a mean magnetic field with the weak turbulence superimposed on the mean field @xcite .
it is found that there is a wide parameter range in which the resulting diffusion coefficient is smaller than that from a simplistic qlt in the low-@xmath0 regime .
we compare a specified configuration of the filaments to an astrophysical jet including agn jets and discuss the correct treatment of what the particle injection threshold in the present context could be .
we then apply the derived coefficient for calculations of the dsa timescale and the achievable highest energy of accelerated particles in that environment . as a matter of convenience
, we also show some generic scalings of the highest energy , taking account of the conceivable energy restrictions for both ions and electrons . in order to systematically spell out the theoretical scenario ,
this paper is divided into two major parts , consisting of the derivation of the diffusion coefficient ( 2 ) and its installation to the dsa model ( 3 ) .
we begin , in 2.1 , with a discussion on the turbulent excitation mechanism due to the cfi , so as to specify a model configuration of the magnetized current filaments .
then in 2.2 , we explicitly formulate the equation that describes particle transport in the random magnetic fluctuations . in 2.3 ,
the power - law spectral index of the magnetic fluctuations is suggested for a specific case . in 2.4 , we write down the diffusion coefficients derived from the transport equation . in 3.1 , we deal with the subject of particle injection , and in 3.2 , we estimate the dsa timescale , which is used to evaluate the maximum energy of an accelerated ion ( 3.3 ) and electron ( 3.4 ) .
finally , 4 is devoted to a discussion of the feasibility and a summary .
in what follows , given the spatial configuration of the magnetized filaments of a bulk plasma jet , we derive the evolution equation for the momentum distribution function of test particles , which is linear to the turbulent spectral intensity , and then extract an effective frequency for collisionless scattering and the corresponding diffusion coefficient from the derived equation .
respecting the macroscopic transport of energetic particles in active galaxies , there is direct / indirect observational evidence that they are ejected from the central core of the galaxies and subsequently transferred , through bipolar jets , to large - scale radio robes in which the kinematic energy considerably dissipates ( e.g. , * ? ? ?
* ; * ? ? ?
* ) . in this picture
, it is expected that the directional plasma flows will favorably induce huge currents in various aspects of the above transport process ( e.g. , @xcite ; an analogous situation also seems to appear in grb jets : e.g. , @xcite ) . because of perfect conductivity in fully ionized plasmas , hot currents driven in , e.g. , the central engine prefer being quickly compensated by plasma return currents
this creates a pattern of the counterstreaming currents that is unstable for the electromagnetic cfi .
as is well known , the pattern is also unstable to electrostatic disturbances with the propagation vectors parallel to the streaming direction , but in the present work , we eliminate the longitudinal modes , so as to isolate the transverse cfi . providing a simple case in which the two uniform currents are carried by electrons , the mechanism of magnetic field amplification due to the cfi
is explained as follows . when the compensation of the counterpropagating electron currents is disturbed in the transverse direction ,
magnetic repulsion between the two currents reinforces the initial disturbance . as a consequence , a larger and larger magnetic field
is produced as time increases .
for the weibel instability as an example , the unstable mode is the purely growing mode without oscillations @xcite , so the temporal variation of magnetic fields is expected to be markedly slow in the saturation regime ( more on these is given in 2.2 and 2.3 ) .
note that a similar pattern of quasi - static magnetic fields can be also established during the collision of electron - positron plasmas @xcite and in a shock front propagating through an ambient plasma with / without initial magnetic fields @xcite .
these dynamics might be involved in the organization of the knotlike features in the fanaroff - riley ( fr ) type i radio jets , which appear to be a shock caused by high - velocity material overtaking slower material ( e.g. , * ? ? ?
similarly , the cumulative impingement could also take place around the hot spots of the fr type ii sources , which arguably reflect the termination shocks .
in fact , the filamentary structure has been observed in the hot spot region of a fr ii source @xcite .
furthermore , estimating the energy budget in many radio lobes implies that the ram pressure of such current - carrying jets is much larger than the energy density of the magnetic fields @xcite ; namely , the jet bulk can be regarded as a huge reservoir of free energy . in this regime ,
the ballistic motion is unlikely to be affected by the self - generated magnetic fields , as is actually seen in the linear feature of jets . as a matter of fact
, the gc region is known to arrange numerous linear filaments , including nonthermal filaments @xcite .
taking these into consideration , we give a simple model of the corresponding current magnetic field system and attempt to unambiguously distinguish the present system from the one that appears in the low-@xmath0 plasmas hitherto well studied . in figure 1 , for a given coordinate , we depict the configuration of the linear current filaments and turbulent magnetic fields of the bulk plasma . recalling that magnetic field perturbations develop in the direction transverse to the initial currents ( e.g. , *
* ) , one supposes the magnetic fields developed in the nonlinear phase to be @xmath7 @xcite , such that the vectors of zeroth - order current density point in the directions parallel and antiparallel to the @xmath8-direction , i.e. , @xmath9 , where the scalar @xmath10 ( @xmath11 ) is nonuniformly distributed on the transverse @xmath12-@xmath13 plane , while uniformly distributed in the @xmath8-direction .
note that for the fluctuating magnetic field vectors , we have used the simple character ( @xmath14 ) without any additional symbol such as `` @xmath15 , '' since the establishment of no significant regular component is expected , and simultaneously , @xmath16 ( @xmath17 ) well embodies the quasi - static current filaments in the zeroth order . for convenience ,
hereafter , the notations `` parallel '' ( @xmath18 ) and `` transverse '' ( @xmath19 ) are referred to as the directions with respect to the linear current filaments aligned in the @xmath8-axis , as they are well defined reasonably ( n.b . in the review of 2.4 , @xmath20 and @xmath21 with the subscript `` @xmath22 '' refer to a mean magnetic field line ) .
it is mentioned that the greatly fluctuating transverse fields could be reproduced by some numerical simulations ( e.g. , * ? ? ?
* ; * ? ? ?
in an actual filamentary jet , a significant reduction of polarization has been found in the center , which could be ascribed to the cancellation of the small - scale structure of magnetic fields @xcite , compatible with the present model configuration . in addition
, there is strong evidence that random fields accompany grb jets ( e.g. , * ? ? ?
the arguments expanded below highlight the transport properties of test particles in such a bulk environment , that is , in a forest of magnetized current filaments .
we are particularly concerned with the stochastic diffusion of the energetic test particles injected into the magnetized current filaments ( for the injection problem , see the discussion in
3.1 ) . as a rule ,
the vlasov equation is appropriate for describing the collisionless transport of the relativistic particles in the turbulent magnetic field , @xmath23 , where @xmath24 and the slow temporal variation has been taken into consideration .
transverse electrostatic fields are ignored , since they preferentially attenuate over long timescales , e.g. , in the propagation time of jets ( see 2.3 ) .
the temporal evolution of the momentum distribution function for the test particles , @xmath25 , can then be described as @xmath26f_{\bf p}=0 \label{eqn:1}\ ] ] for arbitrary particles . here
@xmath27 is the particle charge , for electrons , @xmath28 for positrons , and @xmath29 for ions or nuclei , where @xmath30 and @xmath31 are the elementary charge and the charge number , respectively . ]
@xmath32 is the speed of light , and the other notations are standard
. we decompose the total distribution function into the averaged and fluctuating part , @xmath33 , and consider the specific case in which from a macroscopic point of view , the vector @xmath7 is randomly distributed on the transverse @xmath12-@xmath13 plane @xcite .
taking the ensemble average of equation ( [ eqn:1 ] ) , @xmath34 , then yields @xmath35\delta f_{\bf p}\right > , \label{eqn:2}\ ] ] where we have used @xmath36 . taking account of no mean field
implies that we do not invoke the gyration and guiding center motion of the particles .
subtracting equation ( [ eqn:2 ] ) from equation ( [ eqn:1 ] ) and picking up the term linear in fluctuations , viz . ,
employing the conventional quasi - linear approximation , we obtain @xmath37\left < f_{\bf p}\right>. \label{eqn:3}\ ] ] as usual , equation ( [ eqn:3 ] ) is valid for @xmath38 @xcite . as shown in 3.1
, this condition turns out to be consistent with the aforementioned implication that the injected test particles must be free from the small - scale magnetic traps embedded in the bulk . relating to this , note that to remove the ambiguity of terminologies , the injected , energetic test particles obeying @xmath25 are just compared to the cosmic rays that are shown below to be diffusively accelerated owing to the present scenario . within the framework of the test particle approximation , the back - reaction of the slow spatiotemporal change of @xmath39 to the modulation of @xmath14 ( sustained by the bulk )
is ignored , in contrast to the case for snr environments , where such effects often become nonnegligible ( e.g. , * ? ? ?
* ) . in general
, the vector potential conforms to @xmath40 and @xmath41 .
for the standard , plane wave approximation , we carry out the fourier transformation of the fluctuating components for time and the transverse plane : @xmath42}{\rm d}^3{\cal k } , \label{eqn:4}\ ] ] @xmath43}{\rm d}^3{\cal k } , \label{eqn:5}\ ] ] @xmath44}{\rm d}^3{\cal k } , \label{eqn:6}\ ] ] and @xmath45 , where @xmath46 , @xmath47 , and @xmath48 .
the given magnetic field configuration follows @xmath49 ( @xmath50 ) and @xmath51 . as illustrated in figure 1
, the scalar quantity @xmath52 ( @xmath11 ) is also random on the transverse plane , with no mean value .
making use of equations ( [ eqn:4])([eqn:6 ] ) , equation ( [ eqn:3 ] ) can be transformed into @xmath53^{-1 } \left[{\bf v}\times\left({\bf k}\times{\bf a}_{\cal k}\right)\right ] \cdot{{\partial\left <
f_{\bf p}\right>}\over{\partial{\bf p}}}. \label{eqn:7}\ ] ] on the other hand , the right - hand side ( rhs ) of equation ( [ eqn:2 ] ) can be written as @xmath54 } \left\{\left[{\bf v}\times\left({\bf k}^{\prime}\times { \bf a}_{{\cal k}^{\prime}}\right)\right ] \cdot{\partial\over{\partial{\bf p}}}\right\}\delta f_{\bf p}\right>. \label{eqn:8}\ ] ] substituting equation ( [ eqn:4 ] ) ( involving eq .
[ [ eqn:7 ] ] ) into equation ( [ eqn:8 ] ) , equation ( [ eqn:2 ] ) can be expressed as @xmath55 -\left(\omega+\omega^{\prime}\right)t\right\}}\right .
\left[{\bf v}\times\left({\bf k}^{\prime}\times { \bf a}_{{\cal k}^{\prime}}\right)\right ] \cdot{\partial\over{\partial{\bf p } } } \left\{{{\left[{\bf v}\times\left({\bf k}\times { \bf a}_{\cal k}\right)\right]}\over { \omega-\left({\bf k}\cdot{\bf v}\right ) } } \cdot{\partial\over{\partial{\bf p } } } \right\}\left < f_{\bf p}\right>\right > , \label{eqn:9}\end{aligned}\ ] ] where the definition of the total derivative , @xmath56 , has been introduced . as for the integrand of equation ( [ eqn:9 ] ) ,
it may be instructive to write down the vector identity of @xmath57 from the general expression of equation ( [ eqn:9 ] ) , we derive an effective collision frequency that stems from fluctuating field - particle interaction , as shown below . for convenience ,
we decompose the collision integral ( rhs of eq .
[ [ eqn:9 ] ] ) including the scalar products , @xmath58 , into the four parts : @xmath59 where @xmath60 and @xmath61 in the following notations , the subscripts `` 1 '' and `` 2 '' indicate the parallel ( @xmath18 ) and perpendicular ( @xmath19 ) direction to the current filaments , respectively .
below , as an example , we investigate the contribution from the integral @xmath62 ( see appendix for calculation of the other components ) . for
the purely parallel diffusion involving the partial derivative of only @xmath63 , the first term of the rhs of equation ( [ eqn:10 ] ) does not make a contribution to equation ( [ eqn:9 ] ) . in the ordinary case in which the random fluctuations are stationary and homogeneous
, the correlation function has its sharp peak at @xmath64 and @xmath65 @xcite , that is , @xmath66 where the dirac @xmath15-function has been used . here
note the relation of @xmath67 because we have @xmath68 , where the superscript asterisk indicates the complex conjugate ; this is valid as far as @xmath69 is real , i.e. , @xmath23 is observable . by using equation ( [ eqn:13 ] ) , the integral component @xmath62 can be expressed as @xmath70 , \label{eqn:14}\ ] ] where the relation of @xmath71 has been used . in order to handle the resonant denominator of equation ( [ eqn:14 ] )
, we introduce the causality principle of @xmath72^{-1 } \rightarrow{\cal p}\left[\omega-\left({\bf k}\cdot{\bf v}\right)\right]^{-1 } -i\pi\delta\left[\omega-\left({\bf k}\cdot{\bf v}\right)\right]$ ] , where @xmath73 indicates the principal value @xcite .
one can readily confirm that the real part of the resonant denominator does not contribute to the integration .
thus , we have @xmath74 \left({\bf k}\cdot{\bf v}\right ) { \partial\over{\partial p_{\parallel } } } \left
< f_{\bf p}\right>\right\}. \label{eqn:15}\ ] ] equation ( [ eqn:15 ] ) shows the generalized form of the quasi - linear equation , allowing @xmath75 to be arbitrary functions of @xmath76 and @xmath77 .
, the frequency dependence of the correlation function can be summarized in the form of @xmath78 .
however , this is not the case considered here .
the free - energy source that drives instability is now current flows ; thereby , unstable modes without oscillation ( or with quite slow oscillation ) can be excited . ] in the present circumstances , a typical unstable mode of the cfi is the purely growing weibel mode with @xmath79 in collisionless regimes , although in a dissipative regime the dephasing modes with a finite but small value of @xmath80 are possibly excited @xcite . in the latter case , the spectral lines will be broadened in the nonlinear phase .
nevertheless , it is assumed that the spectrum still retains the peaks around @xmath81 , accompanied by their small broadening of the same order , where @xmath82 , and @xmath83 and @xmath84 are the growth rate and the plasma frequency , respectively . in the special case reflecting the purely growing mode , the spectrum retains a narrow peak at @xmath85 with @xmath86 @xcite .
apparently , the assumed quasi - static properties are in accordance with the results of the fully kinetic simulations @xcite , except for a peculiar temporal property of the rapid coalescence of filaments .
accordingly , here we employ an approximate expression of @xmath87 where @xmath88 .
note that when taking the limit of @xmath89 , equation ( [ eqn:16 ] ) degenerates into @xmath90 . substituting equation ( [ eqn:16 ] ) into equation ( [ eqn:15 ] ) yields @xmath91 \left({\bf k}\cdot{\bf v}\right ) { \partial\over{\partial p_{\parallel } } }
f_{\bf p}\right>\right\}. \label{eqn:17}\ ] ] furthermore , we postulate that the turbulence is isotropic on the transverse plane , though still , of course , allowing anisotropy of the vectors @xmath92 parallel to the @xmath8-axis .
equation ( [ eqn:17 ] ) can be then cast to @xmath93 where @xmath94 , @xmath95 , and @xmath96 . as concerns the integration for @xmath97
, we see that the contribution from the marginal region of the smaller @xmath98 , reflecting narrower pitch angle , is negligible . in astrophysical jets ,
the pitch angle distribution for energetic particles still remains unresolved , although the distribution itself is presumably unimportant .
hence , at the moment it may be adequate to simply take an angular average , considering , for heuristic purposes , the contribution from the range of @xmath99 for a small value of @xmath100 . if one can choose @xmath101 , the above relation , @xmath102 , reflects the off - resonant interaction , i.e. , @xmath103 .
the minimum wavenumber , @xmath104 , is typically of the order of the reciprocal of the finite system size , which is , in the present circumstances , larger than the skin depth @xmath105 .
these ensure the aforementioned relation of @xmath106 ( or @xmath86 ) .
in addition , the off - resonance condition provides an approximate expression of @xmath107 . using this expression , the integral for the angular average
can be approximated by @xmath108 .
this is feasible , on account of the negligible contribution from the angle of @xmath109
. then equation ( [ eqn:18 ] ) reduces to @xmath110 where we have defined the modal energy density ( spectral intensity ) of the quasi - static turbulence by @xmath111 , such that the magnetic energy density in the plasma medium can be evaluated by @xmath112 .
the energy density of the quasi - static magnetic fields , @xmath113 , likely becomes comparable to the thermal pressure of the filaments @xcite .
when exhibiting such a higher @xmath113 level , the bulk plasma state may be regarded as the strong turbulence ; but recall that in the nonlinear cfi , the frequency spectrum with a sharp peak at @xmath85 is scarcely smoothed out , since significant mode - mode energy exchanges are unexpected .
this feature is in contrast to the ordinary magnetohydrodynamic ( mhd ) and electrostatic turbulence , in which a larger energy density of fluctuating fields would involve modal energy transfer .
one of the most remarkable points is that as long as the validity condition of the quasi - linear approximation , @xmath114 , is satisfied ( for details , see
3.1 ) , the present off - resonant scattering theory covers even the strong turbulence regime .
that is , the theory , which might be classified into an extended version of the qlt , does not explicitly restrict the magnetic turbulence to be weak ( for instruction , tsytovich & ter haar [ 1995 ] have considered a generalization of the quasi - linear equation in regard to its application to strong electrostatic turbulence ) .
apparently , this is also in contrast to the conventional qlt for small - angle resonant scattering , which invokes a mean magnetic field ( well defined only for the case in which the turbulence is weak ) in ordinary low-@xmath0 plasmas . in any case , in equation ( [ eqn:19 ] ) we specify the spectral intensity of the random magnetic fields , which are established via the aforementioned mechanism of the electromagnetic cfi .
the closely related analysis in the nonlinear regime was first performed by @xcite , for a simple case in which two counterstreaming electron currents compensate for a uniform , immobile ion background . in the static limit of @xmath115
, they have derived the modal energy densities of fluctuating electrostatic and magnetic fields , by using statistical mechanical techniques .
they predicted the accumulation of magnetic energy at long wavelengths , consistent with the corresponding numerical simulation @xcite .
it was also shown that at long wavelengths , the energy density of a transverse electrostatic field was comparable to the thermal energy density . however , when allowing ion motions , such an electrostatic field is found to attenuate significantly , resulting in equipartition of the energy into magnetic and thermal components @xcite .
that is why we have neglected the electrostatic field in equation ( [ eqn:1 ] ) .
when the spectral intensity of the magnetic fluctuations can be represented by a power - law distribution of the form @xmath116 we refer to @xmath117 as spectral index .
@xcite found that for the transverse magnetic fields accompanying anisotropic current filaments , the spectral index could be approximated by @xmath118 in a wide range of @xmath119 , that is , @xmath120 note that the spectral index is somewhat larger than @xmath121 for the classical mhd context @xcite .
the larger index is rather consistent with the observed trends of softening of filamentary turbulent spectra in extragalactic jets ( e.g. , @xmath122 in cyg a ; * ? ? ? * and references therein ) .
although the turbulent dissipation actually involves the truncation of @xmath123 in the short - wavelength regions , we simply take @xmath124 , excluding the complication . using equation ( [ eqn:20 ] ) and the expression of the magnetic energy density @xmath113 , we find the relation of @xmath125 for @xmath126 .
the spectral details for individual jets ( such as the bend - over scales of @xmath123 , correlation length , and so on ; e.g. , for the heliosphere , see * ? ? ?
* ; * ? ? ?
* ) will render the integration of equation ( [ eqn:22 ] ) more precise , but the related observational information has been poorly updated thus far . for the present purpose , we simply use equation ( [ eqn:22 ] ) , setting @xmath127 , where @xmath128 stands for the radius of the jet , which is actually associated with the radius of a bundle of filaments of various smaller radial sizes ( e.g. , * ? ? ? * ) .
this ensures that the coherence length of the fluctuating force , @xmath129 , is small compared with a characteristic system size , i.e. , the transverse size , as is analogous to the restriction for use of the conventional qlt . in order to evaluate the diffusion coefficients of test particles
, one needs to specify the momentum distribution function , @xmath39 , in equation ( [ eqn:19 ] ) . as is theoretically known , the fermi acceleration mechanisms lead to the differential spectrum of @xmath130 [ or @xmath131 ; @xcite ] , where @xmath132 defines the density of particles with kinetic energy between @xmath133 and @xmath134 .
for the first - order fermi mechanism involving nonrelativistic shock with its compression ratio of @xmath135 , the power - law index reads @xmath136 , accommodated by the observational results . with reference to these
, we have the momentum distribution function of @xmath137 for the ultrarelativistic particles having @xmath138 , such that in the isotropic case , the differential quantity @xmath139 corresponds to @xmath132 defined above ( e.g. , * ? ? ?
then , in equation ( [ eqn:19 ] ) the partial derivative of the distribution function can be estimated as @xmath140(c^{2}/e^{2})\left < f_{\bf p}\right>$ ] , where we have used @xmath141 . making use of this expression and equation ( [ eqn:22 ] ) ,
equation ( [ eqn:19 ] ) can be arranged in the form of @xmath142 . here
@xmath143 reflects an effective collision frequency related to the purely parallel diffusion in momentum space , to give @xmath144\psi_{2}}\over { \pi\alpha}}{{cq^{2}b^{2}r}\over{e^{2 } } } , \label{eqn:23}\ ] ] where we have used the definitions of @xmath145 , and @xmath146 and @xmath147 , whereby @xmath148 .
similarly , one can calculate the other components of the integral @xmath149 as outlined in the appendix and arrange them in the form of @xmath150 . as a result
, we obtained @xmath151 and @xmath152 as would be expected , we confirm a trivial relation of @xmath153 , stemming from the orthogonality in the rhs of equation ( [ eqn:2 ] ) .
now we estimate the spatial diffusion coefficients in an ad hoc manner : @xmath154 .
it is then found that the off - diagonal components , @xmath155 and @xmath156 , include the factor of @xmath157 , implying that these components vanish for an average . for @xmath158 and @xmath159 reflecting the momentum isotropy , the diffusion coefficients can be summarized in the following tensor form : @xmath160 , \label{eqn:26}\end{aligned}\ ] ] where @xmath161 and @xmath162 .
the perpendicular component can be expressed as @xmath163 , where @xmath164 . note the allowable range of @xmath165 for @xmath166 ; particularly , @xmath167 for the expected range of @xmath168 .
it may be instructive to compare the diffusion coefficient of equation ( [ eqn:26 ] ) with that derived from the previously suggested theories including the qlt . in weakly turbulent low-@xmath0 plasmas , the mean magnetic field with its strength @xmath169 ,
which can bind charged particles and assign the gyroradius of @xmath170 , provides a well - defined direction along the field line ; therefore , in the following discussion , we refer , for convenience , to @xmath171 and @xmath172 as the parallel and perpendicular directions to the mean magnetic field , respectively . for a simplistic qlt ,
one sets an ideal environment in which the turbulent alfvn waves propagating along the mean field line resonantly scatter the bound particles , when @xmath173 , where @xmath174 is the parallel wavenumber @xcite .
assuming that the particles interact with the waves in the inertial range of the turbulent spectrum with its index @xmath175 , the parallel diffusion coefficient could be estimated as @xcite @xmath176 for @xmath177 and @xmath178 , where @xmath179 reflects the correlation length of the turbulence and @xmath180 ( @xmath181 ) defines the energy density ratio of the turbulent / mean field . in the special case of @xmath182 , referred to as the bohm diffusion limit @xcite , one gets the ordering @xmath183 , where @xmath184 denotes the bohm diffusion coefficient for ultrarelativistic particles . considering the energy accumulation range of smaller @xmath174 for the kolmogorov turbulence with @xmath185 , @xcite derived a modified coefficient that recovered the scaling of equation ( [ eqn:27 ] ) in the region of @xmath186 .
as for the more complicated perpendicular diffusion , a phenomenological hard - sphere scattering form of the coefficient is @xmath187 in the bohm diffusion limit ; and @xcite suggested a somewhat extended version , @xmath188 $ ] ( referred to as @xmath189 below ) , where @xmath190 is the parallel mean free path ( mfp ) . a significantly improved theory of perpendicular diffusion has recently been proposed by @xcite , including nonlinearity incorporated with the two - dimensional wavevector @xmath191 , whereupon for @xmath185 , @xcite have derived an approximate expression of the corresponding diffusion coefficient , although it still exhibits a somewhat complicated form ( referred to as @xmath192 ) . on the other hand , within the present framework
the gyroradius of the injected energetic particles can not be well defined , because of @xmath193 ( 2.1 and 2.2 ) .
nonetheless , in order to make a fair comparison with the order of the components of equation ( [ eqn:26 ] ) , the variable @xmath169 is formally equated with @xmath194 .
in addition , the correlation length is chosen as @xmath195 , corresponding to the setting in 2.3
. then the ratio of @xmath196 for @xmath197 to equation ( [ eqn:27 ] ) is found to take a value in the range of @xmath198 for the expected values of @xmath117 , @xmath168 . here
@xmath199 and @xmath29 have been introduced .
similarly , we get the scaling of @xmath200 , and @xmath201 for @xmath185 and @xmath202 followed by @xmath203 . furthermore , considering the parameters given in @xcite , which can be accommodated with the above @xmath202 , we also have @xmath204 in the leading order of @xmath192 , proportional to @xmath205 .
these scalings are valid for arbitrary species of charged particles ; for instance , setting @xmath206 reflects electrons , positrons , or protons ( see footnote 1 ) .
particularly , for @xmath207 in equation ( [ eqn:28 ] ) , the efficiency of the present dsa is expected to be higher than that of the conventional one based on the simplistic qlt invoking parallel diffusion @xcite .
this can likely be accomplished for high-@xmath31 particles , as well as electrons with lower maximum energies . here
note that @xmath208 can not take an unlimitedly smaller value with decreasing @xmath133 , since the effects of cold particle trapping in the local magnetic fields make the approximation of no guide field ( eq . [ [ eqn:2 ] ] ) worse ; and the lower limit of @xmath133 is relevant to the injection condition called for the present dsa .
more on these is given in
3 . to apply the dsa model
, one needs the effective diffusion coefficient for the direction normal to the shock front , referred to as the shock - normal direction . for convenience , here we write down the coefficient for the general case in which the current filaments are inclined by an angle of @xmath209 with respect to the shock - normal direction . in the tensor transformation of @xmath210 , where @xmath211
@xmath212 we identify the shock - normal component @xmath213 with @xmath214 .
it can be expressed as @xmath215 or simply as @xmath216 for @xmath217 , where the subscripts @xmath218 , @xmath219 indicate the upstream and downstream regions , respectively .
the expression of equation ( [ eqn:31 ] ) appears to be the same as equation ( 4 ) in @xcite . however , note again that now @xmath18 and @xmath19 refer to the direction of the linear current filaments , compared to an astrophysical jet ( 2.1 and fig .
we consider the particle injection mechanism that makes the present dsa scenario feasible , retaining the validity of the quasi - linear approximation . then , using the diffusion coefficient ( eq .
[ [ eqn:26 ] ] ) , we estimate the dsa timescale for arbitrary species of charged particles and calculate , by taking the competitive energy loss processes into account , at the achievable highest energies of the particles in astrophysical filaments . in the usual dsa context ,
equation ( [ eqn:31 ] ) that calls equation ( [ eqn:26 ] ) determines the cycle time for one back - and - forth of cosmic - ray particles across the shock front , which is used below for evaluation of the mean acceleration time ( 3.2 ; * ? ? ?
here we note that equation ( [ eqn:26 ] ) is valid for a high - energy regime in which the test particles with @xmath133 are unbound to the local magnetic fields , so as to experience the nongyrating motion .
as shown below , this limitation can be deduced from the validity condition of the quasi - linear approximation that has been employed in 2.2 .
using equations ( [ eqn:4 ] ) and ( [ eqn:7 ] ) , the validity condition @xmath220 can be rewritten as @xmath221 \cdot{{\partial\left < f_{\bf p}\right>}\over{\partial{\bf p } } } \right\}\right| , \label{eqn:32}\ ] ] where the off - resonance interaction with the quasi - static fluctuations has been considered ( 2.2 )
. for the momentum distribution function of @xmath222 for the statistically accelerated particles with @xmath138 ( 2.4 ) , the rhs of equation ( [ eqn:32 ] ) is of the order of @xmath223\left < f_{\bf p}\right>$ ] for @xmath224 .
therefore , we see that within the present framework , the quasi - linear approximation is valid for the test particles with an energy of @xmath225 , in a confinement region .
note that this relation ensures the condition that the gyroradius for the local field strength of @xmath226 greatly exceeds the filament size ( coherence length ) of order @xmath129 , namely , @xmath227 [ equivalently , @xmath225 ] , except for a marginal region of @xmath229 .
obviously , this means that in the high - energy regime of @xmath230 , the test particles are not strongly deflected by a local magnetic field accompanying a fine filament with its transverse scale of @xmath129 . on the other hand , in the cold regime of @xmath231 , the test particles
are tightly bound to a local magnetic field having the ( locally defined ) mean strength , violating equation ( [ eqn:2 ] ) .
here it is expected that the bound particles can diffuse along the local field line , and hence , diffusion theories for a low-@xmath0 plasma are likely to be more appropriate , rather than the present theory . summarizing the above discussions
, there seem to exist two distinct energy regimes for the test particles confined in the system comprising numerous magnetized filaments : the higher energy regime of @xmath230 , in which the particles are free from the local magnetic traps , and the lower energy regime of @xmath231 , in which the particles are bound to the local fields , as compared to a low-@xmath0 state .
the hierarchy is illustrated in figure 2 , indicating the characteristic trajectories of those particles .
when shock propagation is allowed , as seen in actual agn jets , the shock accelerator can energize the particles in each energy level . at the moment , we are particularly concerned with ehe particle production by a feasible scenario according to which energetic free particles , unbound to small - scale structure of the magnetized filaments , are further energized by the shock . in this aspect , the particle escape from magnetically bound states , due to another energization mechanism , can be regarded as the injection of preaccelerated particles into the concerned diffusive shock accelerator . if the preaccelerator , as well , is of dsa , relying on the gyromotion of bound particles @xcite , the preaccelerator also calls for the injection ( in a conventional sense ) in a far lower energy level , owing to , e.g. , the maxwellian tail , or the energization of particles up to the energies where the pre - dsa turns on ( for a review , see * ? ? ? * ) . the energy required for this injection , the so - called injection energy ,
could be determined by , e.g. , the competition with collisional resistance . in order to distinguish from this
commonly used definition of `` injection , '' we refer to the corresponding one , owing to the particle escape from the magnetic traps , as the `` transition injection , '' in analogy to the bound - free transition in atomic excitation .
the energy required to accomplish of the transition injection is formally denoted as @xmath232 , where @xmath233 represents a threshold potential energy .
that is , the particles with @xmath27 and the energy exceeding @xmath234 are considered to spaciously meander to experience successive small deflection by the fields of many filaments ( fig .
2 ) , such that the present theory is adequate for describing the particle diffusion .
this scattering property can be compared to that for the conventional qlt in low-@xmath0 regimes : an unperturbed ( zeroth order ) guiding center trajectory of gyrating particles bound to a mean magnetic field must be a good approximation for many coherence lengths of particle scatterer . if both the injection and transition injection work , the multistep dsa can be realized .
in the stage of @xmath235 , many acceleration scenarios that have been proposed thus far ( dsa : e.g. , @xcite ; shock drift acceleration : e.g. , @xcite ; some versions of the combined theories : e.g. , @xcite ; for a review , see , e.g. , @xcite ) can be candidates for the mechanism of the preacceleration up to the energy range of @xmath236 , although before achieving this energy level , the acceleration , especially for electrons , might be , in some cases , knocked down by the energy loss , such as synchrotron cooling , collision with photons , and so on .
the relevant issues for individual specific situations are somewhat beyond the scope of this paper ( observability is discussed in 4 ) . here
we just briefly mention that in the termination regions of large - scale jets where the bulk kinetic energy is significantly converted into the magnetic and particle energies , a conventional dsa mechanism involving large - scale mhd turbulence might work up to ehe ranges ( honda & honda [ 2004b ] for an updated scenario of oblique dsa of protons ) . in the following ,
we focus on the dsa of energetic free particles after the transition injection .
let us consider a typical case of @xmath237 in equation ( [ eqn:31 ] ) , reflecting a reasonable situation that a shock wave propagates along the jet comprising linear filaments . since the vectors of the random magnetic fields are on the plane transverse to the current filaments , this plane is perpendicular to the shock - normal direction .
that is , the shock across the perpendicular magnetic fields is considered . in this case , no irregularity of magnetic surfaces in the shock - normal direction exists , because of @xmath238 . however
, this does _ not _ mean that the particle flux diffusively across the shock surface is in free - streaming ; note that the particles crossing the local fields with nonsmall pitch angles suffer the orthogonal deflection .
anyhow , the injected particles are off - resonantly scattered by the filamentary turbulence , to diffuse , migrating back and forth many times between the upstream and downstream regions of the shock . as a consequence
, a small fraction of them can be stochastically accelerated to very high energy .
this scenario is feasible , as long as the filamentary structure can exist around the discontinuity , as seen in a kinetic simulation for shock propagation @xcite and an actual filamentary jet @xcite .
the timescale of this type of dsa is of the order of the cycle time for one back - and - forth divided by the energy gain per encounter with the shock @xcite . here the cycle time is related to the mean residence time of particles ( in regions i and ii ) , which is determined by the diffusive particle flux across the shock , dependent on @xmath239 ( eq . [ [ eqn:31 ] ] ) . for the moment , the shock speed is assumed to be nonrelativistic .
actually , this approximation is reasonable , since the discrete knots ( for fr i ) and hot spots ( for fr ii ) , which are associated with shocks ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , preferentially move at a nonrelativistic speed , slower than the speed of the relativistic jets ( e.g. , * ? ? ?
when taking the first - order fermi mechanism into consideration for calculation of the energy gain , the mean acceleration time can be expressed as @xcite @xmath240 where @xmath241 and @xmath242 are the flow speed of the upstream and downstream regions in the shock rest frame , respectively .
the present case of @xmath243 ( in eq .
[ [ eqn:31 ] ] ) provides @xmath244 , where @xmath245 is given in equation ( [ eqn:26 ] ) . here
note the relation of @xmath246 , derived from the condition that the current density , @xmath247 , must be continuous across the shock front . when assuming @xmath248 and @xmath249 , we arrive at the result @xmath250 where the definitions of @xmath251 , @xmath252 , @xmath253 , and @xmath254 have been introduced .
equation ( [ eqn:34 ] ) is valid for arbitrary species of particles `` @xmath255 '' having energy @xmath256 , spectral index @xmath257 , and charge @xmath258 .
note that for the plausible ranges of the values of @xmath117 , @xmath257 , and @xmath259 , the value of equation ( [ eqn:34 ] ) does not significantly change .
the @xmath260 dependence is also small , because of @xmath167 , reflecting three - dimensional rms deflection of unbound particles ( 3.1 and fig .
2 ) . in the scaling laws shown below , for convenience we use the typical values of @xmath261 @xcite and @xmath262 ( for the strong shock limit ) , although we indicate , in equation ( [ eqn:39 ] ) , the parameter dependence of the numerical factor . in equation ( [ eqn:34 ] ) for ions ( @xmath263i ) , we set @xmath264 ( see footnote 1 ) and @xmath265 ( e.g. , * ? ? ?
* ; * ? ? ?
* ) . by balancing equation ( [ eqn:34 ] ) with the timescale of the most severe energy loss process
, we derive the maximum possible energy defined as @xmath266 . in the environment of astrophysical filaments including extragalactic jets ,
the phenomenological time balance equation can be expressed as @xmath267 where @xmath268 , @xmath269 , @xmath270 , and @xmath271 stand for the timescales of the shock propagation ( 3.3.1 ; eq .
[ [ eqn:36 ] ] ) , the synchrotron loss for ions ( 3.3.2 ; eq .
[ [ eqn:40 ] ] ) , the photodissociation of the nucleus ( 3.3.3 ; e.g. , eq . [
[ eqn:43 ] ] ) , and the collision of nucleus `` @xmath272 '' with target nucleus `` @xmath273 '' ( 3.3.4 ; e.g. , eq . [ [ eqn:46 ] ] ) , respectively .
in addition , the energy constraint ascribed to the spatial scale , i.e. , the quench caused by the particle escape , should also be taken into account ( 3.3.5 ) .
the individual cases are investigated below . in the actual circumstances of astrophysical jets , the propagation time of a shock through the jet , @xmath268 ,
restricts the maximum possible energy of accelerated particles .
the shock propagation time may be interpreted as the age of knots or hot spots @xcite , which can be crudely estimated as @xmath274 , where @xmath275 represents a distance from the central engine to the knot or hot spot being considered and @xmath276 an average speed of their proper motion . when assuming @xmath277 , we get the scaling @xmath278 for the case in which the shock is currently alive as is observed in agn jets , @xmath268 can not be compared to the `` lifetime '' of the accelerator that is considered in snr shocks ( e.g. , * ? ? ?
it is mentioned that in agn jets , the timescale of adiabatic expansion loss might be estimated as @xmath279 , where @xmath280 and @xmath281 represent the lorentz factor of jet bulk flows and the speed of radial expansion , respectively @xcite .
the fact that the jets are collimating well with an opening angle of @xmath282 means @xmath283 ; thereby , @xmath284 for @xmath285 .
thus , it is sufficient to pay attention to the limit due solely to the shock propagation time .
these circumstances are also in contrast with those in the snrs , where the flows are radially expanding without collimation , and the shock propagation time ( or lifetime ) just reflects the timescale of adiabatic expansion loss ( e.g. , * ? ? ?
* ) . in equation ( [ eqn:35 ] ) ,
let us first consider the case of @xmath286 . by equating ( [ eqn:34 ] ) with ( [ eqn:36 ] )
, we obtain the following expression for the maximum possible energy of an accelerated ion : @xmath287 note the ratio of @xmath288 for the narrow opening angle of agn jets of @xmath289 and not - so - small viewing angle @xmath290 ( e.g. , for the m87 jet , @xmath291 for @xmath292 [ @xcite ] and @xmath293 [ @xcite ] or @xmath294 [ @xcite ] ) . equation ( [ eqn:37 ] ) ( and eq . [ [ eqn:48 ] ] shown below ) corresponds to the modified version of the simple scaling originally proposed by @xcite . concerning the abundance of high-@xmath31 elements and their acceleration to ehe regimes , the following points 14 may be worth noting : 1 .
radial metallicity gradients are expected to be enhanced in elliptical galaxies ( e.g. , * ? ? ?
along with this , a significant increase of heavy elements has been discovered in the central region of the nearby giant elliptical galaxy m87 @xcite , which contains a confirmed jet .
2 . a variety of heavy ions including iron have been detected in a microquasar jet ( ss433 ; * ? ? ?
3 . the haverah park data favor proton primaries below an energy of @xmath295 , whereas they appear to favor a heavier composition above it @xcite .
4 . the recent fly s eye data of @xmath296 are compatible with the assumption of a hadron primary between proton and iron nuclei @xcite . with reference to this observational evidence , we take the possibility of acceleration of ( or deceleration by ) heavy particles into consideration and indicate the charge ( @xmath31 ) and/or atomic number ( @xmath52 ) dependence of the maximum possible energies and loss timescales .
the particles deflected by the random magnetic fields tend to emit unpolarized synchrotron photons , which can be a dominant cooling process . for relativistic ions ,
the timescale can be written as @xmath297 $ ] , where @xmath298 denotes the proton rest mass @xcite . in this expression ,
the energy of an accelerated ion , @xmath299 , can be evaluated by equating @xmath300 with @xmath269 .
that is , we have @xmath301^{1/3}. \label{eqn:38}\ ] ] here the dimensionless factor @xmath302 is given by @xmath303^{1/3 } , \label{eqn:39}\ ] ] and @xmath304 stands for the classical radius of the electron , where @xmath305 is the electron rest mass . substituting equation ( [ eqn:38 ] ) into the expression of @xmath306
, the cooling timescale can be expressed as a function of the physical parameters of the target object . as a result
, we find @xmath307 practically , this expression can be used in equation ( [ eqn:35 ] ) for making a direct comparison with the other loss timescales .
for example , in the fr sources with @xmath308 @xcite , we have @xmath309 , so that the synchrotron cooling loss is ineffective .
it should , however , be noted that in blazars with @xmath310 @xcite , equation ( [ eqn:40 ] ) becomes , in some cases , comparable to equation ( [ eqn:36 ] ) .
when the equality of @xmath311 is fulfilled in equation ( [ eqn:35 ] ) , equation ( [ eqn:38 ] ) just provides the maximum possible energy of the accelerated ion , which scales as @xmath312 the important point is that @xmath313 and @xmath306 are both proportional to @xmath314 , so that the @xmath315 dependence of @xmath316 is canceled out .
this property also appears in the case of electron acceleration attenuated by the synchrotron cooling ( 3.4.1 ) . in equation ( [ eqn:41 ] )
, it appears that for heavier ions , @xmath316 takes a larger value . in the actual situation ,
however , the extremely energetic ions possess a long mfp , and therefore , acceleration may be quenched by the particle escape , as discussed in 3.3.5 . here
we focus on the proton - photon collision that engenders a pion - producing cascade .
the characteristic time of the collision depends on the target photon spectrum @xmath317 in the acceleration site , where @xmath317 is the number density of photons per unit energy interval for photon energy @xmath318 . for @xmath319 ( e.g. , *
* ) , typical for the fr sources @xcite , the timescale can be expressed as @xmath320t_{\rm p , syn}$ ] , where @xmath321 for the average cross section of @xmath322 @xcite , @xmath323 denotes the average energy density of target photons , and @xmath324 ( the subscript `` p '' indicates proton ) .
thus , the expression of @xmath325 includes @xmath326 , i.e. , the energy of the accelerated proton .
this can be evaluated by equating @xmath327 with @xmath325 , to have the form of @xmath328^{1/3 } , \label{eqn:42}\ ] ] where the definition @xmath329 has been introduced . substituting equation ( [ eqn:42 ] ) into the expression of @xmath325 , we obtain the following scaling of the photomeson cooling time :
@xmath330 note that for @xmath331 , we have @xmath332 .
if the equality of @xmath333 is satisfied in equation ( [ eqn:35 ] ) , then equation ( [ eqn:42 ] ) gives the maximum possible energy , which scales as @xmath334 for @xmath335 , equation ( [ eqn:44 ] ) is identical with equation ( [ eqn:41 ] ) for @xmath336 .
the nucleus - nucleus collisions involving spallation reactions can also be a competitive process in high - density regions .
for proton - proton collision , the timescale can be simply evaluated by @xmath337 , where @xmath338 is the number density of target protons , and @xmath339 denotes the cross section in high - energy regimes .
the timescale can be rewritten as @xmath340 it is found that for tenuous jets with @xmath341 , the value of equation ( [ eqn:45 ] ) is larger than the conceivable value of equation ( [ eqn:36 ] ) ; that is , the collisional loss is ineffective . for the collision of an accelerated proton with a nonproton nucleus , the timescale can be evaluated by the analogous notation , @xmath342 , where @xmath343 is the fractional number density of the target nuclei having atomic number @xmath344 . here
we use an empirical scaling of the cross section , @xmath345 , where @xmath346 , although the value of @xmath347 may be an overestimate for very high energy collisions ( e.g. , * ? ? ?
combining @xmath348 with @xmath349 , in general the timescale for collision of a proton with a nucleus of an arbitrary composition can be expressed as @xmath350 where @xmath338 and @xmath343 are both in units of @xmath351 . in equation ( [ eqn:35 ] ) , we consider the case of @xmath352 . by equating ( [ eqn:34 ] ) with ( [ eqn:46 ] )
, we obtain the following expression for the maximum possible energy of an accelerated proton : @xmath353 as for the collision of an arbitrary accelerated nucleus with a target nucleus , we can analogously estimate @xmath271 and @xmath354 . in particular , the heavier nucleus
proton collision is more important , since its timescale @xmath355 is of the order of @xmath356 : for larger @xmath52 and @xmath338 , it can be comparable to the other loss timescales .
for example , the parameters of @xmath357 ( iron ) and @xmath358 lead to @xmath359 . for the case of @xmath360 in equation ( [ eqn:35 ] )
, we have the scaling of @xmath361 , where @xmath362 is of equation ( [ eqn:47 ] ) for @xmath363 .
the particle escape also limits its acceleration ; that is , the spatioscale of the system brings on another energy constraint . relating to this point , in 2.4 we found the relation of @xmath167 , meaning that the anisotropy of the spatial diffusion coefficient is small .
it follows that the radial size of the jet ( rather than @xmath275 ) affects the particle confinement .
recall here that in the interior of a jet the magnetic field vectors tend to be canceled out , whereas around the envelope the uncanceled , large - scale ordered field can appear @xcite .
from the projected view of the jet , on both sides of the envelope the magnetic polarities are reversed .
the spatially decaying properties of such an envelope field in the external tenuous medium or vacuum might influence the transverse diffusion of particles .
the key property that should be recalled is that for @xmath364 distant from a filament , the magnetic field strength is likely to slowly decay , being proportional to @xmath365 @xcite .
it is , therefore , expected that as long as the radial size of the largest filament , i.e. , correlation length , is comparable to the radius of the jet ( 2.3 ) , the long - range field pervades the exterior of the jet , establishing the `` magnetotail '' with the decay property of @xmath366 for @xmath367 .
in fact , in a nearby radio galaxy , the central kiloparsec - scale `` hole '' of the inner radio lobe containing a jet is filled with an ordered , not - so - weak ( rather strong ) magnetic field of the order of @xmath368 @xcite , whose magnitude is comparable to ( or @xmath369 of ) that in the jet @xcite .
presumably , the exuding magnetic field plays an additional role in confining the leaky energetic particles with their long mfp of @xmath370 . in this aspect ,
let us express an effective confinement radius as @xmath371 , where @xmath372 , and impose the condition that the accelerator operates for the particles with the transverse mfp of @xmath373 .
then the equality gives the maximum possible energy in the form of @xmath374 values of @xmath354 ( and @xmath362 ) derived from the time balance equation ( [ eqn:35 ] ) can not exceed that of equation ( [ eqn:48 ] ) .
it appears that equation ( [ eqn:48 ] ) can be compared to the energy scaling derived from , in the simplest model , the maximum gyroradius in a uniform magnetic field . in a manner simiar to
that explained in 3.3 , we find the generic scaling for the achievable highest energy of electrons . in equation ( [ eqn:34 ] ) for electrons ( @xmath263e ) , we set @xmath375 and @xmath376 ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . by balancing equation ( [ eqn:34 ] ) with the timescale of a competitive energy loss process
, we derive the maximum possible energy defined as @xmath377 .
the time balance equation can be written as @xmath378 where @xmath379 , @xmath380 , and @xmath381 stand for the timescales of the synchrotron loss for electrons ( 3.4.1 ; eq .
[ [ eqn:51 ] ] ) , the inverse compton scattering ( 3.4.2 ; eq .
[ [ eqn:54 ] ] ) , and the bremsstrahlung emission loss ( 3.4.3 ; eq . [ [ eqn:57 ] ] ) , respectively . for positrons
the method is so analogous that we omit the explanation . for electrons
, the synchrotron cooling is a familiar loss process , and the timescale can be expressed as @xmath382 . in this expression ,
the energy of an accelerated electron , @xmath383 , can be evaluated by equating @xmath384 with @xmath385 , to give @xmath386^{1/3}. \label{eqn:50}\ ] ] substituting equation ( [ eqn:50 ] ) into the aforementioned expression of @xmath385 , the cooling timescale can be expressed as a function of the physical parameters of the target object : @xmath387 this can be used in equation ( [ eqn:49 ] ) for comparison with the other loss timescales . when the equality of @xmath388 is satisfied in equation ( [ eqn:49 ] ) , equation ( [ eqn:50 ] ) gives the maximum possible energy , which scales as @xmath389 according to the explanation given in 3.3.2 ,
equation ( [ eqn:52 ] ) is independent of @xmath315 ( see also eq . [ [ eqn:41 ] ] ) .
the striking thing is that for plausible parameters , the value of @xmath390 is significantly larger than that obtained in the context of the simplistic qlt invoking the alfvn waves @xcite .
this enhancement is , as seen in equation ( [ eqn:28 ] ) , attributed to the smaller value of the diffusion coefficient for electrons , which leads to a shorter acceleration time , i.e. , a smaller value of equation ( [ eqn:33 ] ) , and thereby to a higher acceleration efficiency . for the case of @xmath391 , the inverse compton scattering of accelerated electrons with target photons can be a dominant loss process .
actually , the environments of agn jets often allow the synchrotron self - compton ( ssc ) and/or external compton ( ec ) processes .
the characteristic time of the inverse comptonization can be estimated as @xmath392 , where @xmath393 and @xmath394 denote the total cross sections in the thomson limit of @xmath395 and the klein - nishina regime of @xmath396 , respectively ( e.g. , * ? ? ?
the expression of @xmath380 includes @xmath383 , which is determined by numerically solving the balance equation of @xmath397 .
because of @xmath398 , the value of @xmath383 is found to be in the region of @xmath399^{1/3 } , \label{eqn:53}\ ] ] in the whole range of @xmath318 .
note that the equality in equation ( [ eqn:53 ] ) reflects the thomson limit of @xmath400 . substituting the value of @xmath383 into the expression of @xmath380
, we can evaluate the scattering time , which takes a value in the range of @xmath401 for a given parameter @xmath402 , the larger value of @xmath323 depresses the lower bound of @xmath380 , though the klein - nishina effects prolong the timescale . it should be noted that the evaluation of @xmath380 along equation ( [ eqn:54 ] ) is , in equation ( [ eqn:49 ] ) , meaningful only for @xmath403 ; that is , the relation of @xmath404 ensures @xmath405 . for the case of @xmath397 in equation
( [ eqn:49 ] ) , @xmath383 , conforming to equation ( [ eqn:53 ] ) , gives the maximum possible energy , which takes the value of @xmath406 for @xmath403 .
again , note that the thomson limit sets the lower bound of @xmath390 .
it is found that the klein - nishina effects enhance the value of @xmath390 in the regime of @xmath407 .
note here that @xmath390 can not unlimitedly increase in actual circumstances but tends to be limited by the synchrotron cooling . combining equation ( [ eqn:52 ] ) with equation ( [ eqn:55 ] ) , therefore , we can express the allowed domain of the variables as follows : @xmath408 note that the upper bound reflects the synchrotron limit . in the critical case of @xmath409 reflecting the energy equipartition
, the generic equation ( [ eqn:56 ] ) degenerates into equation ( [ eqn:52 ] ) .
the bremsstrahlung emission of electrons in the coulomb field of nuclei whose charge is incompletely screened also affects the acceleration .
the timescale can be evaluated by the notation @xmath410 , where @xmath411 is the fractional number density of the target nuclei having charge number @xmath412 and @xmath413 describes the radiation cross section ( e.g. , * ? ? ?
when the screening effects are small , for interaction with a heavy composite we have @xmath414 \sum_{z^{\prime}}n_{z^{\prime}}z^{\prime 2 } \right\}^{-1}~~{\rm s } , \label{eqn:57}\ ] ] where @xmath411 is in units of @xmath351 . in the peculiar environments of high density , enhanced metallicity , and lower magnetic and photon energy densities ,
equation ( [ eqn:57 ] ) may be comparable with equation ( [ eqn:51 ] ) or ( [ eqn:54 ] ) . in ordinary agn jets , however , the corresponding physical parameters seem to be marginal . note also that the bremsstrahlung timescale for ion - ion interactions is larger , by the order of @xmath415 , than the value of equation ( [ eqn:57 ] ) , and found to largely exceed the value of equation ( [ eqn:36 ] ) , namely , the age of the accelerator .
that is why the ion bremsstrahlung has been excluded in equation ( [ eqn:35 ] ) .
the feasibility of the present model could be verified by the measurement of energetic photons emanating from a source , typically , bright knots in nearby agn jets . in any case , the electrons with energy @xmath390 , given in equation ( [ eqn:56 ] ) , emit the most energetic synchrotron photons , whose frequency may be estimated as @xmath416 , where the mean field strength @xmath169 has been compared to the rms strength @xmath315 . for @xmath417 as an example , the frequencies of @xmath418 are found to be achieved for @xmath419 . in the gamma - ray bands ,
however , the energy flux of photons from the synchrotron originator is predicted to be often overcome by that produced by the inverse comptonization of target photons . in this case ,
as far as the condition of @xmath420 is satisfied , the boosted photon energy is given by @xmath421 , independent of the target photon energy @xmath318 , thereby irrespective of ssc or ecs .
this is in contrast to another case of @xmath422 , in which one has @xmath423 dependent on @xmath318 .
apparently , for the extremely high energy ranges of @xmath390 achieved in the present scheme , the former condition is more likely satisfied for a wide range of @xmath318
. therefore , in the circumstances that the source is nearby such that collision with the cosmic infrared background , involving photon - photon pair creation , is insignificant , @xmath424 ( @xmath425 ) just gives the theoretical maximum of gamma - ray energy , although the klein - nishina effects also take part in lowering the flux level .
this means , in turn , that a comparison of the @xmath424 value ( multiplied by an appropriate doppler factor ) with the observed highest energy of the compton emissions might constitute a method to verify the present dsa for electrons .
the case for this method is certainly solidified when the operation of the transition injection ( 3.1 ) is confirmed . making use of the inherent property that the synchrotron photons emitted by electrons having an energy above @xmath426 reduce their polarization ,
the energy hierarchy can be revealed by the polarization measurements , particularly , with wide frequency ranges and high spatioresolution .
according to the reasoning that the critical frequency above which the measured polarization decreases , @xmath427 , ought to be of the order of @xmath428^{2}[|e{\bf b}({\bf r})|/m_{\rm e}c]$ ] , the related coherence length can be estimated as @xmath429^{3}\}^{1/2}$ ] .
note that when the locally defined gyroradius reaches this critical scale , the bound electrons are released . in actual circumstances , @xmath430 and the polarization for a fixed frequency band
are , if anything , likely to increase near the jet surface , where the large - scale coherency could appear ( 3.3.5 )
. this may be responsible for the results of the polarization measurement of a nearby filamentary jet , which indicate a similar transverse dependence @xcite . in the sense of @xmath431 , the transition injection condition for electrons is more restrictive than that for ions .
thus , observational evidence of the present dsa scenario for energetic electrons will , if it is obtained , strongly suggest that the same scenario operates for ion acceleration , providing its finite abundance . to summarize
, we have accomplished the modeling of the diffusive shock accelerator accompanied by the quasi - static , magnetized filamentary turbulence that could be self - organized via the current filamentation instability .
the new theory of particle diffusion relies on the following conditions analogous to those for the conventional qlt : ( 1 ) the test particles must not be strongly deflected by a fine filament but suffer the cumulative small deflection by many filaments , and ( 2 ) the transverse filament size , i.e. , the coherence length of the scatterer , is limited by the system size transverse to the filaments ; whereas , more importantly , it is dependent on neither the gyration , the resonant scattering , nor the explicit limit of the weak turbulence .
we have derived the diffusion coefficient from the quasi - linear type equation and installed it in a dsa model that involves particle injection associated with the bound - free transition in the fluctuating vector potential . by systematically taking the conceivable energy restrictions into account , some generic scalings of the maximum energy of particles have been presented .
the results indicate that the shock in kiloparsec - scale jets could accelerate a proton and heavy nucleus to @xmath432 and @xmath433 ranges , respectively .
in particular , for high-@xmath31 particles , and electrons as well , the acceleration efficiency is significantly higher than that derived from a simplistic qlt - based dsa , as is deduced from equation ( [ eqn:28 ] ) .
consequently , the powerful electron acceleration to @xmath434 ranges becomes possible for the plausible parameters .
we expect that the present theory can be , mutatis mutandis , applied for solving the problem of particle transport and acceleration in grbs @xcite .
the topic is of a cross - disciplinary field closely relevant to astrophysics , high - energy physics , and plasma physics involving fusion science ; particularly , the magnetoelectrodynamics of filamentary turbulence is subject to the complexity of `` flowing plasma . '' in perspective , further theoretical details might be resolved , in part , by a fully kinetic approach allowing multiple dimensions , which goes far beyond the mhd context .
for instruction , we write down the derivation of equations ( [ eqn:24 ] ) and ( [ eqn:25 ] ) for the collisionless scattering of injected test particles by magnetized current filaments having the configuration illustrated in figure 1 . making use of equation ( [ eqn:10 ] ) , @xmath149 for @xmath438 ( eq . [ [ eqn:12 ] ] ) can be explicitly written as where we have used a standard correlation function ( eq . [ [ eqn:13 ] ] ) reflecting random magnetic fluctuations on the transverse plane to the linear current filaments ( see fig . 1 ) .
recalling the causality principle and noticing that the real part does not contribute to the integration , we get again , we use an ad hoc equation ( 16 ) , valid for a quasi - static mode that retains the narrow spectral peak around @xmath445 . assuming that the magnetic turbulence is isotropic on the transverse plane , the angular average of equations ( [ eqn : a4])([eqn : a6 ] ) is carried out . taking account of the off - resonant scattering of particles by the quasi - static random fields gives where the ordering and the definition of @xmath123 are the same as those denoted in 2.2 .
for a given momentum distribution function of @xmath137 for the test particles with an ultrarelativistic energy of @xmath449 , the partial derivatives can be estimated as where @xmath452 and @xmath453 . using equations ( [ eqn : a10 ] ) and ( [ eqn : a11 ] ) , and equation ( [ eqn:22 ] ) concerning the @xmath119-space integration for @xmath124
, one can arrange equations ( [ eqn : a7])([eqn : a9 ] ) in the form of @xmath454 to obtain the expressions of equations ( [ eqn:24 ] ) and ( [ eqn:25 ] ) .
+ asada , k. , kameno , s. , inoue , m. , shen , z .- q . , horiuchi , s. , & gabuzda , d. c. 2000 , in astrophysical phenomena revealed by space vlbi , ed . h. hirabayashi , p. g. edwards , & d. w. murphy ( sagamihara : isas ) , 51 | we expand the off - resonant scattering theory for particle diffusion in magnetized current filaments that can be typically compared to astrophysical jets , including active galactic nucleus jets . in a high plasma @xmath0 region where the directional bulk flow is a free - energy source for establishing turbulent magnetic fields via current filamentation instabilities ,
a novel version of quasi - linear theory to describe the diffusion of test particles is proposed .
the theory relies on the proviso that the injected energetic particles are not trapped in the small - scale structure of magnetic fields wrapping around and permeating a filament but deflected by the filaments , to open a new regime of the energy hierarchy mediated by a transition compared to the particle injection .
the diffusion coefficient derived from a quasi - linear type equation is applied to estimating the timescale for the stochastic acceleration of particles by the shock wave propagating through the jet .
the generic scalings of the achievable highest energy of an accelerated ion and electron , as well as of the characteristic time for conceivable energy restrictions , are systematically presented .
we also discuss a feasible method of verifying the theoretical predictions .
the strong , anisotropic turbulence reflecting cosmic filaments might be the key to the problem of the acceleration mechanism of the highest energy cosmic rays exceeding @xmath1 ( @xmath2 ) , detected in recent air shower experiments . |
the bimodality of gamma - ray burst ( grb ) durations @xcite is indicative of separate progenitor populations for long- and short - duration grbs . while direct observational support exists for the massive star origin of long grbs ( e.g. , @xcite ) , the most popular progenitor model for short grbs
is the coalescence of compact object binaries with neutron star and/or black hole constituents ( ns - ns / ns - bh ; @xcite ) .
one of the key predictions of this model is that some systems will experience large velocity kicks at birth , leading to eventual mergers well outside of the host galaxies , in galactic halos and the intergalactic medium @xcite .
these models predict that @xmath20 of all mergers will occur at offsets of @xmath21 kpc for milky way mass galaxies . in such environments the resulting afterglow emission
is expected to be fainter than for bursts occurring in coincidence with their host galaxies due to the low ambient density @xcite .
a subset of ns - ns binaries ( @xmath22 ) may be formed dynamically in globular clusters @xcite , leading to possible large offsets and low ambient densities despite an absence of kicks .
the distribution of offsets for such binary systems can in principle be calculated from the spatial distribution of globular clusters , and initial predictions are that @xmath23 of these systems will have offsets of @xmath21 kpc ( depending on the mass of the galaxy ; @xcite ) .
an additional expectation is that dynamically - formed binaries will be heavily skewed to lower redshifts due to the additional time delay between the formation and core - collapse of the globular clusters @xcite .
other progenitor systems for short grbs have also been proposed , including young magnetars @xcite , accretion - induced collapse ( aic ) of neutron stars @xcite , and delayed magnetar formation through binary white dwarf mergers or white dwarf aic @xcite .
these models are partially motivated by observations that can not be easily accommodated in the standard ns - ns merger model , such as extended soft @xmath0-ray emission on timescales of @xmath24 s ( e.g. , @xcite ) , or by phenomena such as short - duration giant flares from soft @xmath0-ray repeaters ( e.g. , @xcite ) .
the general expectation is that these alternative progenitors will not experience kicks , and will therefore lead to bursts in coincidence with host galaxies .
the detection of short grb afterglows starting in mid-2005 provided an opportunity to investigate the various progenitor models through a range of observational tests : the redshift distribution @xcite , the host galaxy demographics @xcite , the afterglow properties @xcite , and perhaps most importantly , their locations relative to the host galaxies @xcite .
as of mid-2010 , x - ray and optical afterglows have been detected from 40 and 20 short grbs , respectively , with the latter sample providing accurate sub - arcsecond positions .
of these 20 events , 15 directly coincide with host galaxies with a wide distribution of apparent magnitudes , and redshifts of @xmath25 or beyond ( e.g. , @xcite ) .
however , the remaining five events do not appear to coincide with galaxies , and therefore provide an opportunity to assess the possibility of large progenitor offsets , and to test the validity of the ns - ns merger models .
significant offsets have been claimed previously , in particular for grbs 050509b and 060502b with projected offsets of @xmath26 and @xmath27 kpc , respectively @xcite .
however , in both cases only x - ray positions are available ( @xmath28 and @xmath29 radius , respectively ) , and the error circles contain several galaxies consistent with a negligible offset @xcite . moreover , in the case of grb050509b the x - ray error circle intersected the outer regions of the host , raising the possibility that the progenitor system was formed in , rather than kicked to , the outskirts of the host . this possibility raises a crucial point , namely that a substantial physical offset from the center of the host does not necessarily point to a progenitor kick if the burst still closely coincides with the host light distribution @xcite .
an illustrative example of this point is grb071227 whose optical afterglow position coincides with the outskirts of edge - on disk galaxies , with an offsets of about 15 kpc from the host center @xcite .
large offsets have also been speculated in a few cases with precise optical afterglow positions ( grbs 061201 , 070809 , and 080503 ; @xcite ) .
however , these claims have not been investigated systematically , mainly because they were treated on a case - by - case basis , with probabilistic arguments that prevented conclusive associations .
these cases , combined with the ambiguity inherent to x - ray positions , demonstrate that bursts with optical afterglows are essential for reaching any robust conclusions about progenitor offsets ( due to kicks and/or a globular cluster origin ) . here
we present the first systematic study of short grbs with optical afterglows and no coincident hosts , which combines their afterglow properties with the large - scale environments .
the purpose of this study is to statistically assess the possibility of offsets and to compare this with alternative explanations ( e.g. , a high redshift origin ) . to achieve this goal we set our study in the broader context of short grbs that have optical afterglows _ and _ coincident hosts , as well as short grbs with only x - ray positions .
as we demonstrate throughout the paper , such a combined study is essential since offsets or high redshifts are expected to jointly affect both the afterglow properties and the large - scale environments .
our study also provides a robust statistical assessment of _ a posteriori _ chance coincidence probabilities , and the expected number of spurious associations , which can not be properly assessed for individual bursts .
the paper is organized as follows . in [ sec : obs ] we present deep optical observations of the environments of grbs 061201 , 070809 , 080503 , 090305 , and 090515 , as well as spectroscopic observations of bright galaxies near the burst positions for grbs 061201 , 070809 , and 090515 .
we study trends in the afterglow and prompt emission properties of bursts with and without coincident hosts in [ sec : ag ] , and determine _ a posteriori _ probabilities of chance coincidence as a function of projected angular offset for galaxies near the position of each burst in
[ sec : hosts ] .
we also determine projected physical and host - normalized offsets , and use these in conjunction with the afterglow and prompt emission properties to address two scenarios for the short bursts with optical emission and no coincident bright hosts : ( i ) an origin in faint galaxies at @xmath12 ( [ sec : highz ] ) , or ( ii ) substantial offsets from galaxies at @xmath30 or @xmath7 ( [ sec : kick ] ) . finally , in
[ sec : disc ] we present the offset distribution of all short grbs with optical afterglows ( in the context of scenario ii ) , and compare this distribution with predictions for ns - ns kicks and dynamically - formed ns - ns binaries .
we further investigate whether the circumburst densities that are required for the measured optical magnitudes can be accommodated with a halo or igm origin .
we draw conclusions about the progenitors of short grbs from our systematic study in
[ sec : conc ] .
we include in this investigation all 20 short grbs with optical afterglow detections as of june 2010 .
this is the full subset of events for which sub - arcsecond positions are available .
we stress that this sample represents only about 1/3 of all short grbs discovered to date , and about 1/2 of the sample with x - ray afterglow detections .
thus , it is not a complete sample of short grbs , but it is the only subset for which we can investigate the possibility of large offsets with meaningful statistical significance . as we demonstrate in [ sec : disc ] , we do not expect this sample to be strongly biased with respect to circumburst density , at least for @xmath31 @xmath32 .
the properties of the 20 short grbs , as well as events with deep optical limits , are summarized in table [ tab : shbs ] .
as can be inferred from the table , some of the 20 events with only x - ray detections do not have optical follow - up observations , suggesting that the sample with optical afterglows may be largely representative .
for the purpose of our investigation we define three sub - samples that will be used throughout the paper : ( i ) _ sample 1 _ : short grbs with detected afterglows and coincident host galaxies ( 15 bursts ) ; ( ii ) _ sample 2 _ : the 5 short bursts with detected optical afterglows and no bright coincident hosts ; and ( iii ) _ sample 3 _ : short grbs with detected x - ray afterglows ( from the _ swift _ x - ray telescope ) but no optical detections despite rapid follow - up observations ( 11 bursts ) . for the bursts in _ sample 2 _ we use deep space- and
ground - based optical observations to place limits on the brightness of underlying galaxies , and to assess the probability of chance coincidence for nearby galaxies .
for grb061201 we use _ hubble space telescope _ ( hst ) observations obtained with the advanced camera for surveys ( acs ) in the f814w filter , with a total exposure time of 2244 s @xcite . for grb070809
we use @xmath33-band observations obtained on 2008 january 14 ut with the low dispersion survey spectrograph ( ldss3 ) mounted on the magellan / clay 6.5-m telescope , with a total exposure time of 1500 s. for grb080503 we use hst wide - field planetary camera 2 ( wfpc2 ) observations in the f606w filter from 2009 january 30 ut , with a total exposure time of 4000 s @xcite .
we also use the limits on a coincident host inferred from a deeper stack of hst observations by @xcite , with @xmath34 mag . for grb090305
we use ldss3 @xmath33-band observations obtained on 2010 may 8 ut with a total exposure time of 2400 s. finally , for grb090515 we use @xmath33-band observations obtained with the gemini multi - object spectrograph ( gmos ) mounted on the gemini - north 8-m telescope from 2009 may 15 ut with a total exposure time of 1800 s. the ground - based observations were reduced and analyzed using standard routines in iraf .
the analysis of the hst observations is detailed in @xcite .
the limiting magnitudes for all five observations are listed in table [ tab : nohost ] , and images of the five fields are shown in figures [ fig:061201][fig:090515 ] .
in addition to the imaging observations , we obtained spectroscopic observations of galaxies near the positions of grbs 061201 , 070809 , and 090515 .
our observations of a galaxy located @xmath35 from the afterglow position of grb061201 ( marked `` s4 '' in figure [ fig:061201 ] ) revealed a star forming galaxy at @xmath36 @xcite ; see figure [ fig:061201spec ] . for grb070809 we obtained spectra of two galaxies located @xmath37 and @xmath38 from the optical afterglow position ( marked `` s2 '' and `` s3 '' , respectively in figure [ fig:070809 ] ) using ldss3 on 2008 january 14 ut .
the galaxy at a separation of @xmath37 was previously identified as a star forming galaxy at @xmath39 by @xcite . here
we find that the object at a separation of @xmath38 is an early - type galaxy at @xmath40 , with no evidence for on - going star formation activity ( figure [ fig:070809spec ] ) .
finally , for grb090515 we obtained multi - object spectroscopic observations with ldss3 for nearly 100 galaxies within a @xmath41 field centered on the grb position .
these observations provide redshifts for several galaxies near the host position , including a star forming galaxy at @xmath42 ( @xmath43 offset ; `` s1 '' in figure [ fig:090515 ] ) , an early - type galaxy at @xmath44 ( @xmath45 offset ; s5 ) , and a star forming galaxy at @xmath46 ( @xmath47 offset ; `` s6 '' ) ; see figure [ fig:090515spec ] . in an upcoming paper
we will demonstrate that the galaxy at @xmath44 is a member of a cluster .
we begin our investigation by assessing the distribution of optical afterglow magnitudes for the three samples defined in [ sec : obs ] .
the observed magnitudes and limits as a function of time after the burst are shown in figure [ fig : optag1 ] .
the median observation time for the sample is about 8.5 hr after the burst .
the distribution of detected optical afterglow magnitudes is broad , ranging from @xmath48 to @xmath49 mag .
the limits range from @xmath50 to @xmath51 mag ( with the exception of grb080702 which only has a shallow limit of about 21 ab mag : @xcite ) .
the mean brightness and standard deviation for _ sample 1 _ are @xmath52 mag .
this is substantially fainter , by about an order of magnitude , than the afterglows of long grbs on a comparable timescale ( e.g. , @xcite ) .
it is also remarkably similar to the prediction of @xmath53 mag at @xmath54 hr by @xcite .
since the available limits are at least @xmath55 mag , we conclude that the bursts lacking optical detections are drawn from a population with fainter afterglows .
for the 5 bursts in _ sample 2 _ we find a median and standard deviation of @xmath56 mag , about 1.4 mag fainter than the bursts with coincident hosts .
the cumulative distributions for the three samples normalized to the fiducial time of @xmath57 hr after the burst are shown in figure [ fig : optag2 ] .
a kolmogorov - smirnov ( k - s ) test indicates that there is only an @xmath58 probability that _ sample 1 _ and _ sample 2 _ are drawn from the same underlying distribution of optical afterglow brightnesses .
similarly , the probability that _ sample 1 _ and _ sample 3 _ are drawn from the same underlying distribution is @xmath59 ( an upper limit since the bursts in _ sample 3 _ are not detected in the optical ) .
on the other hand , the probability that _ sample 2 _ and _ sample 3 _ are drawn from the same underlying distribution is high , @xmath60 .
the overall faintness of the optical afterglows in _ sample 2 _ and _ sample 3 _ can be explained in two primary ways .
first , they could result from systematically lower circumburst densities . in the standard afterglow model with @xmath61 , the afterglow flux scales as @xmath62 for a uniform medium @xcite .
thus , a difference of about @xmath63 mag can be explained with a density that is about an order of magnitude lower than for the bursts in _
sample 1_. alternatively , the fainter fluxes may be due to higher redshifts for _ sample 2 _ and _ sample 3 _ compared to _
sample 1 _ since the optical flux also scales as @xmath64 , where @xmath65 is the luminosity distance .
the @xmath63 mag difference corresponds to @xmath66 ( @xmath67 ) for a _ sample 1 _ mean redshift of @xmath68 ( @xmath69 ) .
similarly , the flux also depends on the total energy , with @xmath70 , and therefore a difference of @xmath63 mag can be explained with an energy release lower by about a factor of 3 . the various scenarios ( lower density , lower energy , or higher redshift )
can be further explored through a comparison of the prompt @xmath0-ray emission , and the relation between the optical and x - ray afterglow brightness . in the framework of the standard grb model we do not expect lower densities to impact the prompt emission since it is expected to be produced by internal processes ( shocks or magnetic dissipation ) that do not depend on the external medium . on the other hand , lower energies or higher
redshifts will tend to systematically affect the prompt and afterglow emission . in figure
[ fig : fg_fx_t90 ] we plot the distributions of @xmath0-ray fluence ( @xmath71 ) , afterglow x - ray flux at the fiducial time of 8 hr ( @xmath72 ) , and duration ( @xmath73 ) for the three samples .
these distributions allow us to explore the underlying reason for the difference in optical afterglow brightness between the three samples .
first , we find that the distribution of @xmath71 values for _ sample 1 _ has a mean value that is about a factor of 5 times larger than for _ sample 2 _ and _ sample 3_. this is indicative of higher redshifts for the latter two samples if the isotropic - equivalent energies of all short grbs are similar , or alternatively a lower energy scale if the redshifts are similar . in the same vein , we find that the distributions of @xmath72 values for _ sample 2 _ and _ sample 3 _ have lower means than for _ sample 1_. this result is indicative of overall fainter afterglow emission for the former two samples , and this can again be explained in the context of lower energies or higher redshifts . unlike in the case of the @xmath0-ray emission ,
a lower circumburst density would also account for the fainter x - ray fluxes if @xmath74 .
finally , we find that the durations of the bursts in _ sample 2 _ and _ sample 3 _ are shorter by about a factor of 2 compared to the events with optical afterglows and coincident hosts , although there is substantial scatter in all three samples .
the shorter durations are not trivially explained in the context of lower energies , higher redshifts , or lower densities . to explore this result in more detail
we plot the observed fluence as a function of duration for the events in all three samples ( figure [ fig : t90_fg ] ) .
we find that there is a mild positive correlation between the two quantities , but that the events in _ sample 1 _ appear to have larger fluences at a given duration compared to the events in _ sample 2 _ and _ sample 3_. this is indicative of lower @xmath0-ray fluxes for the latter two samples , possibly as a result of higher redshifts .
higher redshifts will also shift the intrinsic durations of the bursts in _ sample 2 _ and _ sample 3 _ into better agreement with the bursts in _
sample 1_. we therefore conclude that the differences in prompt emission and optical / x - ray afterglow properties are consistent with a higher redshift origin for the bursts in _ sample 2 _ and _ sample 3_. the fainter afterglow emission is also consistent with lower density environments for these two samples , although this does not clearly explain the differences in prompt emission ( at least in the framework of the standard grb model ) .
we return to the discussion of low density versus a high redshift origin in [ sec : hosts ] when we investigate the host galaxy properties .
in addition to the overall faintness of the optical and x - ray afterglows , a substantial difference in density may also be imprinted on the _ ratio _ of optical to x - ray brightness .
this is because the synchrotron cooling frequency depends on density as @xmath75 , and is therefore expected to transition across the x - ray band as the density decreases . for @xmath76
the x - ray and optical bands occupy the same portion of the synchrotron spectrum , with a resulting spectral index of @xmath77 to @xmath78 while for @xmath79 ( i.e. , high density ) , the spectrum between the two bands will be steeper , reaching a maximum value of @xmath80 when @xmath81 . in figure [ fig : opt_xray ] we plot the x - ray flux versus optical magnitude for all three samples .
for each burst the fluxes in the optical and x - rays are taken at similar times after the burst , or extrapolated to a common time .
the correction factors due these extrapolations are generally @xmath82 ( table [ tab : shbs ] ) . for the combined _ sample 1 _ and _ sample 2 _ we find a clear correlation between the fluxes in the two bands , leading to a mean optical to x - ray spectral index of @xmath83 .
this is essentially indistinguishable from the ratio for long grbs , @xmath84 @xcite .
the median values are consistent with @xmath85 , but exhibit dispersion that is likely due to scatter in the values of @xmath86 and/or the location of @xmath87 relative to the x - ray band .
the similarity of @xmath88 for long and short grbs does not necessarily indicate that the densities are similar for the two samples .
in particular , if @xmath87 is located close to the x - ray band for long grbs , while for short grbs @xmath89 ( due to a lower density ) , the effect on @xmath88 will be marginal , particularly within the overall observed scatter .
for example , with @xmath90 the difference in @xmath88 between a model with @xmath87 exactly intermediate between the optical and x - ray bands , and a model with @xmath91 , is @xmath92 .
on the other hand , the scatter resulting from a range of @xmath93 is of the same order , @xmath94 .
similarly , the nearly equivalent median @xmath88 values may indicate that for both grb populations @xmath91 . in this case , the resulting lower limits on @xmath87 therefore prevent the use of @xmath88 as an indicator of density . comparing _ sample 1
_ and _ sample 2 _ , we find no clear difference in @xmath88 ( figure [ fig : opt_xray ] ) .
the same is true for the bursts in _ sample 3 _ , which are all consistent with the same relation given the optical upper limits and a mix of x - ray detections and upper limits .
thus , the ratio of optical to x - ray flux does not allow us to distinguish redshift / energy and density effects between the three samples . to summarize , the optical afterglows of short grb without coincident hosts ( or with only optical limits ) are systematically fainter than those of short grbs with coincident hosts .
the same is true for their x - ray fluxes and @xmath0-ray fluences .
the fainter afterglows may reflect lower densities ( by an order of magnitude ) , but this does not naturally explain the lower @xmath0-ray fluences . alternatively , the fainter afterglows and @xmath0-ray fluences can be explained as a result of higher redshifts ( @xmath1 ) or lower energies ( by about a factor of 3 ) .
we next turn to an analysis of the large - scale environments of the bursts in _ sample 2 _ , partly in comparison to the hosts of bursts in _
sample 1_. as indicated in
[ sec : obs ] , we place limits of @xmath95 mag on the brightness of any galaxy underlying the five short grb positions ( table [ tab : nohost ] ) .
these limits are several magnitudes fainter than the measured brightnesses of short grb hosts at @xmath96 ( _ sample 1 _ ) . to assess the potential that galaxies near each of the 5 bursts are the hosts , we calculate their probability of chance coincidence .
we follow the methodology of @xcite , namely , we determine the expected number density of galaxies brighter than a measured magnitude , @xmath97 , using the results of deep optical galaxy surveys @xcite : @xmath98 the probability for a given separation , @xmath99 , is then given by @xmath100 where we use the fact that for offsets substantially larger than the galaxy size , @xmath101 is the appropriate radius in equation [ eqn : prob ] @xcite . the resulting distributions for each field are shown in figure [ fig : prob1 ] .
we include all galaxies that have probabilities of @xmath102 .
we find that for 4 of the 5 bursts , faint galaxies ( @xmath103 mag ) can be identified within @xmath104 of the afterglow positions , with associated chance coincidence probabilities of @xmath105 ; in the case of grb090515 we do not detect any such faint galaxies within @xmath106 of the afterglow position .
for grb080503 we also include the galaxy at an offset of @xmath107 and @xmath108 mag identified by @xcite based on their deeper stack of hst observations . on the other hand , for 4 of the 5 bursts we find that the galaxies with the lowest probability of chance coincidence , @xmath109 , are brighter objects with offsets of about @xmath110 from the burst positions ; only in the case of grb080503 the lowest chance coincidence is associated with the nearest galaxy ( see @xcite ) . for comparison
we repeat the same analysis for short grbs from _ sample 1 _ which have faint coincident hosts ( grb060121 : 26.0 ab mag ; grb060313 : 26.6 ab mag ; and grb070707 : 27.3 ab mag ) .
the results of the probability analysis are shown in figure [ fig : prob2 ] .
we find that in all three cases , the coincident hosts exhibit the lowest probability of chance coincidence , @xmath111 . only in the case of grb070707
do we find galaxies with @xmath112 arcsec that have @xmath113 .
thus , these three bursts are consistent with negligible offsets from faint galaxies , presumably at @xmath7 .
the use of _ a posteriori _ probabilities to assign _
unique _ galaxy associations is fraught with difficulties .
first , for a given apparent brightness , galaxies located further away from the grb position , potentially due to larger kicks and/or longer merger timescales in the ns - ns merger framework , have higher probabilities of chance coincidence .
since we have no _ a priori _ model - independent knowledge of the range of possible kicks and merger timescales , we can not rule out galaxies at very large offsets for which @xmath114 . indeed , a reasonable constraint of @xmath115 km s@xmath116 and @xmath117 gyr leads to only a weak constraint on the offset of @xmath18 mpc . at @xmath118 ( @xmath69 )
this corresponds to about @xmath119 ( @xmath120 ) , a projected distance at which nearly all galaxies will have a chance coincidence probability of order unity .
a second difficulty is that we are using angular offsets , which ignore the potential wide range of redshifts ( and by extension also luminosities ) of the various galaxies .
for example , if the faint galaxies with small offsets are located at @xmath7 , the corresponding physical offsets are @xmath8 kpc , while if the galaxies at @xmath121 offsets are located at @xmath15 ( see [ sec : obs ] ) , the offsets are only somewhat larger , @xmath122 kpc .
a galaxy at an even lower redshift , @xmath123 , with an offset of 50 kpc will be located about @xmath124 from the grb position and incur a large penalty in terms of chance coincidence probability .
it is important to note , however , that galaxies at lower redshift will generally have brighter apparent magnitudes , partially compensating for the larger angular separations ( equations [ eqn : gal ] and [ eqn : prob ] ) . in only a single case ( grb070809 )
we find a galaxy with @xmath113 at @xmath125 ( which at @xmath126 for this galaxy corresponds to a physical offset of about 100 kpc ) .
a final complication , which is not unique to this subset of events , is that we can only measure projected offsets , @xmath127 .
the measured offsets can be used as lower limits on the actual offsets , while for the overall distribution we can apply an average correction factor of @xmath128 , based on the expectation value for the projection factor , @xmath129 . despite these caveats
we can address the probability that _ all _ of the associations are spurious .
this joint probability is simply the product of the individual probabilities @xcite .
for the faint galaxies at small angular separations the probability that all are spurious associations is @xmath130 , while for the galaxies with the lowest probability of chance coincidence the joint probability is nearly 30 times lower , @xmath131 .
conversely , the probabilities that _ none _ of the associations are spurious are @xmath132 and @xmath133 , respectively . these values indicate the some spurious coincidences may be present for _ sample 2_. indeed , the probabilities that 1 , 2 , or 3 associations are spurious are @xmath134 $ ] and @xmath135 $ ] , respectively . these results indicate that for the faint galaxies it is not unlikely that @xmath136 associations ( out of 5 ) are spurious , while for the brighter galaxies @xmath137 associations may be spurious .
this analysis clearly demonstrates why a joint statistical study is superior to case - by - case attempts to associate short grbs with galaxies at substantial offsets .
we therefore conclude that despite the weaknesses inherent to _ a posteriori _ probabilities , we conclude that there is stronger statistical support for an association of at least some of the 5 bursts in _ sample 2 _ with bright galaxies at separations of @xmath138 , than for an association with the faint galaxies at separations of @xmath139 .
clearly , we can not rule out the possibility that in reality the hosts are a mix of faint and bright galaxies with a range of angular offsets of @xmath140 .
we note that if deeper observations eventually lead to the detection of underlying galaxies ( @xmath141 ) at the level of @xmath142 mag , the associated chance coincidence probabilities will be @xmath143 per object , and the joint probabilities will be only slightly higher than for the bright galaxies with @xmath121 offsets . on the other hand , if we can achieve magnitude limits of @xmath19 mag on any coincident hosts , the resulting probabilities of chance coincidence will be larger than for the offset bright galaxies .
thus , eliminating the possibility of underlying hosts at the level of @xmath19 mag is of the utmost importance .
so far , only grb080503 has been observed to such a depth with no detected coincident host @xcite , but observations of the full sample are required for a robust statistical comparison . as we discuss in [ sec : highz ] below , such deep limits will also reduce the probability of underlying hosts based on redshift arguments . beyond the use of projected angular offsets , we note that the faint galaxies near the burst positions are likely to have projected physical offsets of about @xmath144 kpc . to assess the projected physical offsets for the galaxies with the lowest probability of chance coincidence we measured spectroscopic redshifts in three cases ( grbs 061201 , 070809 , and 090515 ; [ sec : obs ] ) . in the case of grb061201
, the galaxy redshift of @xmath36 leads to a projected physical offset of @xmath145 kpc @xcite . in the case of grb070809 , the lowest probability of chance coincidence
is not associated with the star forming galaxy at @xmath39 identified by @xcite , but instead belongs to the early - type galaxy identified here , which has a redshift of @xmath40 , and hence an offset of @xmath146 kpc . finally , for grb090515 ,
the lowest probability of chance coincidence is associated with a cluster early - type galaxy at @xmath44 , leading to a physical offset of @xmath147 kpc .
thus , two scenarios emerge from our investigation of the large - scale environments : ( i ) the bursts are spatially coincident with currently - undetected galaxies ( with @xmath148 mag ) ; or ( ii ) the bursts have substantial offsets of at least @xmath149 kpc depending on whether they are associated with faint galaxies at small angular separations , or brighter galaxies at @xmath30 ; for the ensemble of 5 events the larger offsets are statistically more likely than the @xmath8 kpc offsets . in the context of large offsets , even larger values may be possible if the bursts originated in galaxies with larger separations and @xmath150 .
these two scenarios echo the possibilities that emerged from the analysis of the afterglow and prompt emission properties ( [ sec : ag ] ) .
distinguishing between these possibilities is clearly of fundamental importance to our understanding of short grbs : the former scenario will point to a population of very faint hosts ( likely at high redshifts ) , while the latter scenario will provide evidence for large offsets ( due to kicks or a globular cluster origin ) and hence ns - ns / ns - bh progenitors for at least some short grbs .
we can place upper limits on the redshifts of the grbs in _ sample 2 _ based on their detections in the optical ( i.e. , the lack of complete suppression by the ly@xmath151 forest ) .
the afterglow of grb061201 was detected in the ultraviolet by the _
swift_/uvot and it is therefore located at @xmath152 @xcite .
the remaining four bursts were detected in the optical @xmath153- or @xmath33-band , and can therefore be placed at @xmath154 ( table [ tab : shbs ] ) .
we place additional constraints on the redshifts of any underlying hosts using the existing sample short grb host galaxies . in figure
[ fig : mags ] we plot the @xmath33-band magnitudes as a function of redshift for the all available short grb hosts from _
sample 1_. for the faint hosts without known redshifts ( grbs 060121 , 060313 , and 070707 ) , we place upper limits on the redshift using optical detections of the afterglows ( table [ tab : shbs ] ) .
a wide range of host magnitudes , @xmath155 mag , is apparent .
we also plot the @xmath33-band magnitudes of long grb hosts @xcite , as well as the @xmath156 phase space that is traced by galaxies with luminosities of @xmath157 l@xmath11 .
we use the appropriate value of l@xmath11 as a function of redshift , taking into account the evolving galaxy luminosity function @xcite .
we find excellent correspondence between the hosts of long and short grbs , and the phase - space traced by @xmath158 l@xmath11 galaxies , at least to @xmath159 .
in the context of these distributions , the available limits for the short grbs in _ sample 2 _ translate to redshifts of @xmath160 if they are 0.1 l@xmath11 galaxies , or @xmath161 if they are l@xmath11 galaxies .
the latter lower limits are comparable to the redshift upper limits inferred from the afterglow detections .
we note that for grb080503 , the limit of @xmath162 mag @xcite places even more stringent limits on the redshift of an underlying @xmath158 l@xmath11 galaxy .
the redshifts of @xmath163 for putative 0.1 l@xmath11 hosts are consistent with the faintness of the optical afterglows , from which we inferred @xmath1 compared to _
sample 1 _ ( [ sec : ag ] ) .
we note , however , that the one known short grb at @xmath12 ( grb090426 ; @xcite ) has a host galaxy luminosity of @xmath164 l@xmath11 , which may suggest that the appropriate redshift lower limits are @xmath161 .
the possibility that the five bursts originated at @xmath161 leads to a bimodal redshift distribution ( figure [ fig : mags ] ) .
nearly all of the bursts in _ sample 1 _ with a known redshift ( 9/10 ) have @xmath25 , with a median of @xmath165 ; the sole exception is grb090426 at @xmath166 .
the three bursts with faint coincident hosts have upper limits of @xmath154 from afterglow detections , while lower limit of @xmath167 can be placed on these hosts if they have @xmath168 l@xmath11 . adding the _ sample 2 _ bursts with the assumption that they have @xmath161
will furthermore result in a population of short grbs with a median of @xmath169 , and leave a substantial gap at @xmath170 ( figure [ fig : mags ] ) .
if the 5 bursts are instead hosted by @xmath171 l@xmath11 galaxies , the inferred lower limits on the redshifts ( @xmath160 ) lead to a potentially more uniform redshift distribution .
it is difficult to explain a bimodal redshift distribution with a single progenitor population such as ns - ns binaries , without appealing to , for example , a bimodal distribution of merger timescales .
another possibility is two distinct progenitor populations , producing bursts of similar observed properties but with distinct redshift ranges .
while these possibilities are difficult to exclude , they do not provide a natural explanation for the short grb population .
a final alternative explanation is that any underlying hosts reside at similar redshifts to the known hosts in _
sample 1 _ ( @xmath16 ) , but have significantly lower luminosities of @xmath172 l@xmath11 .
this scenario would not naturally explain why the bursts in _ sample 2 _ have fainter optical and x - ray afterglows , as well as lower @xmath0-ray fluences .
we therefore do not consider this possibility to be the likely explanation .
while higher redshifts may explain the lack of detected hosts , the fainter afterglows , and the weaker @xmath0-ray fluences of the bursts in _ sample 2 _ , this scenario suffers from several difficulties outlined above .
the alternative explanation is that the bursts occurred at significant offsets relative to their hosts , and hence in lower density environments that would explain the faint afterglow emission ( though possibly not the lower @xmath0-ray fluences ) .
as we demonstrated in [ sec : prob ] , the offsets may be @xmath139 ( @xmath8 kpc ) if the bursts originated in the faint galaxies at the smallest angular separations , or @xmath121 ( @xmath173 kpc ) if they originated in the brighter galaxies with the lowest probability of chance coincidence . below we address the implications of these two possible offset groups through a comparison to the offsets measured for the bursts in _ sample 1 _ ( e.g. , @xcite ) .
we plot the distributions of projected angular offsets for the short grbs with and without coincident hosts in figure [ fig : offset1 ] .
the offsets for _ sample 1 _ have a mean and standard deviation of about @xmath174 and a range of about @xmath175 .
modeled with a log - normal distribution , the resulting mean and width in units of arcseconds are @xmath176 and @xmath177 . if we associate the short bursts in _ sample 2 _ with the galaxies that have the lowest probability of chance coincidence , the resulting distribution has @xmath178 arcsec .
this is clearly distinct from the distribution for _ sample 1_. the effect is less pronounced if we associate the bursts with the galaxies located at the smallest angular offsets ( figure [ fig : offset1 ] ) . even for these separations
the mean and standard deviation are @xmath179 arcsec . as noted in [ sec : prob ] , the projected angular distances may not be the most robust quantity for measuring the offset distribution .
an alternative quantity is the offset normalized by each host s effective radius @xcite .
this quantity takes into account the varying sizes of the hosts due to both intrinsic size variations and redshift effects .
it also gives a better indication of whether the burst coincides with the host light or is significantly offset .
as shown in figure [ fig : offset2 ] , the host - normalized offsets of _ sample 1 _ have a mean and standard deviation of about @xmath180 @xmath181 , and a range of about @xmath182 @xmath181 . a log - normal fit results in a mean of @xmath183 and a width of @xmath184 . the bursts in _ sample 2
_ have much larger host - normalized offsets , with @xmath185 if they originated in the galaxies with the lowest chance coincidence probability . even if we associate the bursts with the nearest faint hosts , the distribution has a mean of about @xmath186 @xmath181 , reflecting the fact that the effective radii of the faint galaxies are smaller than those of the brighter ones
. finally , we plot the projected physical offsets in figure [ fig : offset3 ] .
the mean and standard deviation for _ sample 1 _ are @xmath187 kpc , and a log - normal fit results in a mean of @xmath188 and a width of @xmath189 .
on the other hand , the bursts in _ sample 2 _ have a mean offset of about 19 kpc if they arise in the faint galaxies with small angular separation , or about 40 kpc if they arise in the brighter galaxies , again pointing to distinct distributions .
the distributions of angular , physical , and host - normalized offsets exhibit a clear bimodality if we associate the bursts in _ sample 2 _ with the galaxies at @xmath30 .
this is particularly apparent in the more meaningful quantities , namely physical and host - normalized offsets ( figures [ fig : offset2 ] and [ fig : offset3 ] ) .
the effect is still apparent , though less pronounced in the case of association with the faint galaxies at @xmath7 .
thus , if the offset scenario is correct , the resulting distributions point to a possible bimodality rather than a single continuous distribution of offsets .
the cumulative distributions of physical offsets for _ sample 1 _ alone , and in conjunction with the two possible offset groups for _ sample 2 _ are shown in figure [ fig : offset4 ] .
the combined distributions have a median of about 4 kpc , driven by the bursts with coincident hosts .
however , there is a clear extension to larger physical offsets in the case of association with the brighter galaxies , with about @xmath190 of all objects having @xmath191 kpc .
the cumulative distributions are particularly useful for comparison with ns - ns merger models since predictions exist for both the kick scenario and the globular cluster origin model .
we turn to this discussion below .
we have shown that the lack of host galaxy detections in coincidence with the bursts in _ sample 2 _ indicates either a high redshift ( @xmath12 ) origin , or substantial projected offsets of @xmath192 kpc ( @xmath193 @xmath181 ) .
both scenarios can account for the lack of underlying galaxy detections and the fainter optical / x - ray afterglows ( either due to redshift or density effect ) ; the possibility of large offsets from galaxies at @xmath30 , however , does not naturally explain the lower @xmath0-ray fluences and somewhat shorter durations of the bursts in _
sample 2_. indeed , the lower fluences can be more easily accommodated in the case of undetected high redshift coincident hosts , or in the case of faint hosts with offsets of @xmath8 kpc .
below we test the results of the offset scenario against predictions for ns - ns merger models and dynamically - formed ns - ns binaries in globular clusters .
we also investigate whether the densities expected at offsets of @xmath194 kpc ( i.e. , galaxy halos or the intergalactic medium ) can accommodate the observed optical magnitudes .
finally , we discuss the implications of the various scenarios for short grb energetics .
the sample of short grbs with optical afterglows represents about @xmath195 of all short bursts , and may thus not be fully representative .
one often - discussed bias is that the bursts with optical afterglows require a high circumburst density , and therefore have negligible offsets .
however , from our analysis in this paper it is clear that one explanation for the lack of coincident hosts for the bursts in _ sample 2 _ is indeed large offsets , _ despite their detection in the optical band_. the bursts with only x - ray afterglow detections ( _ swift_/xrt ) have typical positional uncertainties of @xmath196 and therefore lead to a deeper ambiguity about the identity of the hosts .
we have previously argued that most of these bursts are associated with galaxies consistent with a negligible offset ( or as large as tens of kpc ; @xcite ) . in some cases
larger offsets have been advocated based on _ a posteriori _ chance coincidence probabilities ( e.g. , grb050509b : @xcite ; grb060502b : @xcite ) . however , from the analysis in [ sec : prob ] it is clear that while the inferred probabilities are only weakly dependent on the positional accuracy when @xmath197 ( i.e. , galaxies at large offsets ) , a large penalty is incurred for faint galaxies located _ within _ the x - ray error circles since in that case the appropriate value in equation [ eqn : prob ] is @xmath198 . as a result
, the faint galaxies will have chance coincidence probabilities of @xmath150 . to avoid this complication in our comparison to ns - ns model predictions
we restrict the analysis to the sample with optical afterglow positions . in recent work by @xcite
we have shown that the offset distribution for short grbs with optical and/or x - ray positions , and accounting for incompleteness due to bursts with only @xmath0-ray positions , is broadly consistent with the predictions of ns - ns merger models @xcite .
in particular , we concluded that @xmath199 of all short bursts have @xmath200 kpc , while @xmath201 have offsets of @xmath21 kpc .
we repeat the analysis here using the offsets inferred in
[ sec : kick ] for both the faint galaxies at small angular separations and the brighter galaxies at larger separations .
as shown in figure [ fig : offset4 ] , the model predictions have a median of about 6 kpc , compared to about 4 kpc for the observed sample .
on the other hand , the models predict @xmath20 of offsets to be @xmath202 kpc , in good agreement with the observed distribution in both the @xmath8 kpc and @xmath203 kpc scenarios .
we note that the overall smaller offsets measured from the data may be due to projections effects .
indeed , the mean correction factor of @xmath128 nicely reconciles the theoretical and observed distributions . in the previous section we noted a bimodality in the physical and host - normalized offsets for
_ sample 1 _ and _ sample 2 _ ( figures [ fig : offset2 ] and [ fig : offset3 ] ) . in the framework of ns - ns binary kicks
this bimodality may indicate that the binaries generally remain bound to their host galaxies , thereby spending most of their time at the maximal distance defined by @xmath204 ( i.e. , with their kinetic energy stored as potential energy ; @xcite ) .
this would require typical kick velocities of less than a few hundred km s@xmath116 .
we further compare the observed offset distribution to predictions for dynamically - formed ns - ns binaries in globular clusters , with a range of host galaxy virial masses of @xmath205 m@xmath206 @xcite .
these models predict a range of only @xmath207 of all ns - ns mergers to occur within 10 kpc of the host center , in contrast to the observed distribution with about @xmath208 with @xmath209 kpc .
we stress that this result is independent of what offsets we assign to the bursts in _ sample 2 _ since they account for only 1/4 of the bursts with optical afterglows . on the other hand ,
the globular cluster origin may account for the bimodality in the physical and host - normalized offsets ( figures [ fig : offset2 ] and [ fig : offset3 ] ) , with the objects in _ sample 2 _ arising in globular clusters and the objects with coincident hosts arising from primordial ns - ns binaries .
this possibility also agrees with the predicted fraction of dynamically - formed ns - ns binaries of @xmath210 @xcite .
the cumulative offset distributions for _ sample 2 _ alone ( assuming the hosts are the galaxies with the lowest probability of chance coincidence ) is well - matched by the range of predictions for dynamically - formed ns - ns binaries in globular clusters ( figure [ fig : offset4 ] ) . in this scenario , however , the implication is that short grbs outside of globular clusters do not experience kicks as expected for ns - ns binaries since the largest measured offset is only 15 kpc . unless the populations of short grbs with only x - ray or @xmath0-ray positions have fundamentally different offset distributions , we conclude that the measured offsets of short grbs and the predicted offsets for ns - ns kicks are in good agreement , _ if we treat all short grbs with optical afterglows as a single population_. alternatively , it is possible that the bimodal distributions of physical and host - normalized offsets point to a progenitor bimodality , with the bursts in _ sample 2 _ originating in globular clusters .
large offsets are expected to result in low circumburst densities , and we now investigate whether the observed optical magnitudes for _ sample 2 _ can be used to place meaningful constraints on the offsets .
the optical afterglow magnitudes do not provide a direct measure of the circumburst density since they depend on a complex combination of density , energy , and fractions of the burst energy imparted to the radiating electrons ( @xmath211 ) and magnetic fields ( @xmath212 ) . from our comparison of the x - ray and optical afterglows ( [ sec : ag ] ) we were unable to clearly locate the cooling frequency , which in principle can provide additional constraints on the density .
however , we can still gain some insight into the required circumburst densities using a few basic assumptions .
we assume that the @xmath0-ray energy ( @xmath213 ) is a reasonable proxy for the total energy , that @xmath214 , and that @xmath215 .
we further use a fiducial redshift of @xmath68 ( figure [ fig : mags ] ) and a fiducial observation time of 8 hr ( [ sec : ag ] ) . to determine @xmath213 we use the @xmath216 kev fluences measured by _
( table [ tab : shbs ] ) , and determine a correction factor to account for the incomplete energy coverage using bursts that have also been detected by satellites with a broader energy range ( figure [ fig : t90_fg ] ) .
we find a typical correction factor of @xmath217 for the @xmath218 kev range .
for the fiducial redshift of @xmath68 , the mean energy for short grbs with optical afterglows is @xmath219 erg , when we include the correction factor ( figure [ fig : z_egiso ] ) . with these assumptions
we find the following relation between the optical brightness and the circumburst density @xcite : @xmath220 given the typical observed fluxes of about 2 @xmath221jy for _ sample 1 _ , and about 0.6 @xmath221jy for _ sample 2 _ , we infer typical minimum densities of @xmath222 and @xmath223 @xmath32 , respectively .
thus , the observed optical brightnesses can be produced even at very low densities that are typical of the igm .
this indicates that we can not rule out the large offset scenario based on density arguments .
indeed , even if we allow for more typical values of @xmath224 , the resulting densities are @xmath225 and @xmath226 @xmath32 , respectively .
these results compare favorably with predictions for ns - ns kicks , which suggest that most mergers will occur at densities of @xmath227 @xmath32 @xcite .
the distribution of densities in globular clusters is not well known , preventing a meaningful comparison to our inferred minimum densities @xcite .
finally , we investigate the energetics of the bursts in _ sample 2 _ in the context of a high - redshift origin ( @xmath161 ) , an origin in the faint galaxies at @xmath228 separations ( @xmath229 ) , and an origin in the galaxies with the lowest probabilities of chance coincidence ( @xmath230 ) .
the resulting values of @xmath213 are shown in figure [ fig : z_egiso ] .
for the lowest redshift origin , the inferred values are @xmath231 erg , for a @xmath229 origin they are @xmath232 erg , and for a @xmath161 origin they are @xmath233 erg . for comparison , the mean value for the bursts in _ sample 1 _ is @xmath234 erg , with a range of @xmath235 erg .
however , there is a clear redshift dependence for the measured @xmath213 values ( figure [ fig : z_egiso ] ) , with @xmath236 erg at @xmath237 and @xmath238 erg at @xmath239 .
thus , the bursts in _ sample 2 _ generally fit within the known range of isotropic energies regardless of their actual redshift .
indeed , at @xmath169 they exhibit similar values of @xmath240 to that of grb090426 . as a result , we can not use the inferred energy release as a clear discriminant of the redshift range for _ sample 2_. if the bursts in _ sample 2 _ indeed originated at high redshifts , the resulting isotropic - equivalent energies suggest that there is either a large spread in the energy release of short grbs ( at least 2 orders of magnitude ) or a large variations in the ejecta geometry . if the typical energy for short grbs is about @xmath241 erg , as indicated by the nearest events , then the inferred values of @xmath242 at @xmath169 indicate opening angles of @xmath243 .
this is similar to the opening angle of about @xmath244 that was inferred for grb051221 @xcite .
we undertook the first systematic study of short grbs with detected optical afterglows ( and hence sub - arcsecond position ) but no coincident host galaxies to limits of @xmath148 mag .
we find that the optical afterglows of these bursts are fainter by about 1.4 mag compared to the optical afterglows of short grbs with coincident hosts .
they are similarly fainter in x - rays , and have somewhat lower @xmath0-ray fluences and slightly shorter durations .
both samples have similar ratios of x - ray to optical flux , which are moreover similar to the ratios measured for long grbs .
the fainter afterglows of the bursts lacking coincident hosts may be due to lower densities , lower energies , or higher redshifts .
however , we note that only the scenarios with lower energies or higher redshifts naturally explain the faintness of the prompt emission _ in the context of the standard grb model_. this is because in the context of internal processes ( shocks or magnetic dissipation ) the external density should not play a role .
we further use deep optical imaging to determine the probability of chance coincidence for galaxies in the field around each burst , and to place redshift limits for underlying hosts under the assumption that they are drawn from the same distribution of the detected short grb hosts ( @xmath10 l@xmath11 ) .
this analysis leads to the following possible scenarios : ( i ) the underlying hosts are fainter than @xmath245 mag , indicative of redshifts of @xmath163 ( if @xmath246 l@xmath11 ) or @xmath161 ( if @xmath247l@xmath11 ) ; or ( ii ) the hosts are galaxies with substantial offsets either faint galaxies at separations of @xmath228 ( @xmath248 kpc for @xmath7 ) or brighter galaxies , which we find to be located at @xmath30 , with offsets of @xmath249 ( @xmath250 kpc ) . in the former scenario , unless the galaxies have @xmath246 l@xmath11 , the resulting redshift distribution is bimodal with peaks at @xmath16 and @xmath251 .
such a scenario most likely requires a bimodal age distribution in the context of ns - ns mergers , or two distinct progenitor systems dominating at low and high redshifts . while this can not be ruled out by present data , the lack of overlap at @xmath252 is difficult to explain .
we also conclude that it is unlikely that any faint underlying hosts are low - luminosity galaxies ( @xmath253 l@xmath11 ) at similar redshifts to the detected hosts , since this does not naturally explain the difference in afterglow and prompt emission properties
. in the context of the large offsets scenario , the probability of chance coincidence ( both individually and for the sample ) is lower for the brighter galaxies at offsets of @xmath249 than for the faint galaxies at offsets of @xmath228 .
however , it is not unlikely that the true associations are a mix of both populations , since in each case there is a non - negligible probability that @xmath137 of the associations are spurious .
we note that optical positions are crucial for a uniform comparison of the chance coincidence probabilities for faint galaxies with small offsets and bright galaxies with larger offsets ; in the case of only x - ray positions there is an inherent bias ( in the sense of _ a posteriori _ ) against associations with faint galaxies inside the error circle , despite the fact that they are consistent with no offset . from spectroscopic observations for 3 of the 5 bursts we find that the galaxies with the lowest probability of chance coincidence are a star forming galaxy at @xmath36 ( grb061201 ) , an early - type galaxy at @xmath40 ( grb070809 ) , and an early - type cluster member galaxy at @xmath44 ( grb090515 ) .
if these associations are indeed correct , they only slightly alter the host demographics , which are dominated by star forming galaxies @xcite .
however , this does suggest that our present understanding of the relative ratios of star forming and elliptical hosts may be incomplete .
the resulting distributions of angular , physical , and host - normalized offsets for the bursts with and without coincident hosts appear to be distinct , rather than continuous .
however , the joint distribution of projected physical offsets is in good agreement with theoretical predictions for ns - ns binary mergers . on the other hand
, the predicted distribution for dynamically - formed ns - ns binaries in globular clusters provides a much poorer fit to the entire data set , unless they account for only the bursts with large offsets ( _ sample 2 _ ) . in the case of a hybrid population of primordial and dynamically - formed binaries , with the latter accounting for only the large offsets ,
the fact that all the remaining offsets ( _ sample 1 _ ) are @xmath18 kpc , is indicative of no significant kicks .
the large physical offsets also naturally explain the fainter afterglow emission as a result of lower circumburst densities .
the resulting isotropic @xmath0-ray energies match the observed distribution for short grbs with coincident hosts , either at @xmath15 or at @xmath7 .
our conclusion that large offsets of @xmath254 kpc ( corresponding to @xmath193 galactic effective radii ) are a likely explanation for the bursts with optical afterglows and no coincident hosts is of fundamental importance .
the only progenitor model that naturally explains this result in the merger of ns - ns / ns - bh binaries , most likely due to kicks , or possibly with a minor contribution from a globular cluster population ( accounting specifically for the events in _ sample 2 _ ) .
while a conclusive demonstration of a large offset requires an absorption redshift measurement that matches an offset galaxy redshift , the distribution of optical afterglow magnitudes indicates that this will be difficult to achieve .
indeed , an absorption redshift is available for only one likely short grb @xcite .
we end with several important observations .
first , rapid optical follow - up of short grbs with 8-meter class telescopes is essential since observations to a depth of about 25 mag within a few hours after the burst may recover nearly all optical afterglows , regardless of circumburst density ( figure [ fig : optag1 ] ) .
second , short grbs with only x - ray positions are unlikely to provide strong support for either negligible or large offsets due to the appreciable size of the error circles and the fact that the sample with optical afterglows exhibits wide host galaxy diversity , i.e. , events with coincident bright hosts at @xmath255 , events with coincident faint hosts likely at @xmath7 , and events with no coincident hosts likely due to offsets .
third , a meaningful study of short grb offsets requires a statistical approach to mitigate the shortcomings of _ a posteriori _ chance coincidence probabilities , as well as to incorporate the relevant information from afterglow and prompt emission observations .
finally , we note that from a wide range of observations of both the afterglows and host galaxies it appears that the case for ns - ns mergers as the progenitors of short grbs is gaining observational support .
our main result here is that short grbs with optical afterglows and no detected host galaxies are somewhat less likely to be explained with high redshifts or with dwarf galaxy hosts at low redshift .
instead , these bursts likely exploded @xmath8 kpc from galaxies at @xmath229 or tens of kpc from galaxies at @xmath15 . with larger samples
, these possibilities will allow us to improve our understanding of the short grb redshift distribution , the host galaxy demographics , and predictions for gravitational wave detections .
if high redshifts indeed turn out to be prevalent , this will have a significant effect on the possibility of multiple progenitor populations .
we acknowledge helpful discussions with alicia soderberg , ryan chornock , and josh grindlay .
this work was partially supported by swift ao5 grant number 5080010 , and made use of data supplied by the uk swift science data centre at the university of leicester .
lcccccccl 0.08 in + 050709 & 0.07 & 0.161 & @xmath256 & 60.5 & @xmath257 & 34 & 2.3 & 13 + 050724 & 3.0 & 0.257 & @xmath258 & 11.8 & @xmath259 & 12 & 8.4 & 46 + 051221a & 1.4 & 0.546 & @xmath260 & 3.1 & @xmath261 & 3.1 & 5.8 & 78 + 060121 & 2.0 & @xmath262 & @xmath263 & 6.1 & @xmath261 & 7.4 & 8.8 & 911 + 060313 & 0.70 & @xmath264 & @xmath265 & 2.8 & @xmath266 & 2.8 & 10.8 & 1112 + 061006 & 0.42 & 0.438 & @xmath267 & 19.5 & @xmath268 & 14.9 & 2.9 & 11,13 + 061201@xmath269 & 0.80 & @xmath270,0.111 ?
& @xmath271 & 8.6 & @xmath272 & 8.6 & 2.9 & 1415 + 070707 & 1.1 & @xmath262 & @xmath267 & 12.5 & @xmath273 & 11 & 1.9 & 16 + 070714b & 3.0 & 0.923 & @xmath274 & 24.4 & @xmath275 & 23.6 & 0.7 & 17 + 070724 & 0.40 & 0.457 & @xmath276 & 2.35 & @xmath277 & 2.3 & 5.0 & 1819 + 070809@xmath269 & 1.3 & @xmath264,0.473 ? & @xmath278 & 11.0 & @xmath279 & 11 & 0.8 & 2021 + 071227 & 1.8 & 0.381 & @xmath280 & 7.0 & @xmath281 & 7.0 & 1.6 & 13 + 080503@xmath269 & 0.32 & @xmath264 , & @xmath282 & 66.0 & @xmath283 & 25.9 & 0.3 & 22 + 080905 & 1.0 & 0.122 & @xmath284 & 18.4 & @xmath285 & 8.5 & 0.8 & 23 + 090305@xmath269 & 0.40 & @xmath262 , & @xmath286 & 0.45 & @xmath287 & 0.45 & 1.5 & 2425 + 090426 & 1.28 & 2.609 & @xmath288 & 2.0 & @xmath289 & 2.6 & 20.0 & 2627 + 090510 & 0.30 & 0.903 & @xmath290 & 6.4 & @xmath291 & 9.0 & 2.3 & 28 + 090515@xmath269 & 0.04 & @xmath262,0.403 ? & @xmath292 & 4.9 & @xmath293 & 1.9 & 0.1 & 29 + 091109b & 0.30 & & @xmath294 & 3.8 & @xmath295 & 5.7 & 0.5 & 30 + 100117 & 0.30 & 0.92 & @xmath296 & 8.4 & @xmath297 & 8.4 & 0.3 & 31 + + + 050509b & 0.04 & 0.225 & @xmath298 & 9.1 & @xmath297 & 2.1 & 0.7 & 3233 + 050813 & 0.60 & & @xmath299 & 94.9 & @xmath300 & 12.8 & 1.9 & 3436 + 051210 & 1.27 & & @xmath301 & 6.9 & @xmath302 & 19.2 & 1.6 & 11,37 + 060502b & 0.09 & & @xmath303 & 18.6 & @xmath304 & 16.8 & 0.7 & 11,38 + 060801 & 0.50 & 1.130 & @xmath301 & 10.9 & @xmath305 & 12.4 & 0.8 & 11 + 061210 & 0.19 & 0.409 & @xmath265 & 72.4 & @xmath306 & 2.1 & 1.4 & 11 + 061217 & 0.21 & 0.827 & @xmath307 & 38.6 & @xmath308 & 2.8 & 2.0 & 11 + 070429b & 0.50 & 0.902 & @xmath309 & 11.8 & @xmath310 & 4.8 & 0.6 & 39 + 080426 & 1.30 & & @xmath311 & 6.1 & @xmath312 & 7.5 & 2.6 & 40 + 080702 & 0.50 & & @xmath313 & 2.9 & @xmath314 & 12.1 & 12 . & 41 + 081226 & 0.40 & & @xmath315 & 3.2 & @xmath302 & 1.6 & 0.9 & 42 + 100206 & 0.12 & & @xmath284 & 6.9 & @xmath297 & 15.7 & 0.5 & 43 lcccc 0.1 in 061201 & hst / acs & f814w & 2224 & 26.0 + 070809 & magellan / ldss3 & @xmath33 & 1500 & 25.4 + 080503 & hst / wfpc2 & f606w & 4000 & 25.7 + 090305 & magellan / ldss3 & @xmath33 & 2400 & 25.6 + 090515 & gemini - n / gmos & @xmath33 & 1800 & 26.5 -band images of the location of grb070809 .
galaxies near the position of the optical afterglow ( cross - hairs ) are marked .
[ fig:070809],title="fig:",width=288 ] -band images of the location of grb070809 .
galaxies near the position of the optical afterglow ( cross - hairs ) are marked .
[ fig:070809],title="fig:",width=288 ] was found by @xcite based on a deeper stack of hst / wfpc2 observations .
these authors also find that the galaxy marked `` s5 '' is located at @xmath316 , leading to a physical offset of 85 kpc .
[ fig:080503],width=480 ] -band images of the location of grb090305 .
galaxies near the position of the optical afterglow ( cross - hairs ) are marked .
[ fig:090305],title="fig:",width=288 ] -band images of the location of grb090305 .
galaxies near the position of the optical afterglow ( cross - hairs ) are marked .
[ fig:090305],title="fig:",width=288 ] -band images of the location of grb090515 .
galaxies near the position of the optical afterglow ( cross - hairs ) are marked .
note that the object coincident with the cross - hairs is the optical afterglow .
[ fig:090515],title="fig:",width=288 ] -band images of the location of grb090515 .
galaxies near the position of the optical afterglow ( cross - hairs ) are marked .
note that the object coincident with the cross - hairs is the optical afterglow .
[ fig:090515],title="fig:",width=288 ] -ray fluence ( top ) , afterglow x - ray flux at 8 hr ( middle ) , and duration ( bottom ) for the three short grb samples discussed in this paper .
the arrows mark the mean for each sample , indicating that the bursts in _ sample 2 _ and _ sample 3 _ have lower @xmath0-ray fluences , fainter x - ray fluxes , and shorter durations , than the bursts with detected afterglows and coincident hosts .
[ fig : fg_fx_t90],width=480 ] | we investigate the afterglow properties and large - scale environments of several short - duration gamma - ray bursts ( grbs ) with sub - arcsecond optical afterglow positions but no bright coincident host galaxies .
the purpose of this joint study is to robustly assess the possibility of significant offsets , a hallmark of the compact object binary merger model .
five such events exist in the current sample of 20 short bursts with optical afterglows , and we find that their optical , x - ray , and @xmath0-ray emission are systematically fainter
. these differences may be due to lower circumburst densities ( by about an order of magnitude ) , to higher redshifts ( by @xmath1 ) , or to lower energies ( by about a factor of 3 ) , although in the standard grb model the smaller @xmath0-ray fluences can not be explained by lower densities . to study the large - scale environments we use deep optical observations to place limits on underlying hosts and to determine probabilities of chance coincidence for galaxies near each burst . in 4 of the 5 cases the lowest probabilities of chance coincidence ( @xmath2 )
are associated with bright galaxies at separations of @xmath3 , while somewhat higher probabilities of chance coincidence are associated with faint galaxies at separations of @xmath4 . by measuring redshifts for the brighter galaxies in three cases ( @xmath5 )
we find physical offsets of @xmath6 kpc , while for the faint hosts the assumption of @xmath7 leads to offsets of @xmath8 kpc . alternatively
, the limits at the burst positions ( @xmath9 mag ) can be explained by typical short grb host galaxies ( @xmath10 l@xmath11 ) at @xmath12 .
thus , two possibilities exist : ( i ) @xmath13 of short grbs explode @xmath14 kpc or @xmath8 kpc from the centers of @xmath15 or @xmath7 galaxies , respectively , and have fainter afterglows due to the resulting lower densities ; or ( ii ) @xmath13 of short grbs occur at @xmath12 and have fainter afterglows due to their higher redshifts .
the high redshift scenario leads to a bimodal redshift distribution , with peaks at @xmath16 and @xmath17 .
the large offset scenario leads to an offset distribution that is well - matched by theoretical predictions of ns - ns / ns - bh binary kicks , or by a hybrid population with globular cluster ns - ns binaries at large offsets and primordial binaries at offsets of @xmath18 kpc ( indicative of negligible kicks ) .
deeper constraints on any coincident galaxies to @xmath19 mag ( using the _ hubble space telescope _ ) will allow us to better exclude the high - redshift scenario . |
after the experimental realization of bose - einstein condensates ( becs ) in ultracold atoms , a plethora of studies has been devoted to examining and understanding the coherent structures that arise in them @xcite . among these structures , the formation , interactions and dynamics of matter wave dark @xcite and bright solitons @xcite have been a central focus of research both from the experimental and from the theoretical side .
such nonlinear waves were experimentally generated in single - component becs over a decade ago @xcite .
the nature of nonlinear matter waves that can be created in a bec background , depends on the type of the interatomic interactions .
namely , dark solitons can be created in becs with atom - atom repulsion resulting from a positive scattering length , while bright solitons exist in single - component settings with attractive interatomic interactions resulting from a negative scattering length .
in addition to the above single - component context , soliton states can arise also in multi - component settings .
such condensates have been created as mixtures of different spin states of @xmath0rb @xcite and @xmath1na @xcite , and triggered numerous theoretical studies involving soliton complexes .
a prototypical example of the latter is a coupled dark - bright ( db ) soliton state in a highly elongated ( quasi - one - dimensional ) condensate cloud , consisting of a dark soliton in one component and a bright soliton in the second component of a binary bec featuring intra- and inter - species repulsion . since bright solitons are not self - sustained structures in repulsive ( self - defocusing ) media , db solitons are often called symbiotic , that is the dark soliton can be thought of as acting as an effective potential well trapping the bright soliton @xcite .
such symbiotic entities were first observed , and theoretically studied in the context of nonlinear optics @xcite .
however , their experimental realization in the atomic realm @xcite opened a new and highly controllable direction towards a deeper understanding of the dynamics and interactions of these states both with each other as well as with external traps @xcite .
additionally , current state - of - the art experiments offer the possibility of manipulating in a controllable fashion the nonlinear interactions via the well - established technique of feshbach resonances @xcite .
this motivated recent theoretical activities where the static and dynamical properties of dark - bright symbiotic matter waves have been investigated on the level of mean field theory .
mathematically , tuning the interactions corresponds to deviating from the integrable ( manakov ) limit @xcite of the relevant nonlinear schrdinger system , where nonintegrability is introduced when considering arbitrary nonlinearity ( i.e. , interaction ) coefficients .
the latter nonintegrable setting forms also the main focus of the present effort . in this context , despite the nonintegrability , analytical expressions of specific single - db soliton states and lattices thereof have been obtained in @xcite . adding a parabolic trapping potential , it was revealed how the effective restoring force acting on the db soliton depends on the inter - atomic interactions @xcite , verifying that the particle - like nature @xcite of the symbiotic soliton is preserved . in the same spirit , the dependence of the binding energy of a db soliton on the inter - species interactions was found analytically in @xcite , where moreover a proper bound on a phase imprinted in the bright soliton constituent was obtained ( i.e. considering also moving single db states ) , above which a breaking of the symbiotic entity was observed . in our previous work
@xcite the interactions between db matter waves and the consequent formation of bound states for out - of - phase ( anti - symmetric ) bright soliton components has been studied .
based on a two - soliton ansatz of the hyperbolic type @xcite , the full analytical expressions for the interaction energies between two db solitons were obtained for arbitrary nonlinear coefficients , and in the absence of a confining potential .
furthermore , the key intuition that repulsion mediated by the dark solitons at short distances , and attraction mediated by anti - symmetric bright solitons at longer distances , would be counterbalanced leading in turn to a bound state formation , has been enriched by taking into account the significant role of the cross - component interaction energy term .
the crucial dependence of the latter on the inter - species interaction coefficient has been analyzed .
it was shown that anti - symmetric stationary states exist and remain robust for a wide parametric window of inter - species repulsions .
importantly , an exponential instability of the anti - symmetric states was identified upon crossing a critical inter - species repulsion .
the latter was found to be associated with a subcritical pitchfork bifurcation , giving rise to asymmetric stationary states with mass imbalanced bright soliton counterparts . in the present work ,
we extend the aforementioned findings of @xcite upon considering significant deviations from the manakov limit , towards both the immiscible ( i.e. , dominated by inter - species repulsion and thus phase separated in the ground state ) regime , but also towards the largely unexplored miscible ( i.e. , dominated by intra - species repulsion ) regime . to this end
, we investigate the stability and dynamics of the anti - symmetric states , the so - called `` solitonic gluons '' @xcite , as well as the above - mentioned asymmetric modes covering both the miscible and immiscible parameter regime .
in particular , the full excitation spectrum of these symbiotic states is explored in detail by means of a bogolyubov - de gennes linearization analysis @xcite .
it is found that for the asymmetric modes the linearization spectrum is strongly affected when crossing the immiscibility- miscibility threshold , which is in turn directly connected with a rapid change of their density profiles as observed already in @xcite . as a next step
, the stability and dynamics of the anti - symmetric db pairs is explored past the immiscibility - miscibility threshold .
it is found that , in addition to the destabilization in the immiscible regime reported in @xcite , a second critical point occurs deep in the miscible regime , rendering the anti - symmetric state unstable once more .
this instability scenario is found to be associated with a _
pitchfork bifurcation giving rise to another family of mass imbalanced ( asymmetric ) symbiotic structures which are found to be stable . as the inter - species repulsion is decreased
, the miscible character of this regime alters the bright soliton component , resulting into asymmetric pairs with the bright solitons `` living '' on top of a finite background . comparing and contrasting the instability mechanisms in the different parameter regimes , the long - time evolution of the asymmetric and anti - symmetric states is performed numerically at different interaction ratios .
the presentation is structured as follows . in sec .
2 a description of the theoretical model and prior results regarding the existence of the anti - symmetric db soliton pairs is provided .
furthermore , we briefly comment on the methods to be used for the numerical analysis of our findings . sec .
3 contains the results regarding both the stability and the dynamics of asymmetric and anti - symmetric db soliton pairs beyond the integrable limit .
finally , in sec .
4 we summarize our findings and discuss future perspectives .
the model of interest is a system of a two - component bec strongly elongated along the @xmath2-direction , subject to a tight transverse harmonic trap of frequency @xmath3 .
such a mixture can e.g be composed of two different hyperfine states of the same alkali isotope , like @xmath0rb . within mean field theory and after integrating out the frozen transverse degrees of freedom ,
this mixture is described by the following two coupled ( 1 + 1)-dimensional gross - pitaevskii equations ( gpes ) @xcite : @xmath4 in the above equation , @xmath5 denote the mean - field wave functions of the two components normalized to the numbers of atoms @xmath6 , while @xmath7 and @xmath8 are the atomic mass ( identical for both components ) and chemical potentials , respectively .
the effective one - dimensional coupling constants are given by @xmath9 , where @xmath10 denote the three @xmath11-wave scattering lengths that account for collisions between atoms belonging to the same ( @xmath12 ) or different ( @xmath13 ) species .
we restrict our considerations to purely repulsive interactions , i.e. all @xmath14 , and consider an idealized homogeneous setting with no longitudinal trapping potential along the @xmath2-axis . by measuring densities @xmath15 , length , time and energy in units of @xmath16 , @xmath17 , @xmath18 , @xmath19 respectively , and in a second step rescaling space - time coordinates as @xmath20 , @xmath21 , and the densities @xmath22 , the system of eqs .
( [ model ] ) can be written in the following dimensionless form : @xmath23 in the above system of equations we slightly changed the notation , using @xmath24 and @xmath25 , indicating this way that the component @xmath26 ( @xmath27 ) will be supporting dark ( bright ) solitons .
furthermore , @xmath28 is the rescaled chemical potential , while the interaction coefficients are normalized to the scattering length @xmath29 , i.e. @xmath30 , and @xmath31 .
the system of eqs .
( [ deq11])-([deq21 ] ) conserves the total energy , @xmath32 , the rescaled total number of atoms in each component ( @xmath33 and @xmath34 , respectively ) and the rescaled total number of atoms @xmath35 , where : @xmath36 dx , \label{energy } \\
n&=&n_d+n_b=\sum_{i = d , b}\int^{\infty}_{-\infty } dx|\psi_i |^2 . \label{ntot}\end{aligned}\ ] ] in the special case where the nonlinear coefficients are all equal to each other , i.e. @xmath37 , eqs .
( [ deq11])-([deq21 ] ) correspond to the integrable manakov model @xcite .
in such a case , it particularly admits exact single - db - soliton solutions of the form @xcite : @xmath38+i\sin\phi , \label{dark}\\ \psi_b(x , t)&=&\eta { \rm sech}\left[d(x - x_0(t))\right ] \times \exp\left[ikx+i\theta(t)+i(\mu-1)t\right ] , \label{bright}\end{aligned}\ ] ] subject to the boundary conditions @xmath39 , and @xmath40 for @xmath41 . in the aforementioned solutions
, @xmath42 is the so - called soliton s phase angle which fixes the `` grayness '' of the dark soliton , while @xmath43 denotes the amplitude of the bright component .
furthermore , @xmath44 and @xmath45 correspond to the soliton s center and inverse width respectively , while @xmath46 is the wave - number of the bright soliton , associated with the speed @xmath47 of the db soliton , and @xmath48 is its phase .
while such a general exact expression can only be obtained in the manakov limit , one can utilize an approximate ansatz based on the expressions given in eqs .
( [ dark])-([bright ] ) and depart from the integrable limit .
such a method was employed in our very recent work of ref .
@xcite in order to analyze effective interactions between symbiotic matter waves in the general case of different inter - atomic repulsion strengths within each and between the species .
our aim in what follows is to briefly comment on the key results obtained there , to be connected with the numerical findings that will be presented below . in particular ,
a pair of two equal - amplitude dark - bright solitons travelling in opposite directions was considered having the approximate form : @xmath49 \psi_b(x , t ) & = & \eta\ , { \rm sech } x_{-}\ , { \rm e}^{i\left[kx+\theta(t ) + ( \mu-1)t \right]}+ \eta\ , { \rm sech } x_{+}\ , { \rm e}^{i\left[-kx+\theta(t ) + ( \mu-1)t \right]}\ , { \rm e}^{i\delta\theta}. \label{2dbb}\end{aligned}\ ] ] in these expressions @xmath50 where @xmath51 is the distance between the two db solitons , while @xmath52 is the relative phase between the bright solitons . if @xmath53 , the bright solitons are in - phase ( ip ) , while if @xmath54 , the bright solitons are out - of - phase ( op ) or anti - symmetric . within a hamiltonian variational approach
@xcite the approximate ansatz of eqs .
( [ 2dbd])-([2dbb ] ) is substituted into the total energy of the system given by eq .
( [ energy ] ) , and the relevant integrations are performed under the assumption that the soliton velocity is sufficiently small ( @xmath55 , @xmath56 ) .
the total energy then can be decomposed as @xmath57 , where @xmath58 corresponds to the energy of a single db soliton , contributing twice to the total energy , while the remaining three terms , namely @xmath59 , @xmath60 and @xmath61 , account for the interaction between : the two dark solitons , the two bright solitons , and the cross - interaction of a dark soliton in the first component with the bright soliton of the second component , respectively .
analytical expressions for each of these energy contributions as a function of the db soliton distance @xmath62 and for arbitrary nonlinear coefficients can be found in @xcite , together with simplified asymptotic expansions valid for large distances , @xmath63 . between two op dark - bright solitons . from left to right in panels ( * a*)-(*c * ) the value of the inter - species interaction coefficient is increased from @xmath64 to @xmath65 , while @xmath66 and @xmath67 throughout . in all cases both the full and the asymptotic ( valid for large @xmath62 ) expressions
are shown with solid green and dashed yellow lines , respectively .
the equilibrium separation @xmath62 of the stationary db soliton pair as a function of @xmath68 is shown in panel ( * d * ) . in this
latter panel circles in yellow correspond to the equilibrium distance obtained by numerically solving the gpe system of eqs .
( [ deq11])-([deq21 ] ) , while triangles in green denote the variational prediction of the equilibrium , i.e. the respective maximum of @xmath69 . ]
figure [ fig1 ] illustrates our key findings regarding the interaction energy of two out - of - phase db solitons as a function of their distance . here and throughout this work the chemical potential and the intra - species interaction coefficients
are fixed to @xmath67 and @xmath66 @xcite , respectively , while the interatomic interaction coefficient @xmath68 is left to widely vary . in panels ( a)-(c ) both the full and the asymptotic forms of the analytically obtained expressions for the total interaction energy , @xmath70 , are shown with solid green and dashed yellow lines , respectively . as it can be seen in all cases ,
the two results coincide at large distances as they should .
strikingly , in all three cases a pronounced local energy maximum is identified .
the identification of this extremum suggests , if we further take into account the effective negative mass of the db soliton @xcite , the existence of an effective _ stable _ ( with respect to variations in @xmath62 ) fixed point , thus a bound state for the two op dark - bright solitons .
such an anti - symmetric two - db soliton bound state can indeed be identified also in the full gross - pitaevskii system ( for the numerical methods used we refer the reader to the following section ) .
the predictive strength of the variational approximation is directly checked by comparing with the full numerical computation of the respective equilibria @xmath62 on the level of eqs .
( [ deq11])-([deq21 ] ) , and the outcome is illustrated in panel ( d ) , showing good agreement .
it is worth mentioning at this point that in order to evaluate the energy functional we need as input the width @xmath45 and amplitude @xmath43 occuring in eqs .
( [ 2dbd])-([2dbb ] ) .
these are obtained by numerically identifying the exact single - db state at the given @xmath68 and fitting to it the profile known from the integrable limit , i.e. eqs . ( [ dark])-([bright ] ) .
there are thus two main sources of error of this scheme , namely ( i ) the imperfect fit of the @xmath71-@xmath72 profile ( strictly valid only in the integrable limit ) to the single db soliton mode and ( ii ) the limited accuracy of the two - db soliton ansatz of eqs .
( [ 2dbd])-([2dbb ] ) especially at small separations @xmath62 .
for details regarding this and also a discussion of the case of in - phase bright solitons we refer the interested reader to ref .
we note also that as @xmath68 increases ( while @xmath73 and @xmath74 are kept fixed ) , the norm @xmath34 of the bright species is increasingly suppressed and eventually vanishes @xcite , beyond which point there is no bound two - soliton state anymore , since the presence of the bright component is a necessary ingredient for holding the dark solitons together @xcite .
this suppression of the bright norm is compatible with the overall decrease of the total energies from panel ( a ) towards ( b ) and ( c ) in figure [ fig1 ] .
having variationally predicted and numerically confirmed the existence of anti - symmetric stationary dark - bright soliton bound states for a large interval of values of the inter - species interaction parameter @xmath68 ( typically , we study @xmath75 in the present work ) , our main goal in the following is to address their stability and ( where applicable ) decay dynamics . in this section
we briefly comment on the numerical methods to be used so as to obtain stationary symbiotic states consisting of two dark - bright solitons and to determine their stability and time evolution . in the numerical computations that follow ,
we initially obtain stationary solutions of the system of eqs .
( [ deq11])-([deq21 ] ) in the form of @xmath76 and @xmath77 by means of a fixed - point numerical iteration scheme @xcite .
the linear stability of the latter is adressed by using the bogolyubov - de gennes ( bdg ) analysis @xcite . in particular to assess the stability of the obtained fixed points , we substitute the following ansatz into eqs .
( [ deq11])-([deq21 ] ) : @xmath78 in the above equations the asterisk denotes complex conjugation while @xmath79 is the amplitude of infinitesimal perturbations .
the resulting system of equations is then linearized , by keeping only terms of order @xmath80 , and the eigenvalue problem for the eigenfrequencies @xmath81 and eigenmodes @xmath82 is solved numerically . note that an instability occurs if modes with purely imaginary or complex eigenfrequencies are identified @xcite
. since the linearization spectrum is invariant under @xmath83 and @xmath84 , only results in the first quadrant of the complex plane will be shown below .
for the simulation of the time evolution based on eqs .
( [ deq11])-([deq21 ] ) , a fourth order runge - kutta algorithm is employed . in all cases ,
the numerical computations are performed on a spatial grid in the presence of an almost hard - wall super - gaussian potential @xcite that is chosen wide enough for boundary effects to be negligible on small and intermediate time scales .
within the interval @xmath85 . ( * a * ) real part , @xmath86 , of the eigenfrequencies @xmath81 as a function of @xmath68 . the corresponding imaginary part , @xmath87 ,
is shown in panel ( * b * ) . upon increasing @xmath68 there
exist two critical values , @xmath88 , and @xmath89 , indicated with light blue and gray boxes respectively , below and above which the anti - symmetric branch destabilizes .
the trajectories of the two anomalous modes ( see text ) appearing in the spectrum are shown with red stars .
( * c*)-(*d * ) profiles of the anti - symmetric states for @xmath64 , and @xmath65 , i.e. deep in the miscible and immiscible regime , respectively . ]
having verified that stationary anti - symmetric pairs of symbiotic matter waves can be found in a wide range of values of the inter - species repulsion coefficient @xmath68 , a natural next step is to consider the fate of these solutions under small perturbations , providing information on their stability in the different regions of existence .
the latter is explored by using the bdg linearization analysis discussed in section 2.3 . for the presentation of our results
, we will distinguish the miscible ( @xmath90 ) and immiscible ( @xmath91 ) regimes , separated by the miscibility - immiscibility threshold @xcite @xmath92 for our choice of @xmath66 . in figure
[ fig2 ] , the bdg spectrum of the anti - symmetric stationary two dark - bright soliton states is shown as a function of @xmath68 . in particular , both the real @xmath93 and the imaginary parts @xmath94 of the eigenfrequencies @xmath81 are depicted in the top ( a ) and middle ( b ) panels respectively . panels ( c ) and
( d ) depict profiles of the dark and bright wave functions at selected values of @xmath68 in the miscible and immiscible regime , respectively .
two general comments can be made before examining in detail the excitation spectrum .
the most significant one is that within the background spectrum denoted with blue circles there exist two distinguished modes .
the trajectories of the latter are illustrated with red stars .
these modes are the so - called anomalous modes since they posses a so - called negative krein signature @xmath95 @xcite , which for the two - component system considered herein is defined as : @xmath96 the sign of this quantity is a topological property associated with the excited nature of this state and the eigenvectors of such anomalous modes result in a variation of the solitary waves ( as opposed to a variation of the system s background ) .
furthermore , we note in passing that each continuous symmetry of the system corresponds to a pair of zero modes , @xmath97 , in the bdg spectrum .
thus we expect three pairs of such modes related respectively with the conservation of the particle number ( or the @xmath98 gauge invariance ) in each of the two components , and with the translation invariance due to the absence of a confining potential .
this is confirmed in the numerical data .
just by inspecting the trajectories of the two anomalous modes that appear in the spectrum , their very different behavior becomes apparent .
as it is observed in panel ( a ) of figure [ fig2 ] , the higher frequency mode decreases almost monotonically upon increasing @xmath68 .
we were able not only to identify this mode but also to relate it with an out - of - phase vibration of the bound db soliton pair around its equilibrium distance .
the latter can be done by adding the corresponding eigenvector for some value of the inter - species interaction coefficient to the relevant stationary anti - symmetric state ( results not shown here ) .
next , and also in the same panel of figure [ fig2 ] , let us closely follow the trajectory of the lower - lying anomalous mode .
the trajectory of this mode is more complicated than the former one . in particular , starting from the aforementioned reference point , @xmath99 , and increasing the interspecies interactions towards the immiscible regime , i.e. for @xmath100 , it can be seen that there exists a critical point @xmath89 where this mode destabilizes .
this destabilization corresponds to an eigenvalue zero crossing and signals the instability of the anti - symmetric configuration deep in the immiscible regime .
notice the non - zero imaginary part that appears past this critical point shown in panel ( b ) of figure [ fig2 ] .
the existence and destabilization of this mode was found to be related with a symmetry breaking of the bright soliton component , being linked , in turn , to other stationary states that are mass imbalanced with respect to their bright soliton counterpart .
the identification of these asymmetric states which exist below @xmath101 and collide with the anti - symmetric branch at @xmath101 in a subcritical pitchfork bifurcation was established in @xcite .
the following paragraphs will be devoted to a discussion of these asymmetric two - db solutions , before we then return to an analysis of the bdg spectrum of the anti - symmetric pair , figure [ fig2 ] , at small values of @xmath68 .
but for the asymmetric two - db soliton states .
notice in this case the significant modification of the stability properties ( i.e. , a nearly vanishing imaginary part of the eigenfrequency in panel ( b ) ) , as the miscibility - immiscibility threshold @xmath102 is crossed . ]
as mentioned above , the asymmetric and the anti - symmetric two - db states coincide at the bifurcation point @xmath101 . below this critical @xmath68 ,
the anti - symmetric state is stable and there are two symmetry - broken asymmetric solutions which are unstable ( as is characteristic of a subcritical pitchfork bifurcation ) .
indeed , these asymmetric branches can themselves be continued towards much smaller values of @xmath68 and exist both in the immiscible and the miscible regime , see the profiles shown in figure [ fig3 ] ( c ) and ( d ) . in both panels ,
dashed black lines denote the wavefunction of the dark soliton component , while solid green lines depict the corresponding bright soliton counterpart .
to gain further insight regarding the nature of the instability of these asymmetric states , their full bdg spectrum is illustrated in figure [ fig3 ] ( a ) and ( b ) .
once more , both the real @xmath86 and the imaginary @xmath87 parts of the eigenfrequencies @xmath81 are shown as a function of the interspecies interaction . as expected from the above discussion
, only one of the anomalous modes appears in the excitation spectrum of these mass imbalanced states and is depicted with stars in red . replacing the second anomalous mode
, there is now throughout a purely imaginary frequency signaling the instability of the asymmetric branch .
remarkably , as the immiscibility - miscibility threshold is crossed , the growth rate of the instability is drastically decreased , rendering these states only weakly unstable for @xmath103 .
this latter observation is in close contact with the change in the spatial character of these asymmetric states , and also in line with our previous findings @xcite .
namely , as the interspecies interactions decrease towards the miscible region , the asymmetric states change gradually from only weakly asymmetric ( i.e. weakly mass imbalanced with respect to their bright amplitudes ) , to maximally asymmetric ( i.e. almost purely dark / dark - bright bound states ) .
the difference is clearly seen in the profiles provided in panels ( c ) and ( d ) of figure [ fig3 ] .
having identified the unstable linearization eigenmodes , we now turn to the associated decay dynamics .
in particular , figure [ fig4 ] shows the long - time dynamics of the above - obtained stationary asymmetric states for different values of the nonlinear coefficient @xmath68 .
panels ( a)-(d ) correspond to the dark soliton component while the respective bright wave functions are depicted in panels ( e)-(h ) . for @xmath104
shown in panels ( a)-(c ) and ( e)-(g ) respectively , it is observed that as a result of the instability , the solitons move towards each other forming breathing - like structures .
this is accompanied by a redistribution of the bright `` filling '' component between the two dark solitons , in the form of almost complete tunneling back and forth . for parameter values slightly above @xmath105 , as e.g. the one depicted in panels ( c ) and ( g ) , a strong beating phenomenon is observed , clearly evident in the dark soliton component , with the solitons oscillating around a fixed distance from each other forming an almost stationary breather that persists till the end of the propagation . below the above - mentioned threshold and towards the miscible region
the picture becomes progressively more dramatic .
the beating gets much more pronounced with the solitons experiencing more frequent collisions as shown in panels ( b ) and ( f ) .
finally , when entering even deeper in the miscible side illustrated in panels ( a ) and ( e ) , eventually the bound pair fully splits into an essentially empty dark and a dark - bright soliton that are released ( i.e. , are no longer bound by each other ) and propagate towards the outer parts of the simulation domain ( where they are ultimately reflected by the boundaries ) . however , for @xmath106 , i.e. upon increasing @xmath68 towards the immiscible side depicted in panels ( d ) and ( h ) of figure [ fig4 ] , a rather different picture is painted by the symbiotic entities when compared to the asymmetric states presented above . in particular , slightly after the beginning of the propagation , where the asymmetric entity looked quite robust , a dramatic redistribution of the bright soliton s mass occurs .
the latter results in a strong repulsion between the single db soliton formed and the almost empty dark one leading in turn to their subsequent separation . note that a similar decaying mechanism was also observed in @xcite but for the unstable anti - symmetric states ( see also later on in the text ) . .
from left to right the interaction is increased from @xmath64 to @xmath107 , while panels ( * a*)-(*d * ) [ ( * e*)-(*h * ) ] correspond to the densities @xmath108 ( @xmath109 ) of the dark ( bright ) component . ]
we have now characterized the branch of asymmetric two - db soliton modes , which were seen to bifurcate from the anti - symmetric two - db soliton branch in a subcritical pitchfork bifurcation in the immiscible regime , at @xmath89 .
let us now return to the anti - symmetric branch itself and study its fate in the miscible regime .
this regime is also covered in the full excitation spectrum depicted in figure [ fig2 ] , with a typical profile of the state being shown in panel ( c ) .
departing from @xmath110 towards the miscible regime , and in particular by following once again the lower - lying anomalous bdg mode , we observe that as @xmath68 decreases a second critical point exists deep within the miscible side
. the eigenvalue zero crossing occurs at @xmath88 , and is indicated by the solid light blue box .
we note here , that this latter critical point was not discussed in our previous work @xcite , where only larger values of @xmath68 were studied .
the destabilization of the aforementioned mode suggests the existence of a second pitchfork bifurcation .
this is , once again , related with a symmetry breaking of the bright soliton component resulting into mass imbalanced ( i.e. , asymmetric ) two db states . as a function of the inter - species interaction coefficient @xmath68 .
the stable anti - symmetric and asymmetric branches are denoted with solid blue lines , while the unstable anti - symmetric one is shown with dashed green line .
notice that the asymmetric branches exist before the critical point , @xmath88 , but are stable verifying the supercritical nature of the bifurcation .
( * b*)-(*d * ) bdg spectra at three different values of @xmath68 showcasing the stability of the asymmetric states . in all cases
the anomalous modes are illustrated with red stars .
the associated stationary db profiles are depicted in panels ( * e*)-(*g * ) .
( * h*)-(*m * ) spatio - temporal evolution of the stationary asymmetric states shown in panels ( * e*)-(*g * ) showing the dynamical stability of these symbiotic structures .
panels ( * h*)-(*j * ) [ ( * k*)-(*m * ) ] correspond to the dark ( bright ) soliton component.,title="fig : " ] as a function of the inter - species interaction coefficient @xmath68 .
the stable anti - symmetric and asymmetric branches are denoted with solid blue lines , while the unstable anti - symmetric one is shown with dashed green line .
notice that the asymmetric branches exist before the critical point , @xmath88 , but are stable verifying the supercritical nature of the bifurcation .
( * b*)-(*d * ) bdg spectra at three different values of @xmath68 showcasing the stability of the asymmetric states . in all cases the anomalous modes
are illustrated with red stars .
the associated stationary db profiles are depicted in panels ( * e*)-(*g * ) .
( * h*)-(*m * ) spatio - temporal evolution of the stationary asymmetric states shown in panels ( * e*)-(*g * ) showing the dynamical stability of these symbiotic structures .
panels ( * h*)-(*j * ) [ ( * k*)-(*m * ) ] correspond to the dark ( bright ) soliton component.,title="fig : " ] indeed , we were able to identify these _ new _ pairs .
in contrast to the previous instability scenario , these mass imbalanced states exist past the critical point but are _ stable _ , i.e. the pitchfork deep in the miscible domain is found to be _
supercritical_. the corresponding bifurcation diagram is shown in panel ( a ) of figure [ fig5 ] . in order to obtain this diagram
we measure the relative bright imbalance defined as @xmath111 upon varying @xmath68 .
notice that four branches are identified , i.e. three stable ones consisting of two asymmetric and an anti - symmetric branch all denoted by solid blue lines , and one unstable anti - symmetric branch illustrated with dashed green line . to further demonstrate the stability of the new asymmetric symbiotic pairs in panels ( b)-(d ) the bdg spectra
are shown for different values of the inter - species interactions below the associated critical point @xmath88 .
two anomalous modes appear in the linearization spectra of these asymmetric states illustrated with red stars .
the respective stationary wave profiles are depicted in panels ( e)-(g ) , where the dark ( bright ) soliton wavefunction is shown with dashed black ( solid green ) line .
we observe that upon decreasing @xmath68 the asymmetry between the bright solitons increases , as was also the case for the respective asymmetric but unstable states found in the immiscible regime .
furthermore , for @xmath112 a background gradually builds up for the bright solitons as is evident in the stationary state shown in panel ( e ) of figure [ fig5 ] , revealing the miscible character of the regime supporting these states .
our detailed bdg analysis indicates that for all values within the miscible regime that we have checked , the asymmetric states exist as stable configurations and should remain dynamically robust .
this result is verified and highlighted in panels ( h)-(m ) of figure [ fig5 ] where we use as initial condition the stationary asymmetric states depicted in panels ( e)-(g ) .
we note that in all three cases panels ( h)-(j ) [ ( k)-(m ) ] show the evolution of the dark ( bright ) soliton component .
having studied both the static and the dynamical properties of the new asymmetric structures which bifurcate from the op dark - bright states , we now turn our attention to the dynamics of the anti - symmetric db waveforms . . from left to right the interaction is increased from @xmath64 to @xmath113 , and then to @xmath114 .
panels ( * a*)-(*d * ) [ ( * e*)-(*h * ) ] correspond to the densities @xmath108 ( @xmath109 ) of the dark ( bright ) component . ] in particular , we explore the long time evolution of the op db states so as to reveal the decay mechanisms that such a pair suffers from .
the dynamics at different values of @xmath68 are summarized in figure [ fig6 ] , all initiated at equilibrium .
as before , upper row of panels ( a)-(d ) shows the spatio - temporal evolution of the dark soliton counterpart , while panels ( e)-(h ) depict the corresponding bright component . from the stability analysis presented above it
is expected that for all values of the interspecies interaction coefficient @xmath115 the anti - symmetric two dark - bright soliton states exist as stable configurations , and as such should be robust throughout the propagation .
this latter result is confirmed in panels ( c ) and ( g ) of figure [ fig6 ] , for @xmath116 which is slightly above the lower critical point . however , and as anticipated , a very different picture is found below the critical point @xmath117 depicted in panels ( b ) and ( f ) , ( a ) and ( e ) respectively . starting with the former ,
we observe that slightly below @xmath117 the initial stationary state quickly decays and in ( b ) and ( f ) we observe periodic tunneling of the bright component between the two dark solitons , while the dark solitons are only relatively weakly affected here .
it is worth mentioning that similar tunneling dynamics has been identified and interpreted in terms of a bosonic josephson junction model in @xcite , but with the crucial difference that the soliton pair was further supported by the restoring force of a harmonic trap in that work , while in our present setup there is no external potential that would keep the dark solitons in place .
remarkably , in this regime despite the mass exchange , the bound soliton pair does not disintegrate . further decreasing @xmath68 , in panels ( a ) and ( e ) the effects of the instability are more drastic .
the time scale of the bright component oscillations decreases and the dark solitons vibrate more strongly . after some time of almost periodic oscillations , a more irregular type of motion sets in , with both dark solitons (
one filled by most of the bright component , the other almost empty ) eventually moving towards positive @xmath2 while separating and recolliding in the process , suggesting still a kind of effective attractive interaction between them .
this decay mechanism deep in the miscible regime is to be contrasted to the unstable dynamics in the immiscible regime , above @xmath101 . in panels ( d ) and
( h ) we show the time evolution of the anti - symmetric two - db soliton state at @xmath65 . while initially almost no dynamics is visible in the densities , especially no oscillations within the bright component , on intermediate time scales a strong asymmetry in the bright filling builds up ( see again @xcite for a more detailed discussion ) and subsequently the filled and the empty dark soliton split , showing no sign of effective attraction .
notice that the above described decaying mechanism is rather similar to the one observed for the unstable asymmetric states for @xmath106 ( see panels ( d ) and ( h ) of figure [ fig4 ] ) .
in the present contribution we investigated in detail the stability and dynamics of bound pairs of dark - bright symbiotic solitons , which arise as nonlinear matter wave excitations in mixtures of bose - einstein condensates featuring inter - atomic repulsion . in particular , we explored the scenario of differently weighted inter- and intra - species interaction coefficients , breaking the integrability of the relevant nonlinear schrdinger model .
it was argued by means of a recently proposed variational approach @xcite and shown numerically that upon departing from the integrable limit bound states of such symbiotic entities exist for anti - symmetric bright soliton counterparts , the so called solitonic gluons .
these anti - symmetric states were found to be robust within a bounded interval of the inter - species repulsion coefficient @xmath68 , limited by critical points both in the miscible and in the immiscible regime of the model , associated with a supercritical and a subcritical pitchfork bifurcation respectively . below and above these boundaries ,
i.e. deep in the miscible and the immiscible regime , respectively , the anti - symmetric pair becomes unstable .
long - time propagation revealed differences , but also common characteristics of the decay mechanisms in the two domains of instability .
specifically , a striking common feature is the relevance of bright mass transfer between the two dark solitons . in particular , in the miscible domain , we identified new stationary asymmetric states that bifurcate from the anti - symmetric ones in a supercritical pitchfork .
the stability of the new asymmetric states was also dynamically confirmed .
moreover , it was shown that a further decrease of the interspecies repulsion , results in asymmetric states with the bright solitons living on top of a finite bec background , highlighting in this way the miscible character of such bound pairs .
in contrast to the above picture , upon entering the immiscible regime , we had found in our recent work @xcite that the destabilization of the anti - symmetric dark - bright soliton pair is caused by a subcritical pitchfork bifurcation involving an unstable stationary , but as in the corresponding miscible domain also mass - imbalanced dark - bright soliton pair mode . in the present work , we further explored the range of existence and stability of this asymmetric branch , demonstrating its overall instability and its substantial deformation upon entering the miscible regime .
there are many directions that are worth considering in the future along the lines of this work .
in particular , by fixing the intra - species interactions @xmath66 in this work , we have not addressed the fate of the dark - bright soliton pair states when actually approaching the manakov ( integrable ) limit , which would require @xmath37 . in this respect it would be particularly interesting to see if the asymmetric state fully stabilizes upon restoring integrability or maintains its weakly unstable nature . towards this direction , and since in the integrable limit exact solutions are available @xcite , one could link the anti - symmetric and asymmetric states obtained here with the exact families of dark - bright soliton solutions known in the integrable case . establishing a possible connection of this type
would furthermore open a new direction of exploring and understanding the symbiotic soliton pairs , since in such a case one could depart once more from the integrable limit , but having at hand exact analytical expressions for the two - soliton problem rather than the approximate ones constructed only from the single soliton solution , as used in our present work . furthermore , and also in this direction , in the integrable model exact closed form expressions exist not only for static symbiotic states like the ones considered here , but also for moving and scattering ones . based on these , one could hope to get insights into the collisional dynamics of symbiotic entities at least in the vicinity of the integrable limit , paving the way of a detailed understanding of features like the breathing state formation observed here .
studies along these lines are left for future work .
weller , a. ; ronzheimer , j. p. ; gross , c. ; esteve , j. ; oberthaler , m. k. ; frantzeskakis , d. j. ; theocharis , g. ; kevrekidis , p. g. experimental observation of oscillating and interacting matter wave dark solitons .
lett . _ * 2008 * , * 101 * , 130401 - 1130401 - 4 .
theocharis , g. ; weller , a. ; ronzheimer , j. p. ; gross , c. ; oberthaler , m. k. ; kevrekidis , p. g. ; frantzeskakis , d. j. multiple atomic dark solitons in cigar - shaped bose - einstein condensates .
* 2010 * , _ 81 _ , 063604 - 1063604 - 15 .
denschlag , j. ; simsarian , j. e. ; feder , d. l. ; clark , c. w. ; collins , l. a. ; cubizolles , j. ; deng , l. ; hagley , e. w. ; helmerson , k. ; reinhardt , w. p. ; rolston , s. l. ; schneider , b. i. ; phillips , w. d. generating solitons by phase engineering of a bose - einstein condensate .
_ science _ * 2000 * , _ 287 _ , 97 - 101 .
anderson , b. p. ; haljan , p. c. ; regal , c. a. ; feder , d. l. ; collins , l. a. ; clark , c. w. ; cornell e. a. watching dark solitons decay into vortex rings in a bose - einstein condensate .
lett . _ * 2001 * , _ 86 _ , 2926 - 2929 .
strecker , k. e. ; partridge , g. b. ; truscott , a. g. ; hulet , r. g. formation and propagation of matter - wave soliton trains .
_ nat . _ * 2002 * , _ 417 _ , 150 - 153 .
myatt , j. c. ; burt , e. a. ; ghrist , r. w. ; cornell , e. a. ; wieman , c. e. production of two overlapping bose - einstein condensates by sympathetic cooling . _
* 1997 * , _ 78 _ , 586 - 589 .
hall , d. s. ; matthews , m. r. ; ensher , j. r. ; wieman , c. e. ; cornell , e. a. dynamics of component separation in a binary mixture of bose - einstein condensates . _ phys .
lett . _ * 1998 * , _ 81 _ , 1539 - 1542 .
stamper - kurn , d. m. ; andrews , m. r. ; chikkatur , a. p. ; inouye , s. ; miesner , h. -j .
; stenger , j. ; ketterle , w. optical confinement of a bose - einstein condensate . _ phys . rev .
lett . _ * 1998 * , _ 80 _ , 2027 - 2030 .
nistazakis , h. e. ; frantzeskakis , d. j. ; kevrekidis , p. g. ; malomed , b. a. ; carretero - gonzlez , r. bright - dark soliton complexes in spinor bose - einstein condensates .
rev . a _ * 2008 * , _ 77 _ , 033612 - 1033612 - 12 .
yan , d. ; chang , j. j. ; hamner , c. ; hoefer , m. ; kevrekidis , p. g. ; engels , p. ; achilleos , v. ; frantzeskakis , d. j. ; cuevas , j. beating dark - dark solitons in bose - einstein condensates .
_ j. phys .
_ * 2012 * , _ 45 _ , 115301 - 1115301 - 11 .
majed , o. ; alotaibi , d ; carr , l. d. dynamics of vector solitons in bose - einstein condensates .
_ arxiv : 1607.00108v2 _ * 2016*. chen , z. ; segev , m. ; coskun , t. h. ; christodoulides , d. n. ; kivshar , yu . s. coupled photorefractive spatial - soliton pairs .
b _ * 1997 * , _ 14 _ , 3066 - 3077 .
becker , c. ; stellmer , s. ; soltan - panahi , p. ; drscher , s. ; baumert , m. ; richter , e .-
m . ; kronjger , j. ; bongs , k. ; sengstock , k. oscillations and interactions of dark and dark - bright solitons in bose - einstein condensates .
phys . _ * 2008 * , _ 4 _ , 496 - 501 .
middelkamp , s. ; chang , j. j. ; hamner , c. ; carretero - gonzlez , r. ; kevrekidis , p. g. ; achilleos , v. ; frantzeskakis , d. j. ; schmelcher , p. ; engels , p. dynamics of dark - bright solitons in cigar - shaped bose - einstein condensates .
a _ * 2011 * , _ 375 _ , 642 - 646 .
yan , d. ; chang , j. j. ; hamner , c. ; kevrekidis , p. g. ; engels , p. ; achilleos , v. ; frantzeskakis , d. j. ; carretero - gonzlez , r. ; schmelcher , p. multiple dark - bright solitons in atomic bose - einstein condensates .
rev . a _ * 2011 * , _ 84 _ , 053630 - 1053630 - 10 .
lvarez , a. ; cuevas , j. ; romero , f. r. ; hamner , c. ; chang , j. j. ; engels , p. ; kevrekidis , p. g. ; frantzeskakis , d. j. scattering of atomic dark - bright solitons from narrow impurities . _ j. phys .
b _ * 2013 * , _ 46 _ , 065302 - 1065302 - 11 .
roberts , j. l. ; claussen , n. r. ; burke , j. p.jr . ; greene , c. h. ; cornell , e. a. ; wieman , c. e. resonant magnetic field control of elastic scattering in cold @xmath118rb . _ phys .
lett . _ * 1998 * , _ 81 _ , 5109 - 5112 .
thalhammer , g. ; barontini , g. ; de sarlo , l. ; catani , j. ; minardi , f. ; inguscio , m. double species bose - einstein condensate with tunable interspecies interactions .
lett . _ * 2008 * , _ 100 _ , 210402 - 1210402 - 4 .
mertes , k. m. ; merrill , j. w. ; carretero - gonzlez , r. ; frantzeskakis , d. j. ; kevrekidis , p. g. ; hall , d. s. non - equilibrium dynamics and superfluid ring excitations in binary bose - einstein condensates .
lett . _ * 2007 * , _ 99 _ , 190402 - 1190402 - 4 .
egorov , m. ; opanchuk , b. ; drummond , p. ; hall , b. v. ; hannaford , p. ; sidorov , a. i. measurement of @xmath119-wave scattering lengths in a two - component bose - einstein condensate .
_ phys . rev .
* 2013 * , _ 87 _ , 053614 - 1053614 - 10 .
skryabin , d. v. instabilities of vortices in a binary mixture of trapped bose - einstein condensates : role of collective excitations with positive and negative energies .
* 2000*,_63 _ , 013602 - 1013602 - 10 . | the existence , stability , and dynamics of bound pairs of symbiotic matter waves in the form of dark - bright soliton pairs in two - component mixtures of atomic bose - einstein condensates is investigated .
motivated by the tunability of the atomic interactions in recent experiments , we explore in detail the impact that changes in the interaction strengths have on these bound pairs by considering significant deviations from the manakov ( integrable ) limit .
it is found that dark - bright soliton pairs exist as stable configurations in a wide parametric window spanning both the miscible and the immiscible regime of interactions . outside this parameter interval
two unstable regions are identified and are associated with a supercritical and a subcritical pitchfork bifurcation , respectively .
dynamical manifestation of these instabilities gives rise to a redistribution of the bright density between the dark solitons , and also to symmetry - broken stationary states that are mass imbalanced ( asymmetric ) with respect to their bright soliton counterpart .
the long - time dynamics of both the stable and the unstable balanced and imbalanced dark - bright soliton pairs is analyzed . |
the motivation for this work has its origin in recent experimental data by dimauro et al @xcite who studied high order harmonic generation ( hohg ) and above threshold ionization ( ati ) in potassium driven by strong radiation in the wavelength range 32003900 nm .
although hohg and ati have been and continue to be studied extensively , the bulk of the data and theory have concentrated on the noble gases .
experimental convenience has been one of the reasons for this preference , but , at least as far as hohg is concerned , their relatively high ionization potentials and resistance to ionization have also tilted attention in their direction .
the alkali atoms belong to an entirely different class , when it comes to their behavior under strong field excitation . for atomic numbers comparable to the respective noble gas ( potassium versus argon in our context ) , their ionization potential is lower by more than a factor of three .
their excited states and distribution of the oscillator strength for transitions from the ground state are also considerably different .
the energy of the first excited state in potassium is much closer to the ground state than it is in argon . as a consequence ,
if we consider , for example , 12-photon ionization in potassium , five of the photons reach above the first excited state , and the rest seven must be absorbed within the manifold of its excited states .
in contrast , for 12-photon ionization of argon , it takes nine photons to reach the energy range of the first excited state and only the remaining three will involve excitation within the manifold of excited states .
in addition , the wavelength needed for potassium ( about 3000 nm ) is longer by a factor of three than that needed for argon ( about 1100 nm ) .
one might thus expect that , in the process of ionization , an extensive manifold of excited and rydberg states will be strongly driven and perhaps populated .
this should lead to a structure in the ati energy spectrum different in appearance from what we are accustomed to .
one might also anticipate that the behavior should have similarities with that observed in rydberg states driven by microwave fields .
resolution requirements in photoelectron energy analysis do not allow the observation of individual ati peaks in that case , although it is quite feasible to resolve such peaks in potassium driven from its ground state by radiation at mid - infrared wavelengths ; as sheehy et al
. @xcite have shown .
it is the exploration of possible links and similarities between these two situations that induced us to undertake this work .
clearly , the requirements on intensity for saturation are expected to be lower in potassium than in argon . moreover , given the expected participation of manifolds of excited states in potassium , the keldysh parameter as a criterion for the departure from multiphoton ionization may not be as valid as it should be in argon where most of the energy interval from the ground state to the continuum is empty of excited states ; imitating thus better the keldysh model which is essentially based on a ground state and an ionization threshold . the outline of this paper is as follows : the theory , namely the model used to describe the atom , and the two propagation methods used to solve the time - dependent schrdinger equation are briefly presented in section ( [ sec : method ] ) .
section ( [ sec : results ] ) , starts with a presentation of the parameter range of the simulations that follow and a demonstration of convergence by comparing two different methods .
it then moves on to present results in the 12- , 13- , and 14-photon ionization range , and discuss a low - energy plateau in the ati spectrum .
12-photon ionization of hydrogen starting from the 2s is compared to the results from potassium .
we conclude in section ( [ sec : conclusions ] ) , by summarizing the main findings of our results . in the appendix we present some details of the atomic structure model , and compare it with an alternative approach .
potassium , being an alkali atom , can be considered a single electron system for most of the phenomena in which double excitation does not play an essential role .
the first ionization threshold is about 4.34 ev above the ground state ; 18.8 ev are needed to reach the lowest doubly excited state , which leaves us with enough room to study single electron dynamics .
the simplest way to do this is by using a model potential that incorporates the effect of the core electrons , and thus reduces the dynamics to a single electron scheme .
different implementations of this basic idea have been used so far , with some success , in the single active electron approach , pioneered by kulander @xcite , the frozen core calculations @xcite , and other model potential calculations @xcite . for the purpose of studying the dynamics in the mid - infrared ,
it is sufficient to use a simple form of the potential , proposed by hellmann @xcite , which in atomic units is given as : @xmath3 where @xmath4 is 1.989 and @xmath5 is 0.898 .
all formulas that follow are given in atomic units .
the hellmann potential with the above mentioned parameters , results in a ground state energy of 38950 wavenumbers from the first ionization threshold , which is lower than the energy of the actual ground state ( 35010 wavenumbers ) by 11% .
owing to the difference between the ground state energies of the model potential and of the real atom , we also use a scaled wavelength , namely we rescale the energy of the photon needed in an experiment that studies the same process , by the ratio of the model to the theoretical ground state energy .
an alternative approach is to correct the energy of the ground state , and probably some matrix elements to satisfy the oscillator strength sum rules , but this leads to nonlocal modifications in the hamiltonian and is difficult to implement effectively when the propagation is not done in an eigenbasis expansion .
the dynamical part of the problem is treated by solving the resulting time - dependent schrdinger equation in the dipole approximation : @xmath6 \psi({\mbox{\boldmath{$r$ } } } , t),\ ] ] where @xmath7 is the wavefunction describing the outer electron , and depends on the spatial electronic coordinates and on time , @xmath8 is the time - independent field - free atomic hamiltonian , and @xmath9 is the dipole interaction of the atom with the field .
we only use the velocity form of the interaction operator , following detailed studies on the convergence properties of the solution @xcite , which have shown that the expansion of the wavefunction in terms of spherical harmonics can be shortened dramatically if the velocity gauge is used instead of the length gauge . in the velocity gauge
, the dipole operator can be cast in the following form : @xmath10 where @xmath11 is the fine structure constant , @xmath12 is the momentum operator , and @xmath13 is the vector potential which , within the dipole approximation , has no spatial dependence . we choose a convenient form for the pulse envelope , namely a @xmath14 , avoiding the long tails of a gaussian that make the numerics more difficult , without significantly affecting the results .
the explicit form is : @xmath15,\ ] ] where @xmath16 is the amplitude of the vector potential , @xmath17 is the full width at half maximum ( fwhm ) , @xmath18 and @xmath19 the maximum field strength and fundamental frequency respectively .
the solution of the resulting time - dependent equation is written in a system of spherical coordinates and expanded in terms of radial functions and angular spherical harmonics .
this choice is dictated by the central symmetry of the atomic system and has the advantage of requiring the discretization of only one coordinate .
note that this choice leads to efficient algorithms only if the linearly polarized field is not too strong ( compared to the coulomb field ) in which case the global symmetry of the entire system would rather be cylindrical .
writing the solution in the cylindrical system would require discretization of 2 coordinates @xcite greatly increasing the numerical cost of the algorithm .
the problem is therefore treated within `` a box '' ( a sphere in the present case ) whose radius is chosen sufficiently large to contain the expanded atom during the interaction .
part of the procedure for testing convergence consists in ascertaining that the ati spectrum is insensitive to the radius of the box .
we have used two methods of propagating the tdse , namely a propagation onto eigenstates in a box @xcite , and a propagation on a @xmath0-splines basis @xcite .
the expansion of the time - dependent wavefunction on an eigenbasis set reads : @xmath20 where @xmath21 are the field - free box eigenstates of the atom of angular momentum @xmath22 .
since we use linearly polarized light , we only need the @xmath23 magnetic quantum number , as the initial state has @xmath23 , and dipole transition selection rules forbid mixing of other magnetic sublevels .
the time - dependent schrdinger equation is transformed into : @xmath24 with the initial condition @xmath25 .
@xmath26 are the eigenvalues in the box ; @xmath27 are the dipole matrix elements .
thus the problem has been transformed to a set of ordinary differential equations for the coefficients @xmath28 of the wavefunction , which are solved using a high order , explicit propagation technique , namely a fifth - order and sixth - order runge - kutta - verner method .
the ionization yield is calculated by adding up the occupation probabilities of all discretized continuum states at the end of the pulse ; the above threshold ionization ( ati ) spectrum is obtained by the window operator projection technique @xcite . bound state populations are given directly by the square of the norm of the coefficients of equation ( [ eq : psi_basis ] ) .
for the construction of the box - eigenstates that are used as our basis , we use an expansion onto @xmath0-splines , a method that is gaining momentum in many parts of atomic physics as was pointed out in @xcite .
the codes we use are based on ideas published in @xcite , which have been expanded to accommodate the need for large boxes .
it should be stressed that after the basis has been constructed ( i.e. energies and matrix elements have been calculated ) , the rest of the procedure is neutral to the technique for the construction of the basis .
the second approach rests on expanding the radial part of the time - dependent wavefunction directly onto @xmath0-splines @xcite : @xmath29 where in addition to the spherical harmonics , @xmath30 is the @xmath31-th @xmath0-spline of order @xmath32 depending only on the radial coordinate and @xmath33 are time - dependent coefficients to be determined by the solution of the tdse .
again , only @xmath23 magnetic quantum numbers are relevant .
the major difference in this approach is that we need not prediagonalize anything other than the initial state .
substitution of equation ( [ eq : psi_bsplines ] ) into the schrdinger equation ( [ schroedinger_equation ] ) leads to a banded system of differential equations , which is solved by implicit propagation techniques , currently involving a bi - conjugate gradient method with preconditioner .
this second method of propagation scales only linearly with increasing box size : it is thus more efficient when we need a large box .
although our original exploratory calculations have been made with the eigenbasis expansion method , which works quite well for small boxes , most of the results presented in what follows have been obtained through the direct expansion of the time - dependent hamiltonian onto @xmath0-splines .
the only guideline as to what to expect in our study are the experimental data by sheehy et al .
@xcite , in which 12- , 13- , and 14-photon ionization of potassium has been studied , with 3.2 @xmath34 m , 3.6 @xmath34 m , and 3.9 @xmath34 m , 1.5 ps pulses respectively , and intensities close to saturation .
the 1.5 ps pulse is computationally impractical in the tdse framework , thus we have chosen to place our study in the short pulse regime , with the total pulse duration @xmath35 of the @xmath14 pulse being 20 optical cycles , which ( depending on the wavelength ) corresponds to a width of 96 fs to 136 fs for the results that we present .
the wavelengths we use are scaled , to compensate for the inaccuracy of the ground state energy , and are such that 12- , 13- and 14-photon ionization takes place .
an intensity range estimate is obtained by calculating the generalized cross - section through a scaling technique @xcite , using the energy ( 0.295 hartree ) , and radius ( 5.24 a.u . )
obtained with the general hartree - fock code published by froese - fischer @xcite . from the cross - section
, we obtain a saturation intensity estimate by solving @xmath36 , where @xmath37 is the ionization width , and @xmath38 an effective pulse duration , of the order of its fwhm @xcite .
after the first time - dependent calculations , it was established that scaling was underestimating saturation intensity by more than an order of magnitude .
we also calculate the ammosov , delone , krainov ( adk ) rate of tunneling ionization @xcite , which does not depend on the wavelength . solving @xmath39
, we obtain a saturation intensity estimate of about @xmath40 w/@xmath41 , in agreement with the results of the simulations .
for some characteristic wavelengths , corresponding to 12-photon ionization scaled from the experimental wavelength , and the limits of the 13-photon ionization range , we show , in table ( [ tab : parameter : all ] ) , the fwhm duration @xmath17 , the scaled - theory saturation intensity i@xmath42 , the adk theory saturation intensity estimate i@xmath43 , and the upper limit e@xmath44 of the converged ati spectrum for a 3000 atomic units box .
e@xmath44 is calculated by estimating the energy needed for a free electron originally placed at the nucleus to reach the boundary of the box during the pulse ; it is a useful simple estimate of the box size needed to study ati spectra , as has been shown in section ( 5 ) of @xcite .
next , and for two intensities , we show the ponderomotive energy @xmath1 , which is the major component of the shift of the rydberg states and the continuum @xcite .
it is given by : @xmath45 where @xmath46 is the laser intensity , @xmath47 the photon energy and @xmath48 the corresponding wavelength .
for the highest intensities that we use , the ponderomotive energy is a multiple of the photon energy , which is around a third of an ev for the wavelengths in the table .
we also present the keldysh tunneling parameter @xmath49 @xcite , defined by : @xmath50 where @xmath51 is the ionization potential , and @xmath1 the ponderomotive potential . note that for all wavelengths , and for intensities up to @xmath52 w/@xmath41
, @xmath49 is larger than one .
thus , according to the keldysh theory of tunneling ionization , the process lies in the multiphoton regime , and it is meaningful to refer to the order of the transitions involved .
ddccdcdd @xmath48(nm ) & @xmath17(fs ) & @xmath53(w/@xmath41 ) & @xmath54(w/@xmath41 ) & @xmath55(ev ) & @xmath46(w/@xmath41 ) & @xmath1(ev ) & @xmath49 + 2880 & 96 & @xmath56 & @xmath57 & 7.76 & @xmath58 & 0.073 & 5.23 + & & & & & @xmath59 & 0.733 & 1.65 + 3125 & 104 & @xmath60 & @xmath61 & 6.59 & @xmath58 & 0.086 & 4.82 + & & & & & @xmath59 & 0.863 & 1.52 + 3300 & 110 & @xmath62 & @xmath61 & 5.91 & @xmath58 & 0.096 & 4.56 + & & & & & @xmath59 & 0.961 & 1.44 + 2 in figure ( [ fig : level - photon:3300 nm ] ) we show a visual representation of potassium in a 3300 nm field , corresponding to the lowest energy photons in the 13-photon ionization range . a truncated part of the bound atomic levels , of angular momentum up to 4 , is shown , together with the quantum paths leading to 13-photon ionization .
graphs of this type prove to be useful tools in the qualitative analysis of the processes involved in multiphoton phenomena . in our study
, we analyze the wavefunction at the end of the pulse , and thus obtain information on the electron spectrum and ionization .
depending on the physical quantity we are looking for , the parameters needed to achieve convergence vary , making it easier to obtain the value of angle and energy integrated ionization , than the ati spectrum .
we have ensured the convergence of our results , by varying the box size and the grid sampling density , in a way similar to what has been presented in @xcite , and by relying on empirical findings such as the definition of e@xmath63 , or the needed density of discretized continuum states per photon energy , to guide our parameter choice .
we have in addition conducted a further independent test of the numerics involved , by comparing the two different methods , i.e. the expansion in terms of box - eigenstates , or the direct expansion of the radial part in a @xmath0-splines basis .
a sample result , for a demanding quantity such as the ati spectrum in the 13-photon range , is shown in figure ( [ fig : ati : basis - bsp : compare:7e11 ] ) , where the results of the two simulations at 3300 nm , @xmath64 w/@xmath41 and 110 fs are displayed on top of each other .
the direct @xmath0-splines method involves a box of 3000 a.u . , with 3000 linearly sampled @xmath0-splines of order 7 , for each angular momentum up to @xmath65 , parameters that have proven more than sufficient for the method to converge in the range shown .
after the propagation is over , a projection to scattering states is used to obtain the ati spectrum .
the eigenstates expansion involves a box of 2500 a.u .
, with 2500 linearly sampled @xmath0-splines of order 9 for each angular momentum up to @xmath66 for constructing the eigenbasis .
this basis was then truncated to the lowest 1000 basis elements per angular momentum ; using only the 1000 lowest states proved to be sufficient for the energy range presented here , since the higher energy discretized continuum states play a role in the high energy , low - yield part of the spectrum .
the window - operator technique @xcite was used to obtain the photoelectron spectrum after the simulation , and the spectrum was renormalized for the comparison with the @xmath0-splines spectrum . despite the differences in the methods used
, we observe an excellent agreement of the results in figure ( [ fig : ati : basis - bsp : compare:7e11 ] ) , even on a logarithmic scale .
the @xmath0-splines method shows richer structure at the minima between the peaks , demonstrating its superiority at the finest parts of the results . in detailed analysis of the ati results ( too long to be presented here ) ,
we have studied the behavior of the side peaks with intensity and we have noted that different peaks show different shifts with intensity , which helps us in classifying them as either freeman resonances @xcite or bardsley fringes @xcite . note further the clean formation of a plateau in the ati - peak heights , showing up in both methods , and extending over the first 5 to 6 peaks ; we study this plateau later in the paper .
since , in theory , the two methods are related by a unitary transformation of the basis from a spatially localized ( @xmath0-splines ) to a field - free diagonal ( eigenbasis ) representation , the agreement of the results is expected ; however , practice has shown that convergence is strongly affected by the underlying numerics , especially when it comes to ati spectra that stress the subtle parts of our codes .
it is the first time we have used such an elaborate procedure to ascertain the accuracy of our results .
all results presented from now on are from calculations with the direct expansion onto @xmath0-splines , which is more efficient both in computer space and time , when the scale of the simulation increases .
we have compared the results of both methods for all intensities at 3300 nm ( 13-photon range ) and some intensities in the 12-photon range ( 2880 nm ) .
the convergence of all other calculations presented was ensured within the @xmath0-splines propagation method only , using well documented techniques @xcite : variation of box - size , @xmath0-splines density and order , and variation of the number of angular momenta .
we begin the presentation of our results with a study of ion yields , beginning with the 12-photon process at a 2880 nm pulse , of 96 fs temporal width , and intensities ranging from @xmath58 w/@xmath41 to @xmath67 w/@xmath41 . a 2000 a.u
. box with 2000 linearly sampled @xmath0-splines of order 7 , for each angular momentum up to @xmath65 , proves sufficient for obtaining the ion yield . in figure ( [ fig :
yield:2880nm : fit ] ) we present the results , together with a power - law fit to the low - intensity part of the spectrum . in the same figure
we also show the ionization yield estimate obtained by the adk - theory : in this and the following figures where the adk predictions are displayed , we have integrated the adk - rate over a square pulse of maximum intensity equal to the fwhm of the pulse used .
saturation sets in at about @xmath68 w/@xmath41 , substantially higher than the scaled estimate of @xmath58 w/@xmath41 , and in very good agreement with the adk prediction .
the low - intensity spectrum seems to follow a multiphoton perturbative behavior , in agreement with the value of the tunneling parameter @xmath49 in table ( [ tab : parameter : all ] ) , and thus the power - law ; but the least squares fit yields a slope of 7.5 , substantially smaller than the lowest - order perturbation theory expectation of 12 .
the same behavior appears in the experimental results @xcite , where in the 12-photon ionization curve , a slope of 7 has been measured .
note that the adk - theory , which is often used in comparisons to experiments due to its simplicity , markedly fails to predict the low - intensity yield .
2 a scaled system , namely the 12-photon ionization process in hydrogen starting from the metastable 2s state and interacting with 4090 nm light , is used as a test for these results .
these two different systems are compared in figure ( [ fig : compare : k : h2s ] ) , where we show bound states of potassium and hydrogen for the 4 lowest angular momenta , together with the 12-photon quantum paths leading to ionization , in energy units scaled to the photon energy .
the 2s state is chosen instead of the ground state of hydrogen as the initial state , so that the ionization threshold energy and the distribution of the bound states , other than the degeneracies , resemble those of potassium . the resulting ion yield for a 136 fs pulse
is shown in figure ( [ fig : h2s : yield:2880-scal ] ) .
. box with 3500 linearly sampled @xmath0-splines of order 7 for each of the 21 angular momenta is used ; this is certainly an overkill just for obtaining the ionization spectrum , but we also analyzed the ati spectra which need such large boxes due to the long propagation times involved .
we see the same kind of structureless saturated spectrum ; albeit this time the yield is higher for the same intensities ( notice the scales in the figures ) , and the saturation intensity is smaller by a factor of 2 , in accordance with the scaling relations that predict a higher cross - section for hydrogen starting from the 2s .
the adk theory ( shown as the dashed line in the figure ) departs at the lower part of the spectrum , and predicts a smaller saturation intensity .
the low - intensity part again shows a linear dependence in the log - log plot , and this time the slope is 5.7 , even less than what it is in potassium .
the slopes in both potassium and hydrogen 2s , roughly equal the order needed to ionize from the first exited state , the 4p and 3s or 3d respectively , yet , no unambiguous model conforming to all of our data could be constructed .
working in a related context , pont et al .
@xcite have constructed a theory describing multiphoton ionization in a strong field of low frequency @xmath19 , obtaining an asymptotic expansion of the ac quasienergy in powers of @xmath69 .
their paper contains results for the rate of ionization from the 1s state of hydrogen in a circularly polarized 1064 nm field .
when translated into a log - log plot of the rate vs intensity , the rate seems to increase roughly like the eighth power of the intensity , instead of the twelfth power as should be expected if the rate was perturbative .
we move on to 13-photon ionization , where the spectrum shows an unexpected feature . in figure ( [ fig : yield:3300 nm ] ) we show the ion yield versus intensity , using 3300 nm , 110 fs pulses , calculated within a box of 3000 a.u . with 3000 @xmath0-splines per angular momentum up to @xmath65
; for comparison we also display the predictions of the adk theory .
the saturation intensity is similar to the one in the 12-photon case , and again a power - law behavior of the signal with intensity holds for the lowest intensities .
we observe , however , a `` knee '' in the ion - yield curve , for intensities around @xmath70 w/@xmath41 .
this feature resembles a dynamic resonance , rather unexpected given the very high order of the processes involved . 2 in figure ( [ fig : bound :
dist ] ) we plot the distribution of the final bound states for an intensity of @xmath71 w/@xmath41 .
we plot the population probability , versus the principal quantum number for a few of the lowest angular momenta .
most of the population is still in the ground state at this intensity ; most of the excited population is concentrated in the @xmath72 state , and this is true for all lower intensities .
structure appears in the low - angular momentum , low - excited states , then a smooth , flat part , and a bump at the highest excitations .
this bump turns out to be artificial , created by the pseudostates lying between the true bound states and the discretized continuum states .
this was confirmed by observing that by making the box smaller , and thus changing the number of supported bound states , the same effect always appeared in the region just before the continuum .
the smooth region just before the bump is expected , given the small energy differences of these states , compared with the shifts experienced during the pulse . in order to shed more light on the behavior of the `` knee ''
, we performed a series of calculations , for different wavelengths , scanning the entire 13-photon range , from 3125 nm , to 3300 nm , with a step of 25 nm , and with pulses of 20 optical cycles , whose widths correspond to 104 fs for the 3125 nm wavelength , and to 110 fs for 3300 nm . in figure ( [ fig :
yield : scan:13:wl ] ) , we plot the ion yield as a function of the wavelength , for a series of intensities ranging from @xmath58 w/@xmath41 , to @xmath73 w/@xmath41 .
the curves separate in a natural way , since for a fixed wavelength a higher intensity yields more ionization , and we note that for the highest intensities , above about @xmath59 w/@xmath41 , all wavelengths result in practically the same yield . in the graph
we label only the intensities of interest that are in the range of 28@xmath74 w/@xmath41 , where we see a broad resonance shifting to higher wavelengths with increasing intensity ; this suggests a resonance shifting downwards by increasing the intensity . in figure ( [ fig : ion - exc:3150 - 3300 ] )
we plot ( for a few , selected wavelengths ) the ionization curves , together with the total excitation , which is defined as the population in all bound states other than the ground state at the end of the pulse .
the excitation is essentially dominated by the population of the 4p state , for most of the low- to mid - intensity range .
note the smooth variation of the curves with the wavelength ; the results for the other wavelengths of figure ( [ fig : yield : scan:13:wl ] ) , essentially interpolate the ones shown here .
all curves exhibit a `` knee '' which is more pronounced in the excitation spectrum ; after that , ionization mimics excitation in its behavior , whereas for the lower intensities see especially 3150 nm and 3200 nm their behavior may differ substantially .
one can argue that the low intensity part is typical in the multiphoton picture , where the difference in the orders of the processes involved is expected to show up as a difference in excitation compared to ionization . in a floquet picture
, few selected avoided crossings would describe the bulk of the dynamics . as the intensity increases , and higher excitations
play a more important role , a regime is reached , where ionization and excitation are linked , similar to the transition to the classical chaos regime , which has been discussed in the microwave context , for example in @xcite . 2 for all wavelengths , and all intensities below the `` knee '' , the bound state distribution is qualitatively similar to the one in figure ( [ fig : bound : dist ] ) , with the 4p being by far the most populated excited state .
on the basis of figure ( [ fig : level - photon:3300 nm ] ) , this could happen assuming the shifts of the 4s and 4p to be such that the two states are brought closer together when the field is on .
indeed , by calculating the lowest order shift in the presence of the field and at the wavelengths of interest , it turns out that the 4s state shift is negative , and relatively small , whereas the 4p state , repelled by 5s , shifts down by more than three times as much .
thus , for the larger energy photons in the 13-photon range , a resonant excitation of the 4p state during the pulse occurs ; for the lower energy photons of the same range , the shift pushes the states towards each other easing a near - resonant transfer .
the phenomenon of a low excitation playing an essential role in a 13-photon ionization process , is easier to imagine in an atom like potassium , than in hydrogen starting from the ground state , since the excitation lies at less than half the energy that is needed to reach ionization .
it should be noted here that 13-photon ionization simulations in hydrogen , starting from the 2s at a scaled 13-photon wavelength of 4474 nm , display no corresponding characteristic , which can be attributed to the difference in the lowest excitation , as figure ( [ fig : compare : k : h2s ] ) shows .
we close our discussion of ionization curves , by presenting in figure ( [ fig : yield:3510 nm ] ) the results of 14-photon ionization calculations , at a 3510 nm wavelength , with a 117 fs pulse . a 3000 a.u .
box was used , with 3000 @xmath0-splines of order 7 per angular momentum , up to angular momentum @xmath65 .
we also present the adk theory predictions : the departure at the lower intensities is not so dramatic as it was at the shorter wavelengths .
we move on to the presentation of the ati spectra , which , as seen in figure ( [ fig : ati : basis - bsp : compare:7e11 ] ) , exhibit a clean plateau in the low energy range .
this plateau shows up at all wavelengths we have checked , in potassium as well as in the hydrogen simulations starting from the 2s , and it may thus be considered a global feature for mid - infrared wavelengths . in figure ( [ fig : ati : plateau:3300nm : all ] ) a series of ati spectra at 3300 nm , 110 fs @xmath14 pulses , and for selected intensities between @xmath58 w/@xmath41 and @xmath59 w/@xmath41 is shown ; the higher the intensity , the higher the signal shown in the figure . a box of 3000 a.u . , with 3000 @xmath0-splines of order 7 for each of the angular momenta up to @xmath65 is used .
the vertical axis in the figure corresponds to the ionization probability density in units of 1/ev .
the shift of the ati peaks to lower energies with increasing intensity is linear with intensity , and , as expected , is well described by the ponderomotive shift of the ionization threshold .
we note that the extent of the plateau grows , almost in proportion to the intensity . a clearer way of viewing this phenomenon is presented in figure ( [ fig : ati : plateau : peaks ] ) .
we plot the height of the first few ati peaks versus the intensity of the pulse , this time for a 3150 nm wavelength and a width of 105 fs , corresponding again to a 20-cycle pulse .
note the power law ( linear dependence on a log - log plot ) of the first few peak - heights with the intensity .
this indicates that a perturbative process is taking place . as the intensity increases ,
more peaks enter the plateau range and the perturbative picture ceases to hold .
this is demonstrated in figure ( [ fig : ati : plateau : peaks ] ) by the merging of the curves , which implies that for a range of peaks their heights are basically the same . as the intensity increases
, the plateau expands and more curves merge . in figure ( [ fig : ati :
plateau:3300nm : fit ] ) , we show the plateau range plotted versus the intensity for the calculation shown in figure ( [ fig : ati : plateau:3300nm : all ] ) . in this figure
, the plateau range is defined as the abscissa of the interception of a horizontal line joining the first few peaks , and a line over the decreasing peaks of the spectrum .
the fit was made by hand , to manually remove the effect of accidental resonances in the first peaks of the spectrum .
we notice the sharp linear dependence of the plateau range on the intensity . on the same figure ,
we show a fit of the selected points to a linear function of intensity .
the least squares fit gives a plateau range of @xmath75ev @xmath41/w@xmath76ev .
measured in units of @xmath1 for the 3300 nm wavelength , the same equation reads : @xmath77 ev .
this shows a plateau scaling in proportion to @xmath78 . 2 to clarify the relation of the plateau to the ponderomotive energy
, we combine all of the data from the simulations in the 12- , and 13-photon range .
we simplify the definition of the extent of the plateau to a rigorous one , whose extraction from the data is automated , namely , we use the abscissa of the first peak at which the exponential fall of the peak heights begins .
the results are shown in figure ( [ fig : ati : plateau : all ] ) , where for each intensity the plateau range measured in units of @xmath1 is shown for all wavelengths .
the error - bars reflect the imprecision due to cases in which the plateau ends between two consecutive peaks . from the definition of the plateau range that we use
, it follows that the real plateau has the same probability to be located anywhere within our error - bars .
notice the huge uncertainty ( larger than @xmath1 ) at small intensities which reflects the fact that @xmath1 is smaller than the peak spacing .
at higher intensities , the ponderomotive shift grows until it is considerably larger than the peak - spacing , thus shrinking the error - bars to less than @xmath1 .
the data is fitted to a normal distribution with an average value of 2.8 @xmath1 and a standard deviation of 0.5 @xmath1 .
this result relates to the classical theory of the ionization process , as has been first developed in the simpleman s model @xcite .
the main arguments have recently been clearly restated in the appendix of @xcite , although in the context of a rescattering picture @xcite .
the idea rests upon a free - electron maximally gaining 3.17 @xmath1 within a field , returning to its original position .
this happens within the first cycle , or the first period of the electronic motion , subsequent cycles lead to 2.4 @xmath1 , and then to less until the maxima gradually converge to 2 @xmath1 over many cycles , and the average kinetic energy becomes @xmath1 .
a relevant study has been made more than 10 years ago by gallagher in the microwave regime @xcite , where ionization of na rydberg states was studied , and the spectra were explained using the above mentioned theory . in that paper ,
a plot of the extend of the spectrum starting from the 40 s state of na , shows approximate scaling with 3.45 @xmath1 .
an energy of @xmath1 should be subtracted when compared to short - pulse experiments , as the electrons do not keep the extra ponderomotive energy since they can not sample the spatial gradient of the field .
after the observation , in the optical regime , of the long - range plateau extending from 2 to 8 @xmath1 @xcite , the theory has been extended to include backscattering from the nucleus , having thus provided answers to pertinent questions @xcite .
it should be noted that other than the original gallagher experiment , all of these experiments and theories exhibit a steep fall in the region up to @xmath79 , which then stabilizes or falls with a lower slope in the region from 2 to 8 @xmath1 .
this is in contrast to our findings in the mid - infrared regime that show a smooth , almost flat region in the low energy spectra , and an increased downward slope afterwards .
due to the numerical demands on convergence , in the present study we do not analyze the region up to and beyond 8 @xmath1 , and can not therefore establish the presence or not of a second plateau .
we have also confirmed that the plateau characteristic has no relation to the atomic structure involved , by examining the ati spectrum in the scaled problem , namely 12-photon ionization of hydrogen starting from the 2s state .
the data are from the simulations corresponding to the wavelength displayed in figure ( [ fig : compare : k : h2s ] ) , and the parameters used are the same as those used to plot figure ( [ fig : h2s : yield:2880-scal ] ) .
few selected ati spectra are shown in figure ( [ fig : h2s : ati:1 - 2 - 4 - 7e11 ] ) , all of which display the plateau , whose range is estimated as @xmath80 , close to what we obtained from the cumulative data in potassium .
in summary , we have conducted a theoretical study of the dynamics of potassium interacting with a high intensity , mid - infrared , short , laser pulse .
we chose potassium because it has recently been studied experimentally by sheehy et al .
we have performed time dependent calculations in terms of both a direct expansion of the schrdinger equation onto @xmath0-splines , as well as an expansion onto field - free eigenstates within a box , and obtained remarkable agreement between the two methods .
we have studied a 12-photon ionization process , and have observed that the low - intensity , power - law behavior of the ion yield has an exponent much lower than the perturbative expectation of 12 .
we have obtained the same behavior from the study of a scaled system , namely hydrogen starting from the 2s in the 12-photon range . in the 13-photon ionization of potassium
, we noted a `` knee '' structure in the ion yield , and linked it to a similar , more pronounced , behavior in the excitation .
we interpret this feature as a broad dynamical resonance with the lowest excited state .
the ati spectra , in all cases that we have studied , display a clean formation of a plateau in the first few peak - heights .
although the extension of a plateau in the ati spectrum at optical wavelengths from 2 to 8 times the ponderomotive energy has been discussed in the literature , the existence of a clear low - energy plateau has ( to the best of our knowledge ) not been observed in other studies in the optical or uv regime , but has been measured in the microwave regime and interpreted by the simpleman s theory of ionization @xcite . through the analysis of the cumulative data from our simulations , we have determined the extent of the plateau to scale with the ponderomotive energy @xmath1 as @xmath2 , which is compatible with the predictions of the classical theory .
given the much longer wavelength and therefore longer optical period , an initially launched wavepacket will have more time to spread before it backscatters from the core .
this raises the question as to whether backscattering would play an important role at this wavelength .
recall that backscattering for shorter wavelengths has been associated with changes of the slope of the ati spectrum up to around 10 @xmath1 .
a related question of course is whether the initially launched wavepacket might be narrower as a result of which their might not be substantial additional spreading by the time it reaches the core . a definitive evaluation of this aspect would require much more extensive calculations which may be worth undertaking in the future .
we would like to thank dr .
l. f. dimauro for making available their experimental results @xcite prior to publication .
computer time , space and support from the rechnenzentrum garching , and especially john cox and dr .
r. volk is gratefully acknowledged .
one of us ( p. m. ) would like to thank dr .
l. a. a. nikolopoulos for useful discussions .
in the appendix we discuss the model potential used , by comparing it to the accepted structure of potassium and to the alternative frozen core hartree - fock model .
the atomic structure of potassium can be found in several places .
since we can not in a straightforward manner incorporate relativistic effects in our theory , we use the weighted energy levels , which are calculated as : @xmath81 the atomic eigenenergies @xmath82 are taken from @xcite . for easier comparison with the numerical results , we measure the energies from the ionization threshold which is at 35009.814 wavenumbers according to sugar and corliss @xcite .
the multiplet oscillator strengths are calculated using the approximate formula @xcite : @xmath83 the data needed in this formula are taken from @xcite .
note that in this database the quantity that is given is @xmath84 .
the multiplet oscillator strengths are shown in the second column of table ( [ tab : potassium : fosc ] ) .
a few oscillator strengths are also presented in page 300 of @xcite .
they are in agreement with the ones shown in the table .
dddd state & e@xmath85 ( @xmath86 ) & e@xmath87 ( @xmath86 ) & e@xmath88 ( @xmath86 ) + @xmath89 & -35010 & -32253 & -38947 + @xmath90 & -13983 & -13376 & -15761 + @xmath91 & -7560 & -7326 & -8331 + @xmath72 & -21986 & -20972 & -22647 + @xmath92 & -10296 & -10000 & -10696 + @xmath93 & -6005 & -5876 & -6209 + @xmath94 & -13474 & -12755 & -11952 + @xmath95 & -7612 & -7213 & -6735 + @xmath96 & -4824 & -4600 & -4322 + @xmath97 & -6882 & -6860 & -6852 + @xmath98 & -4403 & -4390 & -4384 + @xmath99 & -3057 & -3049 & -3045 + @xmath100 & -4393 & -4390 & -4389 + 2 an alternative approach to the model potential , that originally seemed appealing , is the frozen core hartree - fock method .
its main merit is its _ ab initio _ nature , and the comparatively better description of atomic structure .
extensive theoretical discussion exists on the frozen core hartree - fock method .
we used an implementation that is documented in @xcite .
the work in that paper was concerned with the application of the method to configuration interaction on the frozen core hartree fock basis , whereas here we are interested only in the single outer electron case .
atomic structure is described quite well , as we see in the third column of table ( [ tab : potassium : energies ] ) where we show selected bound state energies , and in columns 3 and 4 of table ( [ tab : potassium : fosc ] ) , where we present bound - bound oscillator strengths in the length and velocity gauge starting from s , p , and d states respectively .
the energies are measured from the first ionization threshold and are expressed in wavenumbers for direct comparison with the available data .
a 500 a.u . box , with 500 @xmath0-splines of order 9 was used for the calculations .
we notice that the agreement between oscillator strengths calculated in the length and velocity gauges is quite good for the cases displayed in the tables .
the major disadvantage of the method is the inconvenient scaling of the time needed to calculate the basis , and the actual size of the basis calculated , which limits us to small cases , up to 1000 a.u .. the time needed by the frozen core hartree fock scales as @xmath101 , although in principle it is limited by a @xmath102 factor .
the size scales with @xmath103 per angular momentum , so that for a @xmath104 basis of 20 angular momenta , we need approximately 2 gb and 35 cpu days in our workstation cluster to perform the structure calculations .
thus it is mainly numerical considerations that dictate the use of the pseudopotential method .
cdddd state & f & f@xmath87(len ) & f@xmath87(vel ) & f@xmath88 + @xmath105 & 1.01 & 1.07 & 1.02 & 0.96 + @xmath106 & 0.009 & 0.01 & 0.008 & 0.026 + @xmath107 & 0.0009 & 0.0012 & 0.0008 & 0.0055 + @xmath108 & 1.49 & 1.52 & 1.49 & 1.42 + @xmath109 & 0.031 & 0.026 & 0.024 & 0.075 + @xmath110 & 1.92 & 1.96 & 1.94 & 1.82 + @xmath111 & 0.18 & 0.18 & 0.17 & 0.19 + @xmath112 & 0.019 & 0.018 & 0.017 & 0.009 + @xmath113 & 0.89 & 0.93 & 0.97 & 0.90 + @xmath114 & 0.0003 & 0.012 & 0.015 & 0.092 + @xmath115 & 0.003 & 0.0002 & 0.0005 & 0.028 + @xmath116 & 0.31 & 0.32 & 0.31 & 0.34 + @xmath117 & 1.20 & 1.25 & 1.29 & 1.00 + @xmath118 & 0.0077 & 0.032 & 0.036 & 0.14 + @xmath119 & 1.48@xmath120 & 0.0037 & 0.0046 & 0.048 + @xmath121 & 0.14 & 0.16 & 0.18 & 0.14 + @xmath122 & 0.0066 & 0.0066 & 0.0078 & 6.75@xmath123 + @xmath124 & 0.76 & 0.88 & 0.88 & 1.061 + @xmath125 & 0.17 & 0.16 & 0.16 & 0.14 + @xmath126 & 0.067 & 0.062 & 0.062 & 0.049 + @xmath127 & 0.30 & 0.34 & 0.35 & 0.25 + @xmath128 & 0.39 & 0.17 & 0.17 & 0.97
+ @xmath129 & 0.14 & 0.62 & 0.62 & 0.18 + 2 the last columns in tables ( [ tab : potassium : energies ] ) , and ( [ tab : potassium : fosc ] ) are the results of the hellmann pseudopotential , as presented in equation ( [ eq : hellmann : definition ] ) , calculated with a box of 2500 a.u . , with 2500 @xmath0-splines of order 9 .
since the potential is @xmath22-independent , length and velocity gauge oscillator strengths agree , without the need of introducing a correction to the dipole operator @xcite .
we calculated the standard deviation of the length and velocity absorption oscillator strengths starting from a specific state .
for all bound states and up to the lower part of the continuum spectrum , which is what interests us , this was less than @xmath130 , which reassures us of the completeness of our description .
the model potential represents the atom less accurately than the hartree - fock does , both energies and oscillator strengths .
nevertheless , its excellent numerical properties , originating from its simplicity , make the large calculations feasible .
scaling techniques help to map the model atom results onto real experiments , but our main aim is to study general features pertaining in the mid - infrared range , thus making the actual atom used of secondary importance
. 2 j. sugar and c. corliss , _ atomic energy levels of the iron - period elements : potassium through nickel _
supplement no . 2 in _ journal of physical and chemical reference data _ ( american chemical society and american institute of physics , national measurement laboratory , national bureau of standards , gaithersburg , maryland 20899 , 1985 ) . | we study the dynamics of the potassium atom in the mid - infrared , high intensity , short laser pulse regime .
we ascertain numerical convergence by comparing the results obtained by the direct expansion of the time - dependent schrdinger equation onto @xmath0-splines , to those obtained by the eigenbasis expansion method .
we present ionization curves in the 12- , 13- , and 14-photon ionization range for potassium .
the ionization curve of a scaled system , namely hydrogen starting from the 2s , is compared to the 12-photon results . in the 13-photon regime ,
a dynamic resonance is found and analyzed in some detail .
the results for all wavelengths and intensities , including hydrogen , display a clear plateau in the peak - heights of the low energy part of the above threshold ionization ( ati ) spectrum , which scales with the ponderomotive energy @xmath1 , and extends to @xmath2 . 2 |
the _ diagonal _ of a multivariate power series with coefficients @xmath3 is the univariate power series with coefficients @xmath4 . particularly interesting
is the class of diagonals of _ rational _ power series ( ie , taylor expansions of rational functions ) . in particular , diagonals of _ bivariate _ rational power series are always roots of nonzero bivariate polynomials ( ie , they are algebraic series ) @xcite .
this property persists for multivariate rational power series , but only in positive characteristic , while the converse inclusion algebraic series being diagonals of rational series always holds @xcite .
as far as we are aware , the first occurrence of this result in the literature is an article of plya s @xcite , which deals with a particular class of bivariate rational functions ; the proof uses elementary complex analysis . along the lines of plya s
approach , furstenberg @xcite gave a ( sketchy ) proof of the general result , over the field of complex numbers ; the same argument has been enhanced later @xcite,@xcite .
three more different proofs exist : a purely algebraic one that works over arbitrary fields of characteristic zero ( * ? ? ?
6.1 ) ( see also ( * ? ? ?
6.3.3 ) ) , one based on non - commutative power series ( * ? ?
5 ) , and a combinatorial proof @xcite that relies on an encoding of the diagonal using unidimensional walks , seen themselves as words of a non - ambiguous context - free language .
various other generalizations are known @xcite .
[ [ polynomial - equations ] ] polynomial equations + + + + + + + + + + + + + + + + + + + + despite the richness of the topic and the fact that most proofs are constructive in essence , we were not able to find in the literature any _ explicit _ algorithm for computing a bivariate polynomial that cancels the diagonal of a general bivariate rational function .
we design in section [ sec : diagonals ] such an algorithm for computing a polynomial equation for the diagonal of an arbitrary bivariate rational function .
we show in proposition [ prop : generic ] that generically , the size of the minimal polynomial for the diagonal of a rational function is exponential in the degree of the input and that our algorithm computes it in quasi - optimal complexity ( theorem [ thm : bound diagonals ] ) .
the algorithm has two main steps that may be of independent interest .
the first step is the computation of a polynomial equation for the residues of a bivariate rational function .
we propose an efficient algorithm for this task , that is a polynomial - time version of bronstein s algorithm @xcite ; corresponding size and complexity bounds are given in theorem [ th : bronstein ] .
the second step is the computation of a polynomial equation for the sums of a fixed number of roots of a given polynomial .
we design an additive version of the platypus algorithm @xcite and analyze it in theorem [ thm : platypus - bound ] .
[ [ recurrences ] ] recurrences + + + + + + + + + + + since it is also classical that algebraic series are differentially finite ( ie , satisfy linear differential equations with polynomial coefficients ) , the coefficients of these bivariate diagonals satisfy linear recurrences and this leads to an optimal algorithm for the computation of their first terms @xcite .
we show however , that computing an annihilating polynomial of the diagonal first is usually not the right approach and that a direct computation of the recurrence @xcite will be more efficient . for completeness
, we mention that in more than two variables , diagonals of rational functions are still differentially finite @xcite and currently the most efficient algorithm in that situation is that based on the griffiths - dwork method @xcite .
[ [ walks ] ] walks + + + + + diagonals of rational functions appear naturally in enumerative combinatorics . in particular , the enumeration of unidimensional walks has been the subject of recent activity , see @xcite and the references therein .
three generating functions of different types of walks are of interest : the generating series @xmath5 of bridges , @xmath6 of excursions and @xmath7 of meanders ( these are defined precisely in section [ sec : walks ] ) .
the algebraicity of these generating functions is classical as well , and related to that of bivariate diagonals . beyond this structural result
, several quantitative and effective results are known .
explicit formulas give the generating functions in terms of implicit algebraic functions attached to the set of allowed steps in the cases of excursions @xcite,@xcite , bridges and meanders @xcite .
moreover , bousquet - mlou gave a tight exponential bound on the degree of the annihilating polynomial in the case of excursions @xcite , while banderier and flajolet designed an algorithm ( called the _ platypus algorithm _ ) computing it @xcite .
our message for these walks is that again , precomputing a polynomial equation is too costly if one is only interested in the enumeration .
instead , we propose to precompute a differential equation for @xmath5 , that has polynomial size only , to use it for expanding @xmath5 , and to recover the expansion of @xmath6 from that of @xmath5 .
for meanders , we compute a polynomial - size differential equation for @xmath8 , from which the expansion of @xmath7 can be computed efficiently .
our algorithms have quasi - linear complexity in the precision of the expansion , while keeping the precomputation step in polynomial complexity ( theorem [ thm : walks ] ) .
[ [ structure - of - the - article ] ] structure of the article + + + + + + + + + + + + + + + + + + + + + + + + after a preliminary section on background and notation , we first discuss several special bivariate resultants of broader general interest in sections [ sec : residues ] and [ sec : summation of residues ] .
next , we consider diagonals , the size of their minimal polynomials and an efficient way of computing annihilating polynomials in section [ sec : diagonals ] .
finally , we turn to walks in section [ sec : walks ] and show how to compute the coefficients of the generating functions of excursions and of meanders efficiently
. a preliminary version of this article has appeared at the issac15 conference @xcite . in the present version ,
we give tight bounds in the main results ( theorems [ thm : platypus - bound ] and [ thm : bound diagonals ] ) , an improved algorithm for the algebraic residues and more detailed proofs throughout . [ [ acknowledgments ] ] acknowledgments + + + + + + + + + + + + + + + this work was supported in part by the project fastrelax anr-14-ce25 - 0018 - 01 .
in this section , that might be skipped at first reading , we introduce notation and technical results that will be used throughout the article . in this article , @xmath9 denotes a field of characteristic 0 , and @xmath10 an algebraic closure of @xmath9 .
we denote by @xmath11_n$ ] the set of polynomials in @xmath11 $ ] of degree less than @xmath12 .
similarly , @xmath13 stands for the set of rational functions in @xmath14 with numerator and denominator in @xmath11_n$ ] , and @xmath15_n$ ] for the set of power series in @xmath15 $ ] truncated at precision @xmath12 . if @xmath16 is a polynomial in @xmath17 $ ] , then its degree with respect to @xmath18 ( resp .
@xmath19 ) is denoted @xmath20 ( resp .
@xmath21 ) .
we take the convention that @xmath22 .
the _ bidegree _ of @xmath16 is the pair @xmath23 .
the notation @xmath24 without any subscript is used for univariate polynomials .
inequalities between bidegrees are component - wise .
the set of polynomials in @xmath17 $ ] of bidegree less than @xmath25 is denoted by @xmath17_{n , m}$ ] , and similarly for more variables .
the _ valuation _ of a polynomial @xmath26 $ ] or a power series @xmath27 $ ] is its smallest exponent with nonzero coefficient .
it is denoted @xmath28 , with the convention @xmath29 .
the _ reciprocal _ of a polynomial @xmath30 $ ] is the polynomial @xmath31 .
if @xmath32 with @xmath33 and @xmath34 for all @xmath35 , the notation @xmath36 stands for the generating series of the _ newton sums _ of @xmath16 : @xmath37 a polynomial is called _ square - free _ when its gcd with its derivative is trivial .
a _ square - free decomposition _ of a nonzero polynomial @xmath38 $ ] , where @xmath39 or @xmath11 $ ] , is a factorization @xmath40 , with @xmath41 $ ] square - free , the @xmath42 s pairwise coprime and @xmath43 .
the corresponding _ square - free part _ of @xmath44 is the polynomial @xmath45 .
if @xmath44 is square - free then @xmath46 .
the coefficient of @xmath47 in a power series @xmath48 $ ] is denoted @xmath49a$ ] . if @xmath50 , then @xmath51 denotes the polynomial @xmath52 .
the exponential series @xmath53 is denoted @xmath54 .
the _ hadamard product _ of two power series @xmath55 and @xmath5 is the power series @xmath56 such that @xmath49a\odot b=[x^n]a\cdot[x^n]b$ ] for all @xmath12 . if @xmath57 is a bivariate power series in @xmath58 $ ] , the _ diagonal _ of @xmath0 , denoted @xmath1 is the univariate power series in @xmath59 $ ] defined by @xmath60 we recall classical complexity notation and facts for later use .
let @xmath9 be again a field of characteristic zero .
unless otherwise specified , we estimate the cost of our algorithms by counting arithmetic operations in @xmath9 ( denoted `` ops . '' ) at unit cost .
the soft - o notation @xmath61 indicates that polylogarithmic factors are omitted in the complexity estimates ( see ( * ? ? ?
25.8 ) for a precise definition ) .
arithmetic size _ of an element of @xmath9 is 1 . that of a univariate polynomial is its degree plus 1 ( ie
, we are considering _ dense _ representations ) . that of tuples of polynomials is the sum or their sizes , and this defines the size for rational functions and multivariate polynomials .
we say that an algorithm has quasi - linear complexity if its complexity is @xmath62 , where @xmath63 is the maximal arithmetic size of the input and of the output . in that case , the algorithm is said to be _ quasi - optimal_. * univariate operations .
* throughout this article we will use the fact that most operations on polynomials , rational functions and power series in one variable can be performed in quasi - linear time .
standard references for these questions are the books @xcite and @xcite , as well as @xcite .
the needed results are summarized in fact [ fact : complexity ] below .
[ fact : complexity ] the following operations can be performed in @xmath64 ops . in @xmath9 : addition , product and differentiation of elements in @xmath11_n$ ] , @xmath13 and @xmath15_n$ ] ; integration in @xmath11_n$ ] and @xmath15_n$ ] ; extended gcd , square - free decomposition and resultant in @xmath11_n$ ] ; multipoint evaluation in @xmath11_n$ ] , @xmath13 at @xmath65 points in @xmath9 ; interpolation in @xmath11_n$ ] and @xmath13 from @xmath12 ( resp .
@xmath66 ) values at pairwise distinct points in @xmath9 ; inverse , logarithm , exponential in @xmath15_n$ ] ( when defined ) ; conversions between @xmath30_n$ ] and @xmath67_n$ ] . * multivariate operations . *
basic operations on polynomials , rational functions and power series in several variables are hard questions from the algorithmic point of view .
for instance , no general quasi - optimal algorithm is currently known for computing resultants of bivariate polynomials , even though in several important cases such algorithms are available @xcite .
multiplication is the most basic non - trivial operation in this setting .
the following result can be proved using kronecker s substitution ; it is quasi - optimal for a fixed number of variables @xmath68 . for polynomials with more complicated monomial supports , or
when the number of variables grows , more sophisticated techniques apply @xcite .
[ fact : multiprod ] for fixed @xmath69 , polynomials in @xmath70_{d_1 , \dots , d_m}$ ] and power series in @xmath71_{d_1 , \dots , d_m}$ ] can be multiplied using @xmath72 ops .
a related operation is multipoint evaluation and interpolation .
the simplest case is when the evaluation points form an @xmath69-dimensional tensor product grid @xmath73 , where @xmath74 is a set of cardinal @xmath75 ; it extends to subgrids of tensor product grids @xcite .
[ fact : multieval]@xcite for fixed @xmath69 , polynomials in @xmath70_{d_1 , \dots , d_m}$ ] can be evaluated and interpolated from values that they take on @xmath76 points that form an @xmath69-dimensional tensor product grid using @xmath72 ops .
again , the complexity in fact [ fact : multieval ] is quasi - optimal for fixed @xmath68 .
a general ( although non - optimal ) technique to deal with more involved operations on multivariable algebraic objects ( eg , in @xmath17 $ ] ) is to use ( multivariate ) evaluation and interpolation on polynomials and to perform operations on the evaluated algebraic objects using facts [ fact : complexity][fact : multieval ] . to put this strategy in practice , the size of the output needs to be well controlled .
we illustrate this philosophy on the example of resultant computation , based on the following easy variation of ( * ? ? ?
* thm . 6.22 ) .
[ fact : resultant ] let @xmath77 and @xmath78 be bivariate polynomials of respective bidegrees @xmath79 and @xmath80 . then , @xmath81 and this is an equality whenever one of @xmath82 or @xmath83 is zero .
[ algo resultant ] let @xmath16 and @xmath44 be polynomials in @xmath84_{d_1 , \dots , d_m , d}$ ]
. then @xmath85 belongs to @xmath70_{d_1 , \dots , d_m}$ ] , where @xmath86 .
moreover , the coefficients of @xmath87 can be computed using @xmath88 ops . in @xmath9 .
the degrees estimates follow from fact [ fact : resultant ] . to compute @xmath87
, we use an evaluation - interpolation scheme : @xmath16 and @xmath44 are evaluated at @xmath89 points @xmath90 forming an @xmath69 dimensional tensor product grid ; @xmath91 univariate resultants in @xmath92_d$ ] are computed ; @xmath87 is recovered by interpolation .
by fact [ fact : multieval ] , the evaluation and interpolation steps are performed in @xmath93 ops .
the second one has cost @xmath94 . using the inequality
@xmath95 concludes the proof .
we conclude this section by recalling two complexity results on bivariate polynomials and rational functions ; for proofs , see @xcite and @xcite .
[ fact : sqfree ] a square - free decomposition of polynomials in + @xmath17_{d_x , d_y}$ ] can be computed using @xmath96 ops .
if @xmath97 $ ] are non - zero coprime polynomials such that @xmath98 and @xmath44 is primitive wrt @xmath19 , then a minimal telescoper for @xmath99 of degree @xmath100 and order at most @xmath101 can be computed using @xmath102 ops , where @xmath103 and @xmath101 is the degree in @xmath19 of any square - free part of @xmath44 .
recall that a minimal telescoper for @xmath99 is a differential operator @xmath104\langle \partial_x\rangle$ ] of minimal order such that @xmath105 with @xmath106 $ ] .
we are interested in a polynomial that vanishes at some or all of the residues of a given rational function .
it is a classical result in symbolic integration that in the case of simple poles , there is a resultant formula for such a polynomial , first introduced by rothstein @xcite and trager @xcite .
this was later generalized by bronstein @xcite to accommodate multiple poles as well .
however , as mentioned by bronstein , the complexity of his method grows exponentially with the multiplicity of the poles .
instead , we develop in this section an algorithm with polynomial complexity .
let @xmath107 be a nonzero element in @xmath108 , where @xmath109 are two coprime polynomials in @xmath92 $ ] .
let also @xmath110 be a divisor of @xmath44 such that @xmath110 and @xmath111 are coprime . in our context
, @xmath110 represents the subset of the roots of @xmath44 at which we want to compute an annihilating polynomial of the residues .
let @xmath112 be a square - free decomposition of @xmath110 .
for @xmath113 , if @xmath114 is a root of @xmath42 in an algebraic extension of @xmath9 , then it is simple and the residue of @xmath115 at @xmath114 is the coefficient of @xmath116 in the laurent expansion of @xmath117 at @xmath118 .
consider the polynomial @xmath119 . since @xmath114 is a simple root of @xmath42
, @xmath120 satisfies @xmath121 and @xmath122 .
therefore , the rational function @xmath123 defined by @xmath124 satisfies @xmath125 and has the advantage of being regular at @xmath118 .
the residue of @xmath115 at @xmath114 may hence be computed as the evaluation at @xmath126 of @xmath127g(y , t)$ ] . if this coefficient is denoted @xmath128 , with polynomials @xmath129 and @xmath130 ,
the residue at @xmath114 is thus a root of @xmath131 .
when the multiplicity of the pole @xmath132 , this is exactly the rothstein - trager resultant .
this computation leads to algorithm [ algo : bronstein ] , which avoids the exponential blowup of the complexity that would follow from a symbolic precomputation of the bronstein resultants .
algorithm * * algebraicresidues**@xmath133 three polynomials @xmath16 , @xmath44 and @xmath110 a divisor of @xmath44 in @xmath92 $ ] such that @xmath110 and @xmath111 are coprime ( @xmath110 can be @xmath44)a polynomial in @xmath134 $ ] canceling the residues of @xmath99 at the roots of @xmath110 compute @xmath135 a square - free decomposition of @xmath110 ; @xmath136 @xmath137 ; @xmath138 ; expand @xmath139 ; write @xmath140 as @xmath141 with @xmath129 and @xmath130 coprime polynomials ; @xmath142 ; @xmath143 [ ex : bronstein ] let @xmath144 be an integer , and let @xmath145 $ ] be the rational function @xmath146 . the poles have order @xmath147 . in this example , the algorithm can be performed by hand for arbitrary @xmath63 : a square - free decomposition has @xmath148 and @xmath149 , the other @xmath42 s being 1
. then @xmath150 and the next step is to expand @xmath151 expanding the binomial series gives the coefficient of @xmath152 as @xmath153 , with @xmath154 the residues are then cancelled by @xmath155 , namely by @xmath156 ( equality is a consequence of the identity ] @xmath157 , which implies @xmath158 , while @xmath159 . ) in our applications , as in the previous example , the polynomials @xmath16 and @xmath44 have coefficients that are themselves polynomials in another variable @xmath18 .
the rest of this section is devoted to the proof of the following .
[ th : bronstein ] let @xmath160 .
let @xmath110 be a divisor of @xmath44 , @xmath161 be a square - free part of it wrt @xmath19 , and denote by @xmath69 the number of factors in the square - free decompositions of @xmath110 .
let @xmath162 be bounds on the bidegree of @xmath163 .
then the polynomial computed by algorithm [ algo : bronstein ] annihilates the residues of @xmath99 at the roots of @xmath110 , has degree in @xmath164 bounded by @xmath165 and degree in @xmath18 bounded by @xmath166 it can be computed in @xmath167 operations in @xmath9 .
note that rewriting the bound under the equivalent form @xmath168 shows that the degree in @xmath18 is bounded by @xmath169 , independently of the multiplicities .
the complexity is also bounded independently of the multiplicities by @xmath170 .
by fact [ fact : resultant ] , the resultant @xmath171 has degree in @xmath164 exactly @xmath172 so that the degree in @xmath164 of the result is bounded by @xmath173 .
the degree in @xmath18 is the sum of the degrees in @xmath18 of all the @xmath171 s . in order to derive a bound on the degree of @xmath171 using fact [
fact : resultant ] , we first consider the degrees in @xmath18 and @xmath19 of @xmath129 and @xmath130 .
the important point is that these degrees do not depend so much on @xmath44 as on its square - free part . in order to quantify this precisely , we first focus on power series expansion of a special type about which we state a few useful lemmas . for a polynomial @xmath174 $ ] and a real number @xmath114 ,
we denote by @xmath175 the subset of @xmath176 $ ] formed of power series that can be written @xmath177 with @xmath178 $ ] and @xmath179 , for all @xmath12 ( recall that @xmath180 , which makes it convenient to allow negative @xmath114 ) .
this notation extends to the case when @xmath18 is a tuple of variables , with @xmath114 replaced by a tuple of real numbers .
the main properties of @xmath175 are summarized as follows .
[ growth ] let @xmath181 $ ] , @xmath182 and @xmath183 $ ] . the set @xmath175 is a subring of @xmath176 $ ] ; let @xmath184 with @xmath185 , then @xmath186 ; the products obey @xmath187 for _ ( 3 ) _ , if @xmath188 and @xmath189 belong respectively to @xmath175 and @xmath190 , then the @xmath12th coefficient of their product is a sum of terms of the form @xmath191 .
therefore , the degree of the numerator is bounded by @xmath192 , whence _ ( 3 ) _ is proved .
property _ ( 1 ) _ is proved similarly , the @xmath12th coefficient of the product being a sum of terms @xmath193 . in property _ ( 2 ) _ , the condition on @xmath194 makes @xmath195 well - defined .
the result then follows from _
( 1)_. [ coro : invpol ] let @xmath196 $ ] be such that @xmath197 .
let @xmath198 be a square - free part of @xmath44 and @xmath199 its total degree in @xmath200 .
then @xmath201 for all @xmath35 , the coefficient of @xmath202 in @xmath44 has degree at most @xmath203 .
thus @xmath204 .
writing @xmath205 and using part _ ( 2 ) _ of lemma [ growth ] with @xmath206 then gives the result when @xmath44 is square - free .
using @xmath207 gives the result for a pure power by part _ ( 1 ) _ of the lemma .
the general case then follows from part _ ( 3 ) _ by induction on the number of parts in the square - free decomposition of @xmath44 , using additivity of degree and total degree .
now , we turn to the fraction @xmath208 , with @xmath209 and @xmath210 .
we use bidegrees with respect to @xmath211 and observe that @xmath212 the total degrees in @xmath213 behave similarly : that of @xmath214 is @xmath215 , while that of @xmath216 is @xmath217 .
corollary [ coro : invpol ] gives @xmath218 from there , part _ ( 3 ) _ of lemma [ growth ] shows that the product of these series belongs to @xmath219 thus the coefficient @xmath140 of @xmath220 in the power series expansion of @xmath221 can be written as @xmath222 with @xmath223 and finally @xmath224 whence @xmath225 fact 4 can now be exploited , leading to a bound on the degree of the resultant : @xmath226 next , we sum over the indices @xmath35 corresponding to factors of @xmath110 .
this leads to the following bound for the degree in @xmath18 of the result @xmath227 this bound being an increasing function of each of the degrees that appear , it is itself upper bounded by replacing any of those degrees by an upper bound . in the context of theorem
[ th : bronstein ] , the bidegrees of @xmath16 , @xmath44 and @xmath110 are bounded by @xmath228 , while those of @xmath198 and @xmath161 are bounded by @xmath162 .
this leads to the bound @xmath229 which rewrites as the bound in the theorem and completes that part of the proof .
by fact [ fact : sqfree ] , a square - free decomposition of @xmath110 can be computed using @xmath96 ops .
we now focus on the computations performed inside the @xmath35th iteration of the loop and write @xmath230 for the bidegree of @xmath42 .
computing @xmath231 requires an exact division of polynomials of bidegrees at most @xmath232 ; this division can be performed by evaluation - interpolation in @xmath233 ops .
similarly , the trivariate polynomial @xmath120 can be computed by evaluation - interpolation wrt @xmath211 in time @xmath234 . by eq . , both @xmath235 and @xmath236 have bidegrees at most @xmath237 , where @xmath238 and @xmath239 .
they can be computed by evaluation - interpolation in @xmath240 ops .
finally , the resultant @xmath241 has bidegree at most @xmath242 , and since the degree in @xmath19 of @xmath243 and @xmath42 is at most @xmath244 , it can be computed by evaluation - interpolation in @xmath245 ops by lemma [ algo resultant ] .
the total cost of the loop is thus @xmath246 , where @xmath247 using the ( crude ) bounds @xmath248 , @xmath249 , @xmath250 and @xmath251 shows that @xmath252 is bounded by @xmath253 which , by using the inequalities @xmath254 and @xmath255 , is seen to belong to @xmath256 , as was to be proved .
this completes the proof of the theorem .
* remark . *
note that one could also use hermite reduction combined with the usual rothstein - trager resultant in order to compute a polynomial @xmath257 that annihilates the residues .
indeed , hermite reduction computes an auxiliary rational function that admits the same residues as the input , while only having simple poles .
a close inspection of this approach provides the same bound @xmath101 for the degree in @xmath19 of @xmath257 , but a less tight bound for its degree in @xmath18 , namely worse by a factor of @xmath101 .
the complexity of this alternative approach appears to be @xmath258 ( using results from @xcite ) , to be compared with the complexity bound from theorem [ th : bronstein ] .
given a polynomial @xmath259 $ ] of degree @xmath63 with coefficients in a field @xmath9 of characteristic 0 , let @xmath260 be its ( not necessarily distinct ) roots in the algebraic closure of @xmath9 . for any positive integer @xmath261 , the polynomial of degree @xmath262 defined by @xmath263 has coefficients in @xmath9 .
this section discusses the computation of @xmath264 summarized in algorithm [ algo : sigma_c ] , which can be seen as an additive analogue of the _ platypus algorithm _ of banderier and flajolet @xcite .
algorithm * * purecomposedsum**@xmath265 a polynomial @xmath16 of degree @xmath63 in @xmath92 $ ] , a positive integer @xmath266the polynomial @xmath267 from eq .
@xmath268 @xmath269 @xmath270 @xmath271 @xmath272f)\odot\sum{n!y^n}\bmod y^{d+1}$ ] @xmath273 we recall two classical formulas for the generating function of the newton sums ( see , eg , @xcite ) , the second one being valid for monic @xmath16 only : @xmath274 truncating these formulas at order @xmath147 makes @xmath36 a representation of the polynomial @xmath16 ( up to normalization ) , since both conversions above can be performed quasi - optimally by newton iteration @xcite . the key for algorithm [ algo : sigma_c ] is the following variant of @xcite .
let @xmath259 $ ] be a polynomial of degree @xmath63 , let @xmath36 denote the generating series of its newton sums and let @xmath275 be the series @xmath276 .
let @xmath277 be the polynomial in @xmath278 $ ] defined by @xmath279\exp\left(\sum_{n \ge 1}{(-1)^{n-1}t_n\frac{z^n}{n } } \right).\ ] ] then the following equality holds @xmath280 by construction , the series @xmath275 is @xmath281 when applied to the polynomial @xmath267 , this becomes @xmath282\prod_{i=1}^{d}{\left(1+z \exp(\alpha_i y)\right)}. \end{aligned}\ ] ] this expression rewrites : @xmath283\exp\left(\sum_{i=1}^{d}\log(1+z \exp({\alpha_i y}))\right)&= [ z^c]\exp\left(\sum_{i=1}^{d}\sum_{m\ge 1}{(-1)^{m-1}\exp({\alpha_i m y})\frac{z^m}{m}}\right)\\ & = [ z^c]\exp\left(\sum_{m\ge 1}{(-1)^{m-1}s(m y)\frac{z^m}{m}}\right ) , \end{aligned}\ ] ] and the last expression equals @xmath284 .
the correctness of algorithm [ algo : sigma_c ] follows from observing that the truncation orders @xmath285 in @xmath19 and @xmath286 in @xmath164 of the power series involved in the algorithm are sufficient to enable the reconstruction of @xmath264 from its first newton sums by
. we will be interested in the case where @xmath16 is a polynomial in @xmath17 $ ] .
then , the coefficients of @xmath264 wrt @xmath19 may have denominators .
we analyze the structure of the coefficients of @xmath267 as elementary symmetric functions of the roots of @xmath16 in order to compute bounds on the bidegree of the polynomial obtained by clearing out these denominators .
the rest of this section proves the following result .
[ thm : platypus - bound ] let @xmath287_{d_x+1,d_y+1}$ ] , and let @xmath288 be a positive integer .
let @xmath289 $ ] denote the leading coefficient of @xmath16 wrt @xmath19 and let @xmath267 be defined as in eq . .
we also denote @xmath290 then @xmath291 is a polynomial in @xmath17 $ ] that cancels all sums @xmath292 of @xmath293 roots @xmath294 of @xmath16 , with @xmath295 , and satisfies @xmath296 moreover , this polynomial can be computed in @xmath297 ops .
these bounds are sharp .
experiments suggest that for generic @xmath16 of bidegree @xmath232 the minimal polynomial of @xmath292 has bidegree precisely @xmath298 . similarly , the complexity result is quasi - optimal up to a factor of @xmath293 only .
we start with the following effective version of a very classical result on symmetric functions ( * ? ? ?
* theorem 6.21 ) .
[ lem : symmetric functions ] let @xmath299 be indeterminates , and @xmath300 be the associated elementary symmetric functions .
let @xmath301 $ ] be a symmetric polynomial satisfying @xmath302 then @xmath16 can be expressed as a polynomial in @xmath300 of total degree at most @xmath63 .
this is a consequence of the form of the matrix of the change of bases from the elementary symmetric functions to the monomial symmetric functions as described for instance in the proof of ( * ? ? ?
* theorem 7.4.4 ) .
since @xmath16 is symmetric and has degree at most @xmath63 with respect to each variable , it can be written as a linear combination of monomial symmetric functions of the form @xmath303 , where @xmath304 for all @xmath35 .
these monomial symmetric functions can in turn be written as linear combinations of elementary symmetric functions of the form @xmath305 where @xmath306 , which is exactly the result of the lemma .
for the proof of the bounds in theorem [ thm : platypus - bound ] , we write @xmath307 let @xmath308 denote the elementary symmetric functions of the @xmath309 s .
then , the elementary symmetric functions of the roots @xmath310 of @xmath267 have degree @xmath311 in each @xmath309 .
therefore , by lemma [ lem : symmetric functions ] , the coefficients of @xmath267 are polynomials of total degree at most @xmath311 in @xmath312 .
from there , the bound on @xmath313 is immediately derived from the classical relations @xmath314 .
the computation is performed by evaluation and interpolation at @xmath315 values of @xmath18 . by fact
[ fact : complexity ] , at each of these values , the computation of the truncated series expansions @xmath316 and @xmath275 in @xmath317_{1+d_y}$ ] have complexity @xmath318 ; so do the computations of @xmath319 and the last step ; the most expensive step is the computation of @xmath0 , which costs @xmath320 ops . in @xmath9 .
since this is executed @xmath321 times , the total cost is @xmath322 .
in this section we turn to our main topic , namely the computation of annihilating polynomials for diagonals of bivariate rational functions .
the algorithm relies on a classical expression of the diagonal as a sum of residues ( see lemma [ lem : diagonal - as - residues ] ) , and on the results of sections [ sec : residues ] and [ sec : summation of residues ] .
the conclusions of the analysis of algorithm [ algo : diagonal ] can be found in theorem [ thm : bound diagonals ] and proposition [ prop : generic ] .
let @xmath323 be a rational function in @xmath324 , whose denominator does not vanish at @xmath325
. then the diagonal of @xmath0 is defined as @xmath326 .
a first basic , but very important , remark is that @xmath327\frac{1}{y}f\!\left(\frac{t}{y } , y\right).\ ] ] when @xmath328 , this coefficient can be viewed as a cauchy integral and computed by the residue formula @xcite . for general @xmath9 ( of characteristic @xmath329 ) , we proceed similarly with a purely algebraic approach , adapted from ( * ? ?
* theorem 6.1 ) .
( the reader who is not interested in the general proof may also skip directly to lemma [ lem : diagonal - as - residues ] . )
the starting point is the partial fraction decomposition of @xmath330 considered as a rational function in @xmath331 : @xmath332 where @xmath333 in particular , @xmath334 is the residue of @xmath335 at @xmath336 for all @xmath337 . by puiseux s theorem
, there exists @xmath338 such that the @xmath339 s and @xmath340 s all lie in the field @xmath341 . in order to apply the operator
@xmath342 $ ] on both sides of equation , it is necessary to find a ring where both the equality and the operator @xmath342 $ ] make sense .
we are going to check that @xmath343 and @xmath342 $ ] computed coefficient - wise are suitable for this .
first , as a rational function , it is immediate that @xmath344 belongs to @xmath345 . in order to expand the right - hand side
, we consider each term separately and distinguish between the cases @xmath346 and @xmath347 .
if @xmath348 , @xmath349 can be written as follows : @xmath350.\ ] ] since @xmath351 , the series @xmath352 actually belongs to @xmath353[[y]]\cong\overline{\kk}[[y]][[t^{1/n}]]$ ] .
hence @xmath354((t^{1/n}))\subset \aa$ ] , and in particular @xmath342f_{i , j}=0 $ ] . on the other hand ,
if @xmath347 then @xmath349 can be expanded directly in @xmath345 as : @xmath355 since @xmath356 and @xmath357 , this last quantity is the sum of a convergent series ( in the sense of formal laurent series ) of elements of @xmath345 , hence belongs to @xmath345 . in this case
we obtain @xmath342f_{i , j}=r_{i,1}$ ] .
we have everything we need to apply @xmath342 $ ] on both sides of equation , leading to the generalization to any base field of characteristic @xmath329 of furstenberg s classical result @xcite .
[ lem : diagonal - as - residues ] if @xmath358 is a rational function in @xmath324 whose denominator does not vanish at @xmath325 , then @xmath359 where @xmath360 is the set of poles of @xmath361 .
the poles @xmath362 such that @xmath363 are called the _ small branches _ of @xmath44 and we denote their number by @xmath364 .
since the elements of @xmath360 are algebraic and finite in number and residues are obtained by series expansion , which entails only rational operations , it follows that the diagonal is algebraic too .
combining the algorithms of the previous section gives algorithm [ algo : diagonal ] that produces a polynomial equation for @xmath1 .
let @xmath144 be an integer , and let @xmath365 be the rational function @xmath366 .
the diagonal of @xmath367 is equal to @xmath368 by the previous argument , it is an algebraic series , which is the sum of the residues of the rational function @xmath369 of example [ ex : bronstein ] over its small branches ( with @xmath18 replaced by @xmath370 ) . in this case , the denominator is @xmath371 .
it has one solution tending to 0 with @xmath370 ; the other one tends to @xmath372 .
thus the diagonal is canceled by the quadratic polynomial .
algorithm * algebraicdiagonal*(a / b ) two polynomials @xmath55 and @xmath5 in @xmath324 , with @xmath373a polynomial @xmath374 $ ] such that @xmath375 @xmath376 , @xmath377 , @xmath378 @xmath379 @xmath380 @xmath381 number of small branches of @xmath44 @xmath382 @xmath383 @xmath384 for an integer @xmath385 , we consider the rational function @xmath386 of bidegree @xmath387 .
the first step of algorithm [ algo : diagonal ] produces @xmath388 of bidegree @xmath389 , whose denominator is irreducible with @xmath63 small branches .
from there , algorithm [ algo : diagonal ] computes a polynomial @xmath390 annihilating @xmath391 , which is experimentally irreducible and whose bidegrees for @xmath392 are @xmath393 , @xmath394 , @xmath395 , @xmath396 . from these values
, it is easy to conjecture that the bidegree is given by @xmath397 of exponential growth in the bidegree of @xmath367 . in general , these bidegrees do not grow faster than in this example . in theorem [ thm : bound diagonals ] below , we prove bounds that are barely larger than the values above .
* sloped diagonals . *
if @xmath398 and @xmath399 are relatively prime positive integers and @xmath400 , then the _ sloped diagonal _ of @xmath0 , @xmath401 is @xmath402 .
direct manipulations show that @xmath403 so that our bounds and algorithm apply almost directly to these more general diagonals .
the rest of this section is devoted to the derivation of bounds on the complexity of algorithm [ algo : diagonal ] and on the size of the polynomial it computes , which are given in theorem [ thm : bound diagonals ]
. * degrees . *
a bound on the bidegree of @xmath404 will be obtained from the bounds successively given by theorems [ th : bronstein ] and [ thm : platypus - bound ] . in order to follow the impact of the change of variables in the first step , we define the _ lower diagonal degree _ and _ upper diagonal degree _ of a polynomial @xmath405 respectively as the integers @xmath406 we collect the properties of interest in the following .
[ lemma : slope ] for any @xmath16 and @xmath44 in @xmath17 $ ] , @xmath407 and @xmath408 ; @xmath409 ; there exists a polynomial @xmath410 $ ] , such that + @xmath411 , with @xmath412 and @xmath413 @xmath414 .
part ( _ 1 ) _ is immediate .
the quantities @xmath415 and @xmath416 are nothing else than @xmath417 and @xmath418 , which makes parts ( _ 2 ) _ and ( _ 3 ) _ clear too . from there , we get the identity @xmath419 for arbitrary @xmath16 and @xmath44 , whence @xmath420 and part ( _ 4 ) _ is a consequence of parts ( _ 1 ) _ and ( _ 3)_. thus , starting with a rational function @xmath421 , with @xmath232 a bound on the bidegrees of @xmath55 and @xmath5 , and @xmath162 a bound on the bidegree of a square - free part @xmath422 of @xmath5 , the first step of the algorithm constructs @xmath423 , with polynomials @xmath16 and @xmath44 and @xmath424 we first explain how to compute the number @xmath293 of small branches of @xmath44 . * small branches . *
it is classical that for a polynomial @xmath425 $ ] , the number of its solutions tending to 0 can be read off its newton polygon ( see , e.g. @xcite ) .
this polygon is the lower convex hull of the union of @xmath426 for @xmath427 such that @xmath428 .
the number of solutions tending to 0 is given by the minimal @xmath19-coordinate of its leftmost points .
since the number of small branches counts only distinct solutions , it is thus given by @xmath429p^\star).\ ] ] the change of variables @xmath430 changes the coordinates of the point corresponding to @xmath431 into @xmath432 .
this transformation maps the vertices of the original newton polygon to the vertices of the newton polygon of the laurent polynomial @xmath433 . multiplying by @xmath434 yields a polynomial and shifts the newton polygon up by @xmath415 ,
thus @xmath435 the number of small branches of the polynomial @xmath44 constructed above is then given by @xmath436 * degree in @xmath437 .
* at this point , there is a slight difference between the cases @xmath438 and @xmath439 .
indeed , in the latter case we have to take the additional small branch at @xmath329 into account .
to do this , we denote by @xmath440 the residue of @xmath335 at @xmath329 .
since @xmath440 is rational , we may compute a polynomial @xmath87 that vanishes only on the residues at non - zero small branches of the denominator of @xmath335 .
if @xmath441 is the polynomial produced by applying algorithm [ algo : sigma_c ] to @xmath442 , then the polynomial @xmath443 cancels @xmath1 . thus we apply algorithm [ algo : bronstein ] to @xmath444 if @xmath438 , and to @xmath445 otherwise . by theorem
[ th : bronstein ] , in both cases we obtain a polynomial @xmath87 of degree @xmath446 , with @xmath447 and applying algorithm [ algo : sigma_c ] gives a polynomial @xmath404 with @xmath448 .
* degree in @xmath370 .
* to bound the degree of @xmath404 in @xmath370 , we can neglect our optimization and apply algorithm [ algo : bronstein ] to @xmath449 or @xmath450 depending on wether @xmath438 or @xmath439 .
indeed , the polynomial @xmath404 obtained this way is clearly a multiple of the one computed by the algorithm . by theorem
[ th : bronstein ] , since the bidegrees of @xmath16 , @xmath451 , @xmath44 and @xmath452 are all bounded by @xmath453 , we compute a polynomial @xmath87 of degree bounded by @xmath454 , where @xmath455 applying theorem [ thm : platypus - bound ] to @xmath442 or @xmath456 depending on the sign of @xmath114 yields in both cases @xmath457 .
* complexity .
* we now analyze the cost of algorithm [ algo : diagonal ] .
the computation of @xmath16 and @xmath44 does not require any arithmetic operation .
next , the computation of @xmath87 and @xmath440 takes @xmath458 ops .
( see the comment after theorem [ th : bronstein ] ) .
the number of small branches is obtained with no arithmetic operation from a square - free decomposition computed in algorithm [ algo : bronstein ] .
the bounds of the discussion above and theorem [ thm : platypus - bound ] show that algorithm [ algo : sigma_c ] uses @xmath459 ops .
finally , if a translation of the variable is needed , it can be performed by evaluation - interpolation in @xmath460 ops .
( one may as well evaluate and interpolate wrt @xmath18 and apply better algorithms for univariate translation @xcite . )
we summarize all the results of this section in the following theorem .
[ thm : bound diagonals ] let @xmath461 be a rational function in @xmath324 with @xmath462 .
let @xmath232 ( resp .
@xmath162 ) be a bound on the bidegrees of @xmath55 and @xmath5 ( resp . a square - free part of @xmath5 ) .
let @xmath463 be defined as in eqs .
( [ eq : dx],[eq : dy],[eq : nb_small ] ) . then there exists a polynomial @xmath374 $ ] such that @xmath464 and @xmath465 algorithm [ algo : diagonal ] computes it in @xmath466 ops .
a general bound on @xmath467 depending only on a bound @xmath468 on the bidegree of the input can be deduced from the above as @xmath469 assume that the denominator of @xmath470 is already partially factored as @xmath471 , where the @xmath339 s are @xmath472 distinct _ rational _ branches among the @xmath293 small branches of @xmath44 .
then their corresponding ( rational ) residues @xmath473 contribute to the diagonal .
the special case where @xmath474 and @xmath475 is exactly the situation that occurred in the discussion on @xmath476 before theorem [ thm : bound diagonals ] , when @xmath439 .
the trick that we used extends directly to the general case : it suffices to apply algorithm [ algo : bronstein ] to @xmath477 , algorithm [ algo : sigma_c ] with @xmath478 roots , and @xmath404 is then recovered through a change of variable . the bounds from theorem [ thm : bound diagonals ] on the bidegree of @xmath404 are slightly pessimistic wrt the variable @xmath370 , but generically tight wrt the variable @xmath437 , as will be proved in proposition [ prop : generic ] below .
we first need a lemma .
[ lemma : galois groups ] let @xmath9 be a field of characteristic @xmath329 , and @xmath479 $ ] be a polynomial of degree @xmath63 , with galois group @xmath480 over @xmath9 .
assume that the roots @xmath260 of @xmath16 are algebraically independent over @xmath481 .
then , for any @xmath266 , the degree @xmath262 polynomial @xmath264 is irreducible in @xmath92 $ ] .
since @xmath482 is a root of @xmath267 , it suffices to prove that @xmath483 has degree @xmath262 over @xmath9 .
the @xmath309 s being algebraically independent , any permutation @xmath484 of all the @xmath309 s that leaves @xmath485 unchanged has to preserve the sets @xmath486 and @xmath487 .
conversely , any such permutation induces an automorphism of @xmath488 that leaves @xmath485 invariant . in other words ,
the galois group of @xmath488 over @xmath483 is equal to @xmath489 it follows that @xmath488 has degree @xmath490 over @xmath483 and degree @xmath491 over @xmath9 , so that @xmath483 has degree @xmath262 over @xmath9 .
[ prop : generic ] let @xmath55 be a polynomial in @xmath492 $ ] . let @xmath493 be non - negative integers , @xmath494 , @xmath495 , and @xmath496,\ ] ] where the @xmath497 and @xmath498 are indeterminates and @xmath499 .
then the polynomial computed by algorithm [ algo : diagonal ] with input @xmath500 is irreducible of degree @xmath501 over @xmath502 $ ] .
first apply the change of variables to obtain @xmath503 , with @xmath504 denote @xmath505 .
then , the polynomial @xmath506 has the form @xmath507 where each of the @xmath508 s is the sum of one of the indeterminates and rational constants .
this implies that the @xmath508 s are algebraically independent over @xmath481 .
therefore , @xmath506 has galois group @xmath509 over @xmath510 and its roots are algebraically independent over @xmath481 @xcite .
this property lifts to @xmath78 @xcite , which thus has galois group @xmath480 and algebraically independent roots , denoted @xmath511 .
now define the polynomial @xmath512 , where @xmath513 if @xmath438 and @xmath514 otherwise . since @xmath44 has simple roots , this is exactly the polynomial that is computed by algorithm [ algo : bronstein ] .
the family @xmath515 is algebraically independent , since any algebraic relation between them would induce one for the @xmath339 s by clearing out denominators .
in particular , the natural morphism @xmath516 is injective , whence an isomorphism .
( here , @xmath517 denotes the galois group of @xmath259 $ ] over @xmath9 . ) since an immediate investigation of the newton polygon of @xmath44 shows that it has @xmath518 small branches , we conclude using lemma [ lemma : galois groups ] and the fact that the translation of the variable does nt change the irreducible character of @xmath404 . proposition [ prop : generic ] should be viewed as an optimality result .
indeed , for a generic rational function @xmath500 as in the proposition , we have @xmath519 , @xmath520 , @xmath521 and @xmath5 has @xmath518 small branches .
this implies that the bound of theorem [ thm : bound diagonals ] for @xmath522 is optimal in this ( generic ) case .
if one believes that random examples should behave like the generic case , then the proposition means that the polynomial computed by algorithm [ algo : diagonal ] will be irreducible most of the time . as an example
, we consider the special case of proposition [ prop : generic ] where @xmath523 . in this case
, @xmath522 is @xmath524 .
we compare this to the following experiment on random examples .
we consider a rational function @xmath525 , where @xmath526 is a dense polynomial of bidegree @xmath468 chosen at random .
for @xmath392 , algorithm * algebraicdiagonal*(@xmath0 ) produces _ irreducible _ outputs with bidegrees @xmath527 , @xmath528 , @xmath529 , @xmath530 , that are matched by the formulas @xmath531 so that the bound on @xmath522 is tight in this case and the irreducibility of the output shows that theorem [ thm : bound diagonals ] can not be improved further .
the key ingredient in the fact that diagonals may have a big minimal polynomial was the possibility to write them as a sum of residues . the same exponential growth as in proposition [ prop : generic ] therefore occurs for other functions bearing this same structure . for instance , constant terms of rational functions in @xmath532 $ ] can also be written as contour integrals of rational functions around the origin and thus by the residue theorem be expressed as a sum of residues .
by contrast , such sums of residues of rational functions always satisfy a differential equation of only polynomial size @xcite .
thus , when an algebraic function appears to be connected to a sum of residues of a rational function , the use of this differential structure is much more adapted to the computation of series expansions , instead of going through a potentially large polynomial .
as an example where this phenomenon occurs naturally , we consider here the enumeration of unidimensional lattice walks , following banderier and flajolet @xcite and bousquet - mlou @xcite .
our goal in this section is to study , from the algorithmic perspective , the series expansions of various generating functions ( for bridges , excursions , meanders ) that have been identified as algebraic @xcite .
one of our contributions is to point out that although algebraic series can be expanded fast @xcite , the precomputation of a polynomial equation could have prohibitive cost .
we overcome this difficulty by precomputing differential ( instead of polynomial ) equations that have polynomial size only , and using them to compute series expansions to precision @xmath2 for bridges , excursions and meanders in time quasi - linear in @xmath2 .
we start with some vocabulary on lattice walks .
a _ simple step _ is a vector @xmath533 with @xmath534 .
@xmath275 is a finite set of simple steps .
a _ unidimensional walk _ in the plane @xmath535 built from @xmath275 is a finite sequence @xmath536 of points in @xmath535 , such that @xmath537 and @xmath538 with @xmath539 . in this case
@xmath12 is called the _ length _ of the walk , and @xmath275 is the _ step set _ of the walk .
the @xmath19-coordinate of the endpoint @xmath540 , namely @xmath541 , is called the final altitude of the walk .
the characteristic polynomial of the step set @xmath275 is @xmath542 ' '' ''
-10pt following banderier and flajolet , we consider three specific families of walks : bridges , excursions and meanders @xcite .
_ bridges _ are walks with final altitude @xmath329 , _ meanders _ are walks confined to the upper half plane , and _ excursions _ are bridges that are also meanders .
figure [ fig : bafl02 ] , taken from @xcite , summarizes these definitions graphically .
we define the full generating power series of walks @xmath543[[x]],\ ] ] where @xmath544 is the number of walks with step set @xmath275 , of length @xmath12 and final altitude @xmath472 .
we denote by @xmath545 ( resp .
@xmath546 , and @xmath547 ) the power series @xmath548 , where @xmath549 is the number of bridges ( resp .
excursions , and meanders ) of length @xmath12 with step set @xmath275 .
we omit the step set @xmath275 as a subscript when there is no ambiguity .
several properties of the power series @xmath550 , @xmath5 , @xmath6 and @xmath7 are classical : @xcite[fact : walks ] the power series @xmath550 , @xmath5 , @xmath6 and @xmath7 satisfy @xmath551 is rational and @xmath552 ; @xmath553 , @xmath554 and @xmath555 are algebraic ; @xmath556w(x , y)$ ] ; @xmath557 . in
what follows , we describe and analyze three methods to compute the power series expansions of @xmath5 , @xmath6 and @xmath7 . in the next two sections , we first study two previously known methods
, then we introduce a new one . from now on ,
we fix a step set @xmath275 , and we denote by @xmath558 ( resp .
@xmath559 ) the largest @xmath560 such that @xmath561 ( resp .
@xmath562 ) .
we also define @xmath563 .
the integer @xmath63 measures the vertical amplitude of @xmath275 ; this makes @xmath63 a good scale for measuring the complexity of the algorithms that will follow .
we assume that both @xmath558 and @xmath559 are positive , since otherwise the study of the bridges , excursions and meanders becomes trivial . * the direct method . *
the combinatorial definition of walks yields a recurrence relation for @xmath544 : @xmath564 with initial conditions @xmath565 if @xmath566 with @xmath567 , and @xmath568 .
if @xmath569 denotes the number of walks of length @xmath12 and final altitude @xmath472 that never exit the upper half plane , then @xmath569 also satisfies recurrence , but with the additional initial conditions @xmath570 for all @xmath571 .
then the bridges ( resp . excursions , meanders ) are counted by the numbers @xmath572 ( resp .
@xmath573 , @xmath574 ) .
one can compute these numbers by unrolling the recurrence relation .
each use of the recurrence costs @xmath575 ops . , and in the worst case one has to compute @xmath576 terms of the sequence ( for example , if the step set is @xmath577 ) .
this leads to the computation of each of the generating series in @xmath578 ops .
this quadratic complexity in @xmath2 is unsatisfactory , and any method that requires the complete expansion of the generating series @xmath551 is bound to be quadratic in @xmath2 .
the two other methods that we are going to present are designed to achieve linear or quasi - linear complexity in @xmath2 . as will be explained ,
this comes at the cost of a precomputation that must be taken into account in the analysis . * using algebraic equations . * in @xcite , a method relying on the algebraicity of @xmath5 , @xmath6 and @xmath7 ( fact [ fact : walks]_(2 ) _ ) ) is suggested .
the series @xmath6 and @xmath7 can be expressed as products in terms of the small branches of the characteristic polynomial @xmath579 ( see ( * ? ? ?
1 ) ) . from there
, a polynomial equation can be obtained using the platypus algorithm @xcite , which computes a polynomial canceling the products of a fixed number of roots of a given polynomial .
given a polynomial equation @xmath580 , another one for @xmath5 can be deduced from the relation @xmath581 as @xmath582 .
once a polynomial equation is known for one of these three series , it can be used to compute a linear recurrence with polynomial coefficients satisfied by its coefficients @xcite . the naive algorithm introduced above
provides a way to compute a sufficiently large number of initial conditions to unroll this recurrence .
( for a quantitative result on the required number of initial conditions , see corollary [ coro : expand algebraic functions ] below . )
this method produces an algorithm that computes the first @xmath2 terms of @xmath5 , @xmath6 and @xmath7 in @xmath583 ops . for this to be an improvement over the naive method for large @xmath2
, the dependence on @xmath63 of the constant in the @xmath584 should not be too large and the precomputation not too costly . indeed , the cost of the precomputation of an algebraic equation is not negligible .
the bound @xmath585 on the degrees of equations for excursions has been obtained by bousquet - mlou , and showed to be tight for a specific family of step sets , as well as generically @xcite .
this bound may be exponentially large with respect to @xmath63 .
empirically , the polynomials for @xmath5 and @xmath7 are similarly large .
the situation for differential equations and recurrences is different : @xmath5 satisfies a differential equation of only polynomial size ( see below ) , whereas ( empirically ) , those for @xmath6 and @xmath7 have a potentially exponential size .
these sizes then transfer to the corresponding recurrences and thereby to the constant in the complexity of unrolling them .
the purpose of theorem [ thm : walks ] below is to give explicitly the polynomial dependence in @xmath63 when using this method , showing at the same time that a true improvement over the naive method can be achieved . with the step set @xmath586 and @xmath587
, the counting series @xmath588 equals @xmath589 experiments indicate that the minimal polynomial of @xmath545 has bidegree @xmath590 , exhibiting an exponential growth in @xmath63 . on the other hand , they show that @xmath545 satisfies a linear differential equation of order @xmath591 and coefficients of degree @xmath592 for even @xmath63 , and @xmath593 for odd @xmath63 .
* new method .
* we now give a method that runs in quasi - linear time ( with respect to @xmath2 ) and avoids the computation of an algebraic equation .
our method relies on the fact that periods of rational functions such as the one in part _ ( 3 ) _ of fact [ fact : walks ] satisfy differential equations of polynomial size in the degree of the input rational function @xcite .
we summarize our results in the following theorem , and then go over the proof in each case individually .
[ thm : walks ] let @xmath275 be a finite set of simple steps and @xmath594 .
the series @xmath595 ( resp . @xmath596 and @xmath597 ) can be expanded at order @xmath2 in @xmath598 ops .
( resp . @xmath599 ops . ) , after a precomputation in @xmath600 ops .
* bridges .
* to expand @xmath553 , we rely on fact [ fact : walks]_(3)_. the formula can be written @xmath601 , the integration path being a circle inside a small annulus around the origin ( * ? ? ?
* proof of th . 1 ) .
moreover , @xmath602 is of the form @xmath99 , where @xmath603 and @xmath604 .
since @xmath16 and @xmath44 are relatively prime and @xmath44 is primitive with respect to @xmath19 , algorithm * hermitetelescoping * ( * ? ? ?
3 ) computes a telescoper for @xmath99 , which is also a differential equation satisfied by @xmath5 .
by fact [ fact : sqfree]_(2 ) _ , the resulting differential equation has order at most @xmath63 and degree @xmath605 , and is computed using @xmath600 ops .
this differential equation can be turned into a recurrence of order @xmath606 in quasi - optimal time ( see the discussion after ( * ? ? ?
2 ) ) . we may use it to expand @xmath607 in @xmath608 ops , once enough initial conditions are known .
again , the initial conditions are computed by means of the direct method .
the only remaining question is the number of initial conditions needed .
indeed , the recurrence may be singular , ie its leading coefficient may have positive integer roots .
if we denote by @xmath114 the largest such root , then we need to compute the first terms of the recurrence up to @xmath609 . in order not to break the flow of reading
, we postpone the discussion on the size of @xmath114 to the next section . for now , we only state the result . [
prop : initial conditions ] let @xmath275 be a set of simple steps , and @xmath610 .
then the largest integer root of the leading term of the recurrence computed by algorithm [ algo : bridges ] is at most @xmath611 see section @xmath612 .
thus , a sufficient number of initial conditions is computed with @xmath613 ops by the direct method , and the total cost of the precomputation is @xmath600 , as announced .
algorithm * walks*(@xmath275 , @xmath2 ) a set @xmath275 of simple steps and an integer @xmath2@xmath614 @xmath615 [ case @xmath616 or @xmath617 [ case @xmath7 ] @xmath619 ( * ? ? ? * fig .
3 ) @xmath620 the recurrence of order @xmath440 associated to @xmath91 @xmath621w(x , y)\bmod x^{r+1}$ ] [ case @xmath622 + @xmath623yw(x , y)/(1-y)\bmod x^{r+1}$ ] [ case @xmath7 ] @xmath624w(x , y)\bmod x^{n+1}$ ] ( from @xmath625 ) @xmath626yw(x , y)/(1-y)\bmod x^{n+1}$ ] ( from @xmath625 ) @xmath627 @xmath628 @xmath629 * excursions . * if @xmath630 is known , it is then possible to recover @xmath631 thanks to fact [ fact : walks]_(4)_. expanding @xmath554 comes down to the computation of the exponential of a series , which can be performed using @xmath632 ops .
( fact [ fact : complexity]_(4 ) _ ) .
* meanders . * as in the case of excursions , the logarithmic derivative of @xmath555 is recovered from a sum of residues by the following .
[ walks - formulas2 ] the series @xmath550 and @xmath7 are related through @xmath633\frac{y}{1-y}w(x , y),\quad m(x ) = \frac{\exp\left(-\int\frac{a(x)}{x}\,\mathrm{d}x\right)}{1-x\gamma(1)}.\ ] ] denote by @xmath634 the small branches of the polynomial @xmath635 .
then @xmath7 is given as ( * ? ? ?
1 ) : @xmath636 on the other hand , @xmath637 where the integral has been taken over a circle around the origin and the small branches . differentiating the equation @xmath638 = 0 with respect to @xmath18 leads to @xmath639 , whence @xmath640 therefore , @xmath641 , finishing the proof .
thus we apply the same method as in the case of the excursions .
we first compute a differential equation for @xmath642 using the method of @xcite .
the computation of the initial conditions for @xmath55 can also be performed naively from its definition as a constant term , by simply expanding @xmath643 .
the formula of the proposition then recovers @xmath555 .
the complexity analysis goes exactly as in the previous case , giving a global cost of @xmath600 ops .
we now come back to the problem of singular recurrences . in our context , the recurrences that we come across have a very specific structure : they are associated to differential resolvents of polynomials .
( the differential resolvent of a polynomial is the least order differential operator canceling all of its roots . )
this structure can be exploited to derive bounds on the singularities of our recurrences .
if @xmath30[y]$ ] is a polynomial , consider the recurrence associated to its differential resolvent @xmath252 .
the leading coefficient of this recurrence is called the indicial polynomial of @xmath252 at 0 .
its largest integer root will be denoted @xmath114 .
the fundamental idea is that there exists a laurent series solution of @xmath252 which has valuation @xmath114 @xcite therefore , it is sufficient to find bounds on the valuations of the solutions of @xmath252 .
this is done in the following theorem .
[ th : valuations ] let @xmath16 be a polynomial in @xmath11[y]$ ] , of bidegree at most @xmath232 , and @xmath252 be the differential resolvent of @xmath16 . then all the laurent series solutions @xmath644 of @xmath252 uniformly satisfy @xmath645 choose a subfamily @xmath646 of the puiseux series roots of @xmath16 that constitutes a basis of the solution space of the resolvent ( in particular , @xmath647 ) .
let @xmath648 be a laurent series solution of the differential resolvent of @xmath16 .
then the fact that @xmath649 for any laurent series @xmath650 $ ] implies that @xmath651 by the multilinearity of the wronskian , the left - hand side of this inequality is nothing more than @xmath652 . on the other hand , the absolute values of the valuations of the @xmath339 s
are bounded by @xmath653 ( because they are slopes of edges in the newton polygon of @xmath16 ) .
a bound for @xmath654 is thus obtained : @xmath655 the proof is then reduced to showing that @xmath656 .
this is very similar to the computations conducted in @xcite .
we start by recalling some facts that are proved there .
there exist polynomials @xmath657 $ ] such that for all @xmath337 and all @xmath658 , the derivative @xmath659 can be expressed as @xmath660 moreover , the polynomials @xmath661 satisfy @xmath662 it follows that @xmath663 $ ] is a polynomial such that @xmath664 $ ] .
we will denote by @xmath87 this last polynomial .
@xmath87 is the determinant of the matrix @xmath665 @xmath87 is an anti - symmetric polynomial in @xmath646 , but @xmath666 is symmetric , as well as @xmath91 , so we can apply lemma [ lem : symmetric functions ] to see that @xmath666 and @xmath91 belong to @xmath14 .
therefore , the equality @xmath667 shows that @xmath668 is the square root of a rational function in @xmath18 .
we are going to use this structure and lemma [ lem : symmetric functions ] to derive the desired bound on the valuation of the wronskian determinant .
if @xmath669 is viewed as a polynomial in @xmath670 $ ] , then @xmath671 and for all @xmath337 , @xmath672 similarly , when @xmath91 is viewed as a polynomial in @xmath670 $ ] , we have : @xmath673 applying lemma [ lem : symmetric functions ] , we deduce that , denoting by @xmath674 the leading coefficient of @xmath77 , @xmath675 where @xmath676 finally , the inequalities @xmath677 and @xmath647 yield @xmath678 which concludes the proof .
we immediately deduce the following corollary on the number of initial conditions required to expand an algebraic power series .
[ coro : expand algebraic functions ] let @xmath679 $ ] be a polynomial of bidegree bounded by @xmath232 .
let @xmath87 be the recurrence associated to the differential resolvent of @xmath16 .
then the largest integer root of the leading coefficient of @xmath87 is at most @xmath680 .
immediate from the theorem and the discussion that precedes it .
we are now able to prove proposition [ prop : initial conditions ] .
( of proposition [ prop : initial conditions ] ) we only treat the case where the recurrence is computed for @xmath5 , and the proof transposes directly to the case of @xmath55 .
let @xmath275 and @xmath63 be as in the proposition , and denote by @xmath16 the minimal polynomial of @xmath5 .
then the recurrence computed by algorithm [ algo : bridges ] is associated to the minimal annihilating differential operator for @xmath5 , which is also the differential resolvent of @xmath16 .
we denote it by @xmath681 .
now since @xmath682w(x , y)/y$ ] , it can be written as a sum of residues similar to formula ( [ eq : diagonal - as - residues ] ) .
if we denote by @xmath87 the polynomial that cancels these residues , then @xmath16 divides @xmath683 for some @xmath293 .
this implies in particular that all the solutions of @xmath681 are linear combinations of the roots of @xmath87 .
thus , if @xmath684 is the differential resolvent of @xmath87 , then all the solutions of @xmath681 are solutions of @xmath684 .
since @xmath550 has bidegree @xmath685 , theorem [ th : bronstein ] and theorem [ th : valuations ] show that all the roots of @xmath16 have valuation at most @xmath611 , and the result follows .
we gave a complete and efficient algorithm that calculates a polynomial equation satisfied by the diagonal of a bivariate rational function in characteristic @xmath329 .
generically , the degree in @xmath437 of the polynomial @xmath686 output by the algorithm is optimal .
the bound on the degree in @xmath370 is not tight . the gap between this bound and the actual degrees is not yet fully understood : it is already present for the rothstein - trager and bronstein resultants .
our complexity results are given in the arithmetic complexity model .
the corresponding study in the binary model remains to be done .
the case of positive characteristic requires different methods and algorithms . in that case , diagonals are algebraic even for rational functions with more than two variables . to the best of our knowledge , these questions have never been studied from the complexity viewpoint .
one possible direction is to try and make effective the proof by furstenberg that these diagonals are algebraic @xcite .
some work has also been done by adamczewski and bell @xcite who among other things studied how the sizes of the polynomial equations satisfied by diagonals vary with the characteristic of the base field . | the diagonal of a multivariate power series @xmath0 is the univariate power series @xmath1 generated by the diagonal terms of @xmath0 .
diagonals form an important class of power series ; they occur frequently in number theory , theoretical physics and enumerative combinatorics .
we study algorithmic questions related to diagonals in the case where @xmath0 is the taylor expansion of a bivariate rational function .
it is classical that in this case @xmath1 is an algebraic function .
we propose an algorithm that computes an annihilating polynomial for @xmath1 .
we give a precise bound on the size of this polynomial and show that generically , this polynomial is the minimal polynomial and that its size reaches the bound .
the algorithm runs in time quasi - linear in this bound , which grows exponentially with the degree of the input rational function .
we then address the related problem of enumerating directed lattice walks .
the insight given by our study leads to a new method for expanding the generating power series of bridges , excursions and meanders .
we show that their first @xmath2 terms can be computed in quasi - linear complexity in @xmath2 , without first computing a very large polynomial equation . |
since a long time electromagnetic two - nucleon knockout reactions have been devised as a preferential tool to investigate two - body correlations in nuclei @xcite . indeed , the probability that a real or virtual photon is absorbed by a pair of nucleons should be a direct measurement of the correlations between the two nucleons .
for an exclusive reaction the transitions amplitudes contain the two - nucleon overlap function ( tof ) between the ground state of the target and the final state of the residual nucleus . in the cross section
the tof gives the two - hole spectral density function , _
i.e. _ @xmath1 which in its diagonal form ( @xmath2 and @xmath3 ) gives the joint probability of removing from the target two nucleons , with momenta @xmath4 and @xmath5 , leaving the residual nucleus in a state with energy @xmath6 with respect to the target ground state .
in an inclusive reaction , integrating the spectral density function over the whole energy spectrum produces the two - body density matrix @xmath7 , that in its diagonal form and in the coordinate representation gives the pair correlation function @xmath8 , _ i.e. _ the conditional probability density of finding in the target a particle at @xmath9 if another one is known to be at @xmath10 .
the two - nucleon overlap , the two - hole spectral function and the two - body density matrix contain information on nuclear structure and correlations .
the problem is that it is not easy to disentangle these quantities from the experimental cross sections and extract clear and unambiguous information on correlations . in order to achieve this goal , along with the experimental work ,
a reliable theoretical model is needed , where all the ingredients are well under control . by correlations we mean what goes beyond the independent - particle model , the mean - field approximation .
of particular interest are the short - range correlations ( src ) , which are produced by the short - range repulsion of the nn - interaction .
src are of particular relevance for pp - pairs and are essential for understanding the short - range and high - momentum properties of the nuclear wave functions
. also very interesting are the tensor correlations ( tc ) , which are due to the strong tensor component of the nn - interaction , that is of intermediate to long - range character .
tc are of particular relevance for pn - pairs . thus src and tc can be investigated for pp- and pn - pairs .
such an investigation would be of great interest , since the short - range repulsion and the tensor character represent the two main features of the nn - interaction . also very important are , however , the so - called long - range correlations ( lrc ) at low energy , which mainly represent the interaction of nucleons at the nuclear surface and are related to the coupling between the single - particle ( s.p . )
dynamics and the collective excitation modes of the nucleus .
src , tc , and lrc are intertwined in the spectral function and in the density matrix and it is not easy to disentangle a specific contribution . a reliable model for cross section calculations requires a careful and consistent treatment of these different types of correlations .
moreover , the cross section is sensitive to the reaction mechanism .
there are two competing processes contributing to two - nucleon knockout .
there is the process where the electromagnetic probe hits , through a one - body ( ob ) current , either nucleon of a correlated pair and both nucleons are then ejected from the nucleus .
two nucleons can not be emitted by an ob current unless they are correlated .
thus , this process is entirely due to correlations .
two nucleons , however , can also and naturally be emitted by two - body ( tb ) currents , due to meson exchanges and @xmath11-isobar excitations , even if they are not correlated .
the role and relevance of these two competing processes can be different in different situations and kinematics , but if situations can be envisaged where the ob current is dominant , these situations are very well suited to investigate correlations
. it must be noted , however , that correlations affect also the process due to the tb currents . in the model
@xcite the transition amplitudes contain three main ingredients : the nuclear current , the two - nucleon scattering wave function , and the tof , which includes correlations .
the two - nucleon wave functions for the initial and final states are consistently treated in the transition amplitudes with the ob and the tb currents .
thus , correlations affect also the contribution of the tb currents and in principle it is possible to obtain information on correlations even when the tb currents are important or dominant in the cross section . moreover , it is not correct to say that the contribution of the ob current is entirely due to correlations in the nuclear wave functions . in the calculation of the transition amplitudes , for the three - body system consisting of the two nucleons and the residual nucleus ,
it is natural to work with center - of - mass ( cm ) coordinates , and in the cm frame a tb operator is obtained in the transition amplitude even with an ob current @xcite .
the relevance of this cm effect depends on the particular situation that is considered and may be large or small , but , independently of that , due to this cm effects there is a contribution to two nucleons emission with an ob current also without correlations in the nuclear wave functions .
the main features of the theoretical approach are outlined in sec .
special emphasis is devoted to the description of the recent improvements performed in comparison with earlier work .
numerical results for the exclusive pp- and pn - knockout reactions off @xmath0o are presented and discussed in sec .
3 also in comparison with the available data .
some conclusions are drawn in sec .
the basic ingredients for calculations of the reaction induced by a real or virtual photon , with momentum , where two nucleons , with momenta @xmath12 and @xmath13 , are emitted from a nucleus , are the transition matrix elements of the charge - current density operator between initial and final nuclear states , _ i.e. _ @xmath14 bilinear products of these integrals give the components of the hadron tensor whose suitable combinations give all the observables available from the reaction process .
the model is based on the two assumptions of an exclusive reaction , where the residual nucleus is left in a discrete eigenstate of its hamiltonian , and of the direct knockout mechanism . as a results of these two assumptions , the matrix elements of eq .
( [ eq : jm ] ) are obtained in the form @xcite @xmath15 equation ( [ eq : jq ] ) contains three main ingredients : the nuclear current @xmath16 , the two - nucleon scattering wave function @xmath17 , and the tof @xmath18 . in the model @xmath18 and @xmath17
are consistently derived from an energy - dependent non - hermitian feshbach - type hamiltonian @xcite .
actually , they are eigenfunctions of this hamiltonian and of its hermitian conjugate at different energies : for an exclusive reaction , we select a channel and then project the initial and final nuclear states in eq .
( [ eq : jm ] ) onto the selected channel subspace ; as a result of the projection , we obtain the optical - model hamiltonian describing the interaction of the two nucleons and the residual nucleus in the eigenstate that we have selected @xcite .
since a fully consistent calculation of @xmath17 and @xmath18 as eigenfunctions of the optical - model hamiltonian would be very difficult , some approximations are used in actual calculations .
the nuclear current operator is the sum of an ob and a tb part .
the ob part contains the longitudinal charge term and the convection and spin currents .
the tb currents are derived from a nonrelativistic reduction of the lowest - order feynman diagrams with one - pion exchange .
therefore we have currents corresponding to the seagull and pion - in - flight diagrams , and to the diagrams with intermediate @xmath11-isobar configurations @xcite .
all these terms contribute to pn - knockout , while the seagull and the pion - in - flight meson - exchange currents ( mec ) do not contribute to pp - emission , at least in the adopted nonrelativistic limit .
the contribution of the tb currents depends on the type of reaction , on the kinematics , on the conditions of the calculations , and , of course , on the treatment of the tb currents .
the explicit expressions of the currents can be found in @xcite .
different parametrizations have been used in previous calculations , in particular for the @xmath11-current .
a fundamental treatment of the @xmath11 in electromagnetic breakup reactions on complex nuclei is presently unavailable and we can only rely on approximative schemes . in our most recent calculations @xcite , the parameters of the @xmath11-current are fixed considering the nn - scattering in the @xmath11-region , where a reasonable description of scattering data is achieved with parameters similar to the ones of the full bonn potential @xcite . for pp - knockout , besides the usual @xmath19-exchange , also @xmath20-exchange is included and the coupling constants are @xmath21 moreover , hadronic form factors are included @xmath22 with @xmath23 , and the cutoffs @xmath24 for pn - knockout , where the tb current contains many more terms , a simpler approach is adopted @xcite where only @xmath19-exchange is included and the parameters for the @xmath11-current are @xmath25 these parameters are able to give , with only @xmath19-exchange , a comparable description of the nn - scattering data . with respect to the pion - in - flight and seagull mec , contributing only to pn - knockout , a dipole cutoff is used of 3 gev , in accordance with the bonn - c potential @xcite which is used also in the calculation of the tof @xcite and of the mutual interaction between the two outgoing nucleons @xcite .
this choice of parameters for the tb currents is called in the following bonn parametrization .
the hadronic form factors defined in eq .
( [ form ] ) are necessary for regularizing the interaction at short distances , where the meson - exchange picture becomes meaningless . in previous calculations
@xcite the same coupling constants as in eq .
( [ formpn ] ) were used with a simpler regularization in coordinate space , both for the mec and the @xmath11-current , which in practice is similar to an unregularized prescription ( @xmath26 ) . in the scattering state
@xmath17 the two outgoing nucleons , 1 and 2 , and the residual nucleus interact via the potential @xmath27 where @xmath28 denotes the interaction between the nucleon @xmath29 and the residual nucleus , described in the model by a complex optical potential , and @xmath30 denotes the mutual interaction between the two outgoing nucleons .
only the contribution of the final - state interaction ( fsi ) due to @xmath28 was included in our first calculations . in this approximation ( dw )
the two - nucleon scattering wave function is given by the product of two uncoupled s.p .
distorted wave functions , eigenfunctions of a phenomenological optical potential fitted to nucleon - nucleus scattering data @xcite .
the contribution of @xmath30 ( nn - fsi ) has been studied within a perturbative approach @xcite .
the main contribution of fsi is given in general by the optical potential , which produces an overall and substantial reduction of the calculated cross sections .
this effect is important and can never be neglected .
nn - fsi gives in general an enhancement of the cross section that depends strongly on the kinematics , on the type of reaction , and on the final state of the residual nucleus .
it is generally larger in pp- than in pn - knockout and in electro- than in photoinduced reactions @xcite . in
many situations the difference between the results obtained in the dw approach and in the more complete approach where also nn - fsi is included ( dw - nn ) is small .
there are , however , also situations where this difference is large .
numerical examples are presented in sec .
the tof @xmath18 contains information on nuclear structure and correlations and represents the most interesting ingredient of the model .
different approaches have been used @xcite to calculate @xmath18 in pp- and pn - knockout from @xmath0o .
@xmath0o is a suitable target for this study , due to the presence of discrete final states in the excitation - energy spectrum of the residual nucleus , both for @xmath31c and @xmath31n , states well separated in energy and that can be separated in experiments with good energy resolution .
experimental data are available for both @xmath0o(e , e@xmath32pp ) @xcite and @xmath0o ( e , e@xmath32pn ) @xcite reactions . in a simpler approach @xcite the tof is given by the product of a coupled and antisymmetrized shell - model ( sm ) pair function and a jastrow - type central and state independent correlation function taken from @xcite . in this approach ( sm - src ) only src are included in the correlation function and the final state of the residual nucleus is a pure two - hole state in the target .
for instance , the ground state of @xmath31c is a ( @xmath33)@xmath34 hole in @xmath0o . in the more sophisticated approaches
@xcite and @xcite , the tof is obtained from the the first calculations of the two - nucleon spectral function of @xmath0o , where src , tc and lrc are included with some approximations but consistently . in both calculations the tof s for low - lying states of the residual nucleus are obtained partitioning the hilbert space onto two subspaces where src and lrc are separately calculated .
lrc and the long - range part of tc are calculated using the self - consistent green s function formalism @xcite in an appropriate harmonic - oscillator ( h.o . ) basis , large enough to account for the main collective features that influence the pair removal amplitudes .
the effects of src due to the central and tensor part at high momenta are included by defect functions , which are solutions of a bethe - goldstone equation where the pauli operator considers only configurations outside the model space where lrc are calculated .
different defect functions are obtained for different relative states and the bonn c nn - potential @xcite is used in the calculations .
the tof is given by a combination of components of relative and cm motion , that are different for different final states , where the coefficients are the two - nucleon removal amplitudes which include lrc , and the defect functions , which contain src , are intertwined in a complicated manner . in the more recent calculation @xcite
some improvements have been included with respect to the previous work @xcite : i ) the non - locality of the pauli operator is computed exactly , resulting in a larger number of defect functions with a more complicate state dependence ; ii ) the evaluation of nuclear structure effects related to the fragmentation of the s.p . strength has been improved by applying a faddeev technique to the description of the internal propagators in the nucleon self - energy . moreover ,
both pp- and pn - pairs are calculated in @xcite , while only the pp - case is considered in @xcite .
the numerical results presented in this contribution are obtained with the more recent tof s from @xcite ( sf - b ) .
in @xcite the defect functions for pn - knockout off @xmath0o are calculated within the framework of the coupled cluster ( cc ) method , using the so - called @xmath35 approximation , and employing the argonne v14 nn - potential @xcite .
also in this calculation ( sf - cc ) the defect functions include src and tc and have a complicate state dependence .
lrc , however , are accounted for in the tof in a simpler way and only knockout of nucleons from the @xmath36 shell is considered .
an alternative procedure is proposed in @xcite , where the tof is obtained from the asympotic behavior of the two - body density matrix ( tdm ) .
this procedure avoids the calculation of the spectral function , but depends strongly and relies on the availability of reliable calculations of the tdm .
the applicability of the procedure is shown in @xcite , where the tof obtained from a simple tdm , which is calculated within the jastrow correlation method @xcite ( jcm ) and incorporates only src , is able to give numerical results that under many aspects are qualitatively similar to the results produced by more sophisticated approaches .
the tdm of @xcite is at present the only tdm available for these calculations
. it would be interesting to check the procedure with a more complete tdm . in the calculation of the transition amplitude of eq .
( [ eq : jq ] ) , for the three - body system consisting of the two nucleons , 1 and 2 , and of the residual nucleus @xmath37 , it is natural to work with cm coordinates @xcite @xmath38 the conjugated momenta are given by @xmath39 where @xmath40 is the momentum of the residual nucleus in the laboratory frame . with the help of these relations , the transition amplitude of eq .
( [ eq : jq ] ) is obtained in the following form @xcite @xmath41 where for the ob current @xmath42 @xmath43 similar expressions are obtained for the tb currents @xcite . in spite of the fact that an ob operator can not affect two particles if they are not correlated , it can be seen from eqs .
( [ eq : jqcm ] ) and ( [ eq : opcm ] ) that in the cm frame the transition operator becomes a tb operator even in the case of an ob nuclear current . only in the limit @xmath44 ,
cm effects vanish and the expression in eq .
( [ eq : jq ] ) becomes zero for a pure ob current when the matrix element is calculated using orthogonalized s.p .
wave functions .
this means that , due to this cm effect , for finite nuclei the ob current can give a contribution to the cross section of two - particle emission even without correlations in the nuclear wave functions @xcite .
independently of the specific prescriptions adopted in the calculations , a conceptual problem arises in the model , where the initial and final states , @xmath18 and @xmath17 , which are eigenfunctions of an energy - dependent optical - model hamiltonian , are , as such , not orthogonal .
indeed , the process involves transitions between bound and continuum states which must be orthogonal , since they are eigenfunctions of the full nuclear many - body hamiltionian .
orthogonality is in general lost in a model when the description is restricted to a subspace where other channels are suppressed .
the description of direct knockout reactions in terms of the eigenfunctions of a complex energy - dependent optical potential considers only partially the contribution of competing inelastic channels .
a consequence of the lack of orthogonality between the initial and the final state is that eq .
( [ eq : jqcm ] ) may contain a spurious contribution , since it does not vanish when the transition operator @xmath45 is set equal to 1 .
the use of an effective nuclear current operator would remove the orthogonality defect besides taking into account space truncation effects @xcite . in the usual approach , however , the replacement of the effective operator by the bare nuclear current operator may introduce a spurious contribution . in our earlier work @xcite the spuriosity
was removed subtracting from the transition amplitude the contribution of the ob current without correlations in the nuclear wave functions .
this prescription , however , subtracts , together with the spuriosity , also the cm effect given by the tb operator in eq .
( [ eq : opcm ] ) , that is present in the ob current independently of correlations in the nuclear wave function , and that is not spurious .
moreover , the definition of a nuclear wave function without correlations is clear in the simpler approach sm - src , where the uncorrelated sm pair function is multiplied by a correlation function , but it is less clear in a more sophisticated approach , like sf - b , where different types of correlations are intertwined in the tof . in this case
the prescription is applied subtracting the part of the tof without the defect functions .
the defect functions , however , include only src , while different correlations are intertwined in the remaining part of the tof that is subtracted .
therefore , with this prescription also part of the contribution of correlations is subtracted .
a more accurate procedure to get rid of the spuriosity is adopted in @xcite , where orthogonality between the initial and final states is enforced by means of a gram - schmidt orthogonalization @xcite . in this approach
each one of the two s.p .
distorted scattering wave functions is orthogonalized to all the s.p .
sm wave functions that are used to calculate the tof , _
i.e. _ , for the tof of @xcite , to the h.o
. states of the basis used in the calculation of the spectral function , which range from the @xmath46 up to the @xmath47-@xmath48 shell , or , for the simpler sm - src approach , to the s.p .
states used to calculate the sm pair function . the procedure to enforce orthogonality between s.p .
bound and scattering states is done accordingly with the definition of the spurious contribution , that in the limit @xmath44 , where cm effects are neglected , the transition amplitude vanishes for a pure ob current sandwiched between orthogonalized s.p .
wave functions .
the orthogonalization procedure adopted in our recent work @xcite removes the spurious contribution and includes all the cm and correlation effects that were subtracted with the previous prescription and that are not spurious .
the relevance of these effects depends on the type of reaction , on the kinematics , and on the treatment of the theoretical ingredients of the model @xcite . in many situations the differences between the results obtained with the two procedures are small .
there are , however , situations where these differences are large , and they are very large in the super - parallel kinematics .
this is a kinematical setting of particular interest , since the recent @xmath0o(e , e@xmath32pp)@xmath31c @xcite and @xmath0o(e , e@xmath32pn)@xmath31n @xcite experiments carried out at mami are both centred on the same super - parallel kinematics .
the super - parallel kinematics was originally proposed @xcite since it is particularly favourable to emphasize short - range effects and has been widely investigated in our work .
recent results @xcite have shown that this kinematic setting is indeed very sensitive to correlations , but it is also very sensitive to all the other ingredients of the model .
such a great sensitivity makes the super - parallel kinematics very interesting although not very suitable to isolate the specific contribution of correlations from other competing contributions .
the achievement of this goal requires experiments in a wider range of kinematics which mutually supplement each other .
these arguments are illustrated in the next section with some numerical examples .
in the super - parallel kinematics the two nucleons are ejected parallel and anti - parallel to @xmath49 and , for a fixed value of @xmath50 and of the energy transfer @xmath51 , it is possible to explore , for different values of the kinetic energies of the outgoing nucleons , all possible values of the recoil - momentum @xmath52 . in fig .
[ fig : fig1 ] the differential cross sections of the @xmath0o(e , e@xmath32pp)@xmath31c reaction are displayed , as a function of @xmath52 , for transitions to different states of @xmath31c : the @xmath53 ground state , the @xmath54 state at 11.31 mev , and the @xmath55 state at 7.67 mev .
the calculations have been performed in the super - parallel kinematics of the mami experiment @xcite , with the sf - b overlap function and the dw approximation for fsi .
different shapes are obtained for the cross sections to the three final states .
the shape of the recoil - momentum distribution is basically driven by the cm orbital angular momentum @xmath56 of the knocked out pair .
different partial waves of relative and cm motion contribute to the tof .
each transition is characterized by different components , with specific values of @xmath56 .
the relative weights of these components , which are given by the two - nucleon removal amplitudes included in the tof , determine the shape of the momentum distribution @xcite .
the results of the orthogonalized approach and of the previous prescription to remove the spuriosity are compared in the figure .
the difference depends on the final state . for the @xmath53 ground state
the orthogonalized approach gives a large enhancement of the cross section calculated with the ob current .
the result is shown in the right panel , where it can be seen that the enhancement is large for values of @xmath52 up to about 300 mev/@xmath57 , and is a factor of about 5 in the maximum region .
it is shown in @xcite that the enhancement is mostly due to the cm effects included in the new calculations and not to the different treatment of the spuriosity or to the restoration of orthogonality between the initial and final - state wave functions .
the difference between the results of the two approaches is only slightly reduced in the final cross sections , shown in the left panel , where also the @xmath11-current is included .
the cross section is dominated by the ob current and only a minor role is played by the @xmath11-current . as a consequence , also
the difference between the results obtained in the orthogonalized approach with the bonn and the unregularized parametrizations , that are also shown in the figure , is small .
a small reduction of the cross section is given by the bonn parametrization .
for the @xmath54 state the difference between the results of the orthogonalized approach and the previous results is practically negligible . in this case
the @xmath11-current gives a significant enhancement of the ob cross section , which is larger with the unregularized than with the bonn parametrization .
for the @xmath55 state the orthogonalized approach gives a significant enhancement of the cross section at low values of @xmath52 and a negligible effect at higher momenta .
the main contribution to the cross section is given by the ob current , the two parametrizations of the @xmath11-current give a negligible difference at low values of @xmath52 and a small reduction at larger momenta is obtained with the bonn parametrization .
the cross sections are very sensitive to correlations and to their treatment in the tof .
an example is shown in fig .
[ fig : fig2 ] , for the reaction @xmath0o(e , e@xmath32pp)@xmath31c to the @xmath53 ground state in the same super - parallel kinematics as in fig .
[ fig : fig1 ] . the cross sections calculated in the orthogonalized approach with three different tof s
are compared : the tof from the spectral function ( sf - b ) , from the simpler approach ( sm - src ) , and from the asymptotic behaviour of a tdm including only jastrow correlations ( jcm ) .
the three cross sections are very different , both in magnitude and shape , and all of them are dominated by the ob current , at least at lower values of @xmath52 .
the effect of the mutual interaction between the two outgoing protons ( nn - fsi ) , which has been neglected in the calculations presented till now , is of particular relevance for the ( e , e@xmath32pp ) reaction in the super - parallel kinematics @xcite .
an example is shown in fig .
[ fig : fig3 ] , where the cross sections calculated in the dw - nn and dw approaches are compared for the reaction @xmath0o(e , e@xmath32pp)@xmath31c to the @xmath53 ground state .
calculations have been performed with the tof from sf - b and the bonn parametrization for the @xmath11-current .
the comparison with the experimental data @xcite is also shown in the figure .
the contribution of nn - fsi is very large , both in the orthogonalized approach and with our previous prescription , and produces a strong enhancement of the cross section that is particularly large just when the cross section calculated in the dw approach is small .
the contribution of nn - fsi is very large at large values of @xmath52 .
the dw - nn cross sections calculated in the orthogonalized and in the previous approaches are very different .
the difference is mainly due to the cm and correlation effects that were neglected in the previous approach .
nn - fsi can be considered an effect of correlations between the two nucleons in the final state .
the comparison with the mami data @xcite in fig . [ fig : fig3 ] emphasizes the important contribution of the cm and correlation effects included in the new orthogonalized calculations , the essential role played by nn - fsi , as well as the need of a careful treatment of different types of correlations , which are consistently included in the tof from sf - b .
a different choice of the theoretical ingredients in figs . [ fig : fig1]-[fig : fig3 ] can give huge differences .
the most refined version of the model , with the best available ingredients ( the dw - nn approach with enforced orthogonality between initial and final states , the tof from sf - b , the bonn parametrization for the tb currents ) is able to give in the best agreement in comparison to data and a reasonable description of the @xmath0o(e , e@xmath32pp ) data . in fig .
[ fig : fig4 ] the differential cross section of the reaction @xmath0o(e , e@xmath32pn)@xmath31n is displayed for the transition to the @xmath58 excited state of @xmath31n at 3.95 mev .
this is the state that is mostly populated in pn - knockout @xcite
. results for this transition are therefore of particular interest .
calculations have been performed , with the tof from sf - b and the dw approximation for fsi , in the same super - parallel kinematics already considered in figs .
[ fig : fig1]-[fig : fig3 ] and realized in the @xmath0o(e , e@xmath32pn)@xmath31n experiment at mami @xcite .
the previous results of @xcite are compared in the figure with the results of the orthogonalized approach .
the cm effects included in the orthogonalized approach give a huge enhancement of the cross section calculated with the ob current .
the result is shown in the right panel of fig .
[ fig : fig4 ] . the enhancement is larger than one order of magnitude at low recoil momenta and is only slightly reduced at higher momenta .
the difference between the cross sections of the two approaches is reduced when the tb currents are added .
the result is shown in the left panel . in the previos calculations @xcite
the cross section is dominated by the tb currents , in particular by the @xmath11-current .
in contrast , the ob current dominates the cross section in the orthogonalized approach and only a small enhancement is given by the tb currents .
the differences obtained in the orthogonalized approach with the bonn and the unregularized parametrizations for the tb currents are appreciable , although not very important .
the bonn parametrization reduces the cross section by at most 30 - 40 % . in the final cross section
the difference between the results of the two approaches remains large , a bit less than one order of magnitude at low momenta , in the maximum region , and it is still sizable , although considerably reduced , at higher momenta .
the sensitivity to the treatment of correlations is shown in fig .
[ fig : fig5 ] , where the cross sections of the reaction @xmath0o(e , e@xmath32pn)@xmath31n to the @xmath58 state calculated in the super - parallel kinematics with the tof s from sf - b , sf - cc , and the simpler sm - src are compared .
the calculations have been performed with the orthogonalized approach , the dw approximation for fsi , and the bonn parametrization for the tb currents .
the three tof s give large differences , both in the shape and in the magnitude of the cross section .
with the simpler sm - src , where only src are taken into account , the contribution of the ob current is negligible and is up to about three orders of magnitude lower than the one obtained with sf - b . when the tb currents are added , the sm - src cross section is enhanced by about one order of magnitude , the difference between the sm - src and sf - b cross sections is reduced but remains very large , up to about two orders of magnitude in the maximum region and somewhat smaller at high values of @xmath52 .
the sm - src cross section is dominated by the tb currents and , as such , it is practically unaffected by the use of orthogonalized s.p . bound and scattering wave functions .
incidentally , we note that for the separate contribution of the ob current the cm effects included in the orthogonalized approach give with sm - src a much smaller effect than with sf - b and sf - cc , whose cross sections are , in both cases , dominated by the ob current .
the dominance of the ob current is due to the very large enhancement of the ob contribution produced , both with sf - b and sf - cc , by the cm effects included the orthogonalized approach . in the previous calculations with sf - b @xcite and sf - cc @xcite the cross section to the @xmath58 state was in both cases dominated by the tb @xmath11-current .
the much larger contribution of the ob current found with sf - b and sf - cc emphasizes the crucial role played by tc , that are very important in pn - emission and are neglected in the simpler sm - src calculation . in the sf - b and sf - cc overlap functions , src and tc
are accounted for consistently in the defect functions , which are calculated in the two tof s within different methods . in fig .
[ fig : fig5 ] the sf - b cross section is generally larger than the sf - cc one . the sf - b result overshoots the sf - cc one up to a factor of 6 in the maximum region . the differences are strongly reduced for values of @xmath59 greater than 100 mev/@xmath57 .
the differences between the two cross sections are due to the different treatment of src , tc and lrc .
a more complete calculation of lrc in an extended shell - model basis is performed in sf - b @xcite , where the normalization of the two - nucleon overlap amplitudes is higher than in sf - cc .
the difference in the shape of the cross section is due to the different mixing of configurations in the two calculations . the results in fig .
[ fig : fig5 ] emphasize the need of a careful and consistent treatment of different correlations in the tof .
the contribution of nn - fsi , that is very large in the super - parallel kinematics for the ( e , e@xmath32pp ) reaction , is strongly reduced in the same kinematics for pn - knockout , and it is quite moderate for the @xmath58 state @xcite .
the strong enhancement of the cross section to the @xmath58 state at low values of @xmath52 , that is due to the cm effects included in the orthogonalized approach , is able to resolve the discrepancies found @xcite in comparison with the experimental data and give a much better agreement @xcite .
a careful comparison with the data will be presented in a forthcoming paper @xcite .
the super - parallel kinematics is particularly sensitive to all the ingredients of the model .
different results can be obtained in different situations .
a suitable choice of kinematics can be helpful to disentangle specific contributions and reduce the uncertainties on the theoretical ingredients .
an example of a different kinematics is shown in fig . [
fig : fig6 ] , where the cross sections of the reactions @xmath0o(@xmath60,pp)@xmath31c@xmath61 and @xmath0o(@xmath60,pn)@xmath31n to the @xmath62 state are displayed , for a coplanar symmetrical kinematics with an incident photon energy @xmath63 mev . in symmetrical kinematics
the two nucleons are ejected at equal energies and equal but opposite angles with respect to @xmath49 , and different values of @xmath52 are obtained changing the scattering angles of the two outgoing nucleons . in this
kinematics the difference between the results of the orthogonalized and the previous prescription are generally small @xcite , and also the effects of nn - fsi are very small .
the cross sections are , however , very sensitive to the parametrization of the tb - currents .
the results of the bonn and the unregularized parametrizations , which are compared in the left panel of the figure , differ by about one order of magnitude .
also for this kinematics the results are very sensitive to the treatment of correlations in the tof .
the cross sections calculated with different tof s , which are compared in the right panels , exhibit very large differences , both in magnitude and shape .
recent improvements in the theortical description of electromagnetic two - nucleon knockout have been reviewed : the sensitivity to the correlations included in the nuclear wave functions , the parametrization used for the tb currents , the effects of final - state interactions , and the role of center - of - mass effects in connection with the problem of the lack of orthogonality between bound and scattering states obtained by the use of an energy - dependent optical potential .
numerical examples have been presented for pp- and pn - knockout off @xmath0o also in comparison with the available data .
the most refined version of the model , with the best available ingredients , is able to give a reasonable , although not perfect , description of the experimental data .
this result is an indication that the main theoretical ingredients contributing to the cross section seem reasonably under control .
the cross sections are very sensitive to correlations and to their treatment .
different type of correlations , of short - range and long - range type , are important and require a careful and consistent treatment .
an essential role is played by tensor correlations in pn - knockout .
the complexity of the necessary model ingredients makes it difficult to extract clear and unambigous information on correlations from one or two `` ideal '' kinematics .
if we want to isolate the contribution of correlations , data are needed in various kinematics which mutually supplement each other .
close collaboration between theorists and experimentalists is necessary to achieve this goal .
o(e , e@xmath32pp)@xmath31c to the low - lying states of @xmath31c : the @xmath53 ground state , the @xmath54 state at 11.31 mev , and the @xmath55 state at 7.67 mev .
the super - parallel kinematics of the mami experiment @xcite is considered , with an incident electron energy @xmath64 mev , @xmath65 mev and @xmath66 mev/@xmath57 .
different values of @xmath67 are obtained changing the kinetic energies of the two outoging nucleons .
positive ( negative ) values of @xmath52 refer to situations where @xmath67 is parallel ( anti - parallel ) to @xmath49 .
the final results given by sum of the ob and tb @xmath11-current are displayed in the left panels , the separate contribution of the ob current is shown in the right panels .
the tof from @xcite ( sf - b ) and the dw approximation for fsi are used in the calculations .
the dotted lines give the results of @xcite , the dashed and solid lines are obtained in the orthogonalized approach with different parametrizations of the tb currents , _
i.e. _ the previous unregularized pararametrization , as in @xcite , ( dashed ) and the bonn parametrization ( solid ) .
[ fig : fig1 ] , width=377,height=453 ] o(e , e@xmath32pp)@xmath31c to the @xmath53 ground state in the same super - parallel kinematics as in fig . 1 .
the final results given by sum of the ob and tb @xmath11-current ( calculated with the bonn parametrization ) are displayed in the left panel , the separate contribution of the ob current is shown in the right panel .
the curves are obtained in the orthogonalized approach with different tof s : sf - b ( solid ) , sm - src ( dashed ) , jcm ( dotted ) .
the dw approximation is used for fsi .
[ fig : fig2 ] , width=377,height=207 ] o(e , e@xmath32pp)@xmath31c to the @xmath53 ground state in the same super - parallel kinematics as in fig . 1 .
solid and dashed lines are obtained with the dw - nn and dw approaches for fsi .
the results of the orthogonalized approach and of the previous prescription to remove the spuriosity are displayed in the left and right panels , respectively .
the tof from sf - b and the bonn parametrization for the @xmath11-current are used in the calculations .
experimental data from @xcite .
[ fig : fig3 ] , width=377,height=207 ] o(e , e@xmath32pn)@xmath31n to the @xmath58 state at 3.95 mev in the same super - parallel kinematics as in fig . 1 .
the proton is emitted parallel and the neutron antiparallel to the momentum transfer .
the final results given by sum of the ob and tb currents are displayed in the left panel , the separate contribution of the ob current is shown in the right panel . line convention as in fig . 1 .
[ fig : fig4 ] , width=377,height=207 ] o(e , e@xmath32pn)@xmath31n to the @xmath58 state at 3.95 mev in the same super - parallel kinematics as in fig . 4 .
the curves are obtained with different tof s : sf - b ( solid ) , sf - cc ( dashed ) , sm - src ( dotted ) .
the final results given by sum of the ob and tb currents are displayed in the left panel , the separate contribution of the ob current is shown in the right panel .
calculations are performed in the orthogonalized approach , with the dw approximation for fsi , and the bonn parametrization for the tb currents .
[ fig : fig5 ] , width=377,height=207 ] o(@xmath60,pp)@xmath31c@xmath61 ( top panels ) and @xmath0o(@xmath60,pn)@xmath31n to the @xmath58 state ( bottom panels ) , in a coplanar symmetrical kinematics at @xmath63 mev , as a function of the scattering angle @xmath60 of the outgoing nucleons . the results obtained with the tof from sf - b for the bonn ( solid ) and the previous unregularized pararametrization ( dashed ) for the tb currents
are compared in the left panels .
the results obtained with the bonn parametrization for different tof s are compared in the right panels : sf - b ( solid ) , sm - src ( dotted ) , jcm ( dot - dashed ) , sf - cc ( dashed ) . the calculations are performed in the orthogonalized approach and with the dw approximation for fsi .
[ fig : fig6 ] , width=377,height=377 ] | recent advances in the description of electromagnetic two - nucleon knockout reactions are reviewed .
the sensitivity to different types of correlations and to their treatment in the nuclear wave functions , the effects of final - state interactions and the role of center - of - mass effects in connection with the problem of the lack of orthogonality between initial bound states and final scattering states obtained by the use of an energy - dependent optical - model potential are discussed .
results are presented for proton - proton and proton - neutron knockout off @xmath0o also in comparison with the available data .
= 1.0 cm = -0.3 cm = 1000 = 5000 * recent advances in the description of electromagnetic two - nucleon knockout reactions * 0.5truecm 0.5truecm 0.5truecm 0.5truecm |
nucleon-@xmath17 dissociation cross sections have been calculated in refs .
@xcite from the gluon-@xmath17 dissociation cross section obtained in short - distance qcd .
@xmath17 can be dissociated by a gluon with energy larger than the quark - antiquark binding energy of @xmath17 @xcite .
as the nucleon energy increases , it generates an increasing number of gluons overcoming the binding so that @xmath17 can be dissociated more easily .
therefore , the nucleon-@xmath17 dissociation cross section increases with total energy of @xmath18 and @xmath17 in the centre - of - mass frame .
this is a result of the short - distance approach in which the operator product expansion of perturbative qcd is applied to heavy quarkonia of small sizes @xcite . a similar energy dependence of @xmath19 and @xmath20 dissociation cross sections has been obtained in ref .
@xcite despite hadron mass corrections to the cross - section formula of peskin and bhanot . including the coulomb potential , the fermi contact term acting as the spin - spin interaction and a colour - independent confining force active only between a quark and an antiquark , cross
sections for @xmath21 , @xmath22 and @xmath23 have been calculated in ref .
@xcite in the quark - interchange mechanism @xcite .
the colour interaction that includes the colour coulomb , spin - spin hyperfine and linear confining interactions has been considered in refs .
@xcite for the dissociation of charmonia in collisions with @xmath16 and @xmath24 mesons .
these quark - model calculations of the quark - interchange mechanism give the result that the cross section increases from a threshold energy for every endothermic reaction from zero , reaching a maximum and decreasing with total energy of the meson and charmonium in the centre - of - mass frame .
this is the energy dependence of the cross section obtained in the quark - interchange approach of the quark potential models @xcite , where the quark interchange mechanism between the two initial mesons breaks the charmonium .
a study of @xmath25 , @xmath26 and @xmath27 at low energies of the mesons was initiated in ref .
meson exchange between the two initial mesons breaks the charmonium and lagrangians with meson couplings are constructed to describe the motion of meson fields .
this is the essence of the meson - exchange approach @xcite .
effective meson lagrangians with different symmetries lead to different cross sections for the same reaction @xcite .
near threshold energies cross sections increase for endothermic reactions and decrease for exothermic reactions . when the total energy of the two initial mesons in the centre - of - mass frame rises far away from threshold , the increase or decrease of the cross sections is due to form factors inserted in three - meson and four - meson vertices of feynman diagrams .
the energy dependence of cross sections for some hadron - charmonium dissociation reactions in vacuum has been studied in the short - distance , the meson - exchange and the quark - interchange approaches . in hadronic matter not only
the energy dependence is interesting , but also the temperature dependence of hadron - charmonium dissociation cross sections is important @xcite . from vacuum to hadronic matter the quark potential , mesonic quark - antiquark relative - motion wave functions , meson masses and cross sections for @xmath16-charmonium and @xmath24-charmonium dissociation change significantly .
the temperature dependence of the cross sections has been shown to be unexpected @xcite .
for instance , even though a @xmath24-charmonium reaction is exothermic in vacuum , it may be endothermic in hadronic matter ; the peak cross section of any @xmath16-charmonium dissociation reaction first decreases with increasing temperature , but then increases rapidly when the temperature approaches the critical temperature .
therefore , we continue here the study of temperature dependence .
we choose kaon - charmonium dissociation reactions as the first objective of the present work .
the energy dependence of the cross sections for the reactions in vacuum has been studied in ref .
@xcite , but in hadronic matter it is not available .
the temperature dependence of the kaon - charmonium dissociation cross sections has not been studied either .
it has been shown by experiments that the ratios @xmath28 , @xmath29 and @xmath30 at midrapidity increase with increasing transverse momentum @xmath31 and the increase is visible only in a certain range of transverse momentum .
for central au+au collisions at @xmath32 gev the ratio @xmath28 ( @xmath29 ) measured by the phenix collaboration is availabe from @xmath33 gev/@xmath34 to 1.95 gev/@xmath34 and it is 0.234 ( 0.221 ) at @xmath35 gev/@xmath34 and 0.55 ( 0.55 ) at @xmath36 gev/@xmath34 at midrapidity @xcite . for central pb+pb collisions at @xmath37 tev the ratio @xmath38 measured by the alice collaboration increases with increasing transverse momentum from @xmath39 , forms a peak at @xmath40 gev/@xmath34 and goes to the value 0.45 at @xmath41 gev/@xmath34 @xcite .
at present , for central au+au collisions at @xmath32 gev the smallest @xmath0 transverse momentum measured by the phenix collaboration is 1.65 gev/@xmath34 and the ratio @xmath30 is 0.46 at @xmath36 gev/@xmath34 at midrapidity @xcite . from the measured ratios we conclude that the @xmath0 species is as rich as the @xmath42 and @xmath43 species in hadronic matter and we need to study @xmath0-charmonium dissociation reactions which form the second objective of the present work . the short - distance approach and the quark - interchange approach
have not been studied for the @xmath0-charmonium dissociation reactions .
the @xmath44 dissociation in vacuum was considered in ref .
@xcite , but no cross sections were presented .
however , we will show that @xmath44 dissociation reactions need to be considered in the study of @xmath17 suppression in hadronic matter .
this paper is organized as follows . in the next section we present cross - section formulas for the charmonium dissociation in collision with a meson . in section
3 kaon - charmonium and @xmath0-charmonium dissociation reactions are listed , numerical cross sections are shown , relevant discussions are given and the numerical results are parametrized in appendix a. using the @xmath0-charmonium dissociation cross sections and the @xmath16-charmonium dissociation cross sections given in ref .
@xcite , we calculate the ratio of the corresponding dissociation rates in hadronic matter .
a summary is in the last section .
in quark degrees of freedom the meson - charmonium dissociation is expressed as @xmath45 where @xmath46 stands for the up quark , down quark or strange quark .
the flavour of the quark @xmath46 may be different from the flavour of the antiquark @xmath47 .
the cross section for @xmath45 is @xcite @xmath48 is the total spin of either the two initial mesons or the two final mesons , and it is conserved in the reaction ; @xmath49 denotes the magnetic projection quantum number of the total spin .
the mandelstam variables are @xmath50 and @xmath51 , where @xmath52 is the four - momentum of meson @xmath53 ; @xmath54 is the angle between the three - dimensional @xmath55 momentum @xmath56 and the three - dimensional @xmath57 momentum @xmath58 in the centre - of - mass frame .
@xmath59 in eq .
( 1 ) stands for the transition amplitude , which takes different values for the two forms of meson - meson scattering , namely , the prior form in fig .
1 and the post form in fig . 2 @xcite .
the scattering in the prior form means that gluon exchange occurs before quark interchange .
the corresponding transition amplitude is @xmath60 , @xmath61 , @xmath62 and @xmath63 , represent the products of quark - antiquark relative - motion wave functions in momentum space and wave functions in the colour space , spin space and flavour space .
@xmath64 is the potential of @xmath46 and @xmath65 and the meaning of @xmath66 and so on can be understood in the same way .
the scattering in the post form contains gluon exchange after quark interchange .
the corresponding transition amplitude is @xmath67 where @xmath68 is a central spin - independent potential and @xmath69 is a spin - spin interaction .
the temperature dependence is explicitly expressed in @xmath68 , @xmath70 \tanh ( ar ) + \frac { \vec { \lambda}_a}{2 } \cdot \frac { \vec { \lambda}_b}{2 } \frac { 6\pi}{25 } \frac { v(\lambda r)}{r } \exp ( -er ) , \label{eq : t vsi}\ ] ] where @xmath71 gev , @xmath72 gev , @xmath73 ^ 6 $ ] gev , @xmath74 gev and @xmath75 in which @xmath76 gev@xmath77 and @xmath78 with @xmath79 ; @xmath80 are the gell - mann matrices for the colour generators of constituent @xmath81 .
the dimensionless function @xmath82 @xcite is @xmath83 \sin \left(\frac { q}{\lambda}x\right ) , \label{eq : v}\end{aligned}\ ] ] where @xmath84 and @xmath85 is the physical running coupling constant at the gluon momentum @xmath86 .
lattice qcd calculations gave a temperature - dependent quark potential at intermediate and large distances @xcite .
in contrast to the linear confinement in vacuum , the quark potential at large distances is independent of the distance and is a plateau . with increasing temperature
the plateau becomes lower and lower and confinement gets weaker and weaker . from a large distance to an intermediate distance the variation of the potential with respect to temperature gets smaller . at @xmath87
fm the variation disappears .
the lattice qcd results ( the above potential ) @xcite give the spin - independent potential in eq .
( 5 ) at intermediate and large distances . at short distances quark interaction is described by perturbative qcd in vacuum .
the spin - independent potential in eq .
( 5 ) at short distances is thus given by one - gluon exchange plus perturbative one- and two - loop corrections in vacuum @xcite .
the temperature dependence of the potential in eq .
( 5 ) comes from the lattice qcd results .
the potential well fits the lattice qcd results at @xmath88 @xcite and is constrained by perturbative qcd at short distances and the lattice qcd results at intermediate and large distances .
we may adjust the parameters @xmath89 and @xmath90 to get a good fit of the lattice qcd results , for example , reduce @xmath89 by 5% and increase @xmath90 by 10% or increase @xmath89 by 5% and reduce @xmath90 by 10% .
but the changes of the two parameters lead to very small changes of meson masses and cross sections for meson - meson reactions @xcite .
hence , the potential has the uncertainty of the values of @xmath89 and @xmath90 , but the uncertainty causes very small changes in meson mass and in cross section .
the expression @xmath91 in the second term of eq .
( 5 ) is obtained in perturbative qcd and arises from one - gluon exchange plus perturbative one- and two - loop corrections in vacuum @xcite .
the factor @xmath92 assures that the potential at small distances is given by perturbative qcd in vacuum . the second term in eq .
( 5 ) is independent of temperature . from a correlator of a very heavy quark - antiquark pair
a time - dependent potential was obtained in quenched lattice qcd @xcite . only at very large times and at @xmath93
the potential agrees with the free energy in the coulomb gauge . using the free energy from lattice calculations as the potential of a charm quark and a charm antiquark in the schrdinger equation correctly describes the nonrelativistic wave function of @xmath17 and reproduces the @xmath17 mass from the qcd sum rule in the vicinity of the critical temperature @xcite .
when the system s temperature is smaller than the critical temperature , the product of the temperature and a meson s entropy is small or negligible in comparison with the quark - antiquark free energy and the free energy can be taken as the quark - antiquark potential to a good approximation @xcite .
the three methods suggest that the free energy obtained in the lattice qcd calculations can be taken as the quark potential in hadronic matter .
the spin - spin interaction @xmath69 arises from perturbative one - gluon exchange plus one- and two - loop corrections @xcite and includes relativistic effects @xcite : @xmath94 where @xmath95 and @xmath96 are the spin and mass of constituent @xmath81 , respectively , and the quantity @xmath97 is related to quark masses by @xmath98+\sigma_{1}^2\left(\frac{2m_am_b}{m_a+m_b}\right)^2 , \label{eq : d}\end{aligned}\ ] ] where @xmath99 gev and @xmath100 . finally , the unpolarised cross section is @xmath101 where @xmath102 ( @xmath103 ) is the @xmath55 ( @xmath104 ) spin and @xmath105 ( @xmath106 ) is obtained by the replacement of @xmath59 in eq .
( 1 ) by @xmath107 ( @xmath108 ) .
@xmath109 is the magnetic projection quantum number of the orbital angular momentum @xmath110 of the meson @xmath104 .
given the charm quark mass @xmath111 gev , the up and down quark masses @xmath112 gev and the strange quark mass @xmath113 gev , the schrdinger equation with the potential in eq . ( 4 ) at @xmath114 reproduces @xcite the experimental masses of @xmath17 , @xmath115 , @xmath116 , the pion , the rho , the kaon , the vector kaon , the @xmath0 meson , the charmed mesons and the charmed strange mesons @xcite . the potential at @xmath114 and the mesonic quark - antiquark relative - motion wave functions obtained from the schrdinger equation reproduce the experimental data of @xmath117-wave @xmath118 elastic phase shifts for @xmath119 scattering in vacuum @xcite .
we establish the notation @xmath120 , @xmath121 , @xmath122 , @xmath123 , @xmath124 and @xmath125 .
the @xmath1-charmonium dissociation includes the reactions : @xmath126 @xmath127 @xmath128 the @xmath129-charmonium dissociation includes @xmath130 @xmath131 @xmath132 cross sections for the @xmath133 dissociation reactions are obtained from the @xmath134 dissociation reactions , for instance , the cross section for @xmath135 is identical to the cross section for @xmath136 . since the @xmath0 meson has the components of @xmath137 , @xmath138 and @xmath139 , the @xmath0-charmonium dissociation includes @xmath137- or @xmath138-induced reactions , viz .
@xmath140 @xmath141 @xmath142 and @xmath139-induced reactions , @xmath143 @xmath144 @xmath145 the valence quark - antiquark wave function of the @xmath0 meson is @xmath146 where @xmath147 ( @xmath148 , @xmath149 ) is the relative - motion wave function of the up ( down , strange ) quark and the up ( down , strange ) antiquark in momentum space ; @xmath150 and @xmath151 are the colour wave function and the spin wave function , respectively .
@xmath147 , @xmath148 and @xmath149 satisfy @xmath152 where @xmath153 ( @xmath154 , @xmath155 ) is the relative momentum of the up ( down , strange ) quark and the up ( down , strange ) antiquark .
the first ( second , third ) term is used in the transition amplitude to obtain the cross sections @xmath156
( @xmath157 , @xmath158 ) for the @xmath137-induced ( @xmath138-induced , @xmath139-induced ) reactions .
the cross section for the @xmath0-charmonium dissociation is @xmath159 the cross section for the production of @xmath160 , @xmath161 or @xmath162 equals the first and second terms while the one for the production of @xmath163 , @xmath164 or @xmath165 equals the third term . solving the schrdinger equation with the potential in eq .
( 4 ) , we obtain temperature - dependent meson masses shown in figs . 1 and 17 in ref .
@xcite and quark - antiquark relative - motion wave functions for charmonia , charmed mesons , charmed strange mesons and the mesons in the ground - state pseudoscalar octet and the ground - state vector nonet . the quark - antiquark relative - motion wave functions used in the transition amplitude are the fourier transforms of the relative - motion wave functions obtained here . since the up quark has the same mass as the down quark , @xmath166 equals @xmath167 but differs from @xmath168 .
then , the cross sections @xmath156 for the @xmath137-induced reactions are the same as @xmath157 for the @xmath138-induced reactions . cross sections for exothermic reactions are infinite at threshold energies and we start calculating the cross sections at @xmath169 with @xmath170 gev . unpolarised cross sections for the @xmath1-charmonium dissociation reactions and @xmath0-charmonium dissociation reactions are displayed in figs .
the cross sections for exothermic reactions decrease very rapidly from infinity when @xmath171 increases from the threshold energies . to indicate this feature of an exothermic reaction
, the rapid decrease should be displayed .
however , when we start plotting the cross sections at the threshold energies plus @xmath172 gev , the decreasing part of the curve for @xmath173 at @xmath174 in fig .
16 disappears . likewise when we start plotting the cross sections at the threshold energies plus @xmath175 gev , the decreasing part of the curve for @xmath176 at @xmath177 in fig .
14 disappears .
therefore , we choose the threshold energies plus @xmath178 gev to start plotting the cross sections so that the rapid decrease is visible . in vacuum
the cross section for @xmath179 obtained from the effective meson lagrangian in ref .
@xcite increases with increasing @xmath171 and reaches a magnitude of about 8 mb . in the present work the cross section for @xmath179 shown by the curve at @xmath114 first increases to a maximum value of about 0.27 mb with increasing @xmath171 and then decreases .
therefore , the magnitude of the present cross section is much smaller than the magnitude obtained in ref .
the @xmath180 dissociation in vacuum was considered in ref .
@xcite , but no cross sections were shown . in ref .
@xcite cross sections for the production of @xmath181 , @xmath182 and @xmath183 in the @xmath1-charmonium dissociation in vacuum were obtained in the quark - interchange mechanism .
the peak cross section of the endothermic reaction @xmath184 is about 0.7 mb at the kinetic energy of the order of 0.44 gev .
the cross section for the exothermic reaction @xmath185 decreases very rapidly and then increases slowly to form a wide peak that corresponds to a cross section of about 1 mb . the exothermic reaction @xmath186 has a peak cross section of about 3 mb at the kinetic energy of the order of 0.17 gev . by comparison
, we obtain a peak cross section of about 0.6 mb at the kinetic energy of the order of 0.44 gev for @xmath184 , no peak cross section for @xmath185 and a peak cross section of about 3.4 mb at the kinetic energy of the order of 0.08 gev for @xmath187 .
we note that we use the average value of the @xmath188 , @xmath189 and @xmath190 masses as the @xmath116 mass and the average value is a little higher than the @xmath189 mass .
the present results are comparable to the @xmath1-charmonium dissociation cross sections in ref .
@xcite . the present work and the quark - interchange model in ref .
@xcite have the same colour matrix elements and the same spin matrix elements , but have different quark potentials that lead to different spatial matrix elements in the transition amplitude . hence
, the energy dependence of the @xmath1-charmonium dissociation cross sections obtained in the present work is not the same as that displayed in ref .
@xcite . in figs . 3 - 11 only @xmath191 , @xmath192 , @xmath193 , @xmath194 and @xmath195 at @xmath114
are exothermic . while temperature goes from @xmath196 to @xmath197 , the increases of the kaon and charmonium radii
cause the increases of the peak cross sections of the @xmath198 and @xmath199 reactions , but the peak cross sections of the @xmath200 reactions decrease .
this relates to the node in the @xmath115 wave function .
the node leads to cancellation between the negative wave function on the left of the node and the positive wave function on the right of the node in the integration involved in the transition amplitude . while the cancellation at @xmath201 is larger than at @xmath202 , the peak cross sections of @xmath191 , @xmath192 and @xmath193 decrease from @xmath202 to @xmath197 . while temperature increases from @xmath201 , the increase of @xmath203 caused by the slow increases of the two initial - meson radii can not overcome the reduction by the weakening confinement and the peak cross sections thus decrease
however , the rapid increases of the kaon and @xmath17 radii from @xmath204 to @xmath205 cause the increase of @xmath203 to overcome the reduction by the weakening confinement , and the peak cross sections of @xmath206 , @xmath207 and @xmath208 rise rapidly from @xmath204 to @xmath205 .
we need to mention the difference between the cross section for @xmath206 and the one for @xmath209 .
the difference arises from the different spins of the charmed mesons and the different spins of the charmed strange mesons between the two reactions .
the masses of the @xmath210 , @xmath211 , @xmath212 and @xmath213 mesons are different and the sum of the @xmath214 and @xmath212 masses is smaller than the sum of the @xmath210 and @xmath213 masses .
hence , the endothermic reaction @xmath206 takes place more easily than @xmath207 and at @xmath215=0 ( 0.65 , 0.75 , 0.85 , 0.9 , 0.95 ) the peak cross section of @xmath206 is 18% ( 47% , 42% , 11% , 135% , 95% ) larger than the one of @xmath207 . the difference between the cross section for @xmath216 and the one for @xmath217 ( @xmath218 and @xmath219 ) can be similarly understood .
the cross sections for the @xmath0-charmonium dissociation reactions have the following behaviour . as shown in figs .
12 - 17 , only three reactions @xmath220 , @xmath221 and @xmath222 are endothermic below some temperature .
the peak cross sections of the reactions increase with temperature from @xmath114 to @xmath223 . as shown in figs .
18 - 23 , only three reactions @xmath224 , @xmath225 and @xmath226 are exothermic below some temperature .
the peak cross sections of the reactions decrease with temperature going from @xmath202 to @xmath227 .
the peak cross sections of @xmath228 , @xmath229 , @xmath230 , @xmath231 , @xmath232 and @xmath233 first rise and then fall with temperature increasing in the region @xmath234 .
since the ratio of the @xmath0 mass to the kaon mass is about 1.1 for @xmath235 , i.e. the two masses are close , it is interesting to compare the peak cross sections of the endothermic @xmath0-charmonium dissociation reactions with those of the endothermic kaon - charmonium dissociation reactions .
the peak cross section of @xmath220 at @xmath236 ( 0.65 , 0.75 , 0.85 ) is 1.64 ( 1.57 , 1.55 , 12.11 ) times the peak cross section of @xmath206 at the same temperature .
the peak cross section of @xmath222 at @xmath236 ( 0.65 , 0.75 , 0.85 ) is 0.56 ( 1.47 , 3.87 , 49.4 ) times the peak cross section of @xmath208 .
this means that for @xmath237 the cross sections for @xmath220 and @xmath222 around their cross - section peaks are larger than the cross sections for @xmath206 and @xmath208 , respectively .
except for @xmath230 and @xmath208 at @xmath238 , the peak cross sections of @xmath239 and @xmath240 are smaller than the ones of @xmath241 and @xmath242 , respectively . since the @xmath0-charmonium dissociation and the @xmath16-charmonium dissociation may produce the same charmed mesons , it is interesting to compare the peak cross sections of the endothermic @xmath0-charmonium dissociation reactions with those of endothermic @xmath16-charmonium dissociation reactions
. cross sections for the @xmath16-charmonium dissociation reactions are given in ref .
@xcite . at any temperature between @xmath243 and 1 the peak cross section of @xmath244 ( @xmath245 ) is more than 1.8 times the one of @xmath246 ( @xmath247 ) . in the @xmath0-charmonium dissociation reactions
the @xmath139 component of the @xmath0 meson leads to the production of @xmath248 , @xmath249 or @xmath250 .
these charmed strange mesons are not produced in the @xmath16-charmonium dissociation reactions .
based on the estimate of the @xmath0-charmonium and @xmath16-charmonium dissociation cross sections , in the following we evaluate the difference between the dissociation rates of charmonium with @xmath16 and @xmath0 in hadronic matter . in hadronic matter the @xmath16 and @xmath0 mesons satisfy the bose - einstein distribution @xmath251 where @xmath252 is the @xmath55 mass .
the @xmath55 number density is @xmath253 where @xmath254 is the spin - isospin degeneracy factor and equals 3 for the pion and 1 for the @xmath0 meson .
the thermal - averaged meson - charmonium dissociation cross section is @xmath255 where @xmath256 is the relative velocity of the @xmath55 meson and the charmonium .
the dissociation rate of the charmonium in the interaction with the @xmath55 meson in hadronic matter is @xmath257 which determines charmonium suppression in mesonic matter @xcite .
the larger the dissociation rate is , the stronger suppression this mesonic matter causes .
the ratio of the dissociation rates of charmonium in the interactions with @xmath0 and @xmath16 is @xmath258 where @xmath259 and @xmath260 are the number densities of @xmath0 and @xmath16 mesons , respectively ; @xmath261 and @xmath262 are the unpolarised cross sections for a @xmath0-charmonium dissociation reaction and a @xmath16-charmonium dissociation reaction , respectively .
we calculate the ratio for the six sets of reactions : @xmath263 @xmath264 @xmath265 @xmath266 @xmath267 @xmath268 results are shown in figs .
the two reactions in each set have the same final states .
the ratio decreases with charmonium momentum increase . for @xmath269 and @xmath270 the ratio in fig .
26 is smaller than 0.018 . for @xmath271 and @xmath272 the ratio in fig .
27 is smaller than 0.82 . for @xmath273 and @xmath274 the ratio in fig .
28 is smaller than 0.08 . for @xmath275 and @xmath276 the ratio in fig .
29 is smaller than 0.72 . for @xmath277 and @xmath278 the ratio at @xmath215=0.65 , 0.75 , 0.9 and 0.95 in fig .
24 is smaller than 0.73 and the ratio at @xmath215=0.85 is larger than 1 when the @xmath17 momentum is smaller than 0.7 gev/@xmath34 . for @xmath279 and @xmath280 the ratio at @xmath215=0.65 and 0.95 in fig .
25 is smaller than 0.51 and the ratio at @xmath215=0.75 , 0.85 and 0.9 is larger than 1 when the @xmath17 momentum is smaller than 0.6 gev/@xmath34 , 3.3 gev/@xmath34 and 6.6 gev/@xmath34 , respectively .
therefore , compared to the dissociation rate of @xmath115 ( @xmath116 ) with @xmath16 , the dissociation rate of @xmath115 ( @xmath116 ) with @xmath0 may be neglected ; but the dissociation rate of @xmath17 with @xmath0 is comparable to the dissociation rate of @xmath17 with @xmath16 at low @xmath17 momenta . to study the @xmath17 suppression in hadronic matter , the @xmath44 dissociation reactions need to be considered .
we have studied the production of @xmath281 , @xmath282 and @xmath283 in the @xmath1-charmonium dissociation and the production of @xmath284 , @xmath285 , @xmath286 , @xmath287 , @xmath288 and @xmath289 in the @xmath0-charmonium dissociation .
the @xmath1-charmonium dissociation includes 9 reactions and the @xmath0-charmonium dissociation includes 18 reactions .
the cross sections for the 27 reactions are calculated with the temperature - dependent quark potential , in the born approximation and in the quark - interchange mechanism .
the cross sections for the @xmath129-charmonium dissociation reactions are identical to the cross sections for the @xmath1-charmonium dissociation reactions .
the numerical cross sections are parametrized as functions of @xmath171 .
the temperature dependence of the peak cross sections of the endothermic @xmath1-charmonium dissociation reactions is closely related to the temperature dependence of the @xmath1 radius , the charmonia radii , the confinement and the node of the @xmath115 wave function . for @xmath290
the cross sections for @xmath284 , @xmath285 and @xmath286 produced in the endothermic @xmath291 reactions around their cross - section peaks are larger than the ones for @xmath281 , @xmath282 and @xmath283 produced in the endothermic @xmath292 reactions , respectively . for @xmath290
the cross sections for the @xmath0-charmonium dissociation reactions are larger than the cross sections for the @xmath16-charmonium dissociation reactions . to the @xmath115 and @xmath116 suppression in hadronic matter the @xmath293 and @xmath294 dissociation reactions
may be neglected , but to the @xmath17 suppression the @xmath44 dissociation reactions need to be taken into account .
this work was supported by the national natural science foundation of china under grant no .
we thank p. mcguire and h. j. weber for careful readings of the manuscript .
the numerical cross sections for the endothermic reactions in figs . 3 - 13 and
18 - 23 can be parametrized as @xmath295 \nonumber \\ & & + a_2 \left ( \frac { \sqrt { s } -\sqrt { s_0 } } { b_2 } \right)^{c_2 } \exp \left [ c_2 \left ( 1-\frac { \sqrt { s } -\sqrt { s_0 } } { b_2 } \right ) \right ] , \nonumber \\\end{aligned}\ ] ] and for the exothermic reactions in figs .
6 - 10 , 12 - 17 and 20 - 21 @xmath296 \right .
\nonumber \\ & & + \left .
a_2 \left ( \frac { \sqrt { s } -\sqrt { s_0 } } { b_2 } \right)^{c_2 } \exp \left [ c_2 \left ( 1-\frac { \sqrt { s } -\sqrt { s_0 } } { b_2 } \right ) \right ] \right\ } , \nonumber \\\end{aligned}\ ] ] with @xmath297^{2 } -4m_{q\bar { q}}^{2}m_{c\bar { c}}^{2}\right\ } , \nonumber \\ \vec { p'}^{2}(\sqrt{s})=\frac{1}{4s}\left\ { \left [ s-\left ( m_{q\bar { c}}^2+m _ { c\bar{q}}^2\right)\right]^2 -4 m _ { q\bar { c } } ^2 m _ { c\bar{q}}^2 \right\}. \nonumber \\\end{aligned}\ ] ] here @xmath298 , @xmath299 , @xmath300 , @xmath301 , @xmath302 and @xmath303 are parameters and @xmath304 is the threshold energy .
the parameter values are listed in tables 1 - 6 .
we also list @xmath305 what is the separation between the peak s location on the @xmath171-axis and the threshold energy and @xmath306 which is the square root of the mandelstam variable at which the cross section is 1/100 of the peak cross section . according to a procedure presented in ref .
@xcite we can get cross sections at any temperature from the quantities listed in tables 1 - 6 .
00 d. kharzeev , h. satz , phys .
b 334 , 155 ( 1994 ) . m.e .
peskin , nucl .
b 156 , 365 ( 1979 ) ; g. bhanot , m.e . peskin , nucl .
b 156 , 391 ( 1979 ) . x .-
xu , d. kharzeev , h. satz , x .-
wang , phys .
c 53 , 3051 ( 1996 ) .
f. arleo , p.b .
gossiaux , t. gousset , j. aichelin , phys .
d 65 , 014005 ( 2001 ) .
k. martins , d. blaschke , e. quack , phys .
c 51 , 2723 ( 1995 ) .
i. bender , h.g .
dosch , h.j .
pirner , h.g .
kruse , nucl .
a 414 , 359 ( 1984 ) .
t. barnes , e.s .
swanson , phys .
d 46 , 131 ( 1992 ) ; e.s . swanson , ann .
220 , 73 ( 1992 ) .
wong , e.s .
swanson , t. barnes , phys . rev .
c 62 , 045201 ( 2000 ) ; phys .
c 65 , 014903 ( 2001 ) .
t. barnes , e.s .
swanson , c .- y .
wong , x .-
xu , phys .
c 68 , 014903 ( 2003 ) .
matinyan , b. mller , phys .
c 58 , 2994 ( 1998 ) .
z. lin , c.m .
ko , phys .
c 62 , 034903 ( 2000 ) ; j. phys .
g 27 , 617 ( 2001 ) .
haglin , phys .
c 61 , 031902 ( 2000 ) ; k.l .
haglin , c. gale , phys .
c 63 , 065201 ( 2001 ) .
y. oh , t. song , s.h .
lee , phys .
c 63 , 034901 ( 2001 ) .
navarra , m. nielsen , m.r .
robilotta , phys .
c 64 , 021901(r ) ( 2001 ) .
l. maiani , f. piccinini , a.d .
polosa , v. riquer , nucl . phys . a 741 , 273 ( 2004 ) .
a. bourque , c. gale , phys .
c 78 , 035206 ( 2008 ) ; 80 , 015204 ( 2009 ) .
wong , phys .
c 65 , 034902 ( 2002 ) .
j. zhou , x .-
xu , phys .
c 85 , 064904 ( 2012 ) .
adler , et al . ,
phenix collaboration , phys .
c 69 , 034909 ( 2004 ) .
b. abelev , et al . ,
alice collaboration , phys .
b 736 , 196 ( 2014 ) .
b. sahlmller , j. phys .
g 34 , s969 ( 2007 ) .
xu , nucl .
a 794 , 210 ( 2007 ) .
mott , h.s.w .
massey , the theory of atomic collisions , clarendon press , oxford , 1965 .
t. barnes , n. black , e.s .
swanson , phys .
c 63 , 025204 ( 2001 ) .
wong , h.w .
crater , phys .
c 63 , 044907 ( 2001 ) .
w. buchmller , s .- h .
h. tye , phys .
d 24 , 132 ( 1981 ) .
f. karsch , e. laermann , a. peikert , nucl .
b 605 , 579 ( 2001 ) . y .-
zhang , x .-
xu , h .- j .
ge , nucl .
phys . a 832 , 112 ( 2010 ) .
a. rothkopf , t. hatsuda , s. sasaki , phys .
108 , 162001 ( 2012 ) .
y. burnier , a. rothkopf , phys .
111 , 182003 ( 2013 ) .
s. h. lee , k. morita , t. song , c. m. ko , phys .
d 89 , 094015 ( 2014 ) .
shen , x .-
xu , chin .
c 39 , 074103 ( 2015 ) .
xu , nucl .
phys . a 697 , 825 ( 2002 ) .
s. godfrey , n. isgur , phys . rev .
d 32 , 189 ( 1985 ) .
k. nakamura , et al .
, particle data group , j. phys .
g 37 , 075021 ( 2010 ) .
e. colton , et al .
d 3 , 2028 ( 1971 ) .
n. b. durusoy , et al .
b 45 , 517 ( 1973 ) . w. hoogland , et al .
b 126 , 109 ( 1977 ) . m. j. losty , et al .
b 69 , 185 ( 1974 ) .
.quantities relevant to the cross sections for the @xmath314 dissociation reactions . @xmath298 and @xmath301 are in units of mb ; @xmath299 , @xmath302 , @xmath305 and @xmath315 are in units of gev ; @xmath300 and @xmath303 are dimensionless . [ cols="<,<,<,<,<,<,<,<,<,<",options="header " , ] | we study kaon - charmonium and @xmath0-charmonium dissociation reactions . the @xmath1-charmonium dissociation and the @xmath0-charmonium dissociation include the following 27 reactions : @xmath2 , @xmath3 and @xmath4 ; @xmath5 , @xmath3 and @xmath4 ; @xmath6 , @xmath3 and @xmath4 ; @xmath7 , @xmath8 , @xmath9 , @xmath10 , @xmath11 and @xmath12 ; @xmath13 , @xmath8 , @xmath9 , @xmath14 , @xmath11 and @xmath12 ; @xmath15 , @xmath8 , @xmath9 , @xmath14 , @xmath11 and @xmath12 .
cross sections for the reactions are calculated in the born approximation , in the quark - interchange mechanism and with a temperature - dependent quark potential .
the temperature dependence of peak cross sections of endothermic reactions is linked to the temperature dependence of quark - antiquark relative - motion wave functions , meson masses and the quark potential .
although the @xmath0 meson and kaon have similar masses , the energy and temperature dependence of the @xmath0-charmonium dissociation cross sections are quite different from those of the @xmath1-charmonium dissociation cross sections . using the @xmath0-charmonium and @xmath16-charmonium dissociation cross sections
, we calculate the ratio of the corresponding dissociation rates in hadronic matter and we find that such rates are comparable at low @xmath17 momenta .
department of physics , shanghai university , baoshan , shanghai 200444 , china keywords : charmonium dissociation cross sections ; quark - interchange mechanism pacs : 25.75.-q ; 24.85.+p ; 12.38.mh |
very massive stars produce violent gravitational explosions at the end of their life . a huge amount of ( anti-)neutrinos of all flavours is emitted during this phenomenon that typically lasts a few seconds .
the neutrino luminosity curve follows closely the collapse , accretion phase and the cooling of the neutron star left .
stars with masses larger than 20 - 25 solar masses can leave a black hole instead , with the neutrino time distribution cut after a few hundreds milliseconds @xcite .
neutrinos emitted in gamma - ray - bursts can also produce a characteristic neutrino signal @xcite .
other interesting features of the supernova dynamics can be extracted from a core - collapse supernova neutrino signal like e.g. information on the shock - wave @xcite .
core - collapse supernova neutrinos have been measured only once so far , during the explosion of the sn1987a located in the large magellanic cloud at 50 kpc from us .
the few events observed in the kamiokande @xcite , imb @xcite and baksan @xcite detectors have brought important confirmation on the theoretical expectations on the neutrino fluxes and information on neutrino properties .
large scale observatories at present under study ( laguna design study financed within the fp7 ) should reach the sensitivity to observe extra - galactic explosions and for the discovery of relic supernova neutrinos .
three technologies are under investigation : water erenkov ( memphys / uno / hyper - k ) , scintillators ( lena ) and liquid argon ( glacier ) @xcite .
about a hundred events are expected in a detector such as borexino , about 10000 in super - kamiokande @xcite and 10@xmath0 in memphys for a typical supernova at 10 kpc @xcite .
while explosions in our galaxy are a rare phenomenon ( typically 1 - 3 per century ) , the rate increases up to 1 per year at a distance as far as 10 mpc @xcite .
relic supernova neutrinos ( also called the diffuse supernova neutrino background ) are neutrinos produced in past supernova explosions .
the relic fluxes are not only sensitive to the supernova dynamics and neutrino properties , but also on the star formation rate .
the latter can be constrained through direct and indirect observations . at redshifts smaller than 1
the rate is rather well known @xcite ; while very recent works try to constrain at redshifts as high as 4 @xcite1 ( z@xmath12 ) . ] .
the experimental present upper limits at 90 @xmath2 c.l .
are of 6.8 @xmath3 10@xmath4 @xmath5 @xmath6s@xmath7 ( 25 mev @xmath8 mev ) and 1.2 @xmath9 @xmath6s@xmath7 ( @xmath10 mev ) come from the lsd @xcite and the super - kamiokande detectors @xcite respectively .
present theoretical predictions show that the relic fluxes are very close to the super - kamiokande limits , so that they might be discovered with large size detectors @xcite . in @xcite
it has been proposed to improve our present relic electron neutrino limit by using the few events associated to electron scattering on oxygen in memphys or on carbon in lena .
on the other hand several hundred events might be measured in glacier thanks to the large neutrino - argon cross section @xcite , as first pointed out in @xcite .
impressive progress has been made in neutrino physics in the last ten years , after the discovery of the phenomenon of neutrino oscillations performed by the super - kamiokande experiment @xcite .
however crucial questions remain open , among which the value of the third neutrino mixing angle , of the cp violating phase(s ) , of the ( majorana versus dirac ) nature of neutrinos and of the absolute mass scale .
the study of neutrinos from the early universe or from stars brings important elements to progress on some of the open issues . in particular
, unknown neutrino properties might have an indirect impact on the supernova environment such as e.g. on the nucleosynthesis of the heavy elements during the r - process or leave an imprint on the time or energy supernova neutrino signal in a detector on earth .
for example in @xcite we have shown that in lead - based detectors one can exploit the measurement of the electron neutrino average energy through one and two neutron emission to extract information on the third neutrino mixing angle .
a lead - based detector ( the halo project ) is planned for construction at snolab @xcite . here
we will discuss possible cp violating effects in the supernova context @xcite . in order to use supernova neutrinos to extract information on neutrino properties , we need to progress on the uncertainties that still affect the supernova neutrino predictions .
in fact , the theoretical calculations of the neutrino fluxes are affected by uncertainties at the neutrinosphere , due to different supernova modelling .
these uncertainties will narrow down once supernova explosions will be successfully obtained for different stellar masses and the relevant microscopic processes included in the neutrino diffusion .
obviously , the number of unknowns can also reduce if planned neutrino experiments progress on some of the neutrino properties like e.g. if double - chooz @xcite or daya - bay @xcite measure the third neutrino mixing angle .
on the other hand , recent important developments have also shed a new light on our understanding of neutrino propagation in dense media in presence of the neutrino - neutrino interaction ( see e.g. @xcite .
in fact , significant differences arise in some cases compared to the standard mykheyev - smirnov - wolfenstein effect @xcite . in the detection process ,
one also need to take into account the knowledge of the detector response .
if we aim at disentangling the different pieces of information , it is obviously essential that we extract as much information as possible from future observations .
while neutral current events exploiting the scattering on electrons or on protons @xcite will measure the total neutrino fluxes , sensitive to the normalisation factors ; charged - current events will be flavour sensitive to both electron neutrino and anti - neutrinos . if electron anti - neutrinos can be detected through their interaction on protons , that is known theoretically at all orders ,
the measurement of electron neutrinos require interaction on nuclei that is affected by theoretical uncertainties . here
we will briefly discuss the need for future measurements with intense low energy neutrino beams .
the possible existence of cp violation in the lepton sector is one of the crucial open issue in neutrino physics that can have tremendous impact on other domains as well .
if the value of the third neutrino mixing angle turns out to be very small , only `` third generation '' accelerator neutrino experiments , namely super - beams , beta - beams and neutrino factories @xcite , will reach the necessary sensitivity to explore this question . and @xmath11 at 1000 km from the neutrinosphere .
the different curves correspond to the cases of normal ( nh ) or inverted ( ih ) hierarchy and large ( sin@xmath12 , la ) or small ( sin@xmath13,sa ) third neutrino mixing angle @xcite .
[ fig : nutau ] ] it is important to search for other strategies to learn about the value of the cp violating phase , like for example with neutrino astronomy @xcite . in a recent work @xcite we have investigated both analytically and numerically possible cp violating effects in the supernova environment .
the neutrino fluxes within the supernova are obviously @xmath14 with the luminosities at the neutrinosphere @xmath15 given by either fermi - dirac or power - law distributions , but for electron neutrino fluxes . in this case @xmath16 and @xmath17 neutrino fluxes are taken different at the neutrinosphere by assuming that their temperatures differ by 1 mev as an example @xcite .
[ fig : nue ] ] but for the electron anti - neutrino fluxes @xcite .
[ fig : anue ] ] analytically one can show that the following equation holds @xcite : @xmath18 if @xmath16 and @xmath17 have equal luminosity at the neutrinosphere ( true at tree level ) , eqs.([e : flux ] ) and ( [ sq2 ] ) imply that there are no cp violation effects on the electron ( anti-)neutrino fluxes for any density profile , as in vacuum .
this conclusion is in agreement with the findings in @xcite .
however higher order corrections to neutrino scattering on matter and/or physics beyond the standard model such as flavour changing interactions can differentiate @xmath19 and @xmath20 .
in this case the electron ( anti-)neutrino fluxes do depend on the cp violating phase eqs.([e : flux])-([sq2 ] ) . we have performed for the first time numerical calculations of the neutrino propagation in a dense environment with @xmath21 ( for details see @xcite ) .
figure [ fig : nutau ] presents the effects on the @xmath17 fluxes as an example .
as one can see the presence of the phase modifies them significantly , even though the effects on the muon and tau neutrino fluxes unfortunately cancel out once one consider neutral current interactions since all neutrino fluxes add .
figure [ fig : nue ] and [ fig : anue ] show the effects on the electron neutrino and anti - neutrino fluxes respectively once we assume that @xmath22 .
such effects are interesting since the electron ( anti-)neutrinos play a specific role both for the nucleosynthesis of heavy elements during the r - process and for a possible signal in a detector where electron ( anti-)neutrinos can be measured through charged - current events . we have also investigated indirect impact on the r - process by calculating how the electron fraction is modified : the effects turn out to be tiny ( figure [ fig : ye ] ) .
finally we have determined the possible impact on the neutrino signal in a detector on earth . while the @xmath21 effect on the total number of events is of the order of 2@xmath23 ,
variations up to 5@xmath2 are obtained for the events as a function of neutrino energy ( figure [ fig : events ] ) . , and to inverted and large ( dot - dashed ) or small ( dotted ) @xmath24 @xcite .
[ fig : ye ] ]
neutrino - nucleus interactions is a topic of current great interest for e.g. understanding the neutrino detector s response , the nucleosynthesis of heavy elements and of the isospin and spin - isospin nuclear response ( see e.g. the reviews @xcite ) .
for example it is shown in @xcite that such measurements furnish a new constraints for neutrinoless double - beta decay predictions that are at present affected by significant variations . in supernova
neutrino observatories various nuclei are exploited as neutrino targets : carbon in scintillator detectors , oxygen in water erenkov , lead in lead - based detectors , argon in liquid argon , deuteron in heavy water .
deuteron is by far the best case since it can be predicted theoretically with errors of a few percent in effective field theory .
carbon is the best studied case experimentally although the inclusive cross sections are still affected by significant variations ( see e.g. @xcite ) .
the only other available measurement is on iron .
theoretical predictions employing microscopic models ( in particular the shell model and the qrpa ) can present discrepancies arising for example from the treatment of the continuum and from different contributions of the forbidden multipoles ( the spin - dipole and higher multipoles ) . future measurements with very intense low energy neutrino beams can provide important information that would put the theoretical predictions on neutrino - nucleus cross sections on a firmer ground .
two avenues are possible : either an intense conventional source ( neutrinos from muon decay - at - rest ) or low energy beta - beams @xcite .
intense conventional sources will be available at @xmath25 sns at los alamos @xcite and at jparc .
a low energy beta - beam facility might be established , in particular if a beta - beam for the search cp violation in the lepton sector is built .
it is shown in @xcite that this might require a devoted storage ring while a less expensive option can be to use one / two detectors at off - axis of the storage ring planned for cp violation @xcite . by varying the lorentz boost of the ions one can try to disentangle the contribution of the forbidden multipoles for which little information is available right now @xcite . besides improving our knowledge of the supernova detector s response ,
other interesting supernova neutrino applications are possible wth low energy beta - beams . in particular combining measurements at different boosts one could reconstruct the neutrino signal from a supernova explosion @xcite .
phase , associated to electron anti - neutrino scattering on protons in a water erenkov detector on earth , such as super - kamiokande ( 22.5 kton fiducial volume ) .
a 100 @xmath2 efficiency is assumed .
the curves show the effects induced by a non - zero cp violating phase , i.e. @xmath26 ( solid ) , @xmath27(dashed ) , @xmath28 ( dot - dashed ) @xcite .
[ fig : events ] ]
supernova neutrino experiments can bring crucial information , if a galactic supernova explosion occurs , or a future large - scale observatory observe an extragalactic event and measure relic supernova neutrinos .
such observations can help unravelling important features of the supernova dynamics , or of the still unknown neutrino properties .
important theoretical developments are also ongoing , in particular in our understanding of neutrino propagation in dense matter , since the neutrino - neutrino interaction included only recently in the calculations produces completely new features in some cases . here
we have discussed the possible indirect and direct ( on a neutrino signal ) impact of cp violation in the lepton sector in the supernova environment .
our numerical calculations show that effects on the electron ( anti-)neutrino fluxes within the supernova can be as large as factors 2 - 4 , if muon and tau neutrinos have different luminosities at the neutrinosphere ( due e.g. to physics beyond the standard model such as flavour changing interactions ) .
possible effects on the electron fraction , relevant for the r - process , are tiny ; while variations up to 5@xmath2 on the events as a function of neutrino energy are found in a detector on earth .
one of the difficulties of this domain is that the information on the supernova dynamics , on the neutrino properties and the star formation rate ( for relic neutrinos ) is all entangled .
it is therefore important to have observatories sensitive both to electron neutrinos and anti - neutrinos , to extract as much information as possible from observations . in this respect , important progress still need to be made on our knowledge of the neutrino - nucleus interaction , necessary for the electron neutrino detection through charged - current reactions .
j. f. beacom , r. n. boyd and a. mezzacappa , _ phys .
* 63 * , 073011 ( 2001 ) [ arxiv : astro - ph/0010398 ] .
g. c. mclaughlin and r. surman , _ phys .
d _ * 75 * , 023005 ( 2007 ) [ arxiv : astro - ph/0605281 ] .
r. c. schirato , g. m. fuller , ( . u. ( .
lanl ) , ucsd and lanl ) , arxiv : astro - ph/0205390 .
_ jcap _ * 0504 * , 002 ( 2005 ) [ arxiv : hep - ph/0412046 ] .
j. p. kneller , g. c. mclaughlin and j. brockman , _ phys .
d _ * 77 * , 045023 ( 2008 ) [ arxiv:0705.3835 [ astro - ph ] ] .
hirata et al . , _ phys .
lett . _ * 58 * , 1490 ( 1987 ) .
d. autiero _ et al .
_ , _ jcap _ * 0711 * , 011 ( 2007 ) [ arxiv:0705.0116 [ hep - ph ] ] . j. f. beacom and p. vogel , _ phys . rev .
d _ * 58 * , 053010 ( 1998 ) [ arxiv : hep - ph/9802424 ] .
s. ando , j. f. beacom and h. yuksel , _ phys .
lett . _ * 95 * , 171101 ( 2005 ) [ arxiv : astro - ph/0503321 ] .
s. ando and k. sato , _ new j. phys . _ * 6 * , 170 ( 2004 ) [ arxiv : astro - ph/0410061 ] and references therein .
h. yuksel , m. d. kistler , j. f. beacom and a. m. hopkins , arxiv:0804.4008 [ astro - ph ] . m. aglietta _ et al .
_ , _ astropart .
phys . _ * 1 * , 1 ( 1992 ) .
m. malek _ et al . _
[ super - kamiokande collaboration ] , _ phys .
lett . _ * 90 * , 061101 ( 2003 ) .
[ arxiv : hep - ex/0209028 ] . c. volpe and j. welzel , arxiv:0711.3237 [ astro - ph ] .
a. g. cocco , a. ereditato , g. fiorillo , g. mangano and v. pettorino , _ jcap _ * 0412 * , 002 ( 2004 ) [ arxiv : hep - ph/0408031 ]
. y. fukuda _ et al .
_ [ super - kamiokande collaboration ] , _ phys . rev . lett . _ * 81 * , 1562 ( 1998 ) [ arxiv : hep - ex/9807003 ] .
j. engel , g. c. mclaughlin and c. volpe , _ phys .
d _ * 67 * , 013005 ( 2003 ) [ arxiv : hep - ph/0209267 ] .
see http://www.snolab.ca/files/science/index.html .
x. guo _ et al .
_ [ daya bay collaboration ] , arxiv : hep - ex/0701029 .
j. t. pantaleone , _ phys .
b _ * 287 * , 128 ( 1992 ) .
s. samuel , _ phys .
d _ * 48 * , 1462 ( 1993 ) .
g. sigl and g. raffelt , _ nucl .
b _ * 406 * , 423 ( 1993 ) .
s. pastor , g. g. raffelt and d. v. semikoz , _ phys .
d _ * 65 * , 053011 ( 2002 ) [ arxiv : hep - ph/0109035 ] .
g. m. fuller and y. z. qian , _ phys .
d _ * 73 * , 023004 ( 2006 ) [ arxiv : astro - ph/0505240 ] .
h. duan , g. m. fuller , j. carlson and y. z. qian , _ phys . rev . d _ * 74 * , 105014 ( 2006 ) [ arxiv : astro - ph/0606616 ] . g. g. raffelt and a. y. smirnov , _ phys .
d _ * 76 * , 125008 ( 2007 ) [ arxiv:0709.4641 [ hep - ph ] ] .
g. l. fogli , e. lisi , a. marrone and a. mirizzi , _ jcap _ * 0712 * , 010 ( 2007 ) [ arxiv:0707.1998 [ hep - ph ] ] .
s. hannestad , g. g. raffelt , g. sigl and y. y. y. wong , _ phys .
d _ * 74 * , 105010 ( 2006 ) [ erratum - ibid .
d * 76 * , 029901 ( 2007 ) ] [ arxiv : astro - ph/0608695 ] .
l. wolfenstein , _ phys .
d _ * 17 * , 2369 ( 1978 ) .
s. p. mikheev and a. y. smirnov , _
nuovo cim .
c _ * 9 * , 17 ( 1986 ) . j. f. beacom , w. m. farr and p. vogel , _ phys . rev .
d _ * 66 * , 033001 ( 2002 ) [ arxiv : hep - ph/0205220 ] . c. volpe , _ j. phys .
g _ * 34 * , r1 ( 2007 ) [ arxiv : hep - ph/0605033 ] .
w. winter , _ phys .
d _ * 74 * , 033015 ( 2006 ) [ arxiv : hep - ph/0604191 ] .
c. volpe , _ j. phys.g _ * 31 * , 903 ( 2005 ) [ arxiv : hep - ph/0501233 ] .
a. b. balantekin and g. m. fuller , _ j. phys
. g _ * 29 * , 2513 ( 2003 ) [ arxiv : astro - ph/0309519 ] . c. volpe , _ nucl .
suppl . _ * 143 * , 43 ( 2005 ) [ arxiv : hep - ph/0409249 ] . c. volpe , n. auerbach , g. colo , t. suzuki and n. van giai , _ phys .
c _ * 62 * , 015501 ( 2000 ) [ arxiv : nucl - th/0001050 ] .
p. zucchelli , _ phys .
b _ * 532 * ( 2002 ) 166 . c. volpe ,
_ j. phys .
g _ * 30 * , l1 ( 2004 ) [ arxiv : hep - ph/0303222 ] .
see http://www.phy.ornl.gov/nusns/ j. serreau and c. volpe , _ phys .
c _ * 70 * , 055502 ( 2004 ) [ arxiv : hep - ph/0403293 ] .
r. lazauskas , a. b. balantekin , j. h. de jesus and c. volpe , _ phys .
d _ * 76 * , 053006 ( 2007 ) [ arxiv : hep - ph/0703063 ] . g. c. mclaughlin , _ phys . rev .
c _ * 70 * , 045804 ( 2004 ) [ arxiv : nucl - th/0404002 ] .
r. lazauskas and c. volpe , _ nucl .
a _ * 792 * , 219 ( 2007 ) [ arxiv:0704.2724 [ nucl - th ] ] . n. jachowicz and g. c. mclaughlin , _ phys .
lett . _ * 96 * , 172301 ( 2006 ) [ arxiv : nucl - th/0604046 ] .
n. jachowicz , g. c. mclaughlin and c. volpe , arxiv:0804.0360 [ nucl - th ] . | core - collapse supernovae are powerful neutrino sources . the observation of a future ( extra-)galactic supernova explosion or of the relic supernova neutrinos might provide important information on the supernova dynamics , on the supernova formation rate and on neutrino properties .
one might learn more about unknown neutrino properties either from indirect effects in the supernova ( e.g. on the explosion or on in the r - process ) or from modifications of the neutrino time or energy distributions in a detector on earth . here
we will discuss in particular possible effects of cp violation in the lepton sector .
we will also mention the interest of future neutrino - nucleus interaction measurements for the precise knowledge of supernova neutrino detector response to electron neutrinos .
address = institut de physique nuclaire orsay , f-91406 orsay cedex , france + e - mail : [email protected],[email protected] |
recent advancement in spintronics @xcite results from an increasing interest in studying the spin - polarized transport through nanostructures @xcite .
significant part of the research effort has been focused on the physical consequences of the spin - orbit interaction ( soi ) and , in particular , on its effect in electron transport .
soi is a relativistic effect which is manifested in a coupling between electron s orbital motion and its spin @xcite .
there are various proposals for spintronic devices in which the soi plays in important role @xcite .
kondo effect is one of the best - studied examples of correlation - driven phenomena in condensed matter systems .
it consists in the screening of the magnetic moment of a localized impurity by the fermi sea of itinerant electrons @xcite .
technological advancement allowed the kondo effect to be studied experimentally in quantum dots @xcite and in magnetic atoms placed on a metallic surface @xcite . in these systems the delocalized electrons are confined to move in two dimensions .
the inversion symmetry is broken by the confining potential and the so - called rashba spin - orbit interaction ( rsoi ) gives rise to a coupling between electron s spin and momentum @xcite .
also , recently , investigation of magnetic impurities on graphene has been initiated @xcite .
thus , a problem is posed : what is the influence of the rsoi on the kondo effect in 2d system ?
several studies have addressed it and the work has been centered on the kondo model .
it is obtained from the generic anderson model by a suitable transformation which integrates out the charge degrees of freedom for sufficiently large coulomb interaction between two electrons on the impurity @xcite . in ref .
10 an effective two - channel kondo model is derived and is predicted that the rsoi does not change the kondo temperature except for some band - width effects . in a later work @xcite , an effective two - channel kondo model with additional dzyaloshinsky - moriya term is derived .
it is argued that the latter interaction produces an exponential increase of the kondo temperature .
recently , zitko and bonca @xcite on the basis of an effective single - band impurity model concluded that rsoi leads to only small variation of the kondo temperature with the strength of the rsoi .
thus , the behavior of the 2d single - impurity anderson ( or kondo ) model with rsoi is not yet fully understood . on the other hand ,
significantly less work has been done on the two - impurity anderson model in a 2d host with rsoi .
the two - impurity anderson model ( tiam ) is the simplest model that describes the interplay between the electron correlations and electron coherence which is the basic physics to be inferred from it .
there are different regimes in the model due to the competition between the direct kondo interaction and the indirect exchange ( the so - called rkky interaction ) .
the former gives rise to the single - impurity kondo effect while the latter is responsible for the relative orientation of the impurities spins .
various approaches have been applied to study the tiam - renormalization group approach @xcite , numerical computations @xcite , perturbative method @xcite , and modified perturbation theory @xcite .
recently , study of the influence of the spin - orbit interaction ( of both rashba and dresselhaus type ) on the direct kondo and rkky interactions has been performed in the framework of the two - impurity kondo model @xcite . in the present work we study the two - impurity anderson model in a 2d host with rsoi .
we use the so - called modified perturbation theory which has been successfully applied to both the single - impurity @xcite and the two - impurity anderson model @xcite . in this method
an approximate electron interaction self - energy is constructed which interpolates between the limits of strong and weak coupling to the band electrons coinciding with the exact expressions in both limits .
more specifically , we construct electron green s functions that are exact up to order @xmath1 .
this approach is not fully exact but gives very good results for only moderate numerical work .
the paper is organized as follows .
section ii introduces the two - impurity anderson model in a 2d system with rashba spin - orbit interaction .
the host part of the hamiltonian is diagonalized and the corresponding quasiparticle energies are obtained .
section iii is devoted to the construction of the impurity green s functions in the framework of the modified perturbation theory and in section iv the results of the paper are summarized .
the hamiltonian of two anderson impurities , coupled to a 2d system with rsoi , is given by @xmath2 with @xmath3 being the hamiltonian of the two - impurity anderson model @xmath4 here , @xmath5 is the annihilation operator for an electron with spin @xmath6 and energy @xmath7 residing on an impurity @xmath8 .
@xmath9 is the annihilation operator for a band electron with wave vector @xmath10 and energy @xmath11 .
@xmath12 is the coulomb repulsion energy between two electrons with opposite spins simultaneously occupying given impurity and @xmath13 is the coupling between the impurities and band electrons .
the particle - number operator is @xmath14 .
the impurities are located in positions @xmath15 with the distance between them being @xmath16 . throughout the paper
we consider only the case of two identical impurities - @xmath17 and @xmath18 .
the rashba spin - orbit interaction is described by the hamiltonian @xmath19 where @xmath20 is its strength and the phase @xmath21 is defined as @xmath22 . in the presence of the rsoi one has to introduce the angular momentum basis for the band electrons . in polar representation
the band electrons operators are developed as @xmath23 where @xmath24 is the orbital magnetic quantum number .
the hamiltonian for the band electrons is given by the sum of the third term in eq .
( [ eq1 ] ) and the rsoi .
it is diagonalized by the canonical transformation @xmath25 which introduces the chirality quantum number @xmath26 .
next , we switch to the even / odd basis states for the impurity electrons ( in the case of identical impurities ) @xmath27 . with all these transformations
the hamiltonian of the system becomes @xmath28 .
@xmath29 is the hamiltonian of the decoupled impurities and band electrons @xmath30 here , @xmath31 and @xmath32 are the renormalized band energies
. the term describing the hybridization between the impurities and the band electrons is cast into the form @xmath33\end{aligned}\ ] ] where @xmath34 and @xmath35 is the @xmath36 order bessel function@xcite .
let us point out that unlike the case of a single anderson impurity the impurities states couple to all orbital channels for the band electrons . in the former case , the impurity states couple only to the @xmath37 channel .
in order to discuss the dynamics of the impurities electrons we introduce the retarded green s functions @xmath38 ( the curly brackets denote the anticommutator ) . note that the impurity green s functions are diagonal in the even / odd basis .
the fourier transforms of the retarded green s functions define the interaction self - energies @xmath39 for the impurities electrons @xmath40^{-1}.\ ] ] @xmath41 is the hartree - fock elastic self - energy @xcite @xmath42 to proceed with further discussion of the physics of the system we are studying we must make some kind of approximation for the interaction self - energies . in this work , we choose to employ the so - called modified perturbation theory . in it
the interaction self - energy is approximated by the ansatz @xmath43 the first term is the contribution of the hartree - fock decoupling of the coulomb interaction term , @xmath44 is the second - order self - energy which is obtained in a perturbation theory with respect to the coulomb repulsion energy @xmath45 , and the coefficients @xmath46 and @xmath47 are to be determined .
the explicit expression for @xmath44 is as follows @xmath48.\end{aligned}\ ] ] the auxiliary quantities are defined as @xmath49f(\omega_2)f(\omega_3)+f(\omega_1)[1-f(\omega_2)][1-f(\omega_3)]$ ] , @xmath50 , and @xmath51 . here
, @xmath52 is the fermi - dirac distribution function and @xmath53 is the hartree - fock spectral function . in the modified perturbation theory ,
the second - order self - energy is calculated using the following expression for the hartree - fock green s functions @xmath54^{-1},\ ] ] the impurity energy level @xmath55 being replaced by an auxiliary energy level @xmath56 which will take into account various renormalizations due to the coulomb interaction and the charge transfer between the impurities and the bands . later on
we shall briefly discuss the ways to determine @xmath56 .
the electron green s functions given by eq .
( [ eq7 ] ) with the interacting self - energy from eq .
( [ eq9 ] ) may be constructed in such a way that the first four moments of the corresponding spectral functions are exactly reproduced@xcite , that is , the @xmath57 we are going to obtain will be exact up to order @xmath1 . to that goal
, we develop the exact green s function @xmath57 [ eq .
( [ eq7 ] ) ] in series with respect to @xmath58 @xmath59 the exact expression for the moment @xmath60 is given by @xmath61 where the operator @xmath62 acts as @xmath63 $ ] ( square brackets denote commutator)@xcite
. next , we develop the approximate green s function @xmath57 given by eqs .
( [ eq7],[eq9],[eq10 ] ) in series with respect to @xmath58 .
this is achieved by developing the explicit expressions for the elastic self - energy @xmath41 [ eq .
( [ eq8 ] ) ] and the second - order self - energy @xmath44 [ eq .
( [ eq10 ] ) ] in series with respect to @xmath58 .
we are able to determine the coefficients @xmath46 and @xmath47 in eq .
( [ eq9 ] ) in such a way that the first four terms in eq .
( [ eq12 ] ) coincide with the first four terms in the expansion of the approximate green s function , that is the approximate green s function reproduces exactly the first four moments @xmath64 .
let us point out that the second - order perturbation theory with respect to @xmath45 ( it corresponds to @xmath65 and @xmath66 ) reproduces only the first two moments .
we shall not give details of the calculations because they are too cumbersome to be presented here .
the calculations along these lines but in the case of the infinite - dimensional hubbard model can be found in ref .
the explicit expressions for @xmath46 and @xmath47 can be cast into the form @xmath67 @xmath68 where @xmath69 is the average number of impurity electrons in the corresponding channel @xmath70 .
it is calculated self - consistently by solving the equations @xmath71 @xmath72 are auxiliary numbers of particles defined as @xmath73 with @xmath74 from eq .
( [ eq11 ] ) . @xmath75 @xmath76 the superscript @xmath77 in the correlation function in eq .
( [ eq18 ] ) means that it is calculated at the hartee - fock level of approximation .
the correlation functions in eqs .
( [ eq17 ] ) , ( [ eq18 ] ) are obtained using the equation - of - motion method and the result is @xmath78 @xmath79 the coefficient @xmath46 [ eq .
( [ eq13 ] ) ] ensures that the high - frequency limit of the exact green s functions [ eq . ( [ eq7 ] ) ] coincides with the high - frequency limit of the corresponding approximate expressions @xcite .
the approximate green s functions have the correct limit of zero coupling between the impurities and the band electrons @xmath80 ( this limit is exactly solvable)@xcite . on the other hand , in the limit of strong coupling to the band electrons @xmath81
, @xmath57 obviously coincides with the corresponding exact expression .
thus , the green s functions we have obtained interpolate between the two exact limits of strong and weak coupling to the band electrons . in order to complete the construction of the impurity green s functions we have to choose the parameter @xmath56 . a discussion on the possible ways to fix its value in the case of the infinite - dimensional hubbard model have shown that none of the possible choices is to be preferred over the others@xcite . in a previous paper@xcite on the two - impurity anderson model ( without spin - orbit interaction )
, we have imposed the friedel sum rule to determine the value of @xmath56 .
the friedel sum rule relates the average impurity charge and the phase shift at the fermi level at zero temperature and represents an exact condition on the @xmath82 behavior of the green s function .
however , its application is confined only to the zero - temperature case while we intent to discuss the nonzero temperature behavior of the model under consideration .
therefore , we choose the value of @xmath56 from the condition @xmath83 .
we have verified that , indeed , there is no significant difference between the results obtained at @xmath84 with the two choices for @xmath56 we have considered .
now the modified perturbation theory for the tiam with rsoi is completed .
let us stress that , unlike the studies of the two - impurity kondo model , no degrees of freedom are integrated out and both the charge and the spin fluctuations are treated on equal footing in our approach . to proceed with the presentation of our results we need to specify the dispersion of the band electrons .
the band electrons have a quadratic dispersion @xmath85 where @xmath86 is the effective mass and @xmath87 is the bottom of the band .
the dispersion of the band electrons with given chirality can be cast into the form @xmath88 with @xmath89 being the rashba wave vector and @xmath90 is the rashba energy . in the presence of the rashba spin - orbit interaction
the band bottom is actually at @xmath91 .
the fermi wave vector is defined as @xmath92 ( we choose the fermi energy to be 0 ) .
we calculate the hartree - fock self - energy [ eq .
( [ eq8 ] ) ] with the dispersion @xmath93 [ eq .
( [ eq21 ] ) ] .
the imaginary parts can be computed analytically but in order to obtain the real parts one has to resort to numerical integration .
the result for the imaginary part @xmath94 is too complicated to be presented here .
we shall only point out that it has a characteristic dependence on the band bottom of the form @xmath95 ( see below)@xcite .
the numerical procedure involves the self - consistent solution of eq .
( [ eq15 ] ) .
the impurity green s functions depend on the values of the correlation functions in eq.([eq19 ] ) which in turn depend on @xmath57 .
therefore , another self - consistent procedure is required in order to compute the value of @xmath57 which is used in eq .
( [ eq15 ] ) .
the procedure is rather effective and only in the case of small impurity separations @xmath96 significant number of steps have to be performed in order to achieve self - consistency . in the following , we present our results for the set of parameters @xmath97 , @xmath98 , @xmath99 with @xmath100 .
note that the second - order perturbation theory in @xmath45 is not expected to work for such large value of @xmath45 .
the choice for the value of @xmath55 does not represent some special point for which a specific behavior can be expected ( recall that the single - impurity anderson model ( siam ) with @xmath97 and with a symmetric density of states for the band electrons is a particle - hole symmetric model ) .
we have pointed out previously @xcite that for tiam with @xmath97 one obtains different physics compared to the case of symmetric siam .
the reason is that in the case of tiam there are two different effective energy levels in the even / odd channel given by @xmath101 as well as two different elastic level widths @xmath102 ( see below ) .
thus , even with a symmetric density of states for the band electrons tiam is not a particle - hole symmetric model . moreover , in the present case , the 2d band electrons with the dispersion @xmath93 [ eq .
( [ eq21 ] ) ] are not actually a particle - hole symmetric system .
these considerations suggest that the behavior of the tiam with rsoi we are going to discuss will not qualitatively depend on any specific value of @xmath55 . in figs . 1 and 2 , we show the spectral functions ( _ i.e. _ @xmath103 ) of the impurities electrons in the even and odd channels at zero temperature @xmath84 .
the results are presented in fig . 1 for rashba energy @xmath104 and for several impurity separations @xmath105 and in fig . 2 for impurity separation @xmath106 and several values of @xmath0 .
there are features in the left - hand side of both figures ( at @xmath107 in fig .
1 ) that result from the dependence of @xmath94 on the band bottom we have mentioned above . as evidenced from the figures , the height and the width of the kondo peaks depend on both @xmath0 and @xmath105 .
the results for the dependence of the height and the width of the kondo peaks on @xmath0 and @xmath105 are presented in the insets of fig . 2 and in fig .
the @xmath105 dependence shown in fig .
3 is qualitatively similar to the dependence of the height and the width of the even / odd kondo peaks on @xmath105 for the 3d tiam without rsoi @xcite .
the height of the even / odd kondo peaks [ fig .
3 , solid lines ] increases / decreases with @xmath105 for both values of @xmath0 and reaches extremum ( maximum / minimum ) for some value of @xmath105 .
semi - quantitative understanding of this behavior has been suggested in ref .
the relevant frequency range for the kondo effect is the range of small @xmath108 ( @xmath108 less than several times the kondo temperature ) and in it one can approximate the effective even / odd elastic level widths @xmath109 by @xmath110 where @xmath111 in the case @xmath112 the effective level width is @xmath113 .
the extremums of this function are given by the zeros of @xmath114 and the first zero is @xmath115 @xcite .
thus , @xmath116 has a minimum and @xmath117 has a maximum at this value of @xmath105 . the data in fig .
3 show that the height of the even / odd kondo peaks reaches a maximum / minimum around @xmath118 in agreement with the previous discussion @xcite .
when @xmath104 the minumum / maximum of @xmath102 is realized for @xmath119 and , indeed , the height of the even / odd kondo peaks has a maximum / minimum in the vicinity of this value .
further , @xmath120 for @xmath121 @xcite and , consequently , for @xmath112 @xmath122 for this value of @xmath105 and both are equal to the elastic level width in the case @xmath123 which corresponds to the doubly - degenerate case of two identical infinitely - separated anderson impurities .
3 shows that the height of the even and of the odd kondo peak is almost equal to 1 in the vicinity of @xmath124 ( recall that the data are normalized to the corresponding results for @xmath123 _ i.e. _ the case of a single anderson impurity ) .
the width of the even kondo peak reaches maximum at @xmath125 for @xmath112 and at somewhat smaller value of @xmath126 for @xmath104 [ fig .
3 , lower panel , the dashed lines ] . in a previous study @xcite ,
we have related the occurrence of this maximum to the known change of the sign of the instantaneous spin - spin correlation function which in the case of 3d tiam without rsoi takes place at @xmath127 @xcite .
we are not aware of previous works that have calculated the instantaneous spin - spin correlation function in the model we are studying .
but , based on our previous experience , we can suppose that in the case of 2d tiam with rsoi the instantaneous spin - spin correlation function changes sign in the vicinity of @xmath125 for @xmath112 and at @xmath126 for @xmath104 being ferromagnetic / antiferromagnetic for lower / higher values of @xmath105 .
the width of the kondo peak in the odd channel is a monotonically increasing function of @xmath105 [ fig .
3 , upper panel , the dashed lines ] . it has a maximum at the same value of @xmath105 at which a minimum of the height of the odd kondo peak is realized ( see above ) . at the same value of @xmath105
there is a maximum of the height and a minimum of the width of the even kondo peak [ fig . 3 ] .
physically , maximum of the height and minimum of the width of the even kondo peak means that it is most difficult for the even kondo resonance to be destroyed by the antiferromagnetic correlations of the impurity spins at the corresponding impurity separation . on the other hand ,
the odd kondo peak has minimal height and maximal width for the same impurity separation , that is , the antiferromagnetic impurity correlations tend to most easily destroy the odd kondo resonance .
the results for the dependence of the height and the width of the even / odd kondo peaks on the rashba energy @xmath0 are presented in the insets of fig .
the @xmath0 dependence is qualitatively the same in both the even and the odd channel .
the height has a maximum at some value of @xmath0 ( different for the even and the odd channel ) and the width is a monotonically increasing function of @xmath0 .
finally , we show the results for the dependence of the kondo temperature @xmath128 on @xmath0 [ fig . 4 ] . the kondo temperature has been determined from the variation of the width of the kondo peak in the even / odd channel with the temperature .
we are not aware of previous works that have considered the reliability of obtaining @xmath128 in the framework of the modified perturbation theory .
nevertheless , we think that this approach provides reasonable predictions for the dependence of @xmath128 on the model parameters . as evidenced from fig . 4 , @xmath128 has a non - monotonic dependence on @xmath0 for sufficiently small impurity separation @xmath129 . for so small impurity separation
there is a significant difference between the kondo temperatures in the even and in the odd channels with @xmath130 .
recall that in the limit @xmath131 ( the two impurities merge at some point ) the odd channel disappears and the even channel corresponds to one impurity with elastic level width @xmath132 . for larger values of @xmath133
the dependence of @xmath128 is almost linear . for larger impurity separations the even and the odd kondo temperature are practically equal and their @xmath0 dependence is monotonic - @xmath128 increases almost linearly with @xmath0 .
thus , we predict that the inclusion of the rsoi will not lead to an exponential increase of @xmath128 . for larger values of @xmath0
the kondo temperature increases almost linearly for any value of @xmath134 similarly to the result obtained in a different approach in the single - impurity anderson model @xcite . to conclude
, we have constructed a modified perturbation theory for the 2d two - impurity anderson model with rashba spin - orbit interaction .
the ansatz interacting self - energy is chosen in such a way that the impurity green s functions are exact up to order @xmath1 .
the height and width of the kondo peaks in the even / odd channels are obtained numerically as functions of the inter - impurity distance and the strength of the rashba spin - orbit interaction .
it is predicted that the kondo temperature will not have an exponential increase with the rashba energy instead having only almost linear dependence on it .
* g. e. w. bauer and l. w. molenkamp ( eds ) , _ focus on spintronics in reduced dimensions _
, new j. phys .
* 9 * ( 2007 ) .
j. fabian , a. matos - abiague , c. erstler , p. stano , and i. zutic , acta physica slovaca * 57 * , 565 ( 2007 ) .
d. d. awschalom , d. loss , and n. samarth ( eds ) , _ semiconductor spintronics and quantum computing _ ( springer - verlag berlin heidelberg 2002 ) .
r. winkler , _ spin - orbit coupling effects in two - dimensional electron and hole systems _
( springer - verlag berlin heidelberg 2003 ) .
a. c. hewson , _ the kondo problem to heavy fermions _
( cambridge univ .
press 1993 ) . d. goldhaber - gordon , h. shtrikman , d. mahalu , d. abusch - magder , u. meirav , and m. a. kastner , nature * 391 * , 156 ( 1998 ) ; s. m. cronenwett , t. h. oosterkamp , and l. p. kouwenhoven , science * 281 * , 540 ( 1998 ) .
j. li , w. d. schneider , r. berndt , and b. delley , phys .
* 80 * , 2893 ( 1998 ) ; v. madhavan , w. chen , t. jamneala , m. f. crommie , and n. s. wingreen , science * 280 * , 567 ( 1998 ) .
e. i. rashba and v. i. sheka , sov .
solid state * 3 * , 1257 ( 1961 ) . v. w. brar et al . , nature phys .
* 7 * , 43 ( 2011 ) .
j. malecki , j. stat .
phys . * 129 * , 741 ( 2007 ) .
m. zarea , s. e. ulloa , and n. sandler , phys .
lett . * 108 * , 046601 ( 2012 ) .
r. zitko and j. bonca , phys .
b * 84 * , 193411 ( 2011 ) .
h. r. krishna - murthy , j. w. wilkins , and k. g. wilson , phys .
b * 21 * , 1003 ( 1980 ) ; c. jayaprakash , h. r. krishna - murthy , and j. w. wilkins , j. appl .
phys . * 53 * , 2142 ( 1982 ) .
r. m. fye , j. e. hirsch , and d. j. scalapino , phys .
b * 35 * , 4901 ( 1987 ) .
g. e. santoro and g. f. giuliani , phys .
b * 49 * , 6746 ( 1994 ) . t. i. ivanov , phys . rev .
b * 62 * , 12 577 ( 2000 ) .
d. f. mross and h. johannesson , phys .
b * 80 * , 155302 ( 2009 ) ; h. johannesson , d. f. mross , and erik eriksson , mod . phys
b , * 25 * , 1083 ( 2011 ) .
a. martin - rodero , f. flores , m. baldo , and r. pucci , solid state commun . * 44 * , 911 ( 1982 ) ; d. meyer , t. wegner , m. potthoff , and w. nolting , physics b * 270 * , 225 ( 1999 ) . in deriving eq .
( [ eq6 ] ) we have used the well known expansion @xmath135 . in order to obtain this result we have used the relations @xmath136
j^2_m(z)=1 \pm j_0(2z)$ ] .
m. potthoff , t. wegner , and w. nolting , phys .
b * 55 * , 16 132 ( 1997 ) .
e. janke , f. emde , and f. lsch , _ tafeln hherer funktionen _
xiii ( b. g. teubner verlagsgesellschaft , stuttgart 1960 ) . | we have studied the two - dimensional two - impurity anderson model with additional rashba spin - orbit interaction by means of the modified perturbation theory .
the impurity green s functions we have constructed exactly reproduce the first four spectral moments .
we discuss the height and the width of the even / odd kondo peaks as functions of the inter - impurity distance and the rashba energy @xmath0 ( the strength of the rashba spin - orbit interaction ) . for small impurity separations
the kondo temperature shows a non - monotonic dependence on @xmath0 being different in the even and the odd channel .
we predict that the kondo temperature has only almost linear dependence on @xmath0 and not an exponential increase with @xmath0 . |
irs - tr 12002 ( sloan & ludovici 2012b ) described the generation of a `` truth '' spectrum of the k giant hr 6348 for use with short - low ( sl ) module of the irs .
truth spectra are fully calibrated spectra of a standard star which can in turn be used to calibrate other spectra .
the approach was to use sl observations of hr 6348 , combined with observations and kurucz models of the a dwarfs @xmath0 lac and @xmath1 umi .
this report describes the procedure for the long - low ( ll ) module and the assembly of full low - resolution spectra using both sl and ll .
the resulting spectra for hr 6348 and hd 173511 can be used as truth spectra for the calibration of all ll observations .
the method for ll is generally similar to that for sl .
the detector settings for ll were altered at the beginning of irs campaign 45 ( 29 october , 2007 ) , requiring different calibrations for data taken through campaign 44 and data taken from campaign 45 to 61 , when the cryogenic _ spitzer _ mission ended .
as described in irs - tr 12002 , we begin the calibration process with a kurucz model of an early a dwarf supplied to the irs team by m. cohen , shifted to the wavelength grids and resolutions of sl and ll , and scaled photometrically to match the red peak - up ( pu ) photometry of @xmath0 lac and @xmath1 umi .
we first generated a calibrated ll spectrum of hr 6348 using data from campaign p in the science verification phase and campaigns 144 during normal science operations .
we then used this spectrum to calibrate all of the other data from the same time period and from those data constructed a calibrated spectrum of hd 173511 .
this star has a spectral class of k5 iii ; its later spectral class leads to a deeper sio absorption band at 8 compared to hr 6348 , and for this reason we have avoided its use to calibrate sl .
it is still a good standard for ll , and because of the limited calibration data available from the change in detector settings for ll in campaign 45 to the end of the cold _ spitzer _ mission , both stars are needed to produce a spectral correction with a high signal / noise ratio .
the procedure follows the general outline used in sl , where we combined the calibrations of hr 6348 using @xmath0 lac and @xmath1 umi order by order , assembled the orders , and then mitigated for any remaining detectable artifacts .
however , the detailed procedure differs in several ways from sl , due to differences in instrumental performance and stellar emission in the two wavelength regimes .
we used all of the hr 6348 spectra observed from irs campaign p through campaign 44 . for sl , we excluded poorly pointed data , but for ll , the throughput is largely unchanged from one pointing to the next because the pointing errors are small compared to the larger size of the ll slit ( 10 ) .
we also changed how we combine the spectra from the two nod positions . in sl ,
we carefully compared the separate spectra , avoiding structure in one nod position not seen in the other . in ll , however , the spectra from the two nods show similar structure on fine wavelength scales , and we have simply averaged them . in sl , we used a spline - smoothing algorithm to force the spectra calibrated with @xmath0 lac and @xmath1 umi to a similar averaged shape , and again carefully avoided structure in one calibration not seen in the other . here
again the ll procedure differs .
both calibrations showed similar degrees of structure , so both were used at all wavelengths .
however , the overall shapes of the two spectra differ , with the spectrum of hr 6348 dropping more steeply to longer wavelengths when calibrated with @xmath0 lac than with @xmath1 umi . from 15 to 37 , the two calibrations have a 3% difference in slope .
figure 1 shows the spectrum of hr 6348 as calibrated with the two a dwarfs in ll order 1 ( ll1 ) .
the calibration based on @xmath0 lac actually falls more steeply than a rayleigh - jeans tail ( which would be horizontal in the units plotted ) and is not physically realistic .
the most likely explanation is that the irs spectrum of @xmath0 lac includes a red excess , which while low - contrast , appears to be real .
it probably arises from a debris disk .
consequently , we use the spline averaging in all three ll orders to force the overall shape of the calibration based on @xmath0 lac to align with the calibration based on @xmath1 umi while preserving the spectral structure on finer wavelength scales .
figure 1 illustrates the construction of the ll1 spectrum of hr 6348 . for wavelengths between 20 and 36
, the structure in the spectrum generally deviates by less than @xmath31% from a rayleigh - jeans tail . moving past 36 ,
the quality of the data grow rapidly worse with increasing wavelength .
the apparent improvement at 40 is misleading . beyond 40
, a second - order spectrum overlaps the first order , leading to contamination of the spectrum with emission from wavelengths of @xmath220 .
figure 2 shows the construction of the ll2 spectrum of hr 6348 .
the procedure is identical to ll1 , with the calibration using @xmath0 lac forced to the general shape of the @xmath1 umi calibration to compensate for the red excess in the former . excluding the endpoints of the spectrum , all of the structure is less than 1% of the continuum , peak - to - peak .
below @xmath217 , the structure is intrinsic to hr 6348 and is due to oh band absorption .
figure 3 illustrates the construction of the spectrum of hr 6348 in the ll - bonus order .
the procedure is identical to that used in the other two orders .
lcc + & * full range ( ) * & * useful range ( ) * + ll2 & 13.9021.27 & 13.9820.54 + ll - bonus & 19.2321.61 & 19.4021.10 + ll1 & 19.7442.27 & 20.5936.85 + figure 4 presents the full ll spectrum of hr 6348 after mitigating for the remaining problems .
a 4400 k engelke function fitted to the range of valid data in ll1 serves as a good reference ( engelke 1992 ) .
we have also used it to replace invalid data at the ends of some of the segments . in ll1 , this includes all data past 37 , where the combination of dropping responsivity in the detector and the 1/@xmath4 dependence of the emission from the star leaves the s / n of our coadded spectrum too low to be trustworthy .
we have also used the fitted engelke function to replace other data outside the useful range of the ll orders where those data are unreliable .
these ranges include the first four wavelength elements of ll1 ( 19.7420.25 ) , and for the ll - bonus order , the first element ( 19.23 ) and the last two ( 21.4321.61 ) .
we have left the remaining data outside the useful ranges alone because they are close enough to the expected values that replacement is unwarranted .
figure 4 plots the ll1 data before and after a correction for residual fringing .
this fringing arises from interference caused by partial delamination of the ll1 filter prior to launch .
it generally cancels out when a spectral correction is applied , but residuals can remain due to differences in pointing between a science target and the suite of calibrators . in hr 6348 , this residual fringing survives at a level of @xmath20.6% .
because the oh band structure apparent in ll2 continues into ll1 with a periodicity similar to the fringing , we can not use the hr 6348 data to determine a fringing solution .
instead , we have used ll1 spectra of two red standards , mrk 231 and iras 07598 + 6508 . for these sources , we fitted the continuum with a sixth - order polynomial and assumed that any deviations were due to fringing .
we reduced the peak - to - peak amplitude of the resulting correction by a factor of two before applying it to hr 6348 to account for the stronger fringing observed in the red standards .
we also limited our correction to wavelengths below 32 because at longer wavelengths , the correction appeared to add more structure and/or noise to the spectrum than it removed .
the engelke function in fig .
4 provides a means of assessing the quality of the data in ll1 , where the largest diversions are due to noise .
the decreasing s / n from 21 to 37 is readily apparent .
below @xmath229 , the data are generally within @xmath30.5% of the engelke function , and beyond @xmath229 , the bounds are about @xmath31.0% . as mentioned above
, the noise beyond 37 has led us to use an engelke function as the best means of estimating the true spectrum of hr 6348 despite its limitation of including no spectral structure .
the ll - bonus spectrum is within 0.5% of the engelke function , although at these wavelengths , the actual spectrum of hr 6348 appears to be deviating from the shape of the engelke function .
this trend continues into ll2 , and because it reflects the differences in the shapes of the spectra of hr 6348 and the a dwarf @xmath1 umi , it indicates that the engelke function does not accurately portray the shape of an early k giant over such a broad wavelength range .
most of the structure in the spectrum of ll2 arises from oh molecules and is probably real .
the exception is the apparent emission feature at 18 , which is likely to be an artifact .
even then , it is smaller than 0.5% .
we conclude that below @xmath229 , the spectroscopic ( point - to - point ) fidelity of our calibrated ll spectrum of hr 6348 is 0.5% or better . beyond
, the spectroscopic fidelity grows worse , with a limit of @xmath21% at 33 and indeterminant beyond 37 .
to produce a full low - resolution spectrum of hr 6348 , we ( 1 ) combined the ll spectrum described here with the sl spectrum described by irs - tr 12002 , ( 2 ) shifted the ll1 spectrum multiplicatively to match the red pu photometry determined by irs - tr 11002 ( sloan & ludovici 2011 ) , and ( 3 ) shifted the remaining orders multiplicatively to eliminate discontinuities between them .
lc + & * multiplicative correction * + sl2 and bonus & 1.0108 + sl1 & 1.0103 + ll2 and bonus & 0.9937 + ll1 & 0.9950 + figure 5 illustrates the process , and table 2 lists the multiplicative corrections applied to each segment .
the correction for ll1 is determined by forcing synthetic photometry from the spectrum to match the actual red pu photometry ( irs - tr 11002 ) .
the normalizations of sl2 to sl1 and ll2 to ll1 take advantage of the fact that the bonus orders are observed simultaneously with the second - order spectra and overlap the first - order spectra .
the normalization between sl1 and ll2 was determined by hand to eliminate any discontinuities across the boundary .
the largest order - to - order correction , between sl1 and ll2 , was 1.7% , with sl1 having to come up to meet ll2 , as is usually the case due to the pointing - induced loss of flux in the smaller sl slit .
we recommend that this spectrum of hr 6348 serve as the truth spectrum to calibrate all of the sl data obtained by the irs , and all ll data prior to the change in detector settings at the beginning of irs campaign 45 .
the limited number of observations of hr 6348 from campaign 45 to the end of the cryogenic mission drives the need for a second standard to calibrate the ll data after the change in detector settings . using hr 6348 as the standard , we calibrated all sl observations of hd 173511 and all ll observations up to campaign 44 ( inclusive ) . following the usual rejection procedures for the sl data ( see irs - tr 12001 , sloan & ludovici 2012a ) and using all of the ll data
, we generated the spectrum shown in gray in figure 6 .
lc + & * multiplicative correction * + sl2 and bonus & 1.0438 + sl1 & 1.0304 + ll2 and bonus & 0.9931 + ll1 & 0.9907 + as with hr 6348 , we have replaced some data at the ends of the orders in the spectrum of hd 173511 .
we use a 3800 k engelke function to replace all data from 35.83 to the red in ll1 .
we replaced the first four elements of ll1 ( 19.7420.25 ) with data from ll2 , and the two red - most elements in ll2 ( 21.1821.27 ) and the two red - most elements in the bonus order ( 21.44 - 21.61 ) with data from ll1 .
the renormalization of the spectrum follows the procedure for hr 6348 . to reproduce the red pu and mips 24- photometry
, we shifted ll1 by almost 1% ( see table 2 in irs - tr 11002 ) .
we then brought the remaining spectral segments into alignment .
table 3 gives the multiplicative corrections .
the 34% shifts in sl , while larger than for hr 6348 , are typical corrections for low - resolution irs spectra . the resulting spectrum should have a spectroscopic fidelity similar to hr 6348 in all orders except ll1 .
the oh bands are stronger in hd 173511 than in hr 6348 , and we have not attempted to determine if the observed structure in ll1 is due to oh or to fringing . despite this uncertainty , the deviations from a smooth spectrum in ll1 are smaller than @xmath30.5% for @xmath5 33 and only slightly worse past that point . beyond 35.8 ,
we have applied a 3800 k engelke function and do not estimate the fidelity .
we recommend the use of this truth spectrum , in conjunction with hr 6348 , to calibrate ll irs observations obtained in campaign 45 or later ( i.e. 2007 october 29 and on ) .
this report describes the construction of a truth spectrum for the early k giant hr 6348 from irs observations of it and the a dwarfs @xmath0 lac and @xmath1 umi .
this spectrum required mitigation for ( 1 ) artifacts in the vicinity of pfund-@xmath0 at 7.45 , ( 2 ) a red excess apparent in the ll spectrum of @xmath0 lac , possibly due to a debris disk , and ( 3 ) fringing in ll1 in the spectrum of hr 6348 . using hr 6348 as a standard , we also calibrated a truth spectrum for the late k giant hd 173511 .
lccrccl + & * intended * & * actual * & * @xmath6 * & * intended * & * actual * & * @xmath6 * + & * @xmath7 ( mjy ) * & * @xmath7 ( mjy ) * & * ( % ) * & * @xmath8 ( mjy ) * & * @xmath8 ( mjy ) * & * ( % ) * + hr 6348 & 235.0 & 235.0 & 0.00 & 209.4 & 209.5 & 0.05 + hd 173511 & 246.9 & 246.7 & @xmath90.08 & 220.0 & 220.2 & 0.09 + table 4 compares the intended flux densities to the actual results of synthetic photometry for irs red peak - up and mips at 24 for the truth spectra of both k giants . in the 2224 range ,
the differences are less than 0.1% , as designed .
the accuracy of the spectra at other wavelengths depends on the accuracy of the shape . to longer wavelengths ,
the spectra follow an engelke function to within @xmath31% or better ( fig . 5 and 6 ) .
to shorter wavelengths , both spectra diverge by more than 1% from the expected continuum shape .
hr 6348 drops more steeply to shorter wavelengths . for hd 173511 ,
if we assume that the local peak at 7 , between the co absorption to the blue and sio absorption the red , is slightly below the actual continuum in that region ( due to overlap of wings of the two molecular bands ) , then this spectrum does not drop steeply enough , perhaps by @xmath22% .
these small errors in the shape of the sl are a natural consequence of small throughput errors from the individual pointings coadded to generate the plotted spectra , and these throughput errors , which are a function of both wavelength and pointing , have been the subject of many previous technical reports ( see irs - tr 12001 for a recent analysis ) .
ultimately , they must lead to limits on the photometric fidelity ( i.e. , overall shape ) of the shape of any spectrum observed with sl . | this report describes the generation of fully calibrated spectra of the k giants hr 6348 and hd 173511 from data obtained with the low - resolution modules of the infrared spectrograph ( irs ) , with an emphasis on the spectra from the long - low ( ll ) module .
the spectra were calibrated using kurucz models and irs observations of the a dwarfs @xmath0 lac and @xmath1 umi .
the calibration process required mitigation for fringing in the first - order ll spectrum and a faint red excess in @xmath0 lac which may arise from a low - contrast debris disk .
the final calibrated spectrum of hr 6348 has a spectroscopic fidelity of 0.5% or better below 29 , with an uncertainty increasing to @xmath21% at 3337 .
the final calibrated spectrum of hd 173511 has a spectroscopic fidelity of @xmath20.5% at all wavelengths below 35.8 . |
over the last decades , v@xmath0o@xmath1 has been investigated extensively by both experiment and theory and still is a topic of great interest ( see , e.g. , refs .
@xcite and the literature given therein ) .
today , v@xmath0o@xmath1 thin films are receiving growing attention , particularly because the thin film geometry appears to strongly modify the electronic properties of the material as is most clearly demonstrated by the famous metal - insulator transition observed in bulk v@xmath0o@xmath1 as a function of temperature , pressure or doping @xcite .
it is known that even a small alteration of the crystal structure , by cr substitution @xcite or application of external pressure @xcite , can have a serious effect on the metal - insulator transition in the bulk material .
therefore , it would not at all be surprising if structural modifications induced by the thin film geometry had similar severe implications .
recently , the observation of a thickness - dependent metal - insulator transition in ultrathin v@xmath0o@xmath1 films was attributed to the increasing @xmath4/@xmath5 lattice parameter ratio ( decreasing @xmath5 , increasing @xmath4 ) because of interaction with the substrate @xcite . however
, this conjecture has been refuted by both x - ray diffraction studies for high quality films @xcite and density functional calculations @xcite .
in fact , the experiments point to a very complex thickness - dependence of the lattice parameters with unsystematic changes of the @xmath4 lattice constant for film thicknesses of @xmath3 1001000 , thereby calling for a more in - depth study of the electronic properties of these films .
indeed , in other systems , like the laalo@xmath1/srtio@xmath1 heterostructure @xcite , it has been shown that interfaces can seriously modify the electron density profile .
a strong dependence of the electronic properties on the layer thickness has been also reported for the hexagonal layer compounds such as @xmath6 @xcite . in the present case a systematic investigation is still missing . in this paper , we aim at clarifying the influence of the vacuum interface on the electronic states of v@xmath0o@xmath1 thin films . at this interface
, two possible scenarii may cause a significant reorganization of the electronic states : ( 1 ) a local modification of the crystal structure and ( 2 ) a charge redistribution due to the broken symmetry at the vacuum interface . in general , charge redistribution due to a surface , similar to heterointerfaces
, may induce a considerable deviation from the bulk properties , see , e.g. , refs .
even though possibility ( 2 ) therefore seems to be promising , it is the aim of this paper to point at the limitations of such a mechanism for film thicknesses exceeding some 15 .
the calculations are based on density - functional theory and the local density approximation ( lda ) .
they were performed using the scalar - relativistic implementation of the standard augmented spherical wave ( asw ) method ( see ref . , chap . 2 of ref . and references therein ) .
in the asw method , the wave function is expanded in atom - centered augmented spherical waves , which are hankel functions and numerical solutions of schrdinger s equation , respectively , outside and inside the so - called augmentation spheres . in order to optimize the basis set
, additional augmented spherical waves were placed at carefully selected interstitial sites .
the choice of these sites as well as the augmentation radii were automatically determined using the sphere - geometry optimization algorithm @xcite .
self - consistency was achieved by a highly efficient algorithm for convergence acceleration @xcite .
the brillouin zone integrations were performed using the linear tetrahedron method with an increasing number of * k*-points in order to check the convergence with respect to the granularity of the * k*-point grid @xcite .
the v@xmath0o@xmath1 surface was simulated by means of a fivefold superstructure of the canonical hexagonal unit cell .
the latter comprises six formula units , i.e. six v layers along the ( hexagonal ) @xmath4 axis .
these layers are separated by o layers , giving rise to vo@xmath7 octahedra . for further details
see , e.g. , refs .
@xcite and the references therein .
the lattice constants and fractional coordinates of the atoms were taken from ref .
the surface was then generated by removing all atoms from one of the five hexagonal unit cells , see fig .
[ fig0 ] , such that one of the resulting surfaces terminates with an o layer .
o@xmath1 surface .
large and small spheres represent the v and o atoms , respectively.,scaledwidth=35.0% ] this vacuum interface is the subject of the subsequent discussion . in order to separate the charge redistribution due to the broken crystal symmetry and the changes in chemical bonding at the vacuum interface from the structural relaxation effects at the v@xmath0o@xmath1 surface we will not include the latter in our considerations .
in fact , structural relaxation is restricted to a very narrow range of about five atomic layers as has been pointed out in a detailed study by kresse _
et al . _
@xcite , which also provides an elaborate overview over the relevant literature .
in contrast , we shall focus on the question whether the charge redistribution coming with the removal of bonding partners could affect the behavior of a much thicker slab of several 1001000 .
consequently , accounting for the near - surface structural modification is , on the one side , dispensable and , on the other side , disadvantageous for a systematic analysis of the decay of charge redistribution effects .
of course , in view of the strong electronic correlations present in @xmath8 , calculations beyond the standard lda seem necessary .
indeed , while lda fails to capture the metal - insulator transition of the bulk material , the latter could be successfully described using lda+dmft @xcite .
however , for the large supercell used in the present simulation of the surface the rather complex lda+dmft calculations would be beyond present day computational facilities .
even more important , the previous lda+dmft studies revealed that the strong electronic correlations come into play via the narrowing of the bands as calculated within the lda , which occurs on going from the crystal structure of the metal to that of the insulator .
hence , there is a clear signature already in the lda bands indicating the metal - insulator transition . as a consequence , although lda has turned out to be insufficient in describing the metal - insulator transition itself , it clearly signals its occurrence . for that reason
, it is well justified to omit a full lda+dmft treatment as long as the lda results are carefully interpreted .
dos for a series of v sites in a v@xmath0o@xmath1 thin film .
numbering of the v atoms corresponds to their distance from the surface.,scaledwidth=45.0% ] + + + due to the octahedral o - coordination of the v atoms in v@xmath0o@xmath1 , the v @xmath9 electronic bands ( located around the fermi level ) are subject to a splitting in two subgroups .
while the partially occupied t@xmath10 manifold is found in the energy range from @xmath11ev to 1.4ev , the unoccupied @xmath12 states extend from @xmath13ev to 3.7ev .
these two groups are clearly observed in fig .
[ fig1 ] , which depicts the site - projected v @xmath9 density of states ( dos ) as obtained for various atoms in the thin film .
atoms are numbered according to their distance from the surface with atom 1 being located directly at the surface .
we find that the shapes of the curves in fig .
[ fig1 ] are closely related to each other . only for atom 1 , which is subject to the strongest interface effects , deviations are visible
since half of the bonding partners are missing and the v o therefore is reduced . in fact , the dos curves of atoms 7 and 11 , which are separated from the surface by about 18 and 28 , respectively , virtually resemble the bulk dos shape .
for this reason , the decay of the interface - induced relaxation of the electronic states is found to be very fast . in order to establish further insight into the above - mentioned decay
, we display in fig .
[ fig2 ] v @xmath9 and o @xmath14 partial doss in a wider energy range in the left column as well as orbital resolved v @xmath9 partial doss in the right column . the nearly perfect splitting of the t@xmath10 manifold into @xmath15 and @xmath16 sub - groups traces back to the trigonal symmetry of the corundum lattice @xcite .
the partial dos of v@xmath0o@xmath1 bulk ( top row of fig .
[ fig2 ] ) is almost identical to that found for a vo@xmath7 octahedron in the thin film some 28 beneath the surface ( center row of fig .
[ fig2 ] ) .
this holds for both the o @xmath14 and the v @xmath9 states .
in contrast , the surface vo@xmath7 octahedron clearly shows distinct electronic relaxation effects of the local electronic states ( bottom row of fig .
[ fig2 ] ) , which can be attributed to the modified chemical @xmath17-@xmath18 bonding at the surface .
in particular , for those v and o atoms , which miss a bonding partner due to the surface , we observe a reduction of the @xmath9 and @xmath14 occupation , respectively , of approximately 0.08 electrons .
we point out again that in obtaining these results we did not take into account the v@xmath0o@xmath1 surface relaxation in order to distinguish between the electronic changes and the short - range structural distortions .
our results entail that the effect of the vacuum interface on the local electronic states , as reflected by the site - projected dos , vanishes at about 15 below the surface .
charge redistribution induced by the surface is therefore restricted to the same narrow surface layer , in which the structural relaxation is observed @xcite . in a next step
we turn to the discussion of @xmath19-resolved quantities and display the electronic bands in fig .
the top row of fig .
[ fig3 ] depicts the band structure of bulk v@xmath0o@xmath1 as well as the non - primitive hexagonal brillouin zone of the corundum lattice of v@xmath0o@xmath1 , which here and in the following is used for representing findings of band structure calculations .
the bulk results serve as a reference for comparison with the surface results given in the bottom left panel of fig .
[ fig3 ] .
of course , the surface band structure comprises a lot more bands due to the simulation of the surface by the fivefold supercell of the surface slab as mentioned above . in this supercell ,
the increased number of different bands reflects the larger number of inequivalent atomic sites .
v ( and o ) atoms with different distances to the vacuum are no longer equivalent from the crystallographic point of view .
furthermore , the dispersion along the high symmetry line @xmath20a is strongly suppressed due to the fact that this line corresponds to the direction perpendicular to the surface . note
that the construction of the surface slab in a fivefold supercell of the hexagonal v@xmath0o@xmath1 cell implies that the length of the line @xmath20a is reduced to a fifth and that the electronic bands are folded back . in order to enable a straightforward comparison with the bulk data , the line @xmath21-a
is thus artificially stretched to the original length in the slab band structure .
having this fact in mind , we eventually find that the surface slab reproduces the characteristics of the bulk v@xmath0o@xmath1 bands .
in particular , there is no sign of an additional state , not present in the bulk , which could be responsible for the anomalous behavior of thin films .
it has been found for v@xmath0o@xmath1 thin films , that the @xmath5 and @xmath4 constants of the corundum lattice deviate slightly , i.e. less than 1% , from their bulk values @xcite .
in addition , we have shown that the influence of the @xmath5 parameter change on the electronic structure is negligible @xcite . here
, we investigate whether modifications of the @xmath4 parameter may play a more important role by performing a surface slab calculation with the @xmath4 constant increased by 1% as compared to the bulk .
the results are depicted in the bottom right panel of fig .
[ fig3 ] , which uses the same representation as before .
comparison with the bottom left panel of fig .
[ fig3 ] shows that an elongation of the @xmath4 axis does not lead to a qualitative change of the electronic states .
the two band structures are , in fact , almost identical .
hence , we obtain that the experimentally observed alterations of the v@xmath0o@xmath1 lattice parameters in a thin film influence the electronic states only marginally , as long as they are not accompanied by a relaxation of the atomic positions .
indeed , a substantial lattice relaxation has been advocated to interpret transport measurements @xcite . to ensure that the thin film electronic structure fully resembles the bulk
a final step is necessary .
we still have to exclude differences in the contribution of the atomic orbitals to the different electronic states , since they would point at a modified chemical bonding .
orbitally weighted band structures are useful tools to address this question . fig .
[ fig4 ] o@xmath1 thin film , located @xmath22 beneath the surface .
the length of the bars shown for each state is proportional to the contributions of this atom .
states with a contribution below a certain threshold are omitted for clarity.,scaledwidth=45.0% ] gives an example of a prototypical v atom , some 28 below the surface .
very similar results are obtained for the neighboring v sites . in fig .
[ fig4 ] , the bars added to each state have a length proportional to the contribution of the @xmath9 @xmath23 orbitals of the atom under consideration .
states with minor @xmath9 @xmath23 admixture are not depicted for clarity . for this reason , some of the bands shown in the bottom left panel of fig .
[ fig3 ] are missing in fig .
we observe that the selected v site contributes to the @xmath9 @xmath23 bands in the entire energy range from @xmath11ev to 1.4ev and thus perfectly reflects the behavior of a bulk atom , which is true for any atom located at least @xmath3 15 below the surface .
charge redistribution as induced by the vacuum interface therefore has to be excluded as a possible source for the novel electronic properties of v@xmath0o@xmath1 thin films .
we finally mention that in the case of a heterosystems , like laalo@xmath1/srtio@xmath1 @xcite , different charge redistributions have been obtained in lda and lda+u calculations .
however , for the present system we do not find such a dependence . in conclusion , our findings entail that neither the observed changes of the @xmath5 and @xmath4 lattice parameters in v@xmath0o@xmath1 thin films nor the response of the electronic system to the presence of the vacuum interface can explain significant deviations from the electronic properties of bulk v@xmath0o@xmath1 .
only next to the surface , i.e. in a narrow layer of at most @xmath24 thickness , a charge redistribution is observed . the same behavior has been found in calculations for other transition metal surfaces @xcite . below the @xmath24
surface layer the electronic states are hardly modified by the vacuum interface .
lda+dmft calculations have shown that the narrowing of the v @xmath9 @xmath23 group of bands seen in the plain lda dos ( which traces back to structural alterations @xcite ) is the essential criterion for obtaining a metal - insulator transition in bulk v@xmath0o@xmath1 @xcite .
because such a narrowing is not observed in our results , they contradict a transition for v@xmath0o@xmath1 thin films of 100 - 1000 thickness , as long as a restriction of the structure relaxation to the surface layer is assumed .
reversely , only further changes of the atomic structure , probably in the entire thin film , can explain the experimental situation .
a detailed study of the crystal structure over the entire thin film is thus required to settle this issue .
the work was supported by the deutsche forschungsgemeinschaft through sfb 484 . | the electronic structure of v@xmath0o@xmath1 thin films is studied by means of the augmented spherical wave method as based on density functional theory and the local density approximation .
we establish that effects of charge redistribution , induced by the vacuum interface , in such films are restricted to a very narrow surface layer of @xmath2 15 thickness . as a consequence , charge redistribution can be ruled out as a source of the extraordinary thickness - dependence of the metal - insulator transition observed in v@xmath0o@xmath1 thin films of @xmath3 100 - 1000 thickness . |
cosmic rays ground - based detectors have been increasingly registering ground - level enhancements ( gles ) in the mev and gev ranges , leading to a very active research area .
ground based instruments detects a large variety of secondary particles produced by external primary protons ( ions ) penetrating the earth s atmosphere @xcite . in most cases ,
gles are linked with the solar activity @xcite , such as the cmes and their associated radiation storms @xcite , that is , solar energetic particles .
the register of gles as a signature of the solar activity has been very low in the current solar cycle 24 . in the previous cycle 23 , sixteen gles were observed @xcite , while only a single event was observed , at least so far , in the current cycle : the gle on may 17 , 2012 , linked with a m5.1-class flare and associated to a cme and radiation storms
. there is , however , one additional event detected with low confidence level by south pole neutron monitor ( nm ) claimed as gle and classified as gle 72 @xcite .
the detection of a gle requires some favourable conditions , such as a geo - effective condition of the solar active region from where the solar flare and its associated cme originate . in most cases only solar active regions , located around the middle west sector of the solar disc ,
has triggered gles .
in addition , the detection of a gle depends of the detector location .
for instance , the gles observed in the solar cycle 23 were observed by detectors located at regions with high latitudes , where the geomagnetic rigidity cutoff is around 1gv or smaller @xcite .
in this paper we report results from a new ground based ( 5 m above sea level ) detector , denominated as the new - tupi telescopes .
they are located at niteroi city , brazil ( @xmath0w , @xmath1s ) , within the south atlantic anomaly ( saa ) , a region characterised by an anomalous low strength geomagnetic field when compared to the geomagnetic dipole field .
the main goal of the new - tupi telescopes is the characterization of space weather effects at ground level .
we have identified particles ( muons ) in excess on new - tupi detector ( a sharp peak ) on oct 29 , 2015 , that correlates with the onset of a radiation storm , as well as with ground - based observations made by detectors located at high latitudes , such as the south pole nm , that is one of the eight nms successfully operated by the bartol research institute @xcite .
the south pole nm is strategically located at south pole , at 90 degrees of latitude , where the geomagnetic rigidity cutoff is 0.1 gev , the lowest in the world .
the signal was also observed in the icetop , an air shower detector located on the top of the icecube neutrino observatory at south pole . despite to the confidence level ,
in all cases not bigger than 2% , it is claimed that the event could be a new gle on the current solar cycle 24 and classified as the gle 73 @xcite .
the event on oct 29 , 2015 was a unusual event because there is no register of x - ray emission due to a solar blast .
however , a partial - halo cme was observed in coronagraph imagery lasco c2 on board of the soho spacecraft @xcite . that happened behind the west limb on the sun , that is , without an apparent magnetic connection with the earth , but even so , the cme has triggered a radiation storms on earth on october 29 , 2015 .
as the solar sector boundary crossing occurred when in the blast s nearby environment on oct 29 , 2015 was filled with energetic particles accelerated by the cme shock waves , we argue that the sep observed at goes and also observed at ground level , was guided from the sun to the earth through the fold of current sheet @xcite
we also reported the effects due to the earth crossing the boundary region causing short - lived geomagnetic disturbances in the imf @xcite observed in the ace spacecraft and a small decrease in the counting rates of some ground level detectors , such as the new - tupi , on oct 29 , 2015 .
in addition , a brief report on the local geomagnetic field effect on the sensitivity of a ground - based detector is presented .
this paper is organized as follows . in section 2
, we give a brief description of the local geomagnetic effect on the sensitivity of ground - based detectors .
the aim is to show why the new - tupi detector has a sensitivity close to the sensitivity of the detectors located at high latitude .
section 3 is devoted to report the observations , including in subsection 3.1 , the wsa - enlil prediction model of the heliosphere , showing of solar wind density on oct 29 , 2015 , and the earth crossing through a fold in the heliospheric current sheet .
spacecraft observations on cme and their associated shock waves , and their effects such as the radiation storm is presented in subsection 3.2 the ground level observations , specially the new - tupi observations and their correlation with the observations reported by space - borne detectors and other ground based experiments is presented in subsection 3.3 . in section 4
we present the effects on the geomagnetic field due to the passage of the earth by the boundary region of the current sheet . a small geomagnetic disturbance , below the minor geomagnetic storm threshold , observed in space - borne detectors and ground level detectors and the section 5 present our conclusions .
the paper has a appendix where a brief description of the new - tupi detector is presented .
secondary particle such as pions are produced in nuclear interactions of solar particles ( above gev energies ) in the upper atmosphere .
thus , a particle such as a electron or a muon produced by pion decay is characterized by a momentum @xmath2 , where @xmath3 is the transverse momentum and @xmath4 is the longitudinal momentum .
particles with a vertical momentum @xmath5 , produced at the altitude z , have a lateral spread at observation level as @xmath6 the transverse momentum is the main cause of the angular divergence of secondary particles .
the lateral spread by this effect is large and increases as the particle s momentum decreases .
in addition , charged particles such as electrons and muons traversing a finite thickness of matter suffer repeated elastic coulomb scattering .
the cumulative effect of these small angle scattering is a net deflection from the original particle direction .
the mean - squared angle of multiple scattering of a muon with energy @xmath7 can be approximated @xcite as @xmath8 this is about @xmath9 degrees for a muon with @xmath10 gev .
however , there is an additional source for the lateral dispersion of secondary charged particles that is neglected in most cases .
this additional lateral effect is due to the earth s magnetic field @xcite .
let us consider a bundle of protons ( @xmath11 gev ) incident vertically at the top of the atmosphere .
the charged particles produced in the atmosphere tend to travel in the downward direction . as the charged particle travels , its initial directions
is shifted by a horizontal distance @xmath12 , in the direction perpendicular to the earth s magnetic field @xmath13 @xmath14 where @xmath15 is the height of the atmosphere where the charged particle is generated , and @xmath16 is the radius of curvature of an ( initially ) vertical charged particle ( with charge @xmath17 ) travelling downward in the atmosphere with momentum @xmath5 .
thus , the deflection by the geomagnetic field of charged particles during their propagation in the atmosphere is caused by the component of the geomagnetic field perpendicular to the particle trajectory .
this effect results in a decrease in the number of collected particles and therefore the detector sensitivity .
the sensitivity of a detector due to geomagnetic field effect is inversely proportional to the transverse component of the geomagnetic field to the particle trajectory . from the horizontal and the vertical components of the geomagnetic field , ( @xmath18 ) , in a given location in which the detector operates , it is possible to obtain the transverse magnetic component for the particle propagation trajectory , for any direction , specified by the zenith @xmath19 and azimuth @xmath20 angles : @xmath21 we have analysed graphically the transverse component , @xmath13 described above for four different sites , where there are particle detectors , with different geomagnetic rigidity cutoff , including the new - tupi location , within the saa region .
the geomagnetic parameters of these sites are summarized in table 1 .
cccccc site & .geomagnetic parameters of the four places analysed here .
[ cols="^ " , ] + south pole & 90.0s & 16.74 & -52.26 & 54.9 & 0.1 + thule & 76.5n , 68.7w & 4.52 & 56.16 & 56.34 & 0.3 + athens & 37.9n , 23.7e & 26.68 & 37.36 & 45.91 & 8.5 + new - tupi & 22.9s , 43.1w & 17.89 & -14.81 & 23.22 & 9.2 + from table 1 , we can see that among the four places , the new - tupi location has the smaller total geomagnetic field , ( b(total ) ) . however , it is also the place with the higher stormer rigidity cutoff ( 9.2 gv ) , that is , in the dipole approximation of the geomagnetic field and that does not include the magnetic anomaly ( saa ) .
if we consider only particles with vertical incidence ( see third column in table 1 ) , a detector located in thule would have the highest sensitivity , because is the place with the smallest magnetic horizontal component . in this case , the sensitivity of a detector located at new - tupi site would be close to the sensitivity of the detector located at south pole . fig .
1 show the perpendicular magnetic component to the particle s trajectory to other directions , that is , as a function of the pointing direction @xmath19 and @xmath20 .
, to the particle s trajectory as a function of the pointing direction @xmath19 ( zenith ) and @xmath20 ( azimuth ) , for four places : south pole , new - tupi , thule and athens , respectively .
, width=415 ] , to the particle s trajectory as a function of the zenith angle , @xmath19 , and fixed azimuth angle : @xmath22 for the two places ( thule and athens ) located in the north hemisphere , and @xmath23 for the two places ( south pole and new - tupi ) located in the south hemisphere , respectively .
, width=302 ] clearly , a funnel effect can be seen .
for detectors installed in the southern hemisphere , the maximum sensitivity is achieved for particles coming from the south ( @xmath24 ) , and for the detectors installed in the northern hemisphere , the maximum sensitivity is achieved for particles coming from the north ( @xmath25 ) . however , this maximum sensitivity is achieved for different zenith angles , as best shown in fig .
2 . in all cases , @xmath13 decrease as the zenith angle increases , up to the point of their maximum sensitivity , @xmath26 . from there , the @xmath13 increases as the zenith angle increases .
we can also see that of increase or decrease rate of the sensitivity with the zenith angle is the same in the south pole , thule and athens .
the exception is the new - tupi location , where the increase ( decrease ) of these rates are less accentuated , that is , an anomalous behaviour .
it is possible to see that for a vertical incidence or near to this , the location of thule presents greater sensitivity , but for zenith angles above @xmath27 , the sensitivity rapidly decreases as the zenith angle increases .
the sensitivity in the south pole increases rapidly with increasing zenith angle , up to 15 degrees , where it reached the maximum sensitivity and from there the sensitivity decreases with increasing zenith angle .
a similar behavior can be seen for the case of tupi location . however , in this case , as already comment , the sensitivity increases more slowly with increasing zenith angle , reaching the maximum sensitivity only for zenith angle of 47 degrees .
we point out that for a vertical incidence , the sensitivity in new - tupi site is very close to the sensitivity at the south pole . among the four places , and for vertical incidence or around it ( @xmath28 )
, the athens s location has the bigger @xmath13 ( the smaller sensitivity ) . in the region
@xmath28 , the sensitivity at athens is smaller than the sensitivity at new - tupi , despite the vertical rigidity cutoff be smaller at athens .
however , this behavior at athens is useful to the study of solar transient events with a relatively high intensity .
this local geomagnetic field effect is taken into account during shower development in the atmosphere , in monte carlo calculations such as the fluka .
on the other hand , the point of first interaction of primary particles is also modulated by the geomagnetic field , this introduces a geomagnetic cutoff , in this case the horizontal component , @xmath13 , to particle trajectory , plays also a important role .
this effect is also included in monte carlo calculations , such as fluka .
heliospheric current sheet ( hcs ) is a thin transition zone where the polarity of the sun s magnetic field changes from plus ( north ) to minus ( south ) .
@xcite . in other words ,
the heliospheric current sheet is a thin region of space with no magnetic polarity and a large electrical current , dividing the inward polarity and outward polarity regions @xcite .
the current sheet is important because charged particles , such as solar particles and low energy cosmic rays , tend to be guided by its folds .
this is the reason why the particle density in the fold is always higher than the density out of it .
for instance , at 1 au the current sheet can produce auroras at polar regions such as the produced by cmes and coronal holes that increase the solar wind .
the heliospheric current sheet passage near earth is a relatively well predicted recurring phenomenon .
a hcs past the earth every 27 days ( rotation rate of the sun ) .
but it can has different structure after one rotation and substructures caused by out flowing transients . in addition , the main orientation can be highly distorted .
the geomagnetic storms caused by these recurring features are relatively weak , in most cases below the minor geomagnetic storm threshold .
3 shows the forecast of conditions in the solar wind density as predicted by the wsa - enlil model on oct 29 , 2015 .
the picture also shows the earth crossing through a fold in the heliospheric current sheet .
such condition has allowed a special magnetic connection between the sun and the earth .
this is evident by the fact that energetic particles from a wide cme that happened in the opposite side on the sun , but near the western limb , managed to reach the earth .
the circular plots are a view from above the north pole of the sun .
the sun is the yellow dot in the center and the earth is the green dot on the right .
the locations of the two stereo satellites are also shown .
clearly , the figure shows the conditions described above . on oct 29 , 2015 ,
the earth facing solar activity remained low , without chances of a noteworthy solar flare , as shown in the upper left - hand corner of figure 2 , where the goes x - ray flux is shown .
there is no sign of a solar flare that could trigger a cme .
however , at 2:24 ut a partial - halo cme was observed in the coronagraph imagery lasco c2 instrument aboard the soho spacecraft .
the cme emission was in the south - west sector of the sun , as show in the upper right - hand corner of figure 2 .
additional information provided by solar dynamics observatory ( sdo ) shows that a noteworthy eruption occurred behind the west limb on the far side of the sun . on the other hand ,
the angular distribution of the principal angle ( counter - clockwise from north ) of the shock waves triggered by cme , provided by cactus catalog , has shown that a fraction of them are located around 270 degrees , that is , on the ecliptic plane , as show in the lower right - hand corner of figure 2 .
this means that particles were accelerated by these shock waves and injected around the ecliptic plane .
indeed , a proton flux increase was observed on oct 29 , 2015 , around 3:00 ut on the ace - sis spacecraft ( at lagrange point l1 ) , as well as , on the geostationary goes satellite , reaching the condition of s1 level , of radiation storm in the noaa storm scale ( proton flux above 10 particles@xmath29 , with energies above 10 mev at 1 au ) , as show in the lower left - hand corner of figure 2 .
this radiation storm was the result of acceleration of particles by shock waves triggered by the cme that occurred from just beyond the sw limb of the sun around 02:19 ut . in temporal coincidence with
the onset of the radiation storm observed on the goes satellite , there was an increase in the counting rate in the south pole nm .
the south pole nm has a strategic location , at latitude @xmath30 south , and it is the detector with the lowest geomagnetic rigidity cutoff ( 0.1 mv ) on earth .
the particle enhancement at south pole nm means the presence of particles around gev energies region , probably in the tail of the energy spectrum .
however , the increase was only up to 1.5% , relative to the background , in the 10 minutes binnig count rate , and the duration of the enhancement was less than 4 hours , as shown in fig . 5 ( bottom panel ) . a search in other detectors at ground level was made and a similar signal was found in new - tupi detector , working in scaler mode , as is shown in fig . 5 ( central panel ) . the new - tupi detector is located in a place where the stormer geomagnetic rigidity cutoff is 9.2 gv .
however , the site is within of the saa central region , and as shown in section 2 , this region has a high sensitivity to detect charged particles .
the confidence level at new - tupi is above 2% , but it is only a sharp peak , a behaviour that was also found at thule nm , a detector located at high latitude on the north hemisphere ( see table 1 ) .
that is , thule nm has also a sharp peak with an increase of up to 2% and temporally coincident with the new - tupi peak .
6 summarizes the situation , where the time profiles 5 minutes binning of the new - tupi , sopo , thule and mcmurdo data are shown .
table 1 presents their coordinates and magnetic parameters .
we would like to point out that in all cases the particle enhancement was below that required to trigger a gle alert .
indeed , two gle alerts have been developed , one of them uses the eight bartol nms stations @xcite .
the signal in three of them ( south pole , thule and macmurdo ) is shown in fig .
the other one is the gle alert system developed by the national and kapodistrian university of athens - cosmic ray group @xcite , which use the esa nm stations ( souvatzoglou et al .
both systems use the same scale . on oct 29 , 2015 ,
no alert was reported , since it is necessary that three or more stations observe an increase equal to or greater than 4% .
the passage of the earth by the heliospheric current sheet is a relatively well predicted recurring phenomenon .
the earth can pass a same hcs every 27 days , that is the rotation rate of the sun , but the hcs can have different structures , as well as substructures caused by out - flowing transients .
in addition , the main orientation can be highly distorted [ e.g. villante et al . , 1979 ] @xcite .
the geomagnetic storms caused by these recurring features are relatively weak , below the minor geomagnetic storm threshold and in most cases are long - lived .
the energy fluctuations of the ions , compressed within the field reversal regions , can be interpreted as alfnic fluctuations and the main effect is a small variation in the @xmath31 component of the imf and can be weakly oriented southward , providing a magnetic reconnection with the earth s magnetic field , oriented northward , several times , but for short periods , with a 2 h average , opening cracks in the magnetosphere , promoting an injection of the solar wind in the upper atmosphere producing auroras in the polar regions .
fig . 7 ( top panel ) shows the results of measurement of the imf carried out with ace - magnetic detector in the lagrange point l1 , on oct 29th , 2015 . as expected , about one hour after of the arrival of the sep in the earth , that is , when the earth began to cross the boundary sector of the heliospheric current sheet , the @xmath31 component also began to be oriented southward , and this coincides with a small fall in the muon counting rate in the new - tupi detector , and in the thule nm , as shown in fig . 7 ( bottom panel ) ,
respectively ) .
however , in all cases the confidence level is not more than 1.5 % .
this `` mini - forbush decrease '' occurred immediately after the sharp peaks observed in both detectors and they represent the particle excess due to the arrival at earth of sep . from fig .
7 we can see also that there are at least other four oscillations in the @xmath31 component on oct 29 , 2015 . during this period
the counting rate remains in the same level after the fall , but with some small oscillations .
thus , the geomagnetic effects caused by this heliospheric boundary crossing was relatively weak , below the minor geomagnetic storm threshold , without an alert through of the geomagnetic indexes , such as the kp and dst .
the sep fluxes derived from the data of ground - based detectors are model dependent and require knowledge of response function .
even so , only the ground - based installations can provide the sufficient geometric factor to measure the low intensities of the solar protons of highest energy .
thus , ground - based measurements are complementary to the measurements performed on spacecraft . to this end
there are several world - wide networks of ground - based detectors , such as the standard nms , that are useful instruments with a huge geometric factor to observe temporal variations of galactic cosmic rays and sep with energies above @xmath32 gev .
in particular , we would like to highlight the bartol s nms located at high latitudes ( polar regions ) , consequently , they have the highest sensibilities .
within the esa neutron monitor database ( nmdb ) we would like to highlight the oulu nm with stable data since 1966 .
the oulu nm , located at finland , has a geomagnetic rigidity cutoff of 0.8 gv @xcite . in the oulu nm web page
, there is an useful gle catalogue @xcite , with information since the gle 15 , occurred on july 07 , 1966 .
the last record in the catalogue is the gle 71 on may 17 , 2012 , the only gle in the current solar cycle 24 . however , there are at least other two events that are being considered as gles .
the first was on january 6 , 2014 , catalogued as the second gle in the solar cycle 24 .
this event was primarily observed by the south pole nm ( increase of @xmath33 ) , whereas a few other neutron monitors recorded smaller increases @xcite .
even so , the event is being classified as the gle 72 .
the second event that is being classified as the gle 73 is the event on oct 29 , 2015 , analysed here . as the previous gles ,
the event was observed only by some ground - level detectors located at high latitudes ( polar regions ) , where the geomagnetic rigidity cutoff is lower .
however , in this case , there is also signal at a ground - based detector at median latitude ( the new - tupi detector ) .
we have shown ( see section 2 ) that despite the relative high rigidity cutoff ( 9.2 gv ) , the detector is located within the saa region , where the local geomagnetic field is the smallest in the world and therefore it allows a detector has high sensitivity to charged particles .
we have also shown ( see section 3.1 ) that the event has occurred when the earth crossed through a fold in the heliospheric current sheet and a fraction of the high energy particles triggered by the solar blast were directed to the earth , through the fold of the heliospheric current sheet . without this condition , probably would not exist the particle enhancements at spacecraft and ground level detectors . in most cases ,
sep observations imply the existence of coronal shocks extending at least @xmath34 @xcite and ip shocks up to @xmath35 at 1 au @xcite .
for instance , in the solar cycle 23 , most ( 80% ) of the cmes correlated with gles were full halo type , with three events non - halos , even so , having widths in the range 167 to 212 degrees @xcite .
whereas , in the event on oct 29 , 2015 , the cme had only an angular width of 114 degrees ( see the upper right corner in fig .
4 ) . in addition ,
as shown in section 3.3 , the particle enhancement observed at ground level , in all cases has a confidence level around 2% or smaller than this .
even so , it is being considered as the gle 73 .
if we used these criteria , we can claim that there are , at least , other two events that can be classified as gles in the current solar cycle , with results already published .
they are the event reported as _ signals at ground level of relativistic solar particles associated with a radiation storm on 2014 april 18 _ @xcite , and the event reported as _ signals at ground level of relativistic solar particles associated to the `` all saints '' filament eruption on 2014 _ @xcite . in both cases , the signals of several nms are reported , including the south pole nm and the tupi telescope ( the predecessor of the new - tupi detector ) . also in both cases , the signals have confidence levels above 3% .
the aim of the new - tupi detector is the study of the space weather effects at ground level , driven by diverse solar transient events .
the new - tupi telescope is located at niteroi city , rio de janeiro state ( @xmath36 ) and at 5 m a.s.l , that is , within the south atlantic anomaly ( saa ) region , and it is the region where the inner van allen radiation belt makes its closest approach to the earth s surface . as a consequence of this behaviour
, the saa region is characterized by an anomalously weak geomagnetic field strength ( less than 28,000 nt ) @xcite .
the new - tupi detector has four detectors assembled on the basis of plastic scintillator ( 150 cm x 75 cm x 5 cm ) within a trapezoidal truncated box , and a hamamatsu , model r877 , with 10 stages , 127 millimeters of diameter photomultiplier is assembled at the truncated part .
the data acquisition system is made on the basis of an national instrument daq with eighth analogical channels , 1 ms / s , 12-bit , simultaneous sampling rate .
only signals above a threshold value , corresponding to energy of about 100 mev deposited by particles ( mostly muons ) that reach the detector , are registered .
the output raw data of each detector is recorded at a rate of one hz .
1 ( left ) shows a photography of the new - tupi detector .
it can operate in the telescope mode , for this purpose , the four scintillators are placed in pairs , with t1 ( top ) and b1 ( bottom ) and t2 and b2 respectively , as shown in fig .
1 ( right ) , this layout allow to obtain the muon flux from three directions , the vertical , west and east , the last two with an inclination around 45 degrees .
the telescopes registers the coincidence rate among t1 and b1 ; t2 and b2 ( both are vertical incidence ) ; and the cross coincidences t1 and b2 ( west incidence ) and the t2 and b1 ( east incidence ) . both vertical and lateral separation between the detectors are 2.83 m. in parallel with the telescope mode , the new - tupi detector can also operate in scaler mode or mode called also as single particle technique @xcite , where the single hit rates of all of the four pmts were recorded once a second . this mode allows to search by particle excess , at fixed time intervals , even when the source is out of the field of view of the telescopes , because the field of view of a single detector is bigger than of field of view of a telescope .
however , the efficiency of particle detection in scaler mode decreases as the zenith angle of the incident particle increases , due to atmospheric absorption .
thus the scaler method is limited to incident particles with a zenith angle not more than 60 degrees .
we acknowledge the space weather prediction center ( swpc ) ; the cactus cor2 cme list and the nmdb database , founded under the european union s fp7 programme ( contract no .
213007 ) for providing data .
the data of the south pole nm ; mcmurdo nm and thule nm are provided by the university of delaware with support from the u.s .
national science foundation under grant ant-0838839 .
this work is supported by the national council for research ( cnpq ) of brazil , under grant 306605/2009 - 0 and fundacao de amparo a pesquisa do estado do rio de janeiro ( faperj ) , under grant 08458.009577/2011 - 81 and e-26/101.649/2011 .
j. a. simpson , j. a. space science reviews , 93 , 11 ( 2000 ) oh et al .
, space weather , 10 , s05004 ( 2012 ) . k. a. firoz et al . , journal of geophysics research , 115 , a09105 ( 2010 ) n. gopalswamy , et al . , arxiv:1205.0688 [ astro-ph.sr ] n. thakur et al .
, arxiv:1406.7172 [ astro-ph.sr ] m. andriopoulou et al . ,
solar physics , 269 , 155 ( 2011 ) .
takao / icetop/01m / gle73.html e. robbrecht and d. berghmans , a@xmath37a , 425 , 1097 ( 2004 ) j. m. wilcox , n. f. ness , journal of geophysical research , 70 , 5793 ( 1965 ) z. svestka , solar physics , 4 , 361 ( 1968 ) s. olbert , phrv , 96 , 1400 ( 1054 ) .
r. de los reyes et at .
et al . , int .
j. of mod .
a , 20 , 7006 ( 2005 ) . c. r. a. augusto et al , astrophys .
j. , 805 , 89 ( 2015 ) .
@xmath38 p. l. israelevich , a. i. ershkovich , and n. a. tsyganenko , j. geophys .
res . , 106 , 25919 ( 2001 ) c. foullon , et al . , solar physics , 259 , 389 ( 2009 ) k. mursula and t. hiltula , geophys .
, 30 , 2135 ( 2003 ) t. kuwabara , et al .
space weather , 4 , s10001 ( 2006 ) .
h. mavromichalaki , advances in solar and solar - terrestrial physics,2012 : 135 - 161 isbn : 978 - 81 - 308 - 0483 - 5 editors : georgeta maris and crisan demetrescu .
u. villante , r. bruno , f. mariani , l.f .
burlaga and n.f .
ness , j. geophys .
res , 84 , 6641 ( 1979 ) . | on oct .
29th , 2015 , the earth crossed through a fold in the heliospheric current sheet .
this is called a `` solar sector boundary crossing '' . under this circumstances ,
a large coronal mass ejection ( cme ) occurred at 2:24 ut , behind the west limb on the sun .
therefore , the boundary crossing occurred when in the blast s nearby environment was filled with energetic particles accelerated by the cme shock waves , spacecraft measurements ( ace and goes ) have shown that in such a case , protons with energies at least up to 30 mev were stored within the range of the sector boundary .
thus , a fraction of the solar energetic particles ( sep ) from cme , reached earth around 03:00 ut in the aftermath of the solar blast , reaching the condition of an s1 ( minor ) radiation storm level .
the effect at ground level was a small increase in the counting rate in some ground based detectors , such as the south pole neutron monitor ( nm ) and a sharp peak observed in the counting rate in the new - tupi detector in rio de janeiro , brazil and thule nm .
the event is being classified as a new gle ( ground level enhancement ) in the current solar cycle , as the gle 73 .
however , in all cases , the counting rate increase is smaller or near than 2% .
the earth crossed through a fold in the heliospheric current sheet also caused a geomagnetic disturbance , below the minor geomagnetic storm threshold , observed in the ace spacecraft and a small decrease in the counting rates of some ground level detectors , such as the new - tupi detector and thuly nm .
details of these observations are reported . |
efficiency of electron heating in high mach number collisionless shocks is one of the outstanding issues of space plasma physics as well as astrophysics .
it is known from observations that quite efficient heating occurs in extremely high mach number shocks like snr ( super nova remnant ) shocks @xcite , while in - situ observations of earth s bow shock mostly show inefficient heating of electrons @xcite .
this is probably due to the fact that a dominant electron heating process and its efficiency through a variety of microinstabilities in shock transition regions strongly depend on mach numbers .
typical mach number of the earth s bow shock is @xmath0 , while that of the snr shocks is of the order of @xmath1 . in the late 1980s a new process of strong and rapid electron heating in extremely high mach number shocks
was proposed @xcite .
6 and 7 discussed the so - called two step instabilities generated by a high velocity reflected ion beam which destabilizes bi ( buneman instability ) followed by ion acoustic instability .
a necessary condition for the bi to be destabilized in the foot of quasi - perpendicular shocks is @xmath2 @xcite , where @xmath3 is the alfvn mach number and @xmath4 the electron plasma beta .
if the above condition is satisfied , rapid electron heating with saturation temperature @xmath5 occurs and the ion acoustic instability sets in leading to further electron heating @xcite .
this condition is easily satisfied in snr shocks , although it is usually not in the earth s bow shock .
the two step instabilities were again shed light on after ref . 9 by utilizing electromagnetic full pic ( particle - in - cell ) simulations .
they improved understandings of the process not only in terms of electron heating but also in terms of a production process of non - thermal electrons , and promoted a number of subsequent simulation studies @xcite .
as a result , understandings of electron heating or acceleration processes initiated by the bi in the context of the shock physics have been extensively developed .
on the other hand , it has been well - known that various microinstabilities are possible to get excited also in relatively low mach number supercritical shocks @xcite . since
most of the microinstabilities are inseparable with electron dynamics , their nonlinear evolutions in self - consistently reproduced shock structures in pic simulations are studied only recently @xcite except for refs .
it is revealed that mtsi ( modified two - stream instability ) becomes dominant in the foot of the earth s bow shock or interplanetary shocks at @xmath6 au @xcite .
simulation studies on nonlinear evolutions of the mtsi in periodic systems were extensively studied in the 1970s and 1980s @xcite , although these simulations imposed some strong assumptions that the code was electrostatic @xcite and unmagnetized ions with small ion - to - electron mass ratio were used @xcite . recently
, some developement without these assumptions are made @xcite .
8 showed in a one - dimensional simulation with realistic mass ratio that long time evolution of the mtsi results in lower cascade of wave spectrum and associated electron heating .
furthermore , it is indicated in two - dimensional simulation that the mtsi can finally survive even if other possible instability like electron cyclotron drift instability is present @xcite .
all these simulation studies show electron heating or acceleration in the long time evolution of the system .
however , as mentioned already , electron heating in the earth s bow shock is seldom observed in - situ .
such a discripancy may arise because of lack of systematic estimate of electron heating rate in simulation studies with more realistic parameters .
for example , all of the above simulation studies on the mtsi assume very low values of squared ratio of electron plasma to cyclotron frequencies , @xmath7 .
importance of @xmath8 in association with electron heating in the long time evolution of the mtsi is unresolved , while @xmath8 is known to be a crucial parameter controlling nonlinear electron heating and acceleration processes in the bi dominant system @xcite .
electron plasma beta is also rather small in most of the previous studies ( @xmath9 ) .
this may exaggerate effects of electron trapping in the nonlinear stage of the mtsi @xcite .
the electron plasma beta in the solar wind is usually several times higher so that importance of the trapping process may be blurred .
hence , in realistic situations expected to be achieved in the typical solar wind conditions an alternative electron heating mechanism and its efficiency are ought to be considered . in this paper , first , electron heating through the mtsi in transition regions of high mach number quasi - perpendicular shocks is reconsidered in section ii . here , quasilinear diffusion is assumed as an alternative electron heating process .
the analysis assumes a broad wave spectrum and small wave amplitudes which are likely to be satisfied in the past simulation studies mentioned above @xcite .
an additional assumption that relative phases among different wave modes are random is also imposed , while this point is currently not evident . under these assumptions , taking second order velocity moments of the so - called quasilinear equations provides evolution equations of electron kinetic energies .
these equations are numerically integrated by imposing the quasilinear assumptions in every instantaneous time steps to obtain saturation kinetic energies as a fuction of the mach number of shocks .
this approach allows to discuss long time evolution of macroscopic quantities including the electron temperature without any restrictions for parameters .
it is intriguing whether the mach number dependence of electron heating throug the mtsi in relatively low mach number regime is different from that through the bi in extremely high mach number regime .
if that is the case , one may expect presence of a critical mach number above which electron heating rate runs up due to the switching of the dominant microinstability @xcite . in section iii ,
the results are compared with 1d pic simulations in which the mtsi is dominantly generated in the foot and effective electron temperature just behind of the shock is measured .
furthermore , post - shock effective electron temperature is measured also in cases that the bi gets excited in the foot . in those cases a rough estimate of the electron temperature is given by using the well - known trapping theory .
finally , a summary and discussions including estimate of the critical mach number are given in section iv .
a foot region of a quasiperpendicular shock is modeled by a plasma composed of incoming electrons at rest , incoming ions , and specularly reflected ions with local approximation .
the ambient magnetic field is along the @xmath10-axis , while the incoming and the reflected ion beams are streaming parallel and antiparallel to a wavevector which is defined in the @xmath11 plane and parallel to the shock normal .
this is essentially the same system discussed in the linear analysis of ref .
since the mtsi grows much faster than ion cyclotron period , ions are assumed to be unmagnetized .
saturation levels of electron kinetic energy through the mtsi are examined as follows . in a vlasov equation ,
@xmath12 variables are expanded around zero - th order slowly varying quantities ( with subscript 0 ) as @xmath13 where @xmath14 , @xmath15 , and @xmath16 are a distribution function , electric charge , and mass of species @xmath17 , @xmath18 is speed of light , @xmath19 and @xmath20 represent magnetic and electric fields , respectively .
the variables with subscript 1 denote fluctuations corresponding to linear oscillations . with a standard approach of quasilinear analysis , an evolution equation of @xmath21
is obtained as follows @xcite .
@xmath22 here , @xmath23 is the volume of integration , @xmath24 the cyclotron frequency , @xmath25 and @xmath26 the velocity components parallel and perpendicular to @xmath27 , @xmath28 the wavevector which is also composed of parallel and perpendicular components @xmath29 , and @xmath30 indicates real wave frequency which is a function of @xmath28 and satisfies the linear dispersion relation . the subscript @xmath17 has been eliminated for simplicity .
only resonant wave - particle interactions corresponding to zero arguments of the delta function , @xmath31 , are taken into account .
moreover , @xmath32,\ ] ] @xmath33,\end{aligned}\ ] ] where @xmath34 , @xmath35 , and @xmath36 represents the @xmath37th order bessel function with argument @xmath38 . when the above expression of @xmath39 is derived , the linear dispersion relation is taken into account .
note that in eqs.([qleq])-([psi2 ] ) parallel and perpendicular directions are defined in terms of @xmath27 .
although this is true for electrons , further note is needed for ions . because ions are assumed to be unmagnetized , distinction of parallel and perpendicular directions in this sense is meaningless .
however , their evolution equations can be formally written in a similar form by defining parallel and perpendicular directions in terms of the ion beam velocity .
this corresponds to a transformation for the rotation of the susceptibility tensor @xcite . to discuss the extended quasilinear evolution of kinetic energies through the mtsi
, we examine parallel and perpendicular kinetic energy densities defined by @xmath40 @xmath41 after some algebra , the following evolution equations of the kinetic energies are obtained .
@xmath42\end{aligned}\ ] ] @xmath43\end{aligned}\ ] ] here , @xmath44 , and @xmath45 means a corresponding value at which the argument of the delta function is zero .
the spectral energy density of the electric field fluctuations evolves as @xmath46 where @xmath47 denotes the linear growth rate of the wave mode having wavevector @xmath28 .
conservation of total energy is written as @xmath48 here , @xmath49 and @xmath50 denote density and bulk velocity of species @xmath17 .
note that @xmath51 for unmagnetized ions and @xmath52 .
further , @xmath30 and @xmath53 are defined at an instantaneous time .
eqs.([paraene])-([enecon ] ) form a closed set of extended quasilinear evolution equations of the system .
we solve the above set of equations numerically in the following subsection .
similar analyses were performed by many authors @xcite . ref .
38 introduced detailed procedure of the method and applied it to estimate of saturation of electromagnetic ion cyclotron instability driven by ion temperature anisotropy .
39 and 40 directly solved the quasilinear equation to discuss electron heating and formation of a high energy ion tail through the ion acoustic instability .
33 , on the other hand , discussed a problem similar to here on the mtsi and gave detailed comparisons of their results with 1d and 2d pic simulations .
however , their analysis includes only nonresonant wave - particle interactions which in turn only slight electron heating which is much less effective compared with a subsequent heating process through nonlinear trapping observed in their pic simulation .
furthermore , they imposed an electrostatic assumption on whole their analyses as well as on their pic simulations . as they pointed out , the electrostatic approximation is valid only when @xmath54 , where @xmath55 denotes the alfvn velocity , and @xmath56 is the electron plasma beta ( ratio of electron thermal to magnetic pressures ) .
usually in solar wind @xmath56 is of the order of @xmath57 so that the above condition reads @xmath58 which is probably satisfied only in transition regions of rather low mach number shocks . since our attention is paid for wide range of mach numbers , we exclude such an assumption . before solving eqs.([paraene])-([enecon ] )
numerically , a few assumptions are imposed .
the distribution functions of each species remain in ( shifted ) maxwellian all the time after ref .
38 . by this assumption , it is easy to obtain @xmath59 and @xmath47 at each time step as instantaneous solutions of the linear dispersion relation .
the actual distribution functions are given by @xmath60 for species @xmath17 . here
@xmath61 and @xmath62 denote parallel and perpendicular thermal velocities . as mentioned before , note that the parallel and perpendicular directions are difined in association with @xmath27 for electrons , while with @xmath63 for incoming and reflected ions .
the bulk velocities of the incoming and reflected ions ( subscripts @xmath64 and @xmath65 , respectively ) satisfy zero current condition so that @xmath66 .
it is also assumed that the response of the ions to the wave fields is electrostatic , while the wave - electron interactions are treated as fully electromagnetic .
therefore , the ion heating is thought to be parallel to @xmath67 or @xmath68 . under these assumptions ,
we obtain the following evolution equations of kinetic energies . @xmath69\end{aligned}\ ] ] @xmath70\end{aligned}\ ] ] @xmath71\end{aligned}\ ] ] here , @xmath72 and @xmath73 , respectively . in this analysis
only the linear resonant wave - praticle interactions are taken into account
. therefore , change of kinetic energy of the reflected ions ( @xmath74 ) can be neglected when we consider the mtsi based on electron - incoming ion interactions .
that is , here , the reflected ions are assumed to be a background component satisfying current and charge neutralities .
reactions of the reflected ions may be essential when influences of the mtsi on the reformation process are considered @xcite .
however , it is out of scope in this paper . in the following
we solve eqs.([eneea])-([enei ] ) , ( [ waveene ] ) , and @xmath75 instead of eq.([enecon ] ) .
eq.([econ ] ) is actually used to determine @xmath76 .
note that contributions from the terms @xmath77 for @xmath78 in eqs.([eneea ] ) and ( [ eneee ] ) are neglected .
the validity of such an assumption will be checked in the following subsection .
, width=302 ] first , let us overview in fig.[fig01 ] time evolutions of the variables calculated in the system for a typical case .
the initial parameters are @xmath79 , and @xmath80 . here , the initial temperatures are isotropic and @xmath81 .
@xmath3 and @xmath82 determine the initial @xmath83 from the zero current condition .
@xmath84 denotes the angle between the ambient magnetic field , @xmath27 , and the wavevector ( the shock normal ) , @xmath85 .
the above parameters are fixed throughout the analysis if not specified .
the field energy is given as appropriately small value : @xmath86 .
for the moment , @xmath87 is fixed to unity just for simplicity .
the field energy exponentially grows in the beginning of the run , although other variables have not been much changed in a visible manner until the field energy reaches at a certain level .
then @xmath83 gradually decreases , while @xmath88 and @xmath89 increase ( @xmath90 ) . at the same time
, the wave growth slows down and it starts damping after @xmath91 .
correspondingly , changing rates of other variables decrease and those variables approach some constant values when the field energy becomes sufficiently small .
such a state is regarded as a saturation state and those constant values are defined as saturation levels .
the saturation state in this system is achieved when the field energy is damped out . in the analysis we check whether the system saturates basically by eye . however , in some special cases where the kinetic energies have not reached at constant values for long time the saturation levels are defined as the values at @xmath92 where @xmath93 is the time at which @xmath83 decreases by @xmath94 from its initial value . since @xmath95 is thought to be a typical time scale of shock reformation , considering the system evolution longer than this time scale may not make sense . in fig.[fig01 ]
@xmath96 keeps almost constant incontrast to @xmath88 or @xmath89 .
this reflects which kinetic effects dominantly work .
the linear growth rate , @xmath97 , can be written by linear superposition of the terms such as @xmath98 where @xmath99 and @xmath100 , respectively . when @xmath101 , contribution from the corresponding term to the growth rate becomes non - negligible . in this sense @xmath102 , and @xmath103
can be indicators how strong the effects of landau and cyclotron dampings of electrons , and inverse ion landau damping are , respectively .
the bottom panel of fig.[fig01 ] represents values of these indicators .
it is now clear that the electron landau damping is the strongest kinetic effect and the electron cyclotron damping hardly works .
the inverse ion landau damping can also work , but may not be so strong as the electron landau damping for this particular case .
this can explain the fact that increase of @xmath88 is most remarkable , while that of @xmath96 is unrecognized .
contributions from other electron cyclotron harmonics ( @xmath104 with @xmath105 ) are further ineffective , and this is confirmed to be true for wide range of parameters ( not shown ) .
therefore , effects of electron cyclotron interactions will be neglected in the remaining of the analysis .
the assumption that @xmath106 is fixed is removed hereafter .
, as a function of ( upper panel ) initial @xmath107 and ( lower panel ) @xmath8 .
the three lines correspond to cases with different initial temperature ratio .
the solid line with circles denotes the case with @xmath108 , the dashed line with squares @xmath109 , and the dotted line with triangles @xmath110 , respectively .
[ fig02],width=302 ] the upper panel of fig.[fig02 ] denotes saturation levels of normalized electron temperature , @xmath111 as a function of the initial @xmath112 when @xmath113 . in calculating the right hand sides of eqs.([eneea])-([enei ] ) , we chose interval of integration with respect to @xmath114 as @xmath115 and grid size in @xmath114 as @xmath116 .
@xmath117 is again fixed here . in both upper and lower panels the solid line with circles ,
the dashed line with squares , and the dotted line with triangles correspond to the cases for @xmath118 , and @xmath119 , respectively .
@xmath120 is essentially independent from @xmath107 for @xmath121 indicating that heating process is similar for relatively high @xmath107 . in the low @xmath107 ,
on the other hand , there is a trend that electron heating is suppressed for @xmath110 .
for sufficiently low @xmath122 , number of resonant particles is very little so that it takes extremely long time to heat electrons .
then after the very long time , heating is suddenly triggered when field energy becomes large enough .
such an unnatural time evolution occurs because of the assumptions that only the resonant wave - particle interactions are taken into account and that the distribution functions are always maxellian .
indeed , it is known from pic simulations that effects of particle trapping , which have not been included in this analysis , become important for low @xmath107 cases @xcite .
the initial electron to ion temperature ratio , @xmath123 , affects the saturation levels . for large ( small ) @xmath124
, @xmath120 is high ( low ) . comparing with the case of @xmath125 , @xmath126 gets closer to unity because @xmath127 gets larger when @xmath123 becomes 1/3 .
hence , the inverse ion landau damping works more effectively so that electron heating becomes less dominant .
the @xmath8 dependence of @xmath120 is shown in the lower panel of fig.[fig02 ] for @xmath128 . as expected from the linear analysis @xcite , dependence on @xmath8
is rather weak .
this implies that using small @xmath8 in simulations does not lead to unrealisitic results for long time evolution of the mtsi . , and ( lower panel )
corresponding wave propagation angles , @xmath129 .
the three black lines correspond to cases with different initial temperature ratio .
their definition is given in the caption of fig.[fig02 ] .
the thick gray solid ( dashed ) line is obtained from eq.([sat_ql ] ) ( eq.([sat - ehole ] ) ) .
[ fig03],width=302 ] fig.[fig03 ] represents @xmath3 dependense for @xmath113 and @xmath128 .
three black lines ( the solid line with circle , the dashed line with squares , and the dotted line with triangles ) in the upper panel denote the maximum saturation levels of @xmath120 in varying @xmath129 again for different initial temperature ratios ( @xmath130 ) . the corresponding values of @xmath129 is indicated in the bottom panel .
@xmath120 does not increase so much , even though the @xmath3 gets larger .
as already mentioned , the system saturates when available field energy is lost and no more waves grow .
such a state may be achieved
when effect of electron landau damping becomes efficient . if it is the case , @xmath131 should be satisfied at the saturation state . here
@xmath132 is a factor of the order of unity and is empirically determined later . by using @xmath133 and @xmath134 , @xmath135 the gray thick solid line in fig.[fig03 ]
is obtained by substituting the values of @xmath129 of the solid line in the bottom panel into eq.([sat_ql ] ) . here , @xmath136 has been assumed .
although dependence on initial tempetarure ratio , @xmath123 , is not trivial in eq.([sat_ql ] ) , this may give a reasonable order of magnitude estimate .
remember that in the mtsi a cross - field ion beam interacts with oblique whistler waves .
the whistler waves which can interact with the high speed ion beam should have appropriately high phase velocities .
the phase velocity of an oblique whistler wave is proportional to @xmath137 @xcite so that a high mach number shock generates high phase velocity , i.e. , less oblique , whistler waves .
this effect appears in the bottom panel . as a result
, @xmath138 does not change so much even the mach number varies , so that @xmath120 remains in the same order for wide mach number regime examined here .
only the dotted line with triangles in the upper panel is weakely decreasing function of the mach number . in this relatively high initial ion temperature case ,
as mentioned earlier , ion kinetic effect seems to be non - negligible . as a result ,
ion heating is more remarkable than in the cases with @xmath118 .
furthermore , because of small growth rate due also to the ion kinetic effect , it takes rather long time for the system to approach a final saturation state .
therefore , the saturation levels were actually estimated at @xmath139 as mentioned before . note
that the condition @xmath140 gives essentially same expression given in ref .
33 which derives saturation energy due to electron trapping .
hence , the expression in eq.([sat_ql ] ) may be applicable in also the low beta case for mtsi . however , because ref .
33 assumed that @xmath141 is constant throughout their analyses , their saturation temperature increases with @xmath142 .
this might be true if the mach number is sufficiently small and the electrostatic approximation is valid .
but it is now clear that constant @xmath141 assumption is invalid at least for @xmath143 .
.[table1]injection plasma parameters for simulation [ cols="^,^,^,^,^,^,^",options="header " , ] here , shock waves are actually reproduced by using one - dimensional pic code and the resultant electron heating through mtsi and bi in the transition region is discussed .
a shock is produced in the simulation domain by using the so - called reflecting wall method .
an upstream plasma consisting of equal numbers of ions and electrons is continuously injected from the left - hand boundary ( @xmath144 ) and carries a uniform magnetic field , @xmath145 , and a motional electric field , @xmath146 , where @xmath147 is the injection flow velocity in the positive @xmath148-direction . at the right - hand boundary the particles
are reflected . as a result of the interaction between the incoming and reflected particles a shock wave
is produced and propagates in the negative @xmath148-direction .
thus the simulation frame is the downstream rest frame .
the grid size is set to be @xmath149 and 200 particles / cell are distributed for both ions and electrons at the injection boundary , where @xmath150 denotes the electron debye length .
physical parameters are summarized in table [ table1 ] .
for runs a , b , and c ( d , e , and f ) , mtsi ( bi ) gets excited in the foot during each reformation cycle . and ( lower left panel ) @xmath151 .
the gray solid lines denote profiles of magnetic field @xmath152 component .
the right panels indicate electron distribution functions in @xmath153 and @xmath154 integrated over space corresponding to the arrowed regions in the left panels .
[ fig04],width=340 ] fig.[fig04 ] shows two examples of snap shots of electron phase space @xmath155 at @xmath156 for run a ( upper left panel ) and @xmath157 at @xmath158 for run f ( lower left panel ) .
the gray solid lines denote profiles of magnetic field @xmath152 component . in both runs velocity distribution functions corresponding to the downstream areas indicated by arrows in the left panels
are represented in the right panels .
nonadiabatic electron heating downstream of the shock is clearly observed .
parallel heating is superior to the adiabatic perpendicular heating in run a , while strong perpendicular heating , or heating parallel to the beam velocity , is seen in run f. the effective temperatures , @xmath159 , just downstream of the shocks are estimated by using the same definition as eqs.([kpara ] ) and ( [ kperp ] ) .
we averaged over the particles distributed in the region of @xmath160 , where @xmath161 is the position of the magnetic overshoot and @xmath162 the ion cyclotron radius defined by upstream flow velocity and ion cyclotron frequency . because of nonstationarity of the shock , @xmath159 is not constant in time .
therefore , @xmath159 is also averaged in time a few shock reformation periods .
the results are plotted in fig.[fig05 ] as a function of the injection mach number , @xmath163 . here , @xmath159 is normalized to the upstream magnetic pressure , @xmath164 , and rewritten as @xmath165 .
for runs a , b , and c , @xmath166 is almost independent on @xmath163 , which is consistent with the extended quasilinear analysis shown in the previous section . the dashed line just below the points a to c
is drawn by using eq.([sat_ql ] ) with assuming @xmath167 , @xmath168 , and @xmath169 .
f are estimated as averaged effective electron temperatures downstream of the shocks reproduced in the pic simulations .
the dashed lines are obtained from the model analyses , eq.([sat_ql ] ) for a @xmath170 c and eq([sat - ehole ] ) for d @xmath170 f , respectively .
[ fig05],width=302 ] on the other hand , @xmath171 efficiently increases with @xmath163 for runs d , e , and f , where bi gets excited in the foot .
nonlinear development of the bi was well studied in the past . especially , ref .
10 analyzed it in detail by performing one dimensional pic simulations with periodic boundary conditions modeling a part of the foot region in a perpendicular shock geometry .
they showed that when @xmath172 , the bi produces well defined electron holes in its early nonlinear stage and some additional heating or acceleration occurs in the course of further long time evolution . although a saturation level of the instability is reduced a little in two dimensional system , electron trapping still plays an important role through oblique modes @xcite .
it is infered from these results that size of the electron holes in phase space may give a lower limit of @xmath171 .
the amplitude of an electron hole was estimated by assuming that electrostatic wave energy density is comparable with drift energy density of electrons in the ion s rest frame @xcite .
for our case , it reads @xmath173 where we assume the bi based on electron - reflected ion interactions which usually grows faster than the bi based on electron - incoming ion interactions so that @xmath174 is used as the relative drift velocity .
as seen in the above expression , @xmath171 increases with @xmath142 which is proportional to the system free energy .
this functional dependence is also derived in refs .
6 , 7 , and 11 .
one can confirm that eq.([sat - ehole ] ) may give a good estimate of the lower limit of @xmath171 by plotting the corresponding values in fig.[fig05 ] for @xmath175 and @xmath176 as the dashed line just below the points d to f. one should note that @xmath3 in eq.([sat - ehole ] ) is not necessarily coincide with @xmath163 in fig.[fig05 ] . because the @xmath3 indicates the local mach number of the foot in the shock rest frame , while the @xmath163 denotes the injection mach number in the simulation frame , i.e. , the downstream rest frame .
it is also hard to define the local mach number of the foot from the simulation data , since the shock is not time stationary .
therefore , the dashed line should be regarded only as a guide .
in the previous section it was confirmed that the extended quasilinear analysis for the mtsi and the trapping analysis for the bi give consistent results with the pic simulations . here
, let us compare the results of these two analysis .
the thick gray dashed line in fig.[fig03 ] denotes @xmath171 obtained from eq.([sat - ehole ] ) for @xmath177 .
this line intersects at @xmath178 with the dotted line with triangles which is @xmath120 for the mtsi with @xmath110 . the gray thick solid line based on eq.([sat_ql ] ) likely to intersect with the gray thick dashed line at @xmath179 . depending on parameters ,
an actual intersection may occur in the range of @xmath180 . if this is written as @xmath181 , the dominant electron heating process possibly switches from the mtsi for @xmath182 to the bi for @xmath183 .
for relatively low mach numbers like in the earth s bow shock , @xmath120 is more or less constant with @xmath184 because the mtsi becomes dominant .
if @xmath185 and @xmath186 are assumed , corresponding electron temperature is @xmath187ev which is consistent with a typical temperature observed downstream of the earth s bow shock .
however , addiabatic heating due to increases of magnetic field and cross shock potential also results in the downstream electron temperature similar to this value .
this may be the reason why remarkable electron heating was seldom observed in near earth shocks in the past .
but one may capture the nonadiabatic parallel electron heating if upstream electron beta is low enough ( @xmath188 ) , since eq.([sat_ql ] ) is still valid for rather low beta cases where the trapping effects become essential as mentioned in section ii .
furthermore , as shown in the upper right panel of fig.[fig04 ] , local temperature anisotropy , @xmath189 , in a transition region may be observed as a result of strong mtsi in some parameter regimes . on the other hand ,
the effective electron temperature significantly increases with being proportinal to @xmath142 in @xmath183 .
although this is consistent with the past studies @xcite , the saturation electron temperature estimated in this study seems to be rather smaller .
this should be because of that the heating process in highly nonlinear stage including the second step ion acoustic instability is neglected .
electron temperature seen in ref .
10 seems to be about one order higher than the estimate given by eq.([sat - ehole ] ) .
this implies that an actual @xmath181 may appear at a little smaller value .
consistency of the results of the extended quasilinear analysis and the pic simulations for the mtsi dominant cases indicates that the resonant wave - particle interactions are essential in electron heating through the mtsi .
in other wards , nonresonant wave - particle interactions which have been neglected in the analysis may not be effective .
it should also be noted that using unrealistically small @xmath8 in pic simulations is justified as far as the mtsi is concerned because of the weak dependence on @xmath8 ( bottom panel of fig.2 ) in contrast to the bi dominant systems .
the transition at @xmath190 might not be so drastic in a realistic case . in the analyses presented here ,
all other possible candidates of microinstabilities have been neglected .
for example , ecdi ( electron cyclotron drift instability ) is one candidate @xcite and it might become important around this critical mach number , although there are some negative indications for the ecdi to become dominant . for instance , it is known that the saturation level of the ecdi is usually not so high @xcite .
in addition , if the ecdi gets excited simultaneously with the mtsi , the mtsi becomes dominant for wide parameter range @xcite .
furthermore , it is shown that the ecdi can be important only when @xmath8 is of the order of unity @xcite .
hence , the mach number regime where the ecdi is dominant seems not to be so wide , but we do not remove the possibility that the ecdi gives some contributions to electron heating around @xmath181 . because the ecdi is insensitive to the electron landau damping which may strongly affect to the bi in this mach number regime .
further , it is known that the mtsi can not get excited in extremely high mach number shocks like @xmath191 @xcite .
there are also other possible microinstabilities in higher dimensional cases @xcite .
contributions from them should be carefully estimated .
some of them have been discussed recently by performing two dimensional pic simulations @xcite .
the multi - dimansionality may give another important contributions to electron heating .
42 pointed out that electrons are accelerated in the so - called rippled structure which is the ion scale structure along the shock surface .
nevertheless , what should be emphasized here is that the mach number dependence of the effective electron temperature is systematically different between high and low mach number regimes . in the present extended quasilinear analysis
the damping of the wave energy in the late stage ( e.g. , @xmath192 in fig.[fig01 ] ) may be an artifact . because of the assumption that the distribution function is always maxwellian , the so - called plateau of the distribution function is never produced in this system .
the field energy might saturate earlier at a certain level if the plateau is produced . in this regard , it might be better that the saturation levels are defined as the values at the time when the field energy becomes the maximum . however , the resultant saturation levels based on the two definitions , i.e. , the saturation temperatures estimated at times corresponding to the maximum and the later sufficiently small field energies , are not so much different from each other . although it is possible to solve the quasilinear equation , eq.([qleq ] ) , directly as done in refs .
39 and 40 , that is the future work .
the explorations of the inner heliosphere will be underway through the bepicolombo mission .
the perihelion point of the mercury is about 0.3au where quite high velocity ipss which have not been decerelated are expected to be observed .
43 estimated the propagation speed of one of the ipss observed in august 1972 and concluded that it approached to @xmath193 km / s at 0.3au from the sun .
more optimistic estimate for the same ips is given in ref .
44 as @xmath194 km / s at 0.3au .
these results imply that the mach number of this ips approached to several tens .
therefore , the transition of the electron heating efficiency at @xmath181 , if present , will be possibly observed in situ @xcite .
99 s. park , s. a. zhekov , d. n. burrows , p. g. garmire , j. l. racusin , and r. mccray , , 1001 ( 2006 ) c. badenes , e. bravo , k. j. borkowski , and i. domnguez , , 358 ( 2003 ) c. badenes , k. j. borkowski , and e. bravo , , 198 ( 2005 ) j. p. hughes , c. e. rakowski , and a. decourchelle , , l61 ( 2000 ) s. l. moses , f. v. coroniti , c. f. kennel , and f. l. scarf , geophys .
* 12 * , 609 ( 1985 ) k. papadopoulos , astrophys .
space sci .
* 144 * , 535 ( 1988 ) p. j. cargill and k. papadopoulos , , l29 ( 1988 ) s. matsukiyo and m. scholer , j. geophys .
* 108 * , smp19 - 1 , 1459 ( 2003 ) n. shimada and m. hoshino , , l67 ( 2000 ) n. shimada and m. hoshino , phys .
plasmas * 11 * , 1840 ( 2004 ) n. shimada and m. hoshino , j. geophys
110 * , a02105 ( 2005 ) m. hoshino and n. shimada , , 880 ( 2002 ) m. e. dieckmann , b. eliasson , p. k. shukla , , 1361 ( 2004 ) m. e. dieckmann , b. eliasson , m. parviainen , p. k. shukla , a. ynnerman , mon . not .
367 * , 865 ( 2005 ) k. g. mcclements , r. o. dendy , m. e. dieckmann , a. ynnerman , and s. c. chapman , j. plasma phys . *
71 * , 127 ( 2005 ) y. ohira and f. takahara , , l171 ( 2007 ) y. ohira and f. takahara , , 320 ( 2008 ) t. amano and m. hoshino , , 190 ( 2007 ) t. amano and m. hoshino , , 244 ( 2009 ) t. amano and m. hoshino , phys .
plasmas * 16 * , 102901 ( 2009 ) t. umeda , m. yamao , and r. yamazaki , , l85 ( 2008 ) m. e. dieckmann and a. bret , , 154 ( 2009 ) k. papadopoulos , agu geophys .
mon . * 34 * , 59 ( 1985 ) c. s. wu , d. winske , y. m. zhou , s. t. tsai , p. rodriguez , m. tanaka , k. papadopoulos , k. akimoto , c. s. lin , m. m. leroy , and c. c. goodrich , space sci . rev . * 37 * , 63 ( 1984 ) m. scholer , i. shinohara , and s. matsukiyo , j. geophys .
108 * , ssh4 - 1 , 1014 ( 2003 ) m. scholer and s. matsukiyo , ann .
* 22 * , 2345 ( 2004 ) s. matsukiyo and m. scholer , adv .
space res . *
38 * , 57 ( 2006a ) muschietti , l. and lembge , b. , adv
. space res . , 37 , 483 ( 2006 ) d. biskamp and h. welter , phys .
* 28 * , 410 ( 1972 ) d. biskamp and h. welter , nucl .
fusion * 12 * , 663 ( 1972 ) d. biskamp and h. welter , j. geophys .
bf 77 , 6052 ( 1972 ) e. ott , j. , b. mcbride , j. , h. orens , and j. , p. boris , phys .
* 28 * , 88 ( 1972 ) j. , b. mcbride , e. ott , j. , p. boris , and j. , h. orens , phys .
fluids * 15 * , 2367 ( 1972 ) m. tanaka and k. papadopoulos , phys .
fluids * 26 * , 1697 ( 1983 ) s. matsukiyo and m. scholer , j. geophys .
111 * , l02101 ( 2006b ) c. f. kennel , and f. engelmann , phys .
fluids * 9 * , 2377 ( 1966 ) t. h. stix , waves in plasmas , aip , chap.17 ( 1992 ) r. c. davidson and j. m. ogden , phys .
fluids * 18 * , 1045 ( 1975 ) o. ishihara and a. hirose , phys . rev .
lett . * 50 * , 1783 ( 1983a ) o. ishihara and a. hirose , phys .
fluids * 26 * , 100 ( 1983b ) o. ishihara , a. hirose , and a. b. langdon , phys .
fluids * 24 * , 452 ( 1981 ) d. burgess , , 316 ( 2006 ) d. f. smart and m. a. shea , j. geophys .
90 * , 183 ( 1985 ) m. dryer , a. eviatar , a. frohlich , a. jacobs , j. h. joseph , and e. j. weber , j. geophys .
80 * , 2001 ( 1975 ) the author thanks m. scholer , t. hada , p. yoon , and r. yamazaki for useful discussions .
the pic simulation was performed by the super computer in isas / jaxa sagamihara .
this work was supported in part by kakenhi(19740304 ) . | efficiency of electron heating through microinstabilities generated in the transition region of a quasi - perpendicular shock for wide range of mach numbers is investigated by utilizing pic ( particle - in - cell ) simulation and model analyses . in
the model analyses saturation levels of effective electron temperature as a result of microinstabilities are estimated from an extended quasilinear ( trapping ) analysis for relatively low ( high ) mach number shocks . here ,
mtsi ( modified two - stream instability ) is assumed to become dominant in low mach number regime , while bi ( buneman instability ) to become dominant in high mach number regime , respectively .
it is revealed that mach number dependence of the effective electron temperature in the mtsi dominant case is essentially different from that in the bi dominant case .
the effective electron temperature through the mtsi does not depend much on the mach number , although that through the bi increases with the mach number as in the past studies .
the results are confirmed to be consistent with the pic simulations both in qualitative and quantitative levels .
the model analyses predict that a critical mach number above which steep rise of electron heating rate occurs may arise at the mach number of a few tens . |
wolf - rayet ( wr ) stars are the evolved descendents of massive stars , with two main flavours : helium - rich wn stars displaying the products of core h burning and carbon - rich wc stars displaying the products of core he burning .
an overview of the number of wr stars in the milky way is provided .
the galactic census tripled between the 1st @xcite and the 6th catalogues ( van der hucht 1981 ) , a century later ( table [ catalogues ] ) , and subsequently doubled by the annex to the 7th catalogue owing to deep narrow - band optical surveys @xcite , the discovery of rich star clusters @xcite and infrared surveys of the galactic centre region @xcite . over the past decade , the number has doubled again , primarily through surveys arising from the advent of large format infrared detectors , with the census totalling 642 , comprising 357 wn stars , 8 wn / c stars , 273 wc stars and 4 wo stars .
[ catalogues ] [ cols="^,<,^",options="header " , ] if we assume that the observed wr distribution is complete to k=8 mag ( though see below ) , @xmath3550 wr dominated stars in the galactic disk reproduce the observed wr distribution .
in addition , over 100 wr stars are currently known with @xmath4 so the total cmz population may be as large as 250 .
such a large population is surprisin given that the cmz accounts for perhaps 5% of the entire milky way star formation rate @xcite , but the presence of three very massive clusters in the galactic centre region suggests wr stars may be disproportionately represented .
a total of 550 + 250 = 800 stars would represent only 82% of the non - dusty wr population since 18% of systems are expected to be dominated by the wr companion in the k - band , so the non - dusty wr population inferred is @xmath3950 . finally , including dusty wc stars , which comprise 15% ( 150 ) of the whole population
, we estimate n(wr ) = 1,100 , providing n(wcd)/n(wc ) is uniform across all galactocentric radii .
a histogram of the predicted jhk@xmath5 magnitudes for this predicted wr population is presented in fig .
[ histogram ] . and gaia g - band magnitudes predicted from the model wr population by @xcite.thick lines represent populations consisting of wn and non - dusty wc stars , while the dotted - line illustrates a model population in which 28% of wc stars are dust - forming with @xmath6 mag . ] based on this global wr population of 1,100 ,
if the lower mass threshold to wr stars is @xmath7 , an average lifetime for the wr phase of 0.25 myr is implied by the current milky way star formation rate of 1.9 @xmath2 yr@xmath8 @xcite and a standard kroupa initial mass function .
this characteristic lifetime is in fair agrement with non - rotation evolutionary models , but is somewhat shorter than rotating model predictions .
if the observed wr distribution is not complete to k=8 mag , the acual wr population will increase . by way of example , the two most recent additions to the on - line catalogue ( v1.13 ) involve wr111 - 13 , a wn6 star from @xcite with k=8.02 mag and wr111 - 14 , a wn78 star from @xcite with k=7.7 mag . if the wr - dominated disk population were 650 instead of 550 , a global population of 1,300 would be inferred , so the current census may be 50% complete .
a brief overview of the wolf - rayet population in the milky way is presented .
the current census of 642 ( v1.13 of online catalogue ) represents a doubling of the population since the annex to the 7th wr catalogue only a decade ago , due to systematic searches from narrow - band surveys at 2@xmath9 m and broad - band near to mid - ir surveys .
a strict lower limit to the binary fraction is 34% ( sb12 systems plus dusty wc stars ) for v@xmath10 15 mag , but the actual fraction will oudoubedly be significantly higher .
spectroscopic monitoring of wr stars in the near - ir is now feasible owing to multi - object ir integral field units ( e.g. kmos at eso s vlt ) . excluding the galactic centre region , @xmath020% of wr stars
are located in star clusters , adding weight to the formation of most massive stars in ob associations , i.e. away from dense clusters .
the current wr census may be 50% complete , in which case the average duration of the wr phase is 0.25 myr for a lower mass limit of 25 @xmath2 to the formation of wr stars .
the majority of systems hosting wr stars are dominated by the he - rich component in the k - band , so these are fundamentally high luminosity ( high mass ) stars .
in contrast , typical progenitors of stripped envelope core - collapse supernovae are relatively low mass stars , having transferred the majority of their h - rich envelope to a close companion .
such systems would not be recognised as conventional wr stars since the he - rich mass donor would be masked by the h - rich mass gainer at optical and near - ir wavelengths . | an overview of the known wolf - rayet ( wr ) population of the milky way is presented , including a brief overview of historical catalogues and recent advances based on infrared photometric and spectroscopic observations resulting in the current census of 642 ( v1.13 online catalogue ) .
the observed distribution of wr stars is considered with respect to known star clusters , given that @xmath020% of wr stars in the disk are located in clusters .
wn stars outnumber wc stars at all galactocentric radii , while early - type wc stars are strongly biased against the inner milky way . finally , recent estimates of the global wr population in the milky way are reassessed , with 1,200@xmath1100 estimated , such that the current census may be 50% complete .
a characteristic wr lifetime of 0.25 myr is inferred for an initial mass threshold of 25 @xmath2 .
* wolf - rayet content of the milky way * + _ * paul a. crowther * _ + _ dept of physics & astronomy , university of sheffield , sheffield s3 7rh , uk _ 2 |
the amount of sequence related data increased dramatically during the past years .
this is due to improvements of high - throughput and computational methods in omics that often yield long lists of gene , protein , or enzyme identifiers ( ids ) . in our laboratory
we process different kinds of sequence based data , e.g. , dna - microarray derived gene - expression data .
the ultimate purpose of any gene - expression experiment is to produce biological knowledge .
independent of the methods used , the result of microarray experiments is , in most cases , a set of genes found to be differentially expressed between two or more conditions under study .
the challenge faced by the researcher is to translate this list of differentially regulated genes into better understanding of the biological phenomena that generate such changes .
a good first step in that direction is the translation of the sequence i d list into a functional profile .
biological pathways can provide key information about the organization of biological systems .
major publicly available biological pathway diagram resources , including the kyoto encyclopedia of genes and genomes ( kegg ) @xcite , genmapp @xcite and biocarta , can be used to allocate sequence data in pathway maps . with this manuscript
we do not intend to present a review about existing solutions but focus on our approach .
our project requires the analysis of sequence cluster lists and extend the analysis to a maximum possible number of organisms .
kegg currently provides adapted maps for over 380 species covering the following molecular interaction and reaction networks : metabolism , genetic information processing , environmental information processing , cellular processes , human diseases . in order to use the kegg pathway database to display and map genes to kegg pathways
, we developed a web - based tool called orfmapper .
orfmapper is an easy - to - use but powerful application that supports data analysis by extracting annotations for given keywords and gene , protein , or enzyme ids , allocating these ids to metabolic pathways , and displaying them on pathway maps .
two color codes can be assigned to the ids , which can , e.g. , represent sequence properties , organism identifiers , or cluster memberships .
these color codes are used in the query output .
the query results are displayed in hypertext format as a web page , prepared for download as tab - delimited raw text , and visualized on colored , hyperlinked kegg metabolic pathway maps that can be downloaded in pdf format .
together with a version optimized for personal digital assistants , orfmapper provides unique functionality with respect to accessing and displaying kegg pathway data .
orfmapper has been entirely developed with php version 4.3.4 , an open source scripting language that is especially suited for internet development .
creation of pdf is performed with fpdf version 1.53 , a freely available php class that allows generating pdf files .
orfmapper runs on a apple mac os x version 10.2 operating system with an apache version 1.3.33 http server .
the processed kegg data are stored in a local relational mysql database version 4.1.13 database .
the database behind orfmapper contains gene identifiers , the annotation , organism , and pathway information , respectively .
the database is updated monthly .
therefore , information from the kegg ftp - server and from the kegg web site are parsed . in order to keep orfmapper working and to avoid user query errors during updates , duplicated tables
are used . upon successful download and processing ,
the updated tables are activated while outdated tables are inactivated .
orfmapper was designed for prompt display of metabolic relations between gene products by the use of kegg pathway maps . a detailed online help guides the beginner through the user interface .
the user has to specify either annotation keywords ( e.g. , `` hydrogenase protein '' or coxa ) , gene ids ( e.g. , kegg , ncbi , uniprot ) , or enzyme ids ( i.e. , ec - numbers ) .
the user input can either be uploaded as an ascii text file , be exported from spreadsheet applications ( e.g. , microsoft excel or openoffice calc ) , or directly pasted into a text area on the web page .
orfmapper is made as flexible as possible in order to handle individual input data formats .
the ids can be listed either vertically or horizontally or mixed .
they can be separated by all typical text delimiters , e.g. , tabulators , spaces , commas and semicolons . placing keywords in quotation marks forces orfmapper to perform a boolean and query . by default ,
all organisms are queried for all entered ids and keywords . in order to restrict output to selected organisms
, it is possible to specify those organisms in the first input row
. this line must be preceded by an angle bracket character `` @xmath0 '' followed by organism names or just parts of organism names ( e.g. , `` droso '' instead `` drosophila melanogaster '' ) .
the organism names must be separated by commas . if no match to an organism name
is found , all organisms are queried . in order to customize visualization , the user may specify colors for individual ids .
therefore , either a color name ( e.g. , yellow , blue , red ) or a hexadecimal rgb code ( e.g. , # ffff00 ) can be appended to ids and keywords with two underscore characters `` '' ( e.g. genenameblue , genename#000080 , keyword1red , `` keyword1 keyword2''green ) .
this colors the enzyme box corresponding to the i d on a kegg pathway map .
likewise , the user can add one additional value to change the box border color .
this is achieved by adding another color preceded by an underscore character to the i d ( e.g. , genenamebluered ) .
coloration is extremely helpful to specify and , in the output , to identify gene products with common properties , such as expression levels or cluster affiliation .
large sets of query data are often stored in spreadsheet applications , e.g. , microsoft excel , openoffice calc , or microsoft access .
thus , we took special care to simplify date import from these applications .
if the data are organized in three columns ( i d , box color , and box border color , respectively ) , then they can directly copy - pasted into orfmapper . upon clicking the _ convert tab _
button , all tabulators are converted to underscores , as required .
orfmapper creates three forms of output : hypertext , raw tab - delimited text , and graphical pdf pathway maps , respectively .
the hypertext query result contains all gene annotations , pathway information , and hyperlinks to kegg pathway maps corresponding to the user defined query ( fig .
[ fig1 ] ) .
this output is sorted by organism names , metabolic categories , pathways , and gene products .
the latter two levels are hyperlinked to the corresponding kegg information pages .
this query result can be downloaded as raw tab - delimited text file for further processing .
the first line of the text file contains the ids given by the user .
all following lines contain the full set of query results with the following entries : sequence or enzyme i d , kegg species : sequence i d , annotation with ec - number and kegg orthology i d , kegg organism i d , species name , kegg pathway map number , metabolic pathway name , box background color , and box border color . upon clicking the document symbol in the hypertext query results
, orfmapper creates a pdf version of the corresponding kegg pathway map .
the graphical pdf map can be saved locally , is scalable , optimized for printing , and includes hyperlinks to kegg metabolite and enzyme information .
if colors were assigned to sequence ids in the query input , the background and borders of enzyme boxes are colored in the pdf maps .
the pdfs are oriented such that the kegg pathway maps fit perfectly either to portrait or landscape paper format .
orfmapper was designed for displaying metabolic pathway oriented information of keywords and nucleotide , protein , or enzyme ids of sequenced organism .
numerous visualization tools for analyzing biological data are available .
orfmapper fills a gap by providing quick access to pathway information via one input field with flexible input formats and output coloration options .
kegg itself provides an integrated tool that can be used to color metabolic pathway objects
. however , orfmapper has a much broader functionality by allowing cross - species queries , giving a more detailed output , hyperlinking individual genes , and converting the colored pathway maps to pdf format retaining hyperlinks . a condensed version of orfmapper requiring less screen space and showing reduced output is devoted to palm - sized pdas .
its screen size is scaled to 240 pixel width and the output of gene annotations is omitted . if equipped with wlan , this allows on the spot information retrieval and mapping of keywords and gene or enzyme ids , e.g. , during research seminars .
orfmappers functionality will continuously be expanded .
while the simple graphical user interface and query syntax will stay unchanged , extensions with respect to the application of functional characters are planned .
we are currently integrating further sequence ids , e.g. , from the protein data bank ( pdb ) .
furthermore , we are planning to facilitate nucleotide and protein sequence querying .
this work is part of the bmbf funded cologne university bioinformatics center ( cubic ) . we like to thank professor d. tautz for generous support , toshiaki katayama from kegg for prompt help , and all beta testers for their valuable comments .
conflict of interest : non declared | computational analyses of , e.g. , genomic , proteomic , or metabolomic data , commonly result in one or more sets of candidate genes , proteins , or enzymes .
these sets are often the outcome of clustering algorithms .
subsequently , it has to be tested if , e.g. , the candidate gene - products are members of known metabolic processes . with orfmapper
we provide a powerful but easy - to - use , web - based database application , that supports such analyses .
all services provided by orfmapper are freely available at http://www.orfmapper.com . |
the development of rechargeable batteries has been an important milestone in the quest for clean sources of energy . among batteries , li - ion batteries with licoo@xmath3 as the cathode material have significant high energy density and found remarkable success in application to portable electronic devices @xcite . however , since licoo@xmath3 suffers from high cost , toxicity , and safety issues , the material is not appropriate for use in electric vehicles . in recent decades ,
substitutional materials for licoo@xmath3 have been investigated , and some manganites are preferred for li - ion batteries since they have low cost , no toxicity , and better safety characteristics .
in fact , limn@xmath3o@xmath4 with a spinel structure has been successfully employed as the cathode in electric vehicles @xcite .
limno@xmath3 and related materials including other transition metal elements with a layered structure are promising cathode materials @xcite .
manganites such as camn@xmath4o@xmath5 and mno@xmath3 are also expected to act as catalysts in photosystems and artificial photosynthesis @xcite .
for application to cathodes and catalysts , _ surface _ electron states play important roles because manganites work as cathodes and catalysts in chemical reactions involving a change of the valence of surface manganese ions mainly from mn@xmath2 to mn@xmath6 .
furthermore , in the case of cathodes , if mn@xmath6 ions are present on the surface , manganese dissolution from the lattice into the electrolyte occurs owing to the disproportion of mn@xmath6 ions into mn@xmath7 and mn@xmath2 ions @xcite .
this reaction is one of the most important factors causing the capacity fading of cathodes in li - ion batteries .
thus , for the material to achieve high efficiency as a cathode or catalyst , it is desirable to retain the mn@xmath6 ions in the bulk while the mn@xmath2 ions reside on the surface .
previous studies approached this problem using first - principles calculations of the charge structure on the surface of manganites . for @xmath8-mno@xmath3 , which is obtained by removing li ions from limn@xmath3o@xmath4 , a portion of mn@xmath2 ions transition to mn@xmath6 ions because of surface reconstruction @xcite .
first - principles calculations have enabled us to consider surface reconstruction or so , but such calculations are limited to small clusters .
therefore , studies of the microscopic mechanisms causing the valence imbalance of manganese ions between the surface and bulk are meaningful when trying to understand increasingly large clusters . to further the above research , we have investigated the effect of the surface on the electronic state in layered manganites using a @xmath9 lattice on the basis of an orbital - degenerate double - exchange model with surfaces @xcite . in ref .
@xcite , in order to clarify the difference in the electron numbers between the surface and bulk , we have concentrated on @xmath0 electron states for some fixed @xmath1 spin patterns . as typical examples , we have chosen two types of @xmath1 spin structures : the c - type antiferromagnetic ( af ) and sheet - like af phases .
these phases were chosen because they have been found in the bulk for some values of @xmath10 , when we include the effect of jahn - teller distortions in the @xmath11 and @xmath12 lattices @xcite . in those cases ,
we have confirmed that the number of mn@xmath2 ions on the surface is always larger than that in the bulk , which is understood by surface - induced friedel oscillations .
note that the behavior of friedel oscillations is different between two @xmath1 spin structures .
thus , it is necessary to carefully investigate the behavior of these oscillations for future applications to cathodes and catalysts . in this paper
, we investigate the behavior of surface - induced friedel oscillations in layered manganites using an orbital - degenerate double - exchange model with surface effects for other types of @xmath1 spin structures .
it is found that the number of mn@xmath2 ions on the surface is always larger than that in the bulk owing to surface - induced friedel oscillations .
furthermore , irrespective of the type of @xmath1 spin structures , we find that the friedel oscillations are enhanced when the fermi - surface curves with better nesting properties exist .
namely , the dimension of the area of ferromagnetic ( fm ) @xmath1 spin structure is not essential to the enhancement of friedel oscillations .
we believe that these results will contribute to the development of highly - efficient cathodes and catalysts . for the calculations in this paper , we use @xmath13 .
herein we introduce the orbital - degenerate double - exchange model @xcite .
this model has been widely used to successfully investigate the physical properties in layered manganites , such as colossal magneto - resistance phenomena @xcite and various types of charge and/or orbital orders in the bulk @xcite . to simplify the model , we consider a widely - used approximation that invokes infinite hund coupling between @xmath0 and @xmath1 electron spins . in this limit
, the @xmath0 electron spin perfectly aligns along the @xmath1 spin direction , reducing the spin degrees of freedom of @xmath0 electrons .
namely , the @xmath0 electrons move only in the fm region of @xmath1 spins , while in the af region , @xmath0 electrons are localized .
the model hamiltonian for this system is given by @xmath14 where @xmath15 ( @xmath16 ) is the annihilation operator for a spinless @xmath0 electron in the @xmath17 ( @xmath18 ) orbitals at site @xmath19 , @xmath20 is the vector connecting nearest - neighbor sites , @xmath21 , @xmath22 is an ising - like @xmath1 spin at site @xmath19 with @xmath23 , and @xmath24 denotes the nearest - neighbor hopping amplitude between @xmath25 and @xmath26 orbitals along the @xmath20 direction with @xmath27 , @xmath28 , and @xmath29 @xcite .
we set the energy , @xmath30 .
we note that the hopping amplitudes between different orbitals of a spinless @xmath0-electron system change sign between along the x - and y- directions , @xmath31 . the second term in eq .
@xmath32 , @xmath10 , is the af coupling between the nearest - neighbor @xmath1 spins and @xmath33 denotes the nearest - neighbor site pair . in this paper
, we do not include explicitly the coulomb interaction terms , but the effect is partially considered by working with a large hund coupling .
namely , we consider a manganite with large bandwidth .
we define @xmath34 as the hole doping and @xmath35 as the electron number of @xmath0 electrons per site with a relation of @xmath36 .
thus , @xmath34 indicates the average number of mn@xmath2 ions per site .
we diagonalize the hamiltonian in a @xmath9 lattice : one layer is composed of a @xmath37 square lattice with periodic boundary conditions along the x- and y - axes . to introduce surfaces to the model
, we consider a stack of @xmath38 layers with open boundary condition ( obc ) along the z - axis . in this paper , we introduce the obc in the sense of @xmath39 , where @xmath40 denotes the wavefunction of a spinless @xmath0 electron along the z - axis and @xmath41 is a layer number .
namely , the first layer along the z - axis is the surface , whereas the @xmath42-th layer denotes the bulk .
spins , respectively .
( a ) the spin structure of the zigzag sheet af state . ( b ) the spin structure in the fm state . ] in this paper , we remark the @xmath0-electron properties on the background of the fixed @xmath1 spin structure . in our previous study , we investigated two @xmath1 spin structures : a c - type af ( c - af ) state and an a - type af ( a - af ) state @xcite . in the c - af state , @xmath0 electrons can move only along the z - axis owing to infinite hund coupling .
thus , the system is regarded as a one - dimensional ( 1d ) chain with the obc along the z - axis . in the a - af state
, @xmath0 electrons can move in the x - z plane , leading to a two - dimensional ( 2d ) sheet with edges . on the other hand , in this paper , we assume two other possible @xmath1 spin structures , zigzag sheet af and fm states , as shown in figs .
1(a ) and 1(b ) , respectively . at least in a @xmath43 lattice with surfaces , it is easy to show that these magnetic states actually appear as the ground states for appropriate values of @xmath10 , but detailed discussion on the stability of ground - state magnetic structures in the large cluster is postponed in future . in the zigzag sheet af state , the fm regions of @xmath1 spins form zigzag structures in the x- and y- directions , whereas @xmath1 spins are ferromagnetic in the z- direction .
thus , the system is regarded as a 2d sheet with edges , where hopping amplitudes change periodically . in the three - dimensional ( 3d ) fm state
, @xmath0 electrons can move in all three dimensions . by using these lattices ,
we evaluate @xmath44 , the number of mn@xmath2 ions per site in layer @xmath41 , and investigate the behavior of the surface - induced friedel oscillations . for several @xmath1 spin structures with different fm region dimensions
, we examine the factor that determines the behavior of the oscillations .
note that @xmath45 and @xmath46 denote the number of mn@xmath2 ions per site on the surface and in the bulk , respectively .
before proceeding to the results of our calculations , we need to briefly review our previous results @xcite .
we observe that @xmath44 clearly exhibits a damped oscillation from the surface ( @xmath47 ) to the bulk ( @xmath48 ) , which is understood to be the surface - induced friedel oscillation , and results in an increase in the number of mn@xmath2 ions on the surface . in the 1d system such as the c - af state ,
the formula for friedel oscillation is expressed as @xcite @xmath49 where @xmath50 is the fermi momentum and @xmath51 is the correlation exponent @xcite . since we do not include coulomb interactions among @xmath0 electrons , @xmath52 , which leads to @xmath53 . on the other hand , in the 2d system such as the a - af state , the formula for friedel oscillations
is modified to @xmath54 where @xmath55 is the z - component of the fermi momentum with good nesting properties along the z - axis .
@xmath56 is the damping power for the oscillations .
we have confirmed that in the 2d system @xmath57 depends on hole doping , @xmath34 , and a relevant orbital , @xmath25 , which leads to @xmath58 . in other words ,
the envelope of the 2d friedel oscillations decays faster than @xmath53 . at this stage ,
the definite expression of @xmath56 is not obtained . however , in general , @xmath45 seems to be increasingly suppressed in higher dimensional systems since the influence of the surface on the system becomes relatively weaker as the dimension of the fm region increases . vs. @xmath41 plot for @xmath59 in the zigzag sheet af state .
the blue dotted curve denotes the envelope expressed by @xmath53 , and the broken horizontal line denotes the value of @xmath34 .
( b ) two fermi - surface curves of the @xmath0-electron system for @xmath59 .
the magnitudes of the fermi momentum along the z - axis are both @xmath60 , which is consistent with the period of the oscillation in ( a ) .
( c ) two fermi - surface curves for @xmath61 .
( d ) efficiency @xmath62 vs. @xmath34 .
@xmath62 is defined as @xmath63 . ] first , let us consider the charge structure in the zigzag sheet af state . in fig .
2(a ) , we show @xmath44 vs. @xmath41 for @xmath59 . since the zigzag sheet af indicates a 2d - like system even though the fm region is not a plane ( see fig .
1(a ) ) , one could naively consider @xmath44 to be expressed by eq .
( [ eq:2d ] ) .
however , as observed in fig .
2(a ) , the envelope of the oscillations is well expressed by @xmath53 .
namely , the friedel oscillation pattern is rather 1d - like in spite of the 2d - like fm region . to investigate one of the factors that determines the value of @xmath56 , let us examine the oscillations from a viewpoint of the fermi - surface structure .
2(b ) , we show two fermi - surface curves at @xmath59 .
note that we depict the fermi - surface curves considering obc along the z - axis .
when we consider the @xmath38 layers stacked along the z - axis with obc , @xmath64 is given by @xmath65 where @xmath66 ( @xmath67 ) is an integer , leading to @xmath68 . in the zigzag sheet af state , we find that the spinless @xmath0-electron system possesses a quasi-1d fermi - surface structure along the z - axis for a certain hole doping , @xmath34 . even in the 2d system ,
when the fermi - surface curves have a quasi-1d structure along the z - axis , the envelope of the oscillations is expressed by @xmath53 as if in the 1d system . the period of the oscillations in fig .
2(a ) is given by @xmath69 , consistent with the @xmath55 shown in fig .
2(b ) and used in eq .
( [ eq:2d ] ) .
2(c ) shows the fermi - surface curves at @xmath61 . for @xmath61
, the fermi - surface curves also retain the quasi 1d structure along the z - axis and we can confirm the value of @xmath56 is almost unity .
the nesting properties are lost for @xmath70 and @xmath71 , so the envelope of the oscillations decays faster than @xmath53 .
to understand the increase in @xmath45 compared with @xmath34 , we evaluate the efficiency @xmath62 , given by @xmath63 @xcite .
this quantity denotes to what degree the mn@xmath2 ions appear on the surface at a given hole doping @xmath34 .
when @xmath45 is enhanced , we expect that the efficiency as cathodes and catalysts is increased .
thus , @xmath62 has been called the `` efficiency '' .
note that @xmath62 is associated with the magnitude of amplitude of the friedel oscillations .
2(d ) shows the plot of @xmath62 as a function of @xmath34 for the zigzag sheet af state .
we find the magnitude of @xmath62 exhibits a peak between @xmath72 and @xmath73 , for which one fermi - surface curve has perfect nesting properties , leading to a van hove singularity .
note that the small oscillations in @xmath62 around @xmath74 are effects due to the size of the lattice .
if we increase the size of the lattice , these oscillations disappear .
thus , we can conclude that when the spinless @xmath0-electron system possesses a fermi - surface structure with decent nesting properties along the z - axis , the value of @xmath56 becomes unity and the amplitude of the friedel oscillations tends to be larger .
namely , the behavior of the friedel oscillations is @xmath75 determined by the dimension of @xmath1 spin structure itself , but the fermi - surface structure of @xmath0-electron system is essentially important . vs. @xmath41 plot for @xmath76 in the 3d fm state .
the broken horizontal line denotes the value of @xmath34 .
( b ) efficiency @xmath62 vs. @xmath34 .
( c ) @xmath77/[1-n(z)]$ ] vs. @xmath34 for @xmath47 and @xmath42 .
( d ) orbital arrangement on the surface x - y plane for @xmath78 .
the orbital size is chosen to be proportional to the orbital density . ]
3(a ) shows the plot of @xmath44 vs. @xmath41 for @xmath76 ; surface - induced friedel oscillations can be observed in the figure .
note that @xmath46 is slightly smaller than @xmath34 .
in contrast to the case of the zigzag sheet af state , the envelope of the oscillations decays faster than @xmath53 , since the nesting properties are weak in the 3d @xmath0-electron system . as observed in fig .
3(b ) , @xmath62 decreases monotonically in the fm state , which can be explained by the fact that the nesting properties in the 3d @xmath0-electron system are enhanced to some extent for smaller @xmath34 . a characteristic point for the fm phase
is the appearance of an orbital arrangement on the surface , in sharp contrast to other @xmath1 spin structures . in fig .
3(c ) , we plot @xmath77/[1-n(z)]$ ] vs. @xmath34 for @xmath47 and @xmath42 , where @xmath79 and @xmath80 denote the average electron number in the @xmath17 and @xmath18 orbitals , respectively . in the bulk ,
@xmath17 and @xmath18 orbitals are occupied with equal weights because cubic symmetry is recovered in the 3d environment . on the surface ,
@xmath17 orbitals are selectively occupied , since @xmath0 electrons on the surface layer can gain kinetic energy by the 2d motion in the x - y plane than the 1d motion along the z - direction .
in fact , as shown in fig .
3(d ) , there is a large density of @xmath17 orbitals with a ferromagnetic - arrangement on the surface , while @xmath81 is relatively small .
namely , electrons on the surface continue to occupy @xmath17 and escape only from @xmath18 orbitals , leading to the overall increase of @xmath45 .
in this paper , we have evaluated @xmath44 , the number of mn@xmath2 ions per site in layer @xmath41 , by analyzing the orbital - degenerate double - exchange model with surfaces for zigzag sheet af and 3d fm states .
we have found that the surface - induced friedel oscillations in @xmath44 are enhanced , when the spinless @xmath0-electron systems possess fermi - surface structures with improved nesting properties along the z - axis , leading to an increase of mn@xmath2 ions on the surface . for cathodes and catalysts , retaining mn@xmath2 ions on the surface and mn@xmath6 in the bulk
is desirable .
subsequently , we can conclude that the @xmath1 spin structures possessing fermi - surface structures with good nesting properties are possibly suitable for high - efficiency cathodes in li - ion batteries and catalysts .
we believe that the present results provide a fundamental understanding for further application of manganites to such systems .
we thank k. hattori , k. kubo , and k. ueda for discussions and comments .
the computation in this work was partly carried out using the facilities of the supercomputer center of institute for solid state physics , university of tokyo . | to investigate valence imbalance phenomena between the surface and bulk in layered manganites , we analyze an @xmath0-orbital degenerate double - exchange model with surfaces for two types of @xmath1 spin structures .
we reconfirm that the surface - induced friedel oscillations occur in the charge structure and that the number of mn@xmath2 ions on the surface is larger than that in the bulk .
this tendency is found to be more significant , when the @xmath0-electron system has fermi - surface structures with better nesting properties .
the behavior of the friedel oscillations depends on the nesting properties of the fermi - surface curves along the direction of the oscillations , and not on the dimension of the system itself .
we believe that these results will be useful for the development of high - efficiency cathodes in li - ion batteries and catalysts . |
consider the following ( random ) game : a deck of @xmath0 cards is divided into several piles . then , for each pile , we leave it intact with probability @xmath1 and remove one card from there with probability @xmath2 ( @xmath3 $ ] is a given parameter ) , independently of other piles .
the cards that were removed are collected to form a new pile .
the order of piles is not important and the piles of size zero are ignored .
the case @xmath4 is trivial ( nothing moves ) and will not be considered . when @xmath5 , this is the game of _ bulgarian solitaire _ ,
made known by martin gardner in @xcite , and studied in @xcite ( cf .
also @xcite for some variations of that game ) .
the `` truly random '' model with parameter @xmath6 is a discrete - time irreducible and aperiodic markov chain on the space of all unordered partitions of @xmath0 ; for obvious reasons , it will be referred to as the _ random bulgarian solitaire_. if the number of cards @xmath0 is a triangular number , i.e. , @xmath7 for some @xmath8 , a remarkable fact is that , starting from any initial configuration , after a finite number of moves the ( deterministic ) bulgarian solitaire will reach the stable configuration formed by piles of sizes @xmath9 . the above result was proved in @xcite ( see the solution to problem 6.10 ) and in @xcite independently , and later it was discovered that the maximal number of moves necessary to enter the stable configuration is @xmath10 , and that that bound is sharp ( see @xcite ) . if @xmath0 is not a triangular number , then such a stable configuration does not exist .
however , it is possible to prove that after at most @xmath11 moves the game will enter into a cycle .
moreover , all the configurations of the cycle are `` almost triangular '' in the following sense . let @xmath12 ; then all the configurations in that cycle can be constructed from the configuration @xmath13 by adding at most one card to each pile , and maybe adding one more pile of size @xmath14 , see @xcite for exact formulations and more details .
thus , we see that bulgarian solitaire `` likes '' triangular configurations , and so we may expect some kind of similar behaviour from the random bulgarian solitaire .
there is no possibility , however , to obtain exact results of the form of those of the previous paragraph , since random bulgarian solitaire is a finite irreducible markov chain , so it visits all its states infinitely many times a.s .
instead , we aim at the results of the following kind : the stationary measure of the set of configurations which are in some sense close to the ( rescaled ) triangular configuration is close to @xmath14 .
this can be regarded as a `` shape theorem '' result even though it is substantially different from most of the shape results appearing in the literature .
( in most cases some time - dependent random set is constructed , and then , when rescaled by time , it converges to some , usually nonrandom , shape .
see e.g. @xcite for results of this kind . )
the results we are aiming at resemble rather those of @xcite . also , let us remark here that the question of how fast the deterministic bulgarian solitaire _ approximates _ the triangle has not been yet studied in the literature . to motivate this question ,
take @xmath7 , and suppose that the initial configuration is @xmath15 , i.e. , the exact triangular configuration is modified by removing one card from the biggest pile and forming one more pile of size @xmath14 with that card .
then , macroscopically this configuration is already quite triangular ; however , if we are aiming to reach @xmath13 , this is the worst possible initial configuration ( the number of moves needed is exactly @xmath10 , cf .
@xcite ) ! here we prove that , whenever the initial configuration is `` reasonable '' ( i.e. , the number of piles is @xmath16 and the number of cards in the biggest pile is also @xmath16 ) , we need only @xmath16 moves of deterministic bulgarian solitaire to make the ( @xmath17-rescaled ) configuration _ close _ to the triangle .
while such a result by itself may not be of great interest , the method of its proof will be an important tool in the course of the proof of the results about random bulgarian solitaire .
now , we introduce some notations and give the formal definition of the process . if @xmath18 is the number of piles in the configuration @xmath19 , we write @xmath20 , where @xmath21 . we denote also by @xmath22 the size of the biggest pile and by @xmath23 the number of cards in the configuration .
let @xmath24 be the operation of putting @xmath25 in the decreasing order and discarding zeros .
now , let @xmath26 be a sequence of i.i.d . random variables such that @xmath27=1-\pr[\xi_1=0]=p$ ] .
then the operator @xmath28 which transforms the configuration @xmath20 in the game of random bulgarian solitaire with parameter @xmath2 is defined by @xmath29 denote also by @xmath30 the result of @xmath31 independent applications of @xmath28 to @xmath19 ; clearly , the process is conservative in the sense that @xmath32 for all @xmath31 .
suppose that @xmath33 .
as remarked above , for @xmath6 the stochastic process @xmath34 is an irreducible aperiodic markov chain with finite state space @xmath35 .
we denote by @xmath36 its stationary measure .
to formulate our results , we need also to find a way to define sets of configurations that are close to a specific triangular configuration . to this end , for two configurations @xmath37 , @xmath38 define the distance @xmath39 by @xmath40 with the convention @xmath41 for all @xmath42 and @xmath43 for all @xmath44 .
next , we define the triangular configuration @xmath45 by @xmath46 , @xmath47 . when @xmath2 is fixed and @xmath48 , we can write @xmath49 , @xmath50 . finally , for @xmath51 ( which may depend on @xmath0 ) define the set @xmath52 of `` roughly triangular '' configurations by @xmath53 let @xmath54 , and define the configuration @xmath55 , where @xmath56 note that @xmath57 ( for example , for @xmath58 we have @xmath59 ) . for the particular case @xmath5
we say that @xmath60 is nondegenerate if it contains the configuration @xmath61 , as well as all the configurations @xmath19 with @xmath62 .
it is easy to see that for any fixed @xmath51 there exists @xmath63 such that @xmath60 is nondegenerate for all @xmath64 , and the same is true when e.g. @xmath65 , @xmath66 .
now we are ready to formulate the main results of this paper .
first , we state the result about the time to approximate the triangular configuration for the deterministic bulgarian solitaire ( i.e. , with @xmath5 ) .
[ det_bs ] take @xmath51 and suppose that @xmath0 is large enough to guarantee that @xmath60 is nondegenerate .
suppose that the initial configuration @xmath67 with @xmath33 has the following properties : @xmath68 and @xmath69 for some @xmath70 .
then there exists @xmath71 such that we have @xmath72 for all @xmath73 . in words
, this result means that if the initial configuration is `` reasonable '' , then the number of moves required to approximate the triangle is @xmath16 .
now , we turn our attention to random bulgarian solitaire : [ ran_bs ] suppose that @xmath6 . then for any @xmath74 there exist positive constants @xmath75 and @xmath76 such that for all @xmath0 @xmath77 in section [ s_fr ] there are some more comments and open problems related to the bulgarian solitaire ( both deterministic and random ) .
also , the reader may find it interesting to look at java simulation of the random bulgarian solitaire ( with @xmath78 ) on the internet page of kyle petersen at + http://people.brandeis.edu/~tkpeters/reach/stuff/reach
this section is organized in the following way . in section [ s_etienne ]
we introduce the notion of _ etienne diagram _ , which is just another way to represent the configurations of the game .
then , we show how the moves of bulgarian solitaire are performed on this diagram and discuss its other properties . in section [ s_proof_det ]
we prove theorem [ det_bs ] , and in section [ s_proof_rand ] we prove theorem [ ran_bs ] ( in section [ s_proof_rand ] some results and technique from sections [ s_etienne ] and [ s_proof_det ] are used , most notably the inequality ( [ iteration ] ) ) . before starting the proofs
, we need to describe another representation of a particular state of ( deterministic ) bulgarian solitaire , which we call an etienne diagram ( cf .
@xcite ) . in this approach
the cards are identified with particles living in the cells of the set @xmath79 , with at most one particle per cell .
we write @xmath80 when the cell @xmath81 is occupied and @xmath82 when the cell @xmath81 is empty .
clearly , @xmath83 is a half - quadrant of @xmath84 , but we would like to visualize @xmath83 in a little bit unconventional way ( see figure [ p1 ] ) : the cell @xmath85 lies in the base and supports the column @xmath86 , while the diagonal @xmath87 goes in the nw direction ( so the rows of @xmath83 are enumerated from right to left ; notice that at this point we deviate from @xcite , where the rows were enumerated from left to right ) .
now , a configuration @xmath20 is represented as follows ( as on figure [ p1 ] ) : we put @xmath88 for @xmath89 and @xmath90 for all other pairs @xmath81 .
from the fact that @xmath21 we immediately deduce that for any @xmath19 @xmath91 and @xmath92 one of the advantages of the representation via etienne diagram is that it makes it more clear how the process approaches the triangular configuration . to see what we mean , first note that the move of the bulgarian solitaire consists in applying the following two substeps to the corresponding etienne diagram ( see figure [ p2 ] ) : *
apply the cyclic shift ( from left to right ) to each row of the diagram ; * if after the shift there is a particle that is placed above an empty cell , then the particle falls there ; this procedure is repeated until no further fall is possible .
speaking formally , let @xmath93 .
then the etienne diagram of @xmath94 is constructed using the following procedure : * for all @xmath95 put @xmath96 for @xmath97 and @xmath98 . *
suppose that for the array @xmath99 we can find @xmath100 such that @xmath101 , @xmath102 . then construct the new array @xmath103 by @xmath104 , @xmath105 , and @xmath106 for @xmath107 .
* repeat the previous procedure with @xmath103 instead of @xmath99 , and so on . at some moment
we will obtain an array @xmath108 for which we can not find @xmath109 such that @xmath110 , @xmath111 .
then for all @xmath112 put @xmath113 .
now , suppose that @xmath114 .
on the etienne diagram the triangular configuration corresponds to the configuration @xmath115 . note also that if the first @xmath116 rows of the diagram are occupied , then they will remain occupied during all the subsequent evolution .
this shows that the falls of particles `` help '' to reach the stable configuration ( more and more rows become all occupied ) .
moreover , in many concrete situations it is possible to know how many moves are needed to fill out some region which was originally empty .
arguments of this kind will be heavily used in the course of the proof of our results .
consider the etienne diagram of a configuration @xmath19 .
since the system is conservative , there is a natural correspondence between particles ( holes ) in that diagram and particles ( holes ) in the diagram of the configuration @xmath117 .
this shows that for each particle ( hole ) on the original diagram we can define its trajectory , i.e. , we know its position after @xmath31 moves of the game .
let @xmath118 be the second coordinate of the particle ( hole ) from @xmath81 after @xmath31 moves , and let @xmath119 be the number of falls ( movements upwards ) that the particle ( hole ) from @xmath81 was subjected to during @xmath31 moves .
that means that , if @xmath120 , then @xmath121 are the coordinates of the particle from @xmath81 after @xmath31 moves , while if @xmath122 , then @xmath123 are the coordinates of the hole from @xmath81 after @xmath31 moves .
it seems to be very difficult to calculate exactly @xmath118 and @xmath119 ( except in trivial situations , when , e.g. , @xmath120 and @xmath124 for all @xmath125 ) .
however , we can establish some relation between these quantities by defining first @xmath126 in words , @xmath127 would be the position of particle ( hole ) from @xmath81 at time @xmath31 if we know that @xmath128 ( the quantities @xmath129 depend also on @xmath19 , but we do not indicate that in our notations ) .
[ jm ] if @xmath122 and @xmath31 is such that @xmath130 , or @xmath120 and @xmath31 is such that @xmath131 , then @xmath132 _ proof .
_ suppose for example that @xmath122 .
denote @xmath133 .
since @xmath31 is such that @xmath130 , we have that @xmath134 .
the lemma then follows from the fact that @xmath118 should be somewhere in between @xmath135 and @xmath136 .
the other case is treated analogously .
next , we define some quantities which concern the geometric structure of the representation via etienne diagram , and prove some relations between them . for @xmath137 define @xmath138 when @xmath139 , we have @xmath140 . using the etienne representation of a configuration @xmath19 , define @xmath141 and put @xmath142 .
the quantity @xmath143 can be thought of as the `` energy '' of the configuration : the bigger is @xmath143 , the `` more distant '' ( not necessarily in the sense of the distance @xmath144 ) is @xmath19 from @xmath61 .
denote also @xmath145 .
the next lemma establishes some elementary properties of the energy @xmath143 .
[ prop_e ] * there exists a constant @xmath146 such that for all @xmath0 and all @xmath147 we have @xmath148 .
* for all @xmath19 it holds that @xmath149 .
moreover , @xmath150 is equal to the number of falls of particles during the second substep of the move of the bulgarian solitaire represented by the etienne diagram ( i.e. , it is equal to @xmath151 in * ( iii ) * ) .
_ define @xmath152 . from ( [ poln_tri ] )
one easily gets that there exists @xmath153 such that for all @xmath147 we have @xmath154 .
the proof of ( i ) then reduces to an elementary computation ( roughly speaking , to compute @xmath143 we have at worst @xmath155 terms , each of order @xmath16 ) . as for the proof of ( ii ) , note first that the operation of cyclic shift does not change the quantities defined in ( [ def_e-])([def_e+ ] ) .
then , it is straightforward to see that each particle s fall decreases @xmath156 by one unit , which concludes the proof of the lemma .
* @xmath159 is the maximal vertical distance between @xmath160 and the holes below @xmath160 ; * @xmath161 is the maximal vertical distance between @xmath160 and the particles above @xmath160 which also lie inside the area indicated by the dashed lines on figure [ p3 ] ; * @xmath162 is the total area covered by the holes below @xmath160 ; * @xmath163 is the total area covered by the particles above @xmath160 . similarly to the energy @xmath143 , all those quantities could be used to measure the deviation of @xmath19 from the `` almost triangular '' configuration @xmath61 . consider also the normalized quantities @xmath164 , @xmath165 , and @xmath166 , @xmath167 .
[ alphas ] for all @xmath168 there exist constants @xmath169 ( depending on @xmath170 ) such that @xmath171 _ proof .
_ it is elementary to obtain the inequalities ( [ a12 ] ) and ( [ a56 ] ) from ( [ pust_tri ] ) .
analogously , to obtain ( [ a34 ] ) and ( [ a78 ] ) , one can use ( [ poln_tri ] ) and the fact that @xmath168 together with the following observation . if @xmath120 for some @xmath172 , then either @xmath173 or @xmath174 . consider a configuration @xmath19 such that @xmath175 . by definition of @xmath160 , there exists a constant @xmath176 such that @xmath177 also , we will always tacitly assume that @xmath178 , i.e. , we will not consider configurations that are `` too close '' to the triangle . in this case there are constants @xmath179 such that @xmath180 ( note also that if @xmath0 is a triangular number , then @xmath181 for any @xmath182 ) . using ( [ a12 ] ) , ( [ a56 ] ) , and ( [ beta12 ] ) , we obtain @xmath183 and , by ( [ a34 ] ) , @xmath184 .
this shows that there is a constant @xmath185 such that @xmath186 for all @xmath168 .
analogously , we obtain that for some @xmath187 it holds that @xmath188 and @xmath189 . by lemma [ prop_e ] ( i ) the quantity @xmath190
is bounded on @xmath191 , so there is @xmath192 such that @xmath193 , which implies that @xmath194 for some @xmath195 .
finally , we use lemma [ prop_e ] ( i ) once again to obtain that there exists @xmath196 such that @xmath197 when @xmath168 .
first , the idea is to prove that after @xmath16 moves , the `` normalized energy '' @xmath190 will decrease by a considerable amount .
consider a configuration @xmath19 with @xmath175 .
abbreviate @xmath198 ; by ( [ minh ] ) , we can find @xmath199 such that @xmath200 and @xmath201 .
moreover , without loss of generality one can suppose that @xmath202 is divisible by @xmath203 .
define also @xmath204 , @xmath205 , and @xmath206 .
define two sets @xmath207 by @xmath208 ( see figure [ p4 ] ) .
note that from ( [ pust_tri ] ) and ( [ poln_tri ] ) it follows that @xmath122 for all @xmath209 and @xmath120 for all @xmath210 .
abbreviate also @xmath211 , @xmath212 .
now , the idea is to consider the evolution of sets @xmath213 at times @xmath214 , @xmath215 .
first , note that @xmath216 , for any @xmath217 and @xmath8 . then , each time we make a complete turn ( i.e. , @xmath218 moves ) a particle which was on the level @xmath219 will be @xmath220 units to the left of its initial position ( provided it did not fall ) .
this shows that there exists @xmath221 such that @xmath222 \cap [ m_1,j_1 ] \big| \geq \frac{2\hh}{5}\ ] ] ( when @xmath223 , by @xmath224 $ ] we mean in fact @xmath225 \cup[{\hat j}_{i'_0,j_2}(k_0i_0),i_0]$ ] ) .
take @xmath226 such that @xmath227\subset \big([{\hat j}_{i'_0,j_2}(k_0i_0),{\hat j}_{i'_0,m_2}(k_0i_0 ) ] \cap [ m_1,j_1]\big)\ ] ] and @xmath228 .
we consider two cases : : at time @xmath229 in the set @xmath230 , j\in[j_3,j_3 - 1+\hh/5]\}\ ] ] there is at least one hole , i.e. , @xmath231 for at least one @xmath232 . in this case , by ( [ pust_tri ] ) no particle can be in the set @xmath233 , j\in[j_4-\hh/5,j_4]\},\ ] ] i.e. , for all @xmath234 we have that @xmath231 .
note that @xmath235 , so the `` image '' of @xmath236 after @xmath237 turns completely covers @xmath238 . on the other hand , @xmath238
must be completely empty , so there should have been a lot of particle falls in order to avoid @xmath238 . in what follows we estimate the minimal number of falls necessary ( and , consequently , we find the minimal amount by which the energy @xmath156 should decrease ) .
define the set @xmath239 , j\in[j_4-\hh/10,j_4]\ } \subset u'_2.\ ] ] for any @xmath240 there is a unique @xmath241 such that @xmath242 , and , by the above observation , @xmath243 , so the cell @xmath244 originally contained a particle . to guarantee that that particle is not in @xmath245 at time @xmath229 , at least one of the following two possibilities must occur : * either @xmath246 , * or @xmath247 , but in this case , by lemma [ jm ] , @xmath248 .
denote @xmath249 ; for the both of the above possibilities , we obtained in fact that @xmath250 .
since the number of cells in the set @xmath245 is at least @xmath251 , the number of particle falls until time @xmath229 should be at least @xmath252 . by lemma [ prop_e ] ( ii ) , it means that , for the case 1 , @xmath253 : there are no holes at time @xmath229 in the set @xmath254 , i.e. , @xmath255 for all @xmath232 . using the duality between holes and particles , this case can be treated quite analogously to the case 1 .
namely , we note first that the `` image '' of @xmath256 after @xmath237 turns completely covers @xmath254 .
so , in order to escape @xmath254 , the holes that are `` candidates '' to be there must make a sufficient number of movements in the upwards direction . in the same way as in the case 1
, one can work out all the details to obtain that ( [ hjk ] ) is valid for the case 2 as well .
we continue proving theorem [ det_bs ] . by ( [ minh ] ) and ( [ hjk ] ) ,
there exist @xmath257 such that @xmath258 where @xmath259 ( the formula ( [ iteration ] ) will play an important role in the proof of theorem [ ran_bs ] as well ) .
consider now the initial configuration @xmath260 ; by lemma [ prop_e ] ( i ) , @xmath261 for some @xmath262 .
fix an arbitrary @xmath51 and define @xmath263 ; then there exists @xmath264 ( depending only on @xmath265 ) such that @xmath266 . by ( [ iteration ] ) this means that @xmath267 where @xmath268 , i.e. , after @xmath16 moves we will arrive to a configuration with small normalized energy @xmath190 .
now we are almost done with the proof of theorem [ det_bs ] , and it remains only to make one small effort : we have to prove that if the energy @xmath269 is small , then either @xmath19 is already close to the triangular configuration @xmath270 ( in the sense of the distance @xmath144 ) , or it will come close to @xmath270 after @xmath16 moves .
define the sets @xmath271 it is elementary to see that , for fixed @xmath272 and for all @xmath0 large enough @xmath273 we need the following [ srezaem_hvosty ] suppose that @xmath274 , and put @xmath275 + 2 $ ] . then @xmath276 .
* suppose that at some moment the configuration belongs to the set @xmath284 .
then if @xmath285 and there are some particles in the set @xmath286 , then at the next moment all those particles will fall one unit . * if @xmath287 and @xmath288 then @xmath289 ( to see this , it is enough to figure out what happens with the configuration @xmath290 after @xmath291 moves , where @xmath290 is defined by @xmath292 whenever either @xmath293 or @xmath294 ) .
now we are ready to finish the proof of theorem [ det_bs ] . from ( [ a12 ] ) ,
( [ a34 ] ) , and ( [ small_energy ] ) we obtain that , if the initial configuration belongs to @xmath295 , then after @xmath16 moves it will be in @xmath296 , where @xmath297 ( @xmath272 is from ( [ small_energy ] ) , @xmath298 from lemma [ alphas ] ) . applying lemma [ srezaem_hvosty ] , we conclude the proof of theorem [ det_bs ] . consider a finite irreducible discrete - time markov chain with state space @xmath299 , transition matrix @xmath300 , and stationary measure @xmath301 .
the following elementary fact will be useful in the course of the proof of theorem [ ran_bs ] : for any @xmath302 and @xmath303 @xmath304 let us describe the main steps of the proof of theorem [ ran_bs ] : * first , in lemma [ reasonable ] we prove ( using ( [ osn_fakt ] ) ) that a typical configuration of the random game should be reasonable , i.e. , the number of piles and the biggest pile should be @xmath16 ; * then , the idea is the following : starting from a reasonable configuration , the macroscopic evolution of the profiles will be very similar in the random game and in the deterministic game where the initial sizes of the piles are @xmath305 times bigger ( indeed , if the initial size of the pile in the random game is @xmath306 , then it will be emptied typically in time @xmath307 ) ; * unfortunately , it seems to be difficult to dominate the stochastic game by the deterministic one directly .
so , we introduce another deterministic process by allowing the immigration of particles to the system on each step . in lemma [ dominacao ]
we prove that the random bulgarian solitaire is in some sense dominated by this new deterministic process ; * it is then possible to see that the process with immigration of particles does not differ much ( when the time interval in question is not too long ) from the deterministic bulgarian solitaire , because the total number of added particles is relatively small , and they can not be very concentrated ( lemma [ nao_fugira ] takes care of the latter statement ) . using this observation and applying inequality ( [ iteration ] ) from the previous section , we obtain that the ( suitably defined ) energy will decrease with large probability after a sufficiently large number of steps ( this is lemma [ iter_rand ] ) ; * the rest of the proof is a straightforward ( although somewhat lengthy ) application of ( [ osn_fakt ] ) and lemma [ iter_rand ] .
so , the first step is to prove that a typical configuration @xmath182 should be `` reasonable '' , i.e. , it should belong to @xmath308 for some @xmath309 : [ reasonable ] * for any @xmath310 there exist positive constants @xmath311 ( depending on @xmath2 ) such that @xmath312 for all @xmath0 . * also , suppose that @xmath313 , where @xmath309 are the quantities from item ( i )
. then there exist @xmath314 and @xmath315 such that for any @xmath316 @xmath317 \geq 1-n^m e^{-\sigma_1{n^{1/2}}}.\ ] ] _ proof . _
we begin by proving ( i ) .
consider the sets @xmath318 note that if @xmath319 , then @xmath320 , i.e. , in @xmath19 there are at most @xmath321 piles with at least @xmath322 cards . clearly ,
if a pile had no more than @xmath322 cards , then there is @xmath323 such that by the time @xmath324 that pile will be empty with probability at least @xmath325 . during the time
@xmath324 only @xmath324 new piles can appear , so , since @xmath326 , for any @xmath327 we have that , @xmath328 \geq 1 - n e^{-c_1{n^{1/2}}}.\ ] ] since @xmath329 , ( [ a1 ] ) also shows that for any @xmath330 we have @xmath331 \leq n e^{-c_1{n^{1/2}}}$ ] . on the other hand ,
if @xmath332 , then clearly @xmath331=0 $ ] , so ( since @xmath333 ) for any @xmath334 we have @xmath335 \leq n e^{-c_1{n^{1/2}}}.\ ] ] now we use ( [ osn_fakt ] ) with @xmath336 and @xmath337 to obtain from ( [ a1 ] ) and ( [ a2 ] ) that for some @xmath338 @xmath339 for @xmath340 define @xmath341 suppose that @xmath342 for some @xmath343 , and let us try to figure out what the configuration @xmath344 should look like . note that * @xmath345 , and moreover @xmath346 for all @xmath347 , so , since @xmath348 , no new pile of size greater than @xmath349 can appear until the moment @xmath350 ; * the evolution of a single pile can be modeled by a random walk on @xmath351 which jumps one unit to the left with probability @xmath2 and holds its position with probability @xmath1 .
this shows that if the size of the pile was less than @xmath352 , then after @xmath350 moves it will be less than @xmath349 with probability at least @xmath353 for some @xmath354 . from the above facts we deduce that for any @xmath342 @xmath355 \leq n e^{-c_3{n^{1/2}}},\ ] ]
as long as @xmath343 . now using ( [ osn_fakt ] ) with @xmath356 and @xmath337
, we obtain that for some @xmath357
so by induction one can show that @xmath359 . summing over @xmath343 and recalling ( [ aaa ] )
, we conclude the proof of the part ( i ) of lemma [ reasonable ] ( with @xmath360 , @xmath361 ) .
to prove the part ( ii ) , first observe that in the proof of ( i ) we have constructed @xmath309 in such a way that for any @xmath313 @xmath362 \leq e^{-c{n^{1/2}}}.\ ] ] to complete the proof of ( ii ) , it is enough to take @xmath363 , @xmath364 ( note that for any @xmath19 we have @xmath365 , @xmath366 ) .
we continue proving theorem [ ran_bs ] .
now , the main idea is the following : first , to dominate the random bulgarian solitaire by a certain deterministic process ( that we will call bulgarian solitaire with immigration of particles ) , and then apply to that process some methods from the proof of theorem [ det_bs ] .
fix @xmath367 , and abbreviate @xmath368 .
denote also @xmath369 $ ] .
for any @xmath19 , let us define configurations @xmath370 in the following way : if @xmath371 , let @xmath372 where @xmath373 are chosen in such a way that @xmath374 and for any @xmath182 we have @xmath375 .
define the operator @xmath376 by @xmath377 i.e. , making the @xmath376-move consists of making a move of deterministic bulgarian solitaire , and then adding @xmath378 cards to the new pile ( so that @xmath379 ) . for the simplicity of notations
, we do not indicate in @xmath376 the dependence on @xmath0 and @xmath380 ; note also that in the above display we do not assume that @xmath175 , so @xmath376 need not apply to only @xmath182 . for two configurations @xmath37 , @xmath38
we say that @xmath381 if @xmath382 and @xmath383 for all @xmath384 .
[ dominacao ] suppose that @xmath175 and @xmath313 ( where @xmath309 are the quantities from lemma [ reasonable ] ) .
then for any @xmath385 there exists @xmath386 such that @xmath387 \geq 1 - n^m e^{-\sigma_2n^{\delta_0}}.\ ] ] _ proof .
_ let us refer to the @xmath18 piles of @xmath19 and @xmath388 as @xmath389 and @xmath390 respectively .
then , the piles born at the moment @xmath95 are referred to as @xmath391 and @xmath392 . using the notation @xmath393 , for @xmath394 , let @xmath395 and @xmath396 stand for the sizes of the piles @xmath397 and @xmath398 at the moment @xmath31 , respectively ( if a pile is emptied at some moment @xmath399 , then we mean that the size remains @xmath400 for all @xmath401 ) .
clearly , the event @xmath402 where we define the event @xmath403 by @xmath404 define also the event @xmath405 ; by lemma [ reasonable ] ( ii ) we know that @xmath406\geq 1 - n^m e^{-\sigma_1{n^{1/2}}}$ ] . on the other hand , @xmath407 \geq \pr[h_i]\pr[\lambda_i\mid h_i , d],\ ] ] where @xmath408 .
now , on @xmath409 we have that @xmath410 for all @xmath411 , and on @xmath412 it holds that @xmath413 . using the large deviation bound for the binomial distribution
, we see that the first term in the right - hand side of ( [ uyi * * ] ) is at least @xmath414 . as for the second term , note that the difference between @xmath415 and @xmath416 is a random walk with drift @xmath400 . since the time that the pile @xmath397 needs to be emptied is @xmath16 , the second term in ( [ uyi * * ] ) is in fact the probability that such a random walk does not deviate from its initial position by more than @xmath417 by time @xmath16 ; clearly , that probability is bounded from below by @xmath418 .
then , it is immediate to deduce lemma [ dominacao ] from ( [ mnbvs ] ) and ( [ uyi * * ] ) .
recall that ( cf .
the proof of lemma [ prop_e ] ) for any configuration @xmath19 we use the notation @xmath419 [ nao_fugira ] suppose that @xmath313 and let @xmath420 be such that @xmath421
. then there exists @xmath422 such that @xmath423 for all @xmath424 .
_ let @xmath425 and denote @xmath426 .
define the triangular configuration @xmath427 by @xmath428 ; then , clearly , @xmath429 .
denote @xmath430 , @xmath431 and define the configuration @xmath432 by @xmath433 . by examining the @xmath376-evolution of @xmath427 on the etienne diagram ,
it is clear that @xmath434 for all @xmath435 .
we then repeat this construction by defining @xmath436 , @xmath437 and the configuration @xmath438 by @xmath439 .
analogously , we obtain that @xmath440 for all @xmath441 .
a simple monotonicity argument then shows that @xmath442 for all @xmath443 .
we have @xmath444 and @xmath445 for some @xmath185 , so @xmath446 for all @xmath447 .
so , since @xmath448 , for some @xmath422 we have @xmath449 for all @xmath0 , thus concluding the proof of lemma [ nao_fugira ] .
[ iter_rand ] fix some @xmath450 and suppose that a configuration @xmath313 is such that @xmath451 for some @xmath452
. then , with @xmath257 from ( [ iteration ] ) , we have for some @xmath453 @xmath454 \geq 1 - e^{-\sigma_3n^{\delta_1}},\ ] ] where @xmath455 .
moreover , ( [ eq_l2.8 ] ) remains true when @xmath456 is substituted by @xmath457 , for any @xmath458 $ ] .
_ first , each particle added to @xmath459 changes the energy @xmath156 by at most @xmath16 , so we have for some constants @xmath460 @xmath461 using the same sort of argument and the fact that @xmath462 for any @xmath463 , with the help of lemma [ nao_fugira ] and ( [ asd ] ) we obtain @xmath464 introduce the event @xmath465 . by lemma [ dominacao ]
we have @xmath466 \geq 1-n'_s e^{-\sigma_2n^{\delta_0}},\ ] ] and , since @xmath467 analogously to ( [ asd])([asd ] ) we obtain that on @xmath468
@xmath469 now , we have that @xmath470 , and @xmath471 . since @xmath472
, we obtain the proof of ( [ eq_l2.8 ] ) from ( [ iteration ] ) , ( [ asd ] ) , ( [ fff * * * ] ) , and ( [ fff * * * * ] ) . as for the second claim of lemma [ iter_rand ] , we note that for @xmath473 , by lemma [ prop_e ] ( ii ) it holds that @xmath474 , and then use the same kind of estimates as used above
. now we are ready to finish the proof of theorem [ ran_bs ] . by lemma [ reasonable ] ( i ) there are @xmath311 such that ( [ sdf ] ) holds .
note that there exists @xmath475 such that if @xmath313 , then @xmath476 .
define @xmath477 .
let @xmath478 and @xmath479 for @xmath480 .
take @xmath481 , @xmath74 , and define @xmath482 ( cf .
( [ a12 ] ) and ( [ a34 ] ) ) .
let @xmath483 ; since @xmath484 , by examining the iteration scheme @xmath485 we obtain that there exists @xmath185 such that @xmath486 .
let @xmath487\},\ ] ] and define also @xmath488 , @xmath489 . take any @xmath490 and denote @xmath491 . by ( [ osn_fakt ] ) and
lemma [ reasonable ] ( i ) we can write @xmath492 where @xmath493 now , by lemma [ iter_rand ] , the left - hand side of ( [ linha1 ] ) can be bounded from below as follows : @xmath494 \nonumber\\ & \geq&\pi_{p , n}(\l_n ) ( 1-e^{-\sigma_3n^{\delta_1 } } ) . \label{yyy}\end{aligned}\ ] ] again using lemma [ iter_rand ] , we write @xmath495 using now ( [ yyy ] ) and ( [ yyy ] ) together with the trivial bound @xmath496 , we obtain from ( [ linha1])([linha2 ] ) that for some @xmath338 @xmath497 by induction , we then obtain that there is @xmath354 such that for any @xmath498 @xmath499 so , since @xmath486 , taking summation in ( [ qqq ] ) we obtain for some @xmath357 that @xmath500 now , the last step of the proof of theorem [ ran_bs ] is analogous to what was done in lemma [ srezaem_hvosty ] . note that if @xmath501 , then @xmath502 so if @xmath501 , then @xmath503 , thus showing that @xmath504 define @xmath505 take any @xmath313 , and denote @xmath506 ( recall that @xmath375 ) . using ( [ osn_fakt ] ) , we write @xmath507 where @xmath508 observe that if @xmath509 and @xmath272 is small enough , then after @xmath510 moves there will be no particles in the set @xmath511 , with probability at least @xmath512 for some @xmath195 .
so , for the left - hand side of ( [ lll1 ] ) we can write @xmath513 on the other hand , the same argument implies that @xmath514 and the bound @xmath515 is trivial .
so , using ( [ oc_en ] ) and ( [ lhs1 ] ) , we obtain from ( [ lll1 ] ) that @xmath516 abbreviate @xmath517 and define @xmath518\big\},\end{aligned}\ ] ] @xmath519 , @xmath520 .
analogously to ( [ linha1])([linha2 ] ) and ( [ lll1])([lll2 ] ) , we write @xmath521 where @xmath522 the following fact can be deduced from ( [ poln_tri ] ) : if @xmath523 and @xmath524 for some @xmath525 , then @xmath526 .
then , by examining the evolution of @xmath459 on the etienne diagram and using lemma [ dominacao ] , it is elementary to obtain that for any @xmath527 , @xmath528 @xmath529 \geq 1 - e^{-c_7{n^{1/2}}}.\ ] ] using that fact , one can bound the left - hand side of ( [ zzz1 ] ) from below by @xmath530 and the term @xmath531 can be bounded from above by @xmath532 . then , it is straightforward to write @xmath533 , @xmath534 .
denoting now @xmath535 , analogously to ( [ qqq])([qqq ] ) we obtain @xmath536 summing over @xmath537 and taking ( [ jsueh ] ) and ( [ oc_h0 ] ) into account , we finally obtain that for some @xmath538 ( depending on @xmath539 ) @xmath540 since @xmath481 and @xmath74 is arbitrary , we complete the proof of theorem [ ran_bs ] ( note that @xmath541 for @xmath542 ) .
a natural question that one may ask is : starting from an initial configuration @xmath19 with @xmath543 , how many steps ( of the deterministic game ) are necessary to reach @xmath544 where @xmath545 as @xmath48 . from the proof of theorem [ det_bs ] it can be deduced that if @xmath546 , @xmath547 , then @xmath548 moves suffice . however , this result is only nontrivial when @xmath549 ( since @xmath155 moves are always enough to reach the `` exact '' triangle ) , and even then it is almost certainly far from being precise . also , loosely speaking , theorem [ ran_bs ] shows that the typical deviation from the triangle is of order at most @xmath550 .
again , we do not believe that that result is the best possible one .
in fact , the author has strong reasons to conjecture that the typical deviation should be of order @xmath551 ; however , the proof of that is still beyond our reach .
the author is thankful to pablo ferrari for many useful discussions about the random bulgarian solitaire , and to ira gessel , who posed the problem of finding the limiting shape for that model during the open problems session at the conference _ discrete random walks 2003 _ ( ihp , paris ) .
also , the author thanks the anonymous referees for careful reading of the manuscript and useful comments and suggestions . | we consider a stochastic variant of the game of bulgarian solitaire @xcite .
for the stationary measure of the random bulgarian solitaire , we prove that most of its mass is concentrated on ( roughly ) triangular configurations of certain type . +
* keywords : * shape theorem , triangular configuration , markov chain , stationary measure departamento de estatstica , instituto de matemtica e estatstica , universidade de so paulo , rua do mato 1010 , cep 05508090 , so paulo sp , brasil . + e - mail : [email protected] |
in contrast with the great success of lattice gauge theory , lattice fermions remain a long - standing problem .
naive discretization causes the species doubling problem @xcite . the situation does not change
regardless of how the lattice spacing is reduced as long as the space - time derivative is modeled as a naive difference .
many attempts have been made to fix the doubling problem @xcite .
wilson removed doublers at low energy by introducing an interaction that mixes left- and right - handed fermions @xcite .
however , unwanted degeneracy persist at high energy and chiral symmetry is explicitly broken . to fix these problems ,
kaplan modified wilson s fermion with an extra dimension @xcite and succeeded in realizing approximate chiral symmetry .
however , the cost of numerical calculations based on it is not cheap . if we find a method to perform such calculations without the extra dimension , calculation time decreases largely and a deeper understanding of quantum field theory becomes possible .
the lattice fermion has another serious problem .
the fermion propagator defined on a lattice deviates from the continuum one even if the doublers are removed with the existing techniques such as kaplan s fermion @xcite .
we need to modify the discretized propagator somehow so that it is close to the continuum one as far as possible .
in general , the extra dimension can be expressed as hopping interactions in a lower - dimensional system .
kaplan s fermion is a formulation with an extra dimension , so there must be a corresponding hamiltonian with no extra dimension .
also , the shape of the fermion propagator can be improved with hopping interactions . in this talk ,
based on a hamiltonian formalism , we introduce ultralocal hopping interactions to remove doublers and improve momentum dependence of fermion energy . ( the word
ultralocal " means that fermion hopping is restricted to a finite range on a real - space lattice @xcite . ) from knowledge of the continuum theory , we know the correct momentum dependence of the energy .
we start from momentum space and go back to real space by way of discrete fourier transform .
@xcite for the details of this work .
let us consider a free dirac fermion on a ( 1 + 1)-dimensional hamiltonian lattice .
@xmath0 where @xmath1 is an index for momentum and @xmath2 is the number of sites .
@xmath3 and @xmath4 are discrete fourier transform of real - space two - component fermion operators @xmath5 and @xmath6 for @xmath7 , respectively .
the system is quantized with anticommutators in the usual way .
periodic boundary conditions are assumed in real space , @xmath8 .
we try to create a real - space hamiltonian with no doubler that reproduces eq .
( [ h1 ] ) . to find real - space representation of @xmath9 ,
let us consider the following function @xmath10 in the limit @xmath11 , the function ( [ sp ] ) goes to @xmath12 .
the function ( [ sp ] ) necessarily has a node at the boundary @xmath13 because it has a periodicity of @xmath14 .
the node is the cause of the doubling problem .
the doubler remains as a singularity at the boundary even if the limit @xmath11 is taken ( by doubler " we mean unwanted energy degeneracy that is not contained in the continuum theory ) .
anyway , the parameter @xmath15 needs to be small for practical formulation because @xmath15 corresponds to the maximum distance of fermion hopping .
in addition to the doubler around the momentum boundary , the function ( [ sp ] ) has another degeneracy .
it oscillates around @xmath12 and has local minima if the summation is truncated with a small @xmath15 . in fourier analysis
, it is called the gibbs phenomenon , which occurs if a function to be expanded has a singularity @xcite .
the oscillation can be removed by replacing eq .
( [ sp ] ) with @xmath16 where @xmath17 @xmath18 is called the lanczos factor @xcite .
the factor almost removes the oscillation of eq .
. however , the doubler modes around the boundary still remain .
we are going to remove them by a trick with hopping interactions .
let us consider the following momentum - space hamiltonian : @xmath19 where @xmath20 and @xmath21 is fermion mass .
we introduce interactions @xmath22 that mix left- and right - handed fermions @xmath23 where @xmath22 are assumed to be nonzero only for @xmath24
. the hamiltonian ( [ h4 ] ) can be diagonalized .
@xmath25 where @xmath26 are energies of one particle states and @xmath27 are transformed variables .
the coefficients @xmath22 are determined so that energy of one - particle states becomes close to the continuum one as far as possible . in fig .
[ fig3 ] , the functions @xmath28 , @xmath29 , and @xmath30 are compared , where @xmath31 .
the function @xmath30 corresponds to energy of one - particle states given by the hamiltonian ( [ h4 ] ) with @xmath32 .
the function @xmath30 agrees well with the correct energy @xmath28 from the continuum theory in the fundamental region @xmath33 .
the hamiltonian ( [ h4 ] ) approximately reproduces the continuum theory without doubler if we identify @xmath34 and @xmath35 .
figure [ fig4 ] shows the diagonal and off - diagonal matrix elements of the transformed @xmath36 in the new basis that diagonalizes the hamiltonian ( [ h4 ] ) with @xmath32 .
the diagonal ( 1,1 ) element of @xmath37 is almost unity at low and intermediate energy @xmath38 and deviates from unity at @xmath39 .
the off - diagonal ( 1,2 ) element of @xmath37 oscillates around zero at @xmath38 and becomes unity at @xmath40 . at low energy , the deviation of the off - diagonal elements from zero
is not large and becomes smaller as the parameter @xmath15 increases . at @xmath38 , the transformed left- and right - handed fermions
have approximately the correct chiral charges @xmath41 and @xmath42 , respectively . the low - energy hamiltonian ( [ h4 ] ) has approximate chiral symmetry because the commutation relation between the hamiltonian ( [ h4 ] ) and chiral charge defined with @xmath37 is almost zero for small @xmath1 .
the errors associated with chiral symmetry can be improved in a systematic way by increasing @xmath15 .
the value used here for the parameter @xmath15 is sufficiently small and does not deny application of the model to actual numerical analysis with a computer .
the real - space hamiltonian and the euclidean action for the dirac spinor is obtained by substituting the discrete fourier transform into eq .
( [ h4 ] ) . as usual ,
gauge symmetry can be implemented by inserting exponentiated gauge fields between the hopping fermions .
the continuum limit @xmath43 is taken with @xmath15 fixed .
the method can be extended to higher - dimensional cases by taking care of the doubler of each direction in the same way . in this talk ,
i have shown how to realize approximate chiral symmetry at low energy .
explicit breaking of chiral symmetry has been compressed to high energy . in future works , it should be precisely checked if insertion of gauge interactions affects chiral properties .
9 k. g. wilson , in _ new phenomena in subnuclear physics _ , erice , 1975 edited by a. zichichi ( plenum , new york , 1977 ) .
j. kogut and l. susskind , phys .
d * 11 * , 395 ( 1975 ) .
s. d. drell , m. weinstein , and s. yankielowicz , phys .
d * 14 * , 487 ( 1976 ) .
d. b. kaplan , phys .
b * 288 * , 342 ( 1992 ) ; m. f. golterman , k. jansen and d. b. kaplan , _ ibid . _ * 301 * , 219 ( 1993 ) m. creutz and i. horvath , phys .
d * 50 * , 2297 ( 1994 ) .
h. neuberger , phys .
b * 417 * , 141 ( 1998 ) ; * 427 * , 353 ( 1998 ) .
m. luscher , phys .
b * 428 * , 342 ( 1998 ) .
w. bietenholz , eur .
j. c * 6 * , 537 ( 1999 ) .
t. sugihara , phys .
d * 68 * , 034502 ( 2003 ) .
g. b. arfken and h. j. weber , mathematical methods for physicists , ( academic press , new york , 1995 ) . | the species doubling problem of the lattice fermion is resolved by introducing hopping interactions that mix left- and right - handed fermions around the momentum boundary .
approximate chiral symmetry is realized on the lattice .
the deviation of the fermion propagator from the continuum one is small . |
the process @xmath1 represents a very important background reaction to searches at the lhc , in particular to @xmath2 production , where the higgs boson decays into a @xmath3 pair .
a successful analysis of this production channel at the lhc requires the knowledge of direct @xmath0 production at nlo in qcd @xcite .
moreover , the calculation of nlo qcd corrections to @xmath4 processes at the lhc represents the current technical frontier .
the complexity of such calculations triggered the creation of prioritized experimenters wishlists where the @xmath0 production ranges among the most wanted candidates @xcite .
the nlo computation for this process has been completed only very recently in @xcite and then confirmed in @xcite where we demonstrated the power of our system based on helac - phegas @xcite , which has , on its own , already been extensively used and tested in phenomenological studies see @xcite , helac-1loop @xcite , which is not yet publicly available , cuttoolspittau / cuttools/ ] @xcite and helac - dipoles @xcite in a realistic computation with six external legs and massive partons . in the following we
briefly summarize the calculation of the virtual and real corrections and present numerical results for the lhc .
a one - loop @xmath5particle amplitude can be expressed in terms of a basis of known @xmath6 , @xmath7 , @xmath8 and @xmath9point scalar integrals : boxes , triangles , bubbles and tadpoles .
the coefficients depend in general on the dimension of space - time , @xmath10 , which , upon expansion around @xmath11 , generates a rational function in the invariants ( rational term ) .
the concept behind modern methods of evaluation , is the direct determination of the coefficients , without recurring to a passarino - veltman reduction . in our case ,
the coefficients and one part of the rational term ( see @xcite ) are extracted via the opp reduction method at the integrand level @xcite , which is implemented in cuttools .
the second part of the rational term coming from the epsilon - dimensional contributions in the numerator is computed with the help of dedicated feynman rules @xcite .
the opp reduction is based on a representation of the numerator of amplitudes , a polynomial in the integration momentum , in a basis of polynomials given by products of the functions in the denominators .
clearly , the cancellation of such terms with the actual denominators will lead to scalar functions with a lower number of denominators . by virtue of the proof provided by the passarino - veltman reduction
, we will end up with a tower of four - point and lower functions , as mentioned before . the determination of the decomposition in the new basis proceeds recursively , by setting chosen denominators on - shell .
this is where the opp method resembles generalized unitarity @xcite .
for most recent applications see @xcite it is important to stress , that working around four dimensions , allows to compute the numerator function in four dimensions .
the difference to the complete result is of the order of @xmath12 , and can therefore be determined a posteriori in a simplified manner . since the calculation of the coefficients of the reduction requires the evaluation of the numerator function for a given value of the loop momentum , the corresponding diagrams can be thought as tree level ( all momenta are fixed ) graphs . to complete the analogy , one needs to chose a propagator and consider it as cut . at this point
the original amplitude for an @xmath13 particle process becomes a tree level amplitude for an @xmath14 particle process .
the advantage is that its value can be obtained by a tree level automate such as helac .
the bookkeeping necessary for a practical implementation is managed by a new software , helac-1loop .
to recapitulate , a complete calculation involves three building components : 1 ) cuttools , for the reduction of integrals with a given numerator to a basis of scalar functions ; 2 ) helac-1loop for the evaluation of the numerator functions for given loop momentum ( fixed by cuttools ) ; 3 ) oneloophameren/ ] @xcite , a library of scalar functions , which provides the actual numerical values of the integrals . at the end
, the 1-loop result is given in the form of real and imaginary parts of the finite term and of the coefficients of the @xmath15 and @xmath16 poles .
the above procedure provides the bare 1-loop amplitude .
renormalization is performed as usual , by evaluating tree level diagrams with counterterms . for our process
, we chose to renormalize the coupling in the @xmath17 scheme , but the mass in the on - shell scheme ( wave function renormalization is done in the on - shell scheme as it must be ) .
notice , that also this part is performed in four dimensions .
this means , that the whole procedure is equivalent to the t hooft - veltman version of dimensional regularization @xcite .
let us emphasise that all parts are calculated fully numerically in a completely automatic manner .
the singularities from soft or collinear gluon emission are isolated via dipole subtraction for nlo qcd calculations @xcite using the formulation for massive quarks @xcite and for arbitrary polarizations @xcite . after combining virtual and real corrections ,
singularities connected to collinear configurations in the final state as well as soft divergencies in the initial and final states cancel for collinear - safe observables automatically after applying a jet algorithm .
singularities connected to collinear initial - state splittings are removed via factorization by pdf redefinitions .
we do not use finite dipoles regularizing the quasi - collinear divergence induced by both top quarks moving in the same direction . due to the large top quark mass
, they are not needed to improve the numerical convergence .
calculations are performed with the help of the helac - dipoles software , which is a complete and publicly available automatic implementation of catani - seymour dipole subtraction and consists of phase space integration of subtracted real radiation and integrated dipoles in both massless and massive cases .
let us stress at this point , that a phase space restriction on the contribution of the dipoles as originally proposed in @xcite is also implemented .
all results presented in the next section have been obtained with @xmath18 parameter .
the phase - space integration is performed with the multichannel monte carlo generator phegas @xcite and optimized via parnihameren/ ] @xcite .
we present predictions for @xmath19 at @xmath20 tev . for the top - quark mass
we take @xmath21 gev , while all other qcd partons , including b quarks , are treated as massless .
final states involving collinear gluons and b - quarks are recombined into collinear - safe jets by means of the @xmath22-algorithm @xcite .
specifically we require two b - quark jets with separation @xmath23 in the rapidity azimuthal - angle plane .
motivated by the search for a @xmath24 signal at the lhc @xcite , we impose the following additional cuts on the transverse momenta and rapidity of the b - quark jets : @xmath25 gev , @xmath26 .
the outgoing ( anti)top quarks are neither affected by the jet algorithm nor by phase - space cuts .
we use cteq6 pdfs @xcite .
most precisely we use the cteq6 m parton distributions at nlo , and the cteq6l1 set at lo .
the suppressed contribution from b quarks in the initial state has been neglected .
scale dependence of the total cross section for @xmath27 at the lhc with @xmath28 .
the blue dashed curve corresponds to the leading order , whereas the red solid one to the next - to - leading order result .
_ , scaledwidth=55.0% ] right panel : distribution of the invariant mass @xmath29 of the @xmath3 pair for @xmath30 at the lhc at lo ( blue dashed line ) and nlo ( red solid line )
. left panel : ratio of the nlo and lo distributions._,title="fig:",scaledwidth=45.0% ] right panel : distribution of the invariant mass @xmath29 of the @xmath3 pair for @xmath30 at the lhc at lo ( blue dashed line ) and nlo ( red solid line ) .
left panel : ratio of the nlo and lo distributions._,title="fig:",scaledwidth=42.5% ] right panel : distribution in the rapidity @xmath31 of the @xmath3 pair for @xmath32 at the lhc at lo ( blue dashed line ) and nlo ( red solid line ) .
left panel : ratio of the nlo and lo distributions._,title="fig:",scaledwidth=45.0% ] right panel : distribution in the rapidity @xmath31 of the @xmath3 pair for @xmath32 at the lhc at lo ( blue dashed line ) and nlo ( red solid line ) .
left panel : ratio of the nlo and lo distributions._,title="fig:",scaledwidth=42.5% ] we start with a presentation of the total cross section at the central value of the scale , @xmath33 at lo and nlo . at the central scale , the full pp cross section receives a very large nlo correction of the order of 77% , which is mainly due to the gluonic initial states .
the full lo and nlo cross sections are given by @xmath34 fb and @xmath35 fb respectively .
varying the scale up and down by a factor 2 in a uniform way changes the cross section by 70% in the lo case , while in the nlo case we have obtained a variation of the order 33% .
subsequently , in fig . [
fig : scale ] we show the result for the scale dependence graphically . in the next step the differential distributions are depicted together with their dynamical k - factors .
invariant mass and rapidity of the two-@xmath36-jet system are presented in fig .
[ fig : mbb ] and in fig .
[ fig : ybb ] respectively .
clearly , the distributions show the same large corrections , which turn out to be relatively constant .
this large scale variation and the size of the corrections themselves , imply that if a meaningful analysis were required in the present setup , additional cuts on extra jets must be introduced in order to reduce the nlo corrections .
a brief summary of the calculations of nlo qcd corrections to the @xmath37 process at the lhc has been presented .
the qcd corrections to the integrated cross sections are found to be very large , changing the leading - order results by about 77% .
the distributions show the same large corrections .
moreover , the shapes of some kinematical distributions change appreciably compared to leading order .
the residual scale uncertainties of the nlo predictions are at the 33% level . work supported in part by the initiative and networking fund of the helmholtz association , contract ha-101 ( `` physics at the terascale '' ) and by the rtn european programme mrtn - ct-2006 - 035505 heptools - tools and precision calculations for physics discoveries at colliders .
10 the atlas collaboration , g. aad _ et .
al . _ , expected performance of the atlas experiment - detector , trigger and physics , arxiv:0901.0512 [ hep - ex ] .
nlo multileg working group collaboration , z. bern _ et . al .
_ , the nlo multileg working group : summary report , arxiv:0803.0494 [ hep - ph ] .
a. bredenstein , a. denner , s. dittmaier and s. pozzorini , nlo qcd corrections to @xmath38 at the lhc , phys .
103 ( 2009 ) 012002 .
g. bevilacqua , m. czakon , c. g. papadopoulos , r. pittau and m. worek , assault on the nlo wishlist : @xmath39 , jhep 0909 ( 2009 ) 109 .
a. kanaki and c. g. papadopoulos , helac : a package to compute electroweak helicity amplitudes , comput .
commun . 132
( 2000 ) 306 . c. g. papadopoulos , phegas : a phase space generator for automatic cross section computation , comput .
commun . 137
( 2001 ) 247 .
a. cafarella , c. g. papadopoulos and m. worek , helac - phegas : a generator for all parton level processes , comput .
commun . 180
( 2009 ) 1941 .
t. gleisberg , f. krauss , c. g. papadopoulos , a. schaelicke and s. schumann , cross sections for multi - particle final states at a linear collider , eur .
j. c34 ( 2004 ) 173 . c. g. papadopoulos and m. worek , multi - parton cross sections at hadron colliders , eur .
j. c50 ( 2007 ) 843
. j. alwall _ et .
al . _ , comparative study of various algorithms for the merging of parton showers and matrix elements in hadronic collisions , eur .
j. c53 ( 2008 ) 473 . c. englert , b. jager , m. worek and d. zeppenfeld , observing strongly interacting vector boson systems at the cern large hadron collider , phys .
d80 ( 2009 ) 035027 .
a. van hameren , c. g. papadopoulos and r. pittau , automated one - loop calculations : a proof of concept , jhep 0909 ( 2009 ) 106 . g. ossola , c. g. papadopoulos and r. pittau , cuttools : a program implementing the opp reduction method to compute one - loop amplitudes , jhep 0803 ( 2008 ) 042 . m. czakon , c. g. papadopoulos and m. worek , polarizing the dipoles , jhep 0908 ( 2009 ) 085 . g. ossola , c. g. papadopoulos and r. pittau , on the rational terms of the one - loop amplitudes , jhep 0805 ( 2008 ) 004 . g. ossola , c. g. papadopoulos and r. pittau , reducing full one - loop amplitudes to scalar integrals at the integrand level , nucl .
b763 ( 2007 ) 147 .
p. draggiotis , m. v. garzelli , c. g. papadopoulos and r. pittau , feynman rules for the rational part of the qcd 1-loop amplitudes , jhep 0904 ( 2009 ) 072 .
m. v. garzelli , i. malamos and r. pittau , feynman rules for the rational part of the electroweak 1- loop amplitudes , arxiv:0910.3130 [ hep - ph ] .
z. bern , l. j. dixon , d. c. dunbar and d. a. kosower , one - loop n - point gauge theory amplitudes , unitarity and collinear limits , nucl .
b425 ( 1994 ) 217 . z. bern , l. j. dixon , d. c. dunbar and d. a. kosower , fusing gauge theory tree amplitudes into loop amplitudes , nucl .
b435 ( 1995 ) 59 .
e. witten , perturbative gauge theory as a string theory in twistor space , commun .
( 2004 ) 189 .
r. britto , f. cachazo and b. feng , generalized unitarity and one - loop amplitudes in n = 4 super - yang - mills , nucl .
b725 ( 2005 ) 275 . z. bern , l. j. dixon and d. a. kosower , on - shell methods in perturbative qcd , annals phys . 322 ( 2007 ) 1587 .
r. k. ellis , w. t. giele and z. kunszt , a numerical unitarity formalism for evaluating one - loop amplitudes , jhep 0803 ( 2008 ) 003 .
w. t. giele , z. kunszt and k. melnikov , full one - loop amplitudes from tree amplitudes , jhep 0804 ( 2008 ) 049 .
r. k. ellis , k. melnikov and g. zanderighi , generalized unitarity at work : first nlo qcd results for hadronic @xmath40 + 3jet production , jhep 0904 ( 2009 ) 077 . c. f. berger _
_ , precise predictions for @xmath40 + 3 jet production at hadron colliders , phys .
102 ( 2009 ) 222001 .
r. keith ellis , k. melnikov and g. zanderighi , @xmath40 + 3jet production at the tevatron , arxiv:0906.1445 [ hep - ph ] . c. f. berger _
et al . _ ,
next - to - leading order qcd predictions for w+3-jet distributions at hadron colliders , arxiv:0907.1984 [ hep - ph ] .
k. melnikov and m. schulze , nlo qcd corrections to top quark pair production and decay at hadron colliders , jhep 0908 ( 2009 ) 049 .
k. melnikov and g. zanderighi , w+3 jet production at the lhc as a signal or background , arxiv:0910.3671 [ hep - ph ] .
g. t hooft and m. j. g. veltman , regularization and renormalization of gauge fieldsnucl .
b44 ( 1972 ) 189 . s. catani and m. h. seymour , a general algorithm for calculating jet cross sections in nlo qcd , nucl .
b485 ( 1997 ) 291 .
s. catani , s. dittmaier , m. h. seymour and z. trocsanyi , the dipole formalism for next - to - leading order qcd calculations with massive partons , nucl .
b627 ( 2002 ) 189 .
z. nagy and z. trocsanyi , next - to - leading order calculation of four - jet observables in electron positron annihilation , phys .
d59 ( 1999 ) 014020 .
z. nagy , next - to - leading order calculation of three jet observables in hadron hadron collision , phys .
d68 ( 2003 ) 094002 .
a. van hameren , parni for importance sampling and density estimation , acta phys .
b40 ( 2009 ) 259 .
s. catani , y. l. dokshitzer and b. r. webber , the @xmath41 perpendicular clustering algorithm for jets in deep inelastic scattering and hadron collisions , phys .
b285 ( 1992 ) 291 . s. catani , y. l. dokshitzer , m. h. seymour and b. r. webber , longitudinally invariant @xmath22 clustering algorithms for hadron hadron collisions , nucl .
b406 ( 1993 ) 187 .
s. d. ellis and d. e. soper , successive combination jet algorithm for hadron collisions , phys .
d48 ( 1993 ) 3160 .
j. pumplin _ et .
_ , new generation of parton distributions with uncertainties from global qcd analysis , jhep 0207 ( 2002 ) 012 .
d. stump _ et .
al . _ , inclusive jet production , parton distributions , and the search for new physics , jhep 0310 ( 2003 ) 046 . | in this contribution , we summarize the results from an nlo computation of the production of @xmath0 in hadronic collisions . the results are obtained by combining two programs : helac-1loop , based on the opp method and the reduction code cuttools , for the virtual one - loop matrix elements and helac - dipoles for the real - emission contributions . selected numerical results
are presented for the lhc .
= by -1 |
understanding the rates and the structure of multi - jet production in hadron
hadron collisions is of primary importance for new physics searches
. production of high transverse momentum jets is a hard process which implies a head - on collision of qcd partons quarks and/or gluons from the small - distance wave functions of initial hadrons .
cross section of a hard collision is small compared with the size of hadron , @xmath6 , with @xmath5 the scale related to transverse momenta of the produced jets , @xmath7 .
therefore , typically it is two partons that experience a hard collision in a given event .
a large angle scattering of these two partons produces two ( or more ) final state partons that manifest themselves as hadron large transverse momentum jets . at the same time
, one can not exclude a possibility that more than one pair of partons happen to collide in a given event , giving rise to a multi - jet event .
a possibility of a double hard collision becomes more important with increase of the energy of the collision where scattering off small @xmath8 partons which have much higher densities becomes possible . in recent years multiparton collisions
have attracted close attention .
following the pioneering work of refs .
@xcite , a large number of related theoretical papers appeared @xcite , based on the parton model and geometrical picture in the impact parameter space .
more recently , this topic has been intensively discussed in view of the lhc program @xcite .
monte carlo event generators that produce multiple parton collisions are being developed @xcite ; theoretical papers exploring properties of double parton distributions and discussing their qcd evolution have appeared @xcite . in our view , however , important elements of qcd that are necessary for theoretical understanding of the multiple hard interactions issue have not yet been properly taken into account by above - mentioned intuitive approaches .
the problem is , sort of , educational : both the probabilistic picture , the mc generator technology is based upon , and the familiar feynman diagram technique , when used in the momentum space , prove to be inadequate for careful analysis and understanding of the physics of multiple collisions . from experience gained by treating standard ( single ) hard processes ,
one became used to a motto that a large momentum transfer scale @xmath5 ensures the dominance of small distances , @xmath9 , in a process under consideration . with the multiple collisions under focus
, however , one has to distinguish two space - time scales : that of _ localization _ of the parton participating in a hard interaction , @xmath10 , and that of _ transverse separation _ , @xmath11 , between the two hard collision vertices .
the latter can be large , of the order of the hadron size , even for large @xmath5 . in order to be able to trace the relative distance between the partons
, one has to use the mixed longitudinal momentum
impact parameter representation which , in the momentum language , reduces to introduction of a mismatch between the transverse momentum of the parton in the amplitude and that of the same parton in the amplitude conjugated . another unusual feature of the multiple collision analysis that may look confusing at the first sight is the fact that even at the tree level _ the amplitude _ describing the double hard interaction process contains additional integrations over longitudinal momentum components ; more precisely over the difference of the ( large ) light - cone momentum components of the two partons originating from the same incident hadron ( see section [ subsec : pt34 ] ) . in the previous short publication @xcite
we have considered production of two pairs of _ nearly back - to - back _ jets resulting from simultaneous hard collisions of two partons from the wave function of one incident hadron with two partons from the other hadron ( `` four - to - four '' processes ) .
as we have shown , this necessitates introduction of a new object a generalized double parton distribution , @xmath3gpd , that depends on a new transverse momentum parameter @xmath12 conjugate to the relative distance between the two partons in the hadron wave function .
generalized double parton distributions provide a natural framework for incorporating longitudinal and transverse correlations between partons in the hadron wave function at the @xmath13 scale , and for tracing the perturbative @xmath5-evolution of the correlations .
the corresponding 4-jet cross section can be expressed in terms of @xmath3gpd s as follows [ eq:1 ] & = & d_a(x_1,x_2 ; ) d_b(x_3,x_4;- ) .
the factor on the second line has dimension of inverse area : = , [ b3 ] where @xmath14 are the corresponding one - parton distributions .
the ratio of the product of two single - inclusive cross sections and the double - inclusive cross section ( @xmath15 ) is often referred to in the literature as `` an effective cross section '' @xmath16 .
we prefer , however , not to look at this quantity as a cross section , since it reflects transversal area of parton overlap as well as longitudinal correlations of the partons . at the same time , it has little to do with the measure of the strength of the interaction , which is what `` cross section '' represents . in a two - parton collision , scattered partons form two nearly back - to - back jets , while additional jets ( should there be any ) tend to be softer and to align with the directions of initial and final partons , because of collinear enhancements due to radiative nature of secondary partons
. such will be typical characteristics of a _
4-jet event _ , in particular . on the other hand , four jets produced as a result of a _
double hard collision _ of two parton pairs would , on the contrary , form _ two pairs _ of _ nearly back - to - back _ jets .
this kinematical preference is in stark contrast with `` hedgehog - like '' configurations of four jets stemming from a single collision and can be used in order to single out double hard collisions experimentally .
such experimental studies were recently carried out by the cdf and d0 collaborations who have studied production of three jets + photon @xcite .
the analysis of the data performed in @xcite using information about generalized parton distribution ( gpds ) obtained from the study of hard exclusive processes at hera has found that observed @xmath17 rates were a factor @xmath18 higher than the expected rates based on a naive model that neglected correlations between partons in the transverse plane .
the use of @xmath3gpd allows one to incorporate such correlations and predict their @xmath5 evolution . on the theory side ,
the back - to - back enhancement has been discussed , at tree level , in a number of studies of various channels ( see , for example , discussion of the 2 jets+@xmath19 in @xcite and references therein ) . in the present paper
we study perturbative radiative effects in the _ differential _
4-jet distribution in the back - to - back kinematics and derive the expression for the corresponding cross section in the leading logarithmic collinear approximation .
it takes into account qcd evolution of the generalized double parton distributions as well as effects due to multiple soft gluon radiation , and turns out to be a direct generalization of the known `` ddt formula '' for back - to - back production of two large transverse momentum particles in hadron collisions @xcite .
we also discuss and treat new specific correlations between transverse momenta of jets due to 3-parton interactions producing 4 jets,``three - to - four '' .
such processes are induced by perturbative splitting of a parton from one of the hadrons , the offspring of which enter double hard collision with two partons from the wave function of the second hadron .
the hard scale of this parton splitting is determined by transverse momentum imbalances of pairs of jets , @xmath20 , @xmath21 , and exhibits specific collinear enhancement in the kinematical region where two jet imbalances practically compensate one another , @xmath22 . consistently taking into account three - to - four parton process
solves a longstanding problem of double counting in treating multi - parton interactions .
discussion of the 2-parton distribution has a long history .
it is commonly defined in the momentum space as a 2-particle inclusive quantity depending on two parton momenta , see @xcite .
being related to ( the imaginary part of ) a certain _ forward _ scattering amplitude , it therefore disregards impact parameter space geometry of the interaction .
exploring properties of 2-parton distributions so defined , an approach to the study of the multiple jet production has been recently suggested in ref .
the reason why this approach has faced difficulties , @xcite , and did not solve , in our view , the problem of systematic pqcd analysis of 4-jet production is clear : it did not incorporate effects due to variations of the transverse separation between the partons information encoded by @xmath3gpd s but not by the 2-parton momentum distributions .
the @xmath3gpd s were recently used in ref .
@xcite for intuitive description of the total 4-jet production cross section .
however , the differential distributions were not discussed in that paper , and not all relevant pqcd contributions were included , so that our results are different from the ones obtained in @xcite . the paper is organized as follows .
in section [ ptanalysis ] we recall the main ingredients of the perturbative analysis based on selection of maximally collinear enhanced contributions in all orders . in section [ sec : eveq ] we present the evolution equation for generalized two - parton distributions .
section [ sec : pt ] is devoted to the perturbative analysis of small - distance correlations between partons .
the main result of the paper the differential distribution of 4-jet production in the back - to - back kinematics is formulated in section [ sec:4324 ] , and the total cross section of two - dijet production is described in section [ total ] .
conclusions and outlook are presented in section [ conclude ] .
the perturbative approach implies that all hardness ( transverse momentum ) scales that characterize the problem are comfortably larger than the intrinsic qcd scale @xmath23 : @xmath24 .
the process under consideration may have up to five hard scales involved .
indeed , in the leading order in @xmath4 , large transverse momentum partons are produced in pairs and have nearly opposite transverse momenta , setting the hard scale @xmath25 . within the parton model framework ( neglecting finite smearing due to intrinsic transverse momenta of incident partons )
, one has @xmath26 .
secondary qcd processes evolution of initial parton distributions and accompanying soft gluon radiation introduce transverse momentum imbalance : @xmath27 which constitutes another hard scale : @xmath28 . for production of four jets in the back - to - back kinematics , this gives four different hard scales . as we shall see below
, in the 3-partons collisions producing four jets yet another scale enters the game : @xmath29 with @xmath30 the total transverse momentum of the 4-jet ensemble . in what follows we consider transverse momenta of all four jets to be of the same order , @xmath31
this is not necessary but helps to avoid complications in the hierarchy of relevant scales .
finally , let us mention that in what follows it will be tacitly implied that fixing these scales from the largest one , @xmath5 , down to smaller ones , @xmath32 and @xmath33 , is not compromised by uncertainties in determination of the transverse momenta of the jets .
the basic 2-jet production cross section scales , asymptotically , as [ eq:22 ] . according to , production of four jets in simultaneous hard collisions of four partons yields [ eq:44 ] r^-2()^2 , with @xmath34 the characteristic distance between the two partons in the hadron wave function . at large @xmath5
this cross section is parametrically smaller than that for production of four well separated jets with transverse momenta @xmath35 in a 2-parton collision : [ eq:24 ] ( with transverse momenta of two out of four jets being integrated over ) .
qualitatively , the production mechanism can be labelled a `` higher twist effect '' .
nevertheless , it may turn out to be essential comparable with the `` leading twist '' @xmath2 , eq .
if one looks at specific kinematics of the 4-jet ensemble .
let @xmath36 be the direction of colliding hadron momenta .
imagine that we are triggering on two jets moving along the @xmath8 and @xmath37 axes in the transverse plane , and look for two accompanying jets inside some solid angles @xmath38 around the @xmath39 and @xmath40 directions .
the production mechanism does not populate this region : the higher order @xmath2 qcd matrix element is enhanced when two final state partons become _ quasi - collinear _ but is perfectly smooth in the back - to - back kinematics .
therefore , its contribution will be suppressed , @xmath41 contrary to the @xmath1 production mechanism which is concentrated in this very kinematical region .
the differential 4-jet production cross section possesses two collinear enhancements . depending on the kinematics of the jets , they are , symbolically , [ eqs:4434 ] [ eq:443444 ] d^(44 ) & & d^2j_3d^2j_4 d , + & & _ 13 ^ 2 q^2 , _ 24 ^ 2 q^2 ; + [ eq:443434 ]
d^(34 ) & & d^2j_3d^2j_4 d , + & & ^2 ^ 2 q^2 , ^2= _ 13 ^ 2_24 ^ 2 .
here @xmath42 is the cross section integrated over the transverse momenta of the `` backward '' jets @xmath43 and @xmath44 .
the integrated cross section @xmath45 contains the squared matrix element of the four - parton production and is of the order of @xmath46 , cf .
at the born level , the jets in pairs are exactly back - to - back , so that @xmath47 in .
to have a non - zero value of the transverse momentum imbalance , one has to have additional large transverse momentum parton(s ) produced . in the second important contribution to the cross section , eq .
, one power of the coupling emerges from the splitting of a parton from one of the incident hadrons into two , and the second power is due to production of an additional final state parton with @xmath48 . in both cases the smallness due to additional powers of the coupling
is compensated by two broad ( logarithmic ) integrations over transverse momentum imbalances as indicated in . in the leading order in @xmath4 , it suffices to have just one parton present with @xmath49 in order to assure @xmath50 . at the same time
, inclusive production of accompanying partons with transverse momenta @xmath51 turns out to be _ suppressed _ in a broad interval @xmath52 , as long as one wants to preserve the collinear enhancement factor @xmath53 in the jet correlation .
this dynamical `` veto '' has two consequences .
first of all , it results in _ reduction _ of the hardness scale of the parton distributions from the natural scale @xmath5 ( scale of the parton distributions in the integrated cross section ) down to the observation - induced scale @xmath54 .
then , it introduces double logarithmic ( dl ) form factors of participating initial state partons , since the transverse momentum of the jet pair can be compensated not only by a hard ( energetic ) parton from inside initial parton distributions but also by a _
soft gluon _ whose radiation did not affect inclusive parton distributions due to real virtual cancellation . the presence of the dl form factors depending on the logarithm of a large ratio of scales , @xmath55 , is typical for the so - called `` semi - inclusive '' processes @xcite .
production of massive lepton pairs in hadron collisions ( the drell yan process ) is a classical example of a two - scale problem . here
enter form factors of colliding quarks that depend on the ratio of the invariant mass @xmath56 to the transverse momentum of the lepton pair , @xmath57 , in the dominant kinematical region @xmath58 : [ eq : ddt2 ] & & = + & & \
{ d_a^q ( x_1 , q_^2 ) d_b^|q(x_2 , q_^2 ) s_q^2(q^2 , q_^2 ) } .
@xmath59 is the double logarithmic qcd quark form factor .
sudakov quark and gluon form factors can be expressed via the exponent of the total probability of the parton decay in the range of virtualities ( transverse momenta ) between the two hard scales : [ eq : sudakovs ] s_q(q^2,^2 ) & = & \{- _ ^2^q^2 _ 0 ^ 1-k / q dz p_q^q(z ) } , + s_g(q^2,^2 ) & = & \{- _ ^2^q^2 _ 0 ^ 1-k / q dz}. here @xmath60 are the non - regularized one - loop dglap splitting functions ( without the `` + '' prescription ) : [ eq : splits ] p_q^q(z ) & = c_f , p_q^g(z ) = p_q^q(1-z ) , p_g^q(z ) & = t_r , p_g^g(z ) = c_a ; the upper limit of @xmath36-integrals properly regularizes the soft gluon singularity , @xmath61 ( in physical terms , it can be looked upon as a condition that the energy of a gluon should be larger than its transverse momentum , @xcite ) .
the case of hadron interactions producing large transverse momentum _ partons _ ( instead of colorless objects like a drell yan pair or an intermediate boson ) is more involved since here the transverse momentum imbalance may be compensated by qcd radiation from the _ final state partons _ too . the azimuthal correlation between two nearly back - to - back large transverse momentum _ particles _ was considered in @xcite .
an analog of the `` ddt formula '' has been derived in the collinear approximation , which expression contained the product of four form factors , two initial parton distributions and two fragmentation functions .
the case when _ jets _ are being reconstructed in the final state is more complicated to analyze as it yields an answer depending on the jet finding algorithm .
the problem has been addressed by banfi and dasgupta in @xcite where a smart way of defining the final state jets was formulated that permitted to write down a resummed qcd formula for soft gluon effects in `` 2 partons @xmath62 2 jets '' cross sections .
collinear logarithms due to hard splittings of the final state partons do not pose a problem : such secondary partons populate the jets .
partons that appear as separate out - of - jet radiation and are relevant for transverse momentum imbalance compensation have to be produced at sufficiently large angles with respect to the jet axis .
this is the domain of large - angle gluon radiation .
production of soft gluons in - between jets is also logarithmically enhanced and induces single logarithmic ( sl ) corrections , @xmath63^n$ ] , that may also be significant and should be resummed in all orders .
contrary to _ collinear enhanced effects _
( that drive evolution of parton distributions and fragmentation functions and determine the sudakov form factors ) , the large - angle gluon radiation can not be attributed to one or another of the partons participating in the hard scattering .
it is coherent and depends on the kinematics and color topology of the hard parton ensemble as a whole .
as a result , resummation of these sl corrections becomes a matrix problem that involves tracing various color states of the parton system , see @xcite and references therein . in the present paper we concentrate on resummation of _ collinear enhanced _ dl and sl terms and
avoid complications due to _ soft _ sl corrections .
this means ignoring color transfer effects in hard interactions .
thus , production of four jets with large transverse momenta @xmath64 and pair imbalances @xmath65 will be equivalent , in our treatment , to production of two colorless drell yan pairs with invariant masses @xmath66 and transverse momenta @xmath20 and @xmath21 .
generalization of the results of @xcite to the case of double parton scattering seems straightforward and should be considered separately .
the @xmath3gpd in the expression for the multiparton production cross section has a meaning of a two body form factor when partons @xmath67 and @xmath68 receive transverse momenta @xmath69 and @xmath70 leaving the hadron intact
. nonrelativistic analogue of this form factor is familiar from the double scattering amplitude in the momentum space representation of the glauber model , see e.g. @xcite . recall that @xcite the scale @xmath69 in @xmath3gpd is conjugate to the relative transverse distance between the two partons in the @xmath3gpd in the impact parameter representation considered in @xcite .
two partons may originate from soft low - scale fluctuations inside the hadron ; they can also emerge from a perturbative splitting of a common parent parton at relatively large momentum scales .
it is clear that these two contributions to @xmath3gpd will have essentially different dependence on the parameter @xmath69 .
the first contribution we will denote [ eq : d2 ] _ [ 2 ] d^bc_a(x_1,x_2 ; q_1 ^ 2,q_2 ^ 2 ; ) , with the subscript @xmath71}$ ] stressing that here the partons @xmath72 and @xmath73 emerge from the no - perturbative wave function of the hadron @xmath74
. it should decrease rapidly at scales larger than a natural scale of short - range parton correlation in a hadron ( this scale may be slightly different for quarks and gluons and could in principle be significantly larger than the @xmath75 as there exists another non - perturbative scale of the chiral symmetry breaking which maybe as large as 700 mev ) .
@xmath3gpd should rapidly decrease for @xmath76 where @xmath77 is the transverse gluonic radius of the nucleon . in the mean field approximation when the correlations are neglected
, it can be approximated by a factorized expression @xcite [ eq:2dfdef ] [ eq:2ddef ] _ [ 2 ] d^b , c_a(x_1,x_2 ; q_1 ^ 2,q_2 ^ 2 ; ) & = & f_g(^2 ; x_1,q_1 ^ 2)f_g(^2 ; x_2,q_2 ^ 2 ) + & & g_a^b(x_1,q_1 ^ 2 ) g_a^c(x_2,q_2 ^ 2 ) , where @xmath78 are the single - parton distributions and the two - gluon form factor @xmath79 can be parametrized as [ eq : fdef ] f_g(^2 ) = ( 1+)^-2 .
the parameter @xmath80 is of the order of 1 gev@xmath81 for @xmath82 and gradually drops with decrease of @xmath8 and with increase of virtuality @xcite .
the second contribution we will denote [ eq : d1 ] _ [ 1 ] d^b , c_a(x_1,x_2 ; q_1 ^ 2,q_2 ^ 2 ; ) , where the subscript @xmath83}$ ] stands as a reminder of the fact that @xmath72 and @xmath73 originate from perturbative splitting of a single parton from the hadron wave function . this contribution is practically @xmath69-independent and should decrease with @xmath69 much more slowly , due to logarithmic pqcd effects .
( a steep power falloff starts only when @xmath84 exceeds the relevant hard scale , @xmath56 . ) by total cross section in the present context we mean the back - to - back 4-jet cross section integrated over pair jet imbalances in the dominant logarithmic region @xmath85 .
we start by noting that the product of two small - distance parton fluctuations , @xmath86 } \!d \times { } _ { [ 1]}\ !
d $ ] , does not contribute to the process we are interested in . indeed , in this case the integral over @xmath69 in eq .
formally diverges and yields a hard scale ( instead of @xmath87 ) in the numerator .
this means a significant contribution to the cross section but not the one we are looking for .
below in section [ sec24 ] we will explicitly verify that a double hard collision of two parton pairs each of which originates from perturbative splitting , lacks the back - to - back enhancement .
in fact , the product @xmath86}\ !
d \times { } _ { [ 1]}\ ! d$ ] corresponds to a one - loop correction to the `` leading twist '' perturbative production of four jets in a hard collision of two partons ( `` two - to - four '' ) whose distribution is _ smooth _ in the back - to - back region and as such _ gets subtracted _ as background .
keeping this in mind , the back - to - back 4-jet production cross section is proportional to the inverse `` interactions area '' @xmath88 described by the expression = ( _ [ 2]d_a ( ) _ [ 2]d_b ( ) + _ [ 2]d_a ( ) _ [ 1]d_b ( ) + _ [ 1]d_a ( ) _ [ 2]d_b ( ) ) , [ s ] where indices @xmath74 , @xmath72 mark two interacting nucleons .
this expression is somewhat symbolic ; a careful analysis of the `` interaction area '' will be carried out below in sec .
[ sec:4324 ] ( see eq . ) . the first term in eq
. we will refer to as a `` four - to - four '' process : two partons from the wave functions of the hadron @xmath74 interact with two partons from the hadron @xmath72 producing four jets .
the second and the third terms in eq .
describe hard collisions of one parton from one hadron with two partons from the second hadron . until recently
, these `` three - to - four '' processes were commonly ignored in the literature ( see , however , @xcite ) . at the same time
, they turn out to be somewhat enhanced . indeed , the contribution due to four - to - four processes to the geometrical factor eq .
is given by [ eq:44est ] f_g^2(^2)f^2_g(^2)=. fast decrease of the product of two squared form factors leads to fast convergence of the integral whose median is positioned at a value as low as @xmath89 .
the case of the three - to - four process is different .
this process corresponds , as we explained above , to interaction of the offspring of the perturbative splitting of a parton from the wave function of one hadron , with two partons from the non - perturbative wave function of the second colliding hadron . on the side of @xmath83}d$ ] , the parameter @xmath69 enters `` perturbative loop '' due to parton splitting and as a result the dependence of @xmath83}d$ ] on @xmath69 turns out to be only logarithmic , that is , parametrically much slower than that of the non - perturbative form factor @xmath90 .
thus , the three - to - four contribution to the double interaction cross section reduces to _ [ 1]d(x_1,x_2 ; ) f^2_g(^2 ) .
_ [ 1]d|_=0 f^2_g(^2 ) , ( where we have neglected the logarithmic @xmath69-dependence of @xmath83}d$ ] ) .
this corresponds to the fact that in the impact parameter space , the distance between partons coming from a perturbative splitting is much smaller than the hadron size , so that the answer is proportional to the density of non - perturbative two - parton correlation at small distances `` in the origin '' : [ eq:34est ] f^2_g(^2 ) = .
comparison with the estimate shows that the contribution to the cross section of the `` interference term '' @xmath91 is enhanced , relative to the @xmath92 process , by the factor 2~5 .
( for the case of the gaussian form factors this enhancement is 15% smaller a factor of 4 . )
this estimate was obtained for the case when all four partons participating in the hard collisions are gluons .
a detailed numerical study of the @xmath93-dependence of the effective interaction area @xmath88 will be presented in the paper under preparation .
so we conclude that the three - to - four processes may provide a sizable contribution to the cross section even if they constitute a small correction to @xmath3gpd .
thus , we represent the generalized double parton distribution @xmath3gpd as a sum of two terms : [ eq:2terms ] d^b , c_a(x_1,x_2 ; q_1 ^ 2,q_2 ^ 2 ; ) & = & _ [ 2 ] d^b , c_a(x_1,x_2 ; q_1 ^ 2,q_2 ^ 2 ; ) + & + & _ [ 1 ] d^b , c_a(x_1,x_2 ; q_1 ^ 2,q_2 ^ 2 ; ) .
the term @xmath94 } d$ ] describes the distribution of two partons from the non - perturbative wave function of the hadron @xmath74 that are _ independently evolved _ to large perturbative scales @xmath95 and @xmath96 according to the standard one - parton evolution equation .
the perturbative evolution involves momentum scales much larger than the hadron wave function correlation scale @xmath97 .
therefore , the evolution practically does not affect the @xmath69-dependence of the two - parton spectrum .
this piece of the @xmath3gpd acquires but a mild additional logarithmic dependence _ at the tail _ of the @xmath69-distribution in addition to a non - perturbative power falloff .
the integral qcd evolution equation for @xmath94 } d^{bc}_a$ ] reads [ eq : termnp ] _ [ 2 ] d^b , c_a ( & x_1 , x_2 ; q_1 ^ 2,q_2 ^ 2 ; ) = s_b(q_1 ^ 2 , q_^2)s_c(q_2 ^ 2 , q_^2 ) _ [ 2]d^b , c_a(x_1,x_2 ; q_0 ^ 2,q_0 ^ 2 ; ) & + _ b _ q_^2^q_1 ^ 2 s_b(q_1 ^ 2,k^2 ) p_b^b(z ) _ [ 2]d^b,c_a(,x_2;k^2,q_2 ^ 2 ; ) & + _ c _
q_^2^q_2 ^ 2 s_c(q_2 ^ 2,k^2 ) p_c^c(z ) _ [ 2]d^b , c_a(x_1,;q_1 ^ 2,k^2 ; ) .
here @xmath60 are the non - regularized one - loop dglap splitting functions and @xmath98 the double logarithmic sudakov parton form factors defined in eq . . the lower limit of the perturbative evolution in eq .
, [ eq : qmindef ] q_^2 = ( q_0 ^ 2 , ^2 ) q_0 ^ 2+^2 , is the only source of additional ( logarithmic ) @xmath69-dependence .
it emerges when @xmath84 exceeds and substitutes the starting non - perturbative scale @xmath13 of the perturbative evolution . the second term in eq .
, @xmath86 } d^{bc}_a$ ] , represents the small - distance correlation between the two partons that emerge from a perturbative splitting of a common parent parton taken from the hadron wave function .
it can be expressed in terms of standard inclusive single - parton distributions as follows [ eq : termpt ] _ [ 1 ] d^b , c_a(x_1,x_2 ; q_1 ^ 2,q_2 ^ 2 ; ) & = _ a,b,c _
q_^2^ g^a_a(y;k^2 , q_0 ^ 2 ) & p_a^b[c](z ) g^b_b(;q_1 ^ 2,k^2 ) g^c_c(;q_2 ^ 2,k^2 ) .
the @xmath69-dependence of @xmath86 } d$ ] is very mild as it emerges solely from the lower limit of the logarithmic transverse momentum integration @xmath99 .
let us analyze the lowest order interaction amplitude shown in fig .
[ fig34 ] that produces a double hard collision and involves parton splitting .
@xmath0,height=207 ] we express parton momenta @xmath100 in terms of the sudakov decomposition using the light - like vectors @xmath101 , @xmath102 along the incident hadron momenta : k_1 & = & x_1p_a + p_b + k _ , k_3 ( x_3-)p_b ; + k_2 & = & x_2p_a - p_b - k _ , k_4 ( x_4+)p_b ; + _ & = & _ 12 = -_34 ( 0 ) ; k_0 ( x_1+x_2)p_a .
here @xmath103 , @xmath104 and @xmath105 are momenta of incoming ( real ) partons , and @xmath106 and @xmath107 virtual ones .
light - cone fractions @xmath93 , i=1, .. 4 , are determined by jet kinematics ( invariant masses and rapidities of jet pair ) .
the fraction @xmath108 that measures the _ difference _ in longitudinal momenta of the two partons coming from the hadron @xmath72 , is arbitrary .
fixed values of the parton momenta @xmath109 and @xmath110 correspond to the plane wave description of the scattering process in which the longitudinal distance between the two scatterings is arbitrary .
this description does not correspond to the physical picture of the process we are interested in . in order to ensure
than the partons @xmath43 and @xmath44 originate form the _ same hadron _ of finite size , we have to introduce an integration over @xmath108 in the _ amplitude _ , in the region @xmath111 .
the feynman amplitude contains the product of two virtual propagators .
the virtualities @xmath112 and the @xmath113 that enter the denominator of the amplitude in terms of the sudakov variables read k_1 ^ 2 = x_1s -k_^2 , k_2 ^ 2 = -x_2s -k_^2 , with @xmath114 and @xmath115 the square of the two - dimensional transverse momentum vector .
a singular contribution we are looking for originates from the region @xmath116 , so that precise form of the longitudinal smearing does not matter and the integral yields n = .
the numerator of the amplitude is proportional to the _ first power of the transverse momentum _ @xmath51 . as a result , the squared amplitude ( and thus the differential cross section ) acquires the necessary factor @xmath117 that enhances the back - to - back jet production .
now we should verify that the diagram of fig .
[ fig24 ] where both incident partons split , and their offspring engage into double hard scattering , does not favor back - to - back jet kinematics .
in other words , it does not lead to a small imbalance factor @xmath117 in the differential cross section .
@xmath2,height=207 ] sudakov decomposition : @xmath118 this is a loop diagram and it contains explicit integration over the loop momentum : @xmath119 to get an enhanced contribution we have to have parton virtualities that enter the denominator of the feynman amplitude to be relatively small , of the order of @xmath120 .
this implies @xmath121 in the essential integration region . adopting this approximation
, we can simplify parton propagators and reduce the longitudinal momentum integrations to the product of two independent integrals : @xmath122 the remaining transverse momentum integration takes the form [ eq : kperpint ] due to gauge invariance the numerator of the diagram the `` vertex factor '' @xmath123 is linear in transverse momenta of the loop partons : @xmath124 .
therefore , the integral produces no more than a logarithmic enhancement factor , @xmath125 , instead of the power back - to - back singularity @xmath126 we were looking for .
so , the diagram fig .
[ fig24 ] with double parton splitting constitutes but a negligible loop correction to the usual `` hedgehog '' 4-jet kinematics typical for @xmath127 qcd processes .
the fact that this loop diagram does not produce a pole singularity in @xmath128 could have been extracted , e.g. , from numerical studies of double @xmath129-boson production in two - parton collisions @xcite and , more generally , of multi - leg parton amplitudes @xcite . the logarithmic character of this correction has been recently confirmed by the systematic study of `` box integrals '' in @xcite .
the presence of the double parton splitting contribution of fig .
[ fig24 ] is being treated in the literature as a source of potential problem of double counting ( see , e.g. , @xcite ) .
the present paper solves this problem .
we have to return now to the @xmath0 process and examine the possibility of producing an additional parton , in collinear enhanced manner , in order to lift off the born level kinematical constraint @xmath130 .
three - to - four amplitudes with extra emission from inside the splitting fork , height=226 ] consider the diagram of fig .
[ fig34 - 1]a .
the momenta of quasi - real colliding partons are @xmath131 and the radiated on - mass - shell parton carries momentum @xmath132 we have three virtual propagators subject to integration over @xmath108 : k_1 & = & x_1p_a + p_b + _ 13 , + k_2 & = & x_2p_a - ( + ) p_b + _ 24 , + & = & ( x_2 + ) p_a -p_b - _ 13 .
closing the contour around the pole @xmath133 , we obtain @xmath134 and [ eq:34-virts ] [ eq:34-virts-1 ] -^2 & = & ( x_2+)s + _ 13 ^ 2 = _ 13 ^ 2 , + [ eq:34-virts-2 ] -k_2 ^ 2 & = & x_2(+ ) s + _ 24 ^ 2 = + ^2 . taken together with the residue of the @xmath108-integration , @xmath135 , eq .
produces the universal factor present in all the amplitudes considered , ( including the diagrams with parton emission off the external lines , see fig .
[ fig34 - 2 ] below ) : [ eq : pdef ] p^-1
= x_1_24 ^ 2 + x_2_13 ^ 2 + ^2 .
we observe that both propagators are enhanced in the back - to - back kinematics .
the amplitude of fig .
[ fig34 - 1]a gives a double collinear enhanced contribution to the cross section in the region [ eq:34-regions ] [ eq:34-region1 ] _
13 ^ 2 _ 24 ^ 2 ^2 .
the inequality corresponds to the following physical picture .
an incident parton @xmath103 splits into @xmath106 and @xmath136 _ early _ , at time @xmath137 corresponding to some comparatively low perturbative scale @xmath20 . at this time
scale the parton @xmath67 collides with @xmath43 , while the parton @xmath136 keeps evolving and scatters off @xmath44 with a much larger momentum transfer @xmath21 .
evolution of the parton @xmath136 in between these two scales is the origin of probable ( logarithmically enhanced ) production of additional parton(s ) .
analogously , the diagram fig .
[ fig34 - 1]b with a parton produced off the virtual line @xmath67 contributes in the complementary kinematical region [ eq:34-region2 ] _
24 ^ 2 _ 13 ^ 2 ^2 . full perturbative analysis of the production of a parton from inside the `` splitting fork '' , fig .
[ fig34 - 1 ] , together with emission off the incoming line `` 0 '' ( to be treated below in sec .
[ sec:324 ] ) is sketched in the appendix .
now that we know the structure of the perturbative corrections to @xmath3gpd , eqs .
, we are in a position to write down the generalization of the ddt formula for ( the first contribution to ) the differential cross section of 4-jet production in nearly back - to - back kinematics .
it reads [ eq : ddt4 ] ^2 & = & \ { _ [ 2]d_a^1,2(x_1,x_2;_13 ^ 2 , _
24 ^ 2 ) _ [ 2]d_b^3,4(x_3,x_4;_13 ^ 2 , _ 24 ^ 2 ) + & & s_1(q^2,_13 ^ 2 ) s_3(q^2,_13 ^ 2 ) s_2(q^2,_24 ^ 2 ) s_4(q^2,_24 ^ 2 ) } .
here @xmath138 is the cross section of double hard parton scattering , and @xmath98 stand for sudakov form factors of four participating partons .
sum over parton species and convolution over @xmath139 as in eq . is implied .
taking derivative over the scale @xmath128 of the function depending on @xmath140 , produces the factor @xmath141 .
differentiating the sudakov form factor of a given parton describes the situation when the jet imbalance is compensated by radiation of a _ soft gluon _ off this parton .
differentiation of the parton distribution corresponds to the situation when a _ hard parton _ takes the recoil .
the differential transverse momentum imbalance distribution due to the cross - terms @xmath94}d \times { } _ { [ 1]}d$ ] contains two pieces .
the first one has the same structure as eq . : [ eq : ddt31 ] ^2 & = & \ { _ [ 1]d_a^1,2(x_1,x_2;_13 ^ 2 , _ 24 ^ 2 ) _ [ 2]d_b^3,4(x_3,x_4;_13 ^ 2 , _ 24 ^ 2 ) + & & s_1(q^2,_13 ^ 2 ) s_3(q^2,_13 ^ 2 ) s_2(q^2,_24 ^ 2 ) s_4(q^2,_24 ^ 2 ) } .
sum over parton species and convolution over @xmath139 is implied as above in eq . . as above , taking the derivatives in eq
. corresponds to fixing transverse momenta of two final state partons that compensate jet pair imbalances : @xmath143 and @xmath144 .
consider now the correlation term @xmath83}d$ ] of eq . .
if we apply the derivatives to the parton distributions @xmath145 in the _ integrand _ of eq . for the correlation term @xmath83}d$ ]
, this contribution will correspond to production of two momentum compensating quanta in the course of evolution of the system of two partons with account of _ small - distance perturbative correlation _ between them . in this case
the scale of the core parton splitting stays smaller than the two external scales @xmath146 , @xmath147 , and is being integrated over .
the correlation term contains an additional option .
namely , instead of creating intermediate state partons that keep evolving up to external scales , the perturbative splitting may produce that very parton that gets engaged in the hard scattering .
this possibility is also contained in eq . : it corresponds to the differentiation of the _ upper limit _ of the virtuality integral in eq . over the _ smaller _ of the two imbalances , @xmath148 .
one of the two parton distribution functions in the integrand then collapses to @xmath149 . taking the second derivative over the _
larger _ imbalance of the second parton distribution produces the contribution described by the diagram of fig .
[ fig34 - 1 ] that we have discussed above in section [ subsection34-add ] .
finally , there is a possibility that _ both _ partons emerging from the perturbative splitting in @xmath83}d$ ] experience hard collisions straight away , without any further evolution .
jet imbalances stemming from such an eventuality are no longer independent but , on the contrary , are strongly correlated . at the `` born level '' one
has @xmath151 with account of additional radiation , the delta - function gets replaced by the pole enhancement of the differential cross section in the _ imbalance of imbalances _ in a specific region of jet momenta when the pair imbalances are practically equal - and - opposite : : @xmath152 importantly , this contribution is also double - collinear enhanced and therefore has to be taken into full consideration .
this eventuality is not incorporated in eq . and should be taken care of separately . in this kinematical region
the parton that carries the compensating transverse momentum @xmath153 can be produced off one of the external legs , as shown in fig .
[ fig34 - 2 ] .
@xmath0 with real parton emission off the external lines , height=226 ] the corresponding contributions to the differential 4-jet production cross section can be combined into the following relatively compact expression : [ eq : ddt32 ] & = _ c p_c^1,2 ( ) s_1(q^2,^2 ) s_2(q^2,^2 ) & \ { s_c(^2,^2 ) s_3(q^2 , ^2 ) s_4(q^2 , ^2 ) _ [ 2]d_b^3,4(x_3,x_4;^2 , ^2 ) } + \ { a b } . once again , integral over @xmath12 on the r.h.s .
is implicit here . in this formula
the splitting of the parton `` c '' from the hadron `` a '' is written explicitly , represented by the splitting function @xmath154 .
the splitting occurs at the virtuality scale @xmath128 .
however , accompanying production of secondary real particles is vetoed starting from a much smaller scale , @xmath155
. this explains appearance of the sudakov form factors of five participating partons in . in the _ impulse approximation _
, there is one parton produced with the transverse momentum @xmath153 compensating the 4-jet imbalance , while transverse momenta of all other real partons are smaller : @xmath156 .
production probability of this parton integrated over its energy ( rapidity ) is embedded into the derivative of the product of three sudakov form factors and of the initial parton distributions depending on @xmath157 as the upper evolution scale . as before in eq .
, by differentiating form factors we obtain soft gluon radiation off three external lines c " ( `` 0 '' ) , `` 3 '' and `` 4 '' .
differentiation of the parton distribution functions describes real parton production due to `` hard '' splittings .
the full answer is given by the sum of eqs .
: [ eq : sumsigmas ] + + .
as we have repeatedly stressed above , double parton collisions constitute but a small correction to 4-jet production if characterized by their contribution to the total cross section .
however , `` higher twist '' four - to - four and three - to - four processes are specific in that they populate the kinematical region of relatively small pair jet imbalances , @xmath159 . being a multi - scale hard process
, back - to - back jet production cross section possesses double logarithmic parton form factors .
the sudakov suppression is the price one pays for having the differential distribution peaked , @xmath160 , which enhancement implies vetoing the bremsstrahlung radiation of gluons with transverse momenta in a broad interval ranging from @xmath161 all the way up to @xmath162 .
the formulas for the differential imbalance spectra were derived in the impulse approximation in which it is a single parton ( soft gluon ) that compensates the transverse momentum imbalance , while radiation of other partons with @xmath161 is strictly vetoed .
however , in the extreme @xmath163 limit the form factor suppression becomes so strong as to override the pole enhancement . in these conditions the impulse approximation is no longer valid , and instead of _ vetoing _ accompanying radiation one has to impose the condition that the real partons carry , in aggregate , the given transverse momentum : @xmath164
this implies exponentiating soft gluon radiation in the _ impact parameter space _ , and evaluating the inverse fourier transform to access the resulting @xmath32-spectrum @xcite .
the result is the _ flattening _ of the spectrum in the @xmath163 limit .
the precise width of this plateau depends on parton form factors as well as on parton distributions , and should be calculated numerically .
however , a rough semi - quantitative estimate of the value of the critical transverse momentum @xmath165 where the spectrum starts to flatten out can be obtained from the following simplified consideration . in the double logarithmic approximation
one can neglect effects due to parton distributions and state that the plateau develops when the product of relevant sudakov form factors just compensates the pole enhancement : @xmath166 from eqs .
we obtain the condition _
c = 1 , where @xmath167 is the sum of the `` squared color charges '' ( casimir operators ) of participating partons . substituting the one - loop expression for the running coupling , we have [ eq:4 ] = ( 1+)^-1 , with @xmath168 the leading coefficient of the qcd @xmath108-function . the estimate of the `` flattening momentum '' follows : [ eq:5 ] p_0 ^ 2 ^2 ( ) ^ , = ( 1+)^-1 . using @xmath169 for @xmath170 light quark flavors , the exponent @xmath171 in eq .
equals @xmath172 for a quark pair ( @xmath173 ) , @xmath174 for a system of a quark and a gluon ( @xmath175 ) , and @xmath176 for two gluons ( @xmath177 ) .
the value @xmath165 characterizes a typical total transverse momentum of the final state produced in a hard 2-parton collision , be it a drell yan pair , @xmath178 or a system of four hadron jets ( or , say , three jets and a large transverse momentum photon ) .
in the case of four- ( three- ) parton collisions , one has _
double _ back - to - back enhancement , accompanied by the _
four _ sudakov form factors , in place of two .
so , the pattern of the spectrum flattening in the @xmath158 limit remains roughly the same .
we conclude the discussion of the differential distribution with two remarks .
as we have seen above , in the back - to - back kinematics the 4-jet production in double parton collisions ( @xmath1 , @xmath0 ) , and due to higher order qcd effects in standard two - parton collisions ( @xmath2 ) , are comparable .
this makes the experimental studies of double hard collisions not an easy task .
to extract mpi , one should learn to reliably subtract the @xmath2 contributions .
this `` background '' should be theoretically predicted and generated at the nlo level ( for the amplitude ) in order to accommodate loop correction effects discussed above in section [ sec24 ] . on the experimental side , the tevatron experiments @xcite have set a fashion of searching for mpi by looking at _ angular correlations _ between transverse momentum imbalances in @xmath179 ( and @xmath180 , @xcite ) events .
such a strategy can well be used to signal the presence of mpi . at the same time
, it is not suited for quantitative analysis as it mixes together contributions of hard and soft physics . in order to extract and study double hard collisions , and thus to get hold of inter - parton correlations inside hadrons
, one should instead base the measurements on the _ values _ ( and relative geometry ) of transverse momentum imbalances .
integrating the differential distribution over imbalances @xmath181 , @xmath182 we obtain the total cross section of the two dijet production in the double parton collision process in the leading collinear approximation : [ stot ] = \ { 1s_4 + } . here
@xmath183 and @xmath184 are @xmath93- and @xmath5-dependent parameters that describe effective interaction areas characterizing 4- and 3-parton collisions , correspondingly .
they are given by the following expressions : s_4 ^ -1 & ( x_1,x_2,x_3,x_4;q^2 ) = \ { _ [ 2]d_a(x_1,x_2;q^2,q^2 ; ) _ [ 2]d_b(x_3,x_4;q^2,q^2;- ) & + _ [ 2]d_a(x_1,x_2;q^2,q^2 ; ) _ [ 1]d_b(x_3,x_4;q^2,q^2;- ) + _ [ 1]d_a(x_1,x_2;q^2,q^2 ; ) _ [ 2]d_b(x_3,x_4;q^2,q^2;- ) } , [ s4 ] and s_3 ^ -1(x_1,x_2,x_3,x_4;q^2 ) & = _ c p_c^1,2 ( ) ^q^2 _
i=1 ^ 4 s_i(q^2,^2 ) & _ [ 2]d^3,4_b(x_3,x_4;^2,^2 ; ) + ( ab ; 1,2 3,4 ) [ s3 ] unlike the case of @xmath1 , at the level of the differential spectrum the contribution @xmath0 due to an interplay of non - perturbative and perturbative two - parton correlations is not given by an expression with full derivatives over two relevant scales ; therefore the final expression for @xmath185 contains logarithmic integration over the hard scale @xmath128 running up to the overall hardness scale of the parton scattering @xmath5 .
in this work we aimed at generalization of the qcd factorization theorem to multiple parton interactions in hadron hadron collisions . such generalization is necessary for carrying out systematic studies both theoretical and experimental of the parton correlations inside hadrons , and allows one to incorporate in a model independent way higher order pqcd effects . in the small-@xmath8 domain ,
generalization of the mpi analysis presented here may be looked upon as realization of the gribov pomeron calculus , with exact account of energy - momentum conservation .
we considered four jet production and have shown that double hard parton collisions play a significant role in the _ back - to - back kinematics_. this kinematics one selects by demanding the transverse momentum imbalances of two pairs of jets , @xmath186 and @xmath187 , to be much smaller than the overall hard scale of the scattering process : @xmath188 .
we were able to analyze and calculate corresponding differential spectra and total cross sections with account of all collinear enhanced pqcd radiative effects giving rise to sudakov form factors and to scaling violations in one - parton and _ generalized double parton distributions _ , @xmath189gpds , that have been introduced in @xcite .
our main results are eqs . for the differential distribution and eq . for the total cross section of 4-jet production in the back - to - back kinematics .
they are derived in the leading logarithmic ( collinear ) approximation .
the formalism developed here provides the framework , and a starting point , for systematic calculation of nll and nlo corrections that may be needed for more careful account of the mpi effects at lhc . in particular , in the present study we did not hunt for single logarithmic corrections due to _ large - angle soft gluon radiation_. these ( formally subleading but potentially important ) logarithmic effects depend on details of the color transfers in hard parton scattering processes
. they should be treated together with a delicate question of algorithmic determination of transverse momenta of the hadron jets .
the corresponding study @xcite exists for the single hard scattering case , @xmath190 , and provides a solid base for analogous treatment of large - angle gluon radiation in double parton collision processes .
we demonstrated that , in addition to the qcd - improved parton model picture of 4-parton collisions , pqcd also reserves an important place to 3-parton collisions producing four jets .
such @xmath0 transitions are the novel hard double scattering processes that are systematically treated in this work for the first time .
@xmath0 parton process appears when a parton from one colliding hadron splits into two partons that enter into double hard interaction with two partons from the ( non - perturbative ) wave function of the second hadron .
perturbative parton splitting occurs at relatively small impact parameters , while the relative impact parameter distance between two hard vertices in a @xmath1 collision can be as large as the transverse size of the hadron .
parametrically , @xmath0 and @xmath1 subprocesses are comparable ( they are of the same order in @xmath4 and have similar structure of the back - to - back enhancement ) . at the same time
, the difference in geometry of the interaction processes results in a _ numerical enhancement _ of @xmath0 with respect to @xmath1 subprocess ( by a numerical factor of 45 ) . as a result
, @xmath0 transitions should contribute significantly to 4-jet production , even if they happen to constitute formally a small correction to a @xmath3gpd .
it is important to stress that the processes where one parton splits perturbatively into two in _ both colliding hadrons _ do not contribute to the 2-dijet production , since the corresponding differential distribution lacks the necessary back - to - back double pole enhancement @xmath191 .
in fact , such configurations represent a one - loop correction to the @xmath2 jet production mechanism and as such belong to the class of the standard single hard interaction processes rather than to mpi .
this result of @xcite has been later confirmed in refs .
the role of @xmath0 parton subprocesses is twofold .
when the perturbative splitting occurs at momentum scales smaller than the pair jet imbalances @xmath128 ( that is , `` deep inside the dglap ladder '' ) , such contributions exhibit the same two - pole enhanced structure as the conventional @xmath1 transitions already discussed in @xcite : @xmath192 the corresponding differential distribution is a direct generalization of the `` ddt formula '' @xcite containing derivatives over transverse momentum imbalances , with eq . describing @xmath1 collisions and eq .
a crosstalk between large- and small - distance two - parton correlators . at the same time , the parton splitting may occur just at the scale @xmath128 and produce two partons that , without any further evolution , interact with two partons from the wave function of the second hadron .
such eventuality also has a necessary back - to - back enhancement but of a different structure : @xmath193 such contributions are absent in the parton model .
they can not be represented as the product of two @xmath3gpds and must be taken into account separately , see eq . .
a detailed numerical investigation of these contributions will be presented elsewhere @xcite .
formulas derived in this work can be used to address various aspects of the lhc physics .
our results are directly applicable to the case of double drell yan processes at lhc .
in particular they can be applied to the recently observed production of heavy quarkonia at the lhc @xcite .
the formalism developed in the present paper should help to improve the accuracy of the prediction of qcd backgrounds in searches for new physics .
it also can be used for further studies of the nucleon structure , in particular of the short - range non - perturbative inter - parton correlations in the nucleon and the nuclei , for the study of the higher twist contributions that are unavailable in the conventional dis processes , etc .
in addition , it would be interesting to extend the formalism developed here to investigate qcd medium effects that manifest themselves in @xmath194 collisions in recent experiments at cern . on the experimental side , observation of the parton splitting processes discussed in this paper
will give the first direct evidence of an interplay between large- and short - distance qcd correlations in hard processes . to achieve this goal one has to look for enhancement of 4-jet production in the kinematical region of jet transverse momenta @xmath195
the authors are grateful to the galileo galilei institute for theoretical physics , florence , for an opportunity to get together and fine tune this work during the _ qcd after the start of lhc _ workshop in september
october 2011 .
here we present the analysis of real parton production in the course of @xmath196 parton splitting , @xmath197 , determining the correlated 4-jet production .
we consider three contributing diagrams : that of fig . [ fig34 - 2](a ) for emission off the initial parton `` 0 '' , and those of fig .
[ fig34 - 1 ] for emission off the offspring partons `` 1 '' and `` 2 '' .
it is straightforward to write down exact expressions for the amplitude using the economic technique of quasi - parton states developed by bukhvostov , frolov , lipatov and kuraev ( bflk ; not to be confused with bfkl ) and described in @xcite . in the bflk technique the quark and ( axial gauge ) gluon green functions are replaced by propagators of two physical @xmath198 helicity states . in general , in this framework additional terms looking like contact four - parton interaction vertices arise .
they , however , do not produce collinear enhanced contributions .
effective three - parton interaction vertices are proportional to the linear combination of the transverse momenta of participating partons weighted with the longitudinal light - cone momentum fractions : k_^= _ i k_j^- _ j k_i^. in the matrix element the vector index is contracted with the gluon polarization vector ( for details see @xcite ) . reproducing the structure of .
emissions from 1 and 2 are coherent , and due to conservation of the color current , @xmath206 , the two expressions coincide . as a result , applies in the entire kinematical region @xmath207 , irrespectively to the value of @xmath203 .
the contributions are related with internal evolution of the @xmath3gpd of the hadron @xmath74 and are contained in eq . .
specifically , in the kinematical region where the imbalances of two jet pairs are significantly different , these contributions correspond to the case when the differentiation over the smallest of the two imbalances applies to the @xmath3gpd and acts upon the integration limit of the correlation @xmath86}d$ ] term in .
b. blok , yu . dokshitzer , l. frankfurt and m. strikman , phys .
d * 83 * , 071501 ( 2011 ) [ arxiv:1009.2714 [ hep - ph ] ]
. n. paver and d. treleani , z. phys .
c * 28 * 187 ( 1985 ) .
m. mekhfi , phys .
rev . d*32 * , 2371 ( 1985 ) . a. del fabbro and d. treleani , phys .
d * 61 * , 077502 ( 2000 ) [ arxiv : hep - ph/9911358 ] ; phys .
d * 63 * , 057901 ( 2001 ) [ arxiv : hep - ph/0005273 ] ; a. accardi and d. treleani , phys .
d * 63 * , 116002 ( 2001 ) [ arxiv : hep - ph/0009234 ] .
proceedings of the 1st international workshop on multiple partonic interactions at the lhc verlug deutshes electronen synchrotron , 2010 .
p. bartalini and l. fano , arxiv:1103.6201 [ hep - ex ] . for
a recent summary see t. sjostrand and p.z .
skands , eur .
j. c * 39 * , 129 ( 2005 ) [ arxiv : hep - ph/0408302 ] . | we study production of two pairs of jets in hadron
hadron collisions in view of extracting contribution of _ double hard interactions _ of three and four partons ( @xmath0 , @xmath1 ) . such interactions , in spite of being power suppressed at the level of the total cross section , become comparable with the standard hard collisions of two partons , @xmath2 , in the _ back - to - back kinematics _ when the transverse momentum imbalances of two pairing jets are relatively small .
we express differential and total cross sections for two - dijet production in double parton collisions through the generalized two - parton distributions , @xmath3gpds @xcite , that contain large - distance two - parton correlations of non - perturbative origin as well as small - distance correlations due to parton evolution .
we find that these large- and small - distance correlations participate in different manner in 4-jet production , and treat them in the leading logarithmic approximation of pqcd that resums collinear logarithms in all orders .
a special emphasis is given to @xmath0 double hard interaction processes that occur as an interplay between large- and short - distance parton correlations and were not taken into consideration by approaches inspired by the parton model picture .
we demonstrate that the @xmath0 mechanism , being of the same order in @xmath4 as the @xmath1 process , turns out to be _ geometrically enhanced _ compared to the latter and should contribute significantly to 4-jet production .
the framework developed here takes into systematic consideration perturbative @xmath5 evolution of @xmath3gpds .
it can be used as a basis for future analysis of nlo corrections to multi - parton interactions ( mpi ) at lhc and tevatron colliders , in particular for improving evaluation of qcd backgrounds to new physics searches . |
in the past idealized models for turbulent fluctuations which can be found in the solar wind plasma or in the interstellar medium have been proposed ( e.g. matthaeus et al .
we are concerned with statistically axisymmetric models of magnetostatic fluctuations @xmath0 that are transverse to a uniform mean magnetic field @xmath1 . if solar wind turbulence is considered , the mean field might be identified with the magnetic field of the sun .
the total magnetic field is a superposition of this mean field and the fluctuations @xmath2 .
whereas we usually approximate the mean field by a constant field aligned parallel to the @xmath3axis ( @xmath4 ) , the turbulent contribution has to be replaced by turbulence models .
some prominant examples are slab , 2d , and two component models that include both slab and 2d contributions ( e.g. matthaeus et al .
1990 ) .
there are recent spacecraft measurements of magnetic correlations in the solar wind ( see e.g. matthaeus et al .
2005 , dasso et al . 2007 ) .
such measurements are very interesting and important since they allow an improved understanding of turbulence .
for instance , characteristic length scales of turbulence such as the correlation length , the bendover scale , and the dissipation scale can be obtained from such observations .
also the investigation of spectral anisotropy by using data from different space craft missions such as wind and ace is possible .
these properties of solar wind turbulence are very important for several investigations ( heating and damping of the solar wind plasma , transport of charged cosmic rays ) .
a further important turbulence property is the turbulence dynamics ( the time dependence of the stochastic magnetic fields ) . in principle ,
data sets from wind and ace can also be used to compute dynamical correlation functions to explore the turbulence dynamics . in a recent article ( shalchi 2008 )
magnetic correlation functions were computed analytically .
such analytical forms of magnetic correlations complement data analysis results such as matthaeus et al .
( 2005 ) and dasso et al .
since we expect that future data analysis work will also allow the investigation of temporal correlation functions , we explore theoretically ( numerically and analytically ) the forms of these eulerian correlations .
these results can be compared with data analysis results as soon as they are available .
the organization of the paper is as follows : in section 2 we define and discuss the basic parameters which are useful for describing turbulence .
furthermore , we explain the slab , the 2d , and the slab/2d compositel model . in section 3
we review different models for the turbulence dynamics . in section 4
we compute eulerian correlation functions numerically and analytically . in section 5
the results of this article are summarized .
the key function in turbulence theory is the two - point - two - time correlation tensor . for homogenous turbulence its components
are r_lm ( , t ) = .
[ s1e1 ] the brackets @xmath5 used here denote the ensemble average .
it is convenient to introduce the correlation tensor in the @xmath6space . by using the fourier representation b_l ( , t ) = d^3 k b_l ( , t ) e^i [ s1e2 ]
we find r_lm ( , t ) = d^3 k d^3 k^ e^i .
[ s1e3 ] for homogenous turbulence we have = p_lm ( , t ) ( - ^ ) [ s1e4 ] with the correlation tensor in the @xmath6space @xmath7 . by assuming the same temporal behaviour of all tensor components , we have p_lm ( , t ) = p_lm ( ) ( , t ) with the dynamical correlation funtion @xmath8 .
( [ s1e3 ] ) becomes than r_lm ( , t ) = d^3 k p_lm ( ) ( , t ) e^i ( ) [ s1e5 ] with the magnetostatic tensor @xmath9 . in this paragraph
we discuss the static tensor @xmath10 defined in eq .
( [ s1e5 ] ) .
matthaeus & smith ( 1981 ) have investigated axisymmetric turbulence and derived a general form of @xmath10 for this special case . in our case
the symmetry - axis has to be identified with the axis of the uniform mean magnetic field @xmath4 . for most applications ( e.g. plasma containment devices , interplanetary medium )
the condition of axisymmetry should be well satisfied .
furthermore , we neglect magnetic helicity and we assume that the parallel component of the turbulent fields is zero or negligible small ( @xmath11 ) . in this case
the correlation tensor has the form p_lm ( ) = a(k_,k _ ) , l , m = x , y [ s1e20 ] and @xmath12 .
the function @xmath13 is controlled by two turbulence properties : the turbulence geometry and the turbulence wave spectrum .
the geometry describes how @xmath13 depends on the direction of the wave vector @xmath14 with respect to the mean field .
there are at least three established models for the turbulence geometry : 1 .
the slab model : here we assume the form a^slab
( k_,k _ ) = g^slab ( k _ ) .
[ s1e21 ] in this model the wave vectors are aligned parallel to the mean field ( @xmath15 ) .
2 . the 2d model : here we replace @xmath13 by a^2d ( k_,k _ ) = g^2d ( k _ ) .
[ s1e22 ] in this model the wave vectors are aligned perpendicular to the mean field ( @xmath16 ) and are therefore in a two - dimensional ( 2d ) plane .
the slab/2d composite ( or two - component ) model : in reality the turbulent fields can depend on all three coordinates of space .
a quasi - three - dimensional model is the so - called slab/2d composite model , where we assume a superposition of slab and 2d fluctuations : @xmath17 . because of @xmath18 , the correlation tensor has the form p_lm^comp ( ) = p_lm^slab ( ) + p_lm^2d ( ) .
[ s4e2 ] in the composite model the total strength of the fluctuations is @xmath19 .
the composite model is often used to model solar wind turbulence .
it was demonstrated by several authors ( e.g. bieber et al .
1994 , 1996 ) that @xmath20 slab / @xmath21 2d should be realistic in the solar wind at 1 au heliocentric distance .
the wave spectrum describes the wave number dependence of @xmath22 . in the slab model
the spectrum is described by the function @xmath23 and in the 2d model by @xmath24 .
as demonstrated in shalchi ( 2008 ) , the combined correlation functions ( defined as @xmath25 ) for pure slab turbulence is given by r_^slab ( z ) = 8 _ 0^ d k _ g^slab ( k _ ) ( k _ z ) [ corrslab ] and the correlation function for pure 2d is r_^2d ( ) = 2 _ 0^ d k _ g^2d ( k _ ) j_0 ( k _ ) .
[ s3e12 ] here @xmath26 is the distance parallel with respect to the mean magnetic field and @xmath27 denotes the distance in the perpendicular direction . to evaluate these formulas we have to specify the two wave spectra @xmath23 and @xmath24 . in a cosmic ray propagation study , bieber et al .
( 1994 ) proposed spectra of the form g^slab ( k _ ) & = & l_slab b_slab^2 ( 1 + k_^2 l_slab^2)^- + g^2d ( k _ ) & = & l_2d b_2d^2 ( 1 + k_^2 l_2d^2)^- with the inertial range spectral index @xmath28 , the two bendover length scales @xmath29 and @xmath30 , and the strength of the slab and the 2d fluctuations @xmath31 and @xmath32 . by requiring normalization of the spectra b^2 = b_x^2 + b_y^2 + b_z^2 = d^3
k [ s2e4 ] we find c ( ) = .
[ s2e5 ] by combining these spectra with eqs .
( [ corrslab ] ) and ( [ s3e12 ] ) the slab correlation function r_^slab ( z ) = 4 c ( ) b_slab^2 l_slab _ 0^ d k _ ( 1+k_^2 l_slab^2)^- ( k _ z ) [ s2e6 ] as well as the 2d correlation function r_^2d ( ) = 4 c ( ) b_2d^2 l_2d _ 0^ d k _
( 1+k_^2 l_2d^2)^- j_0 ( k _ ) [ s3e13 ] can be calculated . in eq .
( [ s3e13 ] ) we have used the bessel function @xmath33 . in shalchi ( 2008 ) such calculations valid for magnetostatic turbulence are presented . for dynamical turbulence the slab and the 2d correlation functions from eqs .
( [ corrslab ] ) and ( [ s3e12 ] ) become r_^slab ( z ) & = & 8 _ 0^ d k _ g^slab ( k _ ) ( k _ z ) ^slab ( k_,t ) + r_^2d ( ) & = & 2 _ 0^ d k _ g^2d ( k _ ) j_0 ( k _ ) ^2d ( k_,t ) .
[ generalcorr ] for the model spectrum defined in the previous paragraph these formulas become r_^slab ( z , t ) & = & 4 c ( ) b_slab^2 l_slab _ 0^ d k _
( 1+k_^2 l_slab^2)^- ( k _ z ) ^slab ( k_,t ) + r_^2d ( , t ) & = & 4 c ( ) b_2d^2 l_2d _ 0^ d k _
( 1+k_^2 l_2d^2)^- j_0 ( k _ ) ^2d ( k_,t ) .
[ corrdyn2 ] to evaluate these equations we have to specify the dynamical correlation functions @xmath34 and @xmath35 which is done in the next section .
in the following , we discuss several models for the dynamical correlation function @xmath8 . in table
[ dyntab ] , different models for the dynamical correlation function are summarized and compared with each other .
.__different models for the dynamical correlation function @xmath36 . here
, @xmath37 is the alfvn speed and @xmath38 is a parameter that allows to adjust the strength of dynamical effects .
the parameters @xmath39 and @xmath40 of the nadt model are defined in eq .
( [ c2s6e3 ] ) . _ _ [ cols="<,<,<",options="header " , ] by comparing the results of this article with spacecraft measurements , we can find out whether modern models like the nadt model is realistic or not .
this would be very useful for testing our understanding of turbulence .
some results of this article , such as eqs .
( [ generalcorr ] ) are quite general and can easily be applied for other turbulence models ( e.g. other wave spectra ) .
_ this research was supported by deutsche forschungsgemeinschaft ( dfg ) under the emmy - noether program ( grant sh 93/3 - 1 ) . as a member of the _ junges kolleg _
a. shalchi also aknowledges support by the nordrhein - westflische akademie der wissenschaften . _
by using abramowitz & stegun ( 1974 ) and gradshteyn & ryzhik ( 2000 ) we can compute analytically the different eulerian correlation functions defined in eqs .
( [ c2s5e9 ] ) and ( [ c2s5e10 ] ) . for the plasma wave model
the results are given in the main part of the paper .
for the damping model of dynamical turbulence we can use _ 0^ d x ( 1+x^2)^- e^-a_i x & = & + & & .
here we used bessel functions @xmath41 and the struve function @xmath42 . by making use of this result
we find for the eulerian correlation function e_^i , dt ( t ) & = & 4 c ( ) b_i^2 + & & with @xmath43 and a_slab & = & + a_2d & = & for the damping model of dynamical turbulence . for the random sweeping model we can employ _
0^ d x ( 1+x^2)^- e^-(a_i x)^2 = u ( , - , a_i^2 ^2 ) with the confluent hypergeometric function @xmath44 . by employing this result
we find for the eulerian correlation function e_^i , rs ( t ) = 2 c ( ) b_i^2 u ( , - , a_i^2 ^2 ) for the random sweeping model . for the nadt model
we only have to explore the 2d fluctuations ( the slab result is trivial and discussed in the main part of the this paper )
( [ c2s5e10 ] ) can be rewritten as e_^2d = 4 c ( ) b_2d^2 the first integral in trivial , the second one can be expressed by an exponential integral function @xmath45 : e_^2d = 4 c ( ) b_2d^2 .
[ ananadt2d ] this is the final result for the eulerian correlation function of the 2d fluctuations . to evaluate this expression for late times ( @xmath46 )
we can approximate the exponential integral function by using e_3 - 1/2 ( 1 ) _
1^ e^- x = .
[ approxen ] here we assumed that the main contribution to the integral comes from the smallest values of @xmath47 , namely @xmath48 . by combining eq .
( [ approxen ] ) with eq .
( [ ananadt2d ] ) we find approximatelly e_^2d 4 c ( ) b_2d^2 e^- corresponding to an exponential behavior of the eulerian correlation function .
for the correlation time scale we find @xmath49 .
a further discussion of these results can be found in the main part of the text . | current spacecraft missions such as wind and ace can be used to determine magnetic correlation functions in the solar wind .
data sets from these missions can , in principle , also be used to compute so - called eulerian correlation functions .
these temporal correlations are essential for understanding the dynamics of solar wind turbulence . in the current article
we calculate these dynamical correlations by using well - established methods .
these results are very useful for a comparison with eulerian correlations obtained from space craft missions . |
the overwhelming majority of the more than 100 gravitationally lensed quasars known so far consist of two quasar images or , due to the ellipticity of the gravitational potential in the lensing mass , four images .
these are galaxy - scale lensed systems , with image separations of @xmath1 , produced by a foreground galaxy acting as a lens .
rarely , small - separation three - image lensed quasars have also been observed , ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , but these contain a central quasar image , close to the center of a lensing galaxy with a central density profile that is shallower than usual . ignoring large - separation lensed quasars produced by galaxy clusters ( e.g. , *
* ) , another way of producing other than two or four quasar images is found in systems with a group of two ( e.g. , * ? ? ? * ) or three lensing galaxies , with a single confirmed example being the six - image b1359 + 154 @xcite . such systems also offer the chance to study in detail the mass distribution in galaxy pairs from the strong lens modeling of the multiple quasar images , as well as the shape of extended arcs from the quasar host galaxy .
an example of a unique lensed quasar in a two - lens system was mg 2016 + 112 ( e.g. , * ? ? ? * ; * ? ? ?
* ; * ? ? ?
* ) , which contains an observed arc that has no optically detected central quasar component , and requires the consideration of both lensing galaxies to explain the observed configuration .
here we show that the recently discovered system sdss j1405 + 0959 @xcite is the second example of such a system , and is also likely to be a three - image lensed quasar system .
this result is a product of an adaptive optics ( ao ) imaging campaign of gravitationally lensed quasars ( * ? ? ?
* rusu et al . , in prep . ) from the sdss quasar lens search ( sqls ; * ? ? ?
* ; * ? ? ?
in section [ sect : previous ] we summarize the previous observations of this object . in section [
sect : aooverall ] we present our adaptive optics observations , data reduction and morphological modeling technique . section [ section : photoz ] addresses the new multi - band color information , and section [ section : lens ] presents our gravitational lens mass models .
we summarize our conclusions in section [ sect : concl ] . throughout this work
, the concordance cosmology with @xmath2 , @xmath3 and @xmath4 is assumed .
sdss j1405 + 0959 was identified in the sdss data release 7 @xcite as a two - component system , one of which is a quasar at @xmath5 .
the system was reported by @xcite as a new gravitationally lensed quasar discovered in the major ukidss sdss cosmic lens survey ( muscles ) . this was based on the spectral similarity of the two components , as well as the decrease in separation with increasing wavelength , suggesting the presence of a lensing galaxy .
based on a tentative detection of the @xmath6 break , as well as possible emission lines , the galaxy redshift was estimated to be @xmath7 .
the system was identified independently as a lensed quasar candidate from the sqls ( see * ? ? ?
* ) , with follow - up spectroscopy and @xmath8 , @xmath9 , @xmath10 band imaging confirming the gravitational lens nature @xcite .
an additional close - by component , also visible in @xcite , in the infrared , was identified as a potential secondary lens ( see figure [ fig:1405closeup ] ) .
adaptive optics observations were performed with the infrared camera and spectrograph ( ircs ; * ? ? ?
* ) at the subaru telescope , in conjunction with the laser guide star adaptive optics system ( lgs@xmath11ao188 ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
ircs was used in the 00528 pixel scale mode , providing a field of view of @xmath12 .
ao188 employs a curvature wavefront sensor and a 188-element bimorph deformable mirror .
the lgs system uses an artificial sodium laser guide star for high - order wavefront sensing @xcite .
observations were initially performed in the @xmath13 band , followed a year later by @xmath14 and @xmath15 band , in order to obtain photometric redshifts for the newly detected objects .
a summary of the exposure times , observation dates , airmass and photometric standard stars is provided in table [ tab : followup - data ] .
[ cols="^,^,<,^,>",options="header " , ] + all four objects are modeled with sersic profiles . here
@xmath16 is the ellipticity , @xmath17 is the effective radius measured along the major axis , and @xmath18 is the sersic index . the position angle is positive from north towards east .
the morphology of g2 is affected little by whether gx is point - like / extended , the results being consistent at @xmath19 , whereas the other objects are unaffected by this .
[ tab:1405morphology ] the technique summarized above assumes that the adaptive optics psf does not vary spatially across the components of the system , which is a reasonable assumption given the small maximum separation @xmath20 of these components .
the use of the two concentric moffat profiles as an analytical psf approximation , instead of , for example , a gaussian representing a diffraction - limited core and a moffat @xcite for the wing was found to provide a better approximation for the stars observed at low strehl ratios in the current adaptive optics observation campaign ( rusu et al .
, in prep . ) .
additional detail about the technique above is presented in rusu et al .
, in prep .
in order to account for systematics introduced by the technique above , the error bars in tables [ tab:1405astrometry ] , [ tab:1405photometry ] and [ tab:1405morphology ] show the largest error obtained from several methods : individual fitting statistical error resulting from the galfit modeling , scatter from using two different hybrid psf core radii ( 10 pixels and 18 pixels ) , and scatter ( where applicable ) from gx and gy being fitted as point source / sersic profile and one / three components , respectively . finally ,
the non - simultaneously observed psf star in @xmath15 band was used to simulate the observed system 100 times , where each simulation differs in terms of the added noise , taken from a portion of the observed image that is devoid of sources .
each simulated image was then modeled in a way similar to the technique described above .
the errors resulting from these simulations are the standard deviations in the 100 values of each best - fitted parameter around the known true value .
these therefore include both systematic and statistical errors , and are typically larger than the errors obtained from the other methods .
an implicit assumption in this method is that the observed psf star can be used as a representation of the actual psfs in the @xmath14 , @xmath15 , and @xmath13 bands , albeit solely for the purpose of estimating realistic error bars .
this assumption is partly justified by the similarity of the analytical parameters describing the four psfs in table [ tab : analpsf ] , in particular _ fwhm1 _ , _ e1 _ , _
fwhm2 _ , the flux ratio and the strehl ratio . as overall results
, we have determined that the morphology of g1 and g2 is consistent with extended objects ( galaxies ) . however , due to the low s / n and close proximity to g2 , it can not be reliably distinguished whether gx is extended or point - like . in case it is fit by a sersic profile
, it has a small effective radius and an unreliably determined sersic index . on the other hand ,
gy consists of three features .
while the central feature is extended , the morphology of the other two can not be constrained , and for the purpose of extracting the astrometry , they have been fitted as point sources .
using the new high resolution observations , photometric redshifts were computed and a color - color plot was drawn for the objects seen in the field of view that is shown in figure [ fig:1405fov ] .
the magnitudes of the isolated galaxies were measured with aperture photometry , and those of g1 , g2 and gx were obtained from profile fitting . as gy seems to consist of 3 objects , which all appear to be red in figure [ fig:1405closeup ] , and are difficult to model analytically due to the low s / n , aperture photometry was performed assuming they all have the same color . for this , all other objects were removed , using the best - fit model obtained with the hybrid psf , and the residuals were covered with nearby blank sky .
three different blank sky covers were considered , in order to account for possible systematics .
an aperture of 30 pixel radius was chosen for the aperture photometry , centered on gyb .
the resulting error bars in all bands are smaller than 0.05 mag . according to the color - color plot ( figure [ fig:1405color ] ) ,
the galaxies seen in figure [ fig:1405fov ] show various colors , indicative of different redshifts rather than physical association into a group of galaxies .
we estimated photometric redshifts from the @xmath21 bands using the template - fitting methods implemented in the publicly available software hyperz @xcite .
as templates , we employed the observed mean spectral energy distributions of galaxies from @xcite .
in particular , g3 , g4 and g6 have a redshift probability ( g3 is shown in figure [ fig:1405photoz ] ) consistent with @xmath22 , which is the estimated redshift of g3 in the sdss . on the other hand , g1 , g2 and g8 have similar colors and redshifts in agreement with @xmath23 , the spectroscopically estimated redshift of g1 .
gx has different estimated colors in the extended / point - source cases , and assuming that it is a galaxy , it would be located at @xmath24 .
more interestingly , gy appears to have a large redshift , consistent with that of the quasar .
in this section , we test various mass models by exploring different scenarios given the complexity of this system
. the most important constraint for strong lensing models is astrometrical .
table [ tab:1405astrometry ] provides high - precision astrometry , as a result of the multi - band adaptive optics observations .
we also used as constraint the flux ratio @xmath25 , which is very consistent between bands , and also with the optical flux ratio in @xcite .
an uncertainty of 10% was set on the observed flux ratio constraint used for the gravitational lens models @xcite . even in the case of gx
being the third quasar image , since its flux is likely to be affected by chromatic processes , we do not use it as a constraint in our model .
the singular isothermal profile has extensively been shown to be a good approximation for the mass distribution in lensing galaxies ( e.g. , * ? ? ?
* ; * ? ? ?
* and references within ) .
we therefore considered a singular isothermal sphere ( sis ) , or the more general power law profile @xmath26 , as our mass models .
we performed all gravitational lens modeling with _ glafic _ @xcite . here , we discuss two mass models @xmath27 assuming that the system consists of only 2 lensed quasar images a and b , and @xmath28 assuming that the system consists of 3 lensed images , a , b , and gx . the latter case is suggested by the insecure morphology of gx ( section [ sect : psf ] ) , as well as the subsequent analysis based on its colors , which we will show in section [ sect:3imag ] .
first , we considered the case when there are only two quasar images .
g2 must be taken into account as a lens , due to its small distance to a and b , which is similar to the separation between them , and its similar luminosity to g1 .
given the agreement between the photometric redshift of g1 and g2 ( figure [ fig:1405photoz ] ) , we assumed that they are both located at the lens redshift @xmath29 , and that the ratio of their velocity dispersions is constrained from their relative luminosities , through the faber - jackson relation @xcite .
we used the 2sis+@xmath30 model , where the external shear @xmath30 was considered at the location of g1 .
we ignored the gravitational lens effect of gx which , assuming it is a galaxy , appears to have different colors and therefore redshift compared to the main deflectors , and is located very close to g2 while being much fainter , making it a comparatively weaker lens .
the parameters of this best - fit model , as well as those described below , are given in table 7 .
next , the case when gx is an additional quasar image was considered . here
, the observed position of image gx is an additional observational constraint .
no flux ratio constraints were used , but the velocity dispersions of both g1 and g2 were fitted as free parameters . a perfect fit ( d.o.f .
@xmath31 ) was obtained , with the velocity dispersion ratio g2/g1 @xmath32 , very close to the value obtained from the faber - jackson relation ( @xmath33 ) .
this result shows that it is indeed possible for this two - galaxy system to produce three quasar images at the observed positions . if an attempt is made to model the observed flux ratio @xmath25 as well , @xmath34/d.o.f .
@xmath35 is obtained . while this is not a very good fit ,
the model might be oversimplified . allowing for a third order perturbation @xcite a perfect fit
is obtained ( model 4 in table 7 ) , where the orientation of the new component is fixed to correspond with that of the shear , in order to avoid a model with negative number of degrees of freedom .
the use of a third order perturbation may be justified by the presence of the crowded environment in which the system is embedded ( figure [ fig:1405fov ] ) . as a last model ,
a power law ( g1 ) + sis ( g2 ) + @xmath30 profile was also considered , with @xmath34/d.o.f.@xmath36 ( model 8 in table 7 ) .
the reason why this is not a perfect fit is that the source is very close to the caustic , which is a curve in the source plane determined from the observed image , lens configuration and lens profile , whose crossing changes the observed image multiplicity ( figure [ fig:1405mass ] ) .
this can be avoided ( i.e. the relative position of the source and caustic can be modified ) if ellipticity is also introduced , in the form of the singular isothermal ellipsoid ( sie ) , at the expense of having negative degrees of freedom : a power law + @xmath30 ( g1 ) + sie ( g2 ) model produces @xmath34/d.o.f.@xmath37 ( model 9 in table 7 ) . for all mass models described here ,
elliptical counterparts were also considered ( table 7 ) , where the orientation of the major axis was fixed to the value obtained from morphological fitting ( table [ tab:1405morphology ] ) , and the same was done for the ellipticity , unless this can not reproduce the number of observed images .
this approach uses the known correlation between the orientations of mass and light profiles in elliptical lens galaxies ( e.g. * ? ? ?
* ; * ? ? ?
* ) , as well as their mass and light ellipticity ( @xcite @xcite ; @xcite @xcite , but see also @xcite @xcite ; @xcite @xcite ) .
representative lensing models obtained in this section are plotted in figure [ fig:1405mass ] . .
[ fig:1405color],width=340 ] .
the vertical lines mark the spectroscopic redshift of the source quasar and of the main lensing galaxy g1 , respectively .
the vertical axis represents the probability of exceeding the @xmath34 value obtained during the fit by chance , as a function of the number of degrees of freedom (= 2 ) .
[ fig:1405photoz],width=317 ] as we have shown above , gx is consistent with the position of a third lensed quasar image .
morphological modeling in the @xmath15 and @xmath13 bands can not distinguish conclusively between a point source and an extended morphology .
furthermore , g2 has an estimated einstein radius of @xmath38 ( table 7 ) , assuming it is located at the redshift of g1 , and the separation between g2 and g3 is only @xmath39 .
since the colors of gx ( figure [ fig:1405photoz ] ) would put it at @xmath24 , in the background of g2 , it would be strongly lensed .
in fact , four or five images of gx would be produced , some of them brighter and at a larger separation from g2 .
such case is clearly inconsistent with the data . in view of the above
, we discuss the possibility of gx being a quasar image , based on its colors .
although the colors are not consistent with those of images a and b , chromatic differences between quasar images are common in the literature ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
such chromatic effects can be caused by dust extinction and microlensing in the lensing galaxy , as well as the intrinsic , time - delayed variability of the source quasar .
it is clear that , assuming gx has a point - like morphology , according to figure [ fig:1405color ] , it must be reddened in comparison to a and b. here we test whether this can be explained by dust extinction .
we assume that neither image a nor b is affected by extinction , since their colors are very similar .
the color difference between a and gx ( hereafter renamed image c , when referring to its flux / colors ) , assumed to be due to extinction , can be used to estimate the extinction , using the equation @xmath40 here , the left hand side is the observed magnitude difference of the two images , and @xmath41 refers each of the three observed bands , @xmath21 .
@xmath42 denotes the model - predicted image flux , and @xmath43 is the extinction in the lensing galaxy rest frame , therefore @xmath44 . to estimate @xmath45
, we use the lensing models from section [ section : lens3 ] , which predict @xmath46 ( ignoring those with large mass ellipticity ) .
the equation can therefore be solved for @xmath47 , obtaining a value of @xmath48 ( @xmath14 ) , @xmath49 ( @xmath15 ) and @xmath50 ( @xmath13 ) .
using an extinction curve from @xcite , and an @xmath51 extinction parameter with the galactic value of @xmath52 ( in agreement with the value found from the sample of lasdttir ) , a visual extinction @xmath53 ( @xmath14 ) , @xmath54 ( @xmath15 ) and @xmath55 ( @xmath13 ) is obtained , which translates to @xmath56/@xmath57 . although rare , such large visual extinction and color excess have been observed in lensing galaxies before ( e.g. , * ? ? ?
* ; * ? ? ?
we therefore conclude that color change due to extinction is plausible .
a problem with this interpretation however is why image b , located similarly close to the larger g1 , also a similar elliptical galaxy , is not similarly affected by extinction .
an alternative explanation is provided by microlensing due to stars in g2 , which can also cause large color differences ( e.g. , * ? ? ?
the small separation between g2 and gx ( @xmath58 effective radius ) makes both extinction and microlensing plausible .
if gx is the third quasar image , it has a negative parity , and is therefore more sensitive to microlensing effects ( e.g. , * ? ? ?
microlensing is also expected to produce a flux change that is decreasing with wavelength , as observed . since
neither a spectrum of gx ( which would also convincingly show whether it is the third quasar image ) , nor multi - epoch observations in the same band are available , the relative effects of extinction and microlensing can not be distinguished using the current data .
on the other hand , intrinsic variability is not expected to play a significant part , since it generally produces much smaller color differences than observed here ( e.g. , * ? ? ?
* ) , and in addition the similar time delays expected between images b and c ( table 7 ) would mean that differences should also be seen between a and b.
lastly , we also investigated whether gx is a chance superposition of a nearby star . such a case
is however excluded due to the observed colors , which are inconsistent with any stars observed in the infrared @xcite . from figure [ fig:1405mass ] , it is immediately apparent that the source is located very close to a caustic ( red curve ) . as the source crosses inside the caustic ,
two additional images are created .
these images will be located close to , and on each side of the critical line ( blue curve ) .
therefore , it is perhaps not coincidental that the observed elongated galaxy gy is located close to the position where the two additional images are predicted to be . since morphologically gyb is not consistent with a point source , but only gya and gyc ( although this is not certain , due to the low s / n ) , the case when gya and gyc are two additional quasar images was considered .
however , this possibility is ruled out because the position of these two objects can not be reproduced , and in addition they are predicted to be brighter than image a. more specifically , two cases were considered , with or without accounting for gyc as an additional sis perturber ( 3sis+@xmath30 and 2sis+@xmath30 , respectively ) , and only the positions of the 5 images were fitted .
this resulted in @xmath591400 for 3 and 4 d.o.f .
, respectively . there is , however , another interpretation for gy .
its photometric redshift suggests that it may be associated with the quasar . as an isolated galaxy at high redshift
, it would be lensed by g1 and g2 into multiple images , a fact which is ruled out by the observed data .
it could , however , be interpreted as the quasar host galaxy . in fact , as the source is very close to the caustic and the host galaxy is an extended object , a portion of the host galaxy can cross the caustic and be significantly magnified , whereas the point - like quasar does not .
indeed , figure [ fig:1405mass ] shows that an extended circular source centered on the source quasar would be lensed into an elongated object whose position , elongation and orientation shows remarkable agreement with gy . to test this interpretation
, we modeled the source as a superposition of a point - like and an extended ( elliptical ) sersic profile , and fitted it to the observed data , using _ glafic_. we used the @xmath13 band , where gy is detected at larger s / n , and we modeled the psf with two concentric moffat profiles .
the reason why the hybrid psf considered in section [ sect : psf ] is not used is that , through the way it was constructed , it already includes a contribution from the extended source , which can not be modeled in this case as a different component .
we kept fixed the mass model parameters determined from fitting the astrometry and flux ratio constraints in section [ section : lens3 ] , and varied only the parameters of the psf and the extended source .
figure [ fig:1405torus ] depicts the result , which shows that the overall flux of gy is indeed well reproduced by the models , in case of two as well as three observed quasar images .
this profile modeling therefore can not solve the issue of the number of point - like images that are observed , given the current observational data .
faint features are still visible at the peaks of gya , gyb and gyc , which are not accounted for with this parametric - based source modeling .
we computed the total magnification of the extended source models as the flux ratio in the image plane and source plane , and found a model - dependent value of @xmath60 .
we also computed the magnification of the modeled gy profile as the ratio of the flux in gy , where all other images were masked , and the region of the source which is mapped into gy ( located inside the intersection area of all the caustics shown in figure [ fig:1405mass ] ) .
here we found a magnification factor of @xmath61 .
the magnification of gy relative to the whole unlensed source is @xmath62 .
[ tab : lensing ] the analysis above shows that the data is fully consistent with gy being the quasar host galaxy , lensed into an elongated arc .
a question that arises with this interpretation is why gy consists of three components .
this might be due to luminous substructure in the host galaxy , irregular overall morphology , a merging system , or small perturbers in the lens plane , which might disturb the arc .
in particular , since gya and gyb are on each side of the critical line , it is plausible that they are mirror - symmetric images of the same feature .
this was successfully checked by using one of the mass models predicting three quasar images , and considering a second point - like source to be lensed into gya and gyb ( figure [ fig:5imag ] ) .
the use of a second source sets tighter constraints on the mass model ( models 6 and 7 in table 7 ) .
three additional images are predicted , and their non - detection can presumably be explained by their close proximity to the quasar images , and their expected small fluxes .
the two sources are separated by @xmath63 in the source plane , or @xmath64 kpc .
this is on the order of one effective radius for typical galaxies at the source redshift ( e.g. , * ? ? ?
we also tried to fit two sources to the mass models predicting two quasar images , but this produced poor fits , with large lens ellipticity for g1 and g2 , at vastly different orientations than the observed light profiles . in conclusion , in view of the photometric redshifts , as well as the good overall fit with the flux distribution expected for an extended source crossing a caustic
, the data implies that gy indeed represents part of the quasar host galaxy .
we refrain however from making quantitative estimates of the physical properties of the host galaxy from the current data .
this is because of the quasar image number ( and therefore the magnification factor ) being unconfirmed , the fragmentation of gy , not accurately reproduced by a single analytical source profile , and the fact that it is difficult to assess whether the analytical psf is reliably disentangled from , or biases the parameters of the extended source
. band .
_ left : _ all objects have been fitted and subtracted with _ glafic _ , using an analytical psf , and without fitting a quasar host galaxy .
_ center _ and _ right : _ a quasar host galaxy modeled as a lensed sersic profile is also fitted , in the cases when either two or three quasar images are observed , successfully removing most of the flux in gy .
objects g2 and gx , in close proximity to each other , are masked , as the three - image lensing model can not reproduce the flux of gx .
[ fig:1405torus],width=317 ] , width=279 ]
we have conducted high resolution , near - infrared , multi - band adaptive optics observations with the subaru telescope of the gravitationally lensed quasar sdss j1405 + 0959 . despite the low strehl ratios , the observations were successful in uncovering unprecedented detail in this system , and in particular in discovering two new components named gx and gy .
we used a new technique to derive the psf , and determine the morphology , astrometry and photometry of the individual components that constitute the observed system .
we showed that gx is likely to be the third quasar image in this system , based on its colors as well as lens modeling .
if confirmed , it will be the first small - separation three - image lensed quasar produced by a pair of lensing galaxies , and in addition will allow the study of the interstellar medium of g2 .
the nature of gx may be tested via high spatial resolution spectroscopy , or alternatively via deep observations with the hubble space telescope ( @xmath65 ) , which produces a better characterized psf , and may distinguish between the point - source or extended morphology of gx .
in addition , such observations would also better separate the extended quasar host emission from the point - like quasar source .
we note that our technique to obtain the psf may be affected by faint host emission underneath the bright quasar image
. this might be responsible for some of the discrepancies in the morphology of g1 and g2 in the three observed bands , since the host galaxy is more prominent at longer wavelengths .
we also showed that gy consists of additional lensed images of an extended region of the quasar host galaxy , based on the observed colors and the match between the observed light distribution and that expected for an extended lensed source , given the best available lensing models .
adaptive optics - assisted or @xmath65 spectroscopy to determine a redshift would unambiguously confirm the nature of gy , as well as allow us to study its physical properties , such as star formation .
the added advantage here is that gy shows emission from the quasar host galaxy alone which is free of contamination from the bright quasar components .
deeper adaptive optics observations , to increase the s / n of gy , would provide hundreds of intensity pixels to further study the mass distribution ( e.g. , * ? ? ?
* ) in this two - galaxy lens system
. this would allow a non - parametric reconstruction of the quasar host galaxy in the source plane , which would be helpful to explain the fragmentation observed in gy and , for example , to study the correlation with the supermassive black hole ( e.g. , * ? ? ?
* ) . as an analogy , the known system most similar to sdss j1405 + 0959 is mg 2016 + 112 , which also contains an observed arc without a central quasar image .
that system also requires two lens galaxies to model , although the second lens is comparatively much weaker , with einstein radius ratio g1/d @xmath66 ( following the notations in the literature ) , whereas for sdss j1405 + 0959 , g2/g1 @xmath67 . as such , the tangential caustics in sdss j1405 + 0959 merge and the two radial caustics
overlap , whereas this does not occur in the case of mg 2016 + 112 , which only produces two quasar images .
the latter system however contains much more astrometric constraints , as radio observations @xcite reveal mirror - symmetric sub - images in both quasar images , as well as in the extended arc . on the other hand , sdss j1405 + 0959 is radio - quiet , based on non - detection in existing radio surveys .
this work was supported in part by the japan society for the promotion of science ( jsps ) kakenhi grant number 26800093 and world premier international research center initiative ( wpi initiative ) , ministry of education , culture , sports , science and technology ( mext ) , japan .
c.e.r . acknowledges the support of the jsps research fellowship . the authors would like to thank sherry suyu for helpful discussions
the authors recognize and acknowledge the very significant cultural role and reverence that the summit of mauna kea has always had within the indigenous hawaiian community .
we are most fortunate to have the opportunity to conduct observations from this superb mountain . | we present the result of subaru telescope multi - band adaptive optics observations of the complex gravitationally lensed quasar sdss j1405 + 0959 , which is produced by two lensing galaxies .
these observations reveal dramatically enhanced morphological detail , leading to the discovery of an additional object 026 from the secondary lensing galaxy , as well as three collinear clumps located in between the two lensing galaxies .
the new object is likely to be the third quasar image , although the possibility that it is a galaxy can not be entirely excluded .
if confirmed via future observations , it would be the first three image lensed quasar produced by two galaxy lenses . in either case
, we show based on gravitational lensing models and photometric redshift that the collinear clumps represent merging images of a portion of the quasar host galaxy , with a magnification factor of @xmath0 , depending on the model .
[ firstpage ] adaptive optics gravitationally lensed quasars
quasar host galaxies |
many mathematical models of physical phenomena deal with systems of volterra integral equations with singular kernels ( e.g. @xcite ) . in this paper , motivated by some nonlinear volterra integral equations arising in quantum mechanics ,
we investigate the properties of convolution operators with kernels possibly more singular than the more known abel ones .
namely , given a generic positive , locally integrable function @xmath0 , we study the action of the operator @xmath7 defined by @xmath8 on intervals @xmath9 $ ] , with @xmath10 ( this assumption being understood in the whole paper ) . precisely , we prove its regularizing effect in hlder and lebesgue spaces and its `` contractive '' effect in sobolev spaces ( where with `` contractive '' we mean that the sobolev norm of @xmath11 on @xmath9 $ ] can be estimated by the norm of @xmath12 times a constant that gets smaller as @xmath13 ) .
it is also worth highlighting that the assumption of local integrability of @xmath0 is the minimum requirement so that definition make sense in general .
in fact , the aim of the paper ( even though some results will require additional hypothesis ) is to work with the least set of assumptions that are necessary in order to detect remarkable effects from the application of the operator @xmath3 . a particular relevance in applications is acquired by the case @xmath14 ( figure [ fig : nu ] ) , where the operator @xmath3 reads @xmath15 we observe that , if we denote by @xmath16 the _ volterra functions _ defined by @xmath17 then @xmath18 coincides with @xmath19 , which is the so - called volterra function _ of order _ @xmath20 ( see @xcite , section 18.3 ) , that is discussed in section [ sec - volt ] .
in addition , recalling that a kernel @xmath21 is said a _ sonine kernel _ if it is a divisor of the unit with respect to the convolution operation , that is , if there exists another kernel @xmath22 such that @xmath23,\ ] ] then , one can prove that @xmath24 is a sonine kernel , with @xmath25 , @xmath26 representing the euler - mascheroni constant ( see eq . ) .
the class of sonine kernels is wide and there are many papers ( see e.g. @xcite and references therein ) , starting with the pioneering one by sonine ( @xcite ) , where embedding theorems for integral operators with kernels displaying singularities at the origin of the type @xmath27 are discussed .
however , we stress that the results proved in the present paper are more general since they take into account also kernels that are more singular in a neighborhood of the origin , such as , indeed , the volterra function @xmath24 , whose asymptotyc expansion near @xmath28 is given by @xmath29 ( see ) .
it is also worth mentioning that a first discussion on the operator @xmath30 is present in @xcite , whereas similar integral operators , but with more regular kernels , have been investigated more recently by @xcite .
more in detail , in @xcite it is analyzed the case of a certain class of _ almost decreasing _ sonine kernels in terms of _ weighted generalized _ hlder spaces , while in @xcite an `` inverse '' operator is discussed within the framework of @xmath31 spaces .
we also recall that in @xcite some relevant features of volterra functions are pointed out , such as asymptotic expansions and some striking relations with the ramanujan integrals .
the interest of the operator @xmath30 is mainly due to its applications in quantum mechanics , and precisely in the study of the schrdinger equation with nonlinear point interactions in @xmath32 .
we recall briefly that a schrdinger equation with a _ linear _ point interaction with strength @xmath33 , placed at @xmath34 , is @xmath35 where @xmath36 is a differential operator with domain @xmath37 \displaystyle \hspace{7 cm } \lim_{{\mathbf{x}}\to { \mathbf{y } } } \phi_\lambda({\mathbf{x } } ) = \left(\alpha+\tfrac{1}{2\pi } \log \tfrac{\sqrt{\lambda}}{2 } + \tfrac{\gamma}{2\pi}\right)q\bigg\ } \end{array}\ ] ] ( @xmath38 denoting the green s function of @xmath39 in @xmath32 ) and action @xmath40 for a complete discussion on the solution of this equation through the theory of self - adjoint extension , we refer the reader to @xcite .
in addition , it is well known that , given an initial datum @xmath41 , the solution of the associated cauchy problem reads @xmath42 where @xmath43 is the the _ propagator _ of the free schrdinger equation in @xmath32 ( with integral kernel @xmath44 ) and @xmath45 ( with a little abuse of notation that is usual in the literature ) is a complex - valued function satisfying the so - called _ charge _
equation @xmath46 now , a _ nonlinear _
point interaction arises when one assumes that the strength of the interaction depends in fact on the function @xmath45 , and in particular when one sets @xmath47 ( @xmath48 , @xmath49 ) in , thus obtaining @xmath50 ( see @xcite ) . since in the nonlinear case
no theory of self - adjoint extensions is available , the relevance of the operator @xmath30 is clear : the well - posedness of the associated cauchy problem is strictly related to the study of the existence and uniqueness of solutions of , which strongly depends on the properties of @xmath30 . even though the application presented above concerns complex - valued functions , this papers only manages real - valued functions .
however , one can check that the results of section [ sob ! ] ( which are actually required in @xcite ) can be easily generalized to complex - valued functions .
another topical example of integral kernels that are included in our general framework are the well known _ abel kernels _ , which correspond to the choice @xmath51 in .
these ones are very important in the theory of _ fractional integration _ and _ generalized differentiation _ ( @xcite ) and , again , in quantum mechanics . in the study of nonlinear point interactions in @xmath52 and @xmath53 , indeed
, the resulting integral equations present the kernel , with @xmath54 , in place of @xmath24 ( see @xcite ) .
finally , we describe briefly the main results of the paper .
they concern , as we told at the beginning , the properties of the operator @xmath3 in hlder spaces , @xmath31 spaces and sobolev spaces .
preliminarily , since it is crucial in the following , we define the integral function of the kernel @xmath0 @xmath55 since @xmath0 is always supposed positive and locally integrable , it turns out that @xmath4 is a positive , increasing and absolutely continuous function with @xmath56 as @xmath57 . in the case of hlder spaces ,
it is well known ( @xcite , theorem 4.2.1 p. 70 ) that when the kernel is @xmath58 , @xmath59 , the operator @xmath3 transforms @xmath60 functions into @xmath61 ones , improving this way the regularity of the modulus of continuity . as a consequence of our main result of section [ sectionhold ] ( theorem [ casoholder ] )
, we will see that more generally the improvement is at least given by the integral function of the kernel : the phenomenon that the power @xmath62 gives as improvement the exponent @xmath63 is therefore true also for any locally integrable kernel which is assumed just equivalent to a decreasing function in a neighborhood of the origin and not blowing too much ( derivative bounded above , for instance ) in its domain
. in the case of @xmath31 spaces , it is well known ( @xcite , theorem 4.1.4 p. 67 ) that when the kernel is @xmath58 , the operator @xmath3 transforms @xmath64 functions , @xmath65 , into @xmath66 functions . to a minor integrability of the kernel
corresponds a minor gain of integrability for @xmath67 , and apparently the gain disappears when the kernel is just @xmath68 . as a consequence of our main result of section [ sectionellepi ] ( theorem [ casolp ] ) , we will show that _ any kernel locally integrable ( again , we assume that it is equivalent to a decreasing function in a neighborhood of the origin ) gives an improvement of integrability , measured in terms of orlicz spaces . the improvement is strictly linked to the orlicz integrability of the kernel , hence it always exists : a classical , remarkable theorem in orlicz spaces theory ( see e.g. @xcite , p. 60 ) tells that any function @xmath68 is always in some orlicz space strictly contained in @xmath68 .
furthermore , in the case @xmath69 , we show ( proposition [ pro : infinity ] ) , under the unique assumption of local integrability , that @xmath3 transforms @xmath70 functions in continuous functions and that the @xmath70 norm of @xmath11 on @xmath9 $ ] is controlled by the norm of @xmath12 times @xmath71 .
_ finally , in the case of sobolev spaces , it is well known ( @xcite , theorem 4.2.2 p. 73 ) that when the kernel is @xmath58 , the operator @xmath3 transforms @xmath72 functions , with @xmath73 , in @xmath74 functions .
analogous results for @xmath75 functions are discussed in @xcite . in this case
, the minor integrability of the kernel yields a minor gain in the sobolev index , which disappears when the kernel is just @xmath68 ( also the preservation of the index is not straightforward ) . as a consequence of our main result of section
[ sob ! ] ( theorem [ contr_lemma ] ) , we will show that when @xmath76 , provided @xmath77 , the sobolev index is in fact preserved and the sobolev norm of @xmath11 is bounded , up to a multiplicative constant , by the norm of @xmath12 times @xmath71 .
furthermore , we will prove that ( almost ) the same result holds for @xmath78 functions , but just in the case @xmath79 ( theorem [ contr_lemma_2 ] ) , and for @xmath80 functions ( theorem [ contr_lemma_3 ] ) .
since the case of a kernel equal to the volterra function @xmath24 ( defined by , figure [ fig : nu ] ) is the most relevant in the applications , it is worth stressing some basic features of @xmath24 . in this way
one can easily see that the abstract results established in the following sections can be actually applied to this kernel .
first , we recall ( see @xcite ) that @xmath24 is analytic for @xmath82 and that @xmath83 , & \quad\text{as } x\to 0\\[1 cm ] \displaystyle { \mathcal{i}}(x ) = e^{x}+\mathcal{o}(x^{-1 } ) , & \quad\text{as } x\to + \infty . \end{array}\ ] ] consequently , the first expansion shows that @xmath84 and that @xmath85 , for any @xmath86 .
is in black , the plot of the first order of the asymptotic expansion of @xmath18 around 0 is dotted.,scaledwidth=40.0% ] also the derivative and the integral function of @xmath24 will play a crucial role in the sequel .
hence , recalling ( see again @xcite ) @xmath87 and @xmath88 , we stress that @xmath89\ ] ] and that @xmath90 furthermore , we can state the following lemma . [ lem : conv ] the function @xmath24 is convex on @xmath91 and admits a positive minimum .
the second part is immediate since @xmath24 is continuous , positive and coercive by . on the other hand , in order to prove the second part , it is sufficient to show @xmath92 ; namely , by , that @xmath93 .
now , following @xcite ( eq . ( 3.1 ) ) and
@xcite , we find that @xmath94 denotes the _ ramanujan _ function ( figure [ fig : erre ] ) .
hence , @xmath95 and , since @xmath96 is _ completely monotonic _
( i.e. , for every @xmath97 , @xmath98 does exist and @xmath99 ) , this entails that @xmath93 and thus that @xmath24 is convex .
.,scaledwidth=40.0% ] it is also convenient to introduce the function @xmath100 ( figure [ fig : enne ] ) . by the properties of @xmath81 , we see that @xmath101 is positive , increasing and absolutely continuous on bounded intervals intervals .
moreover , @xmath102 as @xmath57 , and precisely @xmath103 around zero , scaledwidth=50.0% ] one easily sees that , as one sets @xmath79 in , @xmath4 is equal to @xmath101 . on the other hand , from one
also notes that @xmath101 coincides , up to an additive constant , with @xmath104 .
finally , we point out a relevant property of the operator @xmath30 defined by , which is strictly connected to the fact that @xmath24 is a sonine kernel .
first , define the integral operator @xmath105 then , one notes that , as @xmath106 , @xmath107 is well defined for each function @xmath108 . in addition , one can prove the following result . [ pro : i_identity ] if @xmath109 , then @xmath110.\ ] ] we first observe that one has @xmath111 in @xcite , lemma 32.1 , it is indeed claimed that ( setting @xmath112 therein ) @xmath113 with @xmath114 , and since from @xmath115 , is proved .
now , in the expression @xmath116 we exchange the order of the integration , since @xmath117 \displaystyle = \int_0^{x } \int_0^{x } { \mathcal{i}}(s ) \phi(|x-\sigma - s| ) g(\sigma)ds\,d\sigma = 2 \int_0^{x } \int_0^{x - s } { \mathcal{i}}(s ) \phi(|x-\sigma - s| ) g(\sigma)d\sigma\,ds , \end{array}\ ] ] and using , we conclude that @xmath118 finally , given that an easy change of variable shows @xmath119 , the proof is complete .
we note that , in view of theorem [ contr_lemma_3 ] , if one also assumes @xmath120 in proposition [ pro : i_identity ] , then @xmath121
let @xmath122 and let @xmath0 , @xmath123 be absolutely continuous and positive in @xmath1240,\tau_0]$ ] .
we say that @xmath0 , @xmath123 are @xmath125 if there exist two positive constants @xmath126 such that @xmath1270,\tau_0].\ ] ] of course any function equivalent to @xmath0 in @xmath1240,\tau_0]$ ] is of type @xmath128 , where @xmath129 is absolutely continuous in @xmath1240,\tau_0]$ ] and such that @xmath1300,\tau_0].\ ] ] the statements of this section hold for certain functions @xmath0 which are decreasing in intervals of the type @xmath131 and , more generally , they hold for functions equivalent to decreasing functions in intervals of the type @xmath131 . for the sake of simplicity , the functions @xmath0 in the statements will be always assumed equivalent to decreasing functions in intervals of the type @xmath131 , and the corresponding decreasing functions will be written as products @xmath128 , where @xmath129 is an absolutely continuous function in @xmath1240,\tau_0]$ ] satisfying .
[ primolemma ] let @xmath132 , and let @xmath1330,t])\cap l^1(0,t)$ ] be positive and equivalent to a decreasing function in @xmath131 for some @xmath134 . if @xmath135 then @xmath136 where @xmath4 is defined by . inequalities coming from assumptions of monotonicity of ratios between functions and powers are very well known among researchers working in orlicz spaces .
some proofs of such inequalities work also without the assumption of convexity ( the reader may compare this lemma e.g. with theorem 3 in @xcite or with the results in section 3 of @xcite ) , however , the main feature of is that it has been obtained from assumptions of monotonicity which hold only in a neighborhood of the origin and not in the whole domain of the functions involved ( where , however , at least a boundedness is required ; this assumption appears implicitly in the hypothesis of continuity which holds until the endpoint @xmath137 ) .
we preliminarly note that , since @xmath128 is decreasing in @xmath131 , @xmath138\ , .\ ] ] let @xmath139 , @xmath140 . if @xmath141 , by and respectively , @xmath142 and , if @xmath143 , @xmath144 then again by , @xmath145 and therefore , setting @xmath146}\nu\ , , $ ] @xmath147 finally , if @xmath143 , @xmath148 then , since @xmath149 implies that @xmath4 is increasing , @xmath150 from , , , we get .
the following statement is an immediate consequence of lemma [ primolemma ] .
[ usatodavvero ] in the same assumptions of lemma [ primolemma ] , for any @xmath151 it is @xmath152 uniformly in @xmath153 , @xmath139 .
the next lemma is trivially true for decreasing functions @xmath0 ( see the case ( i ) of the proof ) , and it provides a version of the inequality in case of functions which are decreasing only in a neighborhood of the origin ( however , as in the remark above , one can see that , again , an assumption of boundedness has been made implicitly ) .
[ secondolemma ] if @xmath1330,t])\cap l^1(0,t)$ ] is positive and equivalent to a decreasing function in @xmath131 for some @xmath134 , then @xmath154 we examine the three cases \(i ) @xmath155 \(ii ) @xmath156 \(iii ) @xmath157 case ( i ) : since @xmath128 is decreasing in @xmath158 , @xmath159 case ( ii ) : we have @xmath160 where @xmath161}\nu$ ] .
there are two possibilities : @xmath162 @xmath163 @xmath164 @xmath165 in the case @xmath162 , since @xmath128 is decreasing in @xmath158 , @xmath166 and therefore from @xmath167 in the case @xmath164 , setting @xmath1680,t]}\nu>0 $ ] , @xmath169 and then , using @xmath170 , from we get @xmath171 case ( iii ) : we have @xmath172 and applying case ( i ) with @xmath173 replaced by @xmath174 and case ( ii ) with @xmath175 replaced by @xmath174 , @xmath176 since in our case @xmath177 , it is @xmath178 , hence the first term can be estimated by the right hand side of ; similarly , since @xmath179 , it is @xmath180 and the same conclusion holds for the second term . in next theorem
we are going to consider an assumption on @xmath0 stronger ( as we are going to see ) with respect to that one of lemma [ primolemma ] : in the case @xmath181 ( a similar digression can be done in the general case , replacing @xmath0 by @xmath128 ) we will assume that the positive function @xmath1330,t])\cap l^1(0,t)$ ] is such that @xmath182 is decreasing in @xmath183 for some @xmath134 , @xmath184 , where @xmath63 is a given number in @xmath185 .
it is easy to verify that this latter assumption implies that the function @xmath4 , defined by , is such that @xmath186 and also that @xmath0 is decreasing in @xmath183 ( because @xmath187 is product of positive decreasing functions ) .
let us verify the first assertion .
since @xmath188 decreasing in @xmath183 , it is ( note that , since also @xmath0 is absolutely continuous , their derivatives exist a.e . )
@xmath189 hence , integrating the above inequality in @xmath190 , where @xmath191 , and noting that @xmath192 is decreasing too , we have @xmath193 @xmath194 if we let @xmath195 , since @xmath0 is decreasing , it is @xmath196 and therefore we get @xmath197 from which the assertion follows .
on the other hand , the fact that the assumption is really stronger is shown by the following let @xmath198 , and let @xmath199 so that @xmath200 then @xmath201 is satisfied for @xmath202 , while for the same @xmath203 the function @xmath182 is not decreasing in @xmath183 ( because it is not decreasing in @xmath204 ) . before the statement of the main theorem of this section ,
we observe that the kernels @xmath0 of our interest are such that their difference quotients are bounded above , i.e. there exists a constant @xmath205 such that @xmath206 this property ( which holds automatically , in particular , for all the kernels @xmath0 which are decreasing in the whole @xmath207 ) is expressed in an equivalent way in the assumption below .
[ casoholder ] if @xmath208)$ ] , @xmath59 , @xmath209 , @xmath210 , and if @xmath1330,t])\cap l^1(0,t)$ ] , @xmath211 , is such that @xmath212 @xmath2130,\tau_0[)\ , , \ , 0<b^-\le b(x)\le b^+<\infty\ , , \ ] ] @xmath214 then , setting @xmath215 it is @xmath216 let @xmath217 .
it is @xmath218ds\\ & -\int_{y - x}^y \nu(s - y+x)g(y)ds+\int_{y - x}^y \nu(s - y+x)[g(y)-g(y - s)]ds\\ & = g(y)\int_0^y \nu(s)ds-\int_0^{y - x } \nu(s)[g(y)-g(y - s)]ds-\int_{y - x}^y \nu(s)[g(y)-g(y - s)]ds\\ & -g(y)\int_{y - x}^y \nu(s - y+x)ds+\int_{y - x}^y \nu(s - y+x)[g(y)-g(y - s)]ds\\ & = g(y)[n(y)-n(x)]-\int_0^{y - x } \nu(s)[g(y)-g(y - s)]ds\\ & + \int_{y - x}^y [ \nu(s - y+x)-\nu(s)][g(y)-g(y - s)]ds .
\end{split}\ ] ] therefore @xmath219|+\left\vert\int_0^{y - x } \nu(s)[g(y)-g(y - s)]ds\right\vert\\ & + \left\vert\int_{y - x}^y [ \nu(s - y+x)-\nu(s)][g(y)-g(y - s)]ds\right\vert:=a+b+c .
\end{split}\ ] ] we estimate each term in turn . since @xmath208)$ ] , @xmath210 , and since @xmath4 is increasing and @xmath59 , @xmath220|\\ & \le [ g]_\alpha y^\alpha [ n(y)-n(x)]\\ & = [ g]_\alpha ( y^\alpha - x^\alpha ) [ n(y)-n(x)]+[g]_\alpha x^\alpha [ n(y)-n(x)]\\ & \le [ g]_\alpha ( y - x)^\alpha [ n(y)-n(x)]+[g]_\alpha x^\alpha [ n(y)-n(x ) ] .
\end{split}\ ] ] by the function @xmath128 is decreasing in @xmath183 , hence by @xmath221 @xmath222 and therefore @xmath223_\alpha ( y - x)^\alpha n(y - x)+[g]_\alpha x^\alpha [ n(y)-n(x)].\ ] ] the first term can be estimated by the right hand side of . as to the second term , we begin observing that @xmath224=x^\alpha \frac{n(y)-n(x)}{y - x}(y - x ) = x^\alpha \nu(\xi)(y - x),\ ] ] for some @xmath225 .
we consider first the case @xmath226 , so that @xmath227 ( because @xmath128 is decreasing in @xmath183 ) .
since @xmath151 , by @xmath228 and therefore , since @xmath229 , @xmath230 @xmath231 @xmath232 @xmath233 now we consider the case @xmath234 .
setting @xmath235}x^\alpha\nu ( x ) $ ] , @xmath2360,t]}x^\alpha\nu ( x)>0 $ ] , it is @xmath237 now there are two possibilities : @xmath238 , @xmath239 . in the first possibility , @xmath240 and therefore @xmath241 @xmath242 in the second possibility , @xmath243 and therefore , using @xmath244 , @xmath245 @xmath246 we have therefore shown that also the second term can be estimated by the right hand side of . on the other hand , it is quite easy to check that also @xmath247 can be estimated in the same way , in fact @xmath248ds\right\vert\le \int_0^{y - x } \nu(s)[g]_\alpha s^\alpha ds\ ] ] @xmath249_\alpha\int_0^{y - x } \nu(s ) ( y - x)^\alpha ds= [ g]_\alpha ( y - x)^\alpha n(y - x).\ ] ] it remains to estimate @xmath250 .
it is @xmath251[g(y)-g(y - s)]ds\right|\le [ g]_\alpha \int_{y - x}^y |\nu(s - y+x)-\nu(s)|s^\alpha ds\ ] ] we now use the assumption , so that @xmath252_\alpha \int_{y - x}^y \left[\nu(s - y+x)-\nu(s)\right]s^\alpha ds+[g]_\alpha\int_{y - x}^y k_0(y - x)s^\alpha ds\ , .\ ] ] making the change of variables @xmath253 in the first term , we have @xmath254_\alpha ( y - x)^{\alpha+1 } \int_{1}^{\frac{y}{y - x}}\left [ \nu((y - x)(\sigma-1 ) ) - \nu((y - x)\sigma)\right]\sigma^\alpha d\sigma+[g]_\alpha\int_{y - x}^y k_0(y - x)s^\alpha ds\ ] ] we observe that @xmath255 implies @xmath256 , therefore @xmath257_\alpha ( y - x)^{\alpha+1}\left\ { \int_{1}^{2}\left [ \nu((y - x)(\sigma-1 ) ) - \nu((y - x)\sigma)\right]\sigma^\alpha d\sigma+\right.\\&+ \left.\int_{2}^{\frac{y}{y - x}}\left [ \nu((y - x)(\sigma-1 ) ) - \nu((y - x)\sigma)\right]\sigma^\alpha d\sigma \right\}+[g]_\alpha\int_{y - x}^y k_0(y - x)s^\alpha ds \\ & : = c_{1}+c_{2}+c_{3}\end{split}\ ] ] we observe that the sign of @xmath258 and the sign of @xmath259 are not necessarily positive ; both @xmath258 and @xmath259 will be splitted into more terms , each of them being not necessarily positive ; however , all of them will be shown to be smaller than the right hand side of .
the first term can be estimated as follows : @xmath260_\alpha ( y - x)^{\alpha+1 } \int_{1}^{2}\left [ \nu((y - x)(\sigma-1 ) ) - \nu((y - x)\sigma)\right ] \sigma^\alpha d\sigma\\ & \leqslant 2^{\alpha}[g]_\alpha ( y - x)^{\alpha+1 } \int_{1}^{2}\nu((y - x)(\sigma-1 ) ) d\sigma \\ & = 2^{\alpha}[g]_\alpha ( y - x)^{\alpha } \int_{0}^{y - x}\nu(\tau ) d\tau \\ & = 2^{\alpha}[g]_\alpha ( y - x)^{\alpha } n(y - x)\ , .
\end{split}\ ] ] in order to estimate @xmath259 , we need to use the following inequality , which , as we are going to see , follows easily from the fact that the positive @xmath2610,t])$ ] is decreasing in @xmath183 : @xmath262 where @xmath161}\nu$ ] and @xmath263}\nu>0 $ ] . in order to show ,
let us fix @xmath264 .
if @xmath265 , using that @xmath128 is decreasing in @xmath183 , we have @xmath266 if @xmath267 , @xmath268 the last case is @xmath269 , where the following inequalities hold : @xmath270 @xmath271 using the change of variable @xmath272 , the integration by parts formula , @xmath211 , and lemma [ secondolemma ] , we now estimate @xmath259 as follows : @xmath273_\alpha ( y - x)^{\alpha+1 } \int_{2}^{\frac{y}{y - x}}\left [ \nu((y - x)(\sigma-1 ) ) - \nu((y - x)\sigma)\right]\sigma^\alpha d\sigma\\ & = [ g]_\alpha ( y - x)^{\alpha+1 } \int_{1}^{\frac{x}{y - x}}\left [ \nu((y - x)\tau ) - \nu((y - x)(\tau+1))\right](\tau+1)^\alpha d\tau\\ & = [ g]_\alpha ( y - x)^{\alpha+1 } \int_{1}^{\frac{x}{y - x}}(\tau+1)^\alpha d\left(\frac{1}{y - x}\int_{(y - x)(\tau+1)}^{(y - x)\tau}\nu ( s)ds\right)\\ & = [ g]_\alpha ( y - x)^{\alpha+1 } \left\ { \left .
\frac{(\tau+1)^\alpha}{y - x}\int_{(y - x)(\tau+1)}^{(y - x)\tau}\nu ( s)ds \right\vert^{\tau=\frac{x}{y - x}}_{\tau=1}\right.\\ & \qquad\qquad\qquad\qquad\qquad\qquad\qquad-\left .
\frac{1}{y - x } \int_{1}^{\frac{x}{y - x}}\left(\int_{(y - x)(\tau+1)}^{(y - x)\tau}\nu ( s)ds\right)\alpha ( \tau+1)^{\alpha-1}d\tau \right\}\\ & = [ g]_\alpha ( y - x)^{\alpha } \left\ { \left .
( \tau+1)^\alpha\int_{(y - x)(\tau+1)}^{(y - x)\tau}\nu ( s)ds \right\vert^{\tau=\frac{x}{y - x}}_{\tau=1}\right.\\ & \qquad\qquad\qquad\qquad\qquad\qquad\qquad+\left .
\int_{1}^{\frac{x}{y - x}}\left(\int_{(y - x)\tau}^{(y - x)(\tau+1)}\nu ( s)ds\right)\alpha ( \tau+1)^{\alpha-1}d\tau \right\}\\ & = [ g]_\alpha ( y - x)^{\alpha } \left\ { \left(\frac{y}{y - x}\right)^{\alpha}\int_{y}^{x}\nu(s)ds-2^{\alpha}\int_{2(y - x)}^{y - x}\nu(s)ds\right.\\ & \qquad\qquad\qquad\qquad\qquad\qquad\qquad+\left .
\int_{1}^{\frac{x}{y - x}}\left[\int_{(y - x)\tau}^{(y - x)(\tau+1)}\nu ( s)ds\right]\alpha ( \tau+1)^{\alpha-1}d\tau \right\}\\ & \leqslant [ g]_\alpha ( y - x)^{\alpha}\left\{2^{\alpha}\int_{y - x}^{2(y - x)}\nu(s)ds + c\left(b(\cdot),\nu(\cdot)\right ) \int_{1}^{\frac{x}{y - x}}\left [ ( y - x)\nu ( ( y - x)\tau ) \right]\alpha ( \tau+1)^{\alpha-1}d\tau \right\}\\ & \leqslant [ g]_\alpha ( y - x)^{\alpha}\left\{2^{\alpha}c\left(b(\cdot),\nu(\cdot),t\right)n(y - x)\right.\\ & \qquad\qquad\qquad+\left .
c\left(b(\cdot),\nu(\cdot)\right)(y - x)\alpha\frac{n(y - x)}{y - x}\int_{1}^{\frac{y}{y - x}}(\tau+1)^{\alpha-1}\frac{\nu((y - x)\tau)\,(y - x)}{n(y - x)}d\tau\right\ } \end{split}\ ] ] by corollary [ usatodavvero ] the integral inside the parenthesis is bounded by a constant independent of @xmath274 , depending only on @xmath275 .
hence the estimate becomes @xmath276_\alpha ( y - x)^{\alpha}\left\{2^{\alpha}c\left(b(\cdot),\nu(\cdot),t\right)n(y - x ) + \alpha \textit{c}\left(b(\cdot),\nu(\cdot),t,\alpha+\beta\right ) n(y - x ) \right\}\ ] ] and therefore also @xmath259 is estimated by the right hand side of .
finally , we need to estimate @xmath277 .
we have @xmath278_\alpha\int_{y - x}^y k_0(y - x)s^\alpha ds\le [ g]_\alpha k_0 \int_0^t s^\alpha ds ( y - x)\ ] ] @xmath279_\alpha(y - x)=\textit{c}\left(\nu(\cdot),t,\alpha\right)[g]_\alpha \frac{(y - x)^{1-\alpha}}{n(y - x)}(y - x)^{\alpha}n(y - x)\ ] ] @xmath280_\alpha(y - x)^{\alpha}n(y - x)\ ] ] where the last inequality follows from the fact that from and from @xmath281 it follows that @xmath282 hence @xmath283 is bounded below in @xmath207 by a positive constant , i.e. its reciprocal is bounded above by a positive constant ( depending only on @xmath284 , @xmath0 and @xmath63 ) .
in the case @xmath181 , @xmath285 , @xmath286 , theorem [ casoholder ] gives back theorem 4.2.1 in @xcite .
another interesting case is @xmath287 which satisfies the assumption of theorem [ casoholder ] for _ any @xmath59 , for _ any @xmath184 , with @xmath181 .
of course those positive functions @xmath1330,t])\cap l^1(0,t)$ ] , which are just _ equivalent to the right hand side of only in a neighborhood of the origin , and then not blowing up `` too much '' ( as , for instance , the function @xmath288 in , whose derivative see is again a volterra function which is bounded above ) , are examples for theorem [ casoholder ] and in such cases the resulting regularity for @xmath3 is the same as that one given for . _ _ _ from the proof of theorem [ casoholder ] it is clear that the assumption can be weakened as follows : @xmath289 for a given @xmath0 satisfying the assumptions of theorem [ casoholder ] one may look for the _ best regularity action for @xmath3 , i.e. one may look for the greatest @xmath63 satisfying the assumptions of the theorem or , equivalently , for the smallest @xmath203 satisfying ( in the case of the classical spaces of hlder continuous functions , the inclusions between the spaces are easy and well known ; for a recent book on this topic see @xcite ) : this problem is linked to the notion of boyd indices ( see e.g. @xcite ) . for a short survey including a bibliography on this topic , and for a `` concrete '' way to compute them for explicit examples , see e.g. @xcite,@xcite . _
we begin some background on young s functions and orlicz spaces . in the following a convex function @xmath290 defined on @xmath291
is said to be a young s function if it is convex and such that @xmath292 , @xmath293 for @xmath294 .
this assumption implies that young s functions are strictly increasing and invertible , so that it makes sense to consider its inverse @xmath295 , defined in @xmath291 .
the orlicz space @xmath296 ( here @xmath137 is a fixed positive real number ) is the banach function space of all real - valued ( lebesgue ) measurable functions @xmath297 on @xmath207 such that @xmath298 ( here we use the convention @xmath299 ) . in the special case @xmath300 , @xmath301 , the orlicz space reduces to the familiar lebesgue space . for essentials about orlicz spaces and banach function spaces
the reader may refer to @xcite , sections 2.10.2 and 2.10.3 ( and references therein for extensive treatments ) .
a well known result of the theory is that if two young s functions @xmath290 , @xmath247 are such that @xmath302 then , in spite @xmath303 , @xmath304 may be different , the spaces themselves ( namely , the set of the functions such that the norms are finite ) coincide . in particular , the spaces are completely determined by the values of the young s functions assumed for @xmath175 large .
for measurable functions @xmath305 on @xmath207 , the decreasing rearrangement @xmath306 is defined by the right continuous inverse of @xmath307 , i.e. @xmath308 .
orlicz spaces @xmath296 are rearrangement - invariant : this means , in particular , that the norm is not affected after the action of the decreasing rearrangement operator : @xmath309 .
we may state the following [ casolp ] let @xmath310 , @xmath209 , and let @xmath290 be a young s function . if @xmath311 , @xmath211 , is such that @xmath312 @xmath313 @xmath314 where @xmath315 then , setting @xmath215 it is @xmath316 where @xmath250 is the young s function ( proof of lemma 4.2 in @xcite ) defined by @xmath317 before giving the proof of the theorem , which is a quite easy consequence ( in fact , an application ) of a classical result about fractional integration in orlicz spaces , we highlight a couple of examples which are relevant for this paper .
let @xmath310 and @xmath318 , and let @xmath319 , for all @xmath320 .
then @xmath321 satisfies , hence @xmath250 given by is @xmath322 , and @xmath323 for all @xmath324 .
this special case gives back the refined version of the continuity property for the abel operator in @xmath64 spaces , see theorem 4.1.3 in @xcite .
notice that when @xmath63 approaches @xmath28 , the exponent @xmath325 of integrability of @xmath0 approaches @xmath326 , and the exponent @xmath327 of integrability of @xmath328 approaches @xmath329 , which means no gain of integrability : in the framework of the lebesgue spaces , kernels in @xmath68 which do not possess the higher integrability property ( see e.g. next two examples and , in particular , the volterra function @xmath288 ) are not able to improve the integrability through the operator @xmath67 .
let @xmath310 and let @xmath330 , and let @xmath331 .
then @xmath332 hence @xmath333 satisfies .
the young s function @xmath250 given by is @xmath334 and @xmath335 for all @xmath324 .
it is interesting to note that the kernel does not belong to any lebesgue space @xmath64 with @xmath86 , and that the logarithm in the expression of @xmath250 ( which has a positive power and therefore it is divergent at infinity ) represents an orlicz gain of integrability for @xmath328 . for any kernel considered in theorem [ casolp ]
, the existence of a young function @xmath290 satisfying can be easily established ; moreover , for any @xmath0 the function @xmath328 always enjoys an orlicz gain of integrability with respect to @xmath12 : this is the heart of the following simple result , which is consequence of standard statements of orlicz spaces theory , namely , of the fact that any function @xmath336 is always in some orlicz space @xmath337 strictly contained in @xmath336 ( see e.g. @xcite , p.60 ) , of the hlder s inequality in orlicz spaces ( see e.g. @xcite , 8.11 p. 234 ) @xmath338 where @xmath339 is the young function defined by @xmath340 , of the equivalences ( see e.g. @xcite , ( 7 ) p. 230 and
@xcite , respectively ) @xmath341 @xmath342 and finally of @xmath343 in the assumptions of theorem [ casolp ] , for every @xmath0 there exists a young function @xmath290 satisfying , and therefore @xmath344 is strictly contained in @xmath345 .
let @xmath290 be a young function such that @xmath346 , @xmath347 being increasing and divergent at infinity . by , , respectively ,
@xmath348 as to the second part of the statement , from we get that setting , for large @xmath349 , @xmath350 and @xmath351 ( note that @xmath175 and @xmath352 are large as well ) @xmath353 and using both implications in , we get the assertion .
it is immediate to realize that the statement of theorem [ casolp ] remains true if @xmath0 is replaced by any function equivalent to @xmath0 in a neighborhood of the origin .
hence all the previous remark still holds if @xmath0 is replaced by the function @xmath288 in .
the proof of theorem [ casolp ] will follow as consequence of the following result appeared in sharpley ( ( * ? ? ?
* theorem 3.8 ) ) , in the more abstract setting of general convolution operators , defined in @xcite . in this latter paper
our operator @xmath3 , which goes back to @xcite , is explicitly mentioned as example ( see the end of the section iv therein ) . here
we state it in a more convenient form , and using our notation : [ bysharpley ] suppose that @xmath290 , @xmath247 are young s functions such that @xmath354 and @xmath355 where @xmath250 is the young s function defined by @xmath356 then , @xmath357 where ( @xmath3 is defined by and ) @xmath358 denotes the averaged rearrangement of @xmath0 , defined by @xmath359 _ proof of theorem [ casolp ] . _ setting @xmath360 , is obviously satisfied with equality and @xmath361 ; from we get that if @xmath250 is defined by @xmath362 i.e. if holds , then is satisfied : in fact , from the convexity of the young function @xmath290 , the function @xmath363 is decreasing , hence from we deduce that also the following ones are decreasing : @xmath364 and therefore @xmath365 is increasing , from which is satisfied with @xmath366 .
we are therefore allowed to apply theorem [ bysharpley ] . from and
it follows that for small values of @xmath352 it is @xmath367 , hence , by , for @xmath175 small it is @xmath368 and of course the same conclusion holds ( for a possibly different @xmath369 ) for all @xmath370 .
hence the supremum in the right hand side of is a finite constant @xmath371 , from which follows . finally , the case @xmath69 is considered in the next result , where the regularizing effect of @xmath67 is expressed through continuity .
[ pro : infinity ] if @xmath149 and @xmath311 , then for any @xmath372 there results @xmath373)\ ] ] and @xmath374 where @xmath4 is the integral function of @xmath0 defined by .
recalling and , is immediate .
then , it is left to prove that @xmath11 is continuous .
to this aim , fix @xmath375 and @xmath376x_0,t]$ ] .
easy computations yield @xmath377}(s)g(s)ds+\\[.2 cm ] & \ , -\int_0^t\big(\nu(x_0-s)-\nu(x - s)\big){\mathds{1}}_{[0,x_0]}(s)g(s)ds \end{aligned}\ ] ] and hence @xmath378}(s)|g(s)|ds+ \\ & \,+\int_0^t|\nu(x_0-s)-\nu(x - s)|{\mathds{1}}_{[0,x_0]}(s)|g(s)|ds \\
\leq & \,n(x - x_{0})\|g\|_{l^\infty(0,t)}+\|g\|_{l^\infty(0,t ) } \int_\mathbb{r}|\nu_e(x_0-s)-\nu_e(x - s)|ds , \end{aligned}\ ] ] where @xmath3790,t[\,,\\[.2 cm ] 0 , & \text{if } s\in{\mathbb{r}}\backslash\;]0,t[\ , .
\end{array } \right.\ ] ] therefore , the first term converges to zero by the continuity of @xmath4 and the second term converges to zero by the _
mean continuity property _
( see @xcite ) . since the same holds if @xmath380 , one has @xmath381 , which concludes the proof .
we start with some basics on sobolev spaces with _ fractional _ index .
let @xmath382 and @xmath383 .
we denote by @xmath384 the sobolev space defined by @xmath385_{\dot{h}^\theta(a , b)}^2<\infty\},\ ] ] where @xmath386_{\dot{h}^\theta(a , b)}^2 : = \int_{[a , b]^2}\frac{|g(x)-g(y)|^2}{|x - y|^{1 + 2\theta}}dydx.\ ] ] this is a hilbert space with the natural norm @xmath387_{\dot{h}^\theta(a , b)}^2.\ ] ] when @xmath388 and @xmath389 , @xmath390 can be equivalently defined using the fourier transform ( see @xcite ) ; that is , if we define the fourier transform as @xmath391 then there exist two constants @xmath392 such that @xmath393 [ lem - ext ] let @xmath399 , @xmath400 , and set @xmath401,\\[.3 cm ] g(2t - x ) , & \text{if } x\in\;]t,2t],\\[.3 cm ] 0 , & \text{if } x\in{\mathbb{r}}\backslash[0,2 t ] .
\end{array } \right.\ ] ] the following holds : the proof is a straightforward application of lemma 2.1 in @xcite .
cases @xmath407 are trivial .
consider , then , an arbitrary @xmath408 .
first , we can easily check that @xmath409 and that , with some change of variables , @xmath410_{\dot{h}^\theta(0,2t)}^2 = 2[g]_{\dot{h}^\theta(0,t)}^2 + 2\int_{[0,t]^2}\frac{|g(x)-g(y)|^2}{|x+y-2t|^{1 + 2\theta}}dydx.\ ] ] moreover , since @xmath411 for every @xmath412 ^ 2 $ ] , @xmath413 ^ 2}\frac{|g(x)-g(y)|^2}{|x+y-2t|^{1 + 2\theta}}dydx \leq [ g]_{\dot{h}^\theta(0,t)}^2.\ ] ] hence @xmath414_{\dot{h}^\theta(0,2t)}^2 \leq 4[g]_{\dot{h}^\theta(0,t)}^2 $ ] , so that @xmath415 now , from lemma 2.1 of @xcite , we know that if @xmath4040,1/2[\,$ ] , then there exists @xmath416 such that @xmath417 on the other hand , the same lemma shows that holds even if @xmath4041/2,1[\,$ ] , provided that @xmath210 ( since this entails by definition @xmath418 ) . combining and , the proof is complete .
note that , when @xmath419 , the assumption @xmath210 is meaningful as @xmath12 is continuous by sobolev embeddings ( @xcite ) .
what is more , this requirement is mandatory , since otherwise @xmath420 might not preserve continuity on @xmath52 .
we also stress that the case @xmath421 is not managed by lemma [ lem - ext ] since lemma 2.1 in @xcite is not valid in general for this choice of @xmath422 , due to the failure of hardy inequality ( see e.g. @xcite ) .
fix @xmath423\backslash\{1/2\}$ ] . then , let again @xmath4260,t[\,,\\[.3 cm ] 0 , & \text{if } x\in{\mathbb{r}}\backslash\;]0,t [ \end{array } \right.\ ] ] and , for any @xmath399 , define @xmath427 where @xmath420 is the extension of @xmath12 obtained via lemma [ lem - ext ] . note that applies if either @xmath428 or @xmath419 with , in this second case , the further assumption that @xmath210 . as @xmath429 for
all @xmath430 $ ] , @xmath431 now , by definition @xmath432}(s)ds\,dx.\ ] ] since @xmath433}(s ) = h(s ) - h(s - x)$ ] , where @xmath434 denotes the heaviside function , @xmath435 with @xmath436 . consequently , by well known properties of the fourier transform , @xmath437 thus , noting that @xmath438 and @xmath439 , @xmath440 then , combining , and , @xmath441 moreover , we observe that @xmath442 and , plugging into , that @xmath443 combining with , follows .
we observe that the `` contractive '' effect of @xmath3 , pointed out in the introduction , is in the fact that @xmath444 , as @xmath13 .
it entails that on small intervals the operator `` shrinks '' the norm of the argument function by a factor that gets smaller whenever @xmath137 gets smaller .
note that in the previous theorem , when @xmath419 the assumption @xmath210 can not be removed , since otherwise the result is false .
if one dropped this requirement , indeed , then the statement would imply that @xmath445 belongs to @xmath398 , which can not hold in general .
a remarkable counterexample is given by the case @xmath79 , where one can prove that @xmath446 does not belong to @xmath398 for any @xmath419 ( see lemma [ lem : n ] ) .
the case @xmath421 is far more awkward since no `` extension - to - zero '' result , such as lemma [ lem - ext ] , is available .
however , in the case @xmath79 , we can state an analogous for theorem [ contr_lemma ] . in order to prove it ,
it is though required a further investigation of the behavior of the integral function @xmath24 .
on the other hand , one easily checks that @xmath450 , so that it is left to prove that @xmath451_{\dot{h}^{1/2}(0,t)}<\infty$ ] ( since @xmath452 for all @xmath402 , by @xcite ) .
an easy computation shows that @xmath453_{\dot{h}^{1/2}(0,t)}^2 = 2\int_0^t\int_0^{x/2}\left|\frac{{\mathcal{n}}(x)-{\mathcal{n}}(y)}{x - y}\right|^2dy\,dx+2\int_0^t\int_{x/2}^x\left|\frac{{\mathcal{n}}(x)-{\mathcal{n}}(y)}{x - y}\right|^2dy\,dx.\ ] ] looking at the first integral and recalling that @xmath101 is increasing , we find @xmath454 hence , combining with and , @xmath455 concerning the second integral , jensen inequality yields @xmath456 & \leq \int_0^t\int_{x/2}^x\frac{1}{x - y}\int_y^x{\mathcal{i}}^2(s)ds\,dy\,dx .
\end{aligned}\ ] ] furthermore , since @xmath24 is positive and convex by lemma [ lem : conv ] , it is @xmath457 for every @xmath458 $ ] , so that @xmath459 now , noting that @xmath460 for all @xmath461 and using again , @xmath462 whereas , on the other hand , @xmath463 thus @xmath464 and , summing up , it is @xmath451_{\dot{h}^{1/2}(0,t)}<\infty$ ] , which concludes the proof . since @xmath421 ,
does not hold .
however , defining @xmath420 as in , @xmath468 and hence , arguing as in the proof of theorem [ contr_lemma ] ( with @xmath469 ) , one can check that @xmath470 then , it is left to estimate @xmath471_{\dot{h}^{1/2}(0,t)}$ ] .
first , we note that for every @xmath472 @xmath473 hence , @xmath474_{\dot{h}^{1/2}(0,t)}^2\leq & \ , 4\int_0^t\int_0^x\left|\frac{1}{x - y}\int_y^x{\mathcal{i}}(s)g(x - s)ds\right|^2dy\,dx+\nonumber\\[.3 cm ] & \ , 4\int_0^t\int_0^x\left|\int_0^y{\mathcal{i}}(s)\frac{g(x - s)-g(y - s)}{t - s}ds\right|^2dy\,dx .
\end{aligned}\ ] ] now , one can easily see that , since @xmath372 , @xmath475 & = 2\|g\|_{l^\infty(0,t)}^2[{\mathcal{n}}]_{\dot{h}^{1/2}(0,t)}^2\nonumber\\[.3 cm ] & \leq 2\|g\|_{l^\infty(0,t)}^2\|{\mathcal{n}}\|_{\dot{h}^{1/2}(0,t)}^2 \end{aligned}\ ] ] ( where @xmath476 is finite by lemma [ lem : n ] ) . on the other hand , by jensen inequality and monotonicity of @xmath101 , @xmath477 \hspace{4cm}\displaystyle \leq4{\mathcal{n}}(t)\int_0^t\int_0^x\int_0^y{\mathcal{i}}(s)\left|\frac{g(x - s)-g(y - s)}{x - y}\right|^2ds\,dy\,dx .
\end{array}\ ] ] from the fubini theorem and a change of variables @xmath478 \hspace{4cm}\displaystyle = \int_0^t{\mathcal{i}}(s)\int_0^{t - s}\int_0^x\left|\frac{g(x - s)-g(y - s)}{x - y}\right|^2dy\,dx\,ds\\[.7 cm ] \hspace{4cm}\displaystyle \leq [ g]_{\dot{h}^{1/2}(0,t)}^2{\mathcal{n}}(t ) .
\end{array}\ ] ] consequently , @xmath479 and plugging and into , @xmath480_{\dot{h}^{1/2}(0,t)}^2\leq c\cdot\max\{\|{\mathcal{n}}\|_{\dot{h}^{1/2}(0,t)}^2,{\mathcal{n}}^2(t)\}\left(\|g\|_{l^\infty(0,t)}^2+\|g\|_{\dot{h}^{1/2}(0,t)}^2\right).\ ] ] finally , combining with , follows .
it is also worth stressing that theorem [ contr_lemma_2 ] holds as well for any positive and integrable kernel @xmath0 whose integral function @xmath482 ( such as , for instance , abel kernels ) .
however , since this is a very specific assumption , we preferred to present it in the relevant case of the volterra kernel , where @xmath483 can be clearly shown , leaving to the reader further generalizations . finally , we show that a version of theorem [ contr_lemma ] holds also in @xmath484 .
this result could seem disconnected from the framework of our paper , but nevertheless it further clarifies some specific features of @xmath3 and , then , we mention it for the sake of completeness . recalling and arguing as in the proof of theorem [ contr_lemma ] , one finds that @xmath486,\ ] ] where @xmath487}(s)ds,\ ] ] with @xmath4880,t[\,,\\[.2 cm ] 0 , & \text{if } x\in{\mathbb{r}}\backslash\;]0,t[\ , , \end{array } \right . \quad\text{and}\quad
g_t(x):=\left\ { \begin{array}{ll } g(x ) , & \text{if } x\in\;]0,t[\,,\\[.2 cm ] 0 , & \text{if } x\in{\mathbb{r}}\backslash\;]0,t[\ , .
\end{array } \right.\ ] ] now , recalling that @xmath433}(s)=h(s)-h(s - x)$ ] ( @xmath434 denoting the heaviside function ) and that @xmath489 , reads @xmath490 then , by well known properties of the convolution product @xmath491 on the other hand , by , @xmath492 and thus @xmath493 consequently , since @xmath494 , arguing as before one finds .
the proof of the previous theorem stresses a relevant difference between the cases @xmath80 and @xmath395 , that arises in fact from the lack of further integrability ( of `` power type '' ) of @xmath0 .
indeed , if @xmath0 belongs only to @xmath336 the additional assumption @xmath210 clearly can not be removed in @xmath395 , whereas it is not necessary in @xmath80 .
r. adami , g. dellantonio , r. figari and a. teta , _ the cauchy problem for the schrdinger equation in dimension three with concentrated nonlinearity _
h. poincar anal .
non linaire * 20 * ( 2003 ) , no . 3 , 477500 .
r. adami , g. dellantonio , r. figari and a. teta , _ blow - up solutions for the schrdinger equation in dimension three with a concentrated nonlinearity _
poincar anal .
non linaire * 21 * ( 2004 ) , no . 1 , 121137 . a. erdlyi , w. magnus , f. oberhettinger , and f. g. tricomi , _ higher transcendental functions .
iii _ , robert e. krieger publishing co. , inc . ,
melbourne , fla . , 1981 , based on notes left by harry bateman , reprint of the 1955 original .
s. g. samko , a. a. kilbas and o. i. marichev , _ fractional integrals and derivatives _ , gordon and breach science publishers , yverdon , 1993 , theory and applications , edited and with a foreword by s. m. nikolski , translated from the 1987 russian original , revised by the authors . | for kernels @xmath0 which are positive and integrable we show that the operator @xmath1 on a finite time interval enjoys a regularizing effect when applied to hlder continuous and lebesgue functions and a `` contractive '' effect when applied to sobolev functions . for hlder continuous functions , we establish that the improvement of the regularity of the modulus of continuity is given by the integral of the kernel , namely by the factor @xmath2 . for functions in lebesgue spaces ,
we prove that an improvement always exists , and it can be expressed in terms of orlicz integrability . finally , for functions in sobolev spaces , we show that the operator @xmath3 `` shrinks '' the norm of the argument by a factor that , as in the hlder case , depends on the function @xmath4 ( whereas no regularization result can be obtained ) .
these results can be applied , for instance , to abel kernels and to the volterra function @xmath5 , the latter being relevant for instance in the analysis of the schrdinger equation with concentrated nonlinearities in @xmath6 . _
keywords : _ volterra functions , singular kernels , volterra integral equations , sonine kernels , orlicz integrability .
_ msc 2010 : _ 26a33 , 47g10 , 45e99 , 44a99 , 46e30 . |
there are various measures of complexity in the literature .
a quantitative measure of complexity is useful to estimate the ability of a variety of physical or biological systems for organization .
according to @xcite a complex world is interesting because it is highly structured .
some of the proposed measures of complexity are difficult to compute , although they are intuitively attractive , e.g. the algorithmic complexity @xcite defined as the length of the shortest possible program necessary to reproduce a given object .
the fact that a given program is indeed the shortest one , is hard to prove . in contrast
, there is a class of definitions of complexity , which can be calculated easily i.e. the _ simple _ measure of complexity @xmath0 according to shiner , davison , landsberg ( sdl ) @xcite , and the _ statistical _ measure of complexity @xmath1 , defined by l@xmath2pez - ruiz , mancini , calbet ( lmc ) @xcite . whereas the perfect measure of complexity is as yet unknown , the present work reports analysis of electron densities at atoms .
we specially refer to the shell structure ( periodicity ) , using two of the simplest of complexity measures , sdl and lmc calculated as functions of the atomic number @xmath3 .
this is a continuation of our previous work @xcite .
those measures have been criticized in the literature @xcite and a discussion is presented in section [ sec : sec4 ] .
our calculations are facilitated by our previous experience and results for the information entropy in various quantum systems ( nuclei , atoms , atomic clusters and correlated atoms in a trap - bosons ) @xcite , @xcite .
a remarkable result is the universal property for the information entropy @xmath9 where @xmath10 is the number of particles of the quantum system and @xmath11 , @xmath12 are constants dependent on the system under consideration @xcite .
in fact , if one has a physical model yielding probabilities which describe a system , then one can use them to find @xmath13 and consequently calculate the complexity of the system ( in our case the atom ) as function of @xmath3 .
this was done in @xcite , where we calculated the shannon information entropies in position - space ( @xmath14 ) and momentum - space ( @xmath15 ) and their sum @xmath16 as functions of the atomic number @xmath3 ( @xmath17 ) in atoms .
roothaan - hartree - fock electron wave functions ( rhf ) , for @xmath18 , were employed @xcite . for @xmath19
there is no electron - electron effect .
higher values of @xmath3 ( @xmath20 ) , due to the relativistic effects , are not considered .
analytic wavefunctions are available in @xcite only for @xmath3 up to 54 .
their importance lays in the fact that we can derive accurate wavefunctions in momentum space ; in order to assure that the integration is accurate . in @xcite
we calculated the sdl measure with rhf densities for @xmath21 . in the present work
we calculate lmc measure in the same region of @xmath3 , for the sake of comparison
. one could consider the case of hartree - fock wavefunctions for the extended region @xmath22 @xcite .
it turns out that both cases are satisfactory ( work in progress ) .
in @xcite and the present work , we examine if complexity of an atom is an increasing or decreasing function of @xmath3 .
this is related with the question whether physical or biological systems are able to organize themselves , without the intervention of an external factor , which is a hot subject in the community of scientists interested in complexity .
however , special attention should be paid with respect to the meaning of complexity or structure , which may depend on the system under consideration . in the case of atoms , the overall electrons through pair - wise electron - electron interaction under the external nuclear potential ,
lead to the characteristic electron probability distribution .
the selected information measure which shows up in this manner is employed to estimate complexity .
periodicity is clearly revealed here . in section [ sec : sec2 ]
we review briefly measures of information content and complexity .
section [ sec : sec3 ] contains our numerical results for the lmc measure and discussion . in section [ sec : sec4 ] we comment on the validity of sdl and lmc measures .
the class of measures of complexity considered in the present work have two main features i.e. they are easily computable and are based on previous knowledge of information entropy @xmath13 .
the shannon information entropy in position space @xmath14 is @xmath23 where @xmath24 is the electron density distribution normalized to one .
the corresponding information entropy @xmath15 in momentum space is defined as @xmath25 where @xmath26 is the momentum density distribution normalized to one .
the total information entropy is @xmath27 and it is invariant to uniform scaling of coordinates , i.e. does not depend on the units used to measure @xmath28 and @xmath29 , while the individual @xmath30 and @xmath31 do depend @xcite .
@xmath13 represents the information content of the quantum system ( in bits if the base of the logarithm is 2 or nats if the logarithm is natural ) . for a discrete probability distribution @xmath32
, one defines instead of @xmath13 , the corresponding quantities @xmath33 and @xmath34 @xmath35 and @xmath36 the uniform ( equiprobable ) probability distribution @xmath37 , gives the maximum entropy of the system .
it is noted that the value of @xmath34 can be lowered if there is a constraint on the probabilities \{@xmath38}. another measure of the information content of a quantum system is the concept of information energy @xmath39 defined by onicescu @xcite , who tried to define a finer measure of dispersion distribution than that of shannon information entropy .
onicescu s measure is discussed in @xcite . for a discrete probability distribution ( @xmath40 ) , @xmath39 is defined as @xmath41 while for a continuous one @xmath42 is defined by @xmath43 one can define a measure for information content analogous to shannon s @xmath13 by the relation @xmath44 for three dimensional spherically symmetric density distributions @xmath24 and @xmath26 , in position- and momentum - spaces respectively , one has @xmath45
@xmath46 the product @xmath47 is dimensionless and can be considered as a measure of dispersion or concentration of a quantum system . @xmath13 and @xmath39 are reciprocal .
thus we can redefine @xmath48 as @xmath49 in order to be able to compare @xmath13 and @xmath39 .
landsberg @xcite defined the order parameter @xmath50 ( or disorder @xmath51 ) as @xmath52 where @xmath13 is the information entropy ( actual ) of the system and @xmath53 the maximum entropy accessible to the system .
it is noted that @xmath54 corresponds to perfect order and predictability while @xmath55 means complete disorder and randomness . in @xcite a measure of complexity @xmath0
was defined of the form @xmath56 which is called the `` simple complexity of _ disorder _ strength @xmath57 and _ order _ strength @xmath58 '' .
one has a measure of category i if @xmath59 and @xmath60 , where complexity is an increasing function of _ disorder _ , while category ii is when @xmath60 , @xmath61 and category iii when @xmath62 and @xmath61 where complexity is an increasing function of _
order_. in category ii complexity vanishes at zero _ order _ and zero _ disorder _ and has a maximum of @xmath63 several cases for both @xmath57 and @xmath58 non - negative are shown in fig . 2 of ref .
@xcite , where @xmath0 is shown as a function of @xmath51 . in our previous work @xcite we obtained @xmath64 or @xmath65 as a function of @xmath3 and plotted the dependence of @xmath0 on the atomic number @xmath3 .
we employed @xmath66 according to rigorous inequalities holding to atoms @xcite , which hold for other systems as well ( nuclei , atomic clusters and atoms in a trap bosons ) as verified in @xcite .
these inequalities are @xmath67 the lower and the upper limits can be written , for density distributions normalized to one @xmath68 where @xmath69 is the mean square radius and @xmath70 is the kinetic energy .
we employ in ( [ eq : omega ] ) @xmath71 according to relation ( [ eq : smax ] ) . as an alternative
, we may use instead of @xmath0 the following _ statistical _ measure of complexity @xmath1 due to l@xmath2pez - ruiz , mancini and calbet @xcite defined as @xmath72 where @xmath13 denotes the information content stored in the system ( in our case the information entropy sum @xmath16 ) and @xmath73 is the disequilibrium of the system i.e. the distance from its actual state to equilibrium @xcite defined for a discrete probability distribution @xmath74 as follows @xcite @xmath75 @xmath73 is the quadratic distance of the actual probability distribution @xmath74 to equiprobability . in the continuous case , the rectangular function @xmath76 , where @xmath77 , is the natural extension of the equiprobability distribution of the discrete case .
thus the disequilibrium could be defined as @xmath78 if we redefine @xmath73 omitting the constant adding term in @xmath79 ( which is very small for large @xmath80 ) , the disequilibrium reads now @xmath81 where @xmath73 is positive for every distribution and minimal for the rectangular function which represents the equipartition . for large values of @xmath80 ( @xmath82 )
relation ( [ eq : eq24 ] ) gives @xmath83 another derivation of the formula ( [ eq : d - onedim ] ) for continuous @xmath42 can be found in section 3 of @xcite by using the rnyi generalized entropy @xmath84 where @xmath85 is an index running over all the integer values .
according to @xcite @xmath13 and @xmath73 are the two basic ingredients for calculating complexity . in our 3-dimensional case , we employ instead of ( [ eq : d - onedim ] ) the formula @xmath86 where @xmath87 , @xmath88 are defined in ( [ eq : er ] ) , ( [ eq : ek ] ) .
relation ( [ eq : d-3dim ] ) extends the definition of measure of disequilibrium of the system according to lmc , to our case , where we are interested jointly in position- and momentum - spaces .
it turns out that lmc definition of disequilibrium @xmath73 ( [ eq : d - onedim ] ) is identical to onicescu s formula ( [ eq : oni-1 ] ) for the information energy @xmath39 .
in fact , inspired by the work of l@xmath2pez - ruiz , mancini and calbet , a new interpretation of onicescu information energy may be proposed i.e. it represents the disequilibrium of the system or distance from equilibrium . additionally in our case something new is introduced ,
that is the effect of a delicate balance between conjugate spaces , reflected in the sum @xmath89 and the product @xmath90 .
both @xmath13 and @xmath73 are dimensionless .
the dependence of @xmath0 on @xmath3 for atoms has been calculated recently in @xcite . in the present letter
we calculate @xmath5 employing the same rhf wave functions in the same region @xmath91 , for the sake of comparison .
our results are shown in fig .
[ fig:1 ] and fig .
[ fig:2 ] .
we compare them with @xmath4 shown in fig . 3 of @xcite for @xmath92 .
in all ( six ) cases we observe that the measures of complexity show local minima at closed shells atoms , namely for @xmath3=10 ( ne ) , 18 ( ar ) , 36 ( kr ) .
the physical meaning of that behavior is that the electron density for those atoms is the most compact one compared to neighboring atoms .
the local maxima can be interpreted as being far from the most compact distribution ( low ionization potential systems ) .
this does not contradict common sense and satisfies our intuition .
there are also local minima for @xmath3=24 ( cr ) , 29 ( cu ) , 42 ( mo ) .
those minima are due to a specific change of the arrangement of electrons in shells .
for example , going from @xmath3=24 ( cr ) with electron configuration [ ar]@xmath93 to the next atom @xmath3=25 ( mn ) , with configuration [ ar]@xmath94 , it is seen that one electron is added in an s - orbital ( highest ) .
the situation is similar for @xmath3=29 ( cu ) and @xmath3=42 ( mo ) .
the local minimum for @xmath95 ( pd ) is due to the fact that pd has a @xmath96 electron configuration with extra stability of electron density .
it has no @xmath97 electron , unlike the neighboring atoms .
there are also fluctuations of the complexity measures within particular subshells .
this behavior can be understood in terms of screening effects within the subshell .
the question naturally arises if the values of complexity correlate with properties of atoms in the periodic table .
an example is the correlation of onicescu information content @xmath48 with the ionization potential ( fig . 4 of @xcite ) .
a more detailed / systematic study is needed , which is beyond the scope of the present report .
our calculations in @xcite show a dependence of complexity on the indices of disorder @xmath57 and order @xmath58 . in @xcite , we made a general comment that there are fluctuations of complexity around an average value and atoms can not grow in complexity as @xmath3 increases .
the second part of our comment needs to be modified .
various values of ( @xmath57,@xmath58 ) lead to different trends of @xmath4 i.e. increasing , decreasing or approximately constant .
in addition , in the present letter we compare @xmath5 with @xmath4 and we find a significant overall similarity between the curves @xmath98 and @xmath5 by plotting @xmath5 and @xmath99 in the same fig .
( [ fig:2 ] ) .
the numerical values are different but a high degree of similarity is obvious by simple inspection .
there is also the same succession of local maxima and minima at the same values of @xmath3 .
less striking similarities are observed for other values of ( @xmath100 ) as well , e.g. @xmath101 and @xmath5 .
concluding , the behavior of sdl complexity depends on the values of the parameters @xmath57 and @xmath58 .
the statistical measure lmc displays an increasing trend as @xmath3 increases .
an effort to connect the aforementioned measures , implies that lmc measure corresponds to sdl when the magnitude of disorder @xmath102 and of order @xmath103 . in other words , if one insists that sdl and lmc behave similarly as functions of @xmath3 , then we can conclude that complexity shows an overall increasing behavior with @xmath3 .
their correlation gives for atoms the strength of disorder @xmath104 and order @xmath103 .
a final comment seems appropriate : an analytical comparison of the similarity of @xmath4 and @xmath5 is not trivial . combining equations ( [ eq : omega ] ) and ( [ eq : gamma ] ) we find for the sdl measure @xmath105 while for the lmc one has @xmath106 @xmath13 and @xmath66 depend on @xmath3 as follows @xmath107 ( almost exact fitted expression ) @xcite , while @xmath108 ( a rough approximation ) .
we mention that @xmath30 , @xmath31 , @xmath87 , @xmath88 are known but different functionals of @xmath24 and @xmath26 according to relations ( [ eq : equ1 ] ) , ( [ eq : equ2 ] ) and ( [ eq : er ] ) , ( [ eq : ek ] ) respectively .
it is noted that our numerical calculations were carried out with exact values of @xmath30 , @xmath31 , @xmath87 , @xmath88 , while our fitted expressions for @xmath109 , @xmath110 are presented in order to help the reader to appreciate approximately the trend of @xmath4 and @xmath5 .
complexity is a multi - faceted and context dependent concept , difficult to be quantified .
it is extremely difficult at present to define complexity under all possible circumstances .
instead , one can choose a pragmatic approach and use as a starting point definitions of _ complexity _ found in the literature and attempt to evolve the existing framework .
thus , we have chosen the _
simple _ sdl measure and the _ statistical _ lmc one , which are relatively easy to compute .
the question naturally arises if the above measures do represent the concept of complexity as expected from semantics , intuition or other general criteria .
sdl and lmc measures have also been criticized in several aspects @xcite , some of them are the following : 1 .
all systems with the same disorder @xmath51 have the same @xmath0 i.e. the function @xmath111 is universal . 2 .
landsberg s definition of disorder @xmath64 does not describe properly and does not capture the system s structure , pattern , organization , or symmetries .
3 . the calculation in @xcite of @xmath112 for equilibrium ising systems is not clear . in our previous and present work
we use ground state electronic densities of atoms to evaluate @xmath4 and @xmath5 . at present
it is not possible to answer semantic or sophisticated questions as the ones raised in @xcite and described above .
one can not claim that sdl measure is _ the measure of complexity _ but it is a _ simple _ measure , indicating that it can be used as a starting point . to our knowledge , there are no other studies of _ complexity _ in quantum systems as functions of the number of particles
. the definition of @xmath0 is based on landsberg s disorder @xmath51 in terms of information measures @xmath13 , @xmath113 available as functions of @xmath10 . by definition @xmath50 and @xmath51
are related by @xmath114 .
an alternative approach where @xmath51 and @xmath50 should be considered independent of each other , is an open one . before considering such an extension , we calculate @xmath115 given in ( [ eq : eq19 ] ) , where @xmath13 and @xmath71 show different non - trivial dependence on @xmath10 . in this sense
the universality of @xmath111 is modified : @xmath115 depends on the quantum system under consideration , i.e. atoms @xcite , nuclei , atomic clusters e.t.c .
( work in progress ) . instead of varying the number of particles @xmath10
, one could keep it constant , and study the effect on _ complexity _ of other parameters of the system
. a welcome property of a definition of complexity supporting its validity might be the following : if one complicates the system by varying some of its parameters , and this leads to an increase of the adopted measure of _ complexity _ , then one could argue that this measure describes the complexity of the system properly
. there have been several criticisms regarding the lmc measure of complexity in the sense that it does not deserve the adjective _ statistical _ , it is not a general measure that quantifies structure and it is not an extensive quantity @xcite .
analogous arguments with sdl could be given to support the use of lmc as a first working prototype of complexity to be modified by more sophisticated models .
in @xcite various properties of the sdl measure are examined and a modification is attempted to remove its insensitivity to system differences . however in our present ( quantum ) case of atoms , by employing solely electron density distributions , it is difficult to check these ideas .
in addition , the similarity of the qualitative behavior of @xmath4 and @xmath5 for atoms used as a case study ( although they obey different definitions ) is interesting .
it shows that both measures share some common traits and correlate strongly with the periodicity of the elements in the periodic table .
an important issue is the question if @xmath6 and @xmath1 are true measures of complexity ( structure , pattern etc ) or are just functionals of electron densities .
further research is needed to clarify their region of validity and applicability as measures of complexity .
information entropy is an extensive quantity .
the question arises if @xmath0 and @xmath1 are extensive or intensive quantities , what is their thermodynamic limit , their behavior at the boundaries ( large or small order ) , their properties under scaling transformations e.t.c .
for example , it is stated in @xcite that @xmath1 is neither an intensive nor an extensive thermodynamic variable and it vanishes exponentially in the thermodynamic limit , for all one - dimensional , finite - range systems . however , in a broader context , @xmath0 and @xmath1 can serve as indicators of complexity , based on a probabilistic description . in some cases
it can be applied in a setting wider or different than thermodynamics .
such a case , is the calculation of @xmath1 for various shapes of probability distributions @xmath42 in @xcite , i.e. the rectangular , the isosceles triangle , the gaussian and the exponential probability distributions which are classified according to the corresponding values of @xmath1 .
another case is the present work , where @xmath0 and @xmath1 are estimated for the ground state electron densities , @xmath116 and @xmath117 , of a quantum system ( atom ) at zero temperature . to our knowledge
, such an application is done for the first time for a quantum system , and deserves more detailed investigation ( e.g. for excited states ) .
it is true that there are certain shortcomings of @xmath0 and @xmath1 for various cases in statistical mechanics . in quantum systems further
investigation is needed . our title `` _ _ @xmath118atoms as a testbed _ _ '' describes well our aim .
the work of k. ch .
chatzisavvas was supported by herakleitos research scholarships ( 21866 ) of @xmath119 and the european union . the work of ch . c. moustakidis was supported by pythagoras ii research project ( 80861 ) of @xmath119 and the european union | the _ simple _ measure of complexity @xmath0 of shiner , davison and landsberg ( sdl ) and the _ statistical _ one @xmath1 , according to l@xmath2pez - ruiz , mancini and calbet ( lmc ) , are compared in atoms as functions of the atomic number @xmath3 .
shell effects i.e. local minima at the closed shells atoms are observed , as well as certain qualitative trends of @xmath4 and @xmath5 . if we impose the condition that @xmath6 and @xmath1 behave similarly as functions of @xmath3 , then we can conclude that complexity increases with @xmath3 and for atoms the strength of disorder is @xmath7 and order is @xmath8 . |
this note is about two classical problems in nonparametric statistical analysis of recurrent event data , both formalised within the framework of a simple , stationary renewal process .
we first consider observation around a fixed time point , i.e. , we observe a backward recurrence time @xmath0 and a forward recurrence time @xmath1 .
it is well known that the nonparametric maximum likelihood estimator of the gap - time distribution is the cox - vardi estimator ( cox 1969 , vardi 1985 ) derived from the length - biased distribution of the gap time @xmath2 . however , winter & fldes ( 1988 ) proposed to use a product - limit estimator based on @xmath1 , with delayed entry given by @xmath0 .
keiding & gill ( 1988 ) clarified the relation of that estimator to the standard left truncation problem .
unfortunately this discussion was omitted from the published version ( keiding & gill , 1990 ) .
since these simple relationships do not seem to be on record elsewhere , we offer them here .
the second observation scheme considers a stationary renewal process observed in a finite interval where the left endpoint does not necessarily correspond to an event .
the full likelihood function is complicated , and we briefly survey possibilities for restricting attention to various partial likelihoods , in the nonparametric case again allowing the use of simple product - limit estimators .
winter & fldes ( 1988 ) studied the following estimation problem . consider @xmath3 independent renewal processes in equilibrium with underlying distribution function @xmath4 , which we shall assume absolutely continuous with density @xmath5 , minimal support interval @xmath6 , and hazard @xmath7 , @xmath8 .
the reason for our unconventional choice @xmath9 for the hazard rate belonging to @xmath4 will become apparent later . corresponding to a fixed time ,
say @xmath10 , the backward and forward recurrence times @xmath11 and @xmath12 , @xmath13 , are observed ; their sums @xmath14 are length - biased observations from @xmath4 , i.e. , their density is proportional to @xmath15 .
let @xmath16 denote a generic triple @xmath17 .
we quote the following distribution results : let @xmath18 be the expectation value corresponding to the the distribution @xmath4 , @xmath19 then the joint distribution of @xmath0 and @xmath1 has density @xmath20 , the marginal distributions of @xmath0 and @xmath1 are equal with _ density _ @xmath21 , and
the marginal distribution of @xmath22 has density @xmath23 , the length - biased density corresponding to @xmath5 .
winter and fldes considered the product - limit estimator @xmath24 where @xmath25 is the _ number at risk _ at time @xmath26 .
this estimator is the same as the kaplan - meier estimator for iid survival data @xmath27 left - truncated at @xmath28 ( kaplan & meier 1958 , andersen et al .
winter & fldes showed that @xmath29 is strongly consistent for the _ underlying _ survival function @xmath30 .
we shall show how the derivation of this estimator follows from a simple markov process model similar to the one used by keiding & gill ( 1990 ) to study the random truncation model .
first notice that the conditional distribution of @xmath22 given that @xmath31 has density @xmath32 that is , intensity ( hazard ) @xmath33 , @xmath34 , which is just the hazard @xmath35 corresponding to the underlying distribution @xmath4 left - truncated at @xmath36 .
now define corresponding to @xmath16 a stochastic process @xmath37 on @xmath38 $ ] with state space @xmath39 by @xmath40 note that it takes in successsion the values @xmath10 , @xmath41 and @xmath42 . for @xmath43 , @xmath44 where @xmath45 is the hazard rate of the marginal distribution of @xmath0 . for @xmath46 ( and hence @xmath47 ) @xmath48 by the above result on the conditional hazard of @xmath2 given @xmath0 . for @xmath43 ,
@xmath49 other transitions are impossible .
that these conditional probabilities depend on @xmath50 and @xmath26 only , but not on @xmath51 , @xmath52 , proves that @xmath37 is a markov process with intensities @xmath53 ( the marginal hazard of @xmath0 , equal to the residual mean lifetime function of the underlying distribution @xmath4 ) and @xmath54 see figure 1 .
the markov process framework of keiding & gill ( 1990 ) now indicates that ( ignoring information about @xmath4 in @xmath45 , and just focussing on the transition with rate @xmath9 ) the product limit estimator @xmath29 is a natural estimator of the survivor function @xmath30 of interest , and consistency and asymptotic normality may be obtained as shown by keiding & gill ( 1990 , sec . 5 ) .
note that the backwards intensity @xmath55 the _ backwards _ hazard - rate of a uniform distribution on a bounded interval @xmath56 , @xmath57 . since it has been assumed that @xmath0 has support interval @xmath6 , this shows that the present model _ may not _ be interpreted strictly as a left truncation model , which would require that @xmath58 was the backwards hazard rate of some probability distribution on @xmath6 .
however , this distinction is not important to our discussion .
the fact that @xmath58 does not depend on @xmath4 corresponds to winter and fldes statement that @xmath59 contains no more information than @xmath2 about @xmath4 .
this already follows from sufficiency since the joint density of @xmath59 is @xmath20 .
the likelihood function based on observation of @xmath60 is @xmath61 from which the npmle of @xmath4 is readily derived as @xmath62 that is the cox - vardi estimator in the terminology of winter and fldes ( cox 1969 , vardi 1985 ) .
it follows that the estimator @xmath29 is _ not _ npmle .
the important difference between the situation here and that of the random truncation model studied by keiding & gill ( 1990 , sec .
3 ) is that not only the intensity @xmath63 , but also @xmath64 depends only on the estimand @xmath4 .
as already mentioned , weak convergence of @xmath29 is immediate from keiding & gill ( 1990 , sec . 5 ) .
in particular , in order to achieve the extension to convergence on @xmath65 $ ] it should be required that @xmath66 in the terminology of keiding & gill ( 1990 , sec .
5c ) , and using @xmath67 and @xmath68 the integrability condition translates into @xmath69 or finiteness of @xmath70 where @xmath71 has the underlying ( `` length - unbiased '' ) interarrival time distribution @xmath4 .
it may easily be seen from gill et al .
( 1988 ) that the same condition is needed to ensure weak convergence of the cox - vardi estimator
. a variation of the observation scheme of this section would be to allow also right censoring of the @xmath12 .
this can be immediately included in the markov - process / counting process approach leading to the inefficient product - limit type estimator @xmath29 ; the delayed - entry observations @xmath12 are simultaneously right - censored .
see vardi ( 1985 , 1989 ) and asgharian et al .
( 2002 ) for treatment of the full non - parametric maximum likelihood estimator of @xmath4 , extending the cox - vardi estimator to allow right censoring .
other ad hoc estimators and the rich relationships with a number of other important non - parametric estimation problems are discussed by denby and vardi ( 1985 ) and vardi ( 1989 ) .
we consider again a stationary renewal process on the whole line and assume that we observe it in some interval @xmath72 $ ] determined independently of the process .
nonparametric estimation of the gap time distribution @xmath4 was definitively discussed by vardi ( 1982 ) in discrete time and by soon & woodroofe ( 1996 ) in continuous time .
cook & lawless ( 2007 , chapter 4 ) surveyed the general area of analysis of gap times emphasizing that the assumption of independent gap times is often unrealistic .
we shall here nevertheless work under the assumption of the simplest possible model as indicated above .
because the nonparametric maximum likelihood estimator is computationally involved it may sometimes be useful to calculate less efficient alternatives , and there are indeed such possibilities . under the observation scheme indicated above we may have the following four types of elementary observations + 1 .
times @xmath73 from one renewal to the next , contributing the density @xmath74 to the likelihood .
+ 2 . times from one renewal @xmath75 to @xmath76 , which are right - censored versions of 1 . , contributing factors of the form @xmath77 to the likelihood .
+ 3 . times from @xmath78
to the first renewal @xmath75 ( forward recurrence times ) , contributing factors of the form @xmath79 to the likelihood .
knowledge that no renewal happened in @xmath72 $ ] , actually a right - censored version of 3 . , contributing factors of the form @xmath80 to the likelihood .
+ mcclean & devine ( 1995 ) studied nonparametric maximum likelihood estimation in the conditional distribution given that there is at least one renewal in the interval , i.e. , that there are no observations of type 4 .
our interest is in basing the estimation only on complete or right - censored gap times , i.e. , observations of type 1 or 2 .
when this is possible , we have simple product - limit estimators in the one - sample situation , and we may use well - established regression models ( such as cox regression ) to account for covariates .
pea et al .
( 2001 ) assumed that observation started at a renewal ( thereby defining away observations of type 3 and 4 ) and gave a comprehensive discussion of exact and asymptotic properties of product - limit estimators with comparisons to alternatives , building in particular on results of gill ( 1980 , 1981 ) and sellke ( 1988 ) .
the crucial point here is that calendar time and time since last renewal both need to be taken into account , so the straightforward martingale approach displayed by andersen et al .
( 1993 ) is not available .
pea et al .
also studied robustness to deviations from the assumption of independent gap times .
as noted by aalen & husebye ( 1991 ) in their attractive non - technical discussion of observation patterns , observation does however often start between renewals .
( in the example of keiding et al .
( 1998 ) , auto insurance claims were considered in a fixed calendar period ) . as long as observation starts at a stopping time , inference is still valid , so by starting observation at the first renewal in the interval we can essentially refer back to pea et al .
a more formal argument could be based on the concept of the _ aalen filter _ , see andersen et al .
( 1993 , p. 164 ) .
the resulting product - limit estimators will not be fully efficient , since the information in the backward recurrence time ( types 3 and 4 ) is ignored .
it is important to realize that the validity of this way of reducing the data depends critically on the independence assumptions of the model .
keiding et al .
( 1998 ) , cf . keiding ( 2002 ) for details , used this fact to base a goodness - of - fit test on a comparison of the full nonparametric maximum likelihood estimator with the product - limit estimator .
similar terms appear in another model , called the _ laslett line segment problem _ ( laslett , 1982 ) .
suppose one has a stationary poisson process , with intensity @xmath18 , of points on the real line .
we think of the real line as a calendar time axis , and the points of the poisson process will be called _ pseudo renewal times _ or _ birth times _ of some population of individuals .
suppose the individuals have independent and identically distributed lifetimes , each one starting at the corresponding birth time .
the corresponding calendar time of the end of each lifetime can of course be called a _ death time_.
now suppose that _ all we can observe _ are the intersections of individuals lifetimes ( thought of as time segments on the time axis ) with an observational window @xmath72 $ ] . in particular , we do not know the current age of an individual who is observed alive at time @xmath78 .
again we have exactly the same four kinds of observations : + 1 .
complete _ proper _ lifetimes corresponding to births within @xmath72 $ ] for which death occurred before time @xmath76 .
censored _ proper _ lifetimes corresponding to births within @xmath72 $ ] for which death occurred after time @xmath76 .
complete _ residual _ lifetimes corresponding to births which occurred at an unknown moment before time @xmath78 , and for which death occurred after @xmath78 and before time @xmath76 .
censored _ residual _ lifetimes corresponding to births which occurred at an unknown moment before time @xmath78 , for which death occurred after time @xmath76 , and which are therefore censored at time @xmath76 .
+ the _ number _ of at least partially observed lifetimes ( proper or residual ) is random , and poisson distributed with mean equal to the intensity @xmath18 of the underlying poisson process of birth times , times the factor @xmath81 this provides a fifth , `` poisson '' , factor in the nonparametric likelihood function for parameters @xmath18 and @xmath4 , based on all the available data . maximizing over @xmath18 and @xmath4 ,
the mean of the poisson distribution is estimated by the observed number of partially observed lifetimes .
thus we find that the _ profile likelihood _ for @xmath4 , and the _ marginal likelihood _ for @xmath4 based only on contributions 1.4 .
, are proportional to one another .
nonparametric maximum likelihood estimation of @xmath4 was studied by wijers ( 1995 ) and van der laan ( 1996 ) , cf .
van der laan & gill ( 1999 ) .
some of their results , and the calculations leading to this likelihood , were surveyed by gill ( 1994 , pp . 190 ff . ) .
the nonparametric maximum likelihood estimator is consistent ; whether or not it converges in distribution as @xmath18 tends to infinity is unknown , the model has a singularity coming from the vanishing probability density of complete lifetimes just larger than the length of the observation window corresponding to births just before the start of the observation window and deaths just after its end .
van der laan showed that a mild reduction of the data by grouping or binning leads to a much better behaved nonparametric maximum likelihood estimator . if the amount of binning decreases at an appropriate rate as @xmath18 increases , this leads to an asymptotically efficient estimator of @xmath4 .
this procedure can be thought of as _ regularization _ , a procedure often needed in nonparametric inverse statistical problems , where maximum likelihood can be too greedy .
both `` unregularized '' and regularized estimators are easy to compute with the em algorithm ; and the speed of the algorithm is not so painfully slow as in other inverse problems , since this is still a problem where `` root @xmath3 '' rate estimation is possible .
the problem allows , just as we have seen in earlier sections , all the same inefficient but rapidly computable product - limit type estimators based on various marginal likelihoods .
moreover since the direction of time is basically irrelevant to the model , one can also look at the process `` backwards '' , leading to another plethora of inefficient but easy estimators .
one can even combine in a formal way the censored survival data from a forward and a backward time point of view , which comes down to counting all uncensored observations twice , all singly censored once , and discarding all doubly censored data .
( this idea was essentially suggested much earlier by r.c .
palmer and d.r .
cox , cf .
palmer(1948 ) ) .
the attractive feature of this estimator is again the ease of computation , the fact that it only discards the doubly censored data , and its symmetry under reversing time .
the asymptotic distribution theory of this estimator is of course not standard , but using the nonparametric delta method one can fairly easily give formulas for asymptotic variances and covariances . in practice
one could easily and correctly use the nonparametric bootstrap , resampling from the partially observed lifetimes , where again a resampled complete lifetime is entered twice into the estimate .
the laslett line segment problem has rather important extensions to observation of line segments ( e.g. , cracks in a rock surface ) observed through an observational window in the plane . under the assumption of a homogenous poisson line segment process
one can write down nonparametric likelihoods , maximize them with the em algorithm ; it seems that regularization may well be necessary to get optimal `` root @xmath3 '' behaviour but in principle it is clear how this might be done .
again , we have the same plethora of inefficient but easy product - limit type estimators . van zwet ( 2004 ) studied the behaviour of such estimators when the line segment process is not poisson , but merely stationary .
the idea is to use the poisson process likelihood as a quasi likelihood , i.e. , as a basis for generating estimating equations , which will be unbiased but not efficient , just as in parametric quasi - likelihood .
van zwet shows that this procedure works fine .
coming full circle , one can apply these ideas to the renewal process we first described in this section , and the other models described in earlier sections .
all of them generate stationary line segment processes observed through a finite time window on the line .
thus the nonparametric quasi - likelihood approach can be used there too . since in the renewal process case
we are ignoring the fact that the intensity of the point process of births equals the inverse mean life - time , we do not get full efficiency .
so it is disputable whether it is worth using an inefficient ad - hoc estimator which is difficult to compute when we have the options of soon and woodroofe s fully efficient ( but hard to compute ) full nonparametric maximum likelihood estimator , and the many inefficient but easy and robust product - limit type estimators of this paper .
this research was partially supported by a grant ( ro1ca54706 - 12 ) from the national cancer institute and by the danish natural sciences council grant 272 - 06 - 0442 `` point process modelling and statistical inference '' .
aalen , o.o . & husebye , e. ( 1991 ) . statistical analysis of repeated events forming renewal processes .
_ statistics in medicine _ * 10 * , 12271240 .
keiding , n. ( 2002 ) .
two nonstandard examples of the classical stratification approach to graphically assessing proportionality of hazards . in : _ goodness - of - fit tests and model validity _ ( eds . c. huber - carol , n. balakrishnan , m.s .
nikulin and m. mesbah ) .
boston , birkhuser , 301308 .
sellke , t. ( 1988 ) .
weak convergence of the aalen estimator for a censored renewal process , in _ statistical decision theory and related topics iv _ , ( eds .
s. gupta & j. berger ) , springer - verlag , new york , vol .
2 , 183194 . | nonparametric estimation of the gap time distribution in a simple renewal process may be considered a problem in survival analysis under particular sampling frames corresponding to how the renewal process is observed .
this note describes several such situations where simple product limit estimators , though inefficient , may still be useful . |
for a long time tunneling a particle through an one - dimensional time - independent potential barrier was considered in quantum mechanics as a representative of well - understood phenomena .
however , now it has been realized that this is not the case .
the inherent to quantum theory standard wave - packet analysis ( swpa ) @xcite ( see also @xcite ) , in which the study of the temporal aspects of tunneling is reduced to timing the motion of the center of `` mass '' ( cm ) of the corresponding wave packet to describe the process , does not provide a clear prescription both to interpret the scattering of finite in @xmath0 space wave packets and to introduce characteristic times for a tunneling particle .
the latter is known as the tunneling time problem ( ttp ) which has been of great interest for the last decades . as is known ( see @xcite ) , the main difficulties to arise in interpreting the wave - packet s tunneling are connected to the fact that there is no causal link between the transmitted ( or reflected ) and incident wave packets .
one of the visual consequences of this is that the average particle s kinetic energy for the transmitted , reflected and incident wave packets is different .
for example , in the case of an opaque rectangular barrier , the velocity of the cm of the transmitted wave packet is larger than that of the incident one .
it is evident that this fact needs a proper physical explanation . as was pointed out in @xcite
, it would be strange to interpret the above property of wave packets as the evidence of accelerating a particle ( in the asymptotic regions ) by the static potential barrier .
one has to point also to the well - known hartman effect @xcite related to the acceleration of the cm of the transmitted wave packet , to superluminal velocities ( see also @xcite ) . in many respects ,
the present interpretation of this property of wave packets is still controversial .
note , in the case of wide ( strictly speaking , infinite ) in @xmath0 space wave packets the average kinetic energy of particles , before and after the interaction , is the same .
however , it is evident that a causal link between the transmitted and incident wave packets does not appear in this limiting case .
perhaps , this fact is a basic reason by which many physicists appraise the phase times introduced in the swpa as ill - defined .
at least , a review @xcite devoted to the ttp seems to be the last one in which the swpa is considered in a positive context .
apart from the swpa to deal with the cm of a wave packet , in the same or different setting the tunneling problem , a variety of alternative approaches ( see reviews @xcite and references therein ) to introduce various characteristic times for a tunneling particle have also been developed . among the alternative concepts , of interest
are that of the dwell time @xcite , that of the larmor time @xcite to give the way of measuring the dwell time , and the concept of the time of arrival which is based on introducing either a suitable time operator ( see , for example , @xcite ) or the positive operator valued measure ( see review @xcite ) . besides , of importance are the studies of the temporal aspects of tunneling on the basis of the feynman , bohmian and wigner approaches to deal with the random trajectories of particles ( see , for example , @xcite and references therein ) .
one should also mention the papers @xcite where the ttp is studied beyond the framework of the standard setting the scattering problem .
we have to stress however that in the standard setting , when the initial wave packet may include a zero - momentum component , none of the alternative approaches have led to commonly accepted characteristic times ( see @xcite ) .
the recent papers @xcite , which present new versions of the dwell time ( see @xcite ) and complex tunneling times ( see @xcite ) , evidence too that up to now there are no preferable time concepts for a tunneling particle .
there is an opinion that the ttp , in the standard setting of the one - dimensional scattering problem , is ill - defined , since it does not include the measurement process .
we think that such an opinion is very questionable .
of course , in some cases the measurement process can essentially modify the original scattering one , and hence the study of its possible influence on the temporal aspects of this scattering process can be very useful ( see , for instance , @xcite and the review @xcite where this question is deeply analyzed ) . at the same time , to state that any measurement - independent setting of the ttp is ill - defined is unacceptable , in principle .
otherwise , all hamiltonians to describe a measurement - independent scattering processes ( including the motion of a free particle , and the tunneling process ) would be considered as ones having no physical sense .
we have to stress that any quantum scattering process , like classical one , proceeds , irrespective of our assistance , in some space - time framework . so that quantum theory should give a clear and unambiguous prescription to define both the spatial and temporal limits of this process .
the main question of the ttp , which implies an unique answer , is that of the ( average ) time spent by a quantum particle inside a finite barrier region .
there is a particular case when answering this question is trivial .
we have in mind tunneling a particle through the @xmath1-potential barrier .
indeed , one can _ a priori _ say that this characteristic time should be equal to zero for this potential . for the probability to find
a particle in its barrier region is equal to zero .
note , the ttp is very often ( see , for example , @xcite ) treated as the problem of introducing characteristic times for _ a wave packet _ passing through a quantum - mechanical barrier .
as is seen , from the very outset , this formulation implies timing a lengthy object whose spatial size is comparable with ( or even much more than ) the width of the potential barrier .
such a vision of the ttp is of wide spreading .
therefore it is no mere chance that a nonzero phase transmission time obtained in the swpa for the @xmath1-potential is viewed by many physicists as a fully expected result , which allegedly says about the non - locality of a quantum scattering process .
however , this result , being derived in the swpa on the basis of timing the motion of the cm of a wave packet , is _ a priori _ inconsistent : any selected point of a wave packet should cross instantaneously the point - like support of the @xmath1-potential .
as regards the non - locality of tunneling , the example of the @xmath1-potential shows explicitly that the time spent by a quantum particle in the barrier region provides insufficient information about a quantum one - particle scattering process .
it is useful also to define the time interval when the probability to find a particle crossing through the barrier region is sufficiently large .
the necessity in the additional time scale is associated eventually with the fact that the time of arrival of a particle at some point can be predicted , in quantum theory , with the error amounting to the half - width of the corresponding wave packet .
it is this characteristic time that must be derived , for a particle , with taking into account of the wave - packet s width .
this time , which can be treated as the time of the interaction of a quantum ensemble of particles with the barrier , is always greater than the time spent in the barrier region by each particle of this ensemble .
it is this quantity that must be nonzero for the @xmath1-potential .
so , due to the uncertainty in finding the position of a tunneling particle , the ( average ) time spent by the particle in the barrier region is insufficient to give a full information about its interaction with the barrier .
however , there is once more peculiarity of a quantum description of the one - dimensional scattering of a particle , which drastically complicates solving the ttp . indeed , in classical theory , in timing a scattering particle for a given initial condition
, we deal with the only trajectory of a particle , that corresponds either transmission or reflection
. however , in quantum description we deal with a wave function to include information about both the alternative possibilities .
therefore every physicist setting to the ttp has firstly to resolve the dilemma , whether he has to introduce individual ( transmission and reflection ) times or whether he must solve the ttp with no distinguishing between transmission and reflection .
one should recognize that at present this question is still open .
most of the time concepts , such as the time of arrival as well as the dwell , larmor and phase tunneling times suggest introducing individual characteristic times for transmission and reflection . as is pointed out by nussenzveig ( see @xcite ) , `` [ if some characteristic time ] does not distinguish between reflected and transmitted particles , [ this is ] usually taken as a defect '' . at the same time , nussenzveig himself believes ( ibid ) that `` [ a joint description of the whole ensemble of tunneling particles ] is actually a virtue , since transmission and reflection are inextricably intertwined ; only the characteristic times averaged over transmitted and reflected particles have a physical sense '' .
an intrigue is that there are forcible arguments in both the cases . on the one hand , quantum mechanics , as it stands , indeed provides no prescription to separate to - be - transmitted and to - be - reflected particles at the early stages of scattering .
thus , having no information about the behaviour of both the kinds of particles in the barrier region , it is impossible to find the average time spent , in this region , by particles of each kind . a knowledge about their behavior after the scattering event is insufficient for this purpose .
on the other hand , the final state of a tunneling particle evidences that tunneling consists in fact from two alternative processes - transmission and reflection .
born s formula underlying the statistical interpretation of quantum mechanics fails in this case : the average values of the particle s position and momentum calculated over the whole ensemble of particles , can not be interpreted as the expectation values of these quantities .
we consider that this fact is a poor background for introducing characteristic times averaged over transmitted an reflected particles .
in fact , the above controversy says that usual quantum mechanics does not provide both a joint and separate description of transmitted and reflected particles .
it enables one to study in detail the temporal behavior of wave packets to describe the tunneling process .
however , it gives no basis to extract from these detailed data the expectation values of the particle s position and momentum , as well as to introduce its characteristic times .
its basic tools - born s formula for calculating the expectation values , and the standard timing procedure - proved to be usefulness in studying a tunneling particle .
the main idea of this paper is that in order to learn to calculate expectation values of physical observables for a tunneling particle ( and solve the ttp , on this basis ) one needs to correct our understanding of the nature of a quantum one - particle scattering state and the correspondence principle .
now it is generally accepted that any quantum time - dependent one - particle state can be considered , in principle , as the quantum counterpart to some classical one - particle trajectory .
however , generally speaking , this is not the case . in this approach ,
all quantum one - particle scattering processes described by the schrdinger equation are divided into two classes - combined and elementary .
if the wave packet to describe a quantum one - particle scattering process represents at some time a disconnected object ( or , in other words , when the set of spatial points , where the probability to find a particle is nonzero , is disconnected ) , then we deal with a combined process . otherwise , the process is elementary .
only in the last case , born s formula and the standard timing procedure are applicable . by our approach , only an elementary time - dependent one - particle scattering state can be considered as the quantum counterpart to some classical one - particle trajectory . as regards a combined time - dependent state , it can be associated with several one - particle trajectories .
on the basis of this idea we develop a renewed wave - packet analysis in which we treat the one - particle one - dimensional scattering of a particle on a static potential barrier as a combined process consisting from two alternative ones , transmission and reflection .
we hope that this approach will be useful for a deeper understanding of the nature of quantum one - particle scattering processes and , in particular , the tunneling effect .
the paper is organized as follows . in section [ a1 ]
we pose a complete one - dimensional scattering problem for a particle .
shortcomings of the swpa are analyzed in section [ a12 ] . in section [ a2 ]
we present a renewed wave - packet analysis in which transmission and reflection are treated separately . in section [ a3 ]
we define the average ( exact and asymptotic ) transmission and reflection times and consider the cases of rectangular barriers and @xmath1-potentials . in the last section
some aspects of our approach are discussed in detail .
let us consider a particle tunneling through the time - independent potential barrier @xmath2 confined to the finite spatial interval @xmath3 $ ] @xmath4 ; @xmath5 is the barrier width .
let its in state , @xmath6 at @xmath7 be the normalized function @xmath8 to belong to the set @xmath9 consisting from infinitely differentiable functions vanishing exponentially in the limit @xmath10 .
the fourier - transform of such functions are known to belong to the set @xmath9 as well . in this case
the position and momentum operators both are well - defined . without loss of generality
we will suppose that @xmath11 here @xmath12 is the wave - packet s half - width at @xmath7 ( @xmath13 ) ; @xmath14 and @xmath15 are the operators of the particle s position and momentum , respectively .
since we study a complete scattering , an important restriction should be imposed on the rate of spreading the incident wave packet .
namely , we will suppose that the average velocity @xmath16 is large enough , so that the parts of the incident wave packet lying behind its cm , within the wave - packet s half - width , move toward the barrier together with the cm ; @xmath17 is the particle s mass .
as is known , the formal solution to the temporal one - dimensional schrdinger equation ( ose ) of the problem can be written as @xmath18 in order to solve explicitly this equation we will use here the variant ( see @xcite ) of the well - known transfer matrix method @xcite that allows one to calculate the tunneling parameters , as well as to connect the amplitudes of outgoing and corresponding incoming waves , for any system of potential barriers . let @xmath19 be the energy of a particle . then for the wave function @xmath20 to describe its stationary state in the out - of - barrier regions we have @xmath21 for @xmath22 , and @xmath23 for @xmath24 ; here @xmath25 @xmath26 should be found from the initial condition ; @xmath27 the coefficients entering this solution are connected by the transfer matrix @xmath28 : @xmath29 that can be expressed ( see @xcite ) in terms of the real tunneling parameters @xmath30 , @xmath31 and @xmath32 ; @xmath33;{\mbox{\hspace{5mm}}}p=\sqrt{\frac{r(k)}{t(k)}}\exp\left[i\left(\frac{\pi}{2}+ f(k)-ks\right)\right];\end{aligned}\ ] ] @xmath34 ( the real transmission coefficient ) and @xmath35 ( phase ) are even and odd functions , respectively ; @xmath36 ; @xmath37 ; @xmath38 .
one can easily show that for a particle impinging the barrier from the left @xmath39 we will suppose that the tunneling parameters have already been calculated ( in the case of many - barrier structures , for this purpose one can use the recurrence relations obtained in @xcite just for these real parameters . as is known , solving the ttp is reduced in the swpa to timing a particle beyond the scattering region where the exact solution of the ose approaches the corresponding in or out asymptote @xcite .
thus , definitions of characteristic times in this approach can be done in terms of the in and out asymptotes of the tunneling problem .
note , in asymptote in the one - dimensional scattering problem represents an one - packet object to converge , well before the scattering event , with the incident wave packet .
but out asymptote represents the superposition of two non - overlapped wave packets to converge , at @xmath40 with the transmitted and reflected ones .
it is easy to show that in asymptote @xmath41 and out asymptote @xmath42 can be written , for the problem at hand , as follows @xmath43;\end{aligned}\ ] ] @xmath44 @xmath45;\end{aligned}\ ] ] @xmath46\end{aligned}\ ] ] where @xmath47 for a completed scattering we have @xmath48 it is obvious that the larger is the distance @xmath49 , the more correct is the approximation for @xmath50 at @xmath7 . for the average particle s position ,
well before the scattering event , we have @xmath51 ( hereinafter , for any hermitian operator @xmath52 @xmath53 similar notations are used below for the transmitted and reflected wave packets ) .
the averaging separately over the transmitted and reflected wave packets yields @xmath54 @xmath55 ( hereinafter the prime denotes the derivative with respect to @xmath56 ) .
( [ 63 ] ) ( [ 65 ] ) yield the basis for defining the asymptotic tunneling times in the swpa .
for the following it is convenient to derive again the swpa s tunneling times .
their derivation is known to be based on the standard in quantum mechanics timing procedure which is dictated by the correspondence principle .
namely , by the analogue with classical mechanics where timing the particle s motion is reduced to the analysis of the function @xmath57 ( @xmath0 is the particle s position , @xmath58 is time ) , in quantum mechanics characteristic times for a particle should be derived from studying the temporal dependence of the expectation value of the position of a particle ( or , what is equivalent , from studying the temporal behavior of the cm of a wave packet to describe its state ) .
besides , quantum theory implies calculating the error of the timing , which should be based on the analysis of the temporal dependence of the mean - square deviation for the position operator .
the standard timing procedure is evident to imply that the average value of the particle s position has its primary physical sense ( as the most probable position of a particle ) at all stages of its motion .
for instance , for a free particle whose state is described by a gaussian - like wave packet , this requirement is fulfilled and , as a consequence , no problem arises in timing its motion .
however , an essentially different situation arises in the case of a tunneling particle .
now , following the cm of a wave packet to describe the whole ensemble of tunneling particles becomes meaningless at some stages of scattering .
in particular , after the scattering event , when we deal with two scattered ( transmitted and reflected ) wave packets , the averaging over the whole ensemble of particles has no physical sense . by this reason
the above timing procedure can not be applied in this case . of course , at late times one can attempt to define individual average positions of transmitted and reflected particles
. however , in timing , this implies a separate description of both the subensembles at the first stage of scattering , what is widely accepted to be impossible in conventional quantum mechanics . as a result , it is not clear how to apply the above timing procedure both to the whole ensemble and to its parts , transmitted and reflected particles .
this question remains open in the swpa . in this section , following this approach , we will simply take the incident wave packet as the counterpart to the transmitted ( reflected ) one at the initial stage of scattering .
so , let @xmath59 be the spatial point to lie at some distance @xmath60 ( @xmath61 and @xmath62 ) from the left boundary of the barrier , and @xmath63 be the point to lie at some distance @xmath64 ( @xmath65 ) from its right boundary . following @xcite ,
let us define the difference between the times of arrival of the cms of the incident and transmitted packets at the points @xmath59 and @xmath63 , respectively ( this time will be called below as the `` transmission time '' ) .
analogously , let the `` reflection time '' be the difference between the times of arrival of the cms of the incident and reflected packets at the same point @xmath59 . thus , let @xmath66 and @xmath67 be such instants of time that @xmath68 then , considering ( [ 63 ] ) and ( [ 64 ] ) , one can write the `` transmission time '' @xmath69 ( @xmath70 ) for the given interval in the form @xmath71.\end{aligned}\ ] ] similarly , for the reflected packet , let @xmath72 and @xmath73 be such instants of time that @xmath74 from equations ( [ 63 ] ) , ( [ 65 ] ) and ( [ 10 ] ) it follows that the `` reflection time '' @xmath75 ( @xmath76 ) can be written as @xmath77.\end{aligned}\ ] ] note , the average values of @xmath56 for all three wave packets coincide only in the limit @xmath78 ( i.e. , for particles with a well - defined momentum ) . in the general case these quantities are distinguished .
for example , for a particle whose initial state is described by the gaussian wave packet , when @xmath79 we have @xmath80 @xmath81 let @xmath82 then relations ( [ 100 ] ) and ( [ 101 ] ) can be written in the form ( note that @xmath83 ) @xmath84 as is seen , quantities ( [ 9 ] ) and ( [ 11 ] ) can not serve as characteristic times for a particle . due to the last terms in these expressions
the above times depend on the initial distance between the wave packet and barrier , with @xmath60 being fixed .
these terms are dominant for the sufficiently large distance @xmath49 .
moreover , one of them must be negative .
for example , for the transmitted wave packet it takes place in the case of the under - barrier tunneling through an opaque rectangular barrier , when the difference @xmath85 is sufficiently large .
the numerical modeling of tunneling @xcite shows in this case a premature appearance of the cm of the transmitted packet behind the barrier , what just points to the lack of a causal link between the transmitted and incident wave packets ( see @xcite ) .
as was shown in @xcite , this effect disappears in the limiting case @xmath78 .
for example , in the case of gaussian wave packets the fact that the last terms in ( [ 9 ] ) and ( [ 11 ] ) tend to zero when @xmath86 , with the ratio @xmath87 being fixed , can be proved with help of exps .
( [ 100 ] ) and ( [ 101 ] ) ( note that the limit @xmath86 with a fixed value of @xmath49 is unacceptable in this analysis , because it contradicts the initial condition @xmath88 for a completed scattering ) . however , one has to stress that in the limit @xmath78 the incident and transmitted ( reflected ) wave packets are causally disconnected , as in the general case .
as before , the former can not be considered as the counterpart to the transmitted ( or reflected ) wave packet at the initial stage of scattering .
the inconsistency of introducing the phase times in the swpa is seen explicitly in the particular case of the @xmath1-potential .
as is known , the phase transmission time is nonzero in this case .
this result is usually treated as the evidence of a non - local character of tunneling a particle . however , one has to bear in mind the fact that in the swpa this nonzero time describes the motion of the wave - packet s cm .
thus , this result is clearly erroneous , since any point of a moving wave packet should cross instantly the spatial region of a zero width . at the same time we have to stress that , although exps .
( [ 9 ] ) and ( [ 11 ] ) can not be applied to a particle , they notwithstanding correctly describe the relative motion of the transmitted ( or reflected ) and incident wave packets .
thus , the main problem is that quantum mechanics , as it stands , does not provide a clear prescription to interpret properly the behaviour of these packets as applied to a particle . to clarify
this question is the main goal of our study .
as is seen from the previous section , modeling the tunneling process in terms of wave packets meets a paradoxical situation . on the one hand , since this process is one - particle , the corresponding time - dependent state of a quantum particle is supposed to be the quantum analogue of a classical one - particle trajectory . on the other hand ,
the average values of the particle s position and momentum calculated for this one - particle state , by born s formula , can not be interpreted as the _ expectation _ values of these quantities . however , as is well - known , this formula is an integral part of born s statistical interpretation of a wave function .
therefore the above controversy means that the wave function to describe the one - particle tunneling process can not be associated with the motion of a single quantum particle .
this is seen also from the fact that at late times this function represents the superposition of the transmitted and reflected wave packets .
the naive statistical interpretation of this wave function ( as that to describe the motion of a single particle ) , which ignores the above controversy , leads to the non - local pseudo - effect : transmitted and reflected particles of the ensemble are inextricably intertwined even after the scattering event .
of course , the above fact that born s formula is not applicable to this case says that this `` non - locality '' as well as the above interpretation of wave function have no physical sense .
the above situation is similar to that to arise in the well - known two - slits experiment where the beam of mutually non - interacting , identically prepared particles diffracts on an opaque screen with two slits . as is known , a wave packet to describe this ( one - particle ) process passes , like usual waves , through both the slits . as in the above case , the naive statistical interpretation of this wave function ( as that to describe the stochastic motion of a single particle ) `` permits '' the particle to pass simultaneously through both the slits ; for both the events are inextricably intertwined in this interpretation .
that is , again we arrive at the pseudo - effect of non - locality .
so , scattering a wave packet on the potential barrier as well as its scattering on the slits in an opaque screen , each can not be associated unambiguously with the motion of a single quantum particle .
this is true also for any one - particle quantum scattering process which involves either several sinks or several sources of particles , or several possible channels of motion . in all non - relativistic scattering processes ,
a quantum particle is evident to be an indivisible object . and ,
though we can say nothing , before measuring , about the position of the particle , we can say _ a priori _ that this object can be emitted only by one source of particles , can move only along one scattering channel and can be absorbed only by one sink of particles .
this property should be considered as a distinctive of the motion of a single quantum particle .
it is evident that the behaviour of wave packets in the case of the tunneling process and above diffraction on slits does not guarantee the fulfillment of this property . by our approach ,
all quantum one - particle scattering processes and the corresponding time - dependent states fall into two classes - elementary and combined .
namely , if the above property is fulfilled , we deal with elementary ones .
otherwise , a quantum one - particle scattering process and the corresponding state are combined .
unlike an elementary process , a combined one implies several alternative scattering channels for a particle .
the motion of a particle along each channel should be described by the corresponding wave function to obey the schrdinger equation and the corresponding boundary conditions to distinguish this channel from others .
the wave function to describe a combined process is the sum of those to describe all elementary ones involved in the former .
note , since all elementary processes involved in the same combined one are mutually exclusive , a particle can take part only in one of them .
this means that the whole quantum ensemble of particles involved in the combined scattering process can be uniquely decomposed into the subensembles of particles taking part only in one of the elementary processes : the sum of norms of wave functions to describe the elementary processes is equal to unit .
the number of particles in each sumensemble is evident to be constant in time .
we have to stress that only elementary one - particle states can be viewed as the quantum analogue of classical ones . to calculate the expectation values of the position and momentum of a particle
is meaningful only for elementary processes . in this case one
can calculate the trajectory of the wave - packet s cm .
it can be used then for timing a quantum particle . as regards a combined process
, making use of born s formula has no physical sense in this case .
such a process can be associated with several cm s trajectories , rather than with one .
the number of these trajectories is equal to that of elementary processes involved in the combined one .
so , by our approach tunneling a non - relativistic quantum particle through an one - dimensional static potential barrier is a combined process consisting from two alternative elementary ones , transmission and reflection .
our next step is to find two solutions to the schrdinger equations to describe separately transmission and reflection .
the wave function for transmission will be named further as the transmission wave function ( twf ) , and that for reflection will be named as the reflection wave function ( rwf ) .
the main thing which should be taken into account in finding these solutions is that the rwf should describe only reflected particles , and the twf does only transmitted particles . in both the cases
, stationary solutions should contain one incoming and one outgoing wave . in this paper
we show that such solutions do exist .
so , let @xmath89 and @xmath90 be the searched - for twf and rwf , respectively . in line with subsection [ a21 ] ,
their sum represents the wave function to describe , in the problem at hand , the combined state of the whole ensemble of particles .
hence , from the mathematical point of view our task now is to find such solutions @xmath89 and @xmath90 to the schrdinger equation that for any @xmath58 , @xmath91 where @xmath50 is the full wave function to describe all particles ( see section [ a1 ] ) . in the limit @xmath92 @xmath93 where @xmath94 and @xmath95 are the transmitted and reflected wave packets whose fourier - transforms presented in ( [ 61 ] ) and ( [ 62 ] ) . as is known , searching for the wave functions in the case of the time - independent potential @xmath2 is reduced to the solution of the corresponding stationary schrdinger equation . for a given @xmath56 ,
let us find firstly the functions @xmath96 and @xmath97 for the spatial region @xmath22 . in this region
let @xmath98 @xmath99 where @xmath100 , @xmath101 .
since the rwf describes the state of reflected particles only , the probability flux for @xmath96 should be equal to zero , i.e. , latexmath:[\ ] ] @xmath202 if @xmath203 ; otherwise @xmath204 here the function @xmath205 coincides , for a given @xmath56 , with one of the functions , @xmath206 or @xmath207 , for which @xmath96 is an odd function ( see above ) .
one can easily show that for both the roots @xmath208 a simple analysis in the @xmath56-representation shows that well before the scattering event the average kinetic energy of particles in both subensembles ( with the average wave numbers @xmath209 and @xmath210 ) is equal to that for large times : @xmath211 besides , at early times @xmath212 @xmath213 as it follows from exps .
( [ 73 ] ) and ( [ 74 ] ) , the average starting points @xmath214 and @xmath215 , for the subensembles of transmitted and reflected particles , respectively , differ from that for all particles : @xmath216 the implicit assumption made in the swpa that incident , as well as transmitted and reflected particles start , on the average , from the same point does not agree with this result . of great importance here
is that @xmath214 and @xmath215 are the initial values of @xmath217 and @xmath218 , respectively , which have the status of expectation values of the particle s position .
they behave causally in time .
as regards the average starting point for the whole ensemble of particles , its coordinate is the initial value of @xmath219 which behaves non - causally ; for this quantity is not an expectation value of the particle s position ( see also the last section of this paper ) . by this reason
the ( asymptotic ) phase times obtained in the swpa should be considered as ill - defined quantities , for any wave packets .
let us take into account exps .
( [ 73 ] ) , ( [ 74 ] ) and again analyze the motion of a particle in the above spatial interval covering the barrier region .
in particular , let us calculate the transmission time , @xmath220 , spent ( on the average ) by a particle in the interval @xmath221 $ ] .
it is evident that the above equations for the arrival times @xmath178 and @xmath179 , which correspond to the extreme points @xmath59 and @xmath63 , respectively , read now as @xmath222 considering ( [ 73 ] ) and ( [ 64 ] ) , we obtain from here that now the transmission time is @xmath223 similarly , for the reflection time @xmath224 ( @xmath225 ) , we have @xmath226 considering ( [ 74 ] ) and ( [ 65 ] ) , one can easily show that @xmath227 the inputs @xmath228 ( @xmath229 ) and @xmath230 ( @xmath229 ) will be named below as the asymptotic transmission and reflection times for the barrier region , respectively : @xmath231 @xmath232 here the word `` asymptotic '' points to the fact that these quantities were obtained with making use of the corresponding in and out asymptotes .
unlike the exact tunneling times the asymptotic times can be negative by value .
the corresponding lengths @xmath233 and @xmath234 @xmath235 can be treated as the effective barrier s widths for transmission and reflection , respectively .
let us consider the case of a rectangular barrier ( or well ) of height @xmath147 and obtain explicit expressions for @xmath236 ( now , both for transmission and reflection , @xmath237 since @xmath238 ) which can be treated as the effective width of the barrier for a particle with a given @xmath56 .
besides , we will obtain the corresponding expressions for the expectation value , @xmath239 , of the staring point for this particle : @xmath240 .
it is evident that in terms of @xmath241 the above asymptotic times for a particle with the well - defined momentum @xmath242 read as @xmath243 using the expressions for the real tunneling parameters @xmath31 and @xmath30 ( see @xcite ) , one can show that , for the below - barrier case ( @xmath244 ) , @xmath245 \left[\kappa_0 ^ 2\sinh(\kappa d)-k^2 \kappa d\right ] } { 4k^2\kappa^2 + \kappa_0 ^ 4\sinh^2(\kappa d)}\ ] ] @xmath246 where @xmath247 for the above - barrier case ( @xmath248 @xmath249\left[k^2 \kappa d-\beta \kappa_0 ^ 2\sin(\kappa d)\right ] } { 4k^2\kappa^2+\kappa_0 ^ 4\sin^2(\kappa d)}\ ] ] @xmath250 where @xmath251 @xmath252 if @xmath253 , otherwise , @xmath254 . in both the cases @xmath255 .
it is important to stress that @xmath256 and @xmath257 , in the limit @xmath258 .
this property guarantees that for infinitely narrow in @xmath0-space wave packets the average starting points for both subensembles will coincide with that for all particles .
note , for wells , the values of @xmath241 and , as a consequence , the corresponding asymptotic tunneling times are negative , in the limit @xmath259 , when @xmath260 .
note that for sufficiently narrow barriers and wells , namely when @xmath261 , we have @xmath262 . for the starting point
we have @xmath263 for @xmath244 and @xmath264 , respectively . for wide barriers and wells , when @xmath265 , we have @xmath266 and @xmath267 , for @xmath244 ; and @xmath268 for @xmath154 .
it is important that for the @xmath1-potential , @xmath269 we have @xmath270 .
that is , like the dwell and larmor times , @xmath271 in this case .
thus , though the ensemble of identically prepared particles spends nonzero time to pass through this potential , each quantum particle of the ensemble spends no time in its barrier region .
while the quantum ensemble of particles interacts with the @xmath1-potential , there is a nonzero probability to find a particle near this barrier .
note , unlike the first derivative of @xmath50 with respect to @xmath0 , that of @xmath182 has equal values in the limits @xmath272 .
the average force calculated for a particle in the state @xmath182 is zero , for this potential .
that is , @xmath182 describes that part of the incident wave packet , which does not experience the action of the @xmath1-potential . transmitting particles start , on the average , from the point @xmath273 @xmath274 and moves then freely .
the main idea underlying this paper is that tunneling a particle through an one - dimensional static potential barrier is a combined stochastic process consisting from two alternative elementary ones - transmission and reflection .
we showed that the wave function to describe the tunneling process can be uniquely decomposed into two solutions of the schrdinger equation for the given potential , which describe separately transmission and reflection .
we found both the solutions in the case of symmetric potential barriers and introduced , according to the standard timing procedure , the average ( exact and asymptotic ) transmission and reflection times . by our approach , in the most cases ,
quantum one - particle scattering processes are just combined .
each of them represents a complex stochastic process consisting from several alternative elementary ones .
the decomposition of a combined process into elementary ones can be performed uniquely .
accordingly , the wave function to describe the combined process can be uniquely presented as a sum of those to describe all the elementary ones .
the main peculiarity of combined states is that the averaging over such states of the particle s position and momentum , with help of born s formula intended for calculating the expectation values of physical observables , does not give in reality expectation values of these quantities .
both the average values behave non - causally in time .
strictly speaking , in the case of combined states , the particle s position and momentum ( though their operators are hermitian ) lost their primary status of physical observables ! as a result , timing a particle in such states is meaningless too . only for elementary quantum processes and states
born s formula and the timing procedure are valid . in other words , only for elementary states
the particle s position and momentum ( and other physical quantities with linear hermitian operators ) have their primary status of observables , and , as consequence , there is no problem to time the motion of a particle being in such states .
as is shown , the peculiarity of the wave functions for transmission and reflection is that each of them contains only one incident and only one scattered wave packets . at the same time it is evident
that if a particle was prepared in the combined state @xmath176 , then this packet would be divided by the barrier into two parts .
of course , in the case considered the initial combined state of a particle is @xmath168 but not @xmath176 .
nevertheless , we have to clear up the principal difference taking place between a wave function to describe a combined process and those to describe elementary ones involved in the former .
let , for the given potential , in addition to the problem at hand where the amplitudes of incoming and outgoing waves are , respectively , @xmath275 we have two auxiliary scattering problems with amplitudes @xmath276 and @xmath277 ( the transfer matrix ( [ 50 ] ) is evident to be the same for all three problems )
. note that in the first auxiliary problem the only outgoing wave coincides with the reflected wave arising in ( [ 800 ] ) .
and , in the second one , the only outgoing wave coincides with the transmitted wave in ( [ 800 ] ) .
it is evident that the sum of these two functions results just in that to describe the state of a particle in the original tunneling problem .
as is seen , the main peculiarity of the superposition of these two probability fields is that due to interference their incoming waves in the region @xmath24 fully annihilate each other ( note that in the corresponding reverse motion they are outgoing waves ) .
the corresponding flux of particles is reoriented into the region @xmath278 .
thus , the initial probability fields ( [ 801 ] ) and ( [ 802 ] ) associated with the transmitted and reflected wave packets are radically modified under the superposition . in this case , the wave packet connected causally to the transmitted ( reflected ) one is just @xmath182 ( @xmath192 ) .
thus , we see that the sum of wave functions ( [ 801 ] ) and ( [ 802 ] ) can be presented as that of the stationary - states rwf and twf . as a result of reorienting the probability fields , the squared amplitude of the incoming wave ( in the region @xmath278 ) associated with reflection increases due to interference from the initial value @xmath279 ( @xmath280 ) ( see ( [ 801 ] ) ) to @xmath281 ( @xmath282 ( in the rwf ) . in the case of transmission , the corresponding quantity increases from the initial value @xmath283 ( @xmath284 ) ( see ( [ 802 ] ) ) to @xmath285 ( @xmath286 ( in the twf ) . as is seen , in contrast to probability fields ( [ 801 ] ) and ( [ 802 ] ) , @xmath182 and @xmath192 should be considered as an inseparable pair : they can not evolve separately .
of course , in this case one would doubt the reality of these fields . indeed , they can can not be observed separately . and , besides , being involved in the combined state , they can not be directly observed , at early stages of scattering , because of interference .
however , for any combined process , namely the interference between wave fields to describe elementary sub - processes provides all needed information to justify their existence .
indeed , let @xmath287 be the result of measuring @xmath157 . then , using the distributions @xmath158 and @xmath159 calculated beforehand , we can extract , from the experiment data , the difference @xmath288 to describe interference between the wave fields @xmath182 and @xmath192 .
by our approach , it must have two important properties : 1 ) the integral of this difference over the region @xmath289 must be zero ; 2 ) this difference must be nonzero only for the first stages of scattering , and only for @xmath290 .
the first property means that the whole ensemble of particles , in this scattering problem , can be indeed divided into two subensembles described by the distributions @xmath158 and @xmath159 .
the second property means that one of them is indeed connected causally to the transmitted wave packet , and another evolves causally into the reflected one .
this means , in turn , that the above decomposition is unique .
note , this property is inherent only to combined processes .
the elementary states @xmath182 and @xmath192 , themselves represent decompositions into the orthogonal stationary states .
however , the latter can not be treated as elementary states . for
the interference between them does not have the above two properties .
they can not be separated , in principle .
so , the wave functions @xmath182 and @xmath192 describe two real processes to proceed simultaneously . taking into account the fact that a wave function describes the beam ( or , ensemble ) of identically prepared particles , rather than a single quantum particle , one can interpret the found decomposition of @xmath50 as follows .
namely , @xmath182 describes that part of the beam of mutually non - interacted particles prepared in the state @xmath168 , which is transmitted through the barrier .
similarly , @xmath192 does the reflected part of this beam . at early times
these parts of the beam move in the same spatial region . at this stage of scattering
, the study of the motion of both parts is reduced to the analysis of the interference between them . at all stages
they evolve irrespective of each other , not `` seeing '' their own counterparts ; just @xmath182 ( @xmath192 ) is causally connected to the transmitted ( reflected ) wave packet considered separately , but not @xmath50 . as well as in the superposition of free moving wave packets they do not destroy each other ( after their meeting in some spatial region they move unaltered ) , in the superposition of the modified wave fields @xmath182 and @xmath192 they do not influence each other , too .
it is not surprising that particles of both parts start , on the average , from the spatial points to differ from the average starting point calculated for the whole beam of particles .
firstly , @xmath291 due to interference ; here @xmath292 and @xmath293 are the norms of @xmath182 and @xmath182 , respectively . and , what is more important , among these three average quantities , only @xmath217 and @xmath217 have the physical meaning of the expectation values of the particle s position .
the behaviour of @xmath219 , being averaged over two alternative processes , is not causal .
it can not be interpreted as the expectation value of the particle s position .
the main point of our research is that any combined state represents a superposition of elementary states which are distinguishable . as a consequence , by our approach , the experimental study of the probability density for a particle taking part in a combined process means , in fact ,
the observation of the interference between the elementary states involved in the combined process . however , maxima and minima of the interference pattern behave non - causally . or ,
more correctly , only when we know all information about each elementary state , which just behaves causally , we can unambiguously interpret the behaviour of the interference pattern .
all this takes place , in particular , in the case of tunneling . by this approach ,
the non - causal behaviour of a tunneling wave packet ( which have been pointed out by @xcite ) is explained by the fact that tunneling is a combined process .
an exhaustive explanation of this quantum effect can be achieved only in the framework of a separate description of transmission and reflection . for only these processes
are elementary , and , as a consequence , namely they ( and the corresponding probability densities ) behave causally .
a simple analysis shows that the definitions of the asymptotic tunneling times given in our approach differ essentially from their analogs known in the literature . at this point
we have to note once more that a correct timing of transmitted and reflected particles implies the availability of a complete information about these subensembles of particles at all stages of scattering . making use , in the alternative approaches , of the incident wave packet as
the counterpart to the transmitted ( or reflected ) wave packet at the first stage of scattering is clearly an inconsistent step .
for the former does not connected causally to the transmitted ( or reflected ) one ( see also @xcite ) .
just the wave functions for transmission and reflection found in our approach provides all needed information .
thus , we think that our definitions of tunneling times have a more solid basis .
of course , a final decision in the long - lived controversy in solving the ttp should be made by a reliable experiment . in this connection
, we have to note that the main ideas of such approaches as @xcite , @xcite and others whose formalism involves the peculiarities of the measurement process , may be very useful in the following study of the tunneling and other scattering processes treated as combined ones . indeed ,
the fact that our definitions of the tunneling times do not coincide , for example , with those obtained in @xcite and @xcite does not mean at all that the main ideas underling our and these approaches contradict each other .
rather they are mutually complementary .
we think that namely in combination all these ideas will be useful in studying quantum scattering processes .
so , it would be very useful to define the larmor time and time - of - arrival distribution on the basis of wave functions for transmission and reflection .
it is evident that the influence of an external magnetic field ( or absorbing potential ) on transmitted and reflected particles should be different .
hence , the study of the interference between transmission and reflection , at the first stages of scattering , might permit us to check both our idea of separating these elementary processes and ways @xcite of introducing characteristic times , which differ from the standard timing procedure .
for the first case , for this purpose , the same magnetic field ( or absorbing potential ) might be localized in two equivalent spatial regions lying on the same distance from the midpoint of a symmetric potential barrier . in this case , the symmetry of the original potential remains unaltered , and there is no principal problem to find the wave functions for transmission and reflection .
as regards further development of our approach , it can be applied , in principle , to any potential localized in the finite spatial region . in one dimension
, it is applicable to the potential steps and asymmetric potential barriers .
no principal difficulties should arise also in separating transmission and reflection in the case of quasi - one - dimensional structures , when the potential energy of a particle depends only on one coordinate .
as regards the scattering problem with two slits in the opaque screen , it seems to involve four elementary processes , transmission and reflection for the first and second slits .
besides , scattering a particle on a point - like obstacle , with a spherically symmetrical potential , seems to involve two alternative elementary processes . in this case
there is a plain to separate both the processes .
this plain must be parallel to the vectors @xmath294 and @xmath295 $ ] and pass through the obstacle ; here @xmath296 and @xmath294 are the average position and momentum of a particle calculated for its initial combined state .
of course , in the general case the problem of decomposing some combined process into alternative elementary ones may be technically complicated .
this task should be considered , in every case , separately .
861 hauge e h and stvneng j a 1989 917 wigner e p 1955 145 hartman t e 1962 3427 hauge e h , falck j p and fjeldly t a 1987 4203 teranishi n , kriman a m and ferry d k 1987 _ superlatt . and microstrs . _ * 3 * , 509 collins s , lowe d and barker j r 1987 6213 buttiker m and landauer r 1982 1739 muga
j g , egusquiza i l , damborenea j a , delgado f 2002 042115 - 1 landauer r and martin th 1994 217 olkhovsky v s and recami e 1992 _ phys .
repts . _ * 214 * 339 steinberg a m 1995 2405 muga j g , leavens c r 2000 _ phys
_ * 338 * 353 nussenzveig h m 2002 _ phys .
repts . _ * 364 * 83 smith f t 1960 349 jaworski w and wardlaw d m 1988 2843 leavens c r and aers g c 1989 1202 nussenzveig h m 2000 042107 baz a i 1966 _ yad .
* 4 * 252 rybachenko v f 1966 _ yad .
fiz . _ * 5 * 895 buttiker m 1983 6178 buttiker m 2002 in _ time in quantum mechanics , ( lecture notes in physics vol m72 ) _ ed j g muga , r s mayato and
i l egusquiza ( berlin : springer ) 256 liang z j q , nie y h , liang j j and liang j q 2003 6563 aharonov y , bohm d 1961 1649 brouard s , sala r and muga j g 1994 4312 hahne g e 2003 7149 noh j w , fougeres a and mandel l 1991 1426 hegerfeldt g c , seidel d and muga j g 2003 022111 sokolovski d and baskin l m 1987 4604 kreidl s , grubl g and embacher h g 2003 8851 yamada n 2000 3350 garcia - calderon g , villavicencio j and yamada n 2003 052106 grubl g and rheinberger k 2002 2907 garcia - calderon g and villavicencio j 2001 012107 garcia - calderon g , villavicencio j , delgado f and muga j g 2002 042119 delgado f , muga j g , ruschhaupt a , garcia - calderon g and villavicencio j 2003 032101 goto m , iwamoto h , aquino v m , aquilera - navarro v c and kobe d h 2004 3599 wang z s , kwek l c , lai c h and oh c h 2004 052108 hegerfeldt g c , seidel d , muga j g , and navarro b 2004 012110 olkhovsky v s , petrillo v , and zaichenko1 a k 2004 034103 chuprikov n
l 1992 _ sov .
semicond . _
* 26 * 2040 merzbacher e 1970 _ quantum mechanics _
( john wiley & sons , inc .
new york )
taylor j r 1972 _ scattering theory : the quantum theory on nonrelativistic collisions _
( john wiley & sons , inc .
new york - london - sydney ) chuprikov n
l 1997 _ semicond . _ * 31 * 427
.... \figure{\label{fig1}the $ x$-dependence of $ |\psi_{full}(x , t)|^2 $ ( dashed line ) which represents the gaussian wave packet with $ l_0=7.5 nm$ and the average kinetic particle 's energy $ 0.25 ev$ , as well as $ |\psi_{tr}(x , t)|^2 $ ( open circles ) and $ |\psi_{ref}(x , t)|^2 $ ( solid line ) for the rectangular barrier ( $ v_0=0.3 ev$ , $ a=500 nm$ , $ b=505 nm$ ) ; $ t=0$. } .... .... \figure{\label{fig4}the $ x$-dependence of $ |\psi_{full}(x , t)|^2 $ ( dashed line ) which represents the gaussian wave packet with $ l_0=7.5 nm$ and the average kinetic particle 's energy $ 0.25 ev$ , as well as $ |\psi_{tr}(x , t)|^2 $ ( open circles ) and $ |\psi_{ref}(x , t)|^2 $ ( solid line ) for the rectangular well ( $ v_0=-0.3 ev$ , $ a=500 nm$ , $ b=505 nm$ ) ; $ t=0$. } .... | we present a renewed wave - packet analysis based on the following ideas : if a quantum one - particle scattering process and the corresponding state are described by an indivisible wave packet to move as a whole at all stages of scattering , then they are elementary ; otherwise , they are combined ; each combined process consists from several alternative elementary ones to proceed simultaneously ; the corresponding ( normed ) state can be uniquely presented as the sum of elementary ones whose ( constant ) norms give unit , in sum ; born s formula intended for calculating the _ expectation _ values of physical observables , as well as the standard timing procedure are valid only for elementary states and processes ; only an elementary time - dependent state can be considered as the quantum counterpart to some classical one - particle trajectory . by our approach , tunneling a non - relativistic particle through a static one - dimensional potential barrier is a combined process consisting from two elementary ones , transmission and reflection . in the standard setting of the problem , we find an unique pair of solutions to the schrdinger equation , which describe separately transmission and reflection . on this basis
we introduce ( exact and asymptotic ) characteristic times for transmission and reflection . |
the magnetic field in the solar corona and the inner solar wind is very important because it controls the structure and dynamics of the coronal plasma .
unfortunately it is very difficult to measure the magnetic field in this region directly .
optical observations can be made in the photosphere , and these can be extrapolated outwards into the corona @xcite , but such extrapolations do not agree well with estimates made by extrapolating inwards space - craft measurements at mercury @xcite .
it is possible to estimate the magnetic field in the quiet solar wind by measuring the faraday rotation of linearly polarized radio waves propagating through the region of interest .
thus a good deal of work has been done using the faraday rotation technique ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the faraday rotation of the position angle of the linear polarization ( in cgs units ) is given by @xmath1 .\ ] ] the magnetic field of the inner corona is dominated by closed loops over active regions and open field lines over coronal holes @xcite .
closed loops seldom extend past 2 - 3 @xmath0 and the magnetic field , as defined by bright striations in white light observations , appears to be radial outside of this distance @xcite . at larger distances a tangential component builds up due to rotation of the sun ( hundhausen , 1972 ) , but in the regions of interest for faraday rotation observations
the magnetic field is predominantly radial .
this has an important consequence for faraday rotation observations .
if the corona were spherically symmetric the sign of @xmath2 would reverse at the closest point of approach and the faraday rotation would approach zero .
so one can not use the simple model of a spherically symmetric solar wind - faraday rotation is a measure of the deviation from spherical symmetry .
a further problem with interpreting faraday rotation observations is that the electron density must be known .
fortunately the electron density can be measured by several methods and its average behavior is reasonably well understood ( e.g. * ? ? ?
* ; * ? ? ?
. however it is time variable , and the path integration must be modeled carefully to obtain reliable estimates of the magnetic field . with a pulsar
one can also measure the group delay of the pulse passing through the solar wind .
this is given by @xmath3.\ ] ] we will show that this constraint is very valuable because it allows us to estimate the ratio ( f ) of the instantaneous electron density @xmath4 to the average @xmath5 . in our observations
this factor lay in the range @xmath6 .
correction of the magnetic field estimate by this factor is a significant improvement .
however the group delay due to the solar wind is small and can not be measured with sufficient accuracy for most pulsars .
fortunately the class of pulsars with millisecond periods ( msps ) has very high rotational stability and can be timed with a precision of the order of 100 ns ( e.g. * ? ? ?
this is sufficient to measure the solar wind contribution @xcite .
simultaneous measurements of faraday rotation and group delay can also be performed with suitably equipped spacecraft and such measurements have been very useful , but they not been possible since the helios probes which worked up to 1985 @xcite . coronal modeling can be done most reliably near solar minimum when the magnetic field structure is relatively simple and stable . at this phase of the solar activity cycle
it can be characterized by a `` magnetic pole '' which may be displaced from the rotational pole .
the magnetic field is assumed to be radial everywhere ( and decreases with distance like @xmath7 to conserve flux ) but is opposite in the opposite hemispheres .
thus the magnetic equator is a current - sheet where the field reverses .
the current sheet appears above and below the rotational equator as the sun rotates .
the solar wind velocity is low and the electron density is high in a belt of about @xmath8 around the current sheet .
however , the magnitude of the magnetic field is roughly the same in the fast and slow wind .
figure [ fg : track ] shows the location of the current sheet and the slow wind belt during one of our observations . here
the current sheet , as estimated using the wilcox solar observatory ( wso ) data , is shown as a heavy solid line and the slow wind belt is bounded by dash - dotted lines .
points on the line of sight from the earth to the pulsar at the time of observation are shown as a line of dots for which the spacing is 5@xmath9 subtended at the sun .
one can see that the line of sight crosses the current sheet twice between the earth and the closest point of approach and it passes through two different low and high density regions .
thus there is a current sheet at the magnetic equator where the field reverses .
is shown as a heavy solid line .
the location of the slow wind belt ( @xmath8 about the current sheet ) is indicated by dashed lines .
the path to psr j1022 + 1001 on 2006 aug 25 at 00:20 ut has been extrapolated back to the source surface along the streamlines assuming radial flow and a velocity of 400 km / s .
the source - surface point corresponding to the point of closest approach to the sun along the line of sight is indicated by a triangle and that corresponding to the earth position by @xmath10 . the furthest point from the earth which is shown
subtends an angle of 85@xmath9 at the sun.,title="fig:",width=302][fg : track ] however the location of the current sheet is subject to some error because a current - free potential - field model was used to extrapolate the photospheric observations of the wso into the corona .
also the width of the slow wind belt is not well - defined and the density may vary with time . in this work
we are concerned with the effect of these modeling errors on the final magnetic field estimate , and how these can be reduced using the additional constraint provided by a simultaneous group delay measurement .
this was suggested by @xcite , but they were unable to measure the group delay with sufficient accuracy in their observations .
we do not have enough measurements to make a strong statement about the magnetic field of the solar wind , other than to show consistency with other measurements , but we have sufficient data to show the modeling errors one can expect and to compare them with statistical errors . in the following sections we will discuss : the observations and the primary analysis ; the coronal model and the fitting process ; the magnetic fields estimates and their sensitivity to model errors ; and the prospects for future such observations .
the observations were part of the parkes pulsar timing array ( ppta ) project which commenced in 2004 and has made observations of 20 msps at intervals of 2 - 3 weeks since then @xcite .
here we restrict our analysis to the pulsar which provides the best observations for our purpose , psr j1022@xmath111001 which has ecliptic latitude @xmath12 , but two other msps regularly observed with the ppta , j1730@xmath132304 ( @xmath14 ) and j1824@xmath132452 ( @xmath15 ) are potentially useful .
the observations analysed are those when the line of sight to psr j1022@xmath111001 happened to be close to the sun - they were not deliberately scheduled to be near the sun . as a result
there are few observations useful for this work .
all observations use the center beam of the parkes 20-cm multibeam receiver @xcite at frequencies close to 1400mhz .
all of the data were recorded in full - polarization mode .
the back - end signal processing systems are parkes digital filter bank systems ( pdfb1 to pdfb4 ) .
each observation has a total bandwidth of 256mhz with 1024 channels .
the duration of each observation is 64 minutes and the data were averaged with 1-min sub - integrations .
a 2-min pulsed calibration signal was recorded before each pulsar observation .
the flux density scale was set using observations of hydra a. the cross - coupling between feed probes was measured using observations of psr j0437@xmath134715 covering a wide range of hour angles .
the primary analysis was done using psrchive software @xcite .
initially , 5% of the band edges and radio frequency interference ( rfi ) were automatically removed .
then the data were flux- and polarization - calibrated and the cross - coupling between the feed polarizations was corrected using the pac / pcm algorithm in psrchive . in pulsar analysis
the group delay is quantified by the `` dispersion measure '' dm which is the path integrated electron density in pc @xmath16 .
the faraday rotation is quantified by the `` rotation measure '' rm , which is the term in square brackets in equation 1 in units of rad m@xmath17 .
we will use these measures in our further discussion .
the method of obtaining rm is similar to that discussed in @xcite .
we approached the final result with a series of three successive approximations .
we first adjusted the rm to optimize the linearly polarized intensity integrated over the bandwidth .
then we separated each observation into upper and lower halves of the total band . an improved value of rm was determined using the weighted mean of position angle difference between the two halves of the band for each pulse phase bin .
the rm also includes a time variable ionospheric component .
we corrected it using an ionospheric model which is encoded in 2007 international reference ionophere ( iri ) model .
it was relatively small , ranging from @xmath130.45 to @xmath132.25radm@xmath17 , but not negligible . to obtain the final approximation to rm we used the corrected position angle vs pulse phase measured far from the sun by @xcite as a reference template .
we then found a weighted average of the differences of each observation near the sun to this reference template .
the normalized chi - squared of these weighted averages was typically about 1.3 , indicating that the error estimates used in the weighted fit were reasonably accurate and the position - angle template was a good model when the pulsar was near the sun .
to estimate the dm contribution from the solar wind , we used a simpler procedure .
the dm variations due to the interstellar medium for psr j1022@xmath111001 are smaller than the measurement error @xcite .
so we simply estimated pulse arrival times in the normal way , by fitting a detailed timing model to the entire set of observations , excluding those near the sun .
we then obtained dm@xmath18 by comparing the observed timing residual near the sun with this distant average .
there is a contribution from the ionosphere but , unlike the rm contribution , the ionospheric dm contribution is negligible compared with the measurement error . the final rm and dm measurements used are summarized in table [ tb : obs ] and plotted in figure [ fg : rmdm ] vs distance from the sun . [ cols="^,^,^,^,^,^,^,^,^,^ " , ]
we have identified four primary contributions to the error on magnetic field estimates made using the faraday rotation technique . in order of importance
these are : failure to correct for dm variations ( @xmath1950% ) ; inaccuracy of estimation of the current sheet location ( @xmath1925% ) ; error in the dm measurement ( @xmath1915% ) ; and error in the faraday rotation estimate ( @xmath191% ) .
thus use of simultaneous dm observations can roughly double the precision of the magnetic field estimates .
the value of the ppta for such measurements is clear .
it provides a good timing model for the dm far from the sun and a good template for the position - angle variations far from the sun .
we have plotted our estimates of @xmath20 along with those of other authors in figure [ fg : mag ] .
the estimates of @xcite and @xcite are not plotted because in both cases the authors only provide a measurement of the average line - of - sight component of the magnetic field and do not attempt to estimate @xmath20 .
their estimates of the average line - of - sight components are consistent with other observations but do not put a useful constraint on @xmath20 .
our estimates of @xmath20 are somewhat lower than the others , but are consistent with them , within the error bars .
all the observations are broadly consistent with the helios space craft measurements and an extrapolation like @xmath7 . thus our measurements , although few in number , confirm the very important result that the potential field models of the corona seriously underestimate the magnitude of the magnetic field for @xmath21 .
are plotted as * symbols with error bars .
the extrapolated helios measurements are a solid line @xcite , an average of the @xcite observations ( @xmath22 mg ) are a dot - dashed line and the observations of @xcite are + symbols connected by a dashed line .
the observation of @xcite is an open box and the range given by @xcite is a pair of connected x symbols .
, width=340 ] we did not observe any transients near the sun and would not have been able to analyze them with the technique we used .
however it might be possible to do so if sufficient additional information , such as might be provided by stereo , were available to constrain the geometry of the magnetic field .
we note that faraday rotation is capable of 1% precision by itself , and one might be able to measure rapid variations of magnetic field during transients with this accuracy .
we believe that more extensive observations of psrs j1022@xmath111001 , j1730@xmath132304 , and j1824@xmath132452 near the sun , perhaps involving more telescopes to provide continuous coverage during the closest approach , would be valuable and should be undertaken at the next opportunity .
there are five additional msps with ecliptic latitudes @xmath23 that may be suitable for precision timing : j0030@xmath110451 ; j1614@xmath132230 ; j1721@xmath132457 ; j1802@xmath132124 ; and j1811@xmath1324 .
preliminary observations would be required to confirm that these are `` good timers '' before undertaking coronal measurements .
the cluster of pulsars with a right ascension close to 18:00 would facilitate an intensive observing program in the last two weeks of december each year , but there are insufficient msps to monitor the corona on a regular basis .
xpy is supported by the national natural science foundation of china ( 10803004 ) , the natural science foundation project of chongqing ( cq cstc 2008bb0265 ) and the fundamental research funds for the central universities ( xdjk2012c043 ) .
gh acknowledges support from the chinese academy of sciences cas kjcx2-yw - t09 and nsfc 11050110423 .
the data presented in this paper were obtained as part of the parkes pulsar timing array project .
we thank the collaborators on this project .
the parkes radio telescope is part of the australia telescope which is founded by the commonwealth of australia for operation as a national facility managed by csiro .
r. n. , 2008 , in c. bassa , z. wang , a. cumming , & v. m. kaspi ed .
, 40 years of pulsars : millisecond pulsars , magnetars and more vol . 983 of american institute of physics conference
series , the parkes pulsar timing array project .
pp 584592 w. m. , manchester r. n. , van straten w. , reynolds j. e. , hobbs g. , wang n. , bailes m. , bhat n. d. r. , burke - spolaor s. , champion d. j. , coles w. a. , hotan a. w. , khoo j. , oslowski s. , sarkissian j. m. , verbiest j. p. w. , yardley d. r. b. , 2011 , mnras , 414 , 2087 | the magnetic field of the solar wind near the sun is very difficult to measure directly
. measurements of faraday rotation of linearly polarized radio sources occulted by the solar wind provide a unique opportunity to estimate this magnetic field , and the technique has been widely used in the past . however faraday rotation is a path integral of the product of electron density and the projection of the magnetic field on the path .
the electron density near the sun can be measured by several methods , but it is quite variable . here
we show that it is possible to measure the path integrated electron density and the faraday rotation simultaneously at 6 - 10 @xmath0 using millisecond pulsars as the linearly polarized radio source . by analyzing the faraday rotation measurements with and without the simultaneous electron density observations
we show that these observations significantly improve the accuracy of the magnetic field estimates .
pulsars : general solar wind methods : data analysis |
recently , for the study of a new class of local or graded rings which are cohen macaulay but not gorenstein , _ almost gorenstein _ local or graded rings were defined and have been studied . in this paper , for the further study of almost gorenstein rings , we concentrate on almost gorenstein homogeneous rings and investigate the @xmath0-vectors of almost gorenstein homogeneous rings . originally , the notion of almost gorenstein local rings of dimension one was introduced by barucci and frberg @xcite in the case where the local rings are analytically unramified . after this work , goto , matsuoka and phuong @xcite modified the definition of one - dimensional almost gorenstein local rings , which also works well in the case where the rings are analytically ramified . as a further investigation of almost gorenstein local rings , the definition of almost gorenstein local or graded rings of higher dimension suggested by goto , takahashi and taniguchi @xcite .
we refer the reader to @xcite for the detailed information on almost gorenstein rings .
in addition , matsuoka and murai @xcite have studied the almost gorenstein stanley reisner rings .
they introduce the notion `` uniformly cohen macaulay '' and they prove that a simplicial complex @xmath2 is uniformly cohen
macaulay if and only if there exists an injection from the stanley reisner ring @xmath3 $ ] to its canonical module @xmath4}$ ] .
this is a necessary condition for @xmath3 $ ] to be almost gorenstein ( with @xmath5-invariant 0 ) .
they also characterize the almost gorenstein@xmath6 simplicial complexes of dimension at most 2 .
see ( * ? ? ?
* section 3.3 ) . inspired by these works ( especially @xcite and @xcite ) , we will study higher dimensional almost gorenstein _ homogeneous _ rings .
the main goal of this paper is to give a characterization of almost gorensteinness in terms of the hilbert series , namely , the @xmath0-vector of the cohen macaulay homogeneous ring . for a cohen
macaulay homogeneous domain and the @xmath0-vector @xmath7 of @xmath1 , it is known by the work of stanley ( * ? ? ?
* theorem 4.4 ) that @xmath1 is gorenstein if and only if @xmath8 for @xmath9 .
( see , also ( * ? ? ?
* corollary 4.4.6 ) . )
this says that the `` symmetry '' of the @xmath0-vector of @xmath1 characterizes the gorensteinness of @xmath1 .
hence , it is natural to think of some `` almost symmetry '' of the @xmath0-vector of @xmath1 may characterize the almost gorensteinness of @xmath1 .
although to give a complete characterization might be impossible , we will prove that a certain almost symmetry of the @xmath0-vector of a cohen macaulay homogeneous ring can be a sufficient condition to be almost gorenstein ( theorem [ suff ] ) .
in addition , in the case where the ring is a domain and the socle degree is small , we will characterize the almost gorensteinness in terms of its @xmath0-vector ( theorem [ hvector ] ) .
the structure of this paper is as follows . in section [ pre ] ,
we recall the definition of almost gorenstein graded rings and prepare some propositions for the discussions later . in section [ juubun ]
, we prove that a certain almost symmetry of the @xmath0-vector of the cohen
macaulay homogeneous ring implies the almost gorensteinness .
more precisely , let @xmath1 be a cohen macaulay homogeneous ring and @xmath7 its @xmath0-vector .
we prove that if @xmath8 for @xmath10 , then @xmath1 is almost gorenstein ( theorem [ suff ] ) . in section
[ domain ] , we characterize the almost gorenstein homogeneous domain in terms of its @xmath0-vector . more precisely , for a cohen macaulay homogeneous domain @xmath1 , let @xmath7 be the @xmath0-vector of @xmath1 .
when @xmath11 , the following three conditions are equivalent : ( i ) @xmath1 is almost gorenstein ; ( ii ) @xmath1 is gorenstein ; ( iii ) @xmath12 . moreover ,
when @xmath13 , @xmath1 is almost gorenstein if and only if @xmath14 .
( see theorem [ hvector ] . ) finally , in section [ rei ] , we supply some examples of almost gorenstein homogeneous domains arising from lattice polytopes .
the authors would like to be grateful to shiro goto , naoyuki matsuoka and naoki taniguchi for their helpful comments and instructive discussions .
the author is partially supported by jsps grant - in - aid for young scientists ( b ) @xmath1526800015 .
let @xmath1 be a cohen macaulay homogeneous ring of dimension @xmath16 over an algebraically closed field @xmath17 with characteristic 0 .
let @xmath18 be the @xmath5-invariant of @xmath1 and let @xmath19 be a canonical module of @xmath1 .
note that @xmath20 .
we collect some notation used throughout this paper . * for a graded @xmath1-module @xmath21 , * * let @xmath22 denote the number of elements in a minimal system of generators of @xmath21 as an @xmath1-module ; * * let @xmath23 denote the multiplicity of @xmath21 ; * * note that we have in general the inequality @xmath24 * * let @xmath25 denote the @xmath1-module whose underlying @xmath1-module is the same as that of @xmath21 and whose grading is given by @xmath26 for all @xmath27 ; * * let @xmath28 $ ] denote the hilbert series of @xmath21 , i.e. , @xmath29=\sum_{n \in \zz}(\dim_k m_n ) t^n.\ ] ] * we denote the cohen macaulay type of @xmath1 by @xmath30 . note that @xmath31 . *
we say that @xmath7 is the _
@xmath0-vector _ of @xmath1 if @xmath32=\sum_{n \geq 0}(\dim_k r_n ) t^n=\frac{h_0+h_1t+\cdots+h_st^s}{(1-t)^d}\ ] ] with @xmath33 .
it is well known that each @xmath34 is a nonnegative integer ( since @xmath1 is cohen macaulay ) and @xmath35 .
when @xmath7 is the @xmath0-vector of @xmath1 , the index @xmath36 is called the _ socle degree _ of @xmath1 .
let us recall the definition of the almost gorenstein _ graded _ ring .
_ we say that a cohen macaulay graded ring @xmath1 is _ almost gorenstein _ if there exists an exact sequence @xmath37 of graded @xmath1-modules with @xmath38 , where @xmath39 is an injection of degree 0 . _ for a while
, we consider the condition : @xmath41 this is a necessary condition for @xmath1 to be almost gorenstein .
let @xmath42 .
then @xmath40 is a cohen
macaulay @xmath1-module of dimension @xmath43 if @xmath44 ( see ( * ? ? ?
* lemma 3.1 ) ) .
first , we see that the condition is satisfied in some cases . for example : [ domeinnobaai ] when @xmath1 is a domain , @xmath1 always satisfies the condition .
when @xmath1 is a domain , a canonical module @xmath19 can be taken as an ideal @xmath45 ( ( * ? ? ?
* section 3.3 ) ) .
replace @xmath19 by a canonical ideal @xmath46 of @xmath1 .
consider any @xmath1-homomorphism @xmath47 of degree 0 with @xmath48 .
take @xmath49 . then @xmath50 . since @xmath51 and @xmath52 and @xmath1 is a domain ,
we see @xmath53 or @xmath54 . since @xmath55
, we obtain @xmath53 , i.e. , @xmath56 , as desired .
next , we would like to describe @xmath57 and @xmath58 in terms of some invariants on @xmath1 .
[ myu- ] assume that @xmath1 satisfies
. then @xmath59 .
since @xmath39 is a degree 0 injection and all elements of @xmath60 can be its minimal generators , @xmath61 can be an element of minimal generators of @xmath62 .
hence , the assertion follows from the exact sequence .
[ i- ] assume that @xmath1 satisfies .
let @xmath7 be the @xmath0-vector of @xmath1 .
then we have @xmath63 = \frac{\sum_{j=0}^{s-1}((h_s+\cdots+h_{s - j})-(h_0+\cdots+h_j))t^j}{(1-t)^{d-1}}.\end{aligned}\ ] ] in particular , we have @xmath64 . we know that @xmath65=\sum_{j=0}^sh_jt^j/(1-t)^d$ ] and it is also known ( e.g. , see ( * ? ? ?
* corollary 4.4.6 ) ) that @xmath66=\frac{\sum_{j=0}^s h_{s - j}t^j}{(1-t)^d}. \end{aligned}\ ] ] from the exact sequence , we obtain @xmath67=[[\omega_r(-a)]]-[[r]]$ ] . on the other hand ,
since @xmath21 is a cohen macaulay module of dimension @xmath43 , @xmath67 $ ] looks like @xmath68 , where each @xmath69 is a nonnegative integer .
hence we obtain the equality @xmath70 this implies that @xmath71 for @xmath72 , as required . from this proposition ,
we obtain : [ prop1 ] assume that @xmath1 satisfies . let @xmath7 be the @xmath0-vector of @xmath1
. then we have the inequality @xmath73 for each @xmath74 , and @xmath1 is gorenstein if and only if all the equalities hold , namely , @xmath75 for each @xmath74 .
the above inequality directly follows from and the cohen
macaulayness of @xmath40 .
moreover , since @xmath1 is gorenstein if and only if @xmath76 , the second assertion also follows .
_ corollary [ prop1 ] is just an analogy of stanley s inequality for the @xmath0-vectors of semi - standard cohen
macaulay domains ( ( * ? ? ?
* theorem 2.1 ) ) . in the case of stanley reisner ring of a uniformly cohen macaulay simplicial complex , this inequality is also shown in ( * ? ? ?
* proposition 2.7 ) .
_ [ tokuchou ] the following four conditions are equivalent : * there exists an injection @xmath77 of degree 0 such that @xmath42 satisfies @xmath38 , namely , @xmath1 is almost gorenstein ; * every injection @xmath77 of degree 0 satisfies @xmath38 ; * we have @xmath78 * we have @xmath79 for each @xmath80 .
in particular , @xmath39 does not matter for the almost gorensteinness of @xmath1 .
first , we observe that all the elements of @xmath62 of degree 0 can be an element of minimal system of generators of @xmath62 .
thus , by proposition [ myu- ] and proposition [ i- ] , we obtain the equivalence of ( a ) and ( b ) and ( c ) .
next , the implication ( d ) @xmath81 ( c ) easily follows from the facts @xmath82 and @xmath83 . finally , consider the implication ( c ) @xmath81 ( d ) . from the exact sequence ,
we obtain the exact sequence @xmath84 since @xmath85 for @xmath86 , by taking @xmath87 with @xmath86 for this sequence , we obtain the isomorphism @xmath88 for @xmath86 . in general , for a graded @xmath1-module @xmath21 of dimension @xmath89 , we see that @xmath90 if @xmath91=\sum_{j \in \zz}h_j't^j/(1-t)^e$ ] .
hence , we obtain the inequality @xmath92 for each @xmath80 by proposition [ i- ] . therefore , if @xmath93 for some @xmath94 , since @xmath82 and @xmath83 , we conclude that @xmath95 , as desired .
let @xmath1 be a cohen macaulay homogeneous ring of dimension @xmath16 satisfying the condition and let @xmath7 be the @xmath0-vector of @xmath1 . one of the main results of this paper is the following : [ suff ] if @xmath8 for @xmath96 , then @xmath1 is almost gorenstein .
this theorem should be compared with the following stanley s theorem : [ symm ] assume that @xmath1 is a domain .
then @xmath1 is gorenstein if and only if @xmath8 for @xmath97 .
this theorem says that under the assumption @xmath1 is a domain , the symmetry of the @xmath0-vector of @xmath1 can be a necessary and sufficient condition for @xmath1 to be gorenstein .
on the other hand , theorem [ suff ] says that a certain `` almost symmetry '' of the @xmath0-vector of @xmath1 can be a sufficient condition for @xmath1 to be almost gorenstein .
when @xmath36 is even , @xmath98 is automatically satisfied .
hence , @xmath1 is gorenstein by theorem [ symm ] , in particular , almost gorenstein .
assume that @xmath36 is odd .
let @xmath99 , where @xmath39 is an injection in .
we may show that @xmath38 . by proposition [ i- ] , we have @xmath100=\frac{(h_{(s+1)/2}-h_{(s-1)/2})t^{(s-1)/2}}{(1-t)^{d-1}}.\ ] ] hence
, it is easy to see that @xmath40 satisfies @xmath101 , as required .
assume that @xmath1 satisfies the condition .
let @xmath7 be the @xmath0-vector of @xmath1 . in this section ,
we discuss the condition for @xmath1 to be almost gorenstein in terms of @xmath0-vectors in the case where @xmath36 is small .
first , we observe the case @xmath102 . in the case
@xmath102 , @xmath1 is always almost gorenstein .
moreover , in this case , @xmath1 is gorenstein if and only if @xmath103 . in the statement of ( * ? ? ? * theorem 10.4 ) , the condition @xmath104 is equivalent to @xmath102 and the condition `` @xmath105 is a gorenstein ring '' is automatically satisfied since @xmath1 satisfies ( ( * ? ? ?
* lemma 3.1 ) ) .
thus , ( * ? ? ?
* theorem 10.4 ) directly implies the assertion .
moreover , the condition for @xmath1 to be gorenstein follows from theorem [ symm ] .
next , let us discuss the case @xmath106 .
it is well known that the inequality @xmath107 holds in general .
we say that @xmath1 is _ level _ if @xmath108 .
( see , e.g. , @xcite ) .
one can see from ( * ? ? ?
* lemma 10.2 ) that if @xmath1 is level and @xmath106 , then @xmath1 is almost gorenstein if and only if @xmath1 is gorenstein . before considering the almost gorensteinness of @xmath1 , we give an upper bound for the cohen macaulay types of cohen macaulay homogeneous _ domains _ in theorem [ upper ] .
we will use this for the proof of theorem [ hvector ] .
the most part of the discussions for theorem [ upper ] is derived from ( * ? ? ?
* section 3 ) .
[ upper ] let @xmath1 be a cohen macaulay homogeneous domain .
then @xmath109 take a nonzero divisor @xmath110 of @xmath1 .
then we have @xmath111 and the @xmath0-vectors of @xmath1 and @xmath112 are equal .
hence , by the similar argument to ( * ? ? ?
* lemma 3.1 ) , it suffices to show that for any set of points in uniform position @xmath113 ( @xmath114 ) and the homogeneous coordinate ring @xmath1 of @xmath115 , the inequality holds .
assume that @xmath1 is the coordinate ring of a set of points in uniform position @xmath113 ( @xmath114 ) .
let @xmath36 be the socle degree of @xmath1 .
for @xmath116 , let @xmath117 be the multiplication map , where @xmath118 is the coordinate ring of @xmath119
. then this map @xmath120 is 1-generic ( ( * ? ? ?
* proposition 1.5 ) ) . by the theory of 1-generic map , we have @xmath121 for each @xmath116 ( see @xcite ) .
hence @xmath122 on the other hand , we have @xmath123 .
thus , we see that @xmath124 .
by @xmath125 , we obtain the inequality @xmath126 for @xmath116 .
( note that @xmath127 in this setting . ) since we can see that @xmath128 for each @xmath129 and @xmath130 ( ( * ? ? ?
* lemma 3.9 ) ) , we have @xmath131 . therefore , @xmath132 as an immediate corollary of theorem [ upper ] , we see the following .
note that the following has been already obtained in @xcite .
[ kei_yanagawa ] if the socle degree of @xmath1 is @xmath133 , then @xmath1 is level . in this case
, the inequality implies @xmath134 . on the other hand ,
as mentioned above , we always have @xmath135 .
hence , @xmath136 .
when @xmath11 or @xmath13 and @xmath1 is a domain , we can characterize the almost gorensteinness of @xmath1 in terms of the @xmath0-vector as follows : [ hvector ] assume that @xmath1 is a domain .
* when @xmath11 , the following conditions are equivalent : * * @xmath1 is almost gorenstein ; * * @xmath1 is gorenstein ; * * @xmath12 .
* when @xmath13 , @xmath1 is almost gorenstein if and only if @xmath14 .
\(a ) when @xmath11 , @xmath1 is always level by corollary [ kei_yanagawa ] .
hence , the equivalence ( i ) @xmath137 ( ii ) follows .
moreover , ( ii ) @xmath137 ( iii ) also holds by theorem [ symm ] .
\(b ) let @xmath99 in . from proposition [ myu- ] and proposition [ i- ]
, we have @xmath59 and @xmath138 , respectively . by the inequality given in theorem [ upper ] , we see that @xmath139 hence , if @xmath140 , then @xmath1 is never almost gorenstein by @xmath141 . on the other hand ,
if @xmath14 , then @xmath1 is almost gorenstein by theorem [ suff ] , as required .
in the case @xmath142 , we do not know any characterization of almost gorensteinness in terms of @xmath0-vectors
. however , any characterization of almost gorensteinness using @xmath0-vectors might be impossible .
for example , whether @xmath1 is level or not can not be determined by @xmath0-vectors .
see ( * ? ? ?
* section 3 ) .
the examples described in ( * ? ? ?
* section 3 ) say that the cohen macaulay type can not be determined in terms of @xmath0-vectors in general . hence , by considering corollary [ tokuchou ] , it is natural to think of we can not characterize almost gorensteinness in terms of @xmath0-vectors , either .
_ we remain the following : if exists , find examples of cohen macaulay homogeneous domains @xmath1 and @xmath143 such that * the @xmath0-vectors of @xmath1 and @xmath143 are equal ; * @xmath1 is almost gorenstein ; * @xmath143 is not almost gorenstein . moreover , we are also interested in the question : if @xmath1 is an almost gorenstein homogeneous domain with @xmath144 , then does @xmath145 holds ?
this is true when @xmath11 and @xmath13 by theorem [ hvector ] . on the other hand , when @xmath1 is not a domain , this is not true even in the case @xmath11 .
in fact , by taking the _ ridge sum _ of gorenstein simplicial complexes , we can obtain an almost gorenstein stanley
reisner ring with @xmath146 . for the detail , see @xcite . actually , this is true in general .
[ new ] let @xmath1 be an almost gorenstein homogeneous domain and @xmath7 its @xmath0-vector with @xmath144
. then @xmath145 . from corollary [ tokuchou ] ( d ) , the inequality holds . in particular , we see that @xmath147 on the other hand , since @xmath1 is a domain , we also obtain @xmath148 by for @xmath149 .
therefore , we conslude that @xmath150 , i.e. , @xmath145 .
in this section , we provide some examples of almost gorenstein homogeneous domains which are given by _ ehrhart rings _ of lattice polytopes . before giving examples , let us recall what the ehrhart ring of a lattice polytope is .
let @xmath151 be a lattice polytope , which is a convex polytope all of whose vertices belong to the standard lattice @xmath152 , of dimension @xmath16 .
let @xmath17 be a field .
then we define the @xmath17-algebra @xmath153 $ ] as follows : @xmath154=k [ { \bf x}^\alpha z^n : \alpha
\in n\pc \cap \zz^d , \ ; n \in \zz_{\geq 0 } ] , \end{aligned}\ ] ] where for @xmath155 , @xmath156 denotes a laurent monomial in @xmath157 $ ] and @xmath158 .
this @xmath159 $ ] is a normal cohen macaulay graded domain of dimension @xmath160 , where a grading is defined by @xmath161 for @xmath162 .
the ring @xmath159 $ ] is called the _ ehrhart ring _ of @xmath163 .
note that the ehrhart ring is not necessarily standard graded , but is semi - standard graded , i.e. , @xmath159 $ ] is a finitely generated module over the subring of @xmath153 $ ] generated by all the elements of degree 1 .
hence , the hilbert series of @xmath159 $ ] ( called the _
ehrhart series _ of @xmath163 ) is of the form : @xmath164\ ; ] ] = \frac{\sum_{i=0}^s h_i^ * t^i}{(1-t)^{d+1 } } , \;\;\ ; h_i^
* \in \zz_{\geq 0}.\ ] ] the sequence of the coefficiens @xmath165 appearing in the numerator of the ehrhart series is called the _ @xmath166-vector _ ( or the @xmath167-vector ) of @xmath163 .
it is known that the socle degree @xmath36 of the ehrhart ring of @xmath163 can be written as follows : @xmath168 where @xmath169 denotes the interior of @xmath163 .
thus , in particular , @xmath170 . for more details on ehrhart rings or @xmath166-vectors ,
we refer ( * ? ? ?
* part 2 ) .
first , we supply examples of ehrhart rings whose almost gorensteinness can be guaranteed by theorem [ suff ] .
let @xmath171 be the standard basis for @xmath172 , where each @xmath173 is the @xmath174th unit vector , and let @xmath175 denote the convex hull of a set @xmath176 .
_ let @xmath177 .
then @xmath178 is a lattice polytope of dimension 3 and we see that @xmath179 $ ] is standard graded and @xmath180 .
hence , by theorem [ suff ] , @xmath179 $ ] is almost gorenstein , while this is not gorenstein .
_ moreover , let @xmath181 be a lattice polytope of dimension 5 .
then we see that @xmath182 $ ] is standard graded and @xmath183 .
hence , by theorem [ suff ] again , @xmath182 $ ] is also almost gorenstein but not gorenstein . in general ,
let @xmath184 with @xmath185 and let @xmath186 be the convex hull of @xmath187 .
we compute @xmath188 for small @xmath89 s and the following show the results . @xmath189
the ehrhart rings of these polytopes are standard graded and these are almost gorenstein by theorem [ suff ] .
we expect that @xmath190 $ ] is always standard graded and almost gorenstein for every @xmath89 , although it seems difficult to give a precise proof .
next , we provide other examples of almost gorenstein ehrhart rings .
the examples which we will give are obtained from the lattice polytopes arising from finite partially ordered sets ( posets , for short ) .
let @xmath191 be a poset with a partial order @xmath192 and let @xmath193 this convex polytope @xmath194 is called the _ order polytope _ of @xmath195 .
it is known that @xmath194 is a lattice polytope and the ehrhart ring @xmath196 $ ] is often called the _ hibi ring _ of @xmath195 .
it is known that : * @xmath197 ; * @xmath196 $ ] is standard graded and an algebra with straightening laws on @xmath195 ( @xcite ) ; * the socle degree of @xmath196 $ ] is equal to @xmath198 , where a chain is a totally ordered subset of @xmath195 .
we describe the @xmath166-vector of @xmath194 in terms of @xmath195 .
let @xmath191 be a poset with a _
natural _ partial order @xmath192 , i.e. , if @xmath199 , then @xmath200 .
we say that a map @xmath201 $ ] from @xmath195 to the @xmath202-elements chain _ order - preserving _ if @xmath203 implies @xmath204 .
we can identify an order - preserving map @xmath201 $ ] with a permutation @xmath205 .
the set of all the permutations obtained in this way is denoted by @xmath206 . for example , if @xmath207 is a poset having a natural partial order @xmath208 and @xmath209 , then @xmath210 . here
, we denote @xmath211 instead of the usual notation @xmath212 .
moreover , for @xmath213 , let @xmath214 and let @xmath215 .
( clearly , @xmath216 . )
now , we will consider the following poset : for @xmath220 , let @xmath221 be the poset with the natural partial order @xmath222 , @xmath223 and @xmath224 .
figure [ poset ] below is the hasse diagram of @xmath225 .
first , for the proof of theorem [ hibiring ] , let us observe the set @xmath234 in detail .
let @xmath235 and let @xmath236 .
then @xmath237 . let @xmath238 .
then we see that @xmath239 .
let @xmath240 and let @xmath241 for @xmath242 .
we construct a map @xmath243 for each @xmath242 as follows : fix @xmath244 . *
let @xmath245 and @xmath246 .
since the parities of @xmath247 and @xmath248 are different for each @xmath249 and @xmath250 for each @xmath251 , we obtain that @xmath252 .
let @xmath253 ( resp .
@xmath254 ) be the positive integers with @xmath255 ( resp .
@xmath256 ) such that @xmath257 ( resp .
@xmath258 ) .
* we construct the permutation @xmath259 by defining each @xmath260 inductively as follows : let @xmath261 if @xmath262 or let @xmath263 if @xmath264 , and * * for @xmath265 , let @xmath266 * * for @xmath267 with some @xmath268 , let @xmath269 * * for @xmath270 , let @xmath271 + see example [ rensyuu ] below for this construction .
* we define @xmath272 .
[ rensyuu ] _ let @xmath273 and consider @xmath274
. then @xmath275 and @xmath276 .
thus @xmath277 , @xmath278 and @xmath279 .
we construct @xmath280 as follows : @xmath281 actually , for @xmath282 , we have @xmath283 . _
\(a ) for @xmath267 with some @xmath268 , by the definition of @xmath289 , we can observe that @xmath290 when @xmath289 is odd or @xmath291 when @xmath289 is even .
moreover , we can also observe that for @xmath292 with @xmath293 , when the parities of @xmath289 and @xmath294 are equal , @xmath295 if and only if @xmath296 . similar assertions also hold in the cases where @xmath297 and @xmath270 .
thus , @xmath298 are all different integers with @xmath299 for each @xmath300 .
hence , @xmath301 should define a permutation , i.e. , @xmath284 .
in addition , we also have @xmath302 .
in particular , @xmath285 . on the other hand , we know that if @xmath306 and @xmath307 , then @xmath296 .
in fact , the situation @xmath308 , @xmath307 and @xmath309 may happen if @xmath310 and @xmath311 .
however , this never happens because @xmath312 .
\(c ) since @xmath301 is uniquely determined by @xmath314 , it is obvious that @xmath288 is injective . by the injectivities of @xmath243 and @xmath315 for each @xmath242
, we obtain the bijectivity of @xmath288 for each @xmath174 , as required . given a poset @xmath195 ,
let @xmath316 with @xmath317 for every @xmath318 .
a map @xmath319 is called _ order - reversing _ ( resp . _ strictly order - reversing _ ) if @xmath320 ( resp .
@xmath321 ) for each @xmath322 with @xmath323 ( resp .
@xmath324 ) .
let @xmath325 ( resp .
@xmath326 ) denote the set of all order - reversing ( all strictly order - reversing ) maps @xmath319 with @xmath327 .
it is shown in @xcite that @xmath196 $ ] has a @xmath17-basis consisting of the monomials which look like @xmath328 and it is also shown there that the monomials which look like @xmath329 form a @xmath17-basis of the canonical ideal @xmath330 } \subset k[\oc(p)]$ ] .
let @xmath331 denote the subset corresponding to the minimal set of generators of @xmath330}$ ] .
note that @xmath196 $ ] is standard graded by the grading @xmath332 .
we will show that @xmath340 for each @xmath335 .
assume that @xmath339 can be written as a sum of @xmath341 and @xmath342 , i.e. , @xmath343 for each @xmath318 .
then we have @xmath344 for each @xmath94 . moreover , since @xmath345 , @xmath341 and @xmath327 , we have @xmath346 , equivalently , @xmath347 for each @xmath348 . hence , we obtain @xmath349 for each @xmath350 .
on the other hand , if there is @xmath351 with @xmath352 , then we see that @xmath353 a contradiction to @xmath342 .
hence , @xmath354 for each @xmath351 , i.e. , @xmath355 for each @xmath356 because @xmath341 .
let @xmath357 .
* suppose that @xmath358 .
then we have @xmath359 , a contradiction .
* suppose that @xmath360 .
then we have @xmath361 .
however , as observed above , we also have @xmath362 . since @xmath224 in @xmath225 , we should have @xmath363 , a contradiction .
we are now in the position to give a proof of theorem [ hibiring ] .
work with the same notation as above .
let @xmath99 , where @xmath368 is an injection , and let @xmath227 be the @xmath0-vector of @xmath231 . by the above discussions and lemma [ hodai1 ] , there exists a bijection @xmath288 between @xmath369 and @xmath370
in particular , we have @xmath371 . on the other hand , by proposition [ h_order_poly ]
, we have @xmath372 for each @xmath242 .
since we have @xmath373 and @xmath374 , we conclude that @xmath227 satisfies that @xmath228 for @xmath229 and @xmath230 .
therefore , by proposition [ i- ] , we obtain that @xmath375 in addition , it follows from proposition [ myu- ] and lemma [ hodai2 ] that @xmath376 hence , we see that @xmath377 .
since @xmath378 is also satisfied by , we conclude that @xmath40 satisfies @xmath379 . therefore , by corollary [ tokuchou ] , @xmath231 is almost gorenstein , as desired .
_ let @xmath380 be the poset with the partial order @xmath222 , @xmath223 , @xmath224 and @xmath381 . by the similar discussions as above
, we see that @xmath382 $ ] is also almost gorenstein with @xmath383)=2m-3 $ ] . _ t. hibi , distributive lattices , affine semigroup rings and algebras with straightening laws , in:``commutative algebra and combinatorics '' ( m. nagata and h. matsumura , eds . ) , adv .
_ pure math .
_ * 11 * , north
holland , amsterdam , ( 1987 ) , 93109 . | in this paper , for the development of the study of almost gorenstein graded rings , we discuss some relations between almost gorensteinness of cohen macaulay homogeneous rings and their @xmath0-vectors . concretely , for a cohen macaulay homogeneous ring @xmath1 , we give a sufficient condition for @xmath1 to be almost gorenstein in terms of the @xmath0-vector of @xmath1 ( theorem [ suff ] ) and we also characterize almost gorenstein homogeneous domains with small socle degrees in terms of the @xmath0-vector of @xmath1 ( theorem [ hvector ] ) .
moreover , we also provide the examples of almost gorenstein homogeneous domains arising from lattice polytopes .
-vector , homogeneous domain . ] |
the evolution of the phase - space density of stars in galaxies is determined by the collisionless boltzmann equation , which states that phase - space flow is incompressible .
thus a cloud of stars in phase space becomes more and more distorted as it evolves ; the local or fine - grained density around any point in phase space remains the same , but the coarse - grained density evolves towards a stationary state .
this process , known as phase mixing , is the principal mechanism by which stellar systems reach coarse - grained equilibrium .
the standard analogy is stirring a glass containing 20% rum and 80% coke ; eventually every finite volume element in the glass contains 20% rum and 80% coke , even though infinitesimal volume elements are either 100% rum or 100% coke ( e.g. @xcite ) .
galaxies are not thoroughly phase - mixed , since they are at most a few dynamical times old in their outer parts , and since mergers and star formation continuously add new stars .
there is a variety of direct observational evidence for incomplete phase mixing : ( i ) sharp - edged features ( `` shells '' ) in the outer parts of at least 10% of elliptical galaxies are believed to arise from the recent tidal disruption of small galaxies ( e.g. @xcite ) .
( ii ) proper - motion and radial - velocity surveys of the local halo reveal clumping in phase space ( @xcite ) .
( iii ) the tidally disrupted sagittarius dwarf galaxy ( @xcite ) provides a concrete example of a cloud in the early stages of phase mixing .
( iv ) moving groups in the solar neighborhood may result from dissolution and mixing of star clusters and associations ( @xcite ) .
incomplete phase mixing is likely to play a growing role in the interpretation of observations , for several reasons : radial - velocity and proper - motion surveys are rapidly improving in size and quality ; large telescopes with improved image quality will permit us to examine the surface - brightness structure of elliptical galaxies with high spatial resolution and signal - to - noise ratio ; the hipparcos mission has dramatically improved the precision of measurements of the phase - space distribution of stars in the solar neighborhood , and future space - based astrometric missions such as sim and gaia will provide proper - motion and parallax measurements over much of the galaxy ; and the sloan digital sky survey will soon provide a stellar database of unprecedented size and uniformity ( @xmath3 stars with five - color photometry ) .
virtually all theoretical descriptions of phase mixing have focused on mixing in a two - dimensional phase space , which is the simplest to visualize ( fig .
[ fig : mix ] ) .
this paper discusses some of the geometrical features of partially phase - mixed systems ( we shall call these `` phase structures '' ) with higher dimensionality , and the projection of this geometry into the observables that are measured in a survey .
we shall argue that this geometry can be organized by two integers : the dimension of the phase structure and the dimension of the survey .
a cloud of stars orbiting in an integrable potential ( the `` host galaxy '' ) can be viewed as a surface or manifold of dimension @xmath4 embedded in @xmath5-dimensional phase space ( usually @xmath6 ) .
for example , a cold disk has @xmath7 ( the disk occupies a two - dimensional surface in configuration space , and the velocity at each position is unique ) , collisionless cold dark matter has @xmath8 ( three spatial dimensions but zero random velocity ) , and a hot galaxy has @xmath9 . since real stellar systems always have non - zero thickness and velocity dispersion , we are referring to an `` effective '' or `` coarse - grained '' dimension ; loosely speaking , a structure in a phase space of @xmath5 dimensions has effective dimension @xmath0 if its extent in @xmath0 independent directions is much larger than its extent in the other @xmath10 directions .
because the flow of the cloud through phase space is incompressible , the dimension of the manifold does not change in the initial stages of phase mixing , although the shape of the manifold becomes more and more complicated .
however , the effective dimension can change over longer times , through several distinct processes .
the most obvious is that the phase - mixing scale grows smaller and smaller with time ( see figure [ fig : mix ] ) , and eventually becomes smaller than the coarse - graining scale . at this point
the phase structure has the same ( coarse - grained ) dimension as the phase space , unless all of the stars in the cloud have the same value of one or more isolating integrals of the motion .
gravitational scattering by small - scale irregularities ( e.g. massive objects ) can also lead to diffusion of the phase structure and an increase in effective dimension .
less obviously , phase mixing can also _ decrease _ the number of effective dimensions .
for example , consider the cloud resulting from the disruption of a small hot galaxy .
the motion of the stars in the cloud can be described by a hamiltonian @xmath11 , where @xmath12 are action - angle variables in the host potential .
the equations of motion are = , = h ( ) ( ) . the trajectory of the centroid of the cloud is @xmath13 , where @xmath14 and @xmath15 is approximately the location of the center of the satellite at the disruption time @xmath16 .
similarly , the trajectory of a star in the cloud is @xmath17 , where @xmath18 and @xmath19 .
this can be simplified to ( ( t),(t))=(_c , _
c(t))+(,+t)+o(,)^2 , [ eq : hess ] where h_ik=^2 h(_0)/j_ij_k [ eq : hhh ] is the hessian of the hamiltonian . at large times
the extent of the cloud is dominated by the terms @xmath20 in equation ( [ eq : hess ] ) , so we may write approximately ( t)_0+o ( ) , ( t)_c(t)+t+o ( ) .
[ eq : hessa ] since @xmath21 is symmetric , it is diagonalizable , that is , there exists an orthogonal matrix @xmath22 such that ^-1= ( ) , where @xmath23 , @xmath24 are the eigenvalues of @xmath21 , and @xmath25 .
we now make a canonical transformation to new action - angle variables @xmath26 using the mixed generating function @xmath27 ; thus =s= , = s=^t , = ^t , =. equation ( [ eq : hessa ] ) now simplifies to j_i(t)j_0i+o(j ) , w_i(t)w_ci(t)+_ij_i t+o(w ) .
[ eq : hessb ] this result shows that the shape of the expanding cloud is largely determined by the eigenvalues @xmath28 of the hessian @xmath21 , which are invariant under the orthogonal transformation @xmath22 .
if two of the three eigenvalues are zero ( say @xmath29 ) then the cloud expands only along the @xmath30 direction in phase space , yielding a one - dimensional tidal streamer that grows linearly with time .
if one of the eigenvalues is zero ( say @xmath31 ) then the cloud expands in two dimensions in phase space .
however , if in addition @xmath32 or vice versa , then the expansion will effectively be one - dimensional .
thus the dimension @xmath0 has been reduced from 6 to 1 or 2 , because the cloud expands much faster in some angles than in others , and not at all in action space .
we illustrate these remarks with some examples .
the hamiltonian for the triaxial harmonic oscillator may be written @xmath33 , where @xmath34 is the frequency along axis @xmath35 .
the hessian for this hamiltonian is zero , so the disrupted system does not expand at all : the frequencies are independent of the actions and phase mixing does not occur .
the hamiltonian for the kepler potential is @xmath36 , where @xmath37 and @xmath38 is the semi - major axis .
the eigenvalues of the hessian are @xmath39 , so the expanded cloud is one - dimensional , even if its original state was 6-dimensional
. simulations of disruptions of small galaxies in a spherical logarithmic potential yield one - dimensional clouds as well ( @xcite ) , but the precise conditions determining the effective dimension of the cloud resulting from tidal disruption of a small stellar system have not yet been studied .
we now survey the phase space by measuring @xmath1 of the @xmath5 phase - space coordinates of each star ( we call this @xmath1-dimensional space the `` observable space '' ) , and ask how to detect the @xmath0-dimensional phase structure . for example , an image or set of sky positions has @xmath40 ; a data cube ( positions and radial velocities ) has @xmath41 ; a proper - motion survey has @xmath42 , etc .
a bolometric laboratory detector of cold dark matter measures only their energy and moreover is restricted to a single point in configuration space . if @xmath43 the phase structure will appear as a distinct entity in the observable space , which can be detected by standard techniques : plotting the data on paper ( @xmath40 ) , visualization software ( @xmath41 ) , cluster - finding algorithms , etc .
more specialized and powerful techniques can be used if the phase structure has known properties : for example , the disruption of distant satellite galaxies in a spherical potential leads to phase structures which are great circles on the sky ( @xcite ) . in this discussion
we concentrate the more challenging case when @xmath44 , so that the phase structure covers most or all of the observable space . in this case ,
the most prominent features in the survey will be the singularities of the phase structure , that is , the locations where the projection of the phase structure into the observable space leads to a singularity .
for example , if in figure [ fig : mix ] the observations yield only the horizontal coordinate , then near - singularities arise in the plot at @xmath45 when the curve representing the phase structure is vertical .
we can generalize this simple concept to singularities with higher dimension using catastrophe theory ( @xcite ) .
we begin by setting up a coordinate system @xmath46 on the phase structure that contains the cloud .
we complete the coordinate system of the phase space by @xmath47 , which are chosen so that @xmath48 on the phase structure
. thus the phase structure is specified by = 0 , j=1, ,n ( * u*)=_j = d+1^n u_j^2 . the advantage of specifying the phase structure in terms of the generating function @xmath49 is that @xmath49 is a single - valued function of the phase - space coordinates .
next let @xmath50 be the coordinates of the observable space , and complete the coordinate system of the phase space by @xmath51 ( we call this @xmath52-dimensional space the `` hidden space '' ) . for given values of the observable coordinates , the location or locations of the phase structure
are specified implicitly by the equations = 0 , j=1, ,n - d .
[ eq : sing ] singularities in the observable space arise when there are displacements in the hidden space that leave ( [ eq : sing ] ) unaltered , that is , for which _ k=1^n - d^2(;)y_jy_kdx_k=0 , j=1, ,n - d .
these equations have a non - trivial solution if the determinant of the @xmath53 hessian matrix vanishes , that is , if = 0 .
[ eq : hessmat ] we can restrict our attention to structurally stable singularities ( roughly speaking , ones for which small changes in @xmath49 lead to small changes in the locus of the singularity ) ;
the reason is that unstable singularities represent a subset of measure zero and hence are atypical , at least for @xmath54 ( @xcite ) .
catastrophes are structurally stable singularities of gradient maps .
they are classified by their codimension , which is the dimension of the observable space minus the dimension of the singularity .
thus codimension 1 corresponds to a singular point in a survey with 1 observable , a singular line in a survey with 2 observables , etc .
the corank of the catastrophe is the minimum value of @xmath10 , the dimensionality of the hidden space , which is required for the singularity to occur .
a third classification is by the degree @xmath35 of the singularity : the mean density in a neighborhood of the observable space of radius @xmath55 around the singularity is @xmath56 .
the fold is the only catastrophe with codimension 1 . since the corank of the fold is also 1
, the simplest example occurs in a phase space with two dimensions , one of which is observable .
for example , the generating function ( y;x)=y^3+xy implies that the phase structure is the parabolic curve @xmath57 , and this manifold has a fold catastrophe at @xmath58 . if the linear density on the curve is uniform , the observable density is ( x)(1 - 14x)^1/2 , x<0 , and zero otherwise .
the square - root divergence in density on one side of the singularity is characteristic of a fold , and implies that the degree @xmath59 .
examples of fold catastrophes occur in the projection of a spherical shell onto a plane , rainbows , sunlight sparkling on the sea , twinkling of starlight , gravitational lensing , etc .
the simplest examples of fold catastrophes in phase mixing are shell structures in elliptical galaxies , which are believed to arise from disrupted companion galaxies .
shells are one - dimensional structures that appear in two - dimensional observable space ( the two coordinates on the sky plane ) and hence if they are singularities they must have codimension 1 and hence must be folds .
hernquist and quinn ( 1988 ) carried out simulations of the evolution of disrupted companions in spherical host galaxy potentials , and several of their conclusions can be interpreted in terms of the results we have derived so far .
they find that shell formation requires the accretion of either a dynamically cold or a spatially compact companion ; these are precisely the conditions required so that the cloud dimension @xmath60 ( [ sec : dim ] ) , which in turn is required so that singularities are present in a two - dimensional observable space .
they distinguish `` spatial folding '' ( projection from three - dimensional configuration space to two - dimensional space ) from `` phase - wrapping '' ( projection from phase space into three - dimensional space ) but this distinction is not fundamental : both are projections from phase space into observable space .
finally , they point out that some shells arise from step functions in surface density rather than square - root singularities ; the former are not singularities and hence are not described by catastrophe theory .
the cusp is the only catastrophe with codimension @xmath61 , and has corank 1 .
the cusp is a singular point in a 2-dimensional observable space , a line in 3-dimensional observable space , etc .
for example , consider a phase space with @xmath62 dimensions , two of which are observable .
the generating function ( y;)=y^4+x_2y^2+x_1y [ eq : gencusp ] implies that the phase structure is given by the surface @xmath63 , which is singular ( in the sense of eq .
[ eq : hessmat ] ) along the lines x_1=(-x_2)^3/2 .
[ eq : tan ] each of these lines is a fold catastrophe , and their junction is a cusp catastrophe .
if the surface density of stars is uniform on the phase structure , the surface density in the observable space @xmath64 is given parametrically by ( x_1,x_2)^1/2 , x_1=-y^3-x_2y .
[ eq : csig ] at a distance @xmath65 from the fold ( @xmath66 for @xmath67 and @xmath68 for @xmath69 ( the characteristic fold behavior ) .
the degree of the cusp singularity is @xmath70 . to illustrate the appearance of a cusp catastrophe , the first three panels of figure [ fig : one ] show a survey that samples the surface density ( [ eq : csig ] ) with 300 , 1000 , and 3000 data points in the interval @xmath71 $ ] .
the fourth panel shows a higher - resolution survey with 3000 points in @xmath72 $ ] .
the general cusp in a 2-dimensional observable space is formed from this special case by smooth distortion ( i.e. diffeomorphism ) of the observable space , and still consists of two fold catastrophes meeting with a common tangent , as in equation ( [ eq : tan ] ) .
examples of cusp catastrophes include the caustic curve seen on the surface of coffee in a cup , and the critical point of the van der waals equation of state in the pressure - temperature plane .
one might speculate that some of the dwarf spheroidal satellite galaxies are cusp catastrophes rather than bound equilibrium stellar systems , but their isopleths are approximately elliptical ( @xcite ) and do not resemble the density contours that are expected near a cusp .
there are three catastrophes with codimension @xmath73 , the swallowtail , elliptic umbilic , and hyperbolic umbilic .
the swallowtail has corank 1 , and thus first appears in a phase space with @xmath74 dimensions ; the umbilics have corank 2 and thus require @xmath75 .
all of these singularities are points in a 3-dimensional observable space , lines in 4-dimensional observable space , etc .
just as a cusp occurs at the junction of two fold curves , these occur at the junction of cusp curves , which are connected by fold surfaces .
consider a phase space with @xmath76 dimensions , three of which are observable .
the generating function ( y;)=y^5+x_3y^3+x_2y^2 + x_1y , [ eq : genswallow ] implies that the phase structure is given by @xmath77 , which is singular ( in the sense of eq .
[ eq : hessmat ] ) along the surfaces defined in terms of the parameters @xmath78 by x_1=3y^4+x_3y^2,x_2=-2(2y^3+x_3y ) .
[ eq : tans ] each of these surfaces is a fold catastrophe ; the folds meet at cusp lines defined by x_1=-3y^4,x_2=8y^3 , x_3=-6y^2 .
[ eq : cusps ] the fold surfaces intersect one another along the parabola @xmath79 .
the degree of the swallowtail is @xmath80 .
figure [ fig : two ] shows the fold surfaces and cusp lines for the swallowtail catastrophe .
consider a phase space with @xmath81 dimensions , three of which are observable .
the generating function ( y_1,y_2;)=y_1 ^ 3 - 3y_1y_2 ^ 2-x_3(y_1 ^ 2+y_2 ^ 2)-x_2y_2-x_1y_1 [ eq : genumbe ] implies that the phase structure is given by the surface 3y_1 ^ 2 - 3y_2 ^ 2 - 2x_3y_1-x_1=0 , 6y_1y_2 + 2x_3y_2+x_2=0 . which is singular along the surfaces defined parametrically by x_1=3y_1 ^ 2 - 3y_2 ^ 26y_1(y_1 ^ 2+y_2 ^ 2)^1/2 ,
x_2=-6y_1y_26y_2(y_1 ^ 2+y_2 ^ 2)^1/2 , x_3=(y_1 ^ 2+y_2 ^ 2)^1/2 .
each of these surfaces is a fold catastrophe ; the folds meet at three parabolic cusp lines defined by @xmath82 and confined to planes at angles of @xmath83 that intersect along the @xmath84-axis ( see figure [ fig : three ] ) .
the degree is @xmath85 .
the generating function ( y_1,y_2;)=y_1 ^ 3+y_2 ^ 3-x_3y_1y_2-x_2y_2-x_1y_1 [ eq : genumbh ] implies that the phase structure is given parametrically by x_1=3y_1 ^ 2+_1(y_1y_2 ^ 3)^1/2 , x_2=3y_2 ^ 2+_1(y_1 ^ 3y_2)^1/2 , x_3=_2(y_1y_2)^1/2 , where @xmath86 and @xmath87 .
each of these surfaces is a fold catastrophe ; the folds intersect along the coordinate axes @xmath88 and @xmath89 ( fig .
[ fig : three ] ) .
the cusp line is defined parametrically by x_1=x_2=9y^2 , x_3=6_2y , y>0 .
the degree is @xmath85 .
these exhaust the catastrophes with codimension @xmath90 .
catastrophes with higher codimension are only relevant when the observable space has dimension @xmath91 .
catastrophes of higher codimension become increasingly complicated , and are less important since we have argued in [ sec : dim ] that typical phase structures will often have dimension 3 or less .
it is important to understand the geometry of phase structures before searching for them .
we have focused on two specific aspects of this geometry : the expected dimension of the phase structure in ( usually 6-dimensional ) phase space , and the classification of singularities when a @xmath0-dimensional phase structure is projected into a @xmath0-dimensional survey .
phase mixing of disrupted small , hot galaxies generally leads to 3-dimensional phase structures but the dimensionality can be smaller if the potential of the host galaxy has special symmetries ( e.g. spherical or keplerian ) , or if the eigenvalues of the hessian ( eq . [ eq : hhh ] ) are very different in magnitude .
phase mixing of large , cold galaxies leads to 2-dimensional phase structures . even though phase mixing is a simple process
, there remain many unresolved theoretical issues .
how fast do small - scale gravitational irregularities or large - scale orbital chaos disrupt phase structures ?
do phase structures provide a significant source of relaxation in galaxies ( @xcite ) ?
what are the properties of phase structures in galaxies that we expect from standard models of structure formation ?
what statistical measures can we use to characterize phase structures and their projections ? | partially phase - mixed structures in galaxies occupy a complex surface of dimension @xmath0 in six - dimensional phase space .
the appearance of such structures to observers is determined by their projection into a subspace whose dimensionality @xmath1 is determined by the number of observables ( e.g. sky position , distance , radial velocity , etc . ) .
we discuss the expected dimensionality of phase - space structures and suggest that the most prominent features in surveys with @xmath2 will be stable singularities ( catastrophes ) .
the simplest of these are the shells seen in the outer parts of elliptical galaxies .
= cmcsc10 # 1#2.5pt # 1#2 |
continuous - wave opos have been used for over a decade to study and modify the quantum noise of light .
successful experiments involved creation of squeezed light below the oscillation threshold @xcite , sub shot noise intensity correlations between signal and idler beams @xcite and pump squeezing @xcite .
these experiments have been performed using opos pumped with visible light , generally the second harmonic of nd : yag lasers , and most of them generated non - classical light at a wavelength different from the pump wavelength .
pump squeezing in an opo above threshold @xcite provides a direct and efficient way to generate bright squeezed beams without frequency changes .
the opo acts as a `` quantum noise eater '' , like a kerr medium , by making use of the so - called `` cascaded nonlinearities '' @xcite .
such cascaded non linearities , resulting from the co - existence of sum frequency generation and parametric down conversion occuring in an opo , have been shown to amount to a nonlinear phase - shift on the pump beam which emerges from the cavity after its nonlinear interaction with the signal and idler beams in the parametric crystal .
this effect has been known for a long time to be able to produce significant quantum noise reduction @xcite . in order to have access to a new frequency range in pump squezing
, we have used a quasi phase - matched ( qpm ) material , periodically - poled lithium niobate ( ppln ) , which makes it possible to build low - threshold c.w .
opos operating with a nd : yag laser as a pump and signal and idler beams in the mid - ir @xcite .
the experiment that we present here is the first , to the best of our knowledge , to use qpm material for a quantum noise reduction experiment in the c.w . regime . after a brief description of the experimental set - up
, we will study in section [ meanfields ] the behavior of the mean fields produced by this device and compare it to theoretical predictions . in the last part
, we will present the theoretical predictions as well as the experimental results for the quantum noise , showing 38% noise reduction below the standard quantum limit .
the set - up is shown on fig . [ manip ] . we have used a diode - pumped monolithic nd : yag laser ( lightwave 126 - 1064 - 700 ) as a pump source .
this source delivers a c.w .
beam in a tem@xmath2 mode which exhibits very large excess noise below 20 @xmath3 due to the relaxation oscillation peak noise of the pump semiconductor laser ( left , fig .
[ fc ] ) . in order to obtain noise reduction below the standard quantum limit within the opo cavity bandwidth , we need first to obtain a pump beam at the shot - noise level . for this purpose
, we send the beam through a filtering cavity .
it is a ring three - mirror cavity @xcite to prevent back reflections from perturbing the laser ( fig .
[ manip ] ) .
the input and output mirrors ( m1 , m2 ) are identical plane mirrors with a reflexion coefficient on the order of 99.5 % for the @xmath4-polarisation . the end mirror ( m3 ) is a concave mirror with a radius of curvature of 750 @xmath5 and a reflexion coefficient close to 99.9 % . the measured finesse for the cavity is close to 700 . for a 70 @xmath6 round trip length
the cavity bandwidth is below 1 @xmath3 when the cavity is locked on resonance .
the field fluctuations of the transmitted beam are shot - noise limited above 5 @xmath3 ( right , fig .
[ fc ] ) . due to the non negligible losses on the mirrors ,
we obtain 30 % power transmission as opposed to 80 % theoretical prediction and 75 % in similar experiments where very - high quality mirrors were used @xcite .
when the cavity is locked by a servo loop using a frequency modulation of the nd : yag laser at 40 khz , its output is stable over several tens of minutes .
the mode matching of the pump beam to the opo cavity was 92% without the filtering cavity and over 97 % with the filtering cavity which plays a role of spatial filter .
the half wave plate hwp1 and the polarizing beam splitter pbs1 control the pump intensity sent into the opo cavity .
the pump beam passes through the faraday rotator ( fr ) , and the input polarization is controled by the half wave plate hwp2 . the reflected pump beam , after the interaction with the opo cavity , returns by the same optics and is reflected by pbs1 after the double passage through the faraday rotator . for the local oscillator
, we used part of the input beam reflected by pbs1 . the reflected beam passes by an optical circulator made by a quater wave plate ( qwp1 ) and a mirror ( m4 ) mounted in a piezo - actuator for the phase variation .
the crossed polarized local oscillator and reflected pump beams are then mixed at the homodyne detection system by the polarizing beam splitter pbs2 and the half waveplate hwp3 , and detected by the high efficiency ingaas photodetectors ( epitaxx etx 300 , quantum efficiency 94 % ) .
a pair of brewster plates ( bp ) at the brewster angle for the local oscillator polarization can be introduced in the system , producing a reduction in the reflected pump beam with no noticeable change in the local oscillator .
the triply resonant opo has been described extensively in a previous paper @xcite and we will only recall here its main properties . the cavity is a symmetric cavity with a length of approximately 65 @xmath5 .
the pump coupling mirror ( m5 ) has a reflection coefficient of 87 % , so that the transmission is much larger than the other losses inside the cavity .
the end mirror ( m6 ) is highly reflecting for the pump beam to reduce the losses ( reflection of 99.8 % ) .
the mirrors have large reflection coefficients in the 2 @xmath0 range to ensure small losses for the pump beam due to conversion to signal and idler ( 99.8 % for m5 , 99 % for m6 ) .
they have a radius of curvature of 30 @xmath5 to ensure optimal waist sizes in the crystal ( 36 @xmath0 at 1.06 @xmath0 , 51 @xmath0 at 2.12 @xmath0 ) .
the crystal is a ppln crystal from crystal technologies , with a width of 12 @xmath5 , a length of 19 @xmath5 and a thickness of 0.5 @xmath5 .
both faces of the crystal are antireflection coated for pump , signal and idler frequencies ( residual reflection of 0.6 % for the pump and 0.4 % for signal and idler ) .
the crystal absorption in the infrared region is small ( 0.3 % at 1.06 @xmath0 according to the manufacturer ) .
the crystal is formed of 8 different paths with spatial periodicity varying between 30 and 31.2 @xmath0 .
all the experimental results shown in this paper were made using a 31.1 @xmath0 spatial period for which the exact quasi - phase matching condition for degenerate signal and idler wavelength at 2.12 @xmath0 is obtained with a temperature @xmath7 .
the crystal is placed inside a temperature stabilized oven . in presence of the crystal , the measured finesse for the pump is around 40 , in good agreement with the theoretical expressions while the calculated finesse for signal and idler is around 200 .
the typical oscillation threshold is around 500 @xmath1 with thresholds as low as 300 @xmath1 for a limited time . for continuous operation ,
the opo cavity is locked in the resonance of the 2 @xmath0 output .
the well - known equations for the intracavity mean fields are @xcite @xmath8 where @xmath9 , @xmath10 and @xmath11 are the intracavity pump , signal and idler fields respectively , @xmath12 and @xmath13 are the amplitude transmission and reflection coefficients at the pump frequency , @xmath14 is the nonlinear coupling coefficient , @xmath15 is the common value for the amplitude reflection coefficient at the signal and idler frequencies .
@xmath16 , @xmath17 and @xmath18 are the round - trip phases for pump , signal and idler fields respectively . above threshold ,
equations ( [ eqsignal ] ) and ( [ eqidler ] ) lead to the condition @xcite @xmath19 where @xmath20 is an integer .
each value of @xmath20 is associated with a given mode , _
i.e. _ a well - defined couple of signal and idler frequencies , @xmath21 and @xmath22 .
more precisely , using the sellmeier equation for lithium niobate @xcite and eq .
( [ eqphi ] ) , one can find for any given cavity length @xmath23 , crystal temperature @xmath24 and mode index @xmath20 the values of @xmath21 and @xmath22 . @xmath24 and
@xmath20 being fixed , a minimum of the oscillation threshold is reached at values of the cavity length where both @xmath17 and @xmath18 are multiple of @xmath25 .
let us call @xmath26 the signal - idler double resonance point within a given free spectral range . in order to calculate the oscillation threshold and output fields , we have taken into account the periodical poling .
the usual phase mismatch , @xmath27 , is replaced by @xmath28 where @xmath29 denotes the spatial period of the crystal .
the nonlinear coupling coefficient which can be written in a usual bulk crystal @xmath30 now becomes @xmath31 for an integer number of periods within the crystal of lenght @xmath32 .
if the crystal does not consist of an integer number of periods , the coupling coefficient has an additionnal term like @xmath33 , depending on the added fraction of period .
since the relative contribution of this term will have the order of @xmath29/l , it can be neglected in the calculation of @xmath34 .
however , its contribution to the crystal length will have a much more important effect in the relative phase - shift between the interacting beams .
the effect of such an additional phase - shift is known to increase the threshold in the case of a linear opo cavity @xcite .
this increase will depend on this phase term , reaching 1.9 times the mininum threshold that would be obtained with a perfect phase matching , perfect periodicity of the crystal and no added relative phase . in order to optimize the threshold , it is possible to use crystals with non parallel faces which allow a precise choice of the crystal length @xcite . curves ( d ) , ( e ) and ( f ) on fig .
[ chpm ] show the mean field intensities calculated from these equations as a function of cavity length for three different temperatures and pump intensities . for a given cavity length ,
there are many different modes @xmath20 where the opo can oscillate .
for each one of these modes there will be a different threshold value , and the opo will operate in the mode with the lowest threshold @xcite .
the calculated frequency difference between signal and idler is shown in @xmath35 on traces ( a ) , ( b ) and ( c ) showing that the system can be swept over a large frequency range . during the cavity scanning
, the opo passes through a large number of individual modes , remaining on a given value of the frequency difference only over a limited range of cavity length ( inset , curve ( a ) ) .
when the temperature approaches @xmath36 , the signal and idler frequencies become very close , and the difference between signal and idler indices of refraction becomes very small . in these conditions ,
the resonance lengths @xmath26 of different oscillation modes @xmath20 become very close to each other and the possible oscillating modes overlap . as the opo
is never multimode in steady state operation @xcite , it jumps from one mode to the next ( @xmath37 ) when the cavity length is varied . as a result ,
the opo always oscillate very close to the double resonance configuration .
the mode jumps can be seen as discontinuities on the mean field .
these discontinuies are easily visible on the experimental curves and using an auxiliary fabry - perot , we have observed monomode operation with mode hops at the discontinuities . these discontinuities are not seen on the calculated curves which do not take into account the dynamics of the system while experimentally , since the threshold varies slowly for the different modes , the opo remains on its initial mode even if it is unstable , leading to mode jumps by more than two units .
intracavity signal and idler intensities exhibit a bistable behaviour as a function of cavity length when specific conditions on the pump and signal detunings are met @xcite . in our case , as the opo works almost always with a small detuning of the signal and idler fields , it is difficult to observe any bistable behavior .
nevertheless , due to the parabolic shape of the phase - matching curve in a type - i medium @xcite , for a temperature close to @xmath38 , oscillation is limited by the frequency degeneracy of signal and idler ( @xmath39 ) .
thus the detuning range obtained for the mode @xmath39 is much larger than that obtained for other modes .
such a behaviour may be seen on the left hand side of curve ( f ) . in this region , for an adequate cavity detuning the bistable behavior is expected .
the value of the pump detuning can be selected by the temperature control of the phase matching condition at degeneracy , displacing the position of the sharp side of the signal and idler curve relative to the pump resonance position .
the sharp edge position is critically dependent on the temperature , and a @xmath40 variation is enough to displace the degenerate point out of the pump resonance .
the quantum fluctuations of the pump field have been calculated in ref .
this paper shows that best quadrature squeezing on the pump field increases with pump power and is already significant a few times ( by a factor 3 or 4 ) above threshold and for any value of the detunings .
it increases when approaching the bistability threshold , i.e. for rather important signal - idler and pump detunings .
squeezing occurs on the phase component at exact double resonance @xcite and , for non - zero detunings , on a quadrature component which rotates quickly when this detunning changes . using the detailed calculations of @xcite , it is possible to evaluate the optimum noise reduction as a function of cavity length at a given temperature .
the curves are shown on fig .
it can be seen on these curves that the optimum squeezing is significant on a broad range of detunings and increases with the pump power . on the other hand ,
the intensity noise ( grey line ) exhibits rapid oscillations for small changes of the cavity length ( of the order of tenths of nm ) .
even with a stable electronical locking of the opo cavity , fast small variations around the locking position can not be avoided , thus reducing the measurable intensity squeezing .
actually the most efficient locking is obtained exactly at the triply resonant condition when only quadrature squeezing can be observed .
the quantum noise of quadrature components of the reflected pump beam are measured using a slightly modified homodyne detection procedure . in order to prevent saturation of the detectors ( which occurs around 3 @xmath41 )
, the local oscillator can not be much more powerful than the reflected pump beam .
the difference of the photocurrents fluctuations obtained by the two balanced photodetectors can be written as @xmath42 only when the local oscillator is much stronger than the beam to measure .
@xmath43 denotes the mean local oscillator power and @xmath44 the variance of the reflected pump in the quadrature @xmath45 determined by the phase difference between the local oscillator and the reflected pump mean field .
the measurement , with a spectrum analyzer , of the spectral density of the difference of the photocurrents gives directly the noise spectrum of the quadrature component of angle @xmath45 . in the general case of finite ratio between the mean intensities of the beam to study and of the local oscillator beam
, an additionnal term appears and the difference of the photocurrents is now @xmath46 where we neglect correlations between the two beams ( this is justified since these beams are obtained by splitting a coherent state on a beam splitter ) .
@xmath47 is the mean reflected pump intensity and @xmath48 the variance of the pump beam in a given quadrature .
this extra term can be measured independently .
taking into account the fact that the local oscillator is in a coherent state , its high frequency noise is equal to that of the vacuum noise and can be measured by blocking the local oscillator arm . using this measurement
, we can renormalize the total photocurrent difference ( [ ditotal ] ) and obtain the reflected pump noise .
this correction amounts to approximately 30 % of the shot noise .
the noise measurement is made with a pair of balanced ingaas photodetectors .
the mode matching between the reflected pump and the local oscillator is on the order of 97 % .
the pump intensity incident on the opo is @xmath49 , for a threshold power of 300 @xmath50w .
the reflected pump intensity at the position of the photodetectors was @xmath51 .
the temperature is @xmath52 and the local oscillator power is @xmath53 .
figure [ sqraw ] shows the noise power @xmath54 obtained as the local oscillator phase is scanned ( thick black line ) , as well as the shot noise level measured for the local oscillator and the reflected pump beams ( respectively @xmath55 and @xmath56 ) .
the sum of the shot noise power of both beams is also presented for comparison . the electronic noise level is small ( -102.6 @xmath57 ) and is taken into account in our calculations .
the variance of the quadrature component fluctuations , normalized to the coherent state fluctuations ( @xmath58 ) can be obtained from the noise measurements shown at fig .
[ sqraw ] using the correction presented in eq .
( [ ditotal ] ) .
figure [ sqmes ] shows this normalized noise @xmath59 .
the measured noise reduction is 30 % corresponding to an inferred noise reduction at the output of the opo of 38 % , when losses are taken into account .
to check that we have indeed noise reduction below the standard level , we have introduced losses on the squeezed beam . in order for these losses not to change the operating conditions ( namely the incident pump power , @xmath60 ) ,
we have introduced losses using a pair of glass plates oriented at the brewster angle with respect to the propagation axis ( fig .
[ manip ] ) : in that case , the losses are 46 % for the squeezed beam and not measurable for the local oscillator since the two beams are orthogonally polarized . the dashed grey line of fig .
[ sqmes ] gives the quadrature noise measured using this device .
this line coincides well with the full grey line which represents the measured noise without attenuation ( thick black line ) corrected using the well - known formula @xcite @xmath61 , where @xmath62 is the normalized noise of the attenuated beam , @xmath63 is the normalized noise of the input beam , and @xmath64 is the loss coefficient of the inserted attenuator .
the measured noise reduction is in good agreement with the quadrature noise reduction obtained in our theoretical calculations ( 44 % , fig .
[ sq ] ) . the smaller value for the measured compression may be due to an underestimated value for the intracavity losses at our calculations .
we have reported on the first use of quasi - phase matched materials for cw quantum noise reduction experiments and observed an inferred noise reduction of 38 % below the standard quantum limit using such a material .
the use of quasi - phase matched materials in quantum optics seems very promising as it enlarges the wavelength range where quantum effects can be obtained . | we have used an ultra - low threshold continuous - wave optical parametric oscillator ( opo ) to reduce the quantum fluctuations of the reflected pump beam below the shot noise limit .
the opo consisted of a triply resonant cavity containing a periodically - poled lithium niobate crystal pumped by a nd : yag laser and giving signal and idler wavelengths close to 2.12 @xmath0 and a threshold as low as 300 @xmath1 .
we detected the quantum fluctuations of the pump beam reflected by the opo using a slightly modified homodyne detection technique .
the measured noise reduction was 30 % ( inferred noise reduction at the output of the opo 38 % ) . |
localization involves the process of obtaining physical locations using radio signals .
radio signals are exchanged between radios with known and unknown positions .
radios at known positions are called reference radios .
radios at unknown positions are called blind radios .
localization of blind radios reduces to fitting these measured radio signals to appropriate propagation models .
propagation models express the distance between two radios as a function of the measured radio signals .
these measured radio signals are often modeled as deterministic radio signals with noise using a large variety of empirical statistical models @xcite .
radio localization usually involves non - linear numerical optimization techniques that fit parameters in the propagation model given the joint probability distribution of the ensemble of measured radio signals .
localization precision is usually expressed as the root mean squared error ( rmse ) . in the field of wireless sensor networks ,
the estimation bounds on localization precision are often calculated by the cramr - rao lower bound ( crlb ) from empirical signal models with independent noise @xcite .
in general , localization precision depends on whether the measured radio signals contain phase information .
phase information can only be retrieved from measurements that are instantaneous on a time scale that is short relative to the oscillation period of the signals .
the smallest measurable position difference depends on how well phase can be resolved .
this is usually limited by the speed and noise of the electronics of the system .
less complex and less expensive localization systems are based on measurements of time - averaged power flows or received signal strengths ( rss ) .
time - averaging is usually performed over timespans that are large compared to the coherence time of the radios , so that phase information is lost .
rss localization is an example of such less - complex systems .
hence , when determining the bounds on localization precision , it makes sense to distinguish between the time scales of the signal measurements , i.e. between instantaneous and time - averaged signal and noise processing .
speckle theory @xcite describes the phenomenon of power - flow fluctuations due to random roughness from emitting surfaces .
this theory shows that independent sources generate power - flow fluctuations in the far field that are always cross - correlated over a so - called spatial coherence region .
the linear dimension of this region equals the correlation length .
this correlation length has a lower bound in far - field radiation . in all practical cases of interest ,
the lower bound on the correlation length of this far - field coherence region is of the order of half the mean wavelength of the radiation .
this lower bound is called the diffraction limit or rayleigh criterion @xcite .
it is not obvious that this lower bound on correlation length holds in our wireless sensor network setup with its characteristic small cylindrical antennas .
therefore , we derive this lower bound on the correlation length with the corresponding cross - correlation function from the maxwell equations using the ieee formalism described by @xcite .
our derivation of this lower bound on the correlation length and corresponding cross - correlation function leads to the well - known van cittert - zernike theorem , known in other areas of signal processing like radar @xcite , sonar @xcite and optics @xcite .
our novel experimental setup validates that power - flow fluctuations are cross - correlated over a correlation length of half a wavelength by increasing the density of power - flow measurements from one to 25 power - flow measurements per wavelength . in the field of wireless sensor networks ,
the estimation bounds on localization precision are usually determined by empirical signal models with independent noise over [ space , time ] @xcite .
hence , the correlation length of the radiation is assumed to be infinitely small .
this representation renders practical relevance as long as the inverse of the sampling rate is large compared to the correlation length , so that correlations between measurements are negligibly small .
this paper experimentally and theoretically investigates radio localization in the far field when sufficient measurements are available to reveal the bounds on localization precision .
we use bienayme s theorem @xcite to show that the ensemble of correlated power - flow measurements has an upper bound on the finite number of independent measurements over a finite measuring range .
we show that this finite number of independent measurements depends on the correlation length .
when we account for the finite number of independent measurements in the fisher information , the crlb on localization precision deviates 2 - 3% from our experimental results . we show that this approach provides practically identical results as the crlb for signals with correlated gaussian noise @xcite , given the lower bound on correlation length .
hence , the lower bound on correlation length determines the upper bound on localization precision .
there are a few papers that assume spatially cross - correlated gaussian noise on power - flow measurements @xcite .
these papers consider cross - correlations caused by shadowing that extend over many wavelengths .
@xcite states that their cross - correlation functions do not satisfy the propagation equations .
such cross - correlations have no relation to wavelength of the carrier waves as the diffraction limit is not embedded in the propagation model like we derive .
correlation , coherence and speckle properties of far - field radiation are governed by the spreads of fourier pairs of wave variables that show up as variables in the classical propagation equations of electromagnetic waves .
the products of these spreads of fourier pairs express uncertainty relations in quantum mechanics , which are formulated as bandwidth relations in classical mechanics @xcite and signal processing @xcite .
@xcite mathematically establishes a relationship between fisher information , the crlb and uncertainty .
as uncertainty is lower bounded by diffraction , fisher information is upper bounded and crlb is lower bounded .
our experimental work reveals quantitative evidence for the validation of this theoretical work .
this paper is organized as follows .
section [ sec : experimental_setup ] describes the experimental setup .
section [ sec : theory ] derives the propagation and noise models from first principles for our experimental setup for each individual member of the ensemble of measurements .
section [ sec : model ] describes the signal model , maximum likelihood estimator ( mle ) and crlb analysis for the ensemble of measurements using the results of section [ sec : theory ] .
section [ sec : experimental ] presents the experimental results in terms of spatial correlations and upper bounds on localization precision .
finally , section [ sec : discussion ] provides a discussion and section [ sec : conclusions ] summarizes the conclusions .
[ fig : measurement_setup ] and [ fig : photo ] show the two - dimensional experimental setup .
[ fig : measurement_setup ] shows a square of @xmath0m@xmath1 , which represents the localization surface .
we distinguish between two types of radios in our measurement setup : one reference radio and one blind radio .
the reference radio is successively placed at known positions and is used for estimating the position of the blind radio .
the crosses represent the @xmath2 manual positions of the reference radio ( @xmath3 ) and are uniformly distributed along the circumference of the square ( one position every half centimeter ) . rather than placing a two - dimensional array of reference radios inside the @xmath0m@xmath1 square
, it may suffice to place a significantly smaller number on the circumference of the rectangle and get similar localization performance .
measuring field amplitudes on a circumference rather than measuring across a surface is theoretically expressed by green s theorem as is shown in section [ sec : theory ] . whether sampling power flows on circumferences instead of sampling across two - dimensional surfaces suffices has yet to be verified by experiment .
this paper aims to show the practical feasibility of this novel technique , which was first proposed by @xcite .
m@xmath1 ) to measure rss to the blind radio .
the blind radio is located at the origin . for illustrative purposes
, this figure only shows @xmath4 of the @xmath2 measurement positions.,scaledwidth=48.0% ] [ t ] cm apart .
the inset on the right shows a close - up picture of the radio.,scaledwidth=48.0% ] the red circle represents the position of the blind radio .
we place the blind radio at an unsymmetrical position , namely @xmath5 .
we only use one blind radio and one reference radio to minimize influence of hardware differences between radios .
the blind and reference radios are both main powered to minimize voltage fluctuations .
the blind radio has a power amplifier and broadcasts messages with maximum power allowed by etsi to maximize snr .
both blind and reference radios have an external dipole antenna .
the antennas have the same vertical orientation and are in line - of - sight ( los ) for best reception .
this implies that the polarization is vertically oriented perpendicular to the two - dimensional localization plane .
the length of the antenna is half a wavelength .
its diameter is roughly one twentieth of a wavelength .
we keep the relative direction of the printed circuit boards on which the antennas are mounted constant by realigning them every @xmath6 cm . in order to minimize interference from ground reflections , we place the radios directly on the ground , so that their antennas are within one wavelength height .
the ground floor consists of a reinforced concrete floor covered by industrial vinyl .
we minimize interference from ceiling reflections by placing a @xmath7cm@xmath1 aluminum plate one centimeter above the blind radio antenna .
the ground and the aluminum plate minimize the influence from signals in the z - direction , so that we only have to consider signals in two dimensions .
all reference radio positions are in the far field of the blind radio .
a photo of our setup is shown in fig .
[ fig : photo ] . at each of the @xmath2 reference radio positions ,
the reference and blind radio perform a measurement round .
a measurement round consists of @xmath9 repetitive multiplexed rss measurements to investigate and quantify the measurement noise and apply crlb analysis to this measurement noise .
each measurement round consists of @xmath10 measurement sets that consist of @xmath11 rss measurements on an unmodulated carrier transmitted by the blind radio . between each measurement set of @xmath11 rss measurements ,
the reference and blind radios automatically turn on and off ( recalibrate radios ) .
although we did not expect to find any difference from these two different forms of multiplexing rss signals , our experiments should verify this , which they did .
hence , a measurement round consists of @xmath10 measurement sets , each set consisting of @xmath11 repetitive multiplexed rss measurements .
measurement rounds and measurement sets are represented by @xmath12 .
index @xmath13 identifies the position of the reference radio and thus the measurement round , index @xmath14 identifies the @xmath10 measurement sets , and index @xmath15 identifies the @xmath11 individual rss measurements in each measurement set .
@xmath16 represents the measured power in dbm and is time - averaged over @xmath17 according to the radio chip specification @xcite .
the coherence time is @xmath6 , which implies a coherence length of @xmath18 km .
the averaging time of @xmath17 is a factor of five larger than the specified coherence time of the carrier wave of a typical 802.15.4 radio @xcite .
practically , this means that these power measurements lose all phase information @xcite . in theory , this means that we measure the time - averaged power flow or poynting vector as the cross - sections of the antennas are the same .
the blind radio transmits in ieee 802.15.4 channel @xmath19 in order to minimize interference with wi - fi channels @xmath20 and @xmath21 .
the reference radio performs power - flow measurements in the same channel and sends the raw data to a laptop over usb , which logs the data .
we use matlab to analyze the logged data . between each measurement
round , we change the position of the reference radio by @xmath22 cm and push a button to start a new measurement round .
note that @xmath22 cm is well within the @xmath23 diffraction limit of half the mean wavelength . in summary ,
each measurement set takes one second ; each measurement round takes roughly one minute ( @xmath24 seconds ) .
the experiment consists of @xmath2 measurement rounds and takes in total @xmath25 hours ( @xmath2 minutes @xmath26 @xmath25 hours ) . in practice
, we spend almost three weeks in throughput time to perform this complete data collection .
this section formulates the propagation and noise model of radio localization systems operating in the far field .
this model holds for each individual member of the ensemble of @xmath2 power - flow measurements over space described in section [ sec : experimental_setup ] .
our propagation model is based upon em inverse source theory , where we follow the ieee formalism given by @xcite .
this formalism is needed to derive the lower bound on correlation length of wireless networks using power - flow measurements on a time scale long compared to the coherence length of the hardware .
the measurement configuration of practical interest is composed of the wave - carrying medium , i.e. the atmosphere , a number of source domains and one receiver domain .
our measuring chamber has a variety of contrasting obstacles like a ceiling , walls , a ground floor , pillars , and a small aluminum table hiding the ceiling from the transmitter .
most of these secondary sources are so far away from the receiver that they are neglected in our theoretical representation .
we account for the following domains : * the wave - carrying medium denoted by @xmath27 , with electric permittivity @xmath28 , magnetic permeability @xmath29 and em wave speed @xmath30 .
it is locally reacting , spatially invariant , time invariant and lossless . to locate positions in space , the observer employs the ordered sequence of cartesian coordinates @xmath31 , or @xmath32 , with respect to a given origin @xmath33 , while distances are measured through the euclidean norm @xmath34 .
the fourth dimension is time and is denoted by @xmath35 . *
a transmitter with bounded support @xmath36 with @xmath37 . here
, @xmath38 denotes the spatial support carrying electric currents with volume density @xmath39 , and @xmath40 denotes the spatial support carrying magnetic currents with volume density @xmath41^{-}$].the transmitter transmits unmodulated em currents at a temporal frequency of @xmath42ghz with a temporal frequency bandwidth of @xmath43khz @xcite .
the polarization of the em currents is assumed to be perpendicular to the ground plane of localization space , which is assumed to be the @xmath44 plane . the electromagnetic volume current densities
are usually represented by equivalent surface current densities , which physically relate to the bounded charges and currents in terms of polarization and magnetization @xcite . * volume scatterer(s ) or volume noise with bounded spatial support @xmath45 with @xmath46 . here
, @xmath47 denotes the spatial support carrying electric currents with volume density @xmath48 , and @xmath49 denotes the spatial support carrying magnetic currents with volume density @xmath50^{-}$ ] .
we assume this noise to be negligible in our setup as we identify it with the thermal noise in all contrasting media in the measuring chamber .
* surface scatterer(s ) or surface noise with bounded spatial support @xmath51 with @xmath52 . here
, @xmath53 denotes the surface boundary carrying electric currents with surface density @xmath54 , and @xmath55 denotes the surface boundary carrying magnetic currents with surface density @xmath56^{-}$ ] .
the surface noise is assumed to be caused by surface roughness of the ground floor as described by @xcite .
* a receiver with bounded spatial support @xmath57 in which the electric field strength @xmath58 and the magnetic field strength @xmath59^{-}$ ] are accessible to measurement through the received time - averaged power flow @xmath60^{- } e_{q } dt \text{.}\ ] ] the receiver measures the power flow of an unmodulated carrier with an observation time of @xmath61 , long compared to the coherence time of @xmath6 @xcite .
the mapping sources @xmath62 field is unique if the physical condition of causality is invoked . for the received signal at the point , @xmath63 , of observation , the relation between the em vector and scalar potentials , @xmath64^{- } \right\}$ ] , of the received power flow and the real em currents on the surfaces of the bounded sources is given by @xcite @xmath65^{- } \right\}(\boldsymbol{x } , t ) = g(\boldsymbol{x},t ) \stackrel{(\boldsymbol{x } ) } { * } \stackrel{(t ) } { * } \left\ { j_{k } , \left [ k_{i , j } \right]^{- } \right\}(\boldsymbol{x } , t ) \text{,}\ ] ] where the spatial and temporal convolutions extend over the surfaces of the spatial - temporal supports of the pertaining bounded primary and induced sources located at @xmath66 in the measurement configuration . according to section [ sec : mc ] , the spatial supports are the surfaces of the transmitter and of the ground floor . in two dimensions ,
the surfaces become circumferences of their cylindrical cross sections .
equation links the extended sources to the well - known retarded potentials . in our measurement configuration , where the polarization of transmitted and received signals
are directed perpendicular to the @xmath44 plane , the array of green s functions that couples the sources to their respective potentials takes on the scalar form @xmath67 in , @xmath68 is the ( 3 + 1)-space - time dirac distribution operative at @xmath69 and @xmath70 @xmath71 d \frac{\omega_{0}}{2 \pi } \text{.}\ ] ] the bandwidth , @xmath72 , of temporal frequency spectrum of the hardware is usually mathematically represented by its inverse , the coherence time , @xmath73 , as described by @xcite @xmath74 where we assume a narrow gaussian line shape relative to the carrier frequency as is the case for quasi - monochromatic radiation @xmath75 time and position are connected by the constant speed of light , @xmath76 , as are angular temporal frequency and wavelength @xmath77 @xmath78 to simplify the notation , we represent this time dependence of the source currents in the dirac distribution by limiting the integration interval in to the frequency bandwidth , @xmath72 , @xmath79 \\
\text{sinc } \left ( \delta \omega_{0 } \left(t - \frac{\left| \boldsymbol{x } \right|}{c_{0}}\right ) \right ) \frac{\delta \omega_{0}}{2 \pi}\end{gathered}\ ] ] for @xmath80 , the slowly varying envelope of the sinc - function approaches unity , and we end up with a plane - wave representation leading to a spherical wave in . the sinc - function can be replaced by a normalized gaussian or lorentzian as is applied in @xcite . for our work ,
the only time dependence that is important is that our observation time is long compared to the coherence time so that all phase information is lost .
the convolution integral over the extended spatial supports in assumes its far - field approximation when the point of observation , @xmath63 , is far away from any part of the extended spatial supports , the coordinates of which are denoted by @xmath66 , so that @xmath81 where @xmath82 denotes the unit vector in the direction of observation , @xmath63 .
in the far - field approximation , the potentials in are replaced by the far - field electromagnetic field strengths @xcite with the given polarization along the z - axis @xmath83 where the tilde , @xmath84 , denotes the sum of a macroscopic average of all primary and secondary fields including a small noise term due to the surface roughness of the ground floor @xmath85 in , the far - field green s function is given by substituting in and in .
the transverse far - field magnetic field strength , @xmath86^{-1}$ ] , is directed perpendicular to the z - axis and @xmath82 and equals the far - field electric field strength multiplied by @xmath87 . in the far - field approximation ,
the time - averaged power flow of becomes inversely proportional to @xmath88 with the help of , so that without any noise , reduces to @xmath89 in , @xmath90 denotes the absolute value of the time - averaged poynting vector given by , and @xmath91 denotes the reference power flow at reference far - field distance of , say , @xmath92 m .
equation can be read as the non - logarithmic gain model . in the inverse - source problem formulation pertaining to our measurement configuration
, information is extracted from an ensemble of measured , time - averaged power - flow signals .
this information reveals the nature and location of the scattering volume and surface sources in .
the mapping of the @xmath93 is known to be non - unique .
a detailed analysis in this respect can be found in @xcite . in radio localization
, an algorithm is used that is expected to lead to results with a reasonable degree of confidence .
such an algorithm is based on the iterative minimization of the norm of the mismatch in the response between an assumed propagation model for each individual measurement and an ensemble of observed power - flow signals .
the influence of noise can be accounted for by using an input power - flow signal perturbed by an additive noise signal as is usually done in the log - normal shadowing model ( lnsm ) .
for independent , uncorrelated noise , reduces to the non - logarithmic gain model @xcite @xmath94^{1/2}/10 \ll 1$}.\end{gathered}\ ] ] in , @xmath95 denotes the gaussian variable of the independent zero - mean noise with standard deviation @xmath96^{1/2}$ ] . under the condition that the perturbation is small , we obtain @xmath97^{1/2}/10 \ll 1$}.\ ] ] the linearization of is established by recursive expansion . with this linearization of the perturbed , time - averaged power flow with added noise ,
the nature of the correlations can be derived from , , and and then be combined with and .
we apply the rationale used by van cittert and zernike . according to this rationale , each surface element on the contrasting surfaces
acts as a spatially incoherent point source represented by the green s function of .
the contrasting surfaces observed by the receiver in the far - field in our localization setup are the surfaces of the primary and secondary sources .
the primary source is the blind transmitter , and the main secondary source in our measuring chamber is the ground floor .
the wave vector of the primary source is fixed in size , i.e. @xmath98 , as the primary source is not extended and can be considered as a point source .
the wave vectors @xmath98 of the plane wave amplitudes originating from secondary stochastic point sources induced under grazing incidence all over the ground floor are the same in size .
all these scattered plane waves that are directed towards the vertical antenna of the receiver are again directed parallel to the ground floor .
as the vertical receiving antenna has a circular cross - section , the projections of all these wave vectors on the planes tangential to this cross section give rise to a spatial - frequency bandwidth of @xmath99 .
the expectation value , @xmath100 , of the superposition of these plane waves in the far field at a slightly displaced position , @xmath101 , all of the same amplitude but with random phases follows from , and @xmath102 = \mathbb{e } \left [ g^{\infty}(\boldsymbol{x } + \delta \boldsymbol{x } ) \stackrel{(\boldsymbol{x } + \delta \boldsymbol{x } ) } { * } \delta j_{z}^{s}(\boldsymbol{x } ) \right ] \nonumber \\ & = \mathbb{e } \left [ \exp\left [ i k_{0 } \widehat{\boldsymbol{x } } \cdot \delta \boldsymbol{x } \right ] \delta e_{z}^{\infty}(\boldsymbol{x } ) \right ] \nonumber \\ & = \mathbb{e } \left [ \exp\left [ i k_{0 } \widehat{\boldsymbol{x } } \cdot \delta \boldsymbol{x } \right ] \right ] \mathbb{e } \left [ \delta e_{z}^{\infty}(\boldsymbol{x})\right ] \text { , with } \frac{|\delta \boldsymbol{x}|}{|\boldsymbol{x}| } \ll 1 \text{,}\end{aligned}\ ] ] where we have taken out the time dependence to simplify the notation . in , the third equal sign results from the spatial invariance of the propagating medium and from the fact that the variance in vertical surface roughness is statistically independent of horizontal correlations in the far - field .
the ensemble average of all individual measurements over space of @xmath103 $ ] can be replaced by an average over the fourier conjugated space because of the assumed wide - sense stationary ( wss ) random noise process @xcite . in one dimension
, the expectation value of @xmath103 $ ] equals the sinc - function @xmath104 \right ] = \frac{1}{2 k_{0 } } \int_{-k_{0}}^{k_{0 } } \exp\left [ i k \widehat{\boldsymbol{x } } \cdot \delta \boldsymbol{x } \right ] dk \\ = \text{sinc}(k_{0 } \delta x ) \text{,}\end{gathered}\ ] ] and in two dimensions the jinc - function @xmath105 , with @xmath106 .
the first - order spatial correlation function of the perturbed electric field strengths follows from by multiplying its deterministic form for @xmath107 by @xmath108 and by neglecting the second - order terms of the perturbed field strength .
we then take the expectation value @xmath109 } { \mathbb{e } \left [ \left| e_{z}^{\infty}(\boldsymbol{x } ) \right|^{2 } \right ] } \cong\\ 1 + \frac{\mathbb{e } \left [ \delta \left| e_{z}^{\infty}(\boldsymbol{x } ) \right|^{2 } \right]}{\mathbb{e } \left [ \left| e_{z}^{\infty}(\boldsymbol{x } ) \right|^{2 } \right ] } \text{sinc}(k_{0 } \delta x ) \text{,}\ ] ] where the superscript star , @xmath110 , denotes complex conjugate and the ratio in the right member denotes the normalized variance of the perturbed field .
the sinc- and jinc - functions have their first zeros at @xmath111 and @xmath112 , which are called the diffraction limit or rayleigh criterion in one and two dimensions .
they can be considered as the far - field correlation functions for each individual measurement with a spatial correlation length that equals the diffraction limit .
equation is equivalent with the van cittert - zernike theorem @xcite .
this theorem links far - field spatial correlations of noise to diffraction .
the diffraction limit acts as a fundamental lower bound on the spatial correlation or coherence region for time scales @xmath113 , and hence , as an upper bound on the localization precision as we will see in section [ sec : model ] .
equations - generally allow for extracting the phases from the em field measurements as long as the timescales are short relative to an oscillation period @xmath114 . under those time scales ,
the van cittert - zernike theorem is equivalent with the wiener - khintchine theorem . using and
, we extend the perturbed input signals of and from independent to second - order spatially correlated noise to @xmath115 } { \mathbb{e } \left [ p^{2}(\boldsymbol{x } ) \right ] } \cong 1+\frac{\mathbb{e}\left [ \delta \left| e_{z}^{\infty}(\boldsymbol{x } ) \right|^{4 } \right ] } { \mathbb{e}\left [ \left| e_{z}^{\infty}(\boldsymbol{x } ) \right|^{4 } \right ] } \text{sinc}^{2}(k_{0 } \delta x ) \nonumber \\
\cong 10^{\frac{\mathbb{e}\left[x^2\right ] \text{sinc}^{2}(k_{0 } \delta x)}{10 } } \text { , for } \mathbb{e}\left[x^2\right]^{0.5}/10 \ll
1 \text { , } t_{m } \gg \tau_{c } \text{.}\end{aligned}\ ] ] the normalized scattering cross - sections in and resulting from surface roughness equal the measurement variance @xmath96 $ ] at each individual reference - radio position .
as we shall see in the next subsection , this measurement variance has a fundamental lower bound within the time scale of our measurements ( @xmath113 ) .
although the correlation function derived only holds for correlations in the close neighborhood of each point of observation , can be extended to larger distances when the path losses are accounted for as is usually done in lnsm .
the correlation function adopted in the literature is an exponentially decaying function @xcite .
reference @xcite notes that the exponentially decaying correlation function does not satisfy the propagation equations of em theory .
our correlation functions hold for each individual measurement position of the reference radio .
the extension to an ensemble of measurement positions with a derivation of the measurement variance of this ensemble is given in section [ sec : model ] .
section [ sec : model ] investigates two signal models for this extension and compares the computed measurement variances for the two correlation functions of power flows as a function of the sampling rate .
equation contain two products of the fourier pairs [ angular frequency @xmath116 , time @xmath35 ] and [ spatial frequency @xmath118 , position @xmath119 .
these two sets of conjugate wave variables naturally lead to an uncertainty in physics called the heisenberg uncertainty principle . hence , uncertainty is a basic property of fourier transform pairs as described by @xcite @xmath120 and @xmath121 equations and are the well - known uncertainty relations for classical wave variables of the fourier pairs [ energy , time ] and [ momentum , position ] . in these fourier pairs ,
energy and momentum are proportional to angular frequency and wave number , the constant of proportionality being the reduced planck constant .
although these uncertainty relations originally stem from quantum mechanics , they fully hold in classical mechanics as bandwidth relations . in information theory , they are usually given in the time domain and are called bandwidth measurement relations as described by @xcite . in fourier optics , they are usually given in the space domain as described by @xcite .
heuristically , one would expect the lower bounds on uncertainty in time , @xmath122 , and position , @xmath123 , to correspond to half an oscillation cycle , as half a cycle is the lower bound on the period of energy exchange between free - space radiation and a receiving or transmitting antenna giving a stable time - averaged poynting vector .
when one defines localization performance or precision as the inverse of the rmse , one would expect the localization precision of the ensemble not to become infinitely large by continuously adding measurements .
when the density of measurements crosses the spatial domain of spatial coherence as computed from , the power - flow measurements become mutually dependent .
the link between uncertainties and fisher information was first derived by @xcite . as we show in section [ sec : experimental ] , our experiments reveal convincing evidence for this link .
the first five subsections describe the signal model , mle and crlb usually applied in the field of rss - based radio localization for independent noise @xcite and cross - correlated noise @xcite .
section [ sec : bienayme ] reconciles diffraction and the nyquist sampling theorem with the crlb using cross - correlated noise and bienaym s theorem .
in addition , it introduces the notion of the maximum number of independent rss measurements and relates this to fisher information . in section [ sec : sim ] , simulations verify that the estimator is unbiased and efficient in the setup as described in section [ sec : experimental_setup ] .
we extend the notations introduced in section [ sec : experimental_setup ] and [ sec : theory ] that describe our measurement configuration with the bold - faced signal and estimator vectors usually employed in signal processing .
the experimental setup consists of a reference radio that measures power at @xmath2 positions @xmath124 of an unmodulated carrier transmitted by the blind radio . at each position
, the reference radio performs @xmath125 repetitive multiplexed power measurements @xmath12 .
these measurements are used to estimate the blind radio position , which is located at the origin @xmath126 .
we adopt the empirical lnsm for modeling the power over distance decay of our rss measurements . as the cross - sections of the blind transmitter and reference receiver
are given and equal , the power as well as the power - flow measurements are assumed to satisfy the empirical lnsm @xcite @xmath127 where @xmath128 in and , @xmath129 identifies the power - flow measurement .
@xmath130 denotes the ensemble mean of power - flow measurements at far - field distance @xmath131 in dbm .
@xmath132 represents the power flow at reference distance @xmath133 in dbm .
@xmath134 represents the path - loss exponent .
@xmath135 represents the noise of the model in db due to fading effects .
@xmath135 follows a zero - mean gaussian distribution with variance @xmath136 and is invariant with distance .
equations and are equivalent with the @xmath137 of , where the path - loss exponent equals two .
one usually assumes spatially independent gaussian noise @xcite . in @xcite ,
the cross - correlation between power - flow measurements is independent of wavelength and is modeled with an exponentially decaying function over space by @xmath138 \text{,}\ ] ] where @xmath139 denotes the correlation length .
section [ sec : noise ] shows that the lower bound on correlation length depends on the wavelength and equals half the mean wavelength , @xmath140 . in section [ sec : noise ] , shows that the cross - correlation function of power - flow measurements takes on the form of the diffraction pattern @xcite @xmath141 in section [ sec : crlb ] , we show the implications of using the cross - correlation function of and the cross - correlation function of that satisfies the propagation model derived from first principles .
the non - linear least squares problem assuming the lnsm with correlated noise is denoted by @xmath142 where @xmath143 denotes the vector of power - flow measurements @xmath144 \text{,}\ ] ] where @xmath145 denotes the vector of power - flow measurements calculated by the lnsm as expressed by @xmath146 \text{,}\ ] ] where @xmath147 denotes the vector of distances between the reference radio and the blind radio @xmath148 \text{.}\ ] ] here , @xmath130 denotes power flow at far - field distance @xmath149 as expressed by . in
, @xmath150 denotes the set of unknown parameters , where @xmath151 $ ] .
@xmath152 denotes the covariance matrix .
usually power - flow measurements are assumed to be independent so that @xmath153 , where @xmath154 is the identity matrix @xcite . in case of correlated noise @xcite ,
the elements in the covariance matrix are defined by @xmath155 , where @xmath156 where @xmath157 denotes the correlation distance between reference radio positions @xmath129 and @xmath158 .
equation denotes the correlation coefficients as expressed by and .
usually , the three parameters @xmath159 , @xmath134 , @xmath160 $ ] of the lnsm are calibrated for a given localization setup @xcite .
the blind radio position is assumed to be known when the lnsm is calibrated .
the lnsm assumes that @xmath161 is equal for each rss measurement , so that @xmath132 and @xmath134 can be estimated independent of @xmath161 @xmath162 = \arg \min_{\boldsymbol{\theta } = [ p_{r_{0 } } , \eta ] } v(\boldsymbol{\theta } ) \text{.}\ ] ] in , @xmath163 is defined as the standard deviation between the measurements and the fitted lnsm using @xmath164 and @xmath165 .
equation gives for all our power - flow measurements the best fit when the lnsm parameters are calibrated at @xmath166 we use these calibrated lnsm parameter values to calculate the bias and efficiency of our estimator from simulations .
in addition , we use these calibrated lnsm parameter values to calculate the cross - correlations in the far - field and crlb .
we use the mle as proposed by @xcite , with the physical condition of causality invoked on @xmath134 and with boundary conditions on @xmath167 , to estimate the blind radio position @xmath168
= \arg \min_{\boldsymbol{\theta } = [ x , y , p_{r_{0 } } , \eta ] } v(\boldsymbol{\theta } ) \nonumber \\ & \text{subject to } \nonumber \\ & \eta>0 \nonumber \\ & -1 \leq x \leq 2 \nonumber \\ & -2 \leq y \leq 1 \text{,}\end{aligned}\ ] ] our estimator processes the noise as independent , so that @xmath169 .
section [ sec : sim ] shows that this estimator is unbiased and efficient in our simulation environments with uncorrelated and correlated noise .
crlb analysis provides a lower bound on the spreads of unbiased estimators of unknown deterministic parameters .
this lower bound implies that the covariance of any unbiased estimator , @xmath170 , is bounded from below by the inverse of the fisher information matrix ( fim ) @xmath171 as given by @xcite @xmath172 in , @xmath150 represents the set of unknown parameters , and @xmath173 represents the estimator of these parameters . in case of multivariate gaussian distributions ,
the elements of the fim are given by @xcite @xmath174_{a , b } = \left [ \frac{\partial \mathbf{\bar{p}}}{\partial \theta_{a } } \right]^{t } c^{-1 } \left [ \frac{\partial \mathbf{\bar{p}}}{\partial \theta_{b } } \right ] + \\
\frac{1}{2 } \left [ c^{-1 } \frac{\partial c}{\partial \theta_{a } } c^{-1 } \frac{\partial c}{\partial \theta_{b}}\right]\text{.}\end{gathered}\ ] ] in our case , the covariance matrix is not a function of unknown parameter set @xmath150 so that the second term equals zero .
the elements of the @xmath175 fim associated with the estimator defined by , @xmath176 $ ] , are given in the independent measurement case by @xcite and in the correlated measurement case without nuisance parameters by @xcite .
the elements of @xmath171 consists of all permutations in set @xmath150 .
after some algebra , @xmath177 reduces to @xmath178 where @xmath179 we use to calculate the crlb of the mle expressed by using the calibrated lnsm parameter values given by : @xmath180 $ ] .
the crlb on rmse for unbiased estimators is computed from @xmath181 here @xmath182 represents the trace of the matrix .
we use the calibrated lnsm parameters as given by .,title="fig:",scaledwidth=50.0% ] fig . [
fig : crlb ] shows the lower bound on the rmse calculated by and as a function of the number of rss measurements .
when we assume independent rss measurements , @xmath183 , the rmse decreases to zero with an ever increasing number of independent rss measurements .
this is in accordance with the theoretical bound analysis presented by @xcite .
hence , there is no bound on localization precision with increasing sampling rate .
when we account for the spatial cross - correlations between rss measurements , @xmath184 , the bound on radio localization precision reveals itself when sufficient rss measurements are available .
we evaluate the crlb with the cross - correlation function expressed by the diffraction pattern in and the exponentially decaying cross - correlation function as expressed by .
they basically converge to the same bound on localization precision . apparently , the bound on localization precision depends on the bound on the correlation length rather than on the form of the cross - correlation functions .
the determinant of the covariance matrix starts to decrease when correlations start to increase , until the fisher information and thus localization precision stabilizes on a certain value . when the cross - correlations are equal to the diffraction pattern as expressed by , the covariance matrix becomes singular when the sampling rate goes above a certain threshold , which equals the inverse of the diffraction limit or the nyquist sampling rate .
hence , the multivariate gaussian distribution becomes degenerate and the crlb can not be computed without regularization .
this section connects diffraction to the sampling theorem , covariance and fisher information using bienaym s theorem @xcite .
bienaym s theorem offers a statistical technique to estimate measurement variance of correlated signals .
it states that when an estimator is a linear combination of @xmath15 measurements @xmath185 @xmath186 the covariance of this linear combination of measurements equals @xmath187 and is equivalent to the measurement variance . in
, @xmath188 is a weighting factor , @xmath189 represents the standard deviation of measurement @xmath129 , and @xmath190 represents the correlation coefficient between measurement @xmath135 and @xmath191 . when all measurements have equal variance @xmath192 and equal weights @xmath193 , reduces to @xmath194 in , @xmath195 is the spatially averaged correlation of @xmath196 measurement pairs and is given by @xmath197 and @xmath198 represents the number of measurements that effectively decreases estimator covariance and is given by @xmath199 @xcite shows in the time domain that the cross - correlation of bandwidth limited fluctuations on signals with a periodic wave character is proportional to the point - spread - function ( psf ) of a measurement point . in the space domain ,
the cross - correlation of far - field power - flow measurements takes on the form expressed by the diffraction pattern in .
in addition , @xcite shows that the radius , @xmath200 , of the first zero @xmath201 equals the inverse of the minimum ( nyquist ) sampling rate .
the sampling theorem determines the sampling rate that captures all information of a signal with finite bandwidth . for signals with a periodic wave character measured in the far field ,
noise is cross - correlated over the far - field spatial coherence region of the signal .
this far - field coherence region assumes the far - field diffraction pattern and is equivalent to the psf of power - flow measurements @xcite . the lower bound on the far - field region of spatially correlated signals and noise is equal to the diffraction limit @xcite and to the inverse of the upper bound on spatial frequencies divided by @xmath202 , i.e. @xmath203 . the finite spatial - frequency bandwidth of the far field relates diffraction to the sampling theorem @xcite and whittaker - shannon interpolation formula or cardinal series @xcite .
hence , this coherence region determines the maximum number of independent measurements over a finite measuring range .
this maximum number of independent measurements is also known as the degrees of freedom @xcite , and is given by @xmath198 in .
we decompose into two factors @xmath204 where @xmath205 represents the ratio of the total number of measurements and effective number of measurements , and @xmath206 represents the covariance of the estimator assuming that all measurements are independent .
the global spatial average of correlation coefficients , @xmath195 , as given by can be computed using the correlation function for any localization setup by assuming an ever increasing sampling rate . for our localization setup of fig .
[ fig : measurement_setup ] , we compute @xmath207 processing all @xmath2 rss measurements over space . substituting this value in gives @xmath208 . to account for the effects of spatial correlations on the covariance of non - linear unbiased estimators , we heuristically rewrite into the following postulate @xmath209 where @xmath171 is equal to the fisher information assuming independent measurements .
our postulate rigorously holds when the variance and thus the fisher information of each wave variable at each measurement point is equal , assuming that the crlb holds as defined in .
our postulate of is in agreement with the concept that the degrees of freedom of signals with a periodic wave character is independent of the estimator .
postulate is also in line with that crlb efficiency is approximately maintained over nonlinear transformations if the data record is large enough @xcite .
hence , is a good performance indicator when the correlation length is small relative to the linear dimensions of localization space .
[ fig : crlb ] shows the lower bound on the rmse calculated by and using the two cross - correlation functions expressed by and .
the solid curves represent the results of .
our approach expressed in differs less than 1 mm from the crlb for signals with correlated gaussian noise @xcite over the entire range shown by fig .
[ fig : crlb ] .
the influence of correlated measurements on the crlb is apparent for @xmath210 as converges asymptotically to roughly half the mean wavelength for both cross - correlation functions .
hence , the correlation length and thus the finite number of independent rss measurements determines the bound on localization precision .
note that the effective number of measurements converges to practically identical values for both cross - correlation functions , @xmath208 .
the sinc - squared cross - correlation function expressed by is bandwidth limited , so that the covariance matrix becomes degenerative at roughly the nyquist sampling rate .
the exponentially decaying cross - correlation function expressed by is not bandwidth limited , so that the covariance matrix stays full rank . in this case , the fisher information decreases per individual rss measurement with increasing sampling density until the localization precision stabilizes .
when @xmath211 , provides practically the same results as the crlb for independent measurements ( @xmath2122 mm difference in all cases ) .
hence , one can assume independent measurements as long as the inverse of the sampling rate is large compared to the correlation length .
the bound on estimator covariance becomes of interest when the density of measurements surpasses the bound imposed by the spatial coherence region of far - field radiation of the transmitters .
when the number of measurements goes to infinity , reduces to @xmath213 which does not approach zero .
hence , bienaym s theorem reveals the link of measurement variance to diffraction , the sampling theorem , and fisher information .
we performed @xmath214 simulation runs using the measurement setup described in fig .
[ fig : measurement_setup ] to quantify ( 1 ) the bias and ( 2 ) the efficiency of our estimator in .
we performed two sets of simulations , one with independent noise and one with correlated noise expressed by .
we did not perform simulations with cross - correlation function , because the covariance matrix is singular above the nyquist sampling rate and thus for our measurement setup .
note that these simulations do not provide an experimental validation of the model used .
an estimator is unbiased if @xcite @xmath215
we are interested in the bias of the estimated blind radio position @xmath216 , so we define the bias as @xmath217 simulations show that the bias is of the order of @xmath218 mm for independent and correlated noise .
estimator efficiency is defined as the difference between the covariance of the estimator and crlb .
we quantify this difference by @xmath219 simulations show that the estimator and crlb differs at most by @xmath220 mm for both independent and correlated noise .
hence , our estimator in has a negligible bias and a high efficiency in our simulation environments with independent and correlated noise .
this section presents the experimental results of the two - dimensional measurement setup described in section [ sec : experimental_setup ] .
the first subsection presents the far - field spatial cross - correlation function of the rss signals and estimate the spread of these spatial correlations .
we then determine the spatial fourier transform of these cross - correlations to look for an upper bound at a spatial frequency of @xmath221 , in line with the spatial resolution being bounded by the diffraction limit .
finally , we determine the global rmse of and show its asymptotic behavior to the diffraction limit with the increase of the density of sampling points . we compare this experimentally determined rmse with the rmse determined by the crlb for independent and correlated noise .
we show that the crlb for independent noise underestimates the rmse computed from our measurements .
[ fig : coherence ] shows the spatial cross - covariance function between power flows as a function of distance in wavelengths .
the spatial cross - covariance function is calculated using the deviations from the lnsm expressed by and calibrated propagation parameters expressed by .
hence , spatial cross - covariance and thus cross - correlations are distance independent by cross correlating the deviations from the calibrated lnsm . fig .
[ fig : coherence ] shows that the spatial cross - correlations go to a minimum over a distance of roughly half the mean wavelength , which corresponds to the diffraction limit .
the small difference between the black and red curve indicates that noise resulting from repetitive multiplexed measurements over time is negligible compared to the cross - correlated noise in the far field .
[ t ] million rss measurements .
the red curve shows the measured cross - correlations using @xmath20 rss measurement per reference radio position.,title="fig:",scaledwidth=48.0% ] [ t ] .
theoretically , it is represented by tri(k ) , as it is the fourier transform of .
the vertical line represents the spatial frequency that corresponds to the diffraction limit , which forms an upper bound in the spectrum.,title="fig:",scaledwidth=48.0% ] in the case of fisher information and thus crlb analysis , information is additive when measurements are independent @xcite .
one usually assumes signal models with independent noise@xcite .
hence , localization precision increases with an ever increasing amount of independent measurements over a finite measuring range
. however , when measurements are correlated , information and localization precision gain decrease with increasing correlations .
[ fig : coherence ] shows that when space - measurement intervals become as small as the diffraction limit , measurements become spatially correlated and mutually dependent .
our measurements show that rss signal measurements are spatially correlated over a single - sided region of roughly half the mean wavelength .
it corresponds to the diffraction pattern expressed by derived in section [ sec : noise ] .
therefore , increasing reference radio density beyond one per half a wavelength has a negligible influence on fisher information gain and thus on localization precision .
our experimental results in the next subsection confirm this . in case of independent measurements , fig .
[ fig : coherence ] would show an infinitely sharp pulse ( dirac delta function ) .
[ fig : fft ] shows the measured power spectrum of cross - correlated noise in rss signals , @xmath222 , i.e. the spatial fourier transform of the cross - covariance of fig .
[ fig : coherence ] .
theoretically , it is represented by the spatial fourier transform of .
this figure shows that the energy is mainly located in lower spatial frequencies and it diminishes over a single - sided interval of @xmath223 .
this upper bound corresponds to the diffraction limit .
the nyquist sampling rate provides an estimate of the minimum sampling rate to fully reconstruct the power - flow signal over space without loss of information , which equals the single - sided bandwidth of our power spectrum .
the vertical black line represents the spatial frequency associated with the nyquist sampling rate , which equals @xmath224 . our experimental results in the next subsection
confirm this by showing that the localization precision does not increase by sampling beyond this sampling rate . in case of independent measurements , fig .
[ fig : fft ] would show a uniform distribution .
[ fig : localization ] shows the experimental results of our measurement setup described in section [ sec : experimental_setup ] .
localization precision is given as the inverse of rmse , which is computed from .
[ t ] independent rss measurement per reference radio position ( green curve).,title="fig:",scaledwidth=48.0% ] the red curve in fig .
[ fig : localization ] shows the rmse as a function of the number of measurements per wavelength .
the rmse decreases with increasing number of rss measurements over space until sufficient measurements are available . at a critical measurement density ,
the bound on localization precision becomes of interest .
the rmse ( red and black curves ) converges asymptotically to roughly half the mean wavelength as one would expect from the diffraction limit .
the rmse represented by the red curve is based on processing all @xmath225 rss signal measurements per reference radio position .
the rmse represented by the black curve is based on processing one rss signal measurement per reference radio position .
the negligible difference between the red and black curves shows that the number of repeated rss signal measurements per reference radio has a negligible influence on the rmse .
the crlb for independent measurements starts deviating from the rmse ( red and black curves ) when the sampling rate is increased beyond one rss signal per half the mean wavelength ( see black dotted curve ) , as one would expect from the diffraction limit and the nyquist sampling rate over space @xcite .
spatial correlations between rss signals increase rapidly with increasing rss measurement density beyond one sample per half the mean wavelength ( fig .
[ fig : coherence ] ) .
correlated rss signals can not be considered as independent .
@xcite has shown that fisher information is upper bounded by uncertainty in line with .
as @xmath226 is upper bounded in spatial frequencies by @xmath227 , the spread or uncertainty in position , @xmath228 , is lower bounded .
hence its inverse is upper bounded , as is fisher information .
coherence and speckle theory have shown that uncertainty is lower bounded by the diffraction limit .
hence , the crlb at a sampling density of one sample per half the mean wavelength ( vertical dotted black curve ) should equal the measured bound on rmse .
we define the measured bound on rmse as the rmse processing all 60 million measurements .
our measurements show that the difference between this crlb and the measured bound on rmse is @xmath229-@xmath230% ( 1 mm ) . hence
, our experiments validate the theoretical concepts introduced by @xcite .
our experiments reveal evidence that the crlb can not be further decreased by increasing the number of measurements . on the other hand , at @xmath6 rss signal measurements per wavelength , the rmse is a factor of four higher than the one calculated by the crlb for independent noise .
this difference can not be explained by the difference between the covariance of the estimator and the crlb ( see section [ sec : model ] ) .
the difference can , however , be explained by calculating the degrees of freedom of our localization setup using bienaym s theorem and the lower bound on correlation length .
when one substitutes this lower bound in the and substitutes in , one obtains the asymptotic behavior of the crlb with correlated noise as is shown in fig .
[ fig : crlb ] .
the measured asymptote in fig . [ fig : localization ] deviates @xmath229-@xmath230% from bienaym s theorem and from the crlb for correlated noise .
all @xmath4 million power measurements and matlab files are arranged in a database at linkping university @xcite .
our novel localization setup of using reference radios on the circumference of a two - dimensional localization area instead of setting them up in a two - dimensional array worked well .
this implies that one does not have to know the phases and amplitudes on the closed surfaces around extended sources to reconstruct the positions of these extended sources .
our experiments show that it suffices to measure time - averaged power flows when a localization precision of about half a wavelength is required ( in our case @xmath21 cm ) .
time averaging on a time scale that is long compared to the temporal coherence time is usually employed in rss localization , so that phase information is lost .
we expect such a setup to work well when all radio positions are in los . for a practical implementation in nlos environments
, we refer to @xcite .
the propagation equations of section [ sec : theory ] are than applied to design an efficient setup and algorithm to locate the blind radios . in section [ sec : theory ]
, we distinguished between primary and secondary cross - correlated noise in the far field . in our experimental setup , where signal levels are large compared to noise levels , it is reasonable to neglect thermal and quantum noise . scattering from the surface roughness of the primary source can not provide the the spatial frequency bandwidth of fig .
[ fig : fft ] because of the small size of the transmitting antenna .
we expect that the cross - correlated noise originates from the surface roughness of the large area between the transmitter and receiver .
hence , our @xmath2 time - averaged rss measurements are correlated over space in line with the derived correlation model of as was verified by our experiments .
noise can originate from a variety of sources . in our setup
, the noise level of the antenna plus electronics on a typical @xmath231 radio is about @xmath232dbm below signal level as specified in @xcite .
the reproducible ripples on our rss signals originated in part from small interference effects from undue reflections from mostly hidden metallic obstacles in the measuring chamber . such obstacles like reinforced concrete pillars could not be removed . in a real indoor office environment , these interference effects usually dominate rss signals . in these environments the cross - correlated noise in rss signals is still determined by the diffraction limit , which in turn , determines the bound on localization precision .
multi - path effects and fading do not change the stochastic properties of the secondary extended sources where the noise is generated . as explained in the literature @xcite ,
such models are usually clarified by ray - tracing .
ray - tracing implies the geometrical - optics approximation , so that the wave lenghth is set to zero and diffraction effects are not considered .
when we set the correlation length of the exponentially decaying correlations as expressed by @xcite equal to the diffraction limit , the crlb converges to practically the same bound on localization precision .
it is remarkable that this bound is revealed by our simple experiment using various noise and signal models .
our experiments reveal evidence that the diffraction limit determines the ( 1 ) bound on localization precision , ( 2 ) sampling rate that provides optimal signal reconstruction , ( 3 ) size of the coherence region and uncertainty , ( 4 ) upper bound on spatial frequencies and ( 5 ) upper bound on fisher information and lower bound on the crlb . with the mathematical work of @xcite , a rigorous link between fisher information , crlb and uncertainty was established . as uncertainty
has a lower bound according to coherence and speckle theory , so must fisher information have an upper bound , so that the crlb has a lower bound , all when the noise processes are assumed to be gaussian distributed .
our experiments were able to validate those theoretical concepts .
this is further confirmed by applying bienaym s theorem .
an interesting observation is that despite the fact that power flows have twice the spatial bandwidth than radiation flows determined by amplitudes , their spatial coherence regions are the same in size , as are their degrees of freedom .
our paper describes a novel experimental and theoretical framework for estimating the lower bound on uncertainty for localization setups based on classical wave theory without any other prior knowledge .
bienaym s theorem and existing crlb for signals with correlated gaussian noise reveal our postulate , so that the lower bound on uncertainty corresponds to the upper bound on fisher information .
our experimental results can not be explained by existing propagation models with independent noise .
it took almost three weeks in throughput time to perform the @xmath4 million measurements , generating @xmath233 repetitive multiplexed measurements .
we tried to minimize multi - path effects by avoiding the interfering influence of ground and ceiling reflections . making sure to minimize reflections of other metallic obstacles in the measuring chamber were challenging but could be overcome .
this allowed us to reveal a performance bound in theory and experiment for measurements without phase information .
our novel two - dimensional localization setup where we positioned the reference radios on the circumference of the localization area rather than spreading them out in a two - dimensional array over this area worked well in our los setup .
our measurements show that localization performance does not increase indefinitely with an ever increasing number of multiplexed rss signals ( space and time ) , but is limited by the spatial correlations of the far - field rss signals . when sufficient measurements are available to minimize the influence of measurement noise on localization performance , the bound on localization precision is revealed as the region of spatially correlated far - field radiation noise .
the determination of this region of spatial correlations is straightforward and can be directly calculated from rss signals . within this region of correlated rss signals , assumptions of independent measurements
are invalidated , so that the crlb for independent noise underestimates the bounds on radio localization precision .
this underestimation is removed by accounting for the correlations in the noise .
the bound on the correlations is given by the fundamental bound on the correlation length that we derived from first principles from the propagation equations .
the crlb is linked to the uncertainty principle as measurement variance is directly related to this principle as we showed .
the bounds on the precision of rss- and tof - based localization are expected to be equal and of the order of half a wavelength of the radiation as can be concluded from our experiments and underlying theoretical modeling .
sampling beyond the diffraction limit or the nyquist sampling rate does not further resolve the oscillation period unless near - instantaneous measurements are performed with the a priori knowledge that signal processing is assumed to be based on non - linear mixing .
future research is aimed at the inclusion of strong interference effects such as show up in practically any indoor environment .
the authors would like to thank adrianus t. de hoop , h. a. lorentz chair emeritus professor , faculty of electrical engineering , mathematics and computer sciences , delft university of technology , delft , the netherlands , for his comments on applying green s theorem as a propagation model in electromagnetic theory .
secondly , they would like to thank carlos r. stroud , jr .
of the institute of optics , rochester , ny , for sharing his insight on fourier pairs and uncertainties .
finally , they would like to thank gustaf hendeby and carsten fritsche of linkping university , sweden , for their comments .
35 g. toraldo di francia , `` resolving power and information , '' j. optical soc .
45 , no . 7 , 1955
. a. j. stam , `` some inequalities satisfied by the quantities of information of fisher and shannon , '' inform . and control , vol .
2 , pp . 101112 , 1959 .
a. papoulis , `` error analysis in sampling theory , '' proc .
ieee , vol .
54 , no . 7 , pp .
947 - 955 , july 1966 .
m. gudmundson , `` correlation model for shadow fading in mobile radio systems , '' ieee electronics letters , 21452146 , 7 nov .
h. hashemi , `` the indoor radio propagation channel , '' proc .
ieee , vol .
81 , no . 7 , pp .
943 - 968 , july 1993 .
s. kay , `` fundamentals of statistical signal processing : estimation theory , '' prentice hall , 1993 .
j. s. lee , i. jurkevich , p. dewaele , p. wambacq , a. costerlinck , `` speckle filtering of synthetic aperture radar images : a review '' , remote sensing reviews , vol .
313 - 340 , 1994 .
l. cohen , `` time - frequency analysis , '' prentice hall , 1994 .
e. s. keeping , `` introduction to statistical inference , '' dover publications , 1995 .
m. v. de hoop and a. t. de hoop , `` wavefield reciprocity and optimization in remote sensing , '' proceedings of the royal society of london , pp .
641682 , 2000 . m. a. pinsky , `` introduction to fourier analysis and wavelets , '' amer .
soc . , 2002 .
n. patwari , a. o. h. iii , m. perkins , n. s. correal , and r. j. odea , `` relative location estimation in wireless sensor networks , '' ieee trans .
signal process .
51 , no . 8 , pp . 21372148 , aug .
2003 . wireless medium access control ( mac ) and physical layer ( phy ) specifications for low rate wireless personal area networks ( wpan ) , 802.15.4 - 2003 , 2003 .
j. w. goodman , `` introduction to fourier optics , '' roberts and company publishers , 2004 .
j. w. goodman , `` speckle phenomena in optics : theory and applications , '' roberts and company publishers , 2006 .
n. patwari and a. o. hero iii , `` signal strength localization bounds in ad hoc and sensor networks when transmit powers are random , '' in ieee workshop on sensor array and multichannel processing , july 2006 .
r. a. malaney , `` nuisance parameters and location accuracy in log - normal fading models , '' ieee trans .
wireless commun .
, vol . 6 , pp .
937 - 947 , march 2007 .
n. patwari and p. agrawal , `` effects of correlated shadowing : connectivity , localization , and rf tomography , '' ieee / acm int .
conf . on inf .
processing in sensor networks , april 2008 .
f. gustafsson , f. gunnarsson , `` localization based on observations linear in log range , '' proc .
17th ifac world congress , seoul , 2008 .
y. shen , and m. z. win , `` fundamental limits of wideband localization
part i : a general framework , '' ieee trans .
theory , vol .
4956 4980 , oct . 2010 .
b. j. dil and p. j. m. havinga , `` rss - based self - adaptive localization in dynamic environments , '' int .
internet of things , pp .
5562 , oct .
l. mandel and e. wolf , `` optical coherence and quantum optics , '' cambridge university press , 2013 .
b. j. dil , `` lost in space , '' ph.d .
dissertation , dept .
, twente univ . ,
enschede , netherlands , 2013 .
a.t . de hoop , `` electromagnetic field theory in ( n+1)-spacetime : a modern time - domain tensor / array introduction , '' proceedings of the ieee 101 , pp . 434450 , 2013 .
a. herczyski , `` bound charges and currents , '' american journal of physics , vol .
3 , 202205 , 2013 . g. brooker , `` modern classical optics , '' oxford university press , 2014 .
t. szabo , `` diagnostic ultrasound imaging : inside out , '' 2014 .
j. w. goodman , `` statistical optics , '' wiley , 2015 .
p. lpez - dekker , m. rodriguez - cassola , f. de zan , g. krieger , and a. moreira , `` correlating synthetic aperture radar ( cosar ) , '' ieee trans .
geoscience and remote sensing , no .
99 , pp.1 - 17 , 2015 .
http://www.ti.com/lit/ds/symlink/cc2420.pdf , p. 55 , march 2015 .
b. j. dil , `` database of fundamental bound measurements and matlab files , '' linkping university , sweden , 2015 . | this paper experimentally and theoretically investigates the fundamental bounds on radio localization precision of far - field received signal strength ( rss ) measurements .
rss measurements are proportional to power - flow measurements time - averaged over periods long compared to the coherence time of the radiation .
our experiments are performed in a novel localization setup using 2.4ghz quasi - monochromatic radiation , which corresponds to a mean wavelength of 12.5 cm .
we experimentally and theoretically show that rss measurements are cross - correlated over a minimum distance that approaches the diffraction limit , which equals half the mean wavelength of the radiation .
our experiments show that measuring rss beyond a sampling density of one sample per half the mean wavelength does not increase localization precision , as the root - mean - squared - error ( rmse ) converges asymptotically to roughly half the mean wavelength .
this adds to the evidence that the diffraction limit determines ( 1 ) the lower bound on localization precision and ( 2 ) the sampling density that provides optimal localization precision .
we experimentally validate the theoretical relations between fisher information , cramr - rao lower bound ( crlb ) and uncertainty , where uncertainty is lower bounded by diffraction as derived from coherence and speckle theory .
when we reconcile fisher information with diffraction , the crlb matches the experimental results with an accuracy of 97 - 98% .
radio localization , cramr - rao bounds , fisher information , bienaym s theorem , sampling theorem , speckles , uncertainty principle . |
for over 50 years , measurements of @xmath0 pair production by energetic ions incident on emulsion detectors have yielded events with opening angles at the @xmath4 vertex larger than those expected @xcite for external pair conversion ( epc ) by photons @xcite . on the basis of relatively small data sets ,
el - nadi and badawy @xcite and de boer and van dantzig @xcite proposed in 1988 that such events might represent the decays of neutral bosons with masses considerably larger than generally expected for weinberg - wilczek axions .
since then , additional studies of relativistic heavy - ion interactions with emulsion @xcite have been interpreted in terms of massive neutral boson production .
while such observations have triggered interest @xcite and are mentioned in the review of particle physics section on axion searches @xcite , no comprehensive attempt has been made to assess whether these observations represent new phenomena worthy of further investigation or background effects resulting from well - characterized processes . in this paper , we revisit results from six accelerator studies using emulsion detectors , all of which report observations of @xmath0 pairs that are interpreted as evidence for the existence of neutral bosons with masses between 1.5 and 30 mev/@xmath1 and lifetimes between @xmath5 and @xmath6 s @xcite .
four of these studies @xcite were published after the 1988 proposal that such events might represent the decays of massive neutral bosons @xcite , and hence can be considered tests of this proposal .
the assumptions made regarding background processes are not , however , consistent across these six studies , and the reported neutral boson masses are not identical ; indeed jain and singh @xcite report a continuous spectrum of novel neutral bosons with masses up to 85 mev/@xmath1 .
we have therefore re - analysed these @xmath0 pair production data using a single set of assumptions regarding background processes that are consistent with the experimental data reported by each of these studies .
we also re - analysed the classic cosmic - ray data of anand @xcite and hintermann @xcite as comparisons .
the resulting background - subtracted data sets , taken together , raise the possibility of a spectrum of neutral bosons with energies between 3 and 20 mev/@xmath1 and lifetimes on the order of femtoseconds .
this mass - lifetime window , and in particular the existence of a 9.8 mev/@xmath1 neutral boson decaying to @xmath0 pairs , is consistent with measurements of anomalous internal pair production ( ipc ) in decays of excited states of light nuclei that suggest the existence of both unnatural parity ( pseudoscalar and axial - vector ) @xcite and natural parity ( scalar and vector ) neutral bosons @xcite .
the existence of neutral particles decaying to @xmath0 pairs in this mass - lifetime window was not investigated in early beam - dump measurements @xcite . due to the long beam dumps
, these experiments using high - energy electron and proton beams from laboratories including slac , orsay and fnal could only rule out lighter , longer - lived axion candidates ( @xmath7 ) .
results from a later experiment by bross et al @xcite , using a short beam dump with sensitivity to short - lived bosons with @xmath8 , partly overlap the present lifetime window but do not exclude it .
if multiple neutral particles with similar rest masses exist , it is possible that their individual effects on the electron and muon anomalous magnetic moments would cancel out , potentially explaining the lack of evidence for massive axion - like particles in the available magnetic - moment data .
we conclude that the observations reviewed and re - evaluated here support the existence of fewer massive neutral particles than have been previously claimed @xcite . however , we interpret these data as providing sufficient suggestive evidence for novel massive neutral bosons with femtosecond lifetimes to encourage further experimental investigation .
if confirmed , the existence of such massive neutral bosons would provide _ prima facie _ support for theoretical models that propose axion - like particles with masses in the mev to 10s of mev range @xcite , as are increasingly motivated by cosmological data @xcite .
fig.1 shows a compilation of @xmath0 pair - production events from both cosmic - ray @xcite and accelerator @xcite data sets for which electron energy and opening angle data are available .
the signature of a massive particle @xmath2 decaying to @xmath0 is an opening angle @xmath9 from the pair vertex larger than the borsellino @xcite angle @xmath10 , where @xmath11 = 0.511 mev/@xmath1 and @xmath12 is the total pair energy in mev , expected for epc .
emulsion detectors are ideal for detecting such events , as they allow accurate measurements of both the opening angle and the kinetic energies of the outgoing @xmath0 pair . as can be seen in fig.1 , a large fraction of the observed events fall below the borsellino line .
all such events are consistent with @xmath0 decay of a massless particle , i.e. with epc by photons . in the highest - energy reaction considered here ,
160 a@xmath13gev @xmath14pb on emulsion @xcite , 1,158 out of 1,220 recorded @xmath0 events ( 95% ) have opening angles below the borsellino line ( cf .
1f of @xcite ) and hence are _ prima facie _ consistent with epc ; these points are not plotted in fig .
1 as they would obscure the much smaller numbers of low - angle events recorded from lower - energy experiments .
it has been understood since the pioneering observations of hintermann @xcite that the effective - photon approximation reflected by the borsellino line significantly overpredicts the average opening angle and hence transverse momentum of epc - derived @xmath0 pairs at the pair - production threshold .
qed calculations of pair production cross - sections near threshold correctly reproduce both the observed transverse momenta and the near - threshold cross - section , confirming that the effective - photon approximation is accurate for apparent pair masses above the pair - production threshold @xcite .
previous analyses of these data have considered photons from @xmath15 decay to be the primary source of the epc background evident in fig . 1 @xcite ; however , photons have a mean free path in emulsion of roughly 5 cm @xcite , so photons from @xmath16 decay would be expected to efficiently escape the reaction zone . indeed , jain and singh @xcite estimate that only 73 @xmath0 pairs from @xmath15 decay , distributed isotropically , would be observed within the area of emulsion that they analyzed following 160 a@xmath13gev @xmath14pb bombardment , compared to 1,158 pairs with opening angles consistent with epc actually observed within the forward cone only .
a similar discrepancy between expected epc from the @xmath15 channel and observed events below the borsellino line is evident in the analysis of the 200 a@xmath13gev @xmath17s reaction on emulsion @xcite .
a reconsideration of background processes in these data sets is therefore necessary .
the production of photons as bremsstrahlung radiation following coulomb excitation of beam particles via interactions with target nuclei at large impact parameters has largely been neglected in previous analyses of these data sets , but can be expected to generate significant epc backgrounds in these reactions . the effective energy transfer of such interactions decreases by roughly an order of magnitude for every order of magnitude increase in impact parameter @xcite ; the cross section for projectile excitations in the 100 kev to 10 mev range , below the threshold for projectile breakup
, is therefore expected to be on the order of 10,000 @xmath18 the projectile breakup cross - section .
heavy nuclei coulomb - excited below 10 mev can be expected to de - excite by @xmath19-emission , with average @xmath19 multiplicities on the order of 10 per coulomb excitation event .
photons that are significantly aligned with the projectile momentum ( @xmath20 of the total ) will be lorentz boosted to energies up to the maximum effective photon energy of @xmath21 @xcite . for 200 a@xmath13gev
@xmath17s on emulsion , baroni _ et al . _
@xcite report an average of 2.4 projectile breakup events per meter of penetration ; hence one can expect about @xmath22 @xmath19 s due to bremsstrahlung from coulomb - excited projectiles per meter of target penetration for this reaction , with about @xmath23 @xmath19 s per meter in the forward cone . the nuclear - interaction mean free path for 200 a@xmath13gev @xmath17s in emulsion is about 100 mm @xcite , so one can expect on the order of 6,000 forward @xmath19 s per observed nuclear interaction .
about half of these @xmath19 s will convert within the first 5 cm from the breakup event ; one would therefore expect about 3 epc events per 50 @xmath24 m downstream of an identified 200 a@xmath13gev @xmath17s - emulsion interaction . in the case of the 160 a@xmath13gev @xmath25pb data set
, one would expect about 75 epc events per 50 @xmath24 m downstream of an identified interaction .
the experimental signature of such @xmath19-interactions would be a nearly - uniform distribution of @xmath0 events closely collimated to the pre - interaction projectile trajectory , attenuating linearly with distance @xmath26 from the projectile interaction as photons escape the viewing area of the tracking microscope .
figure 2 shows the distribution of @xmath0 events from 200 a@xmath13gev @xmath17s bombardment of emulsion as a function of distance @xmath26 from the observed projectile - emulsion reaction @xcite .
figure 2 can be compared with the very similar distribution of events from 160 a@xmath13gev @xmath14pb on emulsion , as shown in fig .
1(a ) of @xcite . in both of these studies , events
very near the observed reaction were rejected ; @xmath27 m for the 200 a@xmath13gev @xmath17s data @xcite and @xmath28 m for the 160 a@xmath13gev @xmath14pb data @xcite .
both @xmath26 distributions attenuate by roughly a factor of two per 1,000 @xmath24 m over the entire range scanned , consistent with the 120 @xmath24 m @xcite and 100 @xmath24 m @xcite fields of view of the scanning microscopes used and a lorenz boost of roughly 1,000 @xmath18 for forward - cone bremsstrahlung photons .
these observed distributions support the interpretation of all events at @xmath29 m from the reaction center in both data sets , most of which are below the borsellino line , as due to bremsstrahlung by coulomb - excited projectiles .
photons produced by bremsstrahlung from ag and br nuclei in the emulsion target will not in general be lorenz boosted , and hence will contribute negligibly to the number of epc events observed in the vicinity of the beam track .
reaction products including breakup fragments of the projectile ( mainly @xmath30 particles @xcite ) can be expected to be coulomb - excited as they exit the emulsion ; such fragments will produce bremsstrahlung photons by the mechanism discussed above in proportion to their @xmath31 .
approximately one photon per 2,500 @xmath24 m of beam - fragment track from an identified projectile - target reaction is expected from this secondary mechanism in the case of 200 a@xmath13gev @xmath17s on emulsion .
however , heavy fragments typically exit the reaction zone at relatively high angles from the beam ( cf . figs 3 - 5 to 3 - 14 of @xcite and fig
. 3 of @xcite ) , so bremsstrahlung photons from beam - fragment excitation are not expected to contribute significantly to epc events observed in the immediate vicinity the primary beam track .
photons from @xmath15 may contribute to epc near the observed reaction center ; however , the number of events observed at small @xmath26 compared to the large-@xmath26 tail in the 200 a@xmath13gev @xmath17s and 160 a@xmath13gev @xmath14pb reactions does not appear to scale with the 2 @xmath18 greater @xmath16 production expected @xcite in the latter reaction .
hence bremsstrahlung from coulomb - excited projectiles is expected to be the primary generator of epc events in the data sets considered here .
the cross section for epc by photons is strongly peaked at the production threshold value of @xmath32 , and decreases above threshold as @xmath33 , @xmath34 , which for @xmath35 can be approximated as @xmath36 @xcite . for the 200 a@xmath13gev @xmath17s data , @xmath37 gev ( fig.5 of @xcite ) ,
while for the 160 a@xmath13gev @xmath14pb data , @xmath38 gev ( fig.1f of @xcite ) ; hence the above - threshold epc background spectrum for these data sets can be expected to be identical up to normalization .
the 3.5 a@xmath13gev @xmath39he , @xmath40c and @xmath41ne reactions generate photons with @xmath42 mev @xcite ; the @xmath36 approximation is expected to somewhat over - predict the epc background in this case , but is employed for consistency .
fig.3 shows epc background predictions for @xmath43 mev/@xmath1 for these three reactions , together with invariant mass projections for @xmath43 mev/@xmath1 for the 200 a@xmath13gev
@xmath17s data from table 1 of @xcite , for the 160 a@xmath13gev @xmath14pb data from fig.1f of @xcite computed as described @xcite , and for the 3.5 a@xmath13gev @xmath39he , @xmath40c and @xmath41ne data from fig .
3 - 4 and table 3 - 1 of @xcite .
the lower cutoff of @xmath43 mev/@xmath1 is that chosen by el - nadi _ et al .
_ @xcite to reject threshold @xmath0 production events , and is used to avoid under - prediction of epc by the effective - photon approximation . a bin size of 1 mev/@xmath44 was employed consistently for all data ; this bin size corresponds to the experimental mass resolution for pair masses between 3 and 7 mev/@xmath44 reported by the studies for which tabulated data are available @xcite .
all of the data sets analyzed here represent selections from larger collections of observed by not analyzed events , hence the normalization of the epc predictions , and in particular the ratio of threshold to above - threshold @xmath0 production , can not be calculated reliably from the available data .
normalizations for the epc predictions have therefore been chosen for each reaction to fit the data between 1.3 and 3.3 mev/@xmath1 , i.e. the first two bins of the histograms shown in figs 3 and 4 .
this choice of normalization is consistent with the well - established accuracy of the effective photon approximation above the pair - production threshold .
the alternative of normalizing all data sets at higher mass values would produce order - of - magnitude overpredictions of the low - mass data for which no plausible explanation could be offered .
the normalization of the epc prediction for the 200 a@xmath13gev @xmath17s data with @xmath45 m shown in fig .
3 is consistent with that used by kamel @xcite in analysing direct pair production ( @xmath46 m ) from 200 a@xmath13gev @xmath17s on emulsion .
the rejection of non - epc events as background using this stringent normalization can not be ruled out , especially for the lower - energy data sets . as can be seen in fig .
3 , all of the events with invariant mass below 8 mev/@xmath1 reported in @xcite can be interpreted as due to epc by bremsstrahlung photons , as can events below 6 mev/@xmath1 reported by @xcite and events below 4 mev/@xmath1 reported by @xcite . the statistical significance @xmath47 , where @xmath48 represents counts in the `` signal '' and @xmath49 counts in the `` background '' @xcite varies from @xmath50 for the 9.8 mev/@xmath1 event reported in @xcite to @xmath51 for the 8.8 mev/@xmath1 event reported in @xcite .
figure 4 shows background due to epc for the cosmic - ray data sets of hintermann @xcite and anand @xcite and the 60 a@xmath13gev @xmath52o data set of badawy @xcite , calculated and normalized using the same procedures as used in fig .
3 . similar to the data shown in fg . 3 and those reported in @xcite
, these data sets reveal events above expected epc background for @xmath53 mev/@xmath1 .
pair production by the dalitz channel @xmath54 @xcite has previously been deemed negligible in these data sets based on analyses of the energy partition asymmetries @xmath55 , where @xmath56 is the greater of the two electron energies , across whole data sets @xcite .
the calculated epc backgrounds shown in figs 3 and 4 , and in fig .
2f of @xcite , are therefore taken to represent the backgrounds due to known pair - production processes in these data sets .
background subtraction was carried out for all spectra with a criterion of @xmath57 ; the results of this subtraction are shown in fig .
5 . these background - subtracted spectra are taken to be free of background events from characterized sources .
the events shown in the background - subtracted spectra of fig . 5 , together with one event at 9.8 mev/@xmath1 well - separated from background from @xcite , events at 5.5 , 7.0 and 16.5 mev/@xmath1 well - separated from background from @xcite , and an event at 9.5 mev/@xmath1 well - separated from background from @xcite were added within 1-mev/@xmath1 bins ; the sum spectrum is shown in fig .
6 . while none of the studies from which these data are taken claimed to employ blind analysis methods @xcite and unknown effects due to measurement conditions or to inconsistent assumptions or analysis methods can not be ruled out , the addition of these data sets to generate a single sum spectrum is justified if the published data are taken at face value .
the most straightforward interpretation of this sum spectrum is that it contains mass peaks with full width at half maximum ( fwhm ) of approximately 2 mev/@xmath1 overlaid on a uniform background of approximately one event per bin .
the source of such a uniform background is unknown ; however , there is no experimental motivation for assuming a nonuniform background , and in particular none for assuming a nonmonotonic background . assuming a fwhm of 2 mev/@xmath1 and imposing a criterion of @xmath57 against a uniform 1-event background , peaks can be identified at 3.8 mev/@xmath1 ( 3@xmath3 ) , 5.8 mev/@xmath1 ( 2.5@xmath3 ) , 7.8 mev/@xmath1 ( 4.2@xmath3 ) , 9.8 mev/@xmath1 ( 8.2@xmath3 ) , 15.8 mev/@xmath1 ( 3@xmath3 ) and 18.8 mev/@xmath1 ( 2@xmath3 ) .
the 9.8 @xmath58 1 mev/@xmath1 peak is by far the most significant , and is supported by 8 of the 10 cited studies ( all but @xcite and @xcite ) .
the measurement conditions restrict the neutral particles producing these @xmath0 events to the @xmath5 and @xmath6 s lifetime window .
specific lifetime measurements are reported by @xcite ; reported lifetimes for the 3.8 @xmath58 1 mev/@xmath1 events are @xmath59 to @xmath60 s , for the 5.8 @xmath58 1 mev/@xmath1 events @xmath61 s , for the 7.8 @xmath58 1 mev/@xmath1 events @xmath62 to @xmath63 s , for the 9.8 @xmath58 1 mev/@xmath1 events @xmath64 to @xmath65 s , for the 15.8 @xmath58 1 mev/@xmath1 events @xmath66 to @xmath67 s , and for the 18.8 @xmath58 1 mev/@xmath1 events @xmath68 to @xmath69 s. these measurements are consistent with decay halflives on the order of @xmath70 s for the neutral particles producing these events .
the peaks identified in fig .
6 can be interpreted as indicating the existence of a field @xmath2 outside the standard model , with excitations @xmath71 that behave as neutral bosons with mass @xmath72 .
the @xmath71 have femtosecond lifetimes and decay to @xmath0 pairs .
this interpretation of these events was first proposed in 1988 @xcite ; events consistent with x(9.8 ) , in particular , were identified at that time by de boer and van dantzig @xcite .
emission of a neutral boson consistent with x(9.8 ) has also been advanced as an explanation of anomalous ipc angular distributions in the decays of excited states of @xmath39he , @xmath73be , and @xmath40c @xcite , and well as the forbidden 10.96 mev @xmath74 _ m0 _ decay in @xmath52o @xcite .
if the neutral bosons suggested by the emulsion studies reviewed here may be identified with those suggested by ipc decay studies in light nuclei , @xmath75 assignments of @xmath76 and @xmath77 are indicated @xcite .
the observation of putative @xmath2-bosons in nuclear decays @xcite suggests that at least some of the @xmath2-bosons observed following cosmic ray or heavy - ion beam bombardment of emulsion targets may be produced by the decays of excited states of projectile fragmentation products .
indeed , el - nagdy et al .
@xcite describe 8 events from 200 a@xmath13gev @xmath17s on emulsion that are attributable on the basis of kinematic analysis to projectile - fragment decays .
if all secondary 200 a@xmath13gev @xmath17s events indicative of @xmath2 bosons are interpreted as projectile fragment decays , then assuming a maximum @xmath78 fragment multiplicity of 8 , one obtains a minimal branching ratio @xmath79 for the 1351 identified reactions observed @xcite .
if all secondary 3.6 a@xmath13 gev @xmath40c and @xmath80ne events indicative of @xmath2 bosons are interpreted as projectile fragment decays , one obtains a minimal @xmath81 for the 2600 identified reactions observed @xcite .
these values for @xmath82 are consistent with the range of @xmath82 values observed for individual states of light @xmath83-nuclei , which assume a dominant @xmath84 decay branch for @xmath2 bosons @xcite .
secondary production of @xmath2 bosons by @xmath16 decay can not be ruled out , but can be expected to be significantly less than the production of dalitz channel @xmath0 pairs , which are not observed in the present data sets @xcite . as @xmath16 s with kinetic energies greater than about 15 gev would be expected to escape the detection area unless emitted at very forward angles , the contribution of a possible @xmath85 channel is taken to be negligible .
the cross - section for direct production of @xmath2 bosons in nuclear interactions can be estimated by assuming that all background - subtracted @xmath2-boson events in which a particle decaying to an @xmath0 pair appears to originate directly from the identified projectile interaction represent direct production .
total charged pion multiplicities @xmath86 @xcite and @xmath2-boson multiplicities @xmath87 calculated under this assumption are shown in fig .
the production of charged pions is well represented by a power - law fit , @xmath88 where @xmath89 is the kinetic energy of the projectile ( cf .
fig.4 of @xcite ) .
the multiplicity of direct @xmath2 bosons @xcite is similarly well reproduced by a power law , @xmath90 .
direct production of @xmath2 bosons thus appears to increase very slowly with beam energy , and is increasingly swamped by pion production as the available energy increases .
any new particle decaying to @xmath0 pairs can be expected to affect the anomalous magnetic moment @xmath91 of the electron ; assuming electron - muon universality , the anomalous magnetic moment @xmath92 would be affected as well @xcite .
values for the expected anomaly @xmath93 , where @xmath94 ranges over scalar ( @xmath48 ) , pseudoscalar ( @xmath95 ) , vector ( @xmath96 ) and axial - vector ( @xmath97 ) couplings , calculated using the methods and form factors of schfer @xcite are shown in table 1 , together with widths @xmath98 in ev . as can be seen
, any of the observed @xmath2-bosons considered individually would violate the existing constraints on the values of both @xmath91 , i.e. @xmath99 , and @xmath92 , i.e. @xmath100 @xcite . however , it is also clear that , within the relevant uncertainties on @xmath2-boson mass and width , anomalies contributed by bosons with different couplings could cancel out . hence the current data imply that @xmath2-bosons must carry both parities for any given spin , i.e. that the pseudoscalar and axial - vector @xmath2-bosons implied by decay studies @xcite must be accompanied by scalar or vector @xmath2-bosons of appropriate widths to cancel their contributions to the @xmath101 .
the potential existence of @xmath2-bosons with the decay signature of `` heavy photons '' enables explanations of a number of anomalous results , in addition to those considered above , that have been observed in experiments measuring @xmath0 pair production .
the jacee collaboration has already noted that three presumptive bottom decays can be interpreted as producing @xmath2 bosons in the mass and lifetime range considered here @xcite . in the early 1960 s , tsai - ch _ et al .
_ reported a presumptive n annihilation event , now interpretable as an @xmath102 decay , producing three @xmath16 subsequently decaying to @xmath103 @xcite .
the energies and divergence angles of the @xmath0 pairs produced were sufficiently unusual that these events have been interpreted as evidence for a second neutral pion @xcite . as shown in fig .
8 , this event may instead be evidence for a @xmath104 decay channel , producing one particle below the mass range observable above background here , @xmath105 , @xmath106 , and higher masses of 47 , 66 and 102 mev/@xmath1 .
if this is the case , it would suggest the existence of @xmath2 bosons with @xmath75 of either @xmath107 or @xmath108 to conserve angular momentum and parity , consistent with the conclusion of the @xmath109 considerations outlined above .
while theoretical limits have been placed on the decay @xmath110 @xcite , a potential @xmath111 channel was not considered , and the fraction of apparent @xmath112 decays that may proceed through @xmath2-boson intermediaries is unknown .
it is interesting in this regard that a significant peak at about 10 mev/@xmath1 in the @xmath0 spectrum from @xmath102 decay has been reported by the cleo collaboration ( fig . 2 of @xcite ) .
this peak was dismissed on the basis of geant simulations that indicate a significant background of single-@xmath19 conversions following @xmath113 decay , some of which escape the kinematic constraints used to identify true @xmath102 events @xcite .
however , the approximately 2 @xmath3 difference at 10 mev/@xmath1 between the data and geant simulation shown may indicate evidence for @xmath106 in @xmath102 decay .
anomalous events near this invariant mass have also been observed in 20 gev photoproduction , but were dismissed as insufficiently statistically significant to warrant further study @xcite .
two potential interpretations of the @xmath2-boson spectrum reported here can be suggested .
one is motivated by fayet s approach to @xmath114-symmetry breaking through supersymmetry ( susy ) @xcite .
fayet has proposed a spin-1 @xmath115-boson with double parity and a mass and lifetime consistent with the @xmath2 bosons observed here , but with a dominant decay @xmath116 and a small coupling to leptons .
the observed @xmath2 bosons do not satisfy this latter requirement .
however , a condensate of @xmath115-bosons could , in principle , exhibit both a spectrum of masses with both natural and unnatural @xmath75 values , and a larger coupling to the @xmath0 channel .
preliminary analyses of anomalies in the angular distributions of @xmath0 decays of excited states in the @xmath83-nuclei @xmath73be , @xmath40c and @xmath52o that assume such a condensate of very light ( @xmath117 kev ) bosons suggest that such a mechanism could produce the observed @xmath118 and transition energies . a second line of interpretation is suggested by the relation between @xmath2-boson and pion production discussed above ( fig.7 ) .
assuming bare @xmath119 and @xmath120 quark masses in the mev range @xcite , one can envision the production of weakly - interacting @xmath121 or @xmath122 pairs , in the absence of gluons , with masses and @xmath75 in the range observed here .
such excitations would exist in the low - energy limit of qcd , for which the shapes and strengths of quark potentials are unknown . as indicated by fig.7 , production of such excitations
would be strongly suppressed , relative to pions , above the threshold for gluon production , so all quarks would appear bound by gluons , i.e. as standard model mesons or hadrons , in high - energy experiments .
we have reviewed evidence from heavy - ion bombardment and cosmic - ray studies using emulsion detectors that suggest the existence of a spectrum of neutral @xmath2-bosons with masses between 3 and 20 mev/@xmath1 and lifetimes on the order of femtoseconds . with a conservative and consistent model of the photo - production background imposed on all studies reviewed , 5 putative @xmath2-bosons , with masses of 2.3 , 5.8 , 9.8 , 15.8 and 18.8 mev/@xmath1
are suggested by the combined data .
the statistical significance of individual @xmath2-boson peaks from the combined studies , relative to an assumed uniform one event per mev/@xmath44 background of unknown source , varies between 2 and 8 @xmath3 .
the derived branching ratios for secondary production of such @xmath2 bosons by projectile fragments is consistent with branching ratios for putative @xmath2-bosons observed in decay studies of light nuclei @xcite . under our assumptions , the derived direct production multiplicity for such @xmath2-bosons increases very slowly with projectile kinetic energy as @xmath90 across all projectile energies investigated , indicating that @xmath2-boson production is much smaller than pion production at all energies above the pion production threshold .
consideration of the predicted effects of such @xmath2-bosons on the anomalous magnetic moments of both the electron and muon rules out the possibility of a single massive @xmath2-boson , but leaves open the possibility that @xmath2-boson excitations include both natural and unnatural parity states for each spin . we conclude that the experimental observations reviewed here offer evidence for the existence of new neutral bosons with masses between 3 and 20 mev/@xmath1 and femtosecond lifetimes that encourages dedicated further searches for such particles .
the inferred behaviour of the direct production cross - section indicates that such searches are best carried out below or in the vicinity of the pion production threshold .
fokke w. n. de boer ( 1942 - 2010 ) passed away as this work was nearing completion .
this paper is dedicated to his memory .
the comments of an anonymous referee on a previous version of this paper significantly contributed to its clarity .
borsellino a 1953 phys rev 89 1023 anand b m 1953 proc royal soc ( london ) a220 18 hintermann k 1954 phys rev 93 898 jain p l kazuno m girard b ahmad z 1994 phys rev lett 32 797 jain p l kazuno m girard b 1994 phys rev lett 32 1460 el - nadi m badawy o e 1988 phys rev lett 61 1271 badawy e o 1989 pers comm ( unpublished data ) kamel s 1989 phd dissertation , university of cairo el - nadi m kamel s hussein a abou - moussa z shaat e a salama h osman w 1996 il nuovo cimento 109 1517 kamel s 1996 phys lett
b368 291 jain p l singh g 2007 j phys g : nucl part phys 34 129 ; jain p l 2008 int j mod phys e 17 759 de boer , f w n van dantzig r 1988 phys rev lett 61 1274 el - nagdy m s abdelsalam a badawy b m 2007 second int conf mod trends phys res ( mtpr-06 ) aip 249 maddox j 1988 nature 335 293 perkins d h 1989 phys rev lett 62 2338 de boer , f w n van dantzig r 1989 phys rev lett 62 2639 hagmann c et al .
2008 phys lett 667 459 de boer f w n frhlich o bokemeyer h mller k a _ et al .
_ 1996 phys lett b388 235 ; de boer f w n van dantzig r van klinken j bethge k _ et al .
_ 1997 j phys g : nucl part phys b23 l85 ; de boer f w n bethge k bokemeyer h m buda a _ et al .
_ 1999 nucl phys b72 189 de boer f w n bethge k bokemeyer h m van dantzig r
_ et al . _ 2001 j phys g : nucl part phys 27 l29 stiebing k e de boer f w n frhlich o bokemeyer h m _ et al . _
2004 j phys g : nucl part phys 30 165 krasznahorkay a de boer f w n
csatls m csige l _
_ 2005 int symp on exotic nucl sys ( ens05 ) aip 802 , 135 krasznahorkay a de boer f w n
csatls m csige l _
_ 2006 acta phys polonica b37 239 vitz a krasznahorkay a guyls j csatls m _ et al .
_ 2008 acta phys polonica b39 483 davier m jeanjean j nguyen ngoc h 1986 phys lett b180 295 brown c n et al .
1986 phys rev lett 57 2101 riordan m et al .
1987 phys rev lett 59 755 bjorken j d et al .
1988 phys rev d38 3375 bross a , crisler m , pordes s , volk j , errede s and wrbanek j , 1991 phys rev lett 67 2942 fayet p 1981 nucl phys b187 184 fayet p 2007 phys rev d75 115017 boehm c hooper d silk j casse m paul j 2004 phys rev lett 92 101301 ; boehm c fayet p 2004 b683 219 chen f cline j m fradette a frey a r rabideau c 2009 arxiv:0911.2222 hallman t carroll j dejarnette w mcintyre e k _ et al .
_ 1985 nucl phys a440 697 dabrowska
a holynsky r jurak a _ et al . _
( klm collaboration ) 1993 phys rev d47 1751 adamovich m i aggarwal
m m alexandrov y a _
( emu01 collaboration ) 1991 phys lett b262 369 asakimori k burnett t h cherry m l chevli k _
( jacee collaboration ) 1994 j phys g : nucl part phys 20 1527 ; asakimori k burnett
t h cherry m l chevli k _
( jacee collaboration ) 1999 j phys g : nucl part phys 25 l133 dalitz r h 1951 proc phys soc a64 667 tsai - ch simonin - haudecoeur c 1961 il nuovo cimento 20 1102 ; tsai - ch simonin - haudecoeur c schune - boisbl d brami - dpaux b 1964 il nuovo cimento 31 1376 perkins w 2001 arxiv:0110053 ; perkins w a 2001 int j mod phys a 16 919 | electron - positron pair - production data obtained by bombardment of emulsion detectors with either cosmic rays or projectiles with masses between one and 207 amu and kinetic energies between 18 gev and 32 tev have been re - analysed using a consistent and conservative model of the background from electromagnetic pair conversion .
the combined data yield a spectrum of putative neutral bosons decaying to @xmath0 pairs , with masses between 3 and 20 mev/@xmath1 and femtosecond lifetimes .
the statistical significance against background for these `` @xmath2-bosons '' varies between 2 and 8 @xmath3 .
the cross - section for direct production of @xmath2-bosons increases slowly with projectile energy , remaining over 1,000 times smaller than the pion production cross - section . |
in the past two decades , one of the most important discoveries in modern theoretical physics is the ads / cft correspondence @xcite , relating strongly coupled field theories to weakly coupled gravity .
it provides a powerful tool to study strongly coupled field systems where traditional methods encounter great challenges or even break down .
fruitful results have been obtained by applying this duality to various areas of modern physics , such as condensed matter physics @xcite and qcd @xcite .
one noticeable application of this duality is to study non - equilibrium physics , in particular in the so - called `` holographic thermalization '' topic @xcite which is an atempt to describe the thermalization process of the quark - gluon plasma ( qgp ) produced in heavy ion collisions at the relativistic heavy ion collider ( rhic ) @xcite .
although the near - equilibrium physics of the qgp can be understood using a hydrodynamic description @xcite , the whole process to reach thermal equilibrium is yet not known .
observed data from rhic show that the time scale needed for qgp to reach thermal equilibrium is considerably shorter than expected based on perturbative methods @xcite .
this indicates that the thermalization process of qgp is strongly coupled , and thus motivates us to use ads / cft relation .
according to ads / cft , it is proposed that this thermalization process is dual to the process of a black hole formation in the bulk via gravitational collapse of a scalar field in ads space @xcite , or to a collapse of a thin - shell matter described by an ads - vaidya metric @xcite . to probe the detailed process of the thermalization , local - observables in the boundary field theory , such as energy - momentum tensor and its derivatives , are not sufficient . in refs .
@xcite , in the model of the collapsing of a thin - shell , three non - local observables have been proposed to probe the process , the two - point function , the wilson loop and the entanglement entropy .
according to the ads / cft relation , in the semiclassical limit these three non - local observables in the boundary field theory can be evaluated by calculating certain geometric quantities in the bulk geometry . by studying the time evolution of these three non - local observables , and under einstein gravity frame ,
the thermalization process has thus been studied in detail and some interesting results are found .
first , the saturation time which is needed to reach thermal equilibrium depends on the geometric size of the probe in the boundary field theory . probes with smaller size thermalize faster indicating a `` top - down '' thermalization mechanism , which is very much in contrast with the standard `` bottom - up '' paradigm @xcite from the use of perturbative methods . moreover ,
all probes used show a delay in the onset of the thermalization , an apparent non - analyticity in the end of thermalization , the transition to full thermal equilibrium is instantaneous and these features are independent of dimensionality . finally , for homogeneous initial conditions the entanglement entropy thermalizes slowest , and thus sets a timescale for equilibration that saturates a causality bound .
later , this study is extended to include the effect of chemical potential in the boundary field theory @xcite ( see also @xcite ) , which is usually the case in real heavy ion collision processes . in the gravity side , this is dual to introducing a gauge field and considering charged background geometries .
it is found that the time evolution of the three non - local observables exhibit similar behavior as in chargeless case .
the arise of the chemical potential affects the saturation time : the larger the chemical potential , the longer the saturation time .
also the study has been extended to other situations by considering higher curvature corrections @xcite , non - linear electromagnetic effect @xcite , angular momentum @xcite , noncommutative @xcite and hyperscaling violating geometries @xcite .
further , it is extended to study holographic thermalization of field theories living in curved spacetimes @xcite .
on the other hand , it is believed that einstein gravity should be considered as an effective description of the underlying quantum gravity theory ( string theory for example ) at low energy .
thus , it is interesting to study the effect of stringy corrections on the thermalization process . as we have mentioned above
, some work has been done in this directions by considering higher curvature corrections @xcite .
it is found that the existence of higher curvature corrections considerably affects the thermalization process .
for example , by including the effect of gauss - bonnet terms , the authors in refs .
@xcite found that a larger gauss - bonnet coupling constant will make the saturation shorter , thus leaving the imprint of this stringy effect . in this paper , we would like to consider another kind of stringy effect , the emergence of a dilaton field .
the action we consider is the well - known einstein - maxwell - dilaton action which is a low - energy effective action of string theory @xcite , where the dilaton couples with the maxwell field in a non - trivial way . to use the ads / cft relation and study the effect of the dilaton field
, we should first find asymptotically ads solutions with non - trivial dilaton configuration under this frame .
it is a non - trivial task .
amounts of work have been done to explore various solutions of this gravity theory .
non - trivial coupling between the dilaton and the gauge field yields interesting effects on the available spacetimes @xcite .
however , all of these black hole solutions with non - trivial scalar configuration are not aysmptotically ads . in @xcite , the authors even argued that , with only one liouville - type dilaton potential besides the cosmological constant , there are no dilaton - ds or ads black hole solutions . after lots of efforts ,
an aysmptotically ads black hole solution was found in ref .
@xcite . by choosing an appropriate combination of three liouville - type dilaton potentials
, the authors successfully obtained the static dilaton black hole solutions which are asymptotically ( a)ds in four and higher dimensions .
this solution has some interesting properties @xcite and has shown potential applications in cosmology @xcite .
we will work with this charged dilaton ads black hole , and see if and how the presence of the non - trivial dilaton field affects the thermalization process .
the paper is organized as follows . in the next section ,
we briefly introduce the charged dilaton ads black hole and then generalize it to obtain its vaidya - like version . in sec .
iii , by applying the three non - local observables , we study the detailed process of thermalization to see the influence of the chemical potential and the dilaton field .
the last section is devoted to summary and discussions .
in this section , we briefly introduce the charged dilation black hole in ads space , and obtain its vaidya - like version by introducing external matter source .
we consider the action of @xmath4-dimensional einstein - maxwell - dilaton gravity @xcite @xmath5 where @xmath1 is the coupling constant between the dilaton and the maxwell field . when @xmath6 this action reduces to the usual einstein - maxwell - scalar theory , while for @xmath7 the last term in the bracket represents the well - known dilaton - maxwell coupling that appears in the low energy string action in einstein s frame @xcite .
the dilaton potential is chosen to take the form @xcite @xmath8 e^{-4(n-2)\phi/[(n-1)\alpha]}\nonumber\\ & & + ( n-2)^2 ( n-\alpha^2 ) e^{4\alpha \phi/(n-1 ) } + 4 \alpha^2 ( n-1 ) ( n-2 ) e^{-2(n-2-\alpha^2)\phi/[(n-1)\alpha]}\bigg\},\end{aligned}\ ] ] where @xmath9 is the cosmological constant and @xmath10 the ads radius .
this type of dilaton potential can appear in the compactification of a higher - dimensional theory , including various supergravity models in four dimensions @xcite . from the above action
, we can see that reversing the sign of @xmath1 is equivalent to reverse the sign of the dilaton .
thus , without loss of generality , we can restrict @xmath1 to be non - negative , @xmath11 .
the corresponding equations of motion are @xmath12 they admit a charged dilaton black brane solution @xcite @xmath13 where the constant @xmath14 .
the dilaton and the electromagnetic fields take the form @xmath15 the remaining components of @xmath16 vanish .
the parameters @xmath17 and @xmath18 are related to the physical mass @xmath19 and charge @xmath20 of the black brane , @xmath21 where @xmath22 is the volume of @xmath23 .
hereinafter , without risk of confusion , we simply call the parameters @xmath17 and @xmath18 the mass and charge parameter respectively .
the constant @xmath24 is related to the mass and the charge parameters through the relation @xmath25 we should emphasize here that the solution contains three free parameters : @xmath26 , or equivalently @xmath27 .
when @xmath6 , the solution reduces to the well - known reissner - nordstrm - ads black brane .
the holographic thermalization in this case has been thoroughly discussed in refs .
thus , in the following we will mainly focus on the case with @xmath28 . in this case
, we note that the solution is not real for @xmath29 .
we should exclude this region from spacetime , what can be achieved by defining a new radial coordinate @xmath30 as @xcite @xmath31 the metric becomes @xmath32 with the coordinate @xmath30 now valued in the range @xmath33 .
the dilaton and the electromagnetic field become @xmath34 it can be shown that @xmath35 is a curvature singularity .
the event horizon @xmath36 is the largest root of equation @xmath37 .
depending on the free parameters @xmath27 ( or equivalently @xmath26 ) , there can be two horizons ( inner and outer horizons ) , one degenerate horizon ( extreme case ) and naked singularity .
the hawking temperature on the event horizon is @xmath38 it is not possible to give an analytical expression of @xmath36 .
however , we can express the mass parameter @xmath17 in terms of @xmath36 by solving @xmath37 , which is @xmath39 with @xmath40 . substituting @xmath17 into eq .
( [ temperature ] ) , one obtains the temperature as @xmath41-(n-1)\big[(n-2)\gamma-2\big]\gamma_h\bigg\}.\end{aligned}\ ] ] from it , we can see that when @xmath42 , the temperature is always positive @xmath43 and no extremal limit exists . while @xmath44 , like a reissner - nordstrm black hole , the temperature can be positive ( non - extremal case ) or zero (
extremal case ) . in the latter case ,
if we think of all the parameters but @xmath24 fixed , then the maximum value of @xmath24 corresponding to the extremal case can be obtained by solving @xmath45 , which is @xmath46}{n}\right)^{2/(n-2)}-1\right\}^{-1/2}.\end{aligned}\ ] ] for more discussions on the thermodynamical properties of this black brane , see refs .
@xcite .
according to ads / cft , the above black brane solution is dual to a cft on the boundary with temperature given by eq .
( [ temperature-1 ] ) .
moreover , the chemical potential @xmath47 of the boundary cft is related to the asymptotical value of the temporal part of @xmath48 , i.e. , @xmath49 . by integrating eq . ( [ phi_f ] )
, we can obtain the expression of @xmath50 , @xmath51 where @xmath52 is a constant corresponding to the electrostatic potential at @xmath53 , which is defined such that the gauge field vanishes at the horizon , i.e. , @xmath54 the precise relation between the chemical potential and the gauge field is subtle because there is a dimension mismatch between the chemical potential in the dual field theory and the gauge field @xmath48 defined in the action eq .
( [ action ] ) .
thus , one has to rescale the gauge field as @xmath55 , where @xmath56 is a scale with length unit that depends on the particular compactification . at this time
, the chemical potential and the gauge field has the same unit and the duality relation is @xmath57 since the boundary field theory we considered is conformal , the only relevant parameter we can vary is a dimensionless ratio @xmath0 constructed from the chemical potential and the temperature , i.e. , @xmath58 , @xmath59 now we can see that if @xmath1 and @xmath17 are kept fixed , then by varying @xmath24 which is equivalent to varying the charge parameter @xmath18 , we can explore various values of the ratio @xmath0 in the dual field theory . for @xmath60 , when @xmath24 varies from @xmath61 ( vanishing @xmath52 ) to @xmath62 ( vanishing @xmath63 ) , @xmath0 spans the whole range of values , i.e. , from @xmath64 to @xmath65 .
however , for @xmath66 , the situation becomes complicated .
as we will see later , in this case , as @xmath24 runs over the whole range of values , @xmath0 is bounded in a finite range of values .
our goal is to investigate the thermalization process of the boundary field system under a quantum quench .
according to ads / cft , this process can be holographically modeled by the process of a black hole formation in the bulk which can be described by a vaidya - like metric .
thus , in this section , we would like to get the vaiya - like metric by generalizing the black brane solution to be time - dependent .
it is more convenient to work with a new radial coordinate @xmath67 such that the ads boundary lies at @xmath68 .
thus , the metric becomes @xmath69 introducing ingoing eddington - finkelstein coordinate @xmath70 the above metric becomes @xmath71 usually , the vaidya - like version can be obtained by using time dependent mass and charge parameters @xmath72 , that is , @xmath73 where @xmath17 and @xmath18 are the values of the corresponding parameters at late times .
we can still introduce the parameter @xmath24 , which satisfies the relation eq .
( [ chargeparameter ] ) , to make our expression explicitly . after taking such time - dependent parameters
, we obtain the vaiya - like metric which takes the same form as eq .
( [ metric-2 ] ) but with @xmath74 being now @xmath75 \left(\frac{z}{\sqrt{\l^4+b^2 z^2}}\right)^{n-2 } \gamma^{1-\gamma}.\end{aligned}\ ] ] now , the equations of motion eqs .
( [ eom ] ) do not hold any more and external matter sources need to be introduced , @xmath76 by checking the equation of motion of the dilaton field , we find that it always holds and no more additional external terms need to be introduced . in order to keep simplicity , in this paper
we only consider the case with @xmath77 , that is , the dual field theory we considered is a four - dimensional conformal field theory .
the above vaidya - like metric is a solution of the equations of motion eqs .
( [ eom ] ) provided the external matter source satisfies @xmath78}{2(2+\alpha^2 ) ( \l^4 + b^2 z^2)^{5/2 } } \left(\frac{\l^4}{\l^4 + b^2 z^2}\right)^{-\frac{3\alpha^2}{2(2+\alpha^2 ) } } m \dot{t } ( v)\right\ } \delta_\mu^v \delta_\nu^v,\nonumber\\ 8\pi j^\nu_{(ext ) } & = & \left\{\frac{\sqrt{3 } b z^5}{\l^8 \sqrt{2(2+\alpha^2 ) } } \left(\frac{\l^4}{\l^4+b^2 z^2}\right)^{\frac{2}{2+\alpha^2}}\sqrt{m } \frac{\dot{t}(v)}{\sqrt{t(v)}}\right\ } \delta_z^\nu,\end{aligned}\ ] ] where dot means @xmath79 .
we can see that the in - falling matter is charged dust . as in refs .
@xcite , we consider @xmath80 to take the form @xmath81 where @xmath82 is a small parameter , with typical order of @xmath83 .
that is , the ingoing falling matter is a thin shell .
thus , in the bulk gravity side , we consider the collapsing of an ingoing falling thin - shell to form a black hole . and
this picture is dual to a sudden energy injection into the field system evolving towards thermal equilibrium .
now , we will use the vaidya - like metric constructed in last section to discuss the thermalization process of the dual field system . to investigate the details of the thermalization
, one can use three non - local observables , the two - point coorelation function , the wilson loop and the entanglement entropy , to probe the process .
according to ads / cft , in the saddle approximation these three non - local observable can be evaluated by calculating the corresponding geometric quantities in the bulk geometry .
as we stated above , we mainly focus on the case with @xmath77 . in doing numerical calculations in the following
, we would like to set @xmath84 .
that is , we fix the mass of the black brane formed finally , which is equivalent to fix the amount of energy injected into the dual boundary field system .
and then we would like to see the effect of the chemical potential ( more precisely @xmath0 ) and the coupling constant @xmath1 on the thermalization process . from eq .
( [ massparameter ] ) , we can see that , after the mass parameter @xmath17 is fixed , the event horizon @xmath36 is determined by parameters @xmath1 and @xmath24 through the relation @xmath85 with @xmath86 . when @xmath87 , there is no bound of @xmath24 and it can take arbitrary values . when @xmath7 , the above relation reduces to @xmath88 ,
so @xmath24 should be less than @xmath89 to guarantee the existence of the horizon . while @xmath60 , from eq .
( [ maxb ] ) the maximum value of @xmath24 is @xmath90 the charge parameter @xmath91 then takes values in the range @xmath92 $ ] at this time , with @xmath93 . from eq .
( [ chi ] ) , the ratio of the chemical potential over temperature @xmath0 is @xmath94^{-1},\end{aligned}\ ] ] where we have taken the scale @xmath95 .
the relation between @xmath0 and @xmath96 is plotted in fig .
1 , from where we can see that fixing @xmath1 , @xmath97 and increasing @xmath18 corresponds to increasing @xmath0 .
when @xmath18 varies from @xmath98 to @xmath99 , the value of @xmath0 varies from @xmath100 to @xmath65 .
when @xmath101 , @xmath0 is no longer a monotonically increasing function of @xmath18 .
in fact , as @xmath18 increases from @xmath100 to @xmath65 , @xmath0 first increases and after reaching a maximum value at @xmath102 it begins to decrease . in this case , the value of @xmath0 is bounded in a finite range . for @xmath7 , the charge
@xmath18 should be less than @xmath103 to guarantee the existence of the horizon as we have stated above . from the figure
, we can see that in this case , @xmath0 is still a monotonically increasing function of @xmath18 and approaches a finite constant as @xmath18 approaches @xmath103 . in @xmath77 case for various @xmath1 .
for @xmath7 , @xmath18 should be less than @xmath103 to guarantee the existence of the horizon . when @xmath3 , there is a maximum point of @xmath104 at @xmath102 , which is @xmath105 for @xmath106 and @xmath107 for @xmath108.,scaledwidth=50.0% ] on the boundary at time @xmath109 , we choose two points with spatial separation @xmath110 and compute the correlation function of some operator with large conformal dimension located at the two points .
considering the symmetry of the boundary , we can always choose coordinates such that the two points lie in the @xmath111-axis with coordinates @xmath112 respectively .
other coordinates of the two points are identical . according to the ads / cft duality , in the saddle point approximation the two - point function
can be evaluated by calculating the length of the space - like geodesic connecting the two points .
the geodesic can be parameterized by only two functions : @xmath113 and @xmath114 , with other coordinates unchanged , where we simply denote @xmath115 .
thus , the induced metric on the geodesic is @xmath116 the length functional of the geodesic is @xmath117 to minimize the length of the geodesic , we need to solve the two equations of motion for @xmath113 and @xmath114 respectively .
the expressions are rather involved and we do not give them explicitly here . considering the symmetry
, we have in mind the picture of the geodesic : starting from one of the two points on the boundary @xmath68 , the geodesic extends into the bulk and after reaching the turning point @xmath118 it turns back and finally ends at another point on the boundary . at the turning point @xmath119 , the derivatives vanish , that is , @xmath120 .
it seems that one can impose the boundary conditions at @xmath119 to solve the two equations of motion .
however , the two equations of motion are singular at @xmath119 , and in practice , to make our numerical calculations available , we would rather impose the boundary conditions at the neighborhood of the turning point @xmath119 , @xmath121 where @xmath122 is a small quantity , with typical order of @xmath83 .
the @xmath123 terms are corrections terms and can be derived by solving the two equations of motion around @xmath119 order by order . in our calculations , we only calculate to order @xmath124 .
the two free parameters @xmath125 and @xmath126 characterizing the turning point of the geodesics are determined by the constraint equations @xmath127 where @xmath128 is a uv - cutoff to make the length of the geodesic finite and @xmath109 is the boundary time .
furthermore , note that the integrand in the length functional eq .
( [ lengthfunctional ] ) does not depend on @xmath129 explicitly .
so if we treat @xmath129 as a `` time '' variable then the corresponding hamiltonian is conserved , which yields the following equation @xmath130 by using it , the length functional can be rewritten as @xmath131 the length of the geodesic @xmath132 is a function of the boundary time @xmath109 . by calculating the length at different boundary times , we can get the time evolution of the two - point function . for convenience ,
we define a rescaled length @xmath133 and study its time evolution . to make the length independent of the cutoff @xmath128
, we have subtracted the length at late time in the definition .
@xmath134 depends on three parameters : the boundary time @xmath109 , the dilaton - maxwell coupling constant @xmath1 and the charge @xmath18 . by varying @xmath1 and @xmath18
, we can study the effect of the two parameters on the time evolution of the two - point function . in fig .
2 , we plot the time evolution of the two - point function for various values of @xmath1 @xmath135 and @xmath18 . for comparison
, we also give the result for the @xmath6 case .
the spatial separation of the two points on the boundary is chosen to be @xmath136 . from the figure
, we can see that that the whole process can be divided in the following five successive stages : delay at early times , pre - local - equilibrium with quadratic growth in time , post - local - equilibrium with linear growth , memory - loss - period , and equilibrium .
these features have been observed in various holographic models for thermalization and can be considered to be universal for strongly coupled field systems . as we increase the charge @xmath18 , equivalent to increasing the ratio of the chemical potential over temperature @xmath0 of the dual field system for @xmath137 , the initial absolute value of @xmath138 decreases , meaning that the dual field system is initially closer to thermal equilibrium .
however , larger @xmath18 makes the delay time longer and the growth of @xmath138 slower , thus the saturation time to reach thermal equilibrium increases .
this means , for @xmath137 , when we fix the energy injected , that the saturation time increases as @xmath0 increases .
also , we should emphasize that the dependence of the saturation time on @xmath0 is weak , as we can see from the figure .
this dependence of the saturation time on the ratio of the chemical potential over temperature @xmath0 has been observed in the case with standard maxwell term @xcite and also with non - linear born - infeld term @xcite .
however , when @xmath3 , the situation becomes complicated , as we show in fig .
3 with @xmath108 . in this case , the initial absolute value of @xmath138 is no longer a monotonically decreasing function of the charge @xmath18 .
in fact , as @xmath18 increases , the initial absolute value of @xmath138 first decreases and then after @xmath18 crosses some critical value , the initial absolute value of @xmath138 begins to increase .
more interesting , we can see from the figure that the saturation time is still a monotonically increasing function of @xmath18 . however , from fig . 1 , we can see that for @xmath3 , @xmath0 is no longer a monotonically increasing function of @xmath18 , thus the saturation time is no longer a monotonically increasing function of @xmath0 : when @xmath139 , larger @xmath0 yields longer @xmath140 ; however , when @xmath141 , we have the inverse behavior .
we can see it more clearly in the right panel of fig .
3 . case with @xmath136 .
there is no bound for @xmath18 in this case .
@xmath142 corresponds to the point when @xmath58 reaches its maximum value .
there is a large overlap between the curves with @xmath143 and the one with @xmath144 at early times . _
right : _ saturation time @xmath140 versus the ratio of chemical potential over temperature @xmath0 for @xmath108 case.,title="fig:",scaledwidth=40.0% ] case with @xmath136 .
there is no bound for @xmath18 in this case .
@xmath142 corresponds to the point when @xmath58 reaches its maximum value .
there is a large overlap between the curves with @xmath143 and the one with @xmath144 at early times . _
right : _ saturation time @xmath140 versus the ratio of chemical potential over temperature @xmath0 for @xmath108 case.,title="fig:",scaledwidth=40.0% ] by comparing the four graphs with different values of @xmath1 in figs . 2 and 3
, one can see that the effect of the coupling constant @xmath1 on the saturation time is nearly negligible .
we can see this point more clearly from the left panel of fig .
4 , where we plot the time evolution of the two - point function for various values of @xmath1 and fixed @xmath18 . from the figure , we can see that , for fixed @xmath18 , larger @xmath1 makes the initial state of the dual field system further away from thermal equlibrium , but will make the delay time shorter and the growth of @xmath138 faster thus yielding the saturation time nearly un - affected .
the saturation time is almost independent of the coupling constant @xmath1 . as the one - parameter gravity action eq .
( [ action ] ) corresponds to a one - parameter dual field theory , this phenomenon indicates the universality of the saturation time of this one - parameter strongly coupled field system to reach thermal equilibrium under quantum quench . with @xmath144
the spatial separation of the two points is @xmath136 .
_ right _ : saturation time versus the boundary scale .
the result is the same for various values of @xmath1.,title="fig:",scaledwidth=42.0% ] with @xmath144 .
the spatial separation of the two points is @xmath136 . _
right _ : saturation time versus the boundary scale .
the result is the same for various values of @xmath1.,title="fig:",scaledwidth=40.0% ] the saturation time depends on the boundary scale @xmath110 , as we can see from the right panel of fig .
4 , where we plot the saturation time versus @xmath110 . note that the effect of @xmath1 on the saturation time can be neglected as we stated above , thus this figure is correct for various @xmath1 . as we can see from the figure ,
the saturation time is linear in @xmath110 with slope less than unit when @xmath110 is small , and sub - linear when @xmath110 becomes large .
once again , we can see that the saturation time depends weakly on the charge : larger @xmath18 will have longer saturation time . according to the ads / cft duality , the expectation value of the wilson loop of the boundary field theory , in the saddle approximation , is dual to the area of a two - dimensional extremal surface in the bulk which anchored on the loop at the boundary .
we choose the wilson loop to be a circle and study its time evolution during thermalization .
it turns out that the time evolution of this observable shows a similar behavior as that of the two - point function , which supports our results above .
so to avoid redundancy , we will not show the results for the wilson loop here .
now we turn to use the entanglement entropy to probe the thermalization process . among the three non - local observables , it involves most degrees of freedom of the boundary field system .
so we expect that by studying its evoluton , we can gain more information of the process . on the boundary ,
we choose a spherical subregion of the spatial space constrained by the relation @xmath145 , and calculate its entanglement entropy . according to the conjecture proposed by refs .
@xcite , the entanglement entropy of a subregion of the boundary spatial space is dual to the area of a codimension - two extremal surface anchored on the boundary of the chosen subregion . using spherical coordinates @xmath146 and considering the spherical symmetry
, the codimension - two extremal surface can be parameterized by only two functions , @xmath147 and @xmath148 .
thus , the induced metric on the surface is @xmath149 and the area functional is @xmath150 to get the extremal value of the functional , as done in previous two subsections , we need to solve the two equations of motion , not shown here .
the same boundary conditions as eq .
( [ bnycondition ] ) are imposed .
we define a rescaled area @xmath151 , where @xmath152 is the volume of the chosen subregion , and study its time evolution during thermalization . in fig .
6 , we plot the time evolution of the entanglement entropy for various values of @xmath1 and @xmath18 .
the evolution shows a similar behavior as those of the other two non - local observables : for fixed @xmath1 , the saturation time increases as the charge @xmath18 increases . for @xmath137 ,
this means the ratio of the chemical potential over temperature @xmath0 enhances the saturation time . while for @xmath3
, the situation depends on the charge @xmath18 : when @xmath139 , the effect of @xmath0 on the saturation time is to enhance , while the effect is opposite when @xmath141 . moreover , we see once again that the dilaton - maxwell coupling constant @xmath1 has almost no influence on the saturation time , although it indeed affects the whole process .
these results provide further support for the ones obtained from the other two observables . however , there are some differences comparing to the results of other two non - local observables .
first , the delay time for the entanglement entropy is the shortest .
this is understandable , as entanglement entropy , among the three non - local observables involves most degrees of freedom of the field system and thus is the most sensitive to thermalization .
second , now for all cases with different @xmath1 we show , the behavior of the initial absolute value of @xmath153 as @xmath18 increasing is the same : it decreases as @xmath18 increases .
this is different from results of the other two observalbes , where the initial absolute value of @xmath153 is not a monotonically decreasing function of @xmath18 when @xmath154 .
this is understandable , as the three observables relate to different amounts of degrees of freedom of the boundary field system , and thus may reflect different aspects of the system . as we stated above
, the entanglement entropy involves more degrees of freedom and thus can can reflect more clearly the average state of the field system .
thus it is expected that if all the degrees of freedom of the field system are taken into account , the initial state of the field system gets closer to the thermal equilibrium as @xmath18 increases .
we studied the holographic thermalization process in the presence of a chemical potential and dilaton field .
three non - local observables are used to probe the process , and they together show a rich and consistent picture of the thermalization .
the effect of the chemical potential ( or the charge ) and the dilaton field on the thermalization is explored in detail .
the saturation time increases as the charge @xmath18 increases , which means that the ratio of the chemical potential over temperature @xmath0 enhances the saturation time when @xmath137 .
however , when @xmath1 is larger than @xmath89 , the situation becomes a little complicated : when @xmath18 is smaller than some critical value @xmath102 , the saturation time increases as the @xmath0 increases ; however , when @xmath141 , the behavior is inverse .
this is due to the fact that the @xmath0 is no longer a monotonically increasing function of @xmath18 for @xmath3 .
this is different from cases with standard maxwell term @xcite or non - linear born - infeld term @xcite , where @xmath0 is always a monotonically increasing function of @xmath18 .
we also investigate the effect of the dilaton - electromagnetic coupling constant @xmath1 on the thermalization process . this non - trival coupling yields great effects on the availabe background spactimes .
however , it is remarkable to see that it nearly has no influence on the saturation time .
as different @xmath1 in the action defines a different gravity theory and thus corresponds to a different dual field theory , it indicates the universality of the saturation time of this one - parameter field theories to reach thermal equilibrium under quantum quench .
although it has no effect on the saturation time , it indeed leave imprints on the whole thermalization process .
if we fixed the charge @xmath18 , increasing @xmath1 makes the initial state of the dual field system further away from thermal equlibrium , the delay time shorter and the growth of @xmath138 faster .
it is interesting to see if an physical explanation on this point can be given .
we have modelled the thermalization process via the naive gravitational collapse of a charged thin shell .
it can also be modelled by the collapsing of a scalar field as done in refs .
it is interesting to see if this more realistic model can give us more information about the thermalization process .
we leave it for further investigations .
this work has been supported by fapesp and cnpq ( brazil ) .
j. m. maldacena , _ the large n limit of superconformal field theories and supergravity _ , int .
j. theor .
phys . * 38 * , 1113 ( 1999 ) [ adv .
* 2 * , 231 ( 1998 ) ] [ hep - th/9711200 ] .
s. s. gubser , i. r. klebanov and a. m. polyakov , _ gauge theory correlators from noncritical string theory _ , phys .
b * 428 * , 105 ( 1998 ) [ hep - th/9802109 ] .
e. witten , _ anti - de sitter space and holography _ , adv .
* 2 * , 253 ( 1998 ) [ hep - th/9802150 ] .
s. a. hartnoll , _ lectures on holographic methods for condensed matter physics _ , class .
* 26 * , 224002 ( 2009 ) [ arxiv:0903.3246 [ hep - th ] ] .
c. p. herzog , _ lectures on holographic superfluidity and superconductivity _ , j. phys .
a * 42 * , 343001 ( 2009 ) [ arxiv:0904.1975 [ hep - th ] ] .
j. mcgreevy , _ holographic duality with a view toward many - body physics _ , adv .
high energy phys . * 2010 * , 723105 ( 2010 ) [ arxiv:0909.0518 [ hep - th ] ] .
d. mateos , _ string theory and quantum chromodynamics _ , class .
* 24 * , s713 ( 2007 ) [ arxiv:0709.1523 [ hep - th ] ] .
s. s. gubser and a. karch , _ from gauge - string duality to strong interactions : a pedestrian s guide _ , ann .
nucl . part .
sci . * 59 * , 145 ( 2009 ) [ arxiv:0901.0935 [ hep - th ] ] .
j. casalderrey - solana , h. liu , d. mateos , k. rajagopal and u. a. wiedemann , _ gauge / string duality , hot qcd and heavy ion collisions _ , arxiv:1101.0618 [ hep - th ] .
u. h. danielsson , e. keski - vakkuri and m. kruczenski , _ black hole formation in ads and thermalization on the boundary _ , jhep * 0002 * , 039 ( 2000 ) [ hep - th/9912209 ] .
r. a. janik and r. b. peschanski , _ gauge / gravity duality and thermalization of a boost - invariant perfect fluid _ , phys .
d * 74 * , 046007 ( 2006 ) [ hep - th/0606149 ] .
m. chesler and l. g. yaffe , _ boost invariant flow , black hole formation , and far - from - equilibrium dynamics in n = 4 supersymmetric yang - mills theory _ , phys .
d * 82 * , 026006 ( 2010 ) [ arxiv:0906.4426 [ hep - th ] ] .
d. garfinkle and l. a. pando zayas , _ rapid thermalization in field theory from gravitational collapse _ , phys .
d * 84 * , 066006 ( 2011 ) [ arxiv:1106.2339 [ hep - th ] ] .
d. garfinkle , l. a. pando zayas and d. reichmann , _ on field theory thermalization from gravitational collapse _
, jhep * 1202 * , 119 ( 2012 ) [ arxiv:1110.5823 [ hep - th ] ] . s. bhattacharyya and s. minwalla , _ weak field black hole formation in asymptotically ads spacetimes _ , jhep * 0909 * , 034 ( 2009 ) [ arxiv:0904.0464 [ hep - th ] ] .
s. lin and e. shuryak , _ toward the ads / cft gravity dual for high energy collisions .
gravitationally collapsing shell and quasiequilibrium _ , phys .
d * 78 * , 125018 ( 2008 ) [ arxiv:0808.0910 [ hep - th ] ] . v. balasubramanian , a. bernamonti , j. de boer , n. copland , b. craps , e. keski - vakkuri , b. muller and a. schafer _ et al .
_ , _ thermalization of strongly coupled field theories _ , phys . rev .
lett . *
106 * , 191601 ( 2011 ) [ arxiv:1012.4753 [ hep - th ] ] .
v. balasubramanian , a. bernamonti , j. de boer , n. copland , b. craps , e. keski - vakkuri , b. muller and a. schafer _ et al .
_ , _ holographic thermalization _ , phys .
d * 84 * , 026010 ( 2011 ) [ arxiv:1103.2683 [ hep - th ] ] . j. abajo - arrastia , j. aparicio and e. lopez , _ holographic evolution of entanglement entropy _
, jhep * 1011 * , 149 ( 2010 ) [ arxiv:1006.4090 [ hep - th ] ] . j. aparicio and e. lopez , _ evolution of two - point functions from holography _ , jhep * 1112 * , 082 ( 2011 ) [ arxiv:1109.3571 [ hep - th ] ]
. t. albash and c. v. johnson , _ evolution of holographic entanglement entropy after thermal and electromagnetic quenches _
, new j. phys . * 13 * , 045017 ( 2011 ) [ arxiv:1008.3027 [ hep - th ] ] . v. keranen , e. keski - vakkuri and l. thorlacius , _ thermalization and entanglement following a non - relativistic holographic quench _
d * 85 * , 026005 ( 2012 ) [ arxiv:1110.5035 [ hep - th ] ] .
d. galante and m. schvellinger , _ thermalization with a chemical potential from ads spaces _ , jhep * 1207 * , 096 ( 2012 ) [ arxiv:1205.1548 [ hep - th ] ] .
e. caceres and a. kundu , _ holographic thermalization with chemical potential _ , jhep * 1209 * , 055 ( 2012 ) [ arxiv:1205.2354 [ hep - th ] ] .
v. e. hubeny , m. rangamani and e. tonni , _ thermalization of causal holographic information _ ,
jhep * 1305 * , 136 ( 2013 ) [ arxiv:1302.0853 [ hep - th ] ] . y. z. li , s. f. wu , y. q. wang and g. h. yang , _ linear growth of entanglement entropy in holographic thermalization captured by horizon interiors and mutual information _
, jhep * 1309 * , 057 ( 2013 ) [ arxiv:1306.0210 [ hep - th ] ]
. y. z. li , s. f. wu and g. h. yang , _ gauss - bonnet correction to holographic thermalization : two - point functions , circular wilson loops and entanglement entropy _ , phys .
d * 88 * , 086006 ( 2013 ) [ arxiv:1309.3764 [ hep - th ] ] .
s. a. stricker , _ holographic thermalization in @xmath155 super yang - mills theory at finite coupling _ , eur .
j. c * 74 * , 2727 ( 2014 ) [ arxiv:1307.2736 ] .
w. h. baron and m. schvellinger , _ quantum corrections to dynamical holographic thermalization : entanglement entropy and other non - local observables _ , jhep * 1308 * , 035 ( 2013 ) [ arxiv:1305.2237 [ hep - th ] ]
. i. arefeva , a. bagrov and a. s. koshelev , _ holographic thermalization from kerr - ads _ , jhep * 1307 * , 170 ( 2013 ) [ arxiv:1305.3267 [ hep - th ] ] .
x. zeng and w. liu , _ holographic thermalization in gauss - bonnet gravity _ , physics letters b * 726 * , 481 ( 2013 ) [ arxiv:1305.4841 [ hep - th ] ] .
x. x. zeng , x. m. liu and w. b. liu , _ holographic thermalization with a chemical potential in gauss - bonnet gravity _ ,
jhep * 1403 * , 031 ( 2014 ) [ arxiv:1311.0718 [ hep - th ] ] .
a. giordano , n. e. grandi and g. a. silva , _ holographic thermalization of charged operators _ , arxiv:1412.7953 [ hep - th ] .
w. fischler , s. kundu and j. f. pedraza , _ entanglement and out - of - equilibrium dynamics in holographic models of de sitter qfts _ , jhep * 1407 * , 021 ( 2014 ) [ arxiv:1311.5519 [ hep - th ] ] .
m. alishahiha , a. f. astaneh and m. r. m. mozaffar , _ thermalization in backgrounds with hyperscaling violating factor _
d * 90 * , 046004 ( 2014 ) [ arxiv:1401.2807 [ hep - th ] ] .
p. fonda , l. franti , v. kernen , e. keski - vakkuri , l. thorlacius and e. tonni , _ holographic thermalization with lifshitz scaling and hyperscaling violation _ , jhep * 1408 * , 051 ( 2014 ) [ arxiv:1401.6088 [ hep - th ] ] .
w. fischler , p. h. nguyen , j. f. pedraza and w. tangarife , _ fluctuation and dissipation in de sitter space _ , jhep * 1408 * , 028 ( 2014 ) [ arxiv:1404.0347 [ hep - th ] ] .
j. f. pedraza , _ evolution of nonlocal observables in an expanding boost - invariant plasma _ ,
d * 90 * , 046010 ( 2014 ) [ arxiv:1405.1724 [ hep - th ] ] .
m. alishahiha , m. r. m. mozaffar and m. r. tanhayi , _ evolution of holographic n - partite information _ , arxiv:1406.7677 [ hep - th ] .
s. fischetti , d. kastor and j. traschen , _ non - vacuum ads cosmologies and the approach to equilibrium of entanglement entropy _ , arxiv:1407.4299 [ hep - th ] .
x. x. zeng , x. m. liu and w. b. liu , _ holographic thermalization in noncommutative geometry _ , arxiv:1407.5262 [ hep - th ] .
x. x. zeng , d. y. chen and l. f. li , _ holographic thermalization and gravitational collapse in the spacetime dominated by quintessence dark energy _ , arxiv:1408.6632 [ hep - th ] .
e. caceres , a. kundu , j. f. pedraza and d. l. yang , _ weak field collapse in ads : introducing a charge density _ , arxiv:1411.1744 [ hep - th ] .
g. camilo , b. cuadros - melgar and e. abdalla , _ holographic thermalization with a chemical potential from born - infeld electrodynamics _
, jhep * 1502 * , 103 ( 2015 ) [ arxiv:1412.3878 [ hep - th ] ]
. s. j. zhang , b. wang , e. abdalla and e. papantonopoulos , _ holographic thermalization in gauss - bonnet gravity with de sitter boundary _ , arxiv:1412.7073 [ hep - th ] .
f. gelis , _ the early stages of a high energy heavy ion collision _ , j. phys .
ser . * 381 * , 012021 ( 2012 ) [ arxiv:1110.1544 [ hep - ph ] ] .
e. iancu , _ qcd in heavy ion collisions _ , arxiv:1205.0579 [ hep - ph ] . b. muller and a. schafer , _ entropy creation in relativistic heavy ion collisions _ ,
int . j. mod .
e * 20 * , 2235 ( 2011 ) [ arxiv:1110.2378 [ hep - ph ] ] .
g. policastro , d. t. son and a. o. starinets , _ the shear viscosity of strongly coupled n=4 supersymmetric yang - mills plasma _ , phys .
lett . * 87 * , 081601 ( 2001 ) [ hep - th/0104066 ] .
g. policastro , d. t. son and a. o. starinets , _ from ads / cft correspondence to hydrodynamics _ , jhep * 0209 * , 043 ( 2002 ) [ hep - th/0205052 ] . d. t. son and a. o. starinets , _ viscosity , black holes , and quantum field theory _
sci . * 57 * , 95 ( 2007 ) [ arxiv:0704.0240 [ hep - th ] ] .
s. bhattacharyya , v. e. hubeny , s. minwalla and m. rangamani , _ nonlinear fluid dynamics from gravity _ , jhep * 0802 * , 045 ( 2008 ) [ arxiv:0712.2456 [ hep - th ] ] . r. baier , a. h. mueller , d. schiff and d. t. son , _ bottom up thermalization in heavy ion collisions _ , phys .
b * 502 * , 51 ( 2001 ) [ hep - ph/0009237 ] .
a. h. mueller , a. i. shoshi and s. m. h. wong , _ a possible modified bottom - up thermalization in heavy ion collisions _ , phys .
b * 632 * , 257 ( 2006 ) [ hep - ph/0505164 ] .
j. h. horne and g. t. horowitz , _ rotating dilaton black holes _ , phys .
d * 46 * , 1340 ( 1992 ) [ hep - th/9203083 ] .
g. w. gibbons and k. i. maeda , _ black holes and membranes in higher dimensional theories with dilaton fields _ , nucl .
b * 298 * , 741 ( 1988 ) .
t. koikawa and m. yoshimura , _ dilaton fields and event horizon _ , phys .
b * 189 * , 29 ( 1987 ) .
d. brill and g. t. horowitz , _ negative energy in string theory _
b * 262 * , 437 ( 1991 ) .
d. garfinkle , g. t. horowitz and a. strominger , _ charged black holes in string theory _ , phys .
d * 43 * , 3140 ( 1991 ) [ erratum - ibid .
d * 45 * , 3888 ( 1992 ) ] .
r. gregory and j. a. harvey , _ black holes with a massive dilaton _ , phys .
d * 47 * , 2411 ( 1993 ) [ hep - th/9209070 ] .
g. t. horowitz and a. strominger , _ black strings and p - branes _
b * 360 * , 197 ( 1991 ) .
k. c. k. chan , j. h. horne and r. b. mann , _ charged dilaton black holes with unusual asymptotics _ , nucl .
b * 447 * , 441 ( 1995 ) [ gr - qc/9502042 ] .
g. clement , d. galtsov and c. leygnac , _ linear dilaton black holes _ , phys .
d * 67 * , 024012 ( 2003 ) [ hep - th/0208225 ] .
a. sheykhi , n. riazi and m. h. dehghani , _ magnetic branes in ( n+1)-dimensional einstein - maxwell - dilaton gravity _ , phys .
d * 75 * , 044020 ( 2007 ) [ hep - th/0610086 ] .
a. sheykhi , _ topological born - infeld - dilaton black holes _ , phys .
b * 662 * , 7 ( 2008 ) [ arxiv:0710.3827 [ hep - th ] ] .
s. h. hendi , _ rotating black branes in brans - dicke - born - infeld theory _ , j. math .
phys . * 49 * , 082501 ( 2008 ) [ arxiv:0808.2347 [ gr - qc ] ] . s. j. poletti and d. l. wiltshire , _ the global properties of static spherically symmetric charged dilaton space - times with a liouville potential _ , phys .
d * 50 * , 7260 ( 1994 ) [ erratum - ibid .
d * 52 * , 3753 ( 1995 ) ] [ gr - qc/9407021 ] .
s. j. poletti , j. twamley and d. l. wiltshire , _ charged dilaton black holes with a cosmological constant _ , phys .
d * 51 * , 5720 ( 1995 ) [ hep - th/9412076 ] .
s. mignemi and d. l. wiltshire , _ black holes in higher derivative gravity theories _ , phys .
d * 46 * , 1475 ( 1992 ) [ hep - th/9202031 ] .
c. j. gao and s. n. zhang , _ higher dimensional dilaton black holes with cosmological constant _ , phys .
b * 605 * , 185 ( 2005 ) [ hep - th/0411105 ] .
c. j. gao and s. n. zhang , _ topological black holes in dilaton gravity theory _ , phys .
b * 612 * , 127 ( 2005 ) .
a. sheykhi , _ charged rotating dilaton black strings in ads spaces _ , phys .
d * 78 * , 064055 ( 2008 ) [ arxiv:0809.1130 [ hep - th ] ] .
a. sheykhi , m. h. dehghani and s. h. hendi , _ thermodynamic instability of charged dilaton black holes in ads spaces _ , phys .
d * 81 * , 084040 ( 2010 ) [ arxiv:0912.4199 [ hep - th ] ]
. s. h. hendi , a. sheykhi and m. h. dehghani , _ thermodynamics of higher dimensional topological charged ads black branes in dilaton gravity _ , eur .
j. c * 70 * , 703 ( 2010 ) [ arxiv:1002.0202 [ hep - th ] ] . y. c. ong and p. chen , _ stringy stability of charged dilaton black holes with flat event horizon _ ,
jhep * 1208 * , 079 ( 2012 ) [ erratum - ibid .
* 1501 * , 083 ( 2015 ) ] [ arxiv:1205.4398 [ hep - th ] ] .
c. j. gao and s. n. zhang , _ a universe dominated by dilaton field _ , astro - ph/0605682 .
e. radu and d. h. tchrakian , _ new hairy black hole solutions with a dilaton potential _ , class .
* 22 * , 879 ( 2005 ) [ hep - th/0410154 ] .
s. b. giddings , _ the fate of four - dimensions _ , phys . rev .
d * 68 * , 026006 ( 2003 ) [ hep - th/0303031 ] .
s. ryu and t. takayanagi , _ holographic derivation of entanglement entropy from ads / cft _ , phys .
lett . * 96 * , 181602 ( 2006 ) [ hep - th/0603001 ] .
v. e. hubeny , m. rangamani and t. takayanagi , _ a covariant holographic entanglement entropy proposal _ , jhep * 0707 * , 062 ( 2007 ) [ arxiv:0705.0016 [ hep - th ] ] . | we study holographic thermalization in spacetimes with a chemical potential and a non - trivial dilaton field .
three non - local observables are used to probe the whole process and investigate the effect of the ratio of the chemical potential over temperature @xmath0 and the dilaton - maxwell coupling constant @xmath1 .
it is found that the saturation time is not always a monotonically increasing function of @xmath0 , the situation depends on @xmath1 . when @xmath2 , larger @xmath0 yields longer saturation time , while for @xmath3 , the situation becomes more complex .
more interesting , we found that although @xmath1 indeed has influence on the whole thermalization process , it nearly does not affect the saturation time , which indicates the universality of the saturation time for the dual one - parameter field theories . |
the production of hot , highly - ionized gas in galactic environments is closely related to the input of energy and matter from stars and supernovae into the interstellar medium ( ism ) .
such `` feedback '' can shape the ism on kiloparsec scales in regions with high concentrations of early - type stars . in disk galaxies , with differing pressure gradients in the vertical and radial directions , such energy input is responsible for the production of vertically - extended `` halos '' or `` coronae '' about these systems @xcite .
howk et al .
@xcite have recently completed a study of interstellar o vi in the large magellanic cloud ( lmc ) . because the ionization energy required for its creation ( ip@xmath5 ev ) precludes its production via photoionization by starlight , o vi is a tracer of hot ( @xmath6 k ) , collisionally - ionized gas in galactic environments .
therefore , the howk et al .
study of o vi in the lmc provides fundamental information on the content , distribution , and kinematics of material created by the interactions of stars and supernovae with the ism in the closest and best - studied disk galaxy beyond the milky way .
i summarize the principle results of this study below .
table 1 summarizes the statistical properties of the o vi column densities , @xmath7 , along the 12 sight lines in the howk et al .
@xcite study , while figure [ fig : maps ] compares the distribution of lmc o vi with h@xmath8@xcite and _ rosat _ hard x - ray @xcite views of the lmc .
significant o vi is observed across the whole face of the lmc , although the o vi is patchy : the standard deviation of the measurements is @xmath9 of the mean .
sight lines projected onto known superbubbles ( the four sight lines south of @xmath10 in figure [ fig : maps ] ) show only very modest ( if any ) o vienhancements compared with more quiescent - appearing sight lines ( e.g. , nos .
14 in figure [ fig : maps ] ) .
ll + + & + @xmath11 & @xmath12 @xmath13 + @xmath14 $ ] & @xmath15 @xmath13 + @xmath16 & @xmath17 @xmath13 + for comparison with the milky way values reported by savage et al .
@xcite , we define the quantity @xmath18 , where @xmath19 is the inclination of the lmc . this is the column density projected perpendicular to the plane of the lmc and is equivalent to the @xmath20 measurements in the galaxy
. the first _ fuse _ measurements of galactic halo o vi towards extragalactic objects found @xmath21 ( @xmath9 standard deviation ) . the average lmc and milky way @xmath22 values and their standard deviations are identical .
the observed kinematic profiles of the lmc o vi are quite broad , with breadths implying @xmath23 k. figure [ fig : profiles ] shows the normalized o vi absorption profiles along the 12 sight lines studied and the profiles of a moderate - strength fe ii transition .
the latter traces gas associated with neutral material in the lmc disk .
the o vi is much broader than the fe ii along all sight lines and is systematically blue - shifted from the disk ( by @xmath24 km s@xmath2 on average ) .
thus , the o vi is kinematically decoupled from the bulk of the lmc disk material .
there is gas present at velocities compatible with the outflow of material from the lmc disk along all of the sight lines discussed by howk et al . in only two cases
is there possibly o vi at velocities that may indicate infall .
the _ fuse _ observations reveal interstellar o vi associated with the lmc is present in large quantities across the whole face of the lmc , with an average column density and patchiness identical to those of the galactic halo .
lines of sight projected onto superbubbles and supergiant shells have much the same column densities as lines of sight projected onto quiescent regions .
the lmc o vi absorption is both much broader and shifted to lower absolute velocities than the lower - ionization gas ( e.g. , fe ii ) . for reasons discussed in detail by howk et al .
@xcite , the favored interpretation of these salient aspects of the lmc o vi is that the lmc is surrounded by a hot , highly - ionized halo or corona similar in many respects to that found in the milky way that gives rise to the observed o vi absorption .
several models can explain the physics of the o vi production within a gaseous halo about the lmc , including cooling galactic fountain flows and interface models ( such as turbulent mixing layer or conductive interface models ) .
the cooling fountain model provides an elegant explanation for the similarity of the average milky way and lmc o vi column densities .
though these galaxies differ in oxygen abundance by a factor of @xmath0 , the column density of highly - ionized metals in a cooling flow of hot material is independent of abundance @xcite .
the column density of o vi in the edgar & chevalier @xcite models is a function of the ratio @xmath25 , where @xmath26 is the cooling rate ( in protons @xmath13
s@xmath2 ) , and @xmath27 is the initial density of the flow .
the average lmc o vi column density ( table 1 ) corresponds to a one - sided mass - flow rate from the lmc disk of @xmath28 the adopted density is consistent with estimates of electron densities in supergiant shells and diffuse gas using x - ray observations of the lmc @xcite .
it should be noted , however , that the energy input requirements into the ism and mass flow rates from the disk can be significantly different if the o vi arises in turbulent mixing layers or other interface - type models .
observations of other highly - ionized species ( e.g. , c iv ) will be required to distinguish between the cooling flow and interface models .
99 de avillez , m.a .
2000 , mnras , 315 , 479 edgar , r.j . , & chevalier , r.a .
1986 , , 310 , l27 gaustad , j.e . ,
mccullough , p.r .
, rosing , w. , & van buren , d. 2001 , pasp , in press ( see astro - ph/0108518 ) howk , j.c . ,
sembach , k.r . ,
savage , b.d . ,
massa , d. , friedman , s.d . , &
fullerton , a.w .
2002 , , submitted .
norman , c.a .
, & ikeuchi , s. 1989 , , 345 , 372 points , s.d . , chu , y .- h . ,
snowden , s.l .
, & smith , r.c .
2001 , apjs , 136 , 99 savage , b.d .
, et al . 2000 , , 538 , l27 snowden , s.l . , & petre , r. 1994 , , 436 , 123 | i summarize _ far ultraviolet spectroscopic explorer _ ( _ fuse _ ) observations of interstellar o vi absorption towards 12 early - type stars in the large magellanic cloud ( lmc ) , the closest disk galaxy to the milky way .
lmc o vi is seen towards all 12 stars with properties ( average column densities , kinematics ) very similar to those of the milky way halo , even though o / h in the lmc is lower by a factor of @xmath0 .
sight lines projected onto known lmc superbubbles show little enhancement in o vi column density compared to sight lines towards quiescent regions of the lmc .
the o vi absorption is displaced by @xmath1 km s@xmath2 from the corresponding low - ionization absorption associated with the bulk of the lmc gas .
the lmc o vi most likely arises in a vertically - extended distribution , and i discuss the measurements in the context of a halo composed of radiatively - cooling hot gas . in this case , the mass - flow rate from one side of the lmc disk is of the order @xmath3
m@xmath4 yr@xmath2 . |
the production of lepton pairs in hadron collisions , the drell yan process @xcite , is one of the most powerful tools to probe the structure of hadrons .
its parton model interpretation is straightforward
the process is induced by the annihilation of a quark
antiquark pair into a virtual photon which subsequently decays into a lepton pair .
the drell yan process in proton proton or proton nucleus collisions therefore provides a direct probe of the antiquark densities in protons and nuclei . the naive quark parton model predicts scaling of the drell yan cross section in the variable @xmath3 , where @xmath4 is the invariant mass of the lepton pair and @xmath5 is the hadron
hadron centre - of - mass energy .
apart from the invariant mass distribution @xmath6 one usually studies the distribution of the lepton pairs as function of the feynman - parameter @xmath7 or of the hadron
hadron centre - of - mass rapidity @xmath8 where @xmath9 denotes the four - momentum of the lepton pair in the hadron
hadron centre - of - mass system .
the resulting distributions @xmath10 can be directly related to the @xmath11-dependence of the parton distribution functions in beam and target . moreover , most fixed target experiments have only a limited kinematic coverage in @xmath1 or @xmath2 , such that only these distributions can be measured without extrapolation into experimentally inaccessible regions .
a summary of the experimental data on the drell yan process in unpolarized @xmath12- , @xmath13- , @xmath14- and @xmath15-collisions can be found in @xcite .
these data are used to determine the large-@xmath11 behaviour of antiquark distributions in the proton ( e.g. @xcite ) and to study the parton structure of the pion @xcite . in recent times , the difference between drell yan cross sections in @xmath12 and @xmath16 collisions @xcite has provided valuable information on the asymmetry between the @xmath17- and @xmath18-quark distributions in the proton @xcite . up to now , the only experimental information on the polarized parton distributions in the proton comes from deep inelastic lepton
nucleon scattering experiments @xcite , measuring the polarized structure function @xmath19 .
this structure function probes one particular combination of the polarized quark distributions , its mere knowledge is therefore insufficient for a determination of all different quark and antiquark distributions .
this lack of knowledge is reflected in the results of recent fits of polarized parton distribution functions @xcite .
while the polarized valence quark distributions can be determined with some precision from the experimental data , only little information can be extracted on the total magnitude of the polarized sea quark distribution ; the flavour structure of the polarized quark sea is at present completely undetermined .
experimental information on the polarized antiquark distributions can be obtained from several observables .
charged current exchange in polarized electron
proton scattering can probe the polarized weak structure functions @xcite , which are sensitive to a different combination of quark distributions than their electromagnetic counterparts . combining electromagnetic and weak structure functions in a global parton distribution fit
, it should in principle be possible to disentangle the different quark and antiquark flavours
. a direct probe of the polarized sea quark distributions would be the drell yan process in polarized proton
nucleon collisions @xcite .
such a measurement could be feasible at rhic @xcite or at hera @xcite .
the rather large qcd corrections to the _ unpolarized _ drell yan process suggest , that a reliable interpretation of the drell yan cross section in terms of partonic distribution functions is only possible if qcd corrections are included . this paper aims to derive the @xmath20 corrections to the @xmath1- and @xmath2-distributions of lepton pairs in longitudinally polarized hadron
hadron collisions and to study the impact of these corrections on the drell yan cross section . for the unpolarized drell yan process , qcd corrections have been calculated up to @xmath21 for the invariant mass distribution @xcite and up to @xmath0 for the rapidity and @xmath1-distributions @xcite . for the latter ,
parts of the @xmath21 corrections are known @xcite .
inclusion of the @xmath0corrections changes the drell yan cross section at fixed target energies by up to 50% , the @xmath21 corrections amount up to 10% . at small transverse momenta of the lepton pair ,
the qcd corrections are dominated by the emission of soft gluons .
the effects of soft gluon emission at arbitrary order in perturbation theory have been studied for the drell yan process in @xcite , yielding a resummation of the dominant corrections to all orders .
all these results can in principle be applied to the production of vector bosons at hadron colliders . however , in the case of @xmath22boson production , one has to take into account that @xmath22bosons in hadronic collisions can be identified only in their @xmath23 decay mode , where just the lepton is observed .
all above calculations can not make predictions for the kinematic distributions of the vector boson decay products in the hadron
hadron centre - of - mass frame .
separate calculations of the next - to - leading order qcd corrections to the transverse mass distribution of the @xmath22boson @xcite , to the polarization density matrix of its decay products @xcite and to the production rates of @xmath22 + 0,1 jets @xcite in unpolarized hadronic collisions have been performed .
qcd corrections to the drell yan process in polarized hadron
hadron collisions have been derived up to @xmath0 for the invariant mass distribution in longitudinally @xcite and transversely @xcite polarized collisions .
the @xmath0 corrections to the rapidity distributions are known for longitudinally @xcite and transversely @xcite polarized collisions .
furthermore , a resummation of the effects of soft gluons in longitudinally polarized collisions has been performed in @xcite .
a fully consistent study of the drell yan process at next - to - leading order was until now only possible in the unpolarized case , as the polarized parton distributions could only be determined at leading accuracy . with the recently calculated polarized two
loop splitting functions @xcite , the polarized distributions can now be determined to next - to - leading order from fits @xcite to polarized structure function data . having a complete calculation of the two - loop splitting functions available , it is now furthermore possible to define consistent scheme transformation prescriptions @xcite for parton distributions , splitting functions and parton level cross sections at next - to - leading order .
the scheme dependence of the parton level cross sections for polarized deep inelastic scattering has lead to some discussion @xcite about the partonic interpretation of measurements @xcite of polarized lepton
nucleon scattering .
the scheme dependence of the qcd corrections to the polarized drell yan process has been addressed in @xcite .
this paper is organized as follows : in section [ sec : calc ] , we derive the @xmath0corrections to the @xmath1-distribution of drell yan lepton pairs produced in longitudinally polarized hadron hadron collisions .
we discuss the representation of @xmath24 used in our calculation and motivate our choice of factorization scheme .
furthermore , a brief rederivation of the corrections to the invariant mass and @xmath2-distributions is given .
we numerically estimate the importance of these corrections at fixed target energies in section [ sec : num ] .
section [ sec : conc ] contains a brief summary of our results and concluding remarks . in the appendix ,
we collect several identities used in the calculation .
the qcd corrections to the @xmath1-distributions in the unpolarized drell yan process were originally derived using dimensional regularization by altarelli , ellis and martinelli @xcite . in a later publication ,
kubar _ et al . _
@xcite confirmed the results of @xcite and derived the qcd corrections to the @xmath2-distributions .
the calculation of kubar _ et al .
_ was performed using a non
zero gluon mass as regulator .
the results obtained for the @xmath1-distributions with this method were considerably simpler than the original results of @xcite . in the calculation of the qcd corrections to the @xmath1- and @xmath2-distributions in the polarized drell yan process , we shall work in dimensional regularization and follow closely the method of @xcite . using several identities listed in the appendix , we are able to simplify our results into a form similar to the unpolarized results of @xcite . in the following ,
we give an explicit derivation of the corrections to the @xmath1-distribution .
furthermore , we briefly summarize the results of a rederivation of qcd corrections to the @xmath2-distribution within the same method .
these corrections were originally obtained by weber @xcite as a by - product in the calculation of the soft gluon resummation in the polarized drell yan process . disregarding correlations between the final state lepton plane and the direction of the incoming hadrons ,
it is sufficient to consider the production of an off shell photon of invariant mass @xmath25 , the polarization sum of the outgoing photon can be taken to be @xmath26 .
the lepton pair production cross section can be retrieved by multiplying the results with a factor @xmath27 for the decay @xmath28 .
only the lowest order contribution to the drell yan process ( fig .
[ fig : feynman](a ) ) yields a finite cross section .
all higher order parton level cross sections contain infinities . to make these infinities explicit
, we work in dimensional regularization with @xmath29 dimensions .
this treatment faces the problem of describing @xmath24 in @xmath30 .
we shall use the @xmath24prescription @xcite of t hooft , veltman , breitenlohner and maison ( hvbm ) , which has been consistently used in the derivation of the next - to - leading order corrections to the polarized splitting functions @xcite . in this prescription , one restricts the anticommutation property of @xmath24 to the physical four dimensions , while @xmath24 commutes in the remaining @xmath31 dimensions .
the major drawback of this formalism in the @xmath32scheme is the non conservation of the flavour non singlet axial vector currents @xcite due to a non vanishing first moment of the corresponding non
singlet nlo splitting function @xmath33 . to restore the conservation of this
non singlet axial vector current , a further scheme transformation @xcite of the results obtained in the hvbm formalism is needed @xcite contribution to the polarized drell yan process . ] .
we have to apply the same scheme transformation to our results for the parton - level drell yan cross sections at @xmath0 .
we define the hadron cross sections : @xmath34 where @xmath35 and @xmath36 denote positive and negative hadron helicities . in the parton model picture ,
these cross sections can be expressed as convolutions of ( polarized ) parton distributions @xmath37f_i$ ] with parton subprocess cross sections @xmath37\hat{\sigma}_{ij}$ ] : @xmath38 \sigma = \int \d \xa \d \xb \sum_{ij } \d [ \delta ] \hat{\sigma}_{ij}(\xa,\xb,\mu_f^2 ) \,[\delta ] f_i(\xa,\mu_f^2 ) \,[\delta ] f_j(\xb,\mu_f^2).\ ] ] we have defined @xmath39 where @xmath35 and @xmath36 now denote positive and negative parton helicities .
the lowest order bare cross section for the @xmath40 contribution ( fig .
[ fig : feynman].(a ) ) reads in @xmath29 dimensions : @xmath41 where we define @xmath42 , with @xmath43 being the parton
parton centre - of - mass energy .
the factor @xmath44 is common to all contributions and forms part of the overall normalization of the cross section . the virtual gluon correction ( fig .
[ fig : feynman].(b ) ) to this process yields @xmath45 the correction due to real gluon emission ( fig .
[ fig : feynman].(c , d ) ) takes the form : @xmath46\ ! ,
\label{eq : qqreal}\end{aligned}\ ] ] where we have replaced the parton
parton scattering angle @xmath47 by @xmath48 .
terms proportional to @xmath49 have been omitted if they do not yield a finite contribution to the final result . finally , the contribution from the quark
gluon compton scattering process ( fig .
[ fig : feynman].(e , f ) ) reads : @xmath50 .
\label{eq : qgreal}\end{aligned}\ ] ] using eqs .
( [ eq : plusid ] ) , one can make the poles in @xmath49 in the annihilation and compton contributions explicit : @xmath51 \nonumber \\ & & \hspace{-0.2 cm } + \left[\delta(z_1)+\delta(1-z_1)\right]\ , \bigg [ -\frac{1}{\e } \left [ \frac{2}{(1-z)_{+ } } -1-z + \frac{3}{2 } \delta ( 1-z)\right ] \nonumber \\ & & \hspace{0.1 cm } + ( 1-z ) \left[\ln \frac{(1-z)^2}{z } -3 \right ] -2 \ln \frac{(1-z)^2}{z } + 4 \left(\frac{\ln ( 1-z)}{1-z}\right)_{+ } -\frac{2}{1-z } \ln z\bigg ] \nonumber \\ & & \hspace{-0.2 cm } + \left [ \frac{1}{(z_1)_{+ } } + \frac{1}{(1-z_1)_{+ } } \right ]
\,\left [ -1-z+ \frac{2}{(1-z)_{+}}\right ] -2(1-z)\bigg\ } , \label{eq : qqbare}\\ \delta \hat{\sigma}^{1,c } & = & - \frac{4 \pi^2 \alpha e_q^2}{3m^2 } ( 1+\e ) \ ; \frac{\alpha_s}{2\pi}t_f \left(\frac{4\pi\mu^2 } { m^2}\right)^{\e } \frac{1}{\gamma ( 1-\e ) } \nonumber \\ & & \hspace{-0.6 cm } \times \int_0 ^ 1 \d z_1\ , z\bigg\ { -\frac{1}{\e } \delta(z_1 ) \left [ 2z-1\right ] + \delta(z_1 ) \left [ ( 2z-1 ) \ln \frac{(1-z)^2}{z } + 2 - 2z \right ] \nonumber\\ & & \hspace{-0.2 cm } + \frac{1}{(z_1)_{+ } } \left [ 2z-1\right ] + 2 ( 1-z ) - ( 2-z_1 ) ( 1-z)^2 \bigg\ } .
\label{eq : qgbare}\end{aligned}\ ] ] the @xmath1-differential cross section can now be obtained by multiplying the above expressions by @xmath52\right ) .\ ] ] using the definitions @xmath53 one can fix @xmath54 for the @xmath1-distributions .
the parton level cross sections @xmath55 and @xmath56 can then be rewritten using eqs .
( [ eq : delx1])([eq : delx4 ] ) . the resulting parton level cross sections still contain divergences associated with collinear singularities in the initial state .
these collinear singularities are process independent and are absorbed into the bare parton distribution functions by a mass factorization procedure .
the only distribution relevant in our case is the polarized quark distribution . performing the mass factorization in the @xmath32scheme and making a scheme transformation @xcite to remove spurious terms due to the hvbm prescription for @xmath24 yields for the bare quark distribution up to @xmath0 : @xmath57
\delta q(x / x ' ) \nonumber \\ & & - \frac{\alpha_s}{2\pi } \int_x^1 \frac{\d x'}{x ' } \left [ -\frac{1}{\e } \frac{1}{\gamma(1-\e ) } \left(\frac{4\pi\mu^2}{\mu_f^2}\right)^{\e } \delta p_{qg } ( x ' ) \right ] \delta g(x / x ' ) , \label{eq : mfac}\end{aligned}\ ] ] with @xcite : @xmath58 , \qquad \delta z_{qq}(x ) = -4c_f ( 1-x).\\ \delta p_{qg } ( x ) & = & t_f \left [ 2x-1 \right].\end{aligned}\ ] ] after mass factorization , the polarized cross sections for the drell yan production of lepton pairs at nlo can be expressed in a form similar to the unpolarized cross sections @xcite : @xmath59 \big\ { \delta q_i(x_1,\mu_f^2)\delta\bar{q}_i(x_2,\mu_f^2 ) \nonumber \\ & & \hspace{0.1 cm } + \delta\bar{q}_i(x_1,\mu_f^2)\delta q_i(x_2,\mu_f^2 ) \big\ } \nonumber \\ & & \hspace{-0.2 cm } + \bigg [ \frac{\alpha_s}{2\pi } \frac{\d \delta \hat{\sigma}_{qg}^{(1 ) } } { \d m^2 \d x_f } \left(x_1,x_2 , \frac{m^2}{\mu_f^2}\right ) \delta g(x_1,\mu_f^2 ) \left\ { \delta q_i(x_2,\mu_f^2 ) + \delta \bar{q}_i ( x_2,\mu_f^2 ) \right\ } \nonumber \\ & & \hspace{0.3 cm } + ( 1 \leftrightarrow 2)\bigg ] \bigg\ } , \label{eq : xfmaster}\end{aligned}\ ] ] with @xmath60 \nonumber \\ & & + \bigg(\frac{\delta ( \xa-\xaa)}{x_1 ^ 0+x_2 ^ 0 } \bigg [ \frac{1}{\xb } - \frac{\xbb}{\xb^2 } - \frac{\xbb\,^2+\xb^2}{\xb^2(\xb-\xbb ) } \ln \frac{\xbb}{\xb } \nonumber \\ & & \hspace{0.6 cm } + \frac{\xbb\,^2+\xb^2}{\xb^2 } \left(\frac{\ln ( 1-\xbb/\xb)}{\xb-\xbb}\right)_{+ } + \frac{\xbb\,^2+\xb^2}{\xb^2 } \frac{1}{\left(\xb-\xbb\right)_{+ } } \nonumber\\ & & \hspace{0.6 cm } \ln \frac { ( \xaa+\xbb)(1-\xaa)}{\xaa ( \xaa+\xb ) } \bigg ] + ( 1 \leftrightarrow 2 ) \bigg ) \nonumber \\ & & + \frac{\delta \tilde{g}^a(\xa,\xb)}{\left[(\xa-\xaa)(\xb-\xbb)\right]_{+ } } + \delta\tilde{h}^a ( \xa,\xb ) \nonumber \\ & & + \ln \frac{m^2}{\mu_f^2 } \bigg\ { \frac{\delta ( x_1-x_1 ^ 0)\ , \delta ( x_2-x_2 ^ 0)}{x_1 ^ 0+x_2 ^ 0 } \bigg [ 3 + 2 \ln \frac{1-\xaa}{\xaa } + 2 \ln \frac{1-\xbb}{\xbb } \bigg ] \nonumber\\ & & \hspace{0.6 cm } + \bigg ( \frac{\delta ( \xa-\xaa)}{x_1 ^ 0+x_2 ^ 0 } \frac{\xbb\,^2+\xb^2}{\xb^2}\frac{1}{\left(\xb-\xbb\right)_{+ } } + ( 1 \leftrightarrow 2 ) \bigg ) \bigg\ } \bigg\ } \\ & = & - \frac{\d \hat{\sigma}_{q\bar{q}}^{(1 ) } } { \d m^2 \d x_f } \left(x_1,x_2,\frac{m^2}{\mu_f^2}\right)\ ; , \nonumber\\ \frac{\d \delta \hat{\sigma}_{qg}^{(1 ) } } { \d m^2 \d x_f } \left(x_1,x_2,\frac{m^2}{\mu_f^2}\right ) & = & -t_f\bigg\{\frac{\delta ( \xb-\xbb ) } { ( \xaa+\xbb)\xa^2 } \bigg [ ( 2\xaa -\xa ) \ln\frac{(\xaa+\xbb)(1-\xbb)(\xa-\xaa)}{\xaa\xbb(\xa+\xbb ) } \nonumber \\ & & \hspace{0.6 cm } + 2(\xa-\xaa)\bigg ] + \frac{\delta\tilde{g}^c(\xa,\xb)}{(\xb-\xbb)_{+ } } + \delta\tilde{h}^c(\xa,\xb ) \nonumber \\ & & + \ln \frac{m^2}{\mu_f^2 } \bigg\ { \frac{\delta ( \xb-\xbb ) } { ( \xaa+\xbb)\xa^2 } ( 2\xaa -\xa ) \bigg\}\bigg\}\ ; , \end{aligned}\ ] ] where @xmath61 inserting @xmath62\right)\ ] ] into eqs.([eq : qqbare])([eq : qgbare ] ) and using identities ( [ eq : delx3]),([eq : dely1])([eq : dely3 ] ) , one can derive the @xmath0corrections to the rapidity distribution of the drell yan pairs .
the variables @xmath63 are in this case defined by @xmath64 they are in general different from the @xmath65 and @xmath66 occuring in the @xmath1-distributions .
these corrections have already been calculated in @xcite ; we find numerical agreement with the results ( keeping in mind that a different factorization prescription for the gluonic contribution was used in @xcite ) , our expressions are however considerably simpler : @xmath67 \big\ { \delta q_i(x_1,\mu_f^2)\delta\bar{q}_i(x_2,\mu_f^2 ) \nonumber \\ & & \hspace{0.1 cm } + \delta\bar{q}_i(x_1,\mu_f^2)\delta q_i(x_2,\mu_f^2 ) \big\ } \nonumber \\ & & \hspace{-0.2 cm } + \bigg [ \frac{\alpha_s}{2\pi } \frac{\d \delta \hat{\sigma}_{qg}^{(1 ) } } { \d m^2 \d y } \left(x_1,x_2 , \frac{m^2}{\mu_f^2}\right ) \delta g(x_1,\mu_f^2 ) \left\ { \delta q_i(x_2,\mu_f^2 ) + \delta \bar{q}_i ( x_2,\mu_f^2 ) \right\ } \nonumber \\ & & \hspace{0.3 cm } + ( 1 \leftrightarrow 2)\bigg ] \bigg\ } , \label{eq : ymaster}\end{aligned}\ ] ] with @xmath68 \nonumber \\ & & + \bigg ( \delta ( \xa-\xaa ) \bigg [ \frac{1}{\xb } - \frac{\xbb}{\xb^2 } - \frac{\xbb\,^2+\xb^2}{\xb^2(\xb-\xbb ) } \ln \frac{\xbb}{\xb } \nonumber \\ & & \hspace{0.6 cm } + \frac{\xbb\,^2+\xb^2}{\xb^2 } \left(\frac{\ln ( 1-\xbb/\xb)}{\xb-\xbb}\right)_{+ } + \frac{\xbb\,^2+\xb^2}{\xb^2 } \frac{1}{\left(\xb-\xbb\right)_{+ } } \nonumber\\ & & \hspace{0.6 cm } \ln \frac { 2\xbb(1-\xaa)}{\xaa(\xb+\xbb ) } \bigg ] + ( 1 \leftrightarrow 2 ) \bigg ) \nonumber \\ & & + \frac{\delta g^a(\xa,\xb)}{\left[(\xa-\xaa)(\xb-\xbb)\right]_{+ } } + \delta h^a ( \xa,\xb ) \nonumber \\ & & + \ln \frac{m^2}{\mu_f^2 } \bigg\ { \delta ( x_1-x_1 ^ 0)\ , \delta ( x_2-x_2 ^ 0 ) \bigg [ 3 + 2 \ln \frac{1-\xaa}{\xaa } + 2 \ln \frac{1-\xbb}{\xbb } \bigg ] \nonumber\\ & & \hspace{0.6 cm } + \bigg ( \delta ( \xa-\xaa ) \frac{\xbb\,^2+\xb^2}{\xb^2}\frac{1}{\left(\xb-\xbb\right)_{+ } } + ( 1 \leftrightarrow 2 ) \bigg ) \bigg\ } \bigg\ } \\ & = & - \frac{\d \hat{\sigma}_{q\bar{q}}^{(1 ) } } { \d m^2 \d y } \left(x_1,x_2,\frac{m^2}{\mu_f^2}\right)\ ; , \nonumber \\
\frac{\d \delta \hat{\sigma}_{qg}^{(1 ) } } { \d m^2 \d y } \left(x_1,x_2,\frac{m^2}{\mu_f^2}\right ) & = & -t_f\bigg\{\frac{\delta ( \xb-\xbb ) } { \xa^2 } \bigg [ ( 2\xaa -\xa ) \ln\frac{2(\xa-\xaa)(1-\xbb)}{(\xa+\xaa)\xbb } \nonumber \\ & & \hspace{0.6 cm } + 2(\xa-\xaa)\bigg ] + \frac{\delta g^c(\xa,\xb)}{(\xb-\xbb)_{+ } } + \delta h^c(\xa,\xb ) \nonumber \\ & & + \ln \frac{m^2}{\mu_f^2 } \bigg\ { \frac{\delta ( \xb-\xbb ) } { \xa^2 } ( 2\xaa -\xa ) \bigg\}\bigg\}\ ; , \end{aligned}\ ] ] where @xmath69 as mentioned above , the @xmath0corrections to the @xmath1- and @xmath2-distributions in the unpolarized drell - yan process were derived in @xcite .
a summary of these formulae in the @xmath32scheme in our notation can be found in @xcite . finally ,
integration of eqs.([eq : qqbare])([eq : qgbare ] ) with respect to @xmath70 reproduces after mass factorization using eq .
( [ eq : mfac ] ) the correction terms to the drell yan mass distribution @xcite . these read in the @xmath32scheme : @xmath71 with : @xmath72 \ , , \nonumber \\ & = & - c_{q\barq}^{dy,(1 ) } \left(z,\frac{m^2}{\mu_f^2}\right)\nonumber \\
\delta c_{qg}^{dy,(1 ) } \left(z,\frac{m^2}{\mu_f^2}\right ) & = & -t_f \bigg [ \left(2z-1\right ) \ln \frac{(1-z)^2}{z } + \frac{5}{2}-z-\frac{3}{2 } z^2 + \ln \frac{m^2}{\mu_f^2 } \left\{2z-1\right\ } \bigg ] \ , .\nonumber\end{aligned}\ ] ]
in the following , we will illustrate the impact of the qcd corrections to the polarized drell yan process derived above by numerically evaluating ( [ eq : xfmaster ] ) and ( [ eq : ymaster ] ) for recent parameterizations of the polarized parton distribution functions .
when using the polarized gs parton distribution functions @xcite , we take @xmath73 mev , the corresponding unpolarized cross sections are then evaluated using the unpolarized mrs parton distribution functions set a@xmath74 @xcite .
the polarized grsv distributions @xcite are consistently used in combination with the unpolarized distributions from grv @xcite and for @xmath75 mev .
if not stated otherwise , we shall always use @xmath76 ; the strong coupling constant @xmath77 is evaluated at @xmath78 .
all results in this section are obtained for @xmath79 gev , corresponding to a fixed target experiment in the hera proton beam @xcite with @xmath80 gev . at present
, all polarized parton distributions are fitted to data on the polarized structure function @xmath19 only . as this structure function
is dominated by valence quark contributions for @xmath81 , the knowledge on the polarized sea quark distributions in the large-@xmath11 region is very poor .
this uncertainty is illustrated in fig .
[ fig : fig2 ] , which shows the next - to - leading order predictions ( eq . ( [ eq : signlody ] ) ) for the drell yan asymmetry @xmath82 obtained using the polarized parton distribution functions of @xcite ( gs(a c ) ) and @xcite ( grsvs , v ) . at present , neither magnitude nor sign of this asymmetry can be predicted .
it should therefore be clear , that all numerical results for the polarized drell yan process presented in the remainder of this section do not make solid predictions for observable asymmetries .
the purpose of these numerical calculations is only to illustrate the magnitude of the next - to - leading order corrections and to demonstrate the perturbative stability of the predictions . in the following
, we shall always study the cross section asymmetries @xmath83 as the unpolarized parton distributions are known to some level of accuracy , it is possible to predict the cross sections in the denominator of @xmath84 and @xmath85 .
these cross sections are shown for proton proton collisions ( @xmath79 gev ) in fig .
[ fig : fig3 ] , which is obtained with the mrs(a@xmath74 ) parton distributions @xcite .
the prediction using the grv parton distributions @xcite agrees within 3% over the whole kinematic range .
it can be seen from fig .
[ fig : fig2 ] that the total drell yan asymmetry @xmath86 changes with the invariant mass @xmath4 .
the change is however not uniform over the whole range in @xmath1 or @xmath2 as illustrated in fig .
[ fig : fig4 ] for the polarized gs(a ) parton distributions .
the @xmath1-dependence of the asymmetry can be understood from eq .
( [ eq : xfdef ] ) .
the asymmetry around @xmath87 always probes the polarized parton distributions in beam @xmath88 and target @xmath89 around @xmath90 . if @xmath1 is increased towards its positive kinematic limit , @xmath65 increases towards 1 while @xmath66 decreases only slightly .
decreasing @xmath1 towards the negative kinematic limit yields a small decrease of @xmath65 while @xmath66 increases towards 1 .
the region @xmath91 therefore probes the @xmath11-dependence of the parton distributions in the beam hadron , while @xmath92 is sensitive on the target hadron structure .
moreover , different invariant masses correspond to different @xmath11 intervals probed .
a similar argumentation , based on eq . ( [ eq : ydef ] ) , applies to the @xmath2-dependence of the asymmetry .
the contributions of the individual subprocesses ( @xmath93annihilation at leading and next - to - leading order and quark gluon compton scattering ) to the polarized drell yan cross section are illustrated in fig .
[ fig : fig5 ] .
all curves are obtained with the polarized gs(a ) parton densities and are shown for @xmath94 gev .
all polarized subprocess contributions are normalized to the _ full _ unpolarized cross section at next - to - leading order .
the relative magnitude of the individual contributions is similar to the unpolarized drell yan process .
the @xmath0 correction to the @xmath93annihilation process enhances significantly the lowest order prediction while the quark gluon compton process contributes with a sign opposite to the annihilation process .
however , the relative magnitude of annihilation and compton process depends on the magnitude of the gluon distribution at large @xmath11 , which is completely undetermined at present @xcite .
this uncertainty prevents a sensible prediction of a @xmath95factor between the cross sections at leading and next - to - leading order .
furthermore , fig .
[ fig : fig5 ] shows the prediction for the drell yan asymmetry if both polarized and unpolarized cross section are evaluated using the leading order @xmath96-coefficient function only ( long dashed line ) .
the sizable shift between the leading and next - to - leading order predictions for the drell yan asymmetry seems to contradict the claim @xcite that the next - to - leading order corrections , although large in polarized and unpolarized cross sections , would largely cancel in the drell yan asymmetry .
the change of physical quantities such as cross sections or asymmetries under variations of the ( unphysical ) mass factorization scale @xmath78 enables an estimate of the numerical importance of unknown higher order corrections .
we quantify this uncertainty in fig .
[ fig : fig6 ] , showing the drell yan asymmetry for @xmath97 .
it can be seen that the absolute value of the asymmetry changes less than 0.01 in the central region . even in the region where @xmath1 ( or @xmath2 ) approaches its kinematic limit ,
the absolute value of the asymmetry changes only by a maximum of 0.015 .
this stability is crucial for an estimate of the theoretical error on the polarized sea quark distributions extracted from the polarized drell yan process .
a final point briefly concerns phenomenological aspects of the polarized drell yan process , although this is not the principal aim of the present paper .
figures [ fig : fig7 ] and [ fig : fig8 ] show the drell yan asymmetry in polarized proton proton and proton deuterium collisions for @xmath94 gev as predicted by using the different polarized parton distributions suggested in @xcite and @xcite .
it can clearly be seen that the different predictions vary within a band of @xmath98 in @xmath12 and @xmath99 in @xmath16 collisions .
this uncertainty reflects the present lack of knowledge on the polarization of the light quark sea in the proton at @xmath100 .
the difference between the predictions is furthermore substantially larger than the theoretical uncertainty due to the choice of the factorization scale . due to the smaller overall valence quark polarization in deuterium ,
one finds the asymmetry in @xmath16 collisions to be systematically smaller than in @xmath12 collisions .
although the normalization of the asymmetry in the proton direction ( @xmath101 ) and in the deuteron direction ( @xmath102 ) is slightly different , its shape is still very similar in both regions .
this reflects the assumption of an su(2)@xmath103symmetry of the polarized antiquark sea , which is used in all polarized parton distributions displayed in the figure .
a measurement of the @xmath1-(or @xmath2-)dependence of the drell yan asymmetry in @xmath16 collisions could therefore test this assumption .
we have presented a complete calculation of the @xmath0 corrections to the @xmath1- and @xmath2-dependence of the longitudinally polarized drell yan cross section in the @xmath32scheme .
the results for the @xmath2-dependence and the mass distribution agree ( after suitable change of the factorization scheme ) with earlier results in the literature @xcite .
we have demonstrated that the impact of these corrections on the polarized drell yan cross section is very similar to the impact of the unpolarized corrections .
the corrections due to the individual subprocesses turn out to be sizable .
however , a @xmath95factor between leading and next - to - leading order results can at present not be sensibly predicted due to the uncertainty on the magnitude of the polarized gluon distribution .
the corrections derived in this paper are of particular importance for the extraction of the polarized sea quark distributions from measurements of the drell yan process .
as discussed above , a measurement of the asymmetry as function of @xmath1 or @xmath2 can be directly inferred into the @xmath11-dependence of the polarized sea quark distributions .
the results obtained in this paper can be furthermore immediately keeping in mind the limitations discussed in the introduction applied to the production of massive vector bosons in polarized proton
proton collisions at rhic @xcite . with the knowledge of the next - to - leading order corrections ,
it is furthermore possible to quantify the uncertainty of the theoretical prediction and hence the theoretical error on a measurement of the polarized sea quark distributions .
we have demonstrated that the absolute value of the drell yan asymmetry in the central region varies by less than 0.01 under variation of the mass factorization scale , demonstrating the perturbative stability of the prediction .
this uncertainty is far smaller than the deviation between predictions obtained for different parameterizations of the polarized sea quark distributions .
the @xmath49expansion of the singular terms in the parton level cross sections of section [ sec : calc ] yields @xmath105 integration over @xmath70 yields the following `` + ' ' functions , all defined on the interval @xmath106 $ ] : @xmath107 and @xmath108-functions : @xmath109 furthermore , various products of the above terms appear . by construction
, we have @xmath110 while @xmath111 is fixed by the definition of @xmath1 or @xmath2 respectively . as the parton level cross section is always integrated over the parton momentum fractions @xmath112 and @xmath113 , it is desirable to re - express the above functions in terms of @xmath114 and @xmath115 \frac{\ln(1-x_i^0/x_i)}{x_i - x_i^0 } , \\
\frac{1}{[(\xa-\xaa)(\xb-\xbb)]_{+ } } & \mbox{with } & \int_{\xaa}^{1 } \d \xa \int_{\xbb}^{1 } \d \xb \frac{f(\xa,\xb)}{[(\xa-\xaa)(\xb-\xbb)]_{+ } } \nonumber \\ & & \equiv \int_{\xaa}^{1 } \d \xa\int_{\xbb}^{1 } \d \xb \frac{f(\xa,\xb)-f(\xaa,\xb)-f(\xa,\xbb)+f(\xaa,\xbb)}{(\xa-\xaa)(\xb-\xbb ) } .
\nonumber\end{aligned}\ ] ] we first consider the @xmath1-distributions .
all terms containing @xmath116 can be simplified by @xmath117\right ) = \frac{\xaa \xbb}{\xaa + \xbb}\ , \delta ( \xa-\xaa)\delta(\xb-\xbb ) \label{eq : delx1}\ ] ] and subsequent integration over @xmath70 . for the remaining terms
, we integrate over @xmath70 using the @xmath1-fixing @xmath108-function , which yields @xmath118 as the integrations over the parton momenta @xmath112 and @xmath113 are independent from each other , we can perform them in arbitrary order . for terms containing @xmath119 and @xmath120 only , we choose to integrate @xmath113 first . for terms containing @xmath121 and @xmath122 only ,
@xmath112 is integrated first .
we can then make the following replacements : @xmath123 , \nonumber \\ \delta(1-z_1^ { * } ) & = & \frac{(1-z)\xaa(\xaa+\xb)}{\xaa+\xbb}\,\delta(\xa-\xaa ) , \label{eq : delx2 } \\ \frac{1}{(1-z_1^{*})_{+ } } & = & \frac{(1-z)\xa(\xa+\xb)}{\xa+\xbb } \left [ \frac{1}{(\xa-\xaa)_{+ } } + \delta ( \xa-\xaa ) \ln \frac{(\xaa+\xbb)(1-\xaa ) \xb}{\xaa(\xaa+\xb)(\xb-\xbb)}\right ] .
\nonumber \end{aligned}\ ] ] the above identities for @xmath119 and @xmath124 can be safely applied to terms containing `` @xmath104''functions in @xmath125 .
these can then be rewritten as @xmath126 , \nonumber \\ \delta(\xa-\xaa ) \left(\frac{\ln(1-z)}{1-z}\right)_{+ } & = & \delta(\xa-\xaa ) \bigg [ \xb \left(\frac{\ln ( 1-\xbb/\xb ) } { \xb-\xbb } \right)_{+ } \label{eq : delx3 } \\ & & \hspace{1.5 cm } + \xbb \left ( \li_2(\xbb ) - \frac{\pi^2}{6 } + \frac{1}{2 } \ln^2 ( 1-\xbb ) \right ) \delta ( \xb-\xbb ) \bigg]\nonumber\end{aligned}\ ] ] and similar for @xmath127 .
these identities are however inapplicable to the only remaining term @xmath128 \frac{1}{(1-z)(\xa+\xb)},\ ] ] as @xmath129 and @xmath130 appearing in ( [ eq : delx2 ] ) are not defined for @xmath131 . to overcome this problem , we rewrite @xmath132 and similar for @xmath127 . after subsequent algebraic manipulations and integration over @xmath133 and @xmath134
, we obtain : @xmath135 \frac{1}{(1-z)(\xa+\xb)}= } \nonumber \\ & & \hspace{0.4 cm } \delta ( \xa-\xaa ) \delta ( \xb-\xbb ) \frac{\xaa
\xbb}{\xaa + \xbb } \bigg [ \ln \frac{\xaa}{1-\xaa } \ln
\frac{\xbb}{1-\xbb } - \li_2(\xaa ) -\frac{1}{2 } \ln^2 ( 1-\xaa ) \nonumber \\ & & \hspace{1.5 cm } - \li_2(\xbb ) -\frac{1}{2 } \ln^2 ( 1-\xbb ) + \frac{\pi^2}{2 } \bigg ] \nonumber \\ & & + \ ; \delta ( \xa-\xaa ) \frac{\xaa\xb}{\xaa+\xbb } \bigg [ \frac{1 } { ( \xb-\xbb)_{+ } } \ln \frac{(\xaa+\xbb)(1-\xaa)}{\xaa(\xaa+\xb)}- \left(\frac{\ln ( 1-\xbb/\xb)}{\xb-\xbb}\right)_{+ } \bigg ] \nonumber \\ & & + \ ; \delta ( \xb-\xbb ) \frac{\xa\xbb}{\xaa+\xbb } \bigg [ \frac{1 } { ( \xa-\xaa)_{+ } } \ln \frac{(\xaa+\xbb)(1-\xbb)}{\xbb(\xa+\xbb ) } - \left(\frac{\ln ( 1-\xaa/\xa)}{\xa-\xaa}\right)_{+ } \bigg ] \nonumber \\ & & + \ ; \frac{(\xa + \xb ) \xa\xb}{(\xaa+\xb ) ( \xa+\xbb ) } \ , \frac{1}{[(\xa-\xaa)(\xb-\xbb)]_{+ } } \label{eq : delx4}\end{aligned}\ ] ] analogous relations can be obtained for the @xmath2-distributions , where we can first make the replacement @xmath136\right ) = \xaa \xbb \delta(\xa-\xaa)\delta(\xb-\xbb ) .
\label{eq : dely1}\ ] ] evaluation of the @xmath70integral in the remaining terms yields @xmath137 only the identities ( [ eq : delx3 ] ) are unchanged for the @xmath2-distributions , the remaining identities read : @xmath138 , \nonumber \\ \delta(1-z_1^ { * } ) & = & \frac{(1-z)(1+z)\xaa}{2z}\,\delta(\xa-\xaa ) , \label{eq : dely2 } \\ \frac{1}{(1-z_1^{*})_{+ } } & = & \frac{(1-z)\xa(\xa\xbb+\xaa\xb)}{\xbb(\xa+\xaa ) } \left [ \frac{1}{(\xa-\xaa)_{+ } } + \delta ( \xa-\xaa ) \ln \frac{2\xb\xbb(1-\xaa ) } { \xaa(\xb+\xbb)(\xb-\xbb)}\right ] , \nonumber\end{aligned}\ ] ] and @xmath135 \frac{1}{1-z}= } \nonumber \\ & & \hspace{0.4 cm } \delta ( \xa-\xaa ) \delta ( \xb-\xbb ) \xaa \xbb \bigg [ \ln \frac{\xaa}{1-\xaa } \ln \frac{\xbb}{1-\xbb } - \li_2(\xaa ) -\frac{1}{2 } \ln^2 ( 1-\xaa ) \nonumber \\ & & \hspace{1.5 cm } - \li_2(\xbb ) -\frac{1}{2 } \ln^2 ( 1-\xbb ) + \frac{\pi^2}{2 } \bigg ] \nonumber \\ & & + \ ;
\delta ( \xa-\xaa ) \frac{\xaa\xb(\xb+\xbb)}{2\xbb } \bigg [ \frac{1 } { ( \xb-\xbb)_{+ } } \ln \frac{2\xbb(1-\xaa)}{\xaa(\xb+\xbb)}- \left(\frac{\ln ( 1-\xbb/\xb)}{\xb-\xbb}\right)_{+ } \bigg ] \nonumber \\ & & + \ ; \delta ( \xb-\xbb ) \frac{\xa\xbb(\xa+\xaa)}{2\xaa } \bigg [ \frac{1 } { ( \xa-\xaa)_{+ } } \ln \frac{2\xaa(1-\xbb)}{\xbb(\xa+\xaa ) } - \left(\frac{\ln ( 1-\xaa/\xa)}{\xa-\xaa}\right)_{+ } \bigg ] \nonumber \\ & & + \ ; \frac{(\xa\xbb+\xaa\xb)^2\xa\xb}{\xaa\xbb(\xa+\xaa)(\xb+\xbb ) } \ , \frac{1}{[(\xa-\xaa)(\xb-\xbb)]_{+ } } \label{eq : dely3}\end{aligned}\ ] ] slac - yale collaboration : m.j .
* 37 * ( 1976 ) 1261 ; g. baum _
et al . _ ,
* 45 * ( 1980 ) 2000 ; * 51 * ( 1983 ) 1135 . + emc collaboration : j. ashman _
et al . _ , nucl . phys .
* b328 * ( 1989 ) 1 . + slac - e142 collaboration : p.l .
anthony _ et al .
lett . * 71 * ( 1993 ) 959 ; phys . rev .
* d54 * ( 1996 ) 6620 .
+ smc collaboration : b. adeva _
et al . _ ,
b302 * ( 1993 ) 533 ; * b369 * ( 1996 ) 93 ; d. adams _ et al . _ , phys .
* b329 * ( 1994 ) 399 ; * b336 * ( 1994 ) 125 ; * b357 * ( 1995 ) 248 .
+ slac - e143 collaboration : k. abe _ et al .
* 74 * ( 1995 ) 346 ; * 75 * ( 1995 ) 25 ; * 76 * ( 1996 ) 587 ; phys . lett .
* b364 * ( 1995 ) 61 ; preprint slac - pub-96 - 7242 ( hep - ex/9701004 ) .
m. anselmino , p. gambino and j. kalinowski , torino preprint dftt-44 - 96 ( hep - ph/9607427 ) .
+ m. maul and a. schfer , phys
b390 * ( 1997 ) 437 .
+ m. anselmino _ et al .
_ , proceedings of the workshop `` future physics at hera '' , 1995/96 , eds .
g. ingelman , a. de roeck and g. klanner , desy ( hamburg , 1996 ) , p.827 .
s. gupta , d. indumathi and m.v.n .
murthy , z. phys . *
c42 * ( 1989 ) 493 , erratum * c44 * ( 1989 ) 356
. + h.y .
cheng and s.n .
lai , phys . rev .
* d41 * ( 1990 ) 91 .
+ p. mathews and r. ramachandran , z. phys .
* c53 * ( 1992 ) 305 .
+ p. mathews and v. ravindran , mod .
a7 * ( 1992 ) 2695
. + t. gehrmann and w.j .
stirling , proceedings of the workshop `` future physics at hera '' , 1995/96 , eds .
g. ingelman , a. de roeck and g. klanner , desy ( hamburg , 1996 ) , p.847 .
m. anselmino _ et al .
_ , proceedings of the workshop `` future physics at hera '' , 1995/96 , eds .
g. ingelman , a. de roeck and g. klanner , desy ( hamburg , 1996 ) , p.837 .
korotkov and w.d .
nowak , desy preprint desy97004 ( hep - ph/9701371 ) .
g. altarelli , r.k .
ellis and g. martinelli , nucl .
b143 * ( 1978 ) 521 , erratum * b146 * ( 1978 ) 544 . + j. abad and b.
humpert , phys . lett . * 80b * ( 1979 ) 286 . + j. kubar - andr and f.e .
paige , phys .
d19 * ( 1979 ) 221 .
+ b. humpert and w.l .
van neerven , phys . lett . *
84b * ( 1979 ) 327 , erratum * 85b * ( 1979 ) 471 ; * 89b * ( 1979 ) 69 ; nucl . phys .
* b184 * ( 1981 ) 225 .
contogouris and j. kripfganz , phys .
rev . * d20 * ( 1979 ) 2295
. + a.n .
schellekens and w.l .
van neerven , phys .
* d21 * ( 1980 ) 2619 ; * d22 * ( 1980 ) 1623 .
+ t. matsuura and w.l .
van neerven , z. phys .
* c38 * ( 1988 ) 623 .
+ t. matsuura , s.c .
van der marck and w.l .
van neerven , phys . lett .
* b211 * ( 1988 ) 171 ; nucl . phys .
* b319 * ( 1989 ) 570 .
+ r. hamberg , t. matsuura and w.l .
van neerven , nucl . phys .
* b345 * ( 1990 ) 331 ; * b359 * ( 1991 ) 343 .
van neerven and e.b .
zijlstra , nucl . phys .
* b382 * ( 1992 ) 11 .
davies and w.j .
stirling , nucl .
b244 * ( 1984 ) 337 .
+ c.t.h .
davies , w.j . stirling and b.r .
webber , nucl .
b256 * ( 1985 ) 413 .
+ g. altarelli , r.k .
ellis , m. greco and g. martinelli , nucl .
b246 * ( 1984 ) 12 .
+ g. altarelli , r.k .
ellis and g. martinelli , z. phys .
* c27 * ( 1985 ) 617 .
g. altarelli and g.g .
ross , phys .
* b212 * ( 1988 ) 391 . + r.d .
carlitz , j.c .
collins and a.h .
mueller , phys . lett . *
b214 * ( 1988 ) 229 .
+ g. altarelli and b. lampe , z. phys . *
c47 * ( 1990 ) 315 .
+ g.t . bodwin and j. qiu ,
phys . rev .
* d41 * ( 1990 ) 2755 .
+ w. vogelsang , z. phys .
* c53 * ( 1991 ) 275 . | in this paper we calculate the @xmath0 corrections to the @xmath1- and @xmath2-distributions of lepton pairs produced in collisions of longitudinally polarized hadrons . the numerical importance of these corrections is studied and consequences for the extraction of the polarized sea quark distributions from a measurement of the longitudinally polarized drell yan cross section are discussed .
-1 cm 22.5 cm desy97014 + february 1997 + 3.5 cm * qcd corrections to the longitudinally polarized * .1 cm * drell yan process * 1.cm t. gehrmann .4 cm _
desy , theory group , d-22603 hamburg , germany _
3.cm
pacs : 13.88.+e;13.75.cs;13.85.qk;12.38.bx 0.3 cm keywords : lepton pair production , polarization , hadron hadron collisions . |
pixel lensing , gravitational microlensing of unresolved stars , was first proposed by crotts ( 1992 ) and baillon et al .
classical lensing experiments are restricted to monitoring only fields of resolvable stars , e.g. , toward the galactic bulge and magellanic clouds ( alcock et al .
1995 ; udalski et al . 1995 ; ansari et al . 1995
; alard et al .
by contrast , pixel lensing can overcome this restriction and thus can extend our ability to probe both galactic and extragalactic massive compact objects ( machos ) .
currently , both groups have demonstrated the technique by detecting various types of variable stars ( ansari et al .
1996 ; tomaney & crotts 1996 ) for classical lensing , the light curve is obtained by subtracting the reference flux , @xmath5 , from the amplified flux , @xmath6 : @xmath7 where @xmath8 is the flux of the lensed star after ( or before ) being lensed and @xmath9 is the residual flux from stars not participating in the event .
the amplification is related to the lensing parameters by @xmath10 where @xmath11 is the inverse einstein ring crossing time , @xmath12 is the angular impact parameter in units of angular einstein ring radius , @xmath13 , and @xmath14 is the time of maximum amplification .
the angular einstein ring radius is related to the physical einstein ring radius , @xmath15 , and to the physical parameters of the macho by @xmath16 where @xmath17 , @xmath18 , and @xmath19 are the distances between the observer , source , and lens , and @xmath20 is the mass of the lens . on the other hand ,
it is impossible to measure directly the absolute values of @xmath6 and @xmath5 for the case of pixel lensing because stars can not be resolved .
however , the light curve , @xmath21 , still can be obtained by subtracting the reference image from the image in which a lensing event is occurring .
due to its characteristic differences from classical microlensing observations , additional analysis is required to estimate the pixel lensing event rate . for classical lensing ,
an event with modest maximum amplification ( e.g. , @xmath22 ) can often be detected with high signal - to - noise ratio , @xmath23 , when observed for a few minutes on a 1 m telescope .
this is partly because source stars are nearby , and thus bright , and partly because the residual flux , @xmath24 , is moderate or negligible .
note , however , that @xmath24 is not exactly zero due to the blending of star light in dense fields . for the case of pixel lensing , however , stars are located relatively far away , and thus yield low signal .
additionally , the signal suffers from significant noise from background stars that are not being lensed .
therefore , the event detection rate per observed star is significantly lower than that from classical lensing and thus events can be detected only when either the amplification or the intrinsic flux is very high .
however , the loss of detection rate per star is compensated by the fact that many more stars are monitored .
therefore , one can obtain an equivalent number of total detectable events to that of classical lensing experiments . in this paper ,
i show that one can detect pixel lensing events at a rate @xmath25 with a dedicated 1 m telescope and a @xmath26 ccd . among the events ,
@xmath1 events have high enough @xmath23 to measure the time scale at the 20% level
. the rate can be increased by up to a factor @xmath27 by employing a few improved observational strategies , and thus @xmath28 with measurable time scale could be detected per year .
these strategies are discussed in 7 .
gould ( 1996 ) developed a general formalism for the analysis of pixel lensing .
the reader is referred to this work for a better understanding of the formalism that i use without detailed derivation .
comparison of the formalism with that developed by colley ( 1995 ) is also recommended .
the signal - to - noise ratio during a single observation is given by @xmath29 ^ 2 , \eqno(2.1)\ ] ] where @xmath6 is the flux of the lensed star , @xmath30 is the surface brightness of observed area , @xmath31 is the angular area of the point spread function ( psf ) , @xmath32 is the observation duty cycle , @xmath33 is the mean photon detection rate averaged over the duty cycle .
if the observations are carried out for 4 hrs every night and the camera can detect @xmath34 , then @xmath35 day and @xmath36 , where @xmath37 is the mean exposure time per day .
individual images obtained during the whole event are combined , yielding the total signal - to - noise ratio @xmath38 ^ 2 . \eqno(2.2)\ ] ] by converting the sum @xmath39 ^ 2 $ ] into an integral @xmath40 ^ 2 $ ] , one can obtain a compact form of @xmath23 , @xmath41 ^ 2 \over \int_{-\infty}^{\infty } dt ( \omega^2 t^2 + \beta^2)^{-1 } } , \eqno(2.3)\ ] ] where the factor @xmath42 is evaluated and shown in figure 1 of gould ( 1996 ) .
events are detected either when the star is bright or the amplification is high .
this can be quantified by setting a maximum impact parameter @xmath43 within which a lensing event with flux @xmath6 is detected for a certain threshold signal - to - noise ratio @xmath44 .
the maximum impact parameter is then found by solving equation ( 2.3 ) , i.e. , @xmath43 is the solution of the equation @xmath45 where @xmath46 is the inverse of median einstein time scale .
once @xmath43 is determined , the event rate per angular area @xmath47 is computed by @xmath48 where @xmath49 is the optical depth , @xmath50 is the luminosity function ( hereafter lf ) of m31 bulge stars , and @xmath51 is a normalization factor , @xmath52^{-1}$ ] .
for the computation of the event rate , it is necessary to compute the optical depth [ see eq . ( 2.5 ) ] .
the computation of optical depth toward m31 requires some geometrical transformations since m31 is highly inclined , @xmath53 ( kent 1989 ) .
i define the intrinsic coordinates , @xmath54 , that are centered at the center of m31 and @xmath55 and @xmath56 are measured along and normal to the disk plane , respectively .
these coordinates are related to another set of coordinates , @xmath57 , which have the same center but are measured along the projected major , minor axis , and the line - of - sight , by , @xmath58^{1/2},\ z = | y ' \sin i - z ' \cos i | .
$ ] the adopted distance to m31 is @xmath59 .
there are several populations of lenses when stars in the m31 bulge are monitored .
the first is machos in the galactic halo .
the mean optical depth to a single m31 star lensed by the galactic halo machos is @xmath60 where @xmath61 . for the galactic halo matter distribution ,
i adopt a modified isothermal sphere with a core radius @xmath62 ; @xmath63 where the normalization is @xmath64 , @xmath65 is the solar galactocentric distance , and the core radius is @xmath66 ( bahcall , schmidt , & soneira 1983 ) .
the value of the galactic core radius is uncertain , but it hardly affects the optical depth determination because m31 is located @xmath67 away from the galactic center . with this model ,
i find @xmath68 .
the second lens population is the machos in the m31 halo .
since the rotation curve of m31 is flat at large radii up to @xmath69 from the center ( kent 1989a ) , the m31 halo is modeled by a standard isothermal model with a core radius @xmath70 ; @xmath71 where the core radius is determined from the relation @xmath72 with the total core mass of @xmath73 ( see below ) and the m31 rotation velocity of @xmath74 , i find that @xmath75 .
then the optical depth contributed by m31 halo lenses is computed similarly to equation ( 3.1 ) , and results in @xmath76 .
finally , the rest of the optical depth toward the m31 bulge is contributed by the m31 disk+bulge self - lensing events .
this contribution of optical depth is computed by @xmath77^{-1 } , \eqno(3.5)\ ] ] where @xmath78 , @xmath79 and @xmath80 are the m31 disk and bulge densities , and @xmath81 and @xmath82 are the lower and upper boundaries of mass distribution along the line of sight . the m31 bulge is modeled as an oblate spheroid with its ( unnormalized ) matter density as a function of semimajor axis provided by kent ( 1989b ) .
the best - fitting value of the axis ratio , @xmath83 , is found by fitting the computed surface brightness ( corrected for @xmath84 extinction ) to a @xmath4-band image of m31 bulge ( see 5 for the observations ) . in the central region where the m31 bulge mass dominates ,
the rotation velocity is @xmath85 .
thus i normalize the total m31 bulge mass within @xmath86 kpc by @xmath87 , where @xmath88 and @xmath89 are the adopted rotation speed of the galaxy and galactic bulge mass within @xmath86 kpc ( zhao , spergel , & rich 1995 ) .
this normalization procedure results in @xmath90 , which is in reasonable agreement with the determined value of @xmath91 by kent ( 1989b ) . the m31 bulge is cut off at @xmath92 .
the m31 disk is modeled by the double exponential disk , i.e. , @xmath93 where @xmath94 is the normalization , and the radial and vertical scale heights are @xmath95 and @xmath96 , respectively ( gilmore et al .
1990 ; gould 1994 ) .
this results in @xmath97 .
the resulting optical depth distribution caused by m31 population ( m31 disk+bulge self - lensing plus m31 halo ) is shown as a contour map in figure 1 . the total optical depth both by m31 and galactic halo lenses are obtained by adding @xmath68 to the values marked on the contour .
the contour levels are marked on the map in units of @xmath98 .
i note that the determined optical depth is subject to many uncertainties , especially in the mass distributions . for example , braun ( 1991 ) has argued that there is no need to invoke a massive dark halo component . instead of a halo ,
his model has higher disk and bulge densities , c.f . ,
@xmath99 and @xmath100 within 30 kpc from the center .
pixel lensing experiments may resolve the conflicts in the matter density distribution in m31 from the comparison between the theoretical predictions and future observational result .
not only the optical depth but also the median time scale is required for the computation of the event rate . the time scale is determined by the combination of both the transverse speed , @xmath101 , and the einstein ring radius of the lens , i.e. , @xmath102 . the transverse velocity is related to the lens geometry and velocity of the source , @xmath103 , the lens , @xmath104 , and observer , @xmath105 by @xmath106 .
\eqno(4.1)\ ] ] for m31-m31 self - lensing events , one can approximate equation ( 4.1 ) as @xmath107 since @xmath108 .
then the rms transverse velocity is found by @xmath109 because the cross - term @xmath110 .
here @xmath111 , and @xmath112 is the adopted velocity dispersion of the m31 bulge ( lawrie 1983 ) .
the einstein ring radius is related to the location of the lens system by @xmath113 , assuming all lenses have the same mass . in the computation ,
i assume @xmath114 because the most probable lens population is low mass stars . for other masses @xmath115 .
the probability of finding an event located at a certain location @xmath116 is found by @xmath117 then the time scale probability distribution is computed from @xmath118 by @xmath119 p(d ) .
\eqno(4.3)\ ] ] i find @xmath120 by taking the median value of @xmath121 .
the resulting median time scale @xmath120 for m31 self - lensing events as a function of position @xmath122 shown in units of days in figure 2 as a contour map . for other population events ,
i.e. , galactic halo and m31 halo events , the time scale distributions are similar to that of m31 self - lensing events ( han & gould 1996b ) .
the light from the m31 bulge suffers some extinction by dust in the disk because half the m31 bulge lies in front of the m31 disk while the other half lies behind it .
hence the bulge light from behind the disk is affected by the dust . by experiencing extinction
, the apparent lf of the stars behind the disk will be different from that of stars in front , and thus @xmath123 is affected by the extinction [ see eq . ( 2.5 ) ] .
since the disk is tilted , the total light of stars influenced by the extinction differs from position to position . to quantify the amount of extinction , ccd images in @xmath4 and @xmath124 bands
were obtained on the night of 1995 august 4 using tek2k detector on 1-m united sates naval observatory ritchey chrtien aplanatic telescope , with a plate scale of @xmath125 .
the @xmath124 band filter is centered at @xmath126 and has a band width of @xmath127 .
bias and sky subtraction of the images were carried out using the interactive data reduction package iraf .
i find that the disk of m31 is quite transparent .
let @xmath128 and @xmath129 be the surface brightnesses contributed by stars located behind and in front of the disk , respectively .
then the flux ratio between two bands is given by @xmath130 where @xmath131 is the fraction of light that passes through the disk in each band .
let @xmath132 be the ( instrumental ) ratio of flux in the two bands in the absence of extinction .
this quantity can be determined by measuring the photon counts along the far minor axis where nearly all of the bulge is in front of the disk and where , consequently , extinction has almost no effect on the color .
then , one can rewrite equation ( 5.1 ) as @xmath133 the fraction @xmath134 is related to @xmath135 by @xmath136 which implies more absorption in @xmath4 band , i.e. , @xmath137 .
the surface brightnesses , @xmath128 and @xmath129 , are computed from the mass model in 3 and the color ratio @xmath138
is obtained from the images of the m31 bulge .
then @xmath139 is determined by solving equation ( 5.2 ) numerically using the relation ( 5.3 ) , i.e. , by finding the solution of the equation , @xmath140 in general , equation ( 5.4 ) has two solutions , e.g. , for @xmath141 both @xmath142 ( totally opaque ) and @xmath143 ( totally transparent ) solutions are possible .
for the actual data , one set of solutions has @xmath144 , while the other set has @xmath145 .
if @xmath146 , the m31 bulge surface brightness distribution would be extremely asymmetric with respect to the photometric major axis because most of the light would be extinguished on the near side while only a small fraction would be lost on the far side due to the high inclination of m31 .
in fact , the isophotes are close to symmetric .
the mean extinction is therefore @xmath147 .
because the disk is optically thin , the overall event rate depends only on the mean extinction .
i therefore adopt the simplified model of uniform extinction with @xmath148 , which corresponds to @xmath149 .
the event rate formula in equation ( 2.5 ) is derived under the assumption that the dominant form of noise comes from source stars , i.e. , the poisson photon noise @xmath150 . however , except in the central region , @xmath151 , most of the m31 bulge region has a surface brightness less than the sky brightness , which is @xmath152 averaged over the phases of the moon .
therefore , the noise due to the sky - background , @xmath153 , becomes important .
in addition , the differences in psf from image to image taken at different times cause extra noise ( gould 1996 ) , which i denote @xmath154 .
i define the additional factors in noise due to sky and psf variation by @xmath155 the values of these factors are computed by @xmath156^{1/2 } , \eqno(6.1.2)\ ] ] where @xmath157 and @xmath158 are the surface brightnesses of the sky and m31 , @xmath159 is the rate of photon detection at the fluctuation magnitude , @xmath160 for m31 ( tonry , 1991 ) , @xmath161 is the fractional difference in seeing between images , and @xmath162 is the number of sub - exposures in each exposure time @xmath37 . by dividing the total exposure time into smaller exposures , the noise due to the time variation of psf decreases because @xmath154 is averaged out , while @xmath150 remains the same . as a result of the additional noise , the total noise , @xmath163 , is increased relative to the pure m31 photon noise by a factor , @xmath164 the net influence of additional noise on the event rate is that the required @xmath44 increases by the factor @xmath165
. however , not all the events that satisfy the strict @xmath44 test qualify as detectable events .
additional tests await for them .
one of these tests comes from the duration of an event : if the event lasts for very short time , the event could not be detected however high the @xmath23 is , because the event continues less than the observation interval ( 1 day ) .
this restriction of time scale is quantified in the computation by counting only events with @xmath166 , where the effective duration time is defined by @xmath167 ( gould 1995 ) .
some of these short events might occur during the observation which continues for @xmath168 each night in average .
however , they would not be detected because of another restriction discussed below .
the other restriction comes from the finite size of the source star .
if the source size is bigger than the einstein ring , @xmath169 , the point - source amplification formula in equation ( 1.2 ) is no longer valid at the central region of the light curve , i.e. , when @xmath170 where @xmath171 is the crossing time over the star with a radius @xmath172 .
then the maximum amplification will be overestimated by equation ( 1.2 ) and the event with @xmath173 can not be detected ( gould 1995 ) .
the luminosity at a certain frequency @xmath174 is @xmath175 , where @xmath176 $ ] is the planck function . in the rayleigh - jeans limit ,
e.g. , @xmath177 band @xmath178 , the planck function is approximated by @xmath179 , and thus the luminosity in the @xmath177 band is related to the stellar radius by @xmath180 . here
the frequency is regarded constant over a narrow band .
since the temperature of relevant source star varies within a narrow range , @xmath181 , relative to the luminosity , @xmath182 , the dominant factor in determining stellar radius is the luminosity .
therefore , one can approximate the stellar radius by @xmath183 , where the solar @xmath177-band luminosity is @xmath184 and @xmath177-band lf are provided from the model lf discussed in 6.2 .
i begin by analyzing the event rate in a relatively modest observing program .
the experiment is carried out using a @xmath185 ccd camera with a pixel scale of @xmath186 on a 1 m telescope .
if the pixel scale is too big , the geometric alignment of the images would introduce considerable additional noise ( pixelization error ; gould 1996 ) .
i assume that the camera can detect @xmath187 for an @xmath188 star .
the observation is carried out @xmath189 during the m31 season ( @xmath190 ) .
bad weather will influence the event rate , e.g. , by degrading the @xmath23 and by missing portions of the light curve .
however , the weather pattern is unpredictable and differs dramatically from site to site .
hence , i set the observation time less than what might be being carried out ( @xmath191 ) to compensate the loss of @xmath123 by the bad weather instead of modeling the weather pattern . the total exposure time is subdivided into smaller integration times , @xmath192 , so that the images are not saturated
the angular size of the psf is @xmath193 , assuming typical seeing of @xmath194 .
the fractional seeing difference is kept to @xmath195 by using methods discussed in 7 .
the surface brightness distribution is based on the density models of the m31 bulge and disk and then normalized to the values of walterbos et al .
the lf is modeled primarily by adopting that of galactic bulge stars shown by dotted line in the second panel of figure 3 ( j. frogel , private communication ) however , the lf in the faint end , @xmath196 , is incomplete .
here @xmath197 corresponds to @xmath198 where the lf begins to decline .
if the lf is adopted without corrections , the fraction of bright stars would be overestimated .
the net effect is then the systematic overestimation of event rate because events are more likely to be detected for brighter stars .
thus , i model the faint part of the lf ( @xmath196 ) by adopting the lf of stars in the solar neighborhood determined by wielen , jahreiss , & krger ( 1983 ) for @xmath199 and by gould , bahcall , & flynn ( 1996 ) for @xmath200 . the corrected model lf yields the fluctuation magnitude of @xmath201 , which matches well with tonry s ( 1991 ) determination of @xmath202 before the extinction correction .
the fraction of light from the stars in the corrected part of the lf ( @xmath203 ) is @xmath204 .
the model lf is shown with a solid line in the second panel of figure 3 .
combining all this information , the values of @xmath43 are computed for a given stellar flux @xmath6 by solving equation ( 2.4 ) and taking the additional noise term @xmath165 into consideration .
the determined @xmath43 as a function of the stellar flux is shown in the upper panel of figure 3 for the position @xmath205 at which @xmath206 , @xmath207 , and @xmath208 for the various @xmath44 noted in the figure .
the lf is normalized by the surface brightness of the position .
note that as a star becomes fainter , it can only be detected at higher amplification and that the detection limit decreases as @xmath44 increases .
now one is ready to compute event rate for any adopted @xmath44 with the results of the @xmath49 and @xmath209 distributions determined in 3 and 4 .
the term `` event rate '' is somewhat subjective because its value depends on the required precision and strategy of the experiment . to measure the time scale with better precision , higher @xmath23
is required .
the relation between the precision of measurement and the required @xmath23 is quantified by a factor @xmath210 , where @xmath211 is the fractional accuracy of the time scale measurement .
its value is computed as a function of @xmath12 and shown in figure 1 of gould ( 1996 ) . to determine @xmath44 , i use the mean value , @xmath212 , since @xmath213 is a slowly varying function of @xmath12 except in the region of very close encounter event ( i.e. , @xmath214 ) ; @xmath215 in the range @xmath216 .
the required minimum signal - to - noise ratio is then , @xmath217 first i estimate the event detection rate based on the precision that can distinguish lensing events from flux variation due to statistical noise . in principle
it is possible to select events up to the poisson limit of @xmath218 fluctuations , i.e. , @xmath219 ( gould 1996 ) . to be conservative ,
i use @xmath220 .
the resultant event rate map @xmath221 , @xmath222 , at this @xmath44 is constructed and shown in figure 4 as a contour map .
the contour levels are drawn in units of @xmath223 .
then the total event rate detected over the entire area of the image , @xmath224 , is computed by @xmath225 resulting in @xmath226 .
now i stiffen the qualification of events by counting only those that can provide time scales with an acceptable uncertainty , @xmath227 .
the required signal - to - noise ratio is then @xmath228 determined from equation ( 6.3.1 ) .
the result is @xmath229 .
the total event rate for other values of @xmath44 are shown in figure 5 for individual lens population events .
there are several ways of improving the event rate .
first , one can increase @xmath123 by reducing the seeing difference , @xmath230 , which causes additional noise @xmath231 .
the easiest way to reduce @xmath161 is classifying images ( there will be @xmath232 of them ) with similar seeing , and then subtracting the current image from reference images of similar seeing .
the other method , which is being developed by tomaney & crotts ( 1996 ) , is reducing the seeing differences by convolving the current image so that the resultant psf matches the reference frame .
second , the event rate would increase by carrying out the experiment at more than one place . by observing more often , one could detect events with shorter @xmath233 than could not be detected from a single site . the other effect of multi - observation on @xmath123
is the increase of the time averaged photon detection rate , @xmath33 .
finally , a bigger telescope will detect more events .
the event rates with the additionally adopted observational strategies ( single or combined ) are shown in table 1 .
the values in the second and third columns represent @xmath123 when @xmath234 and 80 , respectively .
the values in the parentheses represent the improved factors in @xmath123 compared to the value obtained under the observational condition described in 6.2 .
if all of these strategies are adopted , the event rate would increase by up to a factor @xmath235 , so one could obtain events with time scale data at a rate @xmath236 . therefore , a few years of observation would provide one with a data set large enough for the analysis of m31 stars and machos . from this analysis , one can obtain important information such as the lf of m31 bulge stars ( gould 1996 ) , the optical depth distribution of machos , and above all the mass spectrum of m31 machos ( han & gould 1996a ) .
the determined event rate is subject to many sources of uncertainties .
these include insufficiently detailed density and velocity distributions and mass spectrum of m31 bulge machos .
additional uncertainties from the crude models of extinction and the lf also propagate to the determination of the event rate . among all the uncertainties ,
however , the most important comes from the stellar mass density and the fraction of dark matter in various forms , which can only be determined by the comparison of the theoretical predictions with the observational results .
the other uncertainties will not seriously affect the estimates given here .
for example , the velocity field is spectroscopically well constrained .
although the model lf at the faint end is very uncertain , stars in this regime make little contribution to the event rate because they are generally too faint to be detected .
thus , the detailed structure of the faint end of the lf is not important . of course , my determination of the amount of extinction , @xmath139 , is not precise enough for other type of studies such as the construction of an extinction map which would require very detailed observation .
however , in the optically thin limit , the event rate depends only on the mean extinction ( and even for a totally opaque disk only falls by a factor @xmath237 ) , so the error induced by uncertain extinction is unlikely to exceed 10% .
in addition , it has been demonstrated by tomaney & crotts ( 1996 ) that systematic errors can be brought under very good control .
for example , they measure the flux variation of a variable found in the central m31 bulge ( surface brightness @xmath238 ) with a @xmath239 flux error equivalent to @xmath240 , within a factor @xmath241 of the photon limit and 8 mag below the flux in a seeing disk .
i would like to thank a. gould and j. frogel for making very helpful comments and kindly providing the galactic bulge luminosity function and to usno for providing telescope time .
this work was supported by a grant ast 94 - 20746 from the nsf .
alard , c. , guibert , j. , bienayme , o. , valls - gabaud , d. , robin , a. c. , terzan , a. , & bertin , e. 1995 , msngr , 80 , 31 alcock , c. et al .
1995 , phrvl , 74 , 2867 ansari , r. et al . 1995 , a&a , 299 , l21 ansari , r. et al . 1996 , 2nd workshop of `` the dark side of the universe : experimental efforts and theoretical frame work '' bahcall , j. a. , schmidt , m. , & soneira , r. m. 1983 , apj , 265 , 730 baillon , p. , bouquet , a. , giraud - hraud , y. , & kaplan , j. 1993 , a&a , 277 , 1 braun , b. 1991 , apj , 372 , 54 colley , w. n. 1995 , aj , 109 , 440 crotts , a. p. s. 1992 , apj , 399 , l43 gilmore , g. , king , i. r. , & van der kruit , p. c. 1990 , the milky way as a galaxy ( university science book , mill valley ) , 35 gould , a. 1994 , apj , 435 , 573 gould , a. 1995 , apj , 455 , 44 gould , a. 1996 , apj , 470 , 000 gould , a. , bahcall , j. a. , & flynn , c. 1996 , apj , 465 , 000 han , c. , & gould , a. 1996 , apj , 467 , 000 han , c. , & gould , a. 1996b , apj , submitted kent , s. 1989a , pasp , 101 , 489 kent , s. 1989b , aj , 97 , 1614 lawrie , d. g. 1983 , apj , 273 , 562 tomaney , a. b. , & crotts , a. p. s. 1996 , apj , submitted tonry , j. l. 1991 , apj , 373 , l1 udalski , a. et al .
1995 , aca , 45 , 237 walterbos , r. a. m. & kennicutt , jr . r. c. 1987 , a&as , 69 , 311 wielen , r. , jahreiss , h. , & krger , r. 1983 , iau coll .
76 : nearby stars and the stellar luminosity function , eds . a. g. d. philip & a. r. upgren , p.163 zhao , h. , spergel , d. n. , & rich , r. m. 1995 , apj , 440 , l13 .
the event rates with improved observational strategies in units of @xmath223 .
the values in the parentheses represent the improved factors compared to that under observational condition described in 6.2 .
the first column shows the additionally adopted strategies , and the values in the second and third columns are those of @xmath123 when @xmath242 and 80 , respectively . [ cols="<,^,^ " , ] * figure 1 * : the resulting optical depth distribution caused by m31 population ( m31 disk+bulge self - lensing plus m31 halo ) is shown as a function of projected position @xmath122 . the total optical depth both by m31 and galactic halo lenses are obtained by adding @xmath68 to the values marked on the contour .
the contour levels are marked on the map in units of @xmath98 .
the lower half of the figure contains the near side of the m31 disk .
* figure 2 * : median time scale distribution as a function of @xmath122 for m31 self - lensing events .
the contour levels are drawn in units of @xmath243 .
other population events , galactic halo and m31 halo events , the time scale distributions are similar to that of m31 self - lensing events .
* figure 3 * : values of @xmath43 as a function of luminosity of m31 bulge stars ( upper panel ) and the adopted lf of m31 bulge stars ( lower panel ) . for the computation of @xmath43 ,
values of @xmath30 and @xmath244 at @xmath245 are used . the basic lf function ( dotted line )
is modeled based on the galactic bulge lf ( j. frogel , private communication ) .
the faint end of the lf ( solid line ) is modeled by combining the lfs of gould et al .
( 1996 ) and wielen et al . | i demonstrate that one can detect pixel gravitational microlensing events at the rate @xmath0 and that some fraction of events ( @xmath1 ) will be good enough to measure time scales at the @xmath2 level if the experiment is carried out toward the central m31 bulge region with a dedicated 1 m telescope and a @xmath3 ccd .
i also show that the event rate could be significantly increased ( up to a factor 6.5 ) if more aggressive observation strategies are adopted .
these aggressive strategies include multiple observation sites , using a bigger telescope , and reduction of seeing differences among images .
a few years of observation could provide a data set big enough to extract important information about the m31 bulge such as the optical depth distribution , the mass spectrum of m31 machos , and the luminosity function , which currently can not be obtained by other methods .
additionally , i find that the extinction of m31 bulge light by dust in the disk is _ not _ important and can be approximated by the model in which only @xmath2 of light in @xmath4 band is extinguished over most of the bulge area .
# 1#2#1 10^#2 = + 0.2 in = 6.0 in = 8.0 in = 0.10 in = 0.5 in = 0.5 in # 1 = 0.4 in = 1 # 1 # 1 ccc 1 & 1 & 1 + 1 & 1 & 1 + 4 & 3 & 8 |
while ferromagnetism is known since antiquity , antiferromagnetism was only discovered in this century .
it is not widely known that the concept of antiferromagnetic order was proposed independently by nel@xcite in 1932 and landau@xcite in 1933 . both
sought to explain the , at that time , puzzling low temperature behavior of the magnetic susceptibility of certain materials : of metals such as cr and mn in the case of nel , and of insulators with layered structure such as the chlorides of cr , bivalent fe , co and ni in the case of landau . while nel correctly suggested the existence of interpenetrating sublattices in cr and mn with opposite magnetization @xcite , landau
equally correctly predicted the existence of stacks of ferromagnetically ordered layers whose magnetization alternates from layer to layer @xcite . in both cases
the total spontaneous magnetization adds up to zero . assuming the interlayer coupling to be weak
, landau @xcite argued that a relatively small magnetic field would be sufficient to modify the mutual orientation of the moments in each layer .
this leads to deviations from the linear dependence of the total moment on the field , i.e. to an anomalous increase of the susceptibility , and finally at high fields to a saturation of the magnetization .
such a behavior was indeed observed by becquerel and van den handel in 1939 @xcite in the mineral mesitite ( carbonate of fe and mg ) at low temperatures .
not being aware of landau s or nel s work they could not explain their observation in terms of ferro- and paramagnetism , and therefore suggested for it the name metamagnetism @xcite .
qualitatively similar , but even more drastic magnetization effects were later observed in many other systems of which fecl@xmath3 and dy@xmath4al@xmath5o@xmath6 ( dag ) are well - studied prototypes @xcite .
these materials are insulators where the valence electrons are localized at the fe and dy ions , respectively .
the resulting local moments order antiferromagnetically and are constrained to lie along an easy axis @xmath7 , implying a strong anisotropy such that a spin
flop transition can not occur .
under the influence of a large magnetic field @xmath8 the staggered magnetization vanishes in a first or second order phase transition , the so called metamagnetic transition .
apart from the thoroughly investigated materials mentioned above , there are also conducting systems in that class , e.g. uranium based mixed systems @xcite , smmn@xmath3ge@xmath3 @xcite , and tbrh@xmath9ir@xmath10si@xmath3 @xcite .
today the term `` metamagnetic transition '' is used in a much wider sense @xcite , namely whenever the homogeneous susceptibility @xmath11 has a maximum at some value @xmath12 , with @xmath13 being strongly enhanced for @xmath14 .
metamagnetism is then found to be a rather common phenomenon which occurs also in spin
flop antiferromagnets ( e.g. the parent compound of high - t@xmath15 superconductivity , la@xmath3cuo@xmath16 @xcite ) , strongly exchange - enhanced paramagnets ( e.g. tibe@xmath3 , yco@xmath3 @xcite ) , heavy fermion and intermediate valence systems ( e.g. ceru@xmath3si@xmath3 , upt@xmath4 @xcite ) . in this paper we will be concerned only with metamagnetism in strongly anisotropic antiferromagnets , several of which are known to have a very interesting @xmath17 phase diagram ( @xmath18 : _ internal _ magnetic field , @xmath19 : temperature ) . in particular ,
in the insulating systems one often finds a tricritical point at which the first order phase transition becomes second order .
theoretical investigations of tricritical points began with the work of landau , who described multicritical behavior within his phenomenological theory of phase transitions@xcite .
clearly , a genuine understanding of the origin of tricritical points etc . requires investigations on a more microscopic level . in the case of
strongly anisotropic metamagnets those investigations where sofar restricted to the insulating systems , such as febr@xmath3 .
they are usually based on the ising model with more than one interaction , in a magnetic field , on a simple cubic lattice @xcite , e.g. @xmath20 where the summations extend over the nearest neighbors ( nn ) and next nearest neighbors ( nnn ) of every site . for @xmath21
one has an antiferromagnetic ( af ) coupling between the @xmath22 nearest neighbor spins and a ferromagnetic coupling between the @xmath23 next nearest neighbors @xcite .
it was pointed out by kincaid and cohen @xcite that in mean field theory a tricritical point @xcite as in fig .
1a exists only for @xmath24 , while for @xmath25 this point separates into a critical endpoint ( ce ) and a bicritical endpoint ( bce ) ( see fig .
1b ) . the latter behavior , especially the finite angle between the two transition lines at ce and the pronounced maximum at the second order line , is qualitatively very similar to that observed in febr@xmath3 @xcite .
however , the first order line between ce and bce has so far not been observed neither experimentally , nor even theoretically when evaluating ( [ gl1 ] ) beyond mean field theory @xcite .
most recently , by measurement of the excess magnetization and anomalous susceptibility loss @xcite , the specific heat @xcite , and by neutron scattering @xcite a strip shaped regime of strong _ non critical _ fluctuations in the @xmath26 plane ( where @xmath27 is the applied field ) of febr@xmath3 was reported which is at least reminiscent of the _ critical _ line ce @xmath28 bce .
this unusual behavior was then also found theoretically by selke @xcite who evaluated ( [ gl1 ] ) and also a more detailed model @xcite for weak ferromagnetic coupling ( @xmath29 ) and large @xmath22 using monte carlo techniques .
there does not yet exist a microscopic theory for the conducting systems such as the uranium based mixed systems @xcite , because this requires a fully quantum mechanical treatment of _ itinerant , _ correlated electrons .
first steps in this direction , where genuine correlation effects were , however , neglected , are the semi phenomenological theories of wohlfarth and rhodes @xcite for metamagnetic phase transitions in paramagnets , and of moriya and usami @xcite for the coexistence of ferro- and antiferromagnetism in itinerant electron systems .
it is the purpose of this paper to examine the origin of metamagnetism in strongly anisotropic antiferromagnets from a microscopic , quantum mechanical point of view . to this end
we will study the hubbard model in the presence of a strong anisotropy direction , with a magnetic field along this direction , since it is the simplest microscopic model that describes insulating _ and _ metallic , spin
localized _ and _ bandlike antiferromagnets with easy axis in an external magnetic field .
_ _ _ _ _ _ in sec .
[ easy ] the underlying model , the hubbard model with easy axis , is introduced and its validity is discussed .
then the dynamical mean field theory which is used to investigate the correlation problem all the way from weak to strong coupling , as well as the quantum monte carlo techniques employed to solve the coupled self
consistency equations , are discussed in sec .
the results obtained for a half filled band and for finite doping are presented in sec .
[ reshf ] and sec .
[ resbhf ] , respectively .
a discussion of the results in sec .
[ dis ] closes the presentation .
the hubbard model @xcite is the generic microscopic model for itinerant and localized antiferromagnetism in correlated electron systems . for nearest neighbor hopping of electrons in the presence of a magnetic field
it has the form @xmath30 where operators carry a hat . from the exact
, analytic solution in dimensions @xmath31 the ( paramagnetic ) ground state of this model at half filling @xmath32 is known to exhibit metamagnetic behavior , i.e. @xmath33 , up to saturation @xcite . however , this is entirely due to the convex shape of the density of states in @xmath31 and occurs even at @xmath34 . in general
the hubbard model in the form of ( [ hub ] ) can not describe metamagnetic behavior .
the hubbard model is isotropic in spin space and hence the antiferromagnetic phase described by it is isotropic , too .
consequently , any finite magnetic field @xmath35 will orient the staggered magnetization @xmath0 _ perpendicular _ to itself .
real antiferromagnets are , however , never isotropic in spin space : the relativistic spin
orbit interaction @xmath36 , where @xmath37 and @xmath38 are operators for the orbital angular momentum and spin , respectively , transfers the anisotropy of _ position _ space ( caused by the broken rotational symmetry of the crystal lattice ) into _ spin _ space , producing one , or more , easy axes which constrain the spins @xcite . in an external magnetic field
such a constraint leads to metamagnetic transitions , with or without a spin
flop depending on the strength of the spin orbit interaction , as explained in sec .
[ intro ] .
a microscopic theory of strongly anisotropic antiferromagnets should ultimately be able to take into account the orbital degeneracy of the electrons and , by including the relativistic spin
orbit interaction @xmath39 in the hamiltonian , to generate an anisotropy axis within the model itself .
at present , this is technically not possible @xcite .
therefore we take the existence of the anisotropy axis @xmath7 for granted : we employ the hubbard model ( [ hub ] ) and constrain the magnetic moments to lie along @xmath40 . by this approach
the kinetic energy and the coulomb interaction are treated microscopically , whereas the relativistic corrections are not .
we note that the relativistic corrections are of the order of @xmath41 ev and are thus small compared to energies of the order of @xmath42 ev for kinetic and coulomb energy .
therefore the existence of @xmath7 and the correlation physics described by the hubbard model ( [ hub ] ) are quite unrelated .
this justifies our approach where the existence of @xmath7 is _ a priori _ assumed @xcite .
for classical spin models ( e.g. the ising model ) it is well known that the weiss molecular field theory becomes exact in the limit of high spatial dimensions ( @xmath44 ) . for lattice electrons
this limit was introduced only recently @xcite . with the proper scaling of the hopping element in ( [ gl2 ] ) , @xmath45 ( @xmath22 = number of nearest neighbors )
, it leads to a quantum mechanical dynamical mean field theory ( dmft ) ; for reviews see @xcite .
the interacting lattice model then reduces to a self consistent single site problem of electrons in an effective medium @xcite , which may be described by a complex , frequency dependent ( i.e. dynamical ) self energy @xmath46 .
this problem is , in fact , equivalent to an anderson impurity model complemented by a self consistency condition @xcite .
there are two limits in which the dmft recovers well - known _ static _
mean field theories : 1 .
_ weak coupling : _ in this situation the effective medium may be approximated by a _ static _
field which is generated by the _ averaged _ densities of the electrons .
this leads to the hartree
fock approximation , e. g. @xmath47 in the homogeneous case , which is expected to give the qualitatively correct behavior at weak coupling .
the averaged densities @xmath48 have to be determined self consistently .
_ _ _ _ 2 .
_ strong coupling at half filling : _ here model ( [ gl2 ] ) can be mapped onto the antiferromagnetic spin 1/2 heisenberg model for which the limit @xmath44 becomes equivalent to the weiss molecular field theory .
the results obtained in these two limits will be presented , and compared to the results for intermediate coupling , in sec .
[ reshf ] .
the dmft has recently provided valuable insight into the physics of strongly correlated electron systems , e.g. the mott
hubbard transition @xcite and transport properties @xcite .
the effect of the magnetic field @xmath18 in ( [ gl2 ] ) for @xmath49 was also studied @xcite . in particular , laloux et al .
@xcite thoroughly investigated the magnetization behavior of the paramagnetic phase , assuming the af order to be suppressed . for @xmath50
they find a first order metamagnetic transition between the strongly correlated metal and the mott insulator at a critical field @xmath51 .
giesekus and brandt @xcite took into account the af order .
they considered the isotropic case where the field orients the staggered magnetization perpendicular to itself , such that a metamagnetic transition can not occur . to investigate the metamagnetic phase transition from an antiferromagnet to a paramagnet we consider a bipartite ( @xmath52-@xmath53 ) lattice and
allow for symmetry breaking with respect to spin @xmath54 and sublattice @xmath55 . the self energy @xmath56 , with matsubara frequencies @xmath57 , @xmath58 ( using the convention @xmath59 ) , and the green function @xmath60 are determined self consistently by two sets of coupled equations @xcite :
@xmath61 here @xmath62 , and the thermal average of some operator @xmath63 $ ] in ( [ gl4 ] ) is defined as a functional integral over the grassmann variables @xmath64 , with @xmath65 { \cal d } [ \psi^ * ] { \cal o [ \psi , \psi^ * ] } e^{a_{\alpha } [ \psi , \psi^*,\sigma , g ] } , \ ] ] in terms of the single site action
@xmath66 \psi_{\sigma n } \nonumber \\ & & - u \int_0^\beta d \tau \psi_\uparrow^ * ( \tau ) \psi_\uparrow ( \tau ) \psi_\downarrow^ * ( \tau ) \psi_\downarrow ( \tau ) \ ; , \label{gl5}\end{aligned}\ ] ] where @xmath67 is the partition function , and @xmath68 is the density of states ( dos ) of the non interacting electrons .
as the results do not much depend on its precise form we choose a half - elliptic dos @xmath69^{1/2}/(2 \pi t^{*2})$ ] .
the constraint @xmath70 is enforced by setting the off - diagonal ( in spin space ) elements of the green function equal to zero : @xmath71 . from now on @xmath72 will set our energy scale , i.e. the total band width of @xmath68 is equal to 4 .
the dyson equation ( [ dys ] ) introduces the lattice into the problem .
it couples @xmath73 and @xmath74sublattices and can be solved by a simple integration , even analytically for the above dos .
the functional integral ( [ fi ] ) however is highly non
trivial since it couples all matsubara frequencies .
georges and kotliar @xcite and jarrell @xcite realized that the action ( [ gl5 ] ) is equivalent to that of an anderson impurity model , and can therefore be treated by standard techniques developed for this model . here
we employ a finite temperature , auxiliary field quantum monte carlo ( qmc ) method @xcite . in this approach
the electron
electron interaction is formally replaced by an interaction of independent electrons with a dynamical , auxiliary field of ising type spins . to this end
the interval @xmath75 $ ] is discretized into @xmath76 steps of size @xmath77 .
equivalently , there is a high energy cut off of matsubara frequencies , i.e. @xmath78 , @xmath79 .
all quantities have to be extrapolated to @xmath80 .
the computer time grows like @xmath81 , restricting @xmath76 to values below @xmath82 and @xmath83 on present supercomputers . for small @xmath76 @xmath84 one can perform a full enumeration ( instead of the monte carlo sampling ) of all @xmath85 possible configurations of the auxiliary field .
we never encountered a minus sign problem , hence no further approximations ( like the `` fix node '' method ) were necessary .
consistency is obtained iteratively as follows : the green function @xmath86 ( omitting indices ) is calculated from some initial self energy , e.g. @xmath87 , by the dyson equation ( [ dys ] ) .
now the new green function @xmath88 is determined by solving ( [ fi ] ) with the qmc method .
finally , the calculation of the new self energy @xmath89 completes one iteration . to improve convergence in the symmetry broken case @xmath90 and
@xmath91 are updated by the dyson equation ( [ dys ] ) after every qmc simulation for one sublattice ( eq .
[ gl4 ] ) . in the symmetry
broken phases , typically @xmath92 iterations with 20000 mc sweeps are necessary to obtain a convergence of @xmath93 .
the calculation of a magnetization curve at @xmath94 takes about 100 hours on a cray - y - mp .
close to a phase transition the convergence is much slower and the statistical errors are larger due to strong fluctuations , in particular in the case of a second order phase transition .
these effects limit the accuracy in the determination of the critical values of the model parameters , e.g. the critical magnetic field ( see sec .
[ ic ] ) . at large @xmath95-values
@xmath96 the monte carlo sampling becomes more and more inefficient due to `` sticking '' problems , i.e. there are two ( or more ) minima in the free energy and the single spin
flip algorithm is no longer able to transfer between them . from the resulting green functions we calculate the densities and the homogeneous and antiferromagnetic magnetization : @xmath97
since we are interested in the question whether the system is insulating or metallic we also determine the electronic compressibility @xmath98 .
it was calculated both by numerical differentiation of @xmath99 and from the two particle correlation functions ( see @xcite and appendix [ sus ] ) .
both results agree within the statistical errors : however , the latter method is much more time consuming .
[ reshf ] for weak coupling ( @xmath100 ) we expect the hartree fock approximation to give an , at least qualitatively@xcite , correct picture of the metamagnetic phase transitions , especially of their order . within this
mean field theory electronic correlations are neglected and the interaction is decoupled as @xmath101 note , that while for @xmath102 the fock term vanishes , @xmath103 , its presence is essential in the case @xmath104 ( see appendix [ perpo ] ) . we confine our investigations to states with homogeneous sublattice magnetization as described by the ansatz @xmath105 applying this hartree
fock decoupling scheme one obtains the effective , one particle hamiltonian @xmath106 .
\label{hhe}\end{aligned}\ ] ] this hamiltonian is diagonalized and the one particle energies @xmath107 are calculated as @xmath108 where @xmath109 denotes the sign of the non interacting electron energy @xmath110 . in the antiferromagnetic phase the dos has a gap of width @xmath111 with square root singularities at its edge . from the one
particle energies ( [ hfe ] ) the grand potential per lattice site ( @xmath112 being the number of lattice sites ) is obtained directly as ( @xmath113 ) @xmath114 the potential @xmath115 has two shallow minima , one corresponding to the paramagnetic state ( @xmath116 ) and the other to the antiferromagnetic state ( @xmath117 ) . by applying a sufficiently strong magnetic field to the antiferromagnetic state , the paramagnetic
minimum becomes the lowest , such that a first order phase transition takes place .
the reason for the occurrence of a first order phase transition is that , within the ( static ) hartree fock approximation , the @xmath118term becomes minimal for largest ( static ) local moments .
therefore pure antiferromagnetic and ferromagnetic order are both energetically favored .
mixed states with @xmath119 and @xmath120 which would occur in the case of second order phase transitions have a small local moment , i.e. a high hartree fock energy , on every second site . to calculate the magnetization curves the self consistent hartree fock equations are obtained from the minimization conditions @xmath121 and @xmath122 : @xmath123 these equations are solved numerically by iteration and integration according to newton cotes rules . within the hartree
fock approximation the metamagnetic phase transition is found to be of _ first _ order for all @xmath19 , even for @xmath124 ( = band width ) ; see fig .
[ hfm ] and [ hfp ] .
hence a tricritical point never occurs . in
this parameter range the quantum monte carlo calculations , however , already show second order transitions in a broad range of temperatures ( see sec .
[ ic ] ) . hence the hartree
fock solution can neither describe the experimental situation , where tricritical points are known to occur , nor the correct behavior of the model for intermediate values of @xmath95 . to estimate the anisotropy energy associated with the easy axis we compare the hartree
fock energies of the configurations with @xmath125 and @xmath126 ( for details see appendix [ perpo ] ) . at half filling and for @xmath95 equal to the band width , the difference between the free energy of these configurations does not exceed a few percent of the band width , i.e. @xmath127 .
in this situation the spin orbit interaction , which can be relatively strong , @xmath128 , indeed leads to a strong anisotropy , i.e. an easy axis @xmath129 , along which @xmath130 is rigidly fixed .
= 8.6 cm = 8.6 cm metamagnetic phase transitions in itinerant , metallic systems where hitherto described by the theory of `` itinerant electron metamagnetism '' ( iem ) . in the case of an antiferromagnetic system in a magnetic field moriya and usami@xcite proposed a landau theory with free energy @xmath131 where @xmath132 and @xmath133 are the homogeneous and staggered susceptibility , respectively , and the coefficients @xmath134 , @xmath135 , @xmath136 , and @xmath137 are the fourth order derivatives of the non - interacting free energy . within the iem theory
the coulomb interaction is treated in random phase approximation .
the corresponding susceptibilities are given as @xmath138 where @xmath139 and @xmath140 are the respective susceptibilities of the non - interacting system .
the random phase approximation for these susceptibilities is equivalent to the hartree
fock scheme described above
. therefore we may ask whether we obtain the iem in the limit @xmath141 ( where @xmath142 ) .
the answer is not straightforward since the prefactors in the expansion ( [ iemlandau ] ) depend , for example , on the lattice structure . on bipartite lattices , as discussed here , they diverge for @xmath143 .
thus an expansion of the free energy in powers of @xmath144 and @xmath145 as assumed in the iem landau theory is not possible in general ( for details see appendix [ icmhf ] ) .
= 8.6 cm in the limit @xmath146 the hubbard model at half filling ( @xmath49 ) is equivalent to an effective heisenberg spin
model @xmath147 where the antiferromagnetic exchange coupling is obtained in second order perturbation theory as @xmath148 .
spin operators are defined as @xmath149 , @xmath150 , and @xmath151 . for this model
the weiss molecular field theory becomes exact in @xmath152 yielding , under the constraint of uniaxial magnetization , the same results as for the ising model . for ising models
metamagnetic phase transitions are well - studied @xcite . in the case of a purely antiferromagnetic next - neighbor coupling ( see ( [ isiaf ] ) )
the phase transitions are of first order only at @xmath153 , but of _ second _ order at all @xmath154 .
the transition line in the @xmath17 phase diagram has indeed the form shown in fig .
1a , but with a tricritical temperature of @xmath155 .
this behavior can be understood already within weiss molecular field theory , where the ground state energy per site is @xmath156 with @xmath157 minimization with respect to @xmath145 and @xmath144 shows that the fully polarized antiferromagnet ( @xmath158 ) has lowest energy for @xmath159 , whereas the fully polarized ferromagnet ( @xmath160 ) is energetically favored for @xmath161 .
thus , by applying a magnetic field a first order transition is induced . at @xmath162 the states
are highly degenerated since all magnetic phases with @xmath163 have the same energy . for @xmath154
this degeneracy is lifted by entropy which disfavors fully polarized phases .
therefore the first order transition at @xmath153 immediately becomes second order for @xmath154 , i.e. @xmath164 .
indeed , a tricritical point at a _
finite _ temperature is only obtained in the case of spin interactions which simultaneously favor both fully polarized antiferromagnetic and ferromagnetic configuration . in particular , adding a ferromagnetic interaction @xmath165 between next nearest neighbors ( nnn ) on a simple cubic lattice as in ( [ isi ] ) stabilizes both ferro- and antiferromagnetic order @xcite . while in the case of effective spin models a ferromagnetic nnn coupling term
is introduced _
ad hoc _ , simply to obtain the first order phase transition , this term naturally arises if we expand the strong coupling perturbation series of the hubbard model to @xmath166 .
however , besides this @xmath165-term there also appear additional _ four_spin terms . for the hypercubic lattice the effective hamiltonian @xmath167 reads@xcite @xmath168 here @xmath169 and @xmath170 are lattice vectors connecting a site to its @xmath22 neighbors , and @xmath171 represents a plaquette .
each plaquette is counted only once : the four sites @xmath172 represent its four corners in clockwise or anticlockwise order .
the constants @xmath52 , @xmath53 and @xmath173 depend on the lattice , and the hermitian operators @xmath174 are defined as @xmath175 the plaquette contribution competes with the ferromagnetic nnn term ( @xmath165 ) and drives the system to second order phase transitions . for the hypercubic lattice
the plaquette contribution is stronger than the ferromagnetic nnn term , yielding second order phase transitions even for @xmath153 .
the same is true for the bethe lattice where , in fact , @xmath176 ( for details see appendix [ asc ] ) . thus in strong coupling perturbation theory the metamagnetic phase transition is of _ second order even at _ @xmath153 .
the perturbation analysis described above demonstrates that the order of the metamagnetic phase transition depends on the coulomb interaction @xmath95 in a delicate way . for small @xmath95
the phase transition is purely of first order and for large @xmath95 of second order .
apparently , the tri- or multicritical point linking these two regimes must be found at intermediate coupling . in this
important , non perturbative regime quantum monte carlo techniques are employed to solve the problem numerically without any further approximation .
the results for the magnetization @xmath13 and the staggered magnetization @xmath177 are shown in fig .
[ mvsh ] for @xmath178 . below the nel temperature a metamagnetic behavior
is clearly seen : for small magnetic fields the magnetization is exponentially suppressed with temperature .
then , towards the metamagnetic phase transition , the susceptibility increases drastically and becomes maximal at the critical field @xmath12 .
second order phase transitions are observed for @xmath179 , whereas the transition is of first order at lower temperatures , i.e. @xmath180 . at the phase transition the order parameter , i.e. the staggered magnetization , vanishes . from the curve
@xmath177 the critical field @xmath12 and also the order of the phase transition is determined by a square root fit for second order transitions and by the mean of the hysteresis for first order transitions . using these values of @xmath12 the phase diagrams , fig .
[ pd ] , for different values of @xmath95 are constructed .
the case @xmath124 ( = band width ) and half filling , fig . [ pd]a , was already presented in ref .
this phase diagram shows both first order ( for @xmath181 ) _ and _ second order phase transitions ( for @xmath182 ) .
the field dependence at intermediate temperatures , i.e. @xmath183 , is more complex : @xmath144 is almost field - independent in region af@xmath184 ( where @xmath185 , decreases sharply at the boundary to region af@xmath186 ( where @xmath187 ) and , upon further increase of the field , vanishes in a second order transition .
although the error bars do not permit an unambiguous interpretation it hence seems that the order parameter decreases by _ two _ consecutive transitions : the first one being of first order or corresponding to an anomaly , and the second one being of second order .
taken together the results seem to correspond to the scenario of fig .
[ scheme]b . for @xmath178 phase transitions of first order
are found for @xmath188 and of second order for @xmath189 ( see fig . [
pd]b ) . here
, the temperature regime with two consecutive transitions , as obtained for @xmath124 , has disappeared or has become very small : the scenario is similar to fig .
[ scheme]a .
the phase diagram for @xmath190 , also displayed in fig .
[ pd]b , shows some features of the phase diagram in fig .
[ scheme]b or @xmath124 , respectively
. in particular one observes a maximum in the second order phase transition line and different slopes for the second and first order line at the crossover point .
however , the numerical data do not indicate the existence of two consecutive transitions .
thus this phase diagram lies in between the scenarios depicted in figs .
[ scheme]a and [ scheme]b .
= 8.6 cm = 8.6 cm = 8.6 cm = 8.6 cm to study the influence of the interaction @xmath95 qualitatively and quantitatively , it would be desirable to calculate @xmath17 phase diagrams at even larger values of @xmath95
. unfortunately the quantum monte carlo approach fails in this case due to the problem of `` sticking '' as mentioned in sec .
therefore , we concentrate on the crossover from the intermediate coupling regime with first _ and _ second order transitions to the weak coupling regime with first order transitions _ only_.
the results for half filling are collected in fig .
[ tcvsu ] showing the @xmath95dependence of two transition temperatures : the top curve is the nel temperature @xmath191 at @xmath192 ( taken from ref .
the lower curve corresponds to the temperature @xmath193 where the second order phase transition line terminates , i.e. it represents either the tricritical or the critical temperature of fig.[scheme ] . for temperatures below @xmath193 a first order metamagnetic phase transition is observed in an external magnetic field .
[ tcvsu ] reveals the crossover from intermediate coupling with first and second order phase transitions to weak coupling with first order transitions only : as @xmath95 decreases the regime with second order phase transitions ( @xmath194 ) shrinks , while the temperature regime for first order transitions remains nearly unchanged up to @xmath195 .
in the preceding sections metamagnetic transitions were investigated in the case of half filling . beyond half filling
the commensurate antiferromagnetic phase remains stable in the parameter regime under consideration ( @xmath196 , @xmath197 ) .
incommensurate spin density waves become stable only in a small density regime at lower temperatures @xcite .
another possible instability of the antiferromagnetic phase away from half filling is phase separation , found within the hartree
fock approximation and in second order perturbation theory at constant order parameter at t=0 @xcite .
however , at least for @xmath198 we do not observe phase separation since the electronic compressibility @xmath199 is finite and positive ( see fig .
[ phassep ] )
. upon doping the magnetization curve changes considerably and hardly indicates the existence of a metamagnetic phase transition ( fig .
[ mvshdop ] ) .
this is due to the fact that in the metallic phase there is no longer a `` slater gap '' at the fermi energy ; therefore the homogeneous susceptibility is not as strongly affected by the antiferromagnetic order as at half filling .
the phase transition is , however , clearly seen in @xmath177 .
= 8.6 cm = 8.6 cm = 8.6 cm from the @xmath144 vs. @xmath18 curve the phase diagram ( fig . [ pddop ] ) is constructed .
the metamagnetic phase transition line is found at lower temperatures and fields compared to half filling .
associated with the metamagnetic phase transition is a change of the electrical resistivity . to study this important effect we calculated the @xmath18dependence of the electronic compressibility @xmath200
this quantity indicates whether the system is metallic or insulating . for an insulator @xmath200
vanishes for @xmath153 and is exponentially small for temperatures lower than the antiferromagnetic gap . in a fermi liquid , on the other hand , @xmath201 is finite since it is proportional to the density of states at the fermi level and hence proportional to the drude conductivity .
= 8.6 cm = 8.6 cm the results for @xmath200 as a function of magnetic field @xmath18 at @xmath178 are shown in fig .
[ comp ] . at half filling , @xmath202 , the compressibility is seen to increase with @xmath18 .
this effect is particular pronounced at low temperatures @xmath203 where @xmath200 is essentially zero at low fields and rises to @xmath204 above the critical field , indicated by an arrow .
hence the metamagnetic phase transition is a transition from an antiferromagnetic insulator to a metal with homogeneous magnetization . at higher temperatures , @xmath205 ,
the compressibility is always finite due to thermal excitations .
we note that at @xmath124 , when the electrons are essentially localized , @xmath200 remains small ( @xmath206 at @xmath207 ; not shown in fig .
[ comp ] ) even above the critical field , indicating an insulator to insulator transition .
the situation is very different at finite doping ( @xmath208 ) . here
[ comp ] shows that @xmath200 _ decreases _ with @xmath18 by approximately 50% as the system goes through the metamagnetic transition from an antiferromagnetic metal to a metal with homogeneous magnetization . _
_ = 8.6 cm
in summary , we investigated the origin of metamagnetism in strongly anisotropic antiferromagnets starting from a microscopic model of strongly correlated electrons , the hubbard model with easy axis , by employing a dynamical mean field theory .
this approach is fundamentally different from previous investigations since we identified and explicitly evaluated the simplest _
electronic , _ i.e. fully quantum mechanical , correlation model that is able to explain the conditions for metamagnetism . for this electronic model
we show unambiguously that at intermediate coupling the phase transition is of first order at low temperatures and of second order near the nel temperature , i.e. the order of the phase transition changes .
this phenomenon , which has been of interest to various communities in classical statistical mechanics for a long time already , is extracted from a model of itinerant electrons . _ _ our approach allows us to describe a broad range of qualitatively different metamagnets within a single model .
while at present this simple model does not permit any quantitative calculation of material properties it does describe itinerant and localized , metallic and insulating metamagnets and the crossover between them .
this crossover is related to two fundamental experimental parameters , i.e. pressure ( related to @xmath209 which decreases with pressure ) and doping . at half
filling the coulomb interaction leads to a crossover from a band insulator to an insulator with localized moments .
thereby the phase transition changes from first order for the bandlike metamagnet to second order for the localized one . only at intermediate couplings
are both first and second order phase transitions observed as found in experiment .
the @xmath17 phase diagram obtained for an intermediate coulomb interaction ( @xmath210 = band width ) is strikingly similar to that of febr@xmath3@xcite or the ising model with @xmath211 @xcite .
we note that in these insulating systems the applicability of a theory which becomes exact for large coordination number is justified by the fact that the af superexchange involves 20 equivalent sites in the two neighboring iron planes @xcite . at smaller values of the coulomb interaction ( @xmath178 )
the temperature regime with second order transitions shrinks and the two step phase transition becomes less pronounced , reproducing the scenario of fig.1a , as observed e.g in fecl@xmath3 @xcite .
the calculations off half filling allow us to investigate the properties of metallic metamagnets , such as the uranium - based mixed - systems @xcite , for which a theory in terms of a correlated electron model is mandatory .
in contrast to the insulating case , the metamagnetic phase transition in the metallic system is hardly visible in the magnetization curve .
this is because there is no longer a gap at the fermi energy .
quite generally , the critical temperatures and fields decrease upon doping .
the metamagnetic transition is accompanied by pronounced changes in the conductivity of the system . the hubbard model with easy axis
can qualitatively describe several scenarios : \(i ) in the insulating , localized regime ( @xmath212 at half filling ) a magnetic field causes a transition from an antiferromagnetic insulator to an insulator with homogeneous magnetization .
\(ii ) at lower @xmath95values ( e.g. @xmath178 ) at half filling an insulator to metal transition occurs at the magnetic field where the af order disappears .
such a phenomenon is observed , for example , in the af phase of la@xmath213ca@xmath10mno@xmath4 , where the resistivity is found to change by several orders of magnitude @xcite .
this is referred to as `` colossal '' magnetoresistance .
we note that la@xmath213ca@xmath10mno@xmath4 shows no strong anisotropy .
therefore our approach can only describe the general features , in particular the existence of the insulator to metal transition .
\(iii ) away from half filling a magnetic field induces a transition from a metallic antiferromagnet to a metal without staggered moment . here the compressibility changes by less than an order of magnitude , e.g. about 50% at @xmath178 , @xmath208 .
a similar effect is found in several strongly anisotropic antiferromagnets , both in multilayers and bulk intermetallic compounds such as updge @xcite . in these systems
the origin of this `` giant '' magnetoresistance is attributed to band structure effects and spin scattering @xcite . by contrast
, our approach stresses the importance of genuine electronic correlation effects .
more detailed investigations , including band degeneracy and spin orbit interaction , may eventually provide even quantitative insight into these interesting and important phenomena .
we acknowledge useful correspondence with n. giordano , g. lander , b. lthi , a. ramirez , and w. wolf , and are grateful to h. capellmann , p. van dongen , w. metzner , h. mller - krumbhaar , j. schlipf , f. steglich , g. stewart and , in particular , to v. dohm , w. kleemann and w. selke for very helpful discussions . in its initial stages
this work was supported in part by the sfb 341 of the dfg and by a grant from the onr , n000149310495 .
quite generally susceptibilities can be obtained from the derivative of the order parameter @xmath214 w.r.t .
the corresponding field @xmath215 : @xmath216 @xmath217 here @xmath218 and @xmath219 lead to the ferromagnetic and antiferromagnetic susceptibilities , @xmath220 to the electronic compressibility , and @xmath221 to the charge density wave susceptibility . from the two self consistency equations ( [ dys ] ) and ( [ gl4 ] ) one obtains two corresponding equations for the derivative of the green function w.r.t .
the variable @xmath215 .
the derivative of the functional integral ( [ gl4 ] ) gives : @xmath222 where @xmath223 is the local two particle correlation function @xmath224 the quantity @xmath225 in ( [ gbar1 ] ) measures the response of the averaged medium to an infinitesimal change of the field @xmath215 .
this dynamical response function is determined by an integral equation in frequency space which does not explicitly depend on momentum .
( note that there are no convolutions in @xmath226 space in the @xmath152 limit as is typical for a mean field theory ) .
this property does not imply , however , that the response function @xmath227 is local , too .
it only indicates that @xmath227 is _ diagonal _ in the momentum @xmath226 .
momentum dependence enters implicitly by the particular @xmath226 dependence of the external field ( @xmath228 in the case of the compressibility or the ferromagnetic susceptibility , and @xmath229 for the staggered susceptibility ) . in the presence of an external field the variables @xmath230 in the dyson equation ( [ dys ] ) are replaced by @xmath231 .
the derivative of the green function yields : @xmath232 with @xmath233 since ( [ gs2 ] ) separates in matsubara frequencies @xmath234 and spin @xmath235 it can be easily solved for @xmath236 : @xmath237 with the @xmath238 array in @xmath239 : @xmath240 whereby @xmath241 and @xmath242 are defined as @xmath243 \frac{\eta^{\sigma}_{b \ , n } } { ( g_{a \ , n}^{\sigma})^2 } & 1 - \frac{\zeta^{\sigma}_{a \ ,
n } } { ( g_{a \ , n}^{\sigma})^2 } \\
\end{array } \right),\\[8pt ] { \mathbf t}^\sigma _ { \ , n}&= & \left ( \begin{array}{cc } \zeta^{\sigma}_{a \ , n } & \eta^{\sigma}_{a \ , n } \\[8pt ] \eta^{\sigma}_{b \ , n } & \zeta^{\sigma}_{b \ , n } \\ \end{array } \right).\end{aligned}\ ] ] now @xmath244 can be eliminated by setting ( [ gbar1 ] ) equal to ( [ gbar2 ] ) , yielding @xmath245 from this equation we determine @xmath246 by numerical inversion of a @xmath247 matrix .
knowing @xmath246 we obtain @xmath244 via ( [ gbar1 ] ) or ( [ gbar2 ] ) and thus the susceptibility ( [ susg ] ) .
[ perpo ] [ hfo ] similar to the derivation of the hartree fock equations for @xmath249 ( see sec . [ wc ] ) we will now investigate the case with perpendicular orientation @xmath250 .
the ansatz for the one particle densities @xmath251 yields in addition to the hartree term a fock term in the decoupling ( [ hde ] ) @xmath252 with this ansatz one readily obtains the effective one particle hamiltonian @xmath253 diagonalizing this hamiltonian yields the one particle energies @xmath254 from these energies the free energy @xmath115 is calculated , and the minimization with respect to @xmath145 and @xmath144 leads to the following hartree fock self consistency equations @xmath255 @xmath256}.\ ] ] as in sec .
[ wc ] these hartree
fock equations are solved numerically .
the itinerant electron metamagnetism theory of moriya and usami @xcite can be derived from the hartree
fock approximation only if the free energy is analytic in @xmath145 and @xmath144 .
since in rpa the hubbard interaction @xmath95 contributes to the free energy analytically ( see ( [ iemlandau ] ) and ( [ rpa ] ) ) any non analytic behavior must be due to the kinetic energy .
its expansion in the order parameter @xmath144 at @xmath153 is analyzed in this section . to calculate the expansion in @xmath144 a staggered magnetic field @xmath257
is introduced : @xmath258 on @xmath52-@xmath53 lattices the one particle energies for this hamiltonian show a gap at @xmath259 with square root singularities at its edge @xmath260 we consider the half filled band , where the staggered magnetization @xmath144 is calculated from the one particle energies .
it shows the following asymptotic non analytic behavior for @xmath261 : @xmath262 similarly the asymptotic behavior of the energy ( [ hhst ] ) is obtained as @xmath263 subtracting the contribution due to @xmath257 , @xmath264 , the asymptotic dependence of the kinetic energy on @xmath144 reads @xmath265 this shows that @xmath266 is non analytic in @xmath144 .
therefore the itinerant electron metamagnetism theory can not be derived from the hartree
fock theory for the hubbard model with an easy axis .
in the limit of strong coupling and half filling the @xmath267 perturbation theory yields the effective spin hamiltonian ( [ hsc ] ) . in the following , we study the metamagnetic phase transition , and especially the order of the transition , for this effective hamiltonian . restricting ourselves to solutions with mixtures of ferromagnetic ( @xmath145 ) and antiferromagnetic ( @xmath144 ) order ,
the ground state energy is a polynomial in @xmath145 and @xmath144 ( @xmath268 ) : @xmath269 one can see that the ferromagnetic next nearest neighbor interaction @xmath173 favors both saturated antiferromagnetism and saturated ferromagnetism rather than ferrimagnetic phases .
by contrast the plaquette term @xmath52 has contributions that support the formation of a ferrimagnetic state . to obtain the ground state
the energy must be minimized with respect to @xmath145 and @xmath144 under the constraints @xmath270 and @xmath271 .
differentiation of @xmath272 with respect to @xmath144 shows that , for fixed @xmath145 , @xmath272 has one maximum at @xmath273 and two minima at @xmath274 for sufficiently strong coupling @xmath95 ( e.g. @xmath275 in the case of the hypercubic lattice with @xmath52=20 , @xmath173=2 and @xmath53=4 ) these minima are outside the constraint @xmath276 .
therefore @xmath272 becomes minimal at the border of the constraint , i.e. for @xmath277 .
replacing @xmath144 by @xmath278 the minimization with respect to @xmath145 readily yields for the ground state @xmath279 1 \hspace{4em } \mbox{for } \hspace{1em } h
\geq \frac { - { c } + { u}^{2 } + { b } } { { u}^{3 } } \\[4pt ] \frac { { a } - 2\,{c } - 2\,{u}^{2 } - 2\,{b } + 2 \,{h}\,{u}^{3 } } { { a } - 4\,{c } } \hspace{2em } { \mathrm else}\\ \end{array } \right . \\ m_{st } & = & 1 - m.\end{aligned}\ ] ] this ground state solution for the effective spin hamiltonian shows a second order metamagnetic phase transition for @xmath280 .
this is the case for the hypercubic lattice ( @xmath281 ) and for the bethe lattice , where @xmath52=0 but @xmath282 , i.e. the next nearest neighbor coupling is antiferromagnetic . in conclusion ,
the strong coupling theory shows second order phase transition for all temperatures , even at @xmath153 .
j. h. v. j. brabers et al .
b * 50 * , 16410 ( 1994 ) .
v. ivanov et al . , j. alloys comp . *
218 * , l19 ( 1995 ) .
t. goto et al . , j. appl .
phys . * 76 * , 6682 ( 1994 ) . for
a review see b. lthi et al .
, j. magn .
magn . mater . *
90 & 91 * , 37 ( 1991 ) .
this hamiltonian allows the study of metamagnetism in strongly anisotropic antiferromagnets such as fecl@xmath3 , febr@xmath3 , which are layered materials with ferromagnetic interactions within the plane and antiferromagnetic interactions between the planes .
adjacent layers belong to different sublattices .
note that for @xmath21 both interactions in ( 1 ) _ favor _ antiferromagnetic ordering on nn sites .
j. hubbard , proc .
r. soc .
london ser .
a * 276 * , 238 ( 1963 ) .
m. c. gutzwiller , phys .
* 10 * , 59 ( 1963 ) .
j. kanamori , prog .
30 * , 275 ( 1963 ) .
m. takahashi , prog .
phys . * 42 * 1098 ( 1969 ) .
in addition to the spin
orbit interaction there is also the dipole
dipole interaction between electronic spins .
the latter is mainly responsible for the `` shape anisotropy '' , i.e. the fact that the direction of the easy axis depends on the shape of the probe .
an introduction into the origins and models of the magnetic anisotropy is given by p. bruno , _ physical origins and theoretical models of magnetic anisotropy _
, in _ magnetismus von festkrpern und grenzflchen _ , 24 .
iff ferienkurs , forschungszentrum jlich , 1993 , isbn 3 - 89336 - 110 - 3 , p. 24.1 ; p. bruno , phys .
b * 39 * , 865 ( 1989 ) .
this procedure is quite similar to _ assuming _ the existence of a lattice on which the hubbard model is defined .
the existence of a lattice and the correlation physics described by the hubbard model are _ a priori _ quite unrelated , and hence we do not need to start on the level of interacting electrons in the continuum . _ _
f. gebhard , _ the mott metal insulator transition _ , springer tracts in modern physics , vol . 137 ( springer , heidelberg , 1997 ) .
for a review see th .
pruschke , m. jarrell and j. k. freericks , adv .
* 44 * , 187 ( 1995 ) .
it can be shown that the hartree
fock results remain qualitatively correct within second order perturbation theory at fixed order parameter : p.
g. j. van dongen , phys .
b * 50 * , 14016 ( 1994 ) .
m. takahashi , j. phys .
c * 10 * , 1289 ( 1977 ) .
g. j. van dongen , phys .
b * 49 * , 7904 ( 1994 ) .
k. held , m. ulmke , and d.vollhardt , mod . phys . lett .
b * 10 * , 203 ( 1996 ) .
j. k. freericks and m. jarrell , phys .
lett . * 74 * , 186 ( 1995 ) .
note that the mixed
phase region in the phase diagram of ref .
[ 21 ] disappears when instead of the applied field @xmath27 the internal field @xmath18 is plotted vs. @xmath19 .
l. hernandez , h. t. diep and d. bertrand , europhys .
* 21 * , 711 ( 1993 ) .
p. schiffer , a. p. ramirez , w. bao , and s .- w .
cheong , phys .
lett . * 75 * , 3336 ( 1995 ) .
p. m. levy and s. zhang , j. magn .
* 151 * , 315 ( 1995 ) . | the microscopic origin of metamagnetism and metamagnetic transitions in strongly anisotropic antiferromagnets is investigated within a quantum mechanical theory of correlated electrons . to this end the hubbard model with staggered magnetization @xmath0 along an easy axis @xmath1 in a magnetic field @xmath2 is studied both analytically and numerically within the dynamical mean field theory ( dmft ) . at intermediate couplings the self consistent dmft equations , which become exact in the limit of large coordination number ,
are solved by finite temperature quantum monte carlo techniques .
the temperature and magnetic field dependence of the homogeneous and staggered magnetization are calculated and the magnetic phase diagram is constructed . at half filling the metamagnetic transitions are found to change from first order at low temperatures to second order near the nel temperature , implying the existence of a multicritical point .
doping with holes or electrons has a strong effect : the system becomes metallic , the electronic compressibility increases and the critical temperatures and fields decrease .
these results are related to known properties of insulating metamagnets such as febr@xmath3 , metallic metamagnets such as updge , and the giant and colossal magnetoresistance found in a number of magnetic bulk systems .
+ pacs : 71.27+a , 75.10.lp , 75.30kz |
self - consistent mean field ( scmf ) models are the most feasible means for the microscopic description of ground state properties and low - energy collective dynamics of nuclei .
these models are also employed for the study of compact stars , since they can be easily extended to include the contributions from hyperons and exotic phenomena as , e.g. , kaon condensation @xcite .
the three most prominent scmf models are the skyrme hartree - fock approach ( shf ) @xcite , the gogny force @xcite , and the relativistic mean field model ( rmf ) @xcite .
we shall focus on shf and rmf in comparison .
shf , as the name suggest , is based on the skyrme energy functional derived from a zero - range effective interaction .
the rmf models are based on an effective lagrangian density which describes the interactions of nucleons through the exchange of the scalar - isoscalar ( @xmath0 ) , vector - isoscalar ( @xmath1 ) and vector - isovector ( @xmath2 ) meson fields .
the form of the energy functional or lagrangian density is given by invariance constraints and plausible physical arguments ( as , e.g. , low - momentum expansion @xcite ) .
however , the parameters of the models are free parameters
. a strict derivation from a microscopic nucleon - nucleon interaction has not yet been too successful at a quantitative level because ab - initio models , in spite of the enormous success over the last years , have still a limited quality if one refrains from adding empirical corrections like a three - body force ( see , e.g. , @xcite ) .
one of the important tasks in the development of mean field models is thus to devise ways of constraining the free parameters of interaction used .
this is done usually by an adjustment to empirical data .
this is , however , rather involved as one does not have a one - to - one correspondence between individual parameters and a specific piece of experimental information .
some parameters are well fixed by known nuclear ground state properties .
others remain only loosely determined leaving leeway for additional data or bias .
therefore , many tens of different parameter sets for shf and rmf have evolved in the course of the development .
a key task is to figure out which features depend on a particular parameterization and which features are more related to the intrinsic structure of the model .
recent extensive explorations of the shf functional have indicated a few of such intrinsic features @xcite .
one key problem is that all shf functionals show a systematic trend to underbinding for super - heavy elements while traditional rmf parameterizations show an opposite trend to overbinding .
it is the aim of the present contribution to study this discrepancy in more detail , in particular with more data and for a broader selection at the side of the rmf . to this end
, we shall compare the binding energy systematics for even - even nuclei computed using the standard and extended versions of the rmf models and the shf models .
the rmf can be grouped into three different classes : ( 1 ) models with non - linear meson self - coupling @xcite , ( 2 ) models with density - dependent nucleon - meson coupling @xcite , ( 3 ) models which combine ( 1 ) and ( 2 ) . we concentrate here on class ( 1 ) . within these non - linear rmf
one can distinguish two sub - classes : ( i ) standard rmf which includes the contributions from the the non - linear self - couplings for the @xmath0 meson field only , and ( ii ) extended rmf model which includes the contributions from the self and/or cross interaction terms for the @xmath0 , @xmath1 and @xmath2 meson fields .
we will discuss both versions of meson coupling .
we have considered four different forces for both the rmf and shf models .
the specific reasons for choosing these forces are as follows .
in particular , we employ the rmf parameterizations : nl3 @xcite , tm1 @xcite , fsugold @xcite , and bsr4 @xcite . the non - linear part of the lagrangian density for these rmf models are not the same . these differences are summarized in table [ tab : rmf ] . the lagrangian density associated with the parameter set nl3 which represents the standard rmf model contains non - linear term only for the @xmath0 meson self - interaction .
the lagrangian density corresponding to the parameter set tm1 is extended to include the self - interaction term also for the @xmath1 mesons .
the fsugold parameter set further includes the @xmath3 cross - interaction term .
the lagrangian density corresponding to the bsr4 parameter set includes all the cross - interaction terms for @xmath0 , @xmath1 and @xmath2 mesons up to quartic order , but , the self - interaction of the @xmath1-meson is switched off . these parameter sets
should enable us to delineate the effects of the meson self- and cross - interactions on the binding energy systematics .
we neglect the contributions from the self interaction of @xmath2 mesons as they can affect the properties of the finite nuclei and neutron stars only marginally @xcite . at the side of shf , we consider the following four parameterizations : ski3 @xcite , sly6 @xcite , bsk4 @xcite and sv - min @xcite .
the form of the shf functional is basically the same for all four cases with the tiny difference that ski3 and sv - min include an isovector contribution to the spin - orbit term @xcite .
the parameterizations differ mainly in the bias and choice of data for the phenomenological adjustment .
the two older parameterizations ski3 and sly6 used a rather small set of fit nuclei .
ski3 included the electro - magnetic form factor and isotope shifts as data while sly6 put emphasis on the neutron rich side and included data from bulk matter .
bsk4 puts emphasis almost exclusively on binding energies , and includes all known even - even nuclei in the data base .
sv - min takes a data set of good `` mean - field nuclei '' which were carefully selected to contain only negligible corrections from collective correlations @xcite .
this amounted to long chains of semi - magic nuclei .
the fit set is much larger than the one for ski3 and sly6 , but much smaller than for bsk4 . on the other hand
, sv - min takes care not only of binding energies but of a larger set of observables ( radii , electro - magnetic form factor , spin - orbit splittings , ... ) .
the four selected forces are thus sufficiently different with respect to these aspects of bias and choice .
we would like to briefly outline the manner in which the corrections to the binding energies arising from the centre of mass motion , pairing and quadrupole correlations are incorporated .
all models include the correction for the spurious center - of - mass energy @xmath4 .
the shf sets compute it from , @xmath5 while the rmf sets use the estimate @xmath6 the values of @xmath7 computed using eqs .
( [ eq : ecm1 ] and [ eq : ecm2 ] ) for the cases with mass number @xmath8 , which is of present interest , may differ at most by @xmath9 mev @xcite .
the pairing is treated within the shf models using a ( density dependent ) zero - range force and smooth cutoff in pairing space @xcite . for the rmf models ,
the contributions from the pairing correlations to the binding energy are evaluated in the constant gap approximation with the gap @xcite , @xmath10 soft nuclei and deformed nuclei can develop substantial contributions from quadrupole correlations @xcite . as an efficient and simple estimate
we include a large part of these correlations in terms of approximate rotational projection expressed through variance of total angular momentum @xmath11 and moment of inertia @xmath12 @xcite .
this is done so far for the shf models . in case of rmf
, we use a simple estimate for the quadrupole correlation energy or rotational correction as , @xmath13 where @xmath14 is the nucleon effective mass in bulk equilibrium matter .
this estimate has been extracted by studying the microscopically computed trends of @xmath15 for a wide variety of shf parameterizations . thus , the manner in which the contributions from the c.m .
correction @xmath4 , from the pairing correlations and from the rotational correction @xmath16 to the binding energy evaluated for the shf and the rmf models are not exactly the same .
we have checked these variants extensively and found that they do not affect the general trends which we aim to discuss . to this end , we may point out that the contributions from the fock and the coulomb exchange terms are ignored in the rmf models , but , included in the shf models .
we first present some numerical details before embarking on our main results .
the rmf equations for nucleons are solved by expanding the dirac spinors in terms of the wave functions for the axially deformed harmonic oscillator potential @xcite .
similar strategy is employed for solving the field equations for the mesons .
the basis space for the nucleons and the mesons are truncated at the major oscillator shells @xmath17 and @xmath18 , respectively .
the reliable solutions for the field equations can be obtained with @xmath19 @xcite .
however , the appropriate choice of @xmath17 depends on the mass number of the nucleus .
for instance , reliable binding energy for @xmath20pb requires number of oscillator shells @xmath21 .
it is quite natural to expect the need for larger basis space for the case of superheavy nuclei which are also well deformed in general .
our prime interest is to look into the binding energy systematics of the superheavy nuclei , it is therefore necessary to ensure that the basis space used is sufficiently large enough to yield meaningful values for the binding energies for these nuclei . in figs .
[ fig : zpe0 ] and [ fig : zpe ] we display the results for the binding energy error or equivalently the mass difference between the measured and the calculated ones , @xmath22 as a function of @xmath17 for some heavy nuclei .
the values of @xmath23 as used for the results plotted in fig .
[ fig : zpe ] are obtained from eq .
( [ eq : erot ] ) .
it can be seen from these figures that we need at least @xmath24 to generate reliable binding energy systematics over wide range of mass number . in table
[ tab : comp ] we compare our results for the binding energies for a few selected nuclei with those of ref .
@xcite obtained for @xmath25 .
we see that our results obtained with the @xmath25 agree reasonably well with those of ref .
though , the pairing gaps used in our paper may be somewhat different .
large differences in the binding energy for the super - heavy nuclei in our work as compared to that of ref .
@xcite are mainly due to the fact that we have used sufficiently large basis space ( see also figs . [
fig : zpe0 ] and [ fig : zpe ] ) .
all the results obtained within the rmf models as presented below correspond to @xmath26 with @xmath27 . in fig .
[ fig : rmf_shf ] we summarize the errors in binding energies ( eq . [ eq : delm ] ) for all experimentally known even - even nuclei .
the upper block collects the results from rmf and the lower block those from shf . as a first impression we note the rather small energy scale of a few mev for the error plots .
all parameterizations shown here provide a good description of known nuclei in the mass range @xmath28 220 .
there are , of course , differences in detail to the extend that in each block the most recent parameterizations perform visibly best .
this concerns bsr4 in the rmf block and bsk4 as well as sv - min in the shf block .
we have indicated the present `` state of the art '' by @xmath29 mev error bars and these three most recent parameterizations stay close to this goal .
it is also found as a general rule that robust spherical nuclei , denoted as `` fit nuclei '' in most parameterizations , are usually somewhat better described than soft or deformed nuclei . at this point it is important to emphasize that the still good description of soft and deformed nuclei is achieved only with the rotational correction and one may reduce the error even more by performing the full collective correlations @xcite .
we also calculate the rms error for the binding energies using the results for all the 513 even - even nuclei as considered in the present work .
these rms errors as given in the parenthesis in the units of mev are : bsr4(2.6),fsugold(6.5),nl3(3.8),tm1(5.9),sv - min(1.6),bsk4(1.1 ) , ski3(2.6),sly6(2.3 ) .
our rms error for the case of bsk4 is higher than @xmath30 mev @xcite . because we use different rotational correction , which probably has some effect for deformed nuclei .
moreover , the basic paper for the bsk4 force uses a different recipe for the cutoff of the pairing space .
their pairing band amounts to 15 mev while we stay typically in the range of 5 mev . that could have a some effect on heavy nuclei . nevertheless , it may be emphasized that these differences in detail do not change the gross trends for the binding energy errors . in particular trends for the binding energy errors for the super - heavy nuclei presented in fig .
[ fig : rmf_shf ] are very much similar to those of ref .
@xcite .
having a closer look at the trends with mass number we see in all cases a growing deviation from the zero line with increasing mass number .
this trend shown here on a much larger data basis confirms what had been observed in @xcite , namely that shf drives to underbinding of super - heavy nuclei while rmf shows just the opposite trend .
this trends become manifest in the region @xmath31 , but develop already for deformed nuclei in the region @xmath32 . within the shf parameterizations
there are significant differences in the quantitative level of the deviation .
bsk4 manages best to minimize the unwanted trend , achieved through the strong bias on binding energies , however , other observables are compromised . in the realm of the rmf
the trend to overbinding is even more dramatic .
this is obvious in the traditional rmf , parameterization nl3 in figure [ fig : rmf_shf ] .
somewhat surprisingly , the situation does not improve with @xmath1 self - couplings in tm1 and the one more cross coupling in fsugold .
it is only bsr4 having the isoscalar and isovector cross - couplings with the @xmath0 fields which allows to produce a much better overall trend in binding energies .
these couplings seem to be a crucial ingredient for success and it is to be noted that shf contains them in the isoscalar and isovector density - dependent terms .
we have looked into the possible consequences of the approximations used in the present work to evaluate the pairing and quadrupole correlation energies for the rmf cases . in fig .
[ fig : svmin - ski3 ] , we plot the shf results for the errors in the binding energies obtained for the sv - min and ski3 forces . for both the cases , the pairing and quadrupole correlation energies are calculated approximately using eqs .
( [ eq : pair ] ) and ( [ eq : erot ] ) , respectively . comparing these results with the corresponding ones as shown in lower block of fig .
[ fig : rmf_shf ] , which are obtained by treating the pairing and quadrupole correlations microscopically , we find that the approximations used to evaluate the pairing and quadrupole correlation energies do not affect the global trends . finally , we compare our results for the bsr4 force for a few heavy nuclei with those obtained using the density dependent meson exchange , ddme2 , force @xcite .
it appears that the bsr4 and ddme2 forces yield similar trends for the binding energy errors for the heavy nuclei . comparing the three best performers , bsr4 , bsk4 and sv - min
, we see that both models are approaching good control over the energies for super - heavy elements . and
yet , there remains an unresolved trend which still is distinctively different between shf and rmf .
the reasons for that are not yet clear .
we try briefly to sort out possible mechanisms .
the defect seems to come from the deformation energy because the large deviation develops with deformation .
this , in turn , localizes the differences in the modeling of the surface energy , and most probably isovector surface energy ( also called surface symmetry energy ) because heavier nuclei have naturally a larger asymmetry .
a direct relation between surface energies and model parameters has not yet been established @xcite . at present
we can only speculate and try to figure out the terms which could have an influence .
although one can map at lowest order @xmath33 the rmf into a skyrme - like functional including proper kinetic and spin - orbit terms @xcite , there remain basic differences between the two classes of models .
the spin - orbit and kinetic terms in the rmf carry effectively a strong density dependence while the corresponding shf terms stay simply linear in @xmath2 .
this yields a different performance with respect to spin orbit splitting and a different density profile in super - heavy elements @xcite , which both are possible sources for different deformation energies .
moreover , the rmf produces a strong link between effective mass and spin - orbit force where one of the consequences is that rmf models produce typically very low effective masses in the range @xmath34 while shf has much more freedom in the effective mass and in adjusting the spin - orbit term independently .
it is certainly a worthwhile task for future work to find out the mechanisms determining the trends as it will shed light on the structure of both approaches , shf and rmf , and indicate missing pieces in the functionals .
in summary , we have calculated the binding energies , for all experimentally known even - even nuclei , within the rmf and the shf models . for the rmf models , we have used four different parameter sets corresponding to different functional forms for the non - linear part of the lagrangian density .
the nl3 parameterization corresponds to the lagrangian density associated with the standard rmf model . while the lagrangian density for the tm1 , fsugold and bsr4 parameterizations correspond to the extended rmf model which include contributions from the self - interaction of the @xmath1 meson and/or the cross - interaction between @xmath0 , @xmath1 and @xmath2 mesons ( see table [ tab : rmf ] ) . for the shf model
we have used four parameter sets , namely , ski3 , sly6 , bsk4 and sv - min .
the form of the shf functional is basically the same for all four cases with the tiny difference that ski3 and sv - min include an isovector contribution to the spin - orbit term @xcite .
we find that all parameterizations of the rmf and shf model considered provide a good description of known nuclei in the mass range @xmath28 220 .
there are , of course , differences in detail to the extent that the most recent parameterizations perform the best .
these parameterizations are bsr4 in the rmf model and bsk4 as well as sv - min in the shf model .
the binding energies for the super - heavy nuclei ( @xmath35 ) for the rmf and shf models are significantly different from the experimental data .
the shf drives to underbinding of super - heavy nuclei while rmf shows just the opposite trend .
this trend on a much larger data basis confirms what had been observed in @xcite .
the absolute errors in the binding energies for the extended rmf model ( bsr4 ) which includes contributions from all the cross - interaction terms are comparable to that for the shf models . other rmf models considered yield much larger values for the absolute errors in the binding energies for the super - heavy nuclei .
.[tab : rmf ] various self - interaction and cross - interaction terms present in the lagrangian density associated with different parameterizations of the rmf models .
the index 1 and 0 is used to indicate whether or not the corresponding term is included . [ cols="^,^,^,^,^,^,^ " , ] + & & & + o16 & & 128.82 & 128.84 + ca40 & 341.91 & 341.96 & 341.99 + ca48 & 415.07 & 415.11 & 415.06 + ni58 & 503.54 & 503.39 & 503.08 + zr90 & 783.41 & 783.08 & 783.06 + sn116 & 986.44 & 986.15 & 986.79 + sn124 & 1048.58 & 1048.15 & 1048.49 + sn132 & 1104.72 & 1104.81 & 1104.77 + pb208 & 1640.16 & 1640.30 & 1640.50 + pb214 & 1662.70 & 1662.02 & 1662.52 + th232 & 1764.99 & 1764.12 & 1767.33 + cf248 & 1857.17 & 1856.48 & 1861.51 + hs264 & & 1927.95 & 1934.68 + | we compare the systematics of binding energies computed within the standard and extended versions of the relativistic mean - field ( rmf ) model and the skyrme hartree - fock ( shf ) model .
the general trends for the binding energies for super - heavy nuclei are significantly different for these models . the shf models tend to underbind the superheavy nuclei , while , rmf models show just the opposite trend .
the extended rmf model seems to provide remarkable improvements over the results obtained for the standard rmf model . |
in this paper we apply the path integration technique recently developed for the statistical sum of the generic time - parametrization invariant gravitational system with the frw metric @xcite to a particular case of the cosmology generated by the conformal field theory ( cft ) with a large number of quantum species .
this problem is motivated by the recently suggested model of initial conditions in cosmology in the form of the microcanonical density matrix @xcite . when applied to cosmology with a large number of fields conformally coupled to gravity
, this theory can be important within the cosmological constant and dark energy problems .
in particular , its statistical ensemble is bounded to a finite range of values of the effective cosmological constant , it generates an inflationary stage and is potentially capable of generating the cosmological acceleration phenomenon within the so - called big boost scenario @xcite . as shown in @xcite , for a spatially closed cosmology with @xmath0-topology the microcanonical statistical sum can be represented by the euclidean quantum gravity path integral , @xmath1\ ; e^{-s[\,g_{\mu\nu},\phi\ , ] } , \end{aligned}\ ] ] over the metric @xmath2 and matter fields @xmath3 which are periodic on the euclidean spacetime with a compactified time @xmath4 ( of @xmath5 topology ) .
the frw metric arises in this path integral as the set of major collective variables of cosmology . under the decomposition of the full set of @xmath6 into the minisuperspace frw sector @xmath7 and inhomogeneous `` matter '' fields
@xmath8 on the background of this metric the path integral can be cast into the form of an integral over a minisuperspace lapse function @xmath9 and a scale factor @xmath10 , @xmath11\ ; e^{-\vargamma[\,a,\,n\ , ] } , \label{1}\\ & & e^{-\vargamma[\,a,\,n ] } = \int d\varphi(x)\ , e^{-s[\,a,\,n;\,\varphi(x)\,]}\ .
\label{2 } \end{aligned}\ ] ] here , @xmath12 $ ] is the euclidean effective action of the fields @xmath13 ( which include also the metric perturbations @xmath14 ) on the frw background , and @xmath15\equiv s[\,g_{\mu\nu},\phi\,]$ ] is the original action rewritten in terms of this minisuperspace decomposition .
it is important that this representation is not a minisuperspace approximation , when all the fields @xmath16 are frozen out .
rather this is the disentangling of the collective degrees of freedom from the configuration space , the rest of which effectively manifest itself in terms of this effective action .
semiclassically the integral ( [ 1 ] ) is dominated by the contribution of the saddle point the periodic solution of the effective friedmann equation of the _ euclidean _ gravity theory @xmath17,$ ] the second equation @xmath18 is the derivative of the friedmann equation and should not be imposed independently of the latter @xcite . ]
@xmath19 here @xmath20 $ ] is taken at this solution and @xmath21 is the preexponential factor accumulating perturbation corrections of the semiclassical expansion . for the cft driven cosmology the instanton solutions
were obtained in @xcite as a function of the primordial cosmological constant , whereas the calculational technique for the one - loop prefactor @xmath21 was developed in @xcite . here
we calculate this prefactor for the set of instantons in a special range of the cosmological constant in which they have a one - fold nature a single oscillation of the scale factor @xmath10 between its minimal and maximal values . in sect.2
we give the description of the full set of instanton solutions of @xcite and compare them with those of @xcite the paper which also considers trace anomaly driven cosmology but differs from our work by setting the problem and the choice of the field - theoretical model .
then after a brief formulation of the calculational method of @xcite in sect.3 we calculate the one - loop prefactor @xmath21 in sect.4 . there we begin with the description of the above mentioned set of one - fold instantons and their various characteristics and then estimate the magnitude of @xmath21 for quantum fields of various spins .
the paper is accomplished with concluding remarks in conclusions .
in the theory with a primordial cosmological constant @xmath22 and a large number of free ( linear ) fields @xmath3 conformally coupled to gravity conformal field theory ( cft ) , @xmath23= -\frac1{16\pi g } \int d^4x\ , g^{1/2}\,\big(r -2\lambda\big ) + s_{cft}[\,g_{\mu\nu},\phi\ , ] , \end{aligned}\ ] ] the effective action @xmath24 $ ] is dominated by the contribution of these fields because they simply outnumber the non - conformal fields including , in particular , the graviton .
then this quantum effective action is exactly calculable as a _ functional _ of histories @xmath25 by the conformal transformation converting ( [ frw ] ) into the static ( einstein universe ) metric with @xmath26 @xcite .
the result reads @xcite @xmath27= \oint d\tau\,n { \cal l}(a , a')+
f(\eta ) , \label{action1}\\ & & \eta=\oint d\tau\,\frac{n}a , \label{eta } \end{aligned}\ ] ] where @xmath28 and the integration runs over the period of @xmath4 on the circle @xmath29 of @xmath5 . here
the effective lagrangian of its local part @xmath30 includes the classical einstein term with the renormalized planck mass @xmath31 and the cosmological constant @xmath32 and contains also the contribution of the conformal anomaly of quantum fields and their vacuum ( casimir ) energy , @xmath33 the constant @xmath34 is determined by the coefficient of the gauss - bonnet term in the overall conformal anomaly of all cft fields @xmath3 @xcite .
a nonlocal part of the action @xmath35 is the free energy of their quasi - equilibrium excitations with the temperature given by the inverse of the conformal time ( [ eta ] ) .
this is a typical boson or fermion sum over field oscillators with energies @xmath36 on a unit 3-sphere .
semiclassically the integral ( [ 1 ] ) is dominated by the saddle points solutions of the effective friedmann equation of the _ euclidean _ gravity theory @xmath37 modified by the quantum @xmath38-term and the radiation term @xmath39 .
the constant @xmath40 here characterizes the sum of the renormalized casimir energy @xmath41 and the energy of the gas of thermally excited particles @xmath42 thus , the amount of thermal radiation in this self - consistent back reaction problem is not put by hands , but determined by the _ bootstrap _ equation ( [ bootstrap ] ) which yields @xmath40 as a functional of the frw geometry in which this radiation evolves .
the inverse temperature @xmath43 the instanton period in units of the conformal time is given by the integral ( [ eta ] ) over the full period of @xmath4 or the @xmath44-multiple of the integral between the two neighboring turning points of the scale factor history @xmath10 , @xmath45 .
this @xmath46-fold nature implies that in the background solution during the full period of its euclidean time the scale factor oscillates @xmath46 times between its maximum and minimum values @xmath47 , @xmath48 , @xmath49 and forms a kind of a garland of @xmath5 topology with oscillating @xmath0 sections . as shown in @xcite ,
these garland - type instantons exist only in the limited range of the cosmological constant @xmath32 , @xmath50 and are weighted in the relevant statistical ensemble by their exponentiated on - shell action according to ( [ z ] ) , where @xmath20 $ ] is taken at the solution of eqs.([efeq])-([bootstrap ] ) .
in particular , a set of instantons with @xmath51 a single oscillation of the scale factor @xmath10 between its minimal and maximal values ( [ apm ] ) occupies a certain continuous range of @xmath52 inside @xmath53 $ ] adjacent to its lower boundary @xcite . the main effect of these exponential weights is a complete suppression of another set of solutions the vacuum hartle - hawking instantons @xcite with no radiation @xmath54 , which represent the euclidean de sitter spacetime .
these purely vacuum contributions are ruled out by their infinite _ positive _ effective action ( cf .
@xmath55-factor in the kinetic @xmath38-terms of the effective action ( [ effaction ] ) which render its integrand diverging to @xmath56 at @xmath57 with @xmath58 and @xmath59 )
. otherwise , these exponential weights are of the same order of magnitude and they are inefficient to select most probable configurations .
this makes important the calculation of the one - loop prefactor @xmath21 which is the main focus of this paper .
the technique for this prefactor was developed for a parametrization invariant system ( [ action1 ] ) with a generic lagrangian @xmath30 restricted only by the order of derivatives of @xmath60 . )
does not contain higher order derivatives of @xmath60 .
this is the result of the special renormalization @xcite , which does not introduce into the minisuperspace sector of einstein theory extra degrees of freedom .
this choice was motivated in @xcite by ghost - free requirements and certain universality properties which , in particular , relate the value of the casimir energy in ( [ effaction ] ) , @xmath61 , to the coefficient @xmath38 of the conformal anomaly @xcite . ] below we apply this technique to the cft diven cosmology of the above type for a set of one - fold instantons with @xmath51 .
cosmological instantons in the trace anomaly driven cosmology were also considered in @xcite .
it should be emphasized , however , that setting the problem and the field - theoretical model itself in this work are essentially different from our approach in @xcite . to begin with
, the solutions of @xcite correspond to the hartle - hawking no - boundary wavefunction and have the @xmath62-topology in contrast to our periodic @xmath5 instantons associated with the microcanonical statistical sum .
therefore , the amount of radiation in @xcite is a freely specifiable constant of integration , while here it is strictly fixed by the bootstrap " equation ( [ bootstrap ] ) .
secondly , the freedom of uv renormalization is not used in @xcite to eradicate higher order derivatives from the friedmann equation , whereas in @xcite this freedom is fixed by the requirement to preserve the classical number of physical degrees of freedom and the stability of the theory against higher - derivative ghosts ( cf .
footnote 2 above ) .
finally , a primordial ( or renormalized ) cosmological constant is assumed to be zero in @xcite , whereas here @xmath22 belongs to a nontrivial range where it parameterizes the istanton solutions . as a result
the so - called double - bubble instantons of @xcite do not exist in our setting .
this automatically removes the collapsing universes found in @xcite by analytic continuation to the physical lorentzian time .
instead we have a thermal version of the hartle - hawking instantons the instanton garlands of @xcite which give rise to expanding lorentzian universes with different values of the primordial cosmological constant @xmath22 .
as discussed in @xcite , they contain a rapidly diluting radiation of conformal particles and have a finite inflationary stage terminating with the decay of @xmath22 if the latter has a composite nature of a slowly varying inflaton field .
this justifies the use these instantons as initial conditions for the universe , capable of formation of the observable large scale structure .
note that in our setting even the vacuum hartle - hawking instantons mentioned above originate from the periodic boundary conditions as a limiting case when a torus @xmath63 gets ripped at the vanishing value of @xmath64 and topologically becomes a 4-sphere @xmath62 .
thus , these round sphere instantons exist just like in @xcite , ) for @xmath65 exactly coincides with its analogue ( 2.18 ) in @xcite under a special choice of the uv counter term @xmath66 which eradicates higher order derivatives from this equation .
but , as mentioned in @xcite , the inclusion of higher derivatives induced by @xmath67 does not destroy a round sphere instanton , while adding a nonvanishing cosmological constant only modifies its effective hubble parameter . ]
but as mentioned above they are completely ruled out by their infinite positive euclidean action @xmath68 @xcite the diverging contribution of the conformal anomaly .
quite interestingly , this property was not observed in @xcite because the action @xmath69 was not calculated there .
the one - loop preexponential factor of the statistical sum ( [ z ] ) was obtained in @xcite for a one - fold instaton background . for a theory with the generic effective action ( [ action1 ] ) it reads as @xmath70 where @xmath71 is
a kind of a specific heat of the thermal bath of matter quanta and @xmath72 is a special functional of the background solution @xmath25 .
as shown in @xcite this quantity also generates the functional determinant @xmath73 of the operator @xmath74 of small disturbances of the canonically normalized mode of the scale factor perturbation @xmath75 propagating on this background implies gauging out the zero mode of this operator , originating from the residual conformal killing symmetry of the action for @xmath76 @xcite .
] , @xmath77 this operator is expressed via the function @xmath78 which simultaneously serves as its zero mode , @xmath79 , and in its turn expresses in terms of the time derivative of the scale factor @xmath80 and the hessian of the lagrangian @xmath30 . thus all the quantities are functionals of this zero mode which itself is easily calculable by eq.([g ] ) for a given instanton background . the quantity @xmath81 expresses in quadratures as a functional of @xmath82 as follows .
for a one - fold instanton the function @xmath83 has one oscillation in the full range of the euclidean time forming a circle .
therefore , it has two zeroes at the antipodal points of this circle @xmath84 corresponding to a maximum and minimum values of the scale factor .
these points mark the boundaries of the half period of the total time range , @xmath85 , at which the periodic function @xmath82 has these two first degree zeros @xmath86 then @xmath81 is given by a combination @xmath87 of boundary values at these points of a special ( non - periodic ) solution @xmath88 of the homogeneous equation @xmath89 with the same operator ( [ operator ] ) .
this solution is defined by the equation @xmath90 where @xmath91 is some fixed point belonging to this half period range of @xmath4 .
the main property of the function @xmath88 is that it is smoothly defined in the half - period range of @xmath4 and @xmath91 , where the integral ( [ psi ] ) is convergent because the roots of @xmath82 do not occur in the integration range of ( [ psi ] ) .
it can not be smoothly continued beyond this half - period , though its limits are well defined for @xmath92 , @xmath93 because the factor @xmath82 tending to zero compensates for the divergence of the integral at @xmath94 . moreover , because of @xmath95 the function @xmath88 is differentiable at @xmath94 , and all the quantities which enter the algorithm ( [ bfi ] ) are well defined .
these properties of @xmath96 guarantee that @xmath72 is independent of an arbitrary choice of the point @xmath91 , which can be easily verified by using a simple relation @xmath97 .
the structure of ( [ prefactor0 ] ) suggests that the effect of the one - loop prefactor can be strong , because the argument of the square root can tend to zero .
though the factor @xmath98 which is proportional to the negative of the specific heat of the thermal bath , @xmath99 is always negative , the sign of @xmath100 is indefinite .
indeed , this quantity can be rewritten as the action functional of the mode @xmath88 in the half period of the euclidean time range @xmath101 its sign depends on the sign of the hessian matrix in ( [ bfi ] ) ( ghost or non - ghost nature of scale factor perturbation @xmath76 in the gravitational action ) . moreover ,
the integral here is also indefinite , because the potential term is not positive - definite .
in fact it is basically negative ( @xmath76 is a tachyon ) , because for an oscillating @xmath102 this function is quasi - harmonic with @xmath103 in the average over the oscillation period . in particular , for a constant @xmath104 , the function @xmath82 is harmonically oscillating , the function @xmath88 is exactly calculable and @xmath100 identically vanishes the case of the second zero mode of @xmath105 complimentary to @xmath82 @xcite , @xmath106 therefore , a nontrivial expression for the prefactor is based on the anharmonicity of @xmath82 and should be analyzed numerically or by some kind of perturbation theory .
now we apply the above technique to the model of the cft driven cosmology described in introduction .
as was shown in @xcite , all saddle - point instantons of this model exist only inside the curvilinear wedge in the two - dimensional plane of the variables @xmath107 primordial cosmological constant @xmath32 and the amount of radiation constant @xmath40 , depicted in fig.1 .
this wedge is bounded by the lower straight - line and upper hyperbolic boundaries , @xmath108 the one - parameter families of k - fold instantons form in this plane the curves interpolating between these two boundaries , and their sequence accumulates at the critical point corresponding to the quantum gravity scale @xmath109 .
each instanton solution is represented by a point on one of these curves with the corresponding @xmath46 the number of oscillations of the scale factor @xmath60 between its minimal and maximal values .
the range of @xmath22 in which these @xmath46-fold instantons exist is given by a projection of their family on the axes of the variable @xmath52 .
here we consider only the one - fold family .
analytic estimates of the prefactor are hard to make even in the limit of a large number of quantum fields @xmath110 and a large value of @xmath38 , because as we will now see this one - loop factor is @xmath111 and is not sensitive to this limit . a natural quantity that could have played
the role of a smallness parameter is the combination @xmath112 which belongs to the bounded range @xmath113 for all these instantons .
as was shown in @xcite for multi - fold instantons with @xmath114 this quantity is indeed small @xmath115 and corresponds to solutions tending to the corner of the curvilinear triangle @xmath116 the point of a new quantum gravity scale @xmath117 .
unfortunately , for the case of @xmath51 instanton which only we consider here @xmath118 and the corresponding prefactor should be calculated numerically . to make this calculation manageable
let us first parameterize the instanton solutions in a more convenient way in terms of the quantity ( [ varepsilon ] ) . from the modified friedmann equation ( [ efeq ] )
we know that the scale factor is a function of time oscillating between the two turning points @xmath119 . in terms of a dimensionless variable @xmath120 this behavior can be expressed as a solution of ( [ efeq ] ) for the time derivative of @xmath10 @xmath121 where the turning points @xmath122 and @xmath123 read in terms of @xmath38 , @xmath40 and @xmath124 as @xmath125 in terms of @xmath124 the curvilinear wedge ( [ wedge ] ) where the instantons are located looks as @xmath126 inside this domain the one - parameter family of solutions interpolating between its boundaries can be labeled by the parameter @xmath127 ranging from @xmath128 at the upper boundary to @xmath129 at the lower boundary @xmath130 together with the definition ( [ varepsilon ] ) this relation is just a relabeling of the solutions from the dimensional variables @xmath131 to the dimensionless @xmath132 having a simple finite range inside a unit quadrangle . in terms of @xmath127
the turning points of the solution @xmath122 and the parameter @xmath133 introduced above take the form @xmath134 with @xmath135 parameterized between the two turning points by the new variable @xmath136 @xmath137 the equation for @xmath80 can be written down as @xmath138^{1/2}+1-d+2d\,x } , \end{aligned}\ ] ] so that the conformal time period of the instanton equals @xmath139^{1/2 } + 1-d+2d\,x\right)^{1/2 } } { ( 1-\varepsilon d+2\varepsilon^2 d x)}. \label{eta10 } \end{aligned}\ ] ] with a very good accuracy for all @xmath124 and @xmath127 the integral here equals @xmath140 , so that @xmath141 . on the other hand ,
the bootstrap equation ( [ bootstrap ] ) for @xmath124 in view of ( [ cvsepsilon ] ) reads @xmath142 in view of the fact that both @xmath143 and @xmath144 for a large number of species @xmath145 , the ratio on the right hand side here is @xmath146 , and the parameter @xmath124 also remains @xmath111 and is not sensitive to the limit of large @xmath145 .
then one can easily check that in the limits of either small , @xmath147 , @xmath148 , or relatively large , @xmath149 , @xmath150 , values of @xmath43 this equation yields the value of @xmath124 close to the middle of its admissible range @xmath151 , @xmath152 .
this confirms that for a one - fold garland instanton @xmath124 never gets either very small or close to one , and numerical simulations become inevitable for the calculation of the prefactor ( [ prefactor0 ] ) with its basic quantity ( [ bfi ] ) .
fortunately , this leads to a simple qualitative picture .
the function ( [ g ] ) for the lagrangian of ( [ effaction ] ) equals in terms of the variable @xmath135 , ( [ z ] ) , and the running integration parameter @xmath136 of eq.([x ] ) @xmath153^{1/2}. \end{aligned}\ ] ] therefore the function ( [ psi ] ) is given by the following integral @xmath154^{1/2 } + 1-d+2d\,y\right)^{3/2 } } { \big[\,\big(1-d\big)^2 + 4d\big(1-\varepsilon\,d\big)\,y + 4\varepsilon\,d^2 y^2\,\big]^{1/2 } } , \end{aligned}\ ] ] where the integration limits equal @xmath155 in the limit @xmath94 ( equivalent to @xmath156 and @xmath157 ) the integral in ( [ gint ] ) is divergent , this divergence being compensated by the factor @xmath82 vanishing at these points . for the calculation of ( [ bfi ] ) we have to take these limits , and their calculation is based on a preliminary integration by parts in @xmath158 which disentangles the divergent part of the integral , @xmath159 at @xmath156 or @xmath160 at @xmath157 . then multiplying the result by @xmath82 and differentiating we get in view of ( [ psipm ] ) ( we take into account that for the cft driven cosmology @xmath161 ) @xmath162 this expression is , of course , independent of the choice of the intermediate point @xmath163 , as is easily seen by differentiation with respect to @xmath163 .
for @xmath164 the function @xmath165 , so that @xmath72 is finite in this limit even despite @xmath166 in the denominator of ( [ bfifinal ] ) .
note that the limit @xmath164 corresponds to the upper hyperbolic boundary of the instanton domain ( [ wedge ] ) on which the cosmological model is static , @xmath167 , and @xmath168 is constant .
so this is the case of the harmonic function @xmath82 in eq.([harmonic ] ) for which formally @xmath169 the case of the second zero mode of the operator @xmath170 , discussed in @xcite .
however , in this limit not only the deviation from harmonicity @xmath171 disappears , but also the amplitude of the function @xmath172 tends to zero , cf .
eq.([g1 ] ) .
therefore , in view of ( [ harmonic ] ) the function @xmath173 acquires a @xmath174 factor , and does not vanish for @xmath164 .
thus , quantum corrections to the preexponential factor are nontrivial even in this limit .
as we will see below , they are smaller than those of @xmath175 but still nonzero . numerical calculation of ( [ g ] ) can be approximately done in the full range of @xmath127 .
first we find from the bootstrap equation ( [ bootstrap1 ] ) the value of @xmath124 as a function of @xmath127 on the relevant one - parameter family of instantons @xmath176 , @xmath177 , and then evaluate with the aid of ( [ bfifinal ] ) , ( [ g ] ) and ( [ specificheat ] ) the preexponential factor on this family @xmath178 we will explicitly demonstrate this procedure on the example of a single gauge vector field . for this field the trace anomaly coefficient @xmath38 and the sum ( [ energy ] ) over its oscillator frequencies @xmath179 , @xmath180 , read as @xmath181 with these expressions and the conformal time @xmath43 given by the integral ( [ eta10 ] ) the numerical solution of the bootstrap equation ( [ bootstrap1 ] ) can be sufficiently accurately fitted by the following function @xmath182 which parameterizes the one - fold family of instantons
. then we compute @xmath72 via ( [ bfifinal])-([g ] ) , which turns out to be negative on this family , and also evaluate the specific heat for the free energy of vector particles on a 3-sphere @xmath183 the use of ( [ p(d ) ] ) for this family of solutions finally yields the plot of @xmath184 depicted on fig.2 .
a characteristic feature of this vector field case is that with @xmath185 and negative @xmath72 the denominator of @xmath184 is smaller than one , but never degenerates to zero .
therefore , the one - loop prefactor stays slightly higher than one , but remains finite and within the perturbation theory range .
similar calculations can be done for pure massless spinor and conformal scalar fields .
the prefactor for these fields remains nearly constant throughout the whole family of one - fold instantons .
for a spinor field the behavior of @xmath186 is qualitatively the same as in the vector field case , whereas for a scalar field the calculations show that @xmath187 and the prefactor turns out to be smaller than one , @xmath188 .
possible mixtures of various spins do not qualitatively change this picture .
the one - loop prefactor of the statistical sum in the cft driven cosmology was obtained for a limited set of saddle - point solutions having one oscillation of the cosmological scale factor .
this limitation is currently explained by the fact that the major ingredient of the calculational algorithm the restricted functional determinant of the operator @xmath170 is known in a closed form ( [ detfstar ] ) only for this simplest set of instantons @xcite .
the resulting preexponetial factor does not present us with new physics , because for all scalar , spinor and vector conformal particles or their combinations the prefactor remains a smooth function @xmath189 .
it justifies the semiclassical expansion , but does not feature phase transitions or strong coupling range beyond perturbation theory restrictions even in the limit of numerous conformal species @xmath190 . at the same time path integral applications in such conformal cosmology suggest the existence of instantons with an arbitrary number of oscillations of the cosmological scale factor .
in particular , for the number of these oscillations tending to infinity this cft driven cosmology approaches a new quantum gravity scale @xmath191 the maximum possible value of the cosmological constant @xcite where the physics and , in particular , the effects of the quantum prefactor become very interesting and important .
thus the extension of the above results to the full set of cosmological instantons looks very promising , as this extension might be relevant to the resolution of the cosmological constant and landscape problems @xcite .
we hope to attain this extension in the foreseeable future .
main difficulty with all known applications of quantum gravity and cosmology is that the lack of uv complete theory forces us to work within the effective field theory trustable only below its cutoff . in this respect
large @xmath145 cft driven cosmology of @xcite seems to be a safe theory because its instanton curvature scale is bounded by @xmath192 which is much lower than the planck scale .
therefore , no non - perturbative extension ( like the hypothetical asymptotic safety of quantum gravity @xcite ) is likely to be needed , and a usual semiclassical expansion is sufficient to handle the problem of cosmological initial conditions . however , as persuasively advocated in @xcite , the cutoff in theories with a large number of quantum species , @xmath193 , also decreases with the growing @xmath145 and the instantons considered above turn out to be very close to this cutoff scale .
this makes important the computation of loop corrections initiated above .
finiteness of the one - loop order and its regularity @xmath189 in the full one - fold instanton range of @xmath22 gives a hope that this property will survive beyond this range and beyond one - loop approximation , and this is a subject of our further studies @xcite .
i am grateful to a.yu.kamenshchik for helpful discussions .
also i wish to express my gratitude to g.dvali for hospitality at the physics department of the ludwig - maximilians university in munich where this work was supported in part by the humboldt foundation .
this work was also supported by the rfbr grant no 11 - 02 - 00512 .
s. weinberg , _ critical phenomena for field theorists _ , in _ understanding the fundamental constituents of matter _
a. zichichi ( plenum press , new york , 1977 ) , phys . rev .
* d 81 * ( 2010 ) 083535 , arxiv:0911.3165 [ hep - th ] ; c. wetterich , phys .
* b 301 * ( 1993 ) 90 . | the path integration technique recently developed for the statistical sum of the microcanonical ensemble in cosmology is applied to the calculation of the one - loop preexponential factor in the cosmological model generated by a conformal field theory with a large number of quantum species the model of initial conditions possibly related to the resolution of the cosmological constant and landscape problems .
the result is obtained for the family of background cosmological instantons with one oscillation of the frw scale factor .
the magnitude of the prefactor is analytically and numerically estimated for fields of various spins conformally coupled to gravity , which justifies the validity of semiclassical expansion for this family of cosmological instantons . |
there has been intense research interest in phase transitions in mass - transport and growth models involving adsorption and desorption , fragmentation , diffusion and aggregation .
these processes are ubiquitous in nature and arise in a large number of seemingly diverse systems such as growing interfaces , colloidal suspensions @xcite , polymer gels @xcite , river networks @xcite , granular materials @xcite , traffic flows @xcite , etc . in these systems ,
different nonequilibrium states arise if the rates of the underlying microscopic processes are varied .
conservation laws also play an important role in determining both the time - dependent behavior and steady states of such systems . as these steady states
are usually not described by the gibbs distribution , they are hard to determine .
however , much insight on this issue has been gained by studying lattice models . due to their simplistic nature
, these models can be treated either exactly or via a mean - field ( mf ) approach @xcite .
they are also simple to implement numerically using monte carlo ( mc ) techniques @xcite . in this paper ,
we present a pedagogical discussion of the modeling and simulation of mass - transport and growth phenomena .
we discuss analytical and numerical techniques in the context of mass - transport models where the elementary move is the fragmentation of mass @xmath0 , and its subsequent diffusion to a neighboring site where it aggregates .
the @xmath0-chip models that we study here are interesting in physical situations where the deposited material consists of polymers .
we study the mf limit of these models , focusing on the steady - state mass distribution [ @xmath1 vs. @xmath2 , which is characterized by @xmath0 branches .
we also compare the mf results with mc simulations in @xmath3 @xmath4 @xmath5 .
this paper is organized as follows . in sec .
[ s2 ] , we present a framework for mass - transport models in terms of the rate ( evolution ) equations for @xmath6 , the probability that a site has mass @xmath7 at time @xmath8 .
we then discuss various systems which can be described within this framework . in sec .
[ s3 ] , we introduce @xmath0-chip models and obtain analytical results for the mf versions of these models . in sec .
[ s4 ] , we present mc results for mass - transport models , and compare them with the corresponding mf solutions .
we conclude this paper with a summary and discussion in sec .
the appendices contain details of calculations and mc procedures .
we consider lattice models of mass transport with the processes of fragmentation , diffusion and aggregation . for simplicity
, we describe the models on a 1-dimensional lattice with periodic boundary conditions .
( the generalization to higher dimensions is straightforward . ) to begin with , masses are placed randomly at each site with an overall mass density @xmath9 .
let @xmath10 be the mass at site @xmath11 at time @xmath8 .
the mass variables assume discrete values 0,1,2,3 , etc .
the evolution of the system is as follows .
a piece of mass @xmath12 chips off a site having mass @xmath13 with rate @xmath14 .
this piece deposits on the right neighbor with probability @xmath15 , or on the left neighbor with probability @xmath16 .
the mass of the chosen neighbor adds up , while that of the departure site decreases , with the total mass of the system remaining conserved .
figure 1 is a schematic depiction of the above model . to facilitate mc simulations ,
the update rules can be rewritten as follows : + 1 ) randomly pick a site @xmath11 at time @xmath8 with mass @xmath17 .
the site is updated as @xmath18 with rate @xmath14 .
+ 2 ) the neighboring sites are updated as @xmath19 with probability @xmath15 , or @xmath20 with probability @xmath16 .
we also study the above models within a mf approximation which keeps track of the distribution of masses , ignoring correlations in the occupancy of adjacent sites .
although the mf theory has this shortcoming , our mc simulations show that it gives an accurate description of the above model , even in the 1-dimensional case .
let @xmath6 denote the probability that a site has mass @xmath7 at time @xmath8 . in the mf limit
, @xmath6 evolves as follows : @xmath21 @xmath22 these equations enumerate all possible ways in which a site with mass @xmath7 may change its mass . the first term on the right - hand - side ( rhs ) of eq .
( [ rate1 ] ) is the `` loss '' of mass @xmath7 due to chipping , i.e. , a site with mass @xmath7 may lose a fragment of mass @xmath23 ( @xmath24 ) to a neighbor . the second term on the rhs represents the loss due to transfer of mass from a neighbor chipping .
the third and fourth terms are the `` gain '' terms which represent the ways in which a site with mass greater ( lesser ) than @xmath7 can lose ( gain ) the excess ( deficit ) to yield mass @xmath7 .
the terms of eq .
( [ rate10 ] ) can be interpreted similarly . in order to ensure that all loss and gain terms have been included in the rate equations ,
it is useful to check the sum rule @xmath25 with some algebra , it can be shown that eqs .
( [ rate1])-([rate10 ] ) do indeed satisfy the above rule .
we provide the steps for this check in appendix a. if other microscopic processes are present , additional terms will have to be included in the rate equations ( [ rate1 ] ) and ( [ rate10 ] ) .
for example , if adsorption of a unit mass at a site occurs with rate @xmath26 , we require additional terms @xmath27 and @xmath28 in eq .
( [ rate1 ] ) , and @xmath29 in eq .
( [ rate10 ] ) . while it is simple to write realistic rate equations by including all relevant microscopic processes , it is often difficult to solve them analytically to obtain either time - dependent or steady - state solutions .
several mf models studied earlier may be obtained as special cases of eqs .
( [ rate1])-([rate10 ] ) by an appropriate choice of the chipping kernel @xmath14 .
some interesting issues which have been addressed in these studies , in addition to obtaining steady - state mass distributions , are the possibilities of phase transitions in these models .
we mention two representative examples to highlight the typical questions which are addressed in this area .
+ 1 ) majumdar et al .
@xcite studied a conserved - mass model in which either a single unit or the entire mass could dissociate from a site .
thus , @xmath14 has the form @xmath30 where @xmath31 is the relative rate of the 1-chip process .
the corresponding rate equations are obtained by substituting eq .
( [ ssm ] ) in eqs .
( [ rate1])-([rate10 ] ) as follows : @xmath32 where @xmath33 . the steady - state mass distributions [ @xmath1 vs. @xmath2 for eqs .
( [ snm1])-([snm0 ] ) were calculated by majumdar et al . as a function of the density @xmath34 of the system .
the relevant analytical techniques are described in sec .
they observed a _ dynamical phase transition _ as @xmath9 was varied ( @xmath31 being fixed ) , with the different phases being characterized by different steady - state distributions . for @xmath35 , @xmath1 decayed exponentially for large @xmath7 .
for @xmath36 , @xmath1 showed a power - law decay , @xmath37 with a universal exponent @xmath38 .
finally , the high - density " phase arising for @xmath39 was characterized by the formation of an infinite aggregate ( at @xmath40 ) .
the aggregate coexisted with smaller clusters , and their mass distribution showed a power - law decay , @xmath37 .
+ 2 ) rajesh et al .
@xcite studied a system of fragmenting and coagulating particles with mass - dependent diffusion rates . in this model ,
@xmath41 the case @xmath42 = 0 corresponds to the model in eq .
( [ ssm ] ) .
the corresponding rate equations are @xmath43 where @xmath44 .
for @xmath45 , rajesh et al .
showed that there is no dynamical phase transition .
the high - density phase with an infinite aggregate disappears , although its imprint in the form of a large aggregate is observed in finite systems .
further , the steady - state probability distribution @xmath1 decays exponentially with @xmath7 for all @xmath9 and @xmath46 . before concluding this discussion , a few words regarding the condensation transition are in order .
the condensation observed in the model in eqs .
( [ snm1])-([snm0 ] ) occurs due to the dynamical rules of evolution and not due to an `` attraction '' between the masses .
though it shares analogies with _ bose - einstein condensation _ ( bec ) , an important difference is that these condensates are formed in real space rather than momentum space as is the case in bec . as a matter of fact
, condensation occurs in a variety of seemingly diverse systems which are governed by nonequilibrium dynamics @xcite .
we now discuss some physical applications of mass - transport models .
the aim here is to stress the general nature of the questions addressed in a variety of physical situations involving mass transport .
recently , there has been much research interest in suspensions of single - domain _ magnetic nanoparticles _ ( mnp ) , which have a wide range of technological applications , e.g. , memory devices , magnetic resonance imaging , targeted drug delivery , bio - markers and bio - sensors @xcite . a major reason for the utility of mnps
is the ease with which they can be detected and manipulated by an external magnetic field .
their response times are strongly size - dependent , thus introducing the possibility of controlling particle sizes to obtain desired response times .
an inherent property of mnp suspensions is cluster formation , due to the presence of attractive interactions of varying strengths between the constituent particles @xcite .
therefore , mass - transport models with fragmentation and aggregation have been traditionally employed to study clustering dynamics in these systems .
the steady - state cluster - size distributions and the average cluster size are determined by the interplay between aggregation ( due to attractive interactions ) and fragmentation ( due to repulsive interactions and thermal noise ) @xcite .
assuming that the number of particles is @xmath47 , and denoting the number of clusters containing @xmath0 particles at time @xmath8 by @xmath48 , the rate equations in the mf approximation are as follows @xcite : @xmath49 in eq .
( [ reqn ] ) , @xmath50 and @xmath51 are the aggregation and fragmentation kernels , respectively .
the aggregation kernel describes the coalescence of two clusters containing @xmath11 and @xmath52 particles to yield a larger cluster with @xmath0 @xmath4 @xmath53 particles . in many models
, it is assumed to have a mass - dependent form , @xmath50 @xmath4 @xmath54 .
this accounts for the reduced mobility of large clusters .
the fragmentation kernel @xmath51 describes the loss of one particle from a cluster with @xmath0 particles , and has also been assumed to have a mass - dependent form , @xmath55 .
equation ( [ reqn ] ) can be rewritten in terms of probability distributions [ cf .
( [ rate1])-([rate10 ] ) ] by introducing a normalization factor , @xmath56 .
our second example is in the context of traffic models . in this context
, we discuss the so - called _ bus route model _
( brm ) @xcite . here
, one is interested in the initial conditions or parameters which result in a clustering of buses or a traffic jam .
the model is defined on a 1-dimensional lattice of size @xmath57 .
each site @xmath11 has two associated variables @xmath58 and @xmath59 : ( i ) if a site @xmath11 is occupied by a bus , @xmath58 @xmath4 1 ; otherwise @xmath58 @xmath4 0 .
( ii ) if a site @xmath11 has passengers , then @xmath59 @xmath4 1 ; otherwise @xmath59 @xmath4 0 .
a site can not have both @xmath58 @xmath4 @xmath59 @xmath4 1 , i.e. , @xmath60 .
if there are @xmath61 buses , the bus density @xmath9 @xmath4 @xmath62 is a conserved quantity .
however , the total number of sites with passengers is not conserved .
the update rules are as follows : ( i ) pick a site @xmath11 at random .
( ii ) if @xmath58 @xmath4 @xmath63 0 , then set @xmath59 @xmath4 1 with rate @xmath64 , i.e. , a passenger arrives at an empty site with rate @xmath64 .
( iii ) if @xmath58 @xmath4 1 and @xmath65 @xmath4 0 , a bus hops onto a site with no passengers ( @xmath66 0 ) with rate @xmath42 , and to a site with passengers ( @xmath67 @xmath4 1 ) with rate @xmath68 .
thus , the variables @xmath69 0 and @xmath70 1 and @xmath71 0 with rate @xmath42 or @xmath68 , as the case may be . usually , @xmath72 as the buses slow down when passengers are being picked up .
a jam in the system is a gap between buses of size @xmath73 , which is stable in the thermodynamic limit .
the mf approximation of this model considers the distribution of gaps @xmath74 , ignoring the time - correlations in the hopping of buses .
it should be noted here that , unlike the mass - transport models described in sec .
[ s2a ] , the brm is asymmetric .
thus , the movement of the buses is unidirectional although the hop rate is proportional to the size of the gap .
these features put the brm in a class of models which are referred to as _ zero - range processes _ ( zrp ) @xcite .
the important property of a zrp is that it yields a steady - state as a product of marginals calculated using well - defined procedures @xcite .
further , mf calculations are exact for this class of models . from simulations of the discrete model , heuristic arguments and mf
theory , oloan et al .
obtain evidence of a jamming transition as a function of the density of buses @xmath9 . in terms of buses and passengers ,
the jam may be interpreted as follows .
an ideal situation is one where the buses are evenly distributed along the route so that each bus picks up approximately the same number of passengers .
jamming or clustering of buses may occur if one of the buses gets delayed due to some fluctuation at a pick - up point .
subsequently , the buses which follow catch up with the delayed bus resulting in a jam !
an important observation here is that the jamming is a consequence of a local , stochastic dynamics which couples the conserved variable ( buses ) and the nonconserved variable ( passengers ) .
the transition is reminiscent of the condensation transition described earlier @xcite , and has also been useful in describing _ clogging _ in the transport of sticky particles down a pipe @xcite . as a final example , we consider the packing of granular materials , which is important in many technological processes .
the crucial issue in these problems is understanding the complex network of forces which is responsible for the static structure and properties of granular materials .
one such system which has been subjected to experiments , simulations and analysis is a pack of spherical beads in a compression cell @xcite .
the bead pack is modeled as a regular lattice of sites , each having a particle of unit mass .
the mechanisms which lead to the formation of force chains in the system are summarized in the rules defined below : ( i ) each site @xmath11 in layer @xmath75 is connected to @xmath47 sites @xmath52 in layer @xmath76 .
( ii ) only vertical forces are considered explicitly .
a fraction @xmath77 of the total weight supported by particle @xmath11 in layer @xmath75 is transmitted to particle @xmath52 in layer @xmath76 .
thus , the weight @xmath78 supported by the particle at the @xmath79 site in layer @xmath75 satisfies the stochastic equation @xmath80 the @xmath81 are independently - distributed random variables which satisfy the constraint @xmath82 @xmath4 1 , required for enforcing the force - balance condition on each particle . in general
, the values of @xmath83 at neighboring sites in layer @xmath75 are not independent .
the mf approximation of this model ignores these correlations . defining a normalized weight variable @xmath84
, we want to obtain the force distribution @xmath85 , i.e. , the probability that a site at depth @xmath75 is subject to a vertical force @xmath86 . within the mf approximation , it is possible to obtain a recursive equation for @xmath85 .
coppersmith et al .
@xcite found that , for almost all distributions of @xmath26 , the distribution of forces decays exponentially .
however , a power - law decay was also observed in some cases .
let us first consider the 1-chip model .
the chipping kernel has the simple form @xmath87 with the above kernel , eqs .
( [ rate1])-([rate10 ] ) become @xmath88 absorbing @xmath31 into the definition of time , these equations simplify to the following form : @xmath89 here , we have defined @xmath90 as the probability of occupancy of a site with mass @xmath91 .
consequently , the probability of a site being empty is @xmath92 .
the above rate equations were obtained earlier in refs .
@xcite and were solved exactly .
we recall this calculation to illustrate the _ generating - function approach _ for obtaining steady - state solutions of such rate equations . defining the generating function @xmath93 @xmath4 @xmath94 , an equation for @xmath95
can be obtained from eq .
( [ rate3 ] ) by multiplying both sides by @xmath96 and summing over @xmath7 : @xmath97+s_{1}z\left[q+p(0,t)\right].\end{aligned}\ ] ] setting @xmath98 , and substituting @xmath99 from the steady - state version of eq .
( [ rate30 ] ) , we obtain @xmath100 the value of @xmath101 is fixed by mass conservation , which requires that @xmath102 = @xmath9 , where @xmath9 is the mass density . putting @xmath103 = @xmath9
, we obtain @xmath104 the steady - state distribution @xmath1 is the coefficient of @xmath96 in @xmath105 , and can be obtained by taylor - expanding @xmath105 about @xmath106 .
this yields @xmath107 for a more complicated function @xmath105 , we can obtain @xmath1 by inverting @xmath105 .
it is useful to illustrate this for the simple form of @xmath105 in eq .
( [ gen1 ] ) .
thus , @xmath108 here , the closed contour @xmath109 encircles the origin in the complex plane counter - clockwise and lies inside the circle @xmath110 .
the integral is calculated using the residue theorem . only those singular points which lie within @xmath109 ( viz .
, @xmath106 , which is a pole of order @xmath7 ) contribute to this evaluation .
the associated residue is @xmath111 thus , the steady - state mass distribution is @xmath112 notice that @xmath113 , so eq . ( [ soln1 ] ) is also valid for @xmath114 .
using eq .
( [ rho1 ] ) , the above mass distribution can be rewritten as @xmath115 where @xmath116 in the case of simple chipping kernels , as in eq .
( [ kernel1 ] ) , the above solution can also be obtained directly from the difference equations ( [ rate3])-([rate30 ] ) by setting the left - hand - side ( lhs ) to zero .
we can then write down expressions for the first few terms of @xmath1 : @xmath117 which is identical to eq .
( [ soln1 ] ) .
again , the mass conservation condition @xmath102 = @xmath9 results in eq .
( [ rho1 ] ) , as expected .
the 1-chip solution in eq .
( [ soln11 ] ) is important because of its universal nature . as a matter of fact , it is a steady - state solution for all mf models where the chipping kernel @xmath14 is independent of the mass of the departure site @xmath7 , @xmath118 . to confirm this , we consider eqs .
( [ rate1])-([rate10 ] ) with @xmath14 replaced by @xmath119 .
the corresponding rate equations are @xmath120 @xmath121 in the steady state , the above equations may be combined to obtain @xmath122 substituting @xmath123 on the rhs of eq .
( [ arbk1ss ] ) , we obtain @xmath124 the first and fourth terms cancel , and the second and third terms cancel , so @xmath125 .
this confirms that @xmath126 is a solution of eqs .
( [ arbk1])-([arbk10 ] ) .
the constants @xmath127 and @xmath128 are fixed by the requirements of probability normalization @xmath129 $ ] and mass conservation [ @xmath130 .
we leave it as an exercise to the reader to verify that these conditions lead to the same values of @xmath127 and @xmath128 as in eq .
( [ a1 ] ) .
next , let us generalize the @xmath131-chip model to a @xmath0-chip model , where @xmath132 .
these @xmath0-chip models are interesting in physical situations where the deposited material consists of polymers or aggregates .
we will see that the steady - state solutions for @xmath0-chip models exhibit a @xmath0-branch structure .
the steady - state distributions for the 2-chip model can be obtained using a procedure similar to the 1-chip model .
the corresponding form of @xmath133 is @xmath134 the rate equations in this case are ( absorbing @xmath31 into time @xmath8 ) @xmath135 here , @xmath136
= @xmath137 is the probability of sites having mass 2 or more .
notice that the kernel in eq .
( [ kernel2 ] ) is independent of the mass of the departure site .
thus , the 1-chip solution is a steady - state solution of eqs .
( [ rate2])-([rate20 ] ) , as can be verified by direct substitution . however , an arbitrary initial condition @xmath138 will not relax to this solution due to the presence of conserved quantities , as we shall see shortly .
the steady - state generating function @xmath105 , obtained in analogy with the @xmath131-chip case , is as follows : @xmath139 where @xmath140 = @xmath141 .
it is straightforward to taylor - expand @xmath105 in eq .
( [ gen2 ] ) and identify @xmath1 .
alternatively , we can obtain @xmath1 using eq .
( [ steady1 ] ) .
thus @xmath142 in this case , the contour @xmath109 encircles the origin counterclockwise , and lies inside the circle @xmath143 = @xmath144 .
the singularities of the integrand @xmath145 in the above equation are @xmath106 ( pole of order @xmath7 ) , @xmath146 ( simple pole ) and @xmath147 = @xmath148 ( simple pole ) .
the second and third poles lie outside @xmath109 , making res @xmath149 the only contributing residue .
this evaluation yields @xmath150s_{2}^{m/2 } + \frac{(s_{1}-s_{2})}{2}\left[1-(-1)^{m}\right]s_{2}^{(m-1)/2}.\end{aligned}\ ] ] thus , the steady - state probability distribution for the 2-chip model is given by @xmath151 here the function @xmath152 is defined as the remainder on division of @xmath7 by @xmath12 .
the first and second terms on the rhs of eq .
( [ steady2 ] ) are the steady - state distributions for even values of @xmath7 @xmath153 $ ] and odd values of @xmath7 @xmath154 $ ] , respectively .
thus , the 2-chip model has a steady - state solution comprising of two branches , both of which have the same exponential decay .
notice that the occupation probabilities for sites with even or odd units of mass are @xmath155 these quantities remain conserved during the evolution because of the nature of the 2-chip move .
the two branches appear as a consequence of these two conserved quantities .
the probabilities of occupancy @xmath101 and @xmath140 are related to the mass density @xmath9 as follows : @xmath156 as before , @xmath9 may be calculated from either @xmath105 or @xmath1 .
the quantities @xmath101 and @xmath140 can be determined in terms of @xmath9 and @xmath157 ( or @xmath158 ) : @xmath159 it should be noted that @xmath160 is always greater than 1 , ensuring @xmath161 . if we choose the initial conditions @xmath138 such that @xmath162 in eq .
( [ s1 ] ) , the branched solution of eq .
( [ steady2 ] ) reduces to the 1-chip exponential solution . alternatively ,
if we substitute @xmath162 in eq .
( [ gen2 ] ) , we recover the generating function of the 1-chip model in eq .
( [ gen1 ] ) . in general , consider the case of @xmath0 units of mass chipping from a site and then aggregating with the mass of a randomly - chosen nearest neighbor : @xmath163 the corresponding rate equations for @xmath6 , obtained by substituting eq .
( [ kernelk ] ) in eqs .
( [ rate1])-([rate10 ] ) , are as follows ( absorbing @xmath31 into @xmath8 ) : @xmath164 here , @xmath165 = @xmath166 is the probability of sites having mass @xmath0 or more . as the kernel in eq .
( [ kernelk ] ) is independent of the mass of the departure site , @xmath123 is a steady - state solution of eqs . ( [ ratek])-([ratek0 ] ) .
however , as in the 2-chip case , an arbitrary initial condition @xmath138 will not relax to this exponential solution due to the presence of @xmath0 conserved quantities : @xmath167 with @xmath12 @xmath4 @xmath168 rather , the steady - state solution will consist of @xmath0 branches .
the steady - state generating function for the @xmath0-chip model is @xmath169 where @xmath170 @xmath4 @xmath171 .
the corresponding probability distribution is as follows : @xmath172 for completeness , we present the derivation of eq .
( [ steadym ] ) for the @xmath173-chip case in appendix b. the @xmath174 s are related to the mass density via the relation @xmath175 thus , the steady - state solution for the @xmath0-chip model consists of @xmath0 branches .
all the branches in the probability distribution of eq .
( [ steadym ] ) decay exponentially with a slope @xmath176 .
the quantities @xmath177 are determined from eq .
( [ rhok ] ) plus the @xmath178 conserved probability sums in the branches : @xmath179 notice that one of the @xmath180 s ( say , @xmath181 ) is not independent because @xmath182 further , appropriately chosen initial conditions resulting in steady - state values @xmath162 , @xmath183, .... ,@xmath184 collapse the @xmath0 branches in eq .
( [ steadym ] ) to the 1-chip solution . before concluding this subsection ,
let us make an observation about the @xmath0-chip lattice model , i.e. , the original model rather than its rate equation " counterpart .
this model has an exponentially large set of disjoint sectors , and configurations in different sectors are not connected by the dynamics . to see this
, we denote the number of particles on a site @xmath11 as @xmath185 .
as @xmath0 particles arrive at or leave this site at a given time , the quantity @xmath186 is conserved .
therefore , the set @xmath187 is conserved by the dynamics , and labels a particular sector .
the number of sectors is @xmath188 , where @xmath47 is the number of lattice sites .
such systems have been referred to as _ many - sector decomposable systems _ by menon et al .
in this section , we present monte carlo ( mc ) results for some of the models discussed earlier .
all simulations were performed on 1-@xmath3 and 2-@xmath3 lattices with periodic boundary conditions .
the lattice sizes were @xmath57 @xmath4 @xmath189 ( in @xmath190 ) and @xmath191 @xmath4 @xmath192 ( in @xmath193 ) .
the data presented here was obtained as an average over @xmath194 independent runs .
the details of the mc procedure are provided in appendix c , so that the reader can implement these models numerically .
first , we present results for chipping kernels which satisfy @xmath195 , discussed at the end of sec .
[ s3.1 ] . in fig . 2 , we plot @xmath1 vs. @xmath7 obtained from 1-@xmath3 mc simulations with three different functional forms of @xmath14 : @xmath196 the mc data sets are numerically coincident with each other as well as the 1-chip solution in eq .
( [ soln11 ] ) ( denoted as a solid line ) , which was obtained from the corresponding mf equations .
the mass density in each case was @xmath197 .
there is excellent agreement between the different data sets , showing that the mc data is described very well by the solutions of the corresponding mf equations , even for @xmath190 .
this feature is also observed in our subsequent results , suggesting that the mf equations are exact in the present context @xcite .
next , we present results for the 2-chip model discussed in sec .
figure 3(a ) shows the steady - state distribution obtained from 1-@xmath3 and 2-@xmath3 mc simulations for initial conditions with @xmath198 , and @xmath199 .
the solid line denotes the result in eq .
( [ steady2 ] ) with values of @xmath101 and @xmath140 evaluated from eq .
( [ s1 ] ) . as our initial condition only had sites with even @xmath7 populated , the steady - state solution is the even-@xmath7 branch of eq .
( [ steady2 ] ) .
figure 3(b ) is similar , but for a mixed initial condition with @xmath200 and @xmath201 .
the branched nature of the solution , resulting in a staircase - type probability distribution , is highlighted in the inset . in fig .
4 , we show the steady - state distributions obtained from 1-@xmath3 and 2-@xmath3 mc simulations of the 3-chip model . in fig .
4(a ) , the initial condition had @xmath202 .
this corresponds to @xmath198 , and all three branches are equally populated . the solid line in fig .
4(a ) denotes the result in eq .
( [ steadym ] ) with @xmath203 and @xmath204 calculated from eq .
( [ 3chs ] ) .
figure 4(b ) is analogous to fig .
4(a ) , but the initial condition now has @xmath205 , @xmath206 , and @xmath207 .
the corresponding value of the average density is @xmath208 .
again , in both sets of figures , the mc simulations agree very well with the corresponding mf result .
let us conclude this paper with a summary and discussion .
there has been much research interest in mass - transport models , which arise in many physical contexts .
therefore , we believe it is appropriate to make this subject accessible to a wider audience . this is the underlying motivation for this paper . for the purposes of this exposition
, we focus on fragmentation - aggregation models with conserved mass , i.e. , there is no ongoing adsorption or desorption .
we consider models with the chipping rate @xmath133 , where @xmath12 is the chipped mass and @xmath7 is the mass of the departure site .
we use the corresponding mean - field ( mf ) equations to obtain the steady - state probability distributions [ @xmath1 vs. @xmath2 for different functional forms of @xmath133 .
we show that a large class of chipping kernels , where @xmath133 is independent of @xmath7 , give rise to an exponentially - decaying distribution : @xmath209 .
this is also the mf solution for the 1-chip model [ cf .
( [ soln11 ] ) ] , where one unit of mass fragments from a site and aggregates with a randomly - chosen nearest - neighbor .
we have also discussed @xmath0-chip models for fragmentation and aggregation .
the resulting steady - state distribution has @xmath0 branches , each of which decays exponentially with the same slope .
this slope is determined by the average density @xmath9 , and the population of the branches in the initial condition @xmath138 for the rate equations ( [ rate1])-([rate10 ] ) .
the initial population in each of the branches is conserved during the evolution , and is also reflected in the steady - state distribution . finally , we compared the mf analytical results with those from monte carlo ( mc ) simulations in @xmath210 . in all cases
, we found that the mf results for @xmath1 vs. @xmath7 were in excellent agreement with the mc results .
this demonstrates that the mf results are exact in the present context .
there are many open research problems in the area of mass transport , aggregation and growth .
these models can be tackled with a wide range of analytical and numerical techniques , which are both simple and elegant .
we hope that this review will motivate further studies of this fascinating area .
the authors would like to thank m. barma , s.n .
majumdar and r. rajesh for fruitful discussions .
gps and vb would like to acknowledge the support of csir grant no .
03(1077)/06/emr - ii .
sp is grateful to the department of science and technology , india for supporting this work through the project _ pattern formation in granular materials_.
we want to show that @xmath211/d t=0 $ ] . using eqs .
( [ rate1])-([rate10 ] ) , @xmath212&=&\frac{d}{dt}\left[\sum_{m=1}^{\infty}p(m , t)+p(0,t)\right]\nonumber\\ & = & -\sum_{m=1}^{\infty}p(m , t)\sum_{m_{1}=1}^{m}g_{m}(m_{1 } ) -\sum_{m=1}^{\infty}p(m , t)\sum_{m_{2}=1}^{\infty}p(m_{2},t)\sum_{m_{1}=1}^{m_{2}}g_{m_{2}}(m_{1})\nonumber\\ & & + \sum_{m=1}^{\infty}\sum_{m_{1}=1}^{\infty}p(m+m_{1},t)g_{m+m_{1}}(m_{1})\nonumber\\ & & + \sum_{m=1}^{\infty}\sum_{m_{1}=1}^{m}p(m - m_{1},t)\sum_{m_{2}=m_{1}}^{\infty}p(m_{2},t)g_{m_{2}}(m_{1})\nonumber\\ & & -p(0,t)\sum_{m_{2}=1}^{\infty}p(m_{2},t)\sum_{m_{1}=1}^{m_{2}}g_{m_{2}}(m_{1 } ) + \sum_{m_{1}=1}^{\infty}p(m_{1},t)g_{m_{1}}(m_{1}).\end{aligned}\ ] ] we regroup terms and write @xmath213\nonumber\\ & & + \left[\sum_{m=1}^{\infty}\sum_{m_{1}=1}^{\infty}p(m+m_{1},t)g_{m+m_{1}}(m_{1 } ) + \sum_{m_{1}=1}^{\infty}p(m_{1},t)g_{m_{1}}(m_{1})\right]\nonumber\\ & & + \sum_{m=1}^{\infty}\sum_{m_{1}=1}^{m}p(m - m_{1},t)\sum_{m_{2}=m_{1}}^{\infty}p(m_{2},t)g_{m_{2}}(m_{1})\\ \label{sum22 } & = & -\sum_{m=1}^{\infty}p(m , t)\sum_{m_{1}=1}^{m}g_{m}(m_{1 } ) -\sum_{m=0}^{\infty}p(m , t)\sum_{m_{2}=1}^{\infty}p(m_{2},t)\sum_{m_{1}=1}^{m_{2}}g_{m_{2}}(m_{1})\nonumber\\ & & + \sum_{m=0}^{\infty}\sum_{m_{1}=1}^{\infty}p(m+m_{1},t)g_{m+m_{1}}(m_{1})\nonumber\\ & & + \sum_{m=1}^{\infty}\sum_{m_{1}=1}^{m}p(m - m_{1},t)\sum_{m_{2}=m_{1}}^{\infty}p(m_{2},t)g_{m_{2}}(m_{1 } ) .\end{aligned}\ ] ] our initial condition for eqs .
( [ rate1])-([rate10 ] ) satisfies @xmath214 .
therefore , we set @xmath215 on the rhs of eq .
( [ sum22 ] ) .
this is justified subsequently as it results in @xmath211/d t=0 $ ] .
thus @xmath216 in the second term on the rhs of eq . ( [ sum4 ] )
, we interchange the order of the summations over @xmath23 and @xmath217 .
this leads to a cancellation of the two terms , proving that @xmath211/d t=0 $ ] .
the generating function for the 3-chip model is [ cf .
( [ genk ] ) ] @xmath218 where @xmath219 . from eq . ( [ steady1 ] )
, we have @xmath220 the closed contour @xmath109 encircles the origin and lies inside the region defined by @xmath221 . as usual , only res @xmath222 contributes to the integral , and is evaluated as @xmath223\bigg|_{z=0 } \nonumber\\ \label{res3c } & = & \frac{(1-s_{1})}{3}\left[1 + 2\cos\left(\frac{2\pi(m-3)}{3}\right)\right]s_{3}^{m/3}\nonumber\\ & & + \frac{(s_{1}-s_{2})}{3}\left[1 + 2\cos\left(\frac{2\pi(m-1)}{3}\right)\right]s_{3}^{(m-1)/3 } \nonumber\\ & & + \frac{(s_{2}-s_{3})}{3}\left[1 + 2\cos\left(\frac{2\pi(m-2)}{3}\right)\right]s_{3}^{(m-2)/3}.\end{aligned}\ ] ] then , the steady - state mass distribution is @xmath224s_{3}^{m/3 } \nonumber\\ & & + \frac{(s_{1}-s_{2})}{3}\left[1 + 2\cos\left(\frac{2\pi(m-1)}{3}\right)\right]s_{3}^{(m-1)/3 } \nonumber\\ & & + \frac{(s_{2}-s_{3})}{3}\left[1 + 2\cos\left(\frac{2\pi(m-2)}{3}\right)\right]s_{3}^{(m-2)/3 } \nonumber\\ \label{3csol3 } & = & ( 1-s_{1}){s_{3}}^{m/3}\delta_{{\rm mod}(m,3),0 } + ( s_{1}-s_{2}){s_{3}}^{\left(m-1\right)/3}\delta_{{\rm mod}(m,3),1}\nonumber\\ & & + ( s_{2}-s_{3}){s_{3}}^{\left(m-2\right)/3}\delta_{{\rm mod}(m,3),2}.\end{aligned}\ ] ] the relation between the mass density @xmath9 and the @xmath174 s can be obtained directly by using eq .
( [ 3csol3 ] ) .
thus @xmath225 as for the 2-chip model , the @xmath174 s can be obtained as a function of @xmath9 and the probability sums on each of the branches in eq .
( [ conserv ] ) .
these result in the following expressions : @xmath226
1 . initializing the lattice : integer masses @xmath228 are placed on the lattice sites in accordance with the chosen @xmath9 such that @xmath229 . a simple procedure for achieving this is as follows : 1 . choose an integer random number nran(@xmath11 ) from the range
@xmath230 $ ] for each site @xmath11 .
2 . assign @xmath231 $ ] , where @xmath232 refers to the integer part of @xmath233 .
2 . chipping and aggregation : a site @xmath11 is chosen at random . if @xmath228 is non - zero , a unit mass chips and aggregates with a randomly - chosen neighbor .
thus @xmath234 , and @xmath235 or @xmath236 with a probability 1/2 . 3 .
repeat step 2 for @xmath57 times , which corresponds to one mc step ( mcs ) .
4 . compute the mass distribution of lattice sites .
repeat steps 2 - 4 for several mcs , storing the mass distribution at intermediate mcs .
repeat steps 2 - 5 for a large number of independent lattice configurations , generated via step 1 .
7 . compute the configuration - averaged steady - state mass distribution , @xmath1 vs. @xmath7 . 1 .
define a matrix @xmath238 of dimension @xmath239 , chosen specific to @xmath14 .
for example , if @xmath240 , the rate of chipping a mass of size 1000 is @xmath241 .
we can assume that chipping of masses greater than 1000 occurs with a very small rate , and hence these events may be ignored .
thus we may set @xmath242 .
initialize @xmath243 to zero .
the row index @xmath7 corresponds to the mass of the lattice site chosen to fragment .
masses greater than @xmath47 may be treated as @xmath47 for the computation of fragmentation rates for reasons discussed in 1 above .
the column index refers to the chipped mass @xmath244 .
thus the matrix @xmath238 has a triangular form , with the non - zero entries calculated using @xmath14 .
for example , if @xmath240 , the rates for the @xmath245 row are [ 1 , 0.25 , 0.1111 , 0.0625 , 0.04 , 0.02777 , 0.0204 , 0.0156 , 0.0123 , 0.01 , 0 , 0 , ... , 0 ] . to connect these to probabilities we normalize these numbers by @xmath246 . 1 .
a site @xmath11 is chosen at random .
if @xmath228 is non - zero , draw a random number @xmath247 in the interval ( 0,1 ) .
2 . go to the @xmath248 row of the table and check the two consecutive entries which sandwich @xmath247 .
the column number of the larger entry is the number of mass particles @xmath12 that chip from @xmath185 .
thus @xmath249 , and @xmath250 or @xmath251 with probability 1/2 .
kuiri , b. joseph , h.p .
lenka , g. sahu , j. ghatak , d. kanjilal and d.p .
mahapatra , observation of a universal aggregation mechanism and a possible phase transition in au sputtered by swift heavy ions " , phys .
* 100 * , 245501 ( 2008 ) . vs. @xmath2 from 1-@xmath3 mc simulations with three different functional forms for the chipping kernel @xmath133 .
the data sets are plotted on a linear - logarithmic scale .
the details of the mc simulations are provided in the text .
the mass density for the initial conditions is @xmath254 .
the solid line denotes the 1-chip solution in eq .
( [ soln11]).,title="fig:",width=453,height=377 ] + vs. @xmath7 on a linear - log scale , from mc simulations in @xmath255 .
the initial conditions were characterized by average density @xmath198 , and @xmath256 .
the solid line denotes the solution in eq .
( [ steady2 ] ) with @xmath257 .
( b ) analogous to ( a ) , but the initial conditions for the mc simulations have a mixture of both even and odd masses .
the corresponding parameter values are @xmath258 .
the solid line denotes the solution in eq .
( [ steady2 ] ) with @xmath259 , @xmath260 .
the inset highlights the staircase structure of the probability distribution .
, title="fig:",width=453,height=377 ] + vs. @xmath7 from mc simulations with @xmath202 , so that @xmath198 .
the solid line denotes the result in eq .
( [ steadym ] ) with @xmath101 , @xmath140 and @xmath204 calculated from eq .
( [ 3chs ] ) as @xmath261 .
( b ) plot of @xmath1 vs. @xmath7 for initial conditions with @xmath262 , so that @xmath208 .
the solid line denotes the result in eq .
( [ steadym ] ) with @xmath263.,title="fig:",width=453,height=377 ] + | we present a review of nonequilibrium phase transitions in mass - transport models with kinetic processes like fragmentation , diffusion , aggregation , etc .
these models have been used extensively to study a wide range of physical problems .
we provide a detailed discussion of the analytical and numerical techniques used to study mass - transport phenomena . |
one of the main tasks of nuclear physics is to understand the in - medium nuclear interactions and the equation of state ( eos ) of nuclear matter .
the uncertainties of the isospin - dependent part of the eos , i.e. , the nuclear symmetry energy ( @xmath10 ) , has hampered our accurate understanding of nuclear matter properties , while it has important ramifications in heavy - ion reactions , astrophysics , and nuclear structures @xcite .
thanks to the great efforts made by nuclear physicists in the past decade , a more stringent constraint on @xmath10 at subsaturation densities has been obtained from various analysis , with the slope parameter of the nuclear symmetry energy so far constrained within @xmath11 mev @xcite , although further verifications are still needed . on the other hand ,
the in - medium isospin splitting of the nucleon effective mass @xmath12 has become a hot topic recently . compared to the bare nucleon mass in free space ,
the in - medium nucleon effective mass comes from the momentum - dependent potential in non - relativistic models ( see , e.g. , chapter 3 of ref .
the isospin splitting of the in - medium nucleon effective mass is thus related to the momentum dependence of the symmetry potential in non - relativistic models @xcite .
it has been found that the isospin splitting of the nucleon effective mass is as important as the nuclear symmetry energy in understanding the isospin dynamics in nuclear reactions @xcite , and has ramifications in the thermodynamic properties of isospin asymmetric nuclear matter as well @xcite .
moreover , the neutron - proton effective mass splitting is actually inter - related to the nuclear symmetry energy through the hugenholtz - van hove theorem @xcite .
for a recent review on the isospin splitting of the nucleon effective mass , we refer the reader to ref .
@xcite .
giant resonances of nuclei serve as a useful probe of nuclear interactions and the eos of nuclear matter at subsaturation densities .
the studies on giant resonances mainly follow two methods , i.e. , the random - phase approximation and the transport model calculations . as a breathing oscillation mode in the radial direction of a nucleus
, the isoscalar giant monopole resonance ( gmr ) is a good probe of the incompressibility of nuclear matter @xcite , while the isoscalar giant quadruple resonance ( isgqr ) , an oscillation mode with quadruple deformation of a nucleus , has been found to be much affected by the isoscalar nucleon effective mass @xmath13 @xcite . on the other hand , the isovector giant dipole resonance ( ivgdr ) and the pygmy dipole resonance ( pdr ) , with the former an oscillation mode between the centers of mass of neutrons and protons and the latter that between the neutron skin and the nucleus core , are valuable probes of the nuclear symmetry energy at subsaturation densities @xcite . since the nuclear symmetry energy acts as a restoring force for the ivgdr
, the main frequency of the ivgdr oscillation , or the centroid energy @xmath14 , is related to @xmath10 at subsaturation densities or its slope parameter @xmath0 @xcite .
the electric dipole polarizability @xmath15 has a strong correlation with the neutron skin thickness @xmath16 @xcite , and @xmath15 times @xmath10 at the saturation density shows a good linear dependence on @xmath0 @xcite .
it was argued that the accurate knowledge of @xmath15 and @xmath17 can help constrain significantly the nuclear symmetry energy at subsaturation densities @xcite .
the effect of the neutron - proton effective mass splitting on the ivgdr properties was realized only recently @xcite .
it is also worth mentioning that the dynamical isovector dipole collective motion in fusion reactions could be a probe of the nuclear symmetry energy as well from transport model studies @xcite .
recently we have improved our isospin- and momentum - dependent interaction ( mdi ) @xcite , which was previously extensively used in the studies of thermodynamic properties of nuclear matter , dynamics of nuclear reactions , and properties of compact stars ( see ref .
@xcite for a review ) . in the improved isospin- and momentum - dependent interaction ( immdi )
@xcite , the momentum dependence of the mean - field potential has been fitted to that extracted from proton - nucleus scatterings up to the nucleon kinetic energy of about 1 gev , and more isovector parameters are further introduced so that the density dependence of the symmetry energy and the momentum dependence of the symmetry potential , or equivalently , the isospin splitting of the nucleon effective mass , can be mimicked separately . in the present study , we are going to investigate the effect of the nuclear symmetry energy and the neutron - proton effective mass splitting on the centroid energy @xmath14 of ivgdr as well as the electric dipole polarizability @xmath15 , by employing the immdi interaction together with the isospin - dependent boltzmann - uehling - uhlenbeck ( ibuu ) transport model . with the experimental data of @xmath14 and @xmath15 from the ivgdr in @xmath2pb available @xcite , we are able to constrain both the slope parameter of the symmetry energy and the isospin splitting of the nucleon effective mass , once the isoscalar nucleon effective mass is well determined .
section [ immdi and giant resonance ] gives a brief introduction to the immdi interaction and the necessary formalisms for ivgdr .
detailed studies on the extraction of @xmath13 and the constraint on @xmath0 and @xmath18 from ivgdr are presented in sec .
[ results ] .
we conclude in sec . [ summary ] .
the potential energy density functional in the asymmetric nuclear matter with isospin asymmetry @xmath6 and nucleon number density @xmath19 from the immdi interaction can be expressed as @xcite @xmath20the single - particle potential of a nucleon with isospin @xmath21 and momentum @xmath22 in the asymmetric nuclear matter with the isospin asymmetry @xmath6 and nucleon number density @xmath19 can be expressed as @xmath23 in the above , @xmath24 and @xmath25 are the number density of neutrons and protons , @xmath26 is the saturation density , and @xmath27 is the isospin asymmetry .
@xmath28 is the phase - space distribution function with @xmath29 being the isospin label of neutrons ( protons ) . the potential energy density of eq .
( [ mdiv ] ) can be derived based on the hartree - fock calculation from an effective interaction with a zero - range density - dependent term and a finite - range yukawa - type term @xcite . in the immdi interaction @xcite ,
the isovector parameters @xmath30 , @xmath31 , and @xmath32 are introduced to vary respectively the slope parameter of the symmetry energy , the momentum dependence of the symmetry potential or the neutron - proton effective mass splitting , and the value of the symmetry energy at the saturation density , via the following relations @xmath33 } , \nonumber \\
c_u(y , z)&=&c_{u0 } + 2(y-2z)\frac{p^2_{f}}{\lambda^2\ln[(4 p^2_{f } + \lambda^2)/\lambda^2 ] } , \nonumber \\\end{aligned}\ ] ] where @xmath34 is the nucleon fermi momentum in symmetric nuclear matter at the saturation density .
we set @xmath35 all through the manuscript .
the values of the parameters @xmath36 , @xmath37 , @xmath38 , @xmath39 , @xmath40 , @xmath41 , @xmath30 , and @xmath31 can be fitted or solved from the saturation density @xmath42 , the binding energy @xmath43 at @xmath42 , the incompressibility @xmath44 , the mean - field potential @xmath45 at infinitely large nucleon momentum at @xmath42 , the isoscalar nucleon effective mass @xmath13 at @xmath42 , the symmetry energy @xmath10 and its slope parameter @xmath0 at @xmath42 , and the isovector nucleon effective mass @xmath46 at @xmath42 , as detailed in [ app ] .
the isovector giant dipole resonance ( ivgdr ) is a collective vibration of protons against neutrons , with the isovector dipole operator defined as @xmath47 where @xmath48 , @xmath49 , and @xmath50 are the number of neutrons , protons , and the mass number of the nucleus , respectively , and @xmath51 is the distance between the centers of mass of protons and neutrons in the nucleus .
the strength function of ivgdr can be calculated from the isovector dipole moment via @xmath52}{\pi \eta } , \end{aligned}\ ] ] where @xmath53 is the fourier transformation of the isovector dipole moment with @xmath54 . in order to initialize the isovector dipole oscillation , the initial momenta of protons and neutrons
are given a perturbation in the opposite direction according to @xcite @xmath55 with @xmath56 being the momentum of the @xmath57th nucleon along @xmath51 , and the perturbation parameter @xmath58 mev / c used in the present study .
one sees that a larger @xmath59 leads to a larger amplitude of the oscillation @xmath60 , while the strength function @xmath61 is independent of @xmath59 as the amplitude is cancelled from the denominator .
once the strength function of ivgdr is obtained , the moments of the strength function can be calculated from @xmath62 both the centroid energy and the electric dipole polarizability can be measured from photoabsorption experiments .
the centroid energy , corresponding to the peak energy in the photoabsorption energy spectrum , is the main frequency of ivgdr .
the electric dipole polarizability , characterizing the response of the nucleus to the external electric field , is related to the photoabsorption cross section @xmath63 via @xcite @xmath64 since both the strength function @xmath61 and the photoabsorption cross section @xmath63 are related to the energy spectrum of the excited states in the nucleus , the centroid energy and the electric dipole polarizability can be expressed in terms of the moments @xmath65 of the strength function as @xmath66 and @xmath67 we note that the relation between @xmath15 and @xmath68 is different from that in ref .
@xcite , because we only consider one - dimensional oscillation , and the definition of the isovector giant dipole operator is different from that used in ref .
@xcite based on a random - phase approximation approach . in the following calculations ,
we fit the time evolution of the isovector giant dipole moment with the function @xmath69 where @xmath70 , @xmath71 , and @xmath72 are fitting constants characterizing the amplitude , the frequency , and the damping time of ivgdr , respectively . with the help of eq .
( [ fit ] ) , the strength function @xmath61 , the moments of the strength function @xmath68 and @xmath73 , the centroid energy @xmath14 of ivgdr , and the electric dipole polarizability @xmath74 can be expressed analytically in terms of the fitting constants respectively as @xmath75 , \notag \\
m_{-1}&=&\frac{-ab}{2\eta\left ( b^{2}+c^{2 } \right ) } , \notag \\
m_{1}&=&\frac{-ab}{2\eta } , \notag \\
e_{-1}&=&\sqrt{m_{1}/m_{-1}}= \sqrt{b^{2}+c^{2 } } , \notag \\ \alpha_{d}&=&2e^{2}m_{-1}=\frac{-2e^{2}ab}{2\eta\left ( b^{2}+c^{2 } \right ) }
. \label{abc}\end{aligned}\ ] ]
in the following study , we employ the ibuu transport model together with the immdi interaction to investigate the giant resonances of nuclei .
the positions of the projectile and target in the ibuu transport model are fixed , i.e. , with zero beam energy .
the initial density distribution is sampled according to that generated from skyrme - hartree - fock calculations with the same physics quantities , such as @xmath0 , @xmath46 , etc .
, listed in table [ para ] .
the initial nucleon momentum is sampled accordingly to the local density from the thomas - fermi approximation .
we generate events from 10 runs with each run 200 test particles .
since the oscillation generally lasts for hunderds of fm / c , in order to improve the stability in the calculation with the momentum - dependent mean - field potential , we use the nuclear matter approximation in the real calculation by taking the phase - space distribution function as @xmath76 and using the analytical expression ( eqs .
( [ mdiu ] ) and ( [ int1 ] ) ) for the momentum - dependent mean - field potential .
this is similar to the spirit of the thomas - fermi approximation in the case that the vibration compared to the stable distribution is small . with this treatment
, we found that up to @xmath77 fm / c only about @xmath78 of the total nucleons become free particles .
we first extract the isoscalar nucleon effective mass from the optical potential .
the single - particle potentials in symmetric nuclear nuclear matter at the saturation density , with different values of the isoscalar nucleon effective mass @xmath79 but same other isoscalar properties of the nuclear interaction , are displayed in fig .
we can see that only the parameter set that leads to @xmath80 , with @xmath81 being the nucleon mass in free space , can describe reasonably well the real part of the optical potential extracted from proton - nucleus scatterings by hama et al .
@xcite . .
the real part of the optical potential extracted from proton - nucleus scatterings by hama et al .
@xcite is shown by scatters for comparison . ] [ cols="^,^,^,^,^,^,^",options="header " , ] [ para ] next , we extract the value of @xmath82 from the isoscalar giant quadruple resonance ( isgqr ) in @xmath2pb , with the operator written as @xmath83 where @xmath84 and @xmath85 are respectively the radial and @xmath32-direction coordinate of the @xmath57th nucleon , and @xmath86 is the spherical harmonic function .
noticing that the following scaling relation in the isgqr is observed ( see , e.g. , ref .
@xcite ) @xmath87 we choose @xmath88 in our simulation to initialize the oscillation .
the value of @xmath89 is close to @xmath90 corresponding to a small vibration with respect to the equilibrium distribution .
again , we found that the value of @xmath89 only affects the amplitude but has almost no effect on the frequency of isgqr .
the time evolution of the isoscalar giant quadruple moment , with different values of the isoscalar nucleon effective mass but same other isoscalar properties of nuclear interaction , is displayed in the left panel of fig .
it is seen that a smaller isoscalar nucleon effective mass @xmath91 leads to a larger frequency of the oscillation . from the fourier transformation of @xmath92 , the linear correlation between the isoscalar nucleon effective mass @xmath91 and the excitation energy of isgqr
is observed in the right panel of fig .
it is found that the result from the isoscalar nucleon effective mass @xmath93 reproduces best the excitation energy @xmath94 mev of isgqr in @xmath2pb extracted from @xmath95-nucleus scattering experiments @xcite .
pb ( a ) and the correlation between the corresponding excitation energy @xmath96 and the isoscalar nucleon effective mass @xmath91 ( b ) . ] the different values of the isoscalar effective mass @xmath79 extracted from the optical potential and from the excitation energy of isgqr in @xmath2pb represent the theoretical uncertainties in the present study . in the following study
, we will thus employ the parameter set i and set ii that lead respectively to @xmath97 and @xmath98 together with different combinations of the isovector parameters @xmath30 and @xmath31 to study the isovector giant dipole resonances of nuclei .
the values of the parameters and the corresponding physics quantities are detailed in table [ para ] .
we first employ the parameter set i with @xmath99 , which is able to reproduce the empirical optical potential as shown in fig .
[ f1 ] , to study the properties of the ivgdr in @xmath2pb .
three different parameter sets , with different combinations of the slope parameter @xmath0 of the symmetry energy and the isovector effective mass @xmath46 detailed as set i(a ) , set i ( b ) , and set i(c ) in table [ para ] , are employed in the study .
the initial momenta of neutrons and protons are modified with the perturbation parameter @xmath59 as detailed in sec .
[ immdi and giant resonance]b .
the oscillation amplitude is proportional to @xmath59 while its frequency is found to be insensitive to the choice of @xmath59 .
the resulting time evolution of the isovector dipole moment is displayed in the left panel of fig .
one sees that the time evolution of @xmath60 follows a good damping oscillation mode as in eq .
( [ fit ] ) . in order to avoid oscillation in the fourier transformation due to the finite @xmath100 in the transport model calculation , a damping factor of @xmath101 with the width @xmath102 mev
is multiplied to the isovector dipole moment @xmath60 in calculating the strength function @xmath61 from eq .
( [ se ] ) as in ref
. @xcite .
this slightly affects the damping coefficient @xmath72 in eq .
( [ fit ] ) but has very small effects on the final results as can be seen from the analytical formulaes of eq .
( [ abc ] ) .
the resulting strength functions from the numerical fourier transformation is displayed in the right panel of fig .
it is seen that with a softer symmetry energy ( set i(b ) ) , the main frequency , i.e. , the centroid energy in the ivgdr , is larger , due to the larger symmetry energy at subsaturation densities acting as a stronger restoring force . on the other hand ,
the centroid energy is sensitive to the isovector nucleon effective mass as well , with a larger isovector effective mass ( set i(c ) ) leading to a smaller centroid energy of ivgdr .
this could be understood since the oscillation frequency is smaller with a larger reduced mass of the system , with the latter attributed to the larger isovector effective mass once the isoscalar effective mass is fixed .
pb with @xmath99 . ] as well as the isovector nucleon effective mass @xmath103 and the centroid energy @xmath14 as well as the electric dipole polarizability @xmath15 from the ivgdr in @xmath2pb , respectively , with the isoscalar nucleon effective mass @xmath99 . ]
the constants @xmath70 , @xmath71 , and @xmath72 can be obtained from fitting the isovector dipole moment @xmath60 according to eq .
( [ fit ] ) , and their detailed values for the three parameter sets are listed in table [ para ] .
the resulting centroid energy @xmath14 and the electric dipole polarizability @xmath74 calculated analytical according to eq .
( [ abc ] ) are also listed in table [ para ] .
it is seen that for a given isovector effective mass @xmath103 , a larger slope parameter @xmath0 leads to a smaller centroid energy @xmath14 and a larger electric dipole polarizability @xmath104 .
analogously , for a given slope parameter @xmath0 , a larger isovector effective mass @xmath103 leads to a smaller centroid energy @xmath14 and a smaller electric dipole polarizability @xmath104 .
the rigorous tool to treat such a problem of multiple experimental results with multiple parameters is the bayesian framework @xcite , which is beyond the scope of the present study . on the other hand
, we found both @xmath14 and @xmath104 are linearly correlated with @xmath0 and @xmath103 , and the corresponding relations based on our transport calculations turns out to be @xmath105 with @xmath0 in mev , @xmath46 in @xmath81 , @xmath14 in mev , and @xmath15 in @xmath106 .
the centroid energy of the ivgdr in @xmath2pb obtained experimentally from the photoabsorption measurement is @xmath107 mev @xcite , while the electric dipole polarizability is @xmath108 @xmath109 measured from photoabsorption cross section by tamii et al .
@xcite and with further correction by subtracting quasideuteron excitations @xcite . with the above experimental data available , the constraints on the slope parameter @xmath0 and the isovector effective mass can be solved from eq .
( [ eq1 ] ) as @xmath110 where the error bars , which originate from the fitting and the statistical error of @xmath70 , @xmath71 , and @xmath72 listed in table [ para ] , are calculated from the error transfer .
we note that the statistical error from 10 runs is much larger than the fitting error in the present study .
besides , we found that the immdi parameterization with the mean values of @xmath0 and @xmath46 in eq .
( [ lmv1 ] ) gives very close results of @xmath14 and @xmath15 compared with the experimental data , justify the linear relation of eq .
( [ eq1 ] ) .
the corresponding isospin splitting of the nucleon effective mass deduced from eq .
( [ mnp ] ) is @xmath111 the above constraint is , however , different from that obtained by analyzing nucleon - nucleus scattering data within an isospin - dependent optical model @xcite .
next , we choose the isoscalar nucleon effective mass to be @xmath93 while keeping other physics quantities the same in the immdi parameterization , and the resulting parameter sets are listed as set ii(a ) , set ii(b ) , and set ii(c ) in table [ para ] with different combinations of the slope parameter @xmath0 of the symmetry energy and the isovector nucleon effective mass @xmath103 . with the same calculation method ,
the time evolution of the isovector dipole moment in @xmath2pb and the corresponding strength function are displayed respectively in the left and right panel of fig .
[ f5 ] . with @xmath70 , @xmath71 , and @xmath72
fitted by eq .
( [ fit ] ) , and the analytical results of the centroid energy @xmath14 and the electric dipole polarizability @xmath15 obtained according to eq .
( [ abc ] ) , we can get the similar linearly relation from transport calculations as @xmath112 with the available experimental data of @xmath14 and @xmath15 , the slope parameter @xmath0 of the symmetry energy and the isovector nucleon effective mass from the constraint of the isqgr and the ivgdr in @xmath2pb are @xmath113 again , we found that the mean values of @xmath0 and @xmath46 reproduce very well the experimental results of @xmath14 and @xmath15 within the statistical error based on our transport model calculations , justifying the linear relation of eq .
( [ eq2 ] ) .
the corresponding neutron - proton effective mass splitting deduced from eq .
( [ mnp ] ) is thus @xmath114 the constraint from both isgqr and ivgdr on the neutron - proton effective mass splitting is consistent with that obtained in ref .
@xcite .
but with the isoscalar nucleon effective mass @xmath115 . ] but with the isoscalar nucleon effective mass @xmath115 . ]
based on an improved isospin- and momentum - dependent interaction ( immdi ) and an isospin - dependent boltzmann - uehling - uhlenbeck ( ibuu ) transport model , we have studied the effect of the slope parameter @xmath0 of the nuclear symmetry energy and the isovector nucleon effective mass @xmath116 on the centroid energy @xmath14 of the isovector giant dipole resonance ( ivgdr ) and the electric dipole polarizability @xmath74 in @xmath2pb .
we found that both @xmath14 and @xmath15 are almost linearly correlated with @xmath0 and @xmath116 . with a given isoscalar nucleon effective mass , we are able to constrain the values of @xmath0 and @xmath46 with the available experimental data of @xmath14 and @xmath15 . from the isoscalar nucleon effective mass @xmath3 constrained by the empirical optical potential
, we obtain a constraint of @xmath117 and @xmath118 , resulting in the isospin splitting of the nucleon effective mass within @xmath119 , with @xmath6 being the isospin asymmetry of nuclear medium . from the isoscalar nucleon effective mass @xmath7
extracted from the excitation energy of the isoscalar giant quadruple resonance ( isgqr ) in @xmath2pb , we obtain a constraint of @xmath8 and @xmath120 , resulting in the isospin splitting of nucleon effective mass within @xmath121 .
the constraint on the neutron - proton effective mass splitting from both isgqr and ivgdr in @xmath2pb is consistent with that from optical model analyses of nucleon - nucleus elastic scatterings .
the uncertainty of the isoscalar nucleon effective mass has hampered our accurate constraint on the neutron - proton effective mass splitting .
we thank chen zhong for maintaining the high - quality performance of the computer facility , and acknowledge helpful communications with zhen zhang .
jx acknowledges support from the major state basic research development program ( 973 program ) of china under contract no .
2015cb856904 and no .
2014cb845401 , the national natural science foundation of china under grant no .
11475243 and no .
11421505 , the `` 100-talent plan '' of shanghai institute of applied physics under grant no .
y290061011 and no .
y526011011 from the chinese academy of sciences , the shanghai key laboratory of particle physics and cosmology under grant no .
15dz2272100 , and the `` shanghai pujiang program '' under grant no .
lwc acknowledges the major state basic research development program ( 973 program ) in china under contract no .
2013cb834405 and no .
2015cb856904 , the national natural science foundation of china under grant no
. 11275125 and no . 11135011 , the `` shu guang '' project supported by shanghai municipal education commission and shanghai education development foundation , the program for professor of special appointment ( eastern scholar ) at shanghai institutions of higher learning , and the science and technology commission of shanghai municipality ( 11dz2260700 ) .
bal acknowledges the national natural science foundation of china under grant no .
11320101004 , the u.s .
department of energy , office of science , under award number de - sc0013702 , and the custipen ( china - u.s .
theory institute for physics with exotic nuclei ) under the us department of energy grant no .
defg02- 13er42025 .
ygm acknowledges the major state basic research development program ( 973 program ) of china under contract no .
2014cb845401 and the national natural science foundation of china under contract nos .
11421505 and 11220101005 .
at zero temperature , the phase - space distribution function can be written as @xmath76 , with @xmath122 being the fermi momentum of nucleons with the isospin label @xmath21 , and the momentum - dependent part of the single - particle potential as well as that in the potential energy density can be integrated analytically as @xcite @xmath123 \right .
+ & & \left.\frac{2p_{f\tau } } { \lambda } -2\left ( \arctan\frac { p+p_{f\tau } } { \lambda } -\arctan\frac { p - p_{f\tau } } { \lambda } \right)\right\}\notag\\\end{aligned}\ ] ] and @xmath124 \nonumber\\ + & & 4\lambda\left[(p_{f\tau}^{3}-p_{f\tau ^{\prime}}^{3})\arctan\left(\frac{p_{f\tau}-p_{f\tau ^{\prime}}}{\lambda}\right)\right.\nonumber \\ - & & \left.(p_{f\tau}^{3}+p_{f\tau ^{\prime}}^{3})\arctan\left(\frac{p_{f\tau}+p_{f\tau ^{\prime}}}{\lambda}\right)\right ] \nonumber\\ + & & \frac{1}{4}[\lambda^{4}+6\lambda^{2}(p_{f\tau}^{2}+p_{f\tau ^{\prime}}^{2})-3(p_{f\tau}^{2}-p_{f\tau ^{\prime}}^{2})^{2 } ] \nonumber\\ \times & & \ln\left[\frac{(p_{f\tau } + p_{f\tau ^{\prime}})^{2}+\lambda ^{2}}{(p_{f\tau } -p_{f\tau ^{\prime}})^{2}+\lambda ^{2}}\right]\}.\end{aligned}\ ] ] the binding energy per nucleon for asymmetric nuclear matter can be expressed as @xmath125 with the kinetic energy per nucleon calculated from @xmath126 \notag\\ & = & \frac{4\pi } { 5 m h^{3 } \rho } \left ( p_{fn}^{5}+p_{fp}^{5 } \right ) , \end{aligned}\ ] ] where @xmath127 is the fermi momentum of neutrons ( protons ) , and @xmath81 is the nucleon mass . by setting @xmath128 and @xmath129
, we can express the binding energy per nucleon for symmetric nuclear matter as @xcite @xmath130.\end{aligned}\ ] ] the saturation density is determined by the zero point of the first - order derivative of the binding energy per nucleon , with the latter expressed as @xcite @xmath131.\notag\end{aligned}\ ] ] the incompressibility of symmetric nuclear matter is defined as @xmath132 , with the second - order derivative of the binding energy per nucleon expressed as @xcite @xmath133.\notag \end{aligned}\ ] ] the symmetry energy by definition can be written as @xcite @xmath134 \\ & + & \frac{c_{u}}{9\rho _ { 0}\rho } \left ( \frac{4\pi } { h^{3 } } \right ) ^{2}\lambda ^{2}\left [ 4p_{f}^{4}-p_{f}^{2}\left(4p_{f}^{2}+\lambda ^{2}\right ) \ln\left(\frac{4p_{f}^{2}+\lambda ^{2}}{\lambda ^{2 } } \right)\right ] \nonumber.\end{aligned}\ ] ] the slope parameter of the symmetry energy at the saturation density is defined as @xmath135_{\rho = \rho _ { 0}}$ ] , with the first - order derivative of the symmetry energy expressed as @xcite @xmath136 \notag\\ & -&\frac{4c_u}{27\rho_0\rho^2}\left ( \frac{4\pi}{h^3}\right)^2 \lambda^2 p_f^4
\left [ \ln \left ( \frac{4p_f^2+\lambda^2}{\lambda^2}\right)+\frac{8p_f^2}{4p_f^2+\lambda^2}\right].\notag\end{aligned}\ ] ] the isoscalar nucleon effective mass @xmath79 is defined as the nucleon effective mass in symmetric nuclear matter at the saturation density , and it can be calculated from the mean - field potential @xmath137 in symmetric nuclear matter via @xmath138 in asymmetric nuclear matter , the isovector nucleon effective mass can be calculated through the following relation @xmath139 with the neutron ( proton ) effective mass defined as @xmath140 keeping the first - order term of the isospin asymmetry @xmath6 , the neutron - proton effective mass splitting is related to the isoscalar and isovector effective mass through the following relation @xmath141 finally , the mean - field potential of a nucleon with infinitely large momentum in symmetric nuclear matter at the saturation density can be expressed as @xmath142 the values of the parameters @xmath36 , @xmath37 , @xmath38 , @xmath39 , @xmath40 , @xmath41 , @xmath30 , and @xmath31 can be obtained from eqs .
( [ e0]-[u0 ] ) , with given @xmath42 , @xmath143 , @xmath44 , @xmath45 , @xmath13 , @xmath144 , @xmath0 , and @xmath46 .
ko , and w. bauer , int .
e * 7 * , 147 ( 1998 ) . _ isospin physics in heavy - ion collisions at intermediate energies _ , eds .
li and w. uuo schrder ( nova science publishers , inc , new york , 2001 ) .
p. danielewicz , r. lacey , and w.g .
lynch , science * 298 * , 1592 ( 2002 ) .
v. baran , m. colonna , v. greco , and m. di toro , phys .
rep . * 410 * , 335 ( 2005 ) .
a.w . steiner , m. prakash , j.m .
lattimer , and p.j .
ellis , phys .
rep . * 411 * , 325 ( 2005 ) .
lattimer and m. prakash , phys . rep . *
442 * , 109 ( 2007 ) .
chen , and c.m .
ko , phys .
rep . * 464 * , 113 ( 2008 ) .
chen , arxiv:1212.0284 [ nucl - th ] .
li and x. han , phys .
b * 727 * , 276 ( 2013 ) .
m. oertel , m. hempel , t. klhn , and s. typel , arxiv:1610.03361 [ astro-ph.he ] .
li , phys .
c * 69 * , 064602 ( 2004 ) .
j. rizzo , m. colonna , and m. di toro , phys .
c * 72 * , 064609 ( 2005 ) . v. giordano _ et al .
c * 81 * , 044611 ( 2010 ) .
feng , phys .
c * 84 * , 024610 ( 2011 ) .
z.q . feng , nucl .
a * 878 * , 3 ( 2012 ) .
zhang , m.b .
tsang , z.x .
li , and h. liu , phys .
b * 732 * , 186 ( 2014 ) .
w.j . xie and f.s .
zhang , phys .
b * 735 * , 250 ( 2014 ) .
kong , y. xia , j. xu , l.w .
chen , b.a .
li , and y.g .
ma , phys .
c * 91 * , 047601 ( 2015 ) .
b * 697 * , 246 ( 2011 ) .
j. xu , l.w .
chen , and b.a .
li , phys .
c * 91 * , 014611 ( 2015 ) . c. xu , b.a .
li , and l.w .
chen , phys .
c * 82 * , 054607 ( 2010 ) .
li and l.w .
chen , mod .
a * 30 * , 1530010 ( 2015 ) ; b.a .
cai , l.w .
chen , and x.h .
li , nuclear science and techniques , * 27 * , 141 ( 2016 ) .
j. blaizot _
et al . _ ,
phys . a * 591 * , 435 ( 1995 ) .
youngblood , h.l .
clark , and y.w .
lui , phys .
lett . * 82 * , 691 ( 1999 ) .
b. k. agrawal , s. shlomo , and v. kim au , phys .
c * 68 * , 031304 ( 2003 ) . g. col _
et al . _ ,
c * 70 * , 024307 ( 2004 ) .
todd - rutel and j. piekarewicz , phys .
lett . * 95 * , 122501 ( 2005 ) .
a. bohr and b.r .
mottelson , _ nuclear stucture _ ,
i and ii ( w. a. benjamin inc . , reading , ma , 1975 ) .
o. bohigas , a.m. lane , and j. martorell , phys . rep . * 51 * , 267 ( 1979 ) .
blaizot , phys . rep . * 64 * , 171 ( 1980 ) .
p. klpfel , p .-
reinhard , t.j .
brvenich , and j.a .
maruhn , phys .
c * 79 * , 034310 ( 2009 ) .
x. roca - maza _ et al .
c * 87 * , 034301 ( 2013 ) .
z. zhang and l.w .
chen , phys .
c * 93 * , 034335 ( 2016 ) .
a. klimkiewicz _
c * 76 * , 051603(r ) ( 2007 ) .
l. trippa , g. col , and e. vigezzi , phys .
c * 77 * , 061304(r ) ( 2008 ) .
a. carbone _ et al .
c * 81 * , 041301(r ) ( 2010 ) .
reinhard and w. nazarewicz , phys .
c * 81 * , 051303(r ) ( 2010 ) .
lett . * 107 * , 062502 ( 2011 ) .
j. piekarewicz , b.k .
agrawal , g. col , w. nazarewicz , n. paar , p .-
reinhard , x. roca - maza , and d. vretenar , phys .
c * 85 * , 041302(r ) ( 2012 ) . x. roca - maza , g. pozzi , m. brenna , k. mizuyama , and g. col , phys . rev .
c * 85 * , 024601 ( 2012 ) .
d. vretenar , y.f .
niu , n. paar , and j. meng , phys .
c * 85 * , 044317 ( 2012 ) .
v. baran , b. frecus , m. colonna , and m. di toro , phys .
c * 85 * , 051601(r ) ( 2012 ) .
x. roca - maza , m. brenna , g. col , m. centelles , x. vias , b.k .
agrawal , n. paar , d. vretenar , and j. piekarewicz , phys . rev .
c * 88 * , 024316 ( 2013 ) . c. tao
_ et al . _ ,
c * 87 * , 014621 ( 2013 ) ; c. tao _ et al .
_ , nuclear science and techniques , * 24 * , 030502 ( 2013 ) .
g. col , u. garg , and h. sagawa , eur .
j. a * 50 * , 26 ( 2014 ) .
x. roca - maza , x. vias , m. centelles , b.k .
agrawal , g. col , n. paar , j. piekarewicz , and d. vretenar , phys .
c * 92 * , 064304 ( 2015 ) .
z. zhang and l.w .
chen , phys .
c * 93 * , 031301(r ) ( 2015 ) .
h. zheng , s. burrello , m. colonna , and v. baran , phys .
c * 94 * , 014313 ( 2016 ) . v. baran , d.m .
brink , m. colonna , and m. di toro , phys .
lett . * 87 * , 182501 ( 2001 ) . v. baran , m. cabibbo , m. colonna , m. di toro , and n. tsoneva , nucl .
a * 679 * , 373 ( 2001 ) .
c * 72 * , 064608 ( 2005 ) . v. baran , c. rizzo , m. colonna , m. di toro , and d. pierroutsakou , phys .
c * 79 * , 021603(r ) ( 2009 ) .
c * 81 * , 047602 ( 2010 ) .
c * 88 * , 047602 ( 2013 ) ; s.q .
_ , nuclear science and techniques , * 25 * , 030501 ( 2014 ) .
das , s. das gupta , c. gale , and b.a .
li , phys .
c * 67 * , 034611 ( 2003 ) .
l.w . chen , c.m .
ko , and b.a .
li , phys .
lett . * 94 * , 032701 ( 2005 ) .
l.w . chen , c.m .
li , c. xu , and j. xu , eur .
j. a * 50 * , 29 ( 2014 ) .
m. buenerd , j. phys .
colloques * 45 * , c4 - 115 ( 1984 ) ; d.h .
youngblood , p. bogucki , j.d .
bronson , u. garg , y. w. lui , and c.m .
rozsa , phys .
c * 23 * , 1997 ( 1981 ) ; s. brandenburg , ph.d .
thesis , university of groningen , 1985 ; d.h .
youngblood , y .- w .
lui , h.l .
clark , b. john , y. tokimoto , and x. chen , phys .
c * 69 * , 034315 ( 2004 ) . | with a newly improved isospin- and momentum - dependent interaction and an isospin - dependent boltzmann - uehling - uhlenbeck transport model , we have investigated the effects of the slope parameter @xmath0 of the nuclear symmetry energy and the isospin splitting of the nucleon effective mass @xmath1 on the centroid energy of the isovector giant dipole resonance and the electric dipole polarizability in @xmath2pb . with the isoscalar nucleon effective mass @xmath3 constrained by the empirical optical potential , we obtain a constraint of @xmath4 and @xmath5 , with @xmath6 being the isospin asymmetry of nuclear medium . with the isoscalar nucleon effective mass @xmath7 extracted from the excitation energy of the isoscalar giant quadruple resonance in @xmath2pb
, we obtain a constraint of @xmath8 and @xmath9 . |
the chromomagnetic operator of the @xmath2-type is defined as follows : @xmath3 in the effective hamiltonian is convention dependent and @xmath4 is the wilson coefficient . ] @xmath5 + .. $ ] , there is an additional factor of @xmath6 , whose origin can be understood from the minimal flavour symmetry of the standard model ( sm ) .
the operator is therefore effectively of mass dimension six and thus on the same footing as the four - fermi interactions .
the motivation for our work @xcite is twofold : first , to provide an estimate of the chromomagnetic matrix element ; and second , to compare it with the qcd factorisation ( qcdf ) calculation .
the matrix element is of importance for isospin asymmetries @xcite , as the photon can be emitted from the spectator quark , and for testing for weak phases in @xmath7 @xcite , which may be related to the relatively large direct cp - violation in @xmath8 @xcite .
the comparison with qcdf is not natural as lcsr are not tailored around the heavy quark expansion and contain additional contributions not inherent in the qcdf calculation .
possibly the most surprising outcome of our investigations is the appearance of a complex anomalous threshold on the physical sheet in the correlation function used to extract the matrix element .
we aim to compute the following matrix element @xmath9 where @xmath10 stands for a light vector meson of the @xmath11,etc .- type and the star indicates that the photon @xmath12 can be off - shell , i.e. @xmath13 . the formalism allows us to extract the matrix element above with the @xmath0-meson replaced by a @xmath1-meson as well as the @xmath10-meson replaced by a light pseudoscalar of the @xmath14,etc .- type . throughout this write - up we shall refer mostly to the @xmath15 transition and replacements for the other decays are considered implied .
( [ eq : matel ] ) decomposes into the following transverse lorentz structures is inserted to absorb trivial factors due to the @xmath16 wave functions . @xmath17 for @xmath18 in @xmath19 , @xmath20 in all other transitions into @xmath21 and @xmath18 , and @xmath22 otherwise . ]
@xmath23 { \cal a}^{*\rho}(p ) & = & k_g \ , g_t ( q^2 ) p_t^\rho \;. \end{aligned}\ ] ] ( @xmath24 where @xmath25 ) @xmath26 the prefactor @xmath27 is chosen such that @xmath28 and @xmath29 parallel the standard form factors @xmath30 and @xmath31 in the sense that the amplitude reads @xmath32 and likewise in the pseudoscalar case . under the replacement @xmath33 ,
i.e. @xmath34 , at our level of approximation , the matrix element transforms as follows , @xmath35 by virtue of parity conservation of qcd .
the matrix element is extracted from the following correlation function : ) , though from a physical point of view this does not make sense for an off - shell photon . ]
@xmath36 where @xmath37 plays the role of the interpolating current for the @xmath0-meson . at leading order in @xmath38
there are a total of twelve graphs .
we divide these into those where the gluon connects to the spectator quark ( s ) and those where it connects to the non - spectator quark ( ns ) : @xmath39 the four diagrams denoted by @xmath40 to @xmath41 in fig.[fig : diaa](top , middle ) contribute to @xmath42 whereas the diagrams at the bottom of the same figure correspond to the @xmath43-contributions . the latter factorise into a function of @xmath44 times standard vector , axial or tensor form factors .
the function @xmath45 , in terms of an expansion in powers of @xmath46 and logarithmic terms , has been obtained in the inclusive case in @xcite .
the two diagrams where the photon is emitted from the spectator quark and the gluon connects to the non - spectator quark are not shown .
they are expected to be small since no fraction of the @xmath47-rest mass is transmitted to the energetic photon .
to @xmath41 to which we refer to as spectator corrections ( bottom ) non - spectator corrections .
crosses denote possible places for photon emission.,width=345 ] to @xmath41 to which we refer to as spectator corrections ( bottom ) non - spectator corrections .
crosses denote possible places for photon emission.,width=345 ] the diagrams are computed via the light - cone operator product expansion ( lc - ope ) , since the correlation function is believed to be dominated by light - like distances .
schematically this amounts to to a convolution between a perturbatively calculable hard scattering kernel @xmath48 and light - cone distribution amplitudes ( da ) @xmath49 : @xmath50 the sum extends over increasing orders of twist , defined as the dimension of the operator minus its spin projection , and the variable @xmath51 stands for the momentum fraction of the strange quark in the light meson .
we limit ourselves to leading twist-2 .
for the important steps of selecting cuts for the dispersion relation and borel transformation we refer the reader to the main paper @xcite .
we shall quote here an intermediate expression , @xmath52 that relates an on - shell matrix element to an integral over an off - shell correlation function .
@xmath53 is a closed path on the physical riemann sheet that does not contain any singularities such as poles and cuts .
@xmath54 is of the same type except that it contains the pole of the matrix element . at this stage
the relation is exact but admittedly rather cryptic . in sum rules
the analytic structure of the correlation functions is usually such that the singularities are on the real line . in the case at hand though it happens that there is an anomalous threshold extending into the complex plane .
let us explain in more detail : after the reduction to scalar integrals the passarino - veltman function @xmath55 ( where @xmath56 ) appears in the expression for @xmath57 , c.f . fig . [
fig : triangle ] .
the fact that this function has an anomalous threshold extending into the lower half - plane , for @xmath58 and appropriate momentum fraction @xmath51 , can be seen in various ways .
first , using the explicit result valid on the real line we can show by uniqueness of analytic continuation from the real line that there must be singularities in the complex plane @xcite .
second , by setting @xmath59 in the dispersion integral we see that its path is deformed into the lower complex plane by analytic continuation of @xmath60 to its physical value @xcite .
the analytic structure of this function is depicted in fig .
[ fig : triangle ] for a simplified set of variables .
further to that , the work of klln and wightman @xcite shows that anomalous thresholds are present in the corresponding triangle function of the full theory using a minimal number of axioms .
this almost implies that the anomalous thresholds are present in the full theory : the loophole is that the reduction to a scalar object in the non - perturbative case is not as efficient as for the one - loop case ; thus it could in principle be that the contributions cancel , however this is unlikely and in any case not relevant to our discussion .
the important point is that the @xmath61 $ ] and thus the rest the anomalous branch cut is well above the @xmath62-pole , and the anomalous cut is part of what is usually called the continuum contribution in sum rules . in the following paragraph
we aim to discuss to what extent these anomalous thresholds are surprising and what their meaning is in the hope of clarifying to the reader some of our brief argumentation above .
pv - function .
the conventions are the same as in looptools @xcite and feyncalc @xcite .
( right ) analytic structure of @xmath63 in the @xmath64-plane ; more precisely this corresponds to the physical sheet .
the leading landau singularities of this triangle function are @xmath65 .
as discussed in the text @xmath66 is on the physical sheet whereas @xmath67 is not.,width=326 ] the crucial point is that the correspondence between matrix elements and correlation functions is complicated when the number of legs increases .
let us begin by discussing the simplest case .
for a two - point function of gauge invariant operators , a dispersion representation is in one - to - one correspondence with the insertion of a complete set of states as is explicit in the celebrated klln - lehmann representation @xcite and derivations thereof .
thus the analytic structure in the complex plane of the four momentum invariant has a cut and poles on the real line starting from the lowest state in the spectrum . for correlation functions with three or more fields
, there is no such direct relation .
it seems preferable to think of the analytic structure ( singularity structure ) as a fundamental part of the correlation function rather than as the insertion of a complete set of states . for the dispersion relation , which is essentially an application of cauchy s integral theorem ,
it is immaterial whether the singularities are of the normal or anomalous type .
normal thresholds are related to unitarity , that is to say to the insertion of a complete set of states .
anomalous thresholds do not have such an interpretation .
in fact , the anomalous threshold of the triangle graph in perturbation theory corresponds to the leading landau singularity which in turn amounts to setting all the propagators on the mass shell .
one out of the two solutions @xmath68 of the ( leading ) landau equations turns out to be the end of the anomalous threshold ( fig .
[ fig : triangle ] ) , whereas the other one is not on the physical sheet .
we should add that determining which riemann - sheet a complex singularity is on is generally a difficult problem .
we refer the reader to reference @xcite for some explicit analytic results .
the results of @xmath57 and @xmath69 are collected in appendices a and c of that reference . amongst all the possible flavour transitions , only four are characteristically different depending on whether the initial meson is neutral or charged and whether it is of the beauty or charm type .
various subparts of this class , at @xmath70 , are collected in tab .
[ tab : ratios ] .
the ratio of @xmath71 to @xmath72 can be understood as a radiative correction and is proportional to @xmath73 times other factors of @xmath74 .
further semi - quantitative insight can be gained by considering the heavy quark scaling of the matrix elements which is @xmath75 except for diagrams @xmath76 which scale as @xmath77 $ ] as discussed in section [ sec : qcdfac ] .
projecting out the quark charges the respective ratios of @xmath78 to @xmath79 follow the ratio of heavy quark scaling and @xmath80 surprisingly well .
@xmath81 , width=230 ] , width=230 ] and @xmath82 for charged ( c.f . fig .
[ fig : plotsm ] ) and uncharged @xmath0 meson .
these plots make it clear why one the @xmath83-functions are not referred to as form factors as they do not reveal a particular structure but parameterise a multitude of effects amongst which long distance contributions play a definite rle .
the results are valid in the region of @xmath84 above @xmath85 , that is to say outside the resonance region and say @xmath86 below @xmath87.,width=230 ] and @xmath82 for charged ( c.f .
[ fig : plotsm ] ) and uncharged @xmath0 meson .
these plots make it clear why one the @xmath83-functions are not referred to as form factors as they do not reveal a particular structure but parameterise a multitude of effects amongst which long distance contributions play a definite rle .
the results are valid in the region of @xmath84 above @xmath85 , that is to say outside the resonance region and say @xmath86 below @xmath87.,width=230 ] we consider it worthwhile to quickly mention how the various projections of the das contribute to the matrix elements and how they are interrelated . at twist-2
there are seven contributions : @xmath88 g_t^{(s ) } & = & g_t^{(p)}(q^2 ) \;,\end{aligned}\ ] ] it turns out that ( [ eq : gdec ] ) can be fully reconstructed by knowing the three subparts @xmath89 , @xmath82 and @xmath90 .
thus there are four relations : @xmath91 , @xmath92 , @xmath93 and @xmath94 .
the third relation assures a finite decay width in the limit @xmath95 ( as employed here ) @xcite .
the fourth relation is of the large energy effective theory ( leet)-type as found for form factors @xcite , whose origin is explained in appendix a of @xcite using the third relation and a generic ansatz .
furthermore , in the ultra - relativistic approximation @xmath95 , the projections @xmath90 @xmath96 are proportional to each other modulo a replacement of the corresponding da .
thus the results can be understood qualitatively from the plots of @xmath97 and @xmath98 c.f .
[ fig : plotsm],[fig : plots ] as the da of @xmath99 and @xmath100 hardly differ in practice
. the uncertainties , which are added in quadrature , are estimated to be between @xmath101 and @xmath102 @xcite , for the @xmath103 and @xmath104-transitions respectively , depending on the hadronic input .
it worthwhile to point out that the uncertainties are higher for the charm transitions because of the low scale for @xmath38 .
we shall summarise here a few points of the discussion in chapter 5 of @xcite on comparing the contributions of diagrams @xmath76 of fig . [
fig : diaa ] in the qcdf and lcsr approaches .
these diagrams form a well defined subset as they are isospin dependent .
parameterising @xmath105 as @xmath106 the qcdf @xcite and lcsr @xcite results read @xmath107 x_\perp^{lcsr}(u ) & = & \int_{m_b^2}^{s_0 } ds \ , e^{\frac{m_b^2-s}{m^2 } } \rho(s , u ) \;,\end{aligned}\ ] ] with @xmath108 and @xmath109 \;.\ ] ] the endpoint divergence in qcdf arises as follows : assuming an endpoint behaviour @xmath110 , which is true at every finite order in the gegenbauer expansion , it is readily seen that @xmath111 is logarithmically divergent ( at the endpoint @xmath112 ) . on a purely technical level
this happens because two propagators behave as @xmath113 c.f .
[ fig : endpoint ] . in lcsr
there is only one propagator with manifest @xmath114-behaviour ( c.f .
[ fig : endpoint ] ) and there is no such term hidden in the loop as it would correspond to a power ir - divergence , and it is known that in four dimensions ir - singularities , whether they are soft or collinear , are at worst of logarithmic nature , e.g. @xcite .
thus the worst behaviour that we can get is @xmath115 which is integrable , i.e. does not show an endpoint - divergence .
inspection or evaluation eq .
( [ eq : xperp ] ) gives an even milder behaviour with @xmath116 for which we see no particular reason .
the comparison between qcdf and lcsr can be sharpened if the heavy quark scaling @xmath117 , @xmath118 and @xmath119 , as suggested in @xcite , is applied : @xmath120 the logarithmic term indicates that this expression is not expandable in inverse powers of the heavy quark mass .
in fact , the expansion of the density around @xmath121 reveals that @xmath122 in our results .
are the lcsr and the qcdf to be seen on an equal footing when the leading heavy quark scaling behaviour is considered ?
the answer must be no as the former has a sizeable imaginary part whereas the latter is real ( at leading order in @xmath38 ) @xcite .
this difference is due to the fact that lcsr contain additional contributions which are not present in the qcdf computation .
an interpretation of the complex phase as a long distance phenomenon is given in the caption of fig .
[ fig : phase_p ]
. in regards to the discussion above it is seems worthwhile to point out that even though qcdf and lcsr are both based on lc - expansions in terms of hard kernels and das they differ on a conceptual level .
qcdf computes physical processes in a direct way and is in that sense very transparent .
lcsr is of an indirect nature as correlation functions are computed in which the matrix element in question appears as a residue of a pole .
the matrix element is then extracted by manipulating the sum rule ( borel transformation ) and considering appropriate kinematical limits . in this sense lcsr
is akin to lattice - qcd extraction of matrix elements . in both figures .
( left ) diagram used in lcsr .
( right ) diagram used in qcdf .
thus @xmath123 and @xmath124 at worst , as explained in the text.,width=345 ] -particle cut in in terms of a long - distance hadronic process .
the state @xmath125 is any state , single or multiparticle , of spin zero with @xmath126 and @xmath127 quantum numbers .
the latter is a source for the strong ( cp - even ) phase that we obtain for the @xmath83-functions.,width=288 ]
in this write up we have reported on the computation of matrix elements of heavy to light meson transitions induced by the @xmath128 operator using lcsr .
we focused on the appearance of a complex anomalous threshold and on comparing our results with those previously computed in qcdf .
a characteristic feature of the results is a large strong phase reflecting the long distance physics contained in these matrix elements .
rz is grateful for the support of an advanced stfc fellowship and to the organisers of the 4th capri workshop on flavour physics and to many colleagues for inspiring discussions .
computer manipulation and numerics are performed with the aid of looptools @xcite and feyncalc @xcite .
m. dimou , j. lyon and r. zwicky , `` exclusive chromomagnetism in heavy - to - light fcncs , '' phys .
d * 87 * ( 2013 ) 074008 arxiv:1212.2242 [ hep - ph ] .
a. l. kagan and m. neubert , `` isospin breaking in @xmath129 gamma decays , '' phys .
b * 539 * ( 2002 ) 227 [ hep - ph/0110078 ] .
t. feldmann and j. matias , `` forward backward and isospin asymmetry for @xmath130 decay in the standard model and in supersymmetry , '' jhep * 0301 * ( 2003 ) 074 [ hep - ph/0212158 ] .
y. y. keum , m. matsumori and a. i. sanda , `` cp asymmetry , branching ratios and isospin breaking effects of @xmath129 gamma with perturbative qcd approach , '' phys .
d * 72 * ( 2005 ) 014013 [ hep - ph/0406055 ] .
p. ball , g. w. jones and r. zwicky , `` b > v gamma beyond qcd factorisation , '' phys .
d * 75 * ( 2007 ) 054004 [ hep - ph/0612081 ] .
j. lyon and r. zwicky , `` isospin asymmetries in @xmath131 and @xmath132 in and beyond the standard model , '' arxiv:1305.4797 [ hep - ph ] .
g. klln , `` on the definition of the renormalization constants in quantum electrodynamics , '' helv .
acta * 25 * ( 1952 ) 417 .
+ h. lehmann , `` on the properties of propagation functions and renormalization contants of quantized fields , '' nuovo cim .
* 11 * ( 1954 ) 342 . | we report the computation of the matrix element of the chromomagnetic operator of the flavour changing neutral current ( fcnc)-type between a @xmath0- or @xmath1-meson state and a light hadron and off - shell photon .
the computation is carried out by using the method of light - cone sum rules ( lcsr ) .
it is found that the matrix element exhibits a large strong phase for which we give a long distance interpretation .
the analytic structure of the correlation function in use admits a complex anomalous threshold on the physical sheet , the meaning and handling of which within the sum rule approach is discussed .
we compare our results to qcd factorisation for which spectator photon emission is end - point divergent . |
most of analysis in wireless networks consider a 2d analytical model of network .
however since the needs are increasing in terms of qos and performance , and because of the scarcity of the frequency bandwidth , forecast of performance needs to be more and more accurate .
this is the reason why more accurate analytical models are developed and in particular 3d models of wireless networks which should be closer to a real network .
a large literature has been developed about the modeling of wireless networks @xcite @xcite @xcite @xcite @xcite @xcite @xcite @xcite @xcite , in the aim to analyze the capacity , throughput , coverage , and more generally performances for different types of wireless systems such as sensors , cellular and ad - hoc ones .
these models of networks are based on i ) stochastic geometry : the transmitters are distributed according to a spatial poisson process , ii ) hexagonal pattern : the transmitting base stations constitute a regular infinite hexagonal grid , iii ) a fluid approach : the interfering transmitters are replaced by a continuum .
the hexagonal wireless network model is the most used one ( @xcite-@xcite-@xcite ) .
this model seems rather reasonable for regular deployments of base stations .
however , it is a two dimensional one : it does not take into account the height of antennas , and the propagation in the three dimensions . moreover , from an analytical point of view , it is intractable . therefore , extensive computations are needed to establish performance . among the techniques developed to perform such computations , monte carlo simulations are widely used in conjunction with this model @xcite or numerical computations in hexagonal networks @xcite .
more generally , most of the works focus on 2d models of wireless networks .
nevertheless , 3d models of wireless networks were developed and analyzed in term of capacity @xcite @xcite .
authors of @xcite present a 3d geometrical propagation mathematical model . based on this model , a 3d reference model for mimo mobile to mobile fading channels
is proposed . in @xcite , authors propose an exact form of the coverage probability , in two cases : i ) interferers form a 3d poisson point process and ii ) interferers form a 3d modified matern process .
the results established are then compared with a 2d case .
authors of @xcite develop a three - dimensional channel model between radar and cellular base stations , in which the radar uses a two - dimensional antenna array and the base station uses a one - dimensional antenna array .
their approach allows them to evaluate the interference impact .
the 3d wireless model for sensor networks developed in @xcite allows the authors to develop an analysis where the coverage can be improved .
the tilt angle is the corner stone of an algorithm which uses the 3d model they propose , : in this paper , we develop a three dimensional analytical model of wireless networks .
we establish a closed form formula of the sinr .
we show that the formula allows analyzing wireless networks with more accuracy than a 2d wireless approach .
comparisons of performance coverage and qos results , established with this 3d spatial fluid model and with a classical 2d model , show the wide interest of using this approach instead of a 2d approach .
the paper is organized as follows . in section [ model ] , we develop the 3d network model . in section [ 3dfluid ] , the analytical expression of the sinr by using this 3d analytical fluid model
is established . in section [ valid3d ] ,
the validation of this analytical 3d model is done by comparison with monte carlo simulations .
section [ conclusion ] concludes the paper .
we consider a wireless network consisting of @xmath0 geographical sites , each one composed by 3 base stations .
each antenna covers a sectored cell .
we focus our analysis on the downlink , in the context of an ofdma based wireless network , with frequency reuse 1 .
let us consider : * @xmath1 the set of geographic sites , uniformly and regularly distributed over a two - dimensional plane . * @xmath2 the set of base stations , uniformly and regularly distributed over the two - dimensional plane .
the base stations are equipped with directional antennas : @xmath3= 3 @xmath0 . *
the antenna height , denoted @xmath4 .
* @xmath5 sub - carriers @xmath6 where we denote @xmath7 the bandwidth of each sub - carrier . * @xmath8 the transmitted power assigned by the base station @xmath9 to sub - carrier @xmath10 towards user @xmath11 . * @xmath12 the propagation gain between transmitter @xmath9 and user @xmath11 in sub - carrier @xmath10 .
we assume that time is divided into slots .
each slot consists in a given sequence of ofdma symbols . as usual at network level
, we assume that there is no inter - carrier interference ( ici ) so that there is no intra - cell interference .
the total amount of power received by a ue @xmath11 connected to the base station @xmath13 , on sub - carrier @xmath10 is given by the sum of : a useful signal @xmath14 , an interference power due to the other transmitters @xmath15 and thermal noise power @xmath16 .
we consider the sinr @xmath17 defined by : @xmath18 as the criterion of radio quality .
we investigate the quality of service and performance issues of a network composed of sites equipped with 3d directional transmitting antennas .
the analyzed scenarios consider that all the subcarriers are allocated to ues ( full load scenario ) .
consequently , each sub - carrier @xmath10 of any base station is used and can interfere with the ones of other sites .
all sub - carriers are independent , we can thus focus on a generic one and drop the index @xmath10 .
let us consider the path - gain model @xmath19 , where @xmath20 is a constant , @xmath21 is the distance between a transmitter @xmath22 and a receiver @xmath11 , and @xmath23 2 is the path - loss exponent .
the parameter @xmath24 is the antenna gain ( assuming that receivers have a 0 dbi antenna gain ) .
therefore , for a user @xmath11 located at distance @xmath25 from its serving base station @xmath13 , the expression ( [ sinr ] ) of the sinr can be expressed , for each sub - carrier ( dropping the index @xmath10 ) : @xmath26 where : * @xmath27 is the transmitted power , * @xmath28 is the pattern of the 3d transmitting antenna of the base station @xmath13 , and @xmath29 is the maximum antenna gain .
* @xmath30 is the horizontal angle between the ue and the principal direction of the antenna @xmath9 , * @xmath31 is the vertical angle between the ue and the antenna @xmath9 ( see fig . [ antennes3d-2 ] ) , * @xmath32 , where @xmath33 represents the projection of @xmath25 on the ground . the gain @xmath34 of an antenna in a direction @xmath35 is defined as the ratio between the power radiated in that direction and the power that radiates an isotropic antenna without losses .
this property characterizes the ability of an antenna to focus the radiated power in one direction . the parameter @xmath29 ( [ sinrdirect ] ) is particularly important for a beamforming analysis .
let notice that it is determined by considering that the power , which would be transmitted in all directions for a non directive antenna ( with a solid angle of @xmath36 ) , is transmitted in a solid angle given by the horizontal and the vertical apertures of the antenna . in the ideal case where the antenna emits in a cone defined by @xmath37 and @xmath38 ,
the gain is given by @xmath39 .
in our analysis , we conform to the model of @xcite for the antenna pattern ( gain , side - lobe level ) .
the antenna pattern which is applied to our scheme , is computed as : @xmath40,\end{aligned}\ ] ] where @xmath41 and @xmath42 correspond respectively to the horizontal and the vertical antenna patterns .
the horizontal antenna pattern used for each base station is given by : @xmath43,\end{aligned}\ ] ] where : * @xmath44 is the half - power beamwidth ( 3 db beamwidth ) ; * @xmath45 is the maximum attenuation .
the vertical antenna direction is given by : @xmath46,\end{aligned}\ ] ] where : * @xmath47 is the downtilt angle ; * @xmath48 is the 3 db beamwidth . .
it receives a useful power from antenna @xmath13 and interference power from antenna @xmath9 . ]
each site is constituted by 3 antennas ( 3 sectors ) .
therefore , for any site @xmath49 of the network , we have : @xmath50 where @xmath51 and @xmath52 represent the angles relative to the antenna @xmath53 for the site @xmath49 .
for the sake of simplicity , in expression ( [ sinrdirect ] ) we do a sum on the base stations ( not on the sites ) and denote @xmath30 and @xmath31 the angles relative to the antenna @xmath9 . for a ue at the distance @xmath54 from the antenna @xmath9
, the vertical angle can be expressed as : @xmath55 for interfering antennas , it can be noticed that since @xmath56 , we have @xmath57 , and @xmath58 . therefore the vertical antenna pattern ( [ eq : v_pattern ] ) can be written as : @xmath59 \nonumber \\ & \approx & -\min \left [ 12 \left ( \frac{\phi_{tilt}}{\phi_{3db } } \right)^{2 } , a_m \right ] \nonumber \\ & = & g_{v_{db}},\end{aligned}\ ] ] where @xmath60 $ ] ( i.e. a constant ) . and the antenna gain can be expressed as : @xmath61 \nonumber \\ & = & -\min \left [ -a_{h_{db}}(\theta ) - g_{v_{db } } , a_m \right ] \nonumber \\ & = & -\min \left [ -a_{h_{db}}(\theta ) , a_m + g_{v_{db } } \right ] + g_{v_{db } } \nonumber \\ & = & b_{db}(\theta)+ g_{v_{db}},\end{aligned}\ ] ] where @xmath62 $ ] .
so we have : @xmath63 \\
g_{v_{db } } = - \min\left[12\left(\frac{\phi_{tilt}}{\phi_{3db}}\right)^2,a_m\right ] \end{array } \right.\end{aligned}\ ] ] therefore , we establish that in this case , the vertical antenna gain only depends on the angle @xmath64 .
the main assumption of the fluid network modeling @xcite @xcite @xcite consists in replacing a fixed finite number of enbs by an equivalent continuum of enbs spatially distributed in the network .
therefore , the transmitting power of the set of interfering enbs of the network is considered as a continuum field all over the network . considering a density @xmath65 of sites @xmath0 and following the approach developed in @xcite @xcite @xcite ,
let us consider a ue located at @xmath66 in the area covered by the enb @xmath13 .
since each site is equipped by 3 antennas , we can express the denominator of ( [ sinrdirect ] ) as : @xmath67 where the integral represents the interference due to all the other sites of the network , and the discrete sum represents the interference due to the 2 antennas ( enb ) co - localized with the enb @xmath13 .
the index @xmath68 holds for these 2 antennas .
this can be further written as : @xmath69 since for the other enb of the network , the distance @xmath70 , we have @xmath71 , and the interference can be approximated by using ( [ eq : hv_pattern2 ] ) : @xmath72 the approach developed in @xcite @xcite allows to express @xmath73 as @xmath74 , where @xmath75 represents the intersite distance ( isd ) .
we refer the reader to @xcite @xcite @xcite for the detailed explanation and validation through monte carlo simulations .
therefore , ( [ interference2 ] ) can be expressed as : @xmath76 for a ue located at @xmath77 ( dropping the index @xmath13 ) , relatively to its serving enb , the inverse of the sinr ( [ sinrdirect ] ) is finally given by the expression : @xmath78 where the index @xmath68 holds for the 2 antennas ( enb ) co - localized with the serving antenna ( enb ) .
the closed form formula ( [ sinrfluid2 ] ) allows the calculation of the sinr in an easy way .
first of all , it only depends on the distance of a ue to its serving enb ( not on the distances to each enb ( sinr expression ( [ sinrdirect])).this formula highlights the network characteristic parameters which have an impact on the sinr ( path loss parameter , inter - site distance , antenna gain ) . a simple numerical calculation is needed , due to the tractability of that formula .
the sinr allows calculating the maximum theoretical achievable throughput @xmath79 of a ue @xmath11 , by using shannon expression . for a bandwidth @xmath7
, it can be written : @xmath80 * remarks : * in the case of realistic wireless network systems , it can be noticed that the mapping between the sinr and the achievable throughput are established by the mean of _ level curves_. +
the validation of the analytical formula ( [ sinrfluid2 ] ) consists in the comparison of the results established by this formula , to the ones established by monte carlo simulations .
let us consider : * a hexagonal network composed of sectored sites ; * three base stations per site ; * the 2d model : the antenna gain of a transmitting base station is given in db by : @xmath81,\ ] ] where @xmath44 = @xmath82 and @xmath83 db ; * the 3d model : the antenna gain of a transmitting base station is given by expressions ( [ eq : hv_pattern ] ) ( [ eq : h_pattern ] ) ( [ eq : v_pattern ] ) ; * analyzed scenarios corresponding to realistic situations in a network : * * urban environment : inter site distance isd = 200 m , 500 m and 750 m , * * antennas tilts : @xmath84 , @xmath85 , @xmath86 .
user equipments are randomly distributed in a cell of a 2d hexagonal based network ( fig .
[ hexagonaljpg ] ) .
this hexagonal network is equipped by antennas which have a given height ( 30 m and 50 m in our analysis ) , in the third dimension .
monte carlo simulations are done to calculate the sinr for each ue .
we focus our analysis on a typical hexagonal site .
the cumulative distribution function ( cdf ) of the sinr can be established by using these simulations .
these curves are compared to the ones established by using the analytical formula ( [ sinrfluid2 ] ) to calculate the sinr values . moreover , the sinr values established by the two ways are drawn on figures representing a site with three antennas .
we present two types of comparisons .
we first establish the cdf of the sinr .
indeed , the cdf of sinr provides a lot of information about the network characteristics : the coverage and the outage probability , the performance distribution , and the quality of service that can be reached by the system . as an example ,
figure 3 shows that for an outage probability target of 10% , the sinr reaches -8 db , which corresponds to a given throughput .
a second comparison , focused on the values of the sinr at each location of the cell , establishes a map of sinr over the cell . and
@xmath87 axes represent the coordinates , in meters .
the intersite distance in this example is 750 m. ] for the validation , we compare the two methods by considering realistic values of network parameters . an urban environment with realistic parameters of propagation is simulated @xcite .
different tilts and apertures are considered .
the scenarios , summarized in tab .
[ tab_scenariofigures ] , show that the 3d fluid analytical model and the simulations provide very close values of sinr : .scenarios and figures [ cols="^,^,^,^,^,^,^",options="header " , ] the figures of scenario 1 ( fig .
3 ) , scenario 2 ( fig . 5 and 6 ) ,
scenario 3 ( fig .
8) , scenario 4 ( fig .
10 ) , scenario 5 ( fig .
12 ) and scenario 6 ( fig .
14 ) show that the analytical model ( blue curves ) and the simulations ( red curves ) provide very close cdf of sinr curves .
the figures of scenario 1 ( fig . 4 ) ,
scenario 2 ( fig .
7 ) , scenario 3 ( fig .
9 ) , scenario 4 ( fig .
11 ) , scenario 5 ( fig . 13 ) and scenario 6 ( fig .
15 ) represent the values of sinr in each location of a cell , where the @xmath88 and @xmath87 axes represent the coordinates ( in meters ) .
these figures show that the analytical model ( right side ) and the simulations ( left side ) provide very close maps of sinr . , a vertical aperture @xmath89 and an horizontal aperture @xmath90 . ] , a vertical aperture @xmath89 and an horizontal aperture @xmath90 . ] , a vertical aperture @xmath89 and an horizontal aperture @xmath91 . ] ) where @xmath92 , @xmath89 and @xmath93 . ] , a vertical aperture @xmath89 and an horizontal aperture @xmath93 . ] , a vertical aperture @xmath89 and an horizontal aperture @xmath90 . ] , a vertical aperture @xmath89 and an horizontal aperture @xmath90 . ] , a vertical aperture @xmath89 and an horizontal aperture @xmath94 . ] , a vertical aperture @xmath89 and an horizontal aperture @xmath94 . ] , a vertical aperture @xmath95 and an horizontal aperture @xmath93 . ] , a vertical aperture @xmath95 and an horizontal aperture @xmath93 . ] , a vertical aperture @xmath89 and an horizontal aperture @xmath93 . ] , a vertical aperture @xmath89 and an horizontal aperture @xmath93 . ]
the aim of our analysis is to propose a model allowing to evaluate the performance , quality of service and coverage reachable in a cell whose standard antennas are replaced by 3d antennas , taking into account the height of the antenna and its tilt .
therefore the whole antenna energy is focused on the cell .
this implies that the angle @xmath48 has to be lower than @xmath47 , otherwise ues belonging to other cells could be served by this antenna .
the validation process was done according to this constraint .
however , the analytical closed - form formula ( [ sinrfluid2 ] ) allows to establish cdf of sinr very closed to simulated ones , for the different values of @xmath47 , vertical apertures @xmath48 and horizontal apertures @xmath96 , as soon as @xmath97 .
moreover , the sinr maps given by simulations and by the formula are also very closed .
therefore , the formula is particularly well adapted for 3d wireless networks analysis .
we develop , and validate , a three dimensional analytical wireless network model .
this model allows us to establish a closed form formula of the sinr reached by a ue at any location of a cell , for a 3d wireless network .
the validation of this model , by comparisons with monte carlo simulations results , shows that the two approaches establish very close results , in terms of cdf of sinr , and also in terms of sinr map of the cell .
this model may be used in the aim to analyze wireless networks , in a simple way , with a higher accuracy than a classical 2d approach .
jean - marc kelif , marceau coupechoux , philippe godlewski , a fluid model for performance analysis in cellular networks , eurasip journal on wireless communications and networking , vol .
2010 , article i d 435189 , doi:10.1155/2010/435189 | in this article we develop a three dimensional ( 3d ) analytical model of wireless networks .
we establish an analytical expression of the sinr ( signal to interference plus noise ratio ) of user equipments ( ue ) , by using a 3d fluid model approach of the network .
this model enables to evaluate in a simple way the cumulative distribution function of the sinr , and therefore the performance , the quality of service and the coverage of wireless networks , with a high accuracy .
the use of this 3d wireless network model , instead of a standard two - dimensional one , in order to analyze wireless networks , is particularly interesting .
indeed , this 3d model enables to establish more accurate performance and quality of services results than a 2d one . |
the last unknown mixing angle @xmath0 has been measured by daya bay @xcite@xcite and other reactor @xcite@xcite and accelerator @xcite@xcite experiments .
it opens a new era of neutrino experiments , to determine mass hierarchy and search for charge - parity ( cp ) violation in the neutrino oscillation .
the large value of @xmath0 makes both measurements easier .
there are a couple of proposed long baseline accelerator experiments@xcite @xcite@xcite which have abilities to measure both mass hierarchy and cp phase .
on the other hand , reactor experiments at a medium baseline can still play important roles , since the disappearance of reactor electron antineutrinos does nt reply on the cp phase , which can give a clean measurement of mass hierarchy .
besides , current and future reactor experiments have abilities to improve the precision of four mixing parameters significantly , which can be used to validate the unitarity of the mixing matrix .
it was pointed out the sensitivities of @xmath5 with 3 years data are 0.008 , 0.03 , 0.02 at 90% c.l . for daya bay , double chooz and reno , respectively @xcite .
daya bay released the new result of electron antineutrino oscillation amplitude and frequency based on a spectral measurement at the nufact 2013 conference@xcite @xcite .
the 1-@xmath6 precision of @xmath5 has been improved to around 10% , and the future precision can reach to be better than 4% .
daya bay also , for the first time , measured the neutrino mass splitting @xmath7 using reactor antineutrinos , and the result is consistent with the one from accelerator neutrinos @xcite .
reno s current precision of @xmath5 is about 16% @xcite and it s expected to be improved to 7% assuming the systematic uncertainty can reach to 0.005 .
the combined analysis of neutron capture on different targets in the double chooz far detector gives about 30% precision on @xmath5 and it s going to be improved significantly with near detector running in 2014@xcite .
mass hierarchy can be determined by the precision energy spectrum measurement of reactor neutrinos , and looking for the interference between two oscillation frequency components driven by @xmath8 and @xmath7 , respectively @xcite@xcite , where @xmath9 represents the mass difference of two neutrino mass eigenstates @xmath10 and @xmath11 . at a baseline of about 60 km ,
more than 20 oscillation cycles can be seen . in this case , a fourier transform of the @xmath12 spectrum , where @xmath13 is the baseline and @xmath14 is neutrino energy , can enhance the information of oscillation frequencies and thus improve the sensitivity of mass hierarchy @xcite@xcite . in the @xmath15 spectrum after the fourier transform , the relative amplitude and position of the peak and valley
give the discrimination power of mass hierarchy , without any precondition of the absolute measurement of @xmath8 or @xmath7 @xcite .
on the other hand , adding the absolute value of @xmath15 can bring an additional sensitivity .
a @xmath16 analysis shows that , with an improved measurement of @xmath17 ( which is an approximation of @xmath8 or @xmath7 ) by the accelerator experiments in the near future , we can achieve additional @xmath18 by @xmath3 9 and 4 for the 1% and 1.5% relative errors of the @xmath17 measurement@xcite .
the sensitivity is affected by the experimental conditions , such as the detector energy response and distribution of multiple reactor cores .
since the oscillation is very fast along with energy , the precision measurement of the energy spectrum requires @xmath1@xmath2 energy resolution and no less than 50,000 detected antineutrino interactions @xcite . in an ideal case of a single reactor core and a single detector ,
the baseline was optimized at about 52 km @xcite .
multiple reactors cores with different baselines can reduce the oscillation structure . according to ref .
@xcite , the baseline difference ca nt be more than 500 m. the energy nonlinearity is a challenge for the liquid scintillator detectors because it s difficult to determine the quenching effect and the cherenkov contribution . however , since there are multiple oscillations which can be observed in the @xmath12 spectrum , and each of them carries the same information of @xmath15 , this redundancy can be used to measure the energy scale at different energies , which is called self - calibration .
@xcite uses a quadratic function to fit the residual nonlinearity and shows that the sensitivity is almost not affected after the self - calibration .
the jiangmen underground neutrino observatory ( juno ) is a multipurpose neutrino - oscillation experiment designed to determine neutrino mass hierarchy and precisely measure oscillation parameters by detecting reactor antineutrinos from the yangjiang and taishan nuclear power plants , observe supernova neutrinos , study the atmospheric , solar neutrinos and geo - neutrinos , and perform exotic searches , with a 20 kiloton ( kton ) liquid scintillator detector of unprecedented 3% energy resolution ( at 1 mev ) at 700-meter deep underground and has other rich scientific possibilities .
the major goal of juno is determining neutrino mass hierarchy by precisely measuring the energy spectrum of reactor electron antineutrinos at a distance of 53 km from the reactors .
the relative measurement can reach to a sensitivity of @xmath4 in the idea case of a single reactor and a single detector , and @xmath19 considering the spread of reactor cores and uncertainties of the detector response .
if the absolute value of @xmath17 measured from accelerator experiments is included with a precision of 1% , the sensitivity of mass hierarchy can be improved to @xmath20 and @xmath4 in the idea and real case , respectively , as shown in figure [ juno_mh ] .
juno is going to improve the precision of @xmath21 , @xmath7 and @xmath22 to be better than 1% . considering the precision of @xmath23
can be measured to @xmath34% by daya bay , the unitarity of the neutrino mixing matrix can be probed to 1% level . the expected precision of mixing parameters by juno is listed in table [ mix_par ] . the mass hierarchy sensitivity for juno .
dash lines represent the relative measurement .
solid lines include the absolute value of @xmath17 with a precision of 1% . ]
lll & current&juno + @xmath21 & @xmath33% & @xmath30.6% + @xmath7 & @xmath35% & @xmath30.6% + @xmath22 & @xmath36% & @xmath30.7% + @xmath24 & @xmath320% & n / a + @xmath23 & @xmath34% in a near future & @xmath315% + juno is located in kaiping , jiangmen , in southern china , as shown in figure [ juno_site ] .
it s about 53 km from the yangjiang and taishan nuclear power plants , both of which are under construction .
the planned total thermal power of these reactors is 36 gw . there is no other nuclear power plant within 200 km .
a 270 m high granite mountain provides good shielding of cosmic muons , which are the major sources of backgrounds . to further suppress muon induced backgrounds , the detector is designed to be located deep underground through a tunnel , and the total overburden
will be 700 m rock .
experiment construction will start in 2013 and complete in 2019 , including a tunnel , an underground experiment hall , a water pool , a central detector , a muon tracking detector , and some ancillary facilities .
experiment site of juno .
the previous site is near daya bay .
that s the reason this experiment was known as daya bay ii .
the current site is in jiangmen city . ]
the central detector concept includes two concentric spherical tanks located in a water pool , as shown in figure [ juno_det ] .
the inner acrylic tank is filled with 20 kton linear alkylbenzene ( lab ) based liquid scintillator ( ls ) .
the outer stainless steel tank is filled with 6 kton mineral oil as buffer to protect ls from radioactivities .
there are around 15,000 20 " photomultiplier tubes ( pmts ) installed in the internal surface of the steel tank .
since it s extremely difficult to build both large tanks at the same time , there are other options of the detector design .
option 1 removes the steel tank .
the acrylic tank is directly put into water , which brings large pressure difference .
mineral oil is replaced by water in this option .
pmts can be installed in a steel frame in water .
option 2 removes the acrylic tank .
instead , small acrylic boxes filled with mineral oil can be made and installed as modules to contain single pmt or a group of pmts .
there are pipes at the back of each module for mineral oil filling and cabling .
the leakage through cables is the major concern .
option 3 uses a balloon to replace the acrylic tank .
balloon is relatively cheap for construction and quick for installation .
experiences from borexino and kamland are very encouraging .
there are many technical details of film materials need to be considered , such as the transparency , radon permeability , and the leak check .
option 4 is actually a fall back plan for option 3 .
if the balloon is failed , the liquid scintillator has to be directly filled into the steel tank . in this case ,
pmts are immersed in the liquid scintillator hence need special protection . on the other hand ,
since there is no mineral oil or water buffer , radioactivities from pmt will increase the trigger rate to more than 1 mhz , which ca nt be handled by the data acquisition system .
an online data reduction algorithm was developed to reduce the trigger rate to less than 1 khz based on the charge pattern detected by pmts .
however , the energy resolution is affected because of high probability of the overlap between antineutrino signals and radioactivities . a detector concept for juno . ]
the water pool protects the central detector from natural radioactivity in surrounding rocks .
it also serves as a water cherenkov detector after being equipped with pmts , to tag cosmic muons .
there is another muon tracking detector on top of the water pool , used to improve muon detection efficiency and to get better muon tracking .
3% energy resolution at 1 mev corresponds to 1,200 photon electrons per mev , which is a much better performance than the state of the art detector such as borexino@xcite and kamland@xcite .
the technical challenges are new type of pmts with high efficiency and highly transparent liquid scintillator .
the quantum efficiency of the photocathode made of super bialkali is expected to reach to 35% .
the traditional dynode can be replaced by the micro channel plate ( mcp ) , which has near 4@xmath25 acceptance , receiving not only the transmission light but also the refection light with reflection photocathode at the bottom of pmt , hence largely improve the collection efficiency . a prototype of mcp - pmt has been made and under test .
the ideal arrangement of 20 `` pmts can reach to more than 80% coverage . a mixture of 8 '' pmts and 20 " pmts was considered , which can reach to the similar coverage as the ideal case , while the smaller pmt can provide better timing for event vertex reconstruction .
an idea of adding reflection cones into the clearance was studied . with two thin acrylic panels with air gap , for uniformly distributed events ,
mc simulation shows @xmath36% increase on the total number of pe . besides , reflecting to local pmts wo nt impact on the vertex reconstruction . to reach a high transparency for the liquid scintillator ,
the production of lab has been improved and the attenuation length of the raw liquid reached to 20.5 m at 430 nm wavelength .
there are various of equipments produced for the purification , such as molecular distillation , vacuum distillation and filtration with the al@xmath26o@xmath27 column .
currently , the attenuation length is increased to 24 m at 430 nm .
other methods such as lowering the temperature or optimizing the fluor concentration were studied to improve the light yield .
the reactor electron antineutrino interacts with the proton via the inverse @xmath28-decay ( ibd ) reaction in the liquid scintillator , and releases a positron and a neutron .
the positron deposits its energy quickly , providing a prompt signal .
the energy of positron carries most of the kinetic energy of the neutrino .
the neutron is captured by a proton after an average time of 200 @xmath29s , then releases a 2.2 mev gamma , providing a delayed signal .
the coincidence of prompt - delayed signals provides a distinctive antineutrino signature .
the estimated ibd reaction rate is 40/day .
the dominated background is accidental coincidence , coming from two uncorrelated background radiation interactions that randomly satisfy the energy and time correlation for the ibd antineutrino selection .
it s designed to be less than 10% of ibd signals and can be precisely measured in data .
other major backgrounds are introduced by cosmic muons , including cosmogenic @xmath28-n isotope @xmath30li/@xmath31he and fast neutrons . both of them are less than 1% after appropriate muon veto .
juno got great support from chinese academy of sciences with the strategic priority research program , which has been approved on february 1 , 2013 .
the construction period is from 2013 to 2019 . and the experiment will be in operation in 2020 .
there were two get - together meetings in january and july in 2013 , and the next meeting will be in the experiment site jiangmen the next january .
collaborators are welcome .
reno-50 was proposed in south korea@xcite .
it plans an underground detector consisting of 18 kton ultra - low - radioactivity liquid scintillator and 15,000 20 " pmts , at 50 km away from the hanbit nuclear power plant .
the scientific goals include high precision measurement of @xmath32 and @xmath33 , determination of mass hierarchy , study neutrinos from reactors , the sun , the earth , supernova , and any possible stellar objects .
the total budget is 100 million dollars for 6 years construction .
facility and detector construction will be from 2013 to 2018 , and data taking will be started in 2019 .
ongoing reactor neutrino experiments in the world played important roles in measuring the smallest neutrino mixing angle @xmath0 .
they will continue running to improve the precision of @xmath34 to be better than 4% .
next generation reactor experiments will focus on mass hierarchy determination and precision measurement of mixing parameters .
science case is strong with significant technical challenges .
juno was proposed a few years ago ( known as daya bay ii ) , now boosted by the large @xmath0 .
funding has been approved from chinese academy of sciences .
a similar proposal was from korea known as reno-50 .
both of them are going to start data taking around 2020 .
an _ et al . _
[ daya bay collaboration ] , phys .
rev . lett . * 108 * , 171803 ( 2012 ) .
[ daya bay collaboration ] , chin .
c * 37 * , 011001 ( 2013 ) .
[ reno collaboration ] , phys .
lett . * 108 * , 191802 ( 2012 ) .
[ double chooz collaboration ] , phys .
lett . * 108 * , 131801 ( 2012 ) .
[ t2k collaboration ] , phys .
lett . * 107 * , 041801 ( 2011 ) .
p. adamson _
. _ [ minos collaboration ] , phys . rev . lett . * 107 * , 181802 ( 2011 ) .
et al . _ [ hyperk working group ] , arxiv:1109.3262
. t. akiri _
[ lbne collaboration ] , arxiv:1110.6249 .
s. bertolucci _
et al . _ ,
_ , jhep 0911:044 ( 2009 ) .
s. kettell , plenary talk at nufact 2013 , beijing .
s. jetter , parallel talk at nufact 2013 , beijing .
p. adamson _
[ minos collaboration ] , phys .
* 110 * , 251801 ( 2013 ) .
h. seo , talk at win 2013 , natal , brazil .
r. sharankova , parallel talk at nufact 2013 , beijing .
s. t. petcov and m. piai , phys .
b * 533 * , 94 ( 2002 ) .
s. choubey , s. t. petcov , and m. piai , phys .
d * 68 * , 113006 ( 2003 ) .
l. zhan , y. wang , j. cao , and l. wen , phys .
d * 78 * , 111103(r ) ( 2008 ) .
j. learned , s. dye , s. pakvasa , and r. svoboda , phys .
d * 78 * , 071302 ( 2008 ) .
yu - feng li , jun cao , yifang wang , and liang zhan , phys .
d * 88 * , 013008 ( 2013 ) .
l. zhan , y. wang , j. cao , and l. wen , phys .
d * 79 * , 073007 ( 2009 ) .
g. alimonti , _ et al .
_ , [ borexino collaboration ] , astropart.phys . * 16 * 205 - 234 ( 2002 ) .
k. eguchi _
et al . _ , [ kamland collaboration ] , phys .
lett . * 90 * 021802 ( 2003 ) .
soo - bong kim , talk at the international workshop on reno-50 , june 13 - 14 , 2013 , seoul . | the measurement of the neutrino mixing angle @xmath0 opens a gateway for the next generation experiments to measure the neutrino mass hierarchy and the leptonic cp - violating phase .
future reactor experiments will focus on mass hierarchy determination and the precision measurement of mixing parameters .
mass hierarchy can be determined from the disappearance of reactor electron antineutrinos based on the interference effect of two separated oscillation modes .
relative and absolute measurement techniques have been explored .
a proposed experiment juno , with a 20 kton liquid scintillator detector of @xmath1@xmath2 energy resolution , @xmath3 53 km far from reactors of @xmath3 36 gw total thermal power , can reach to a sensitivity of @xmath4 considering the spread of reactor cores and uncertainties of the detector response .
three of mixing parameters are expected to be measured to better than 1% precision .
there are multiple detector options for juno under investigation .
the technical challenges are new type of pmts with high efficiency and highly transparent liquid scintillator .
funding has been approved from chinese academy of sciences .
a similar proposal was from korea known as reno-50 .
both of them are going to start data taking around 2020 . |
geometrical - optical illusions ( gois ) have been discovered in the xix century by german psychologists ( oppel 1854 @xcite , hering , 1878 , @xcite ) and have been defined as situations in which there is an awareness of a mismatch of geometrical properties between an item in object space and its associated percept @xcite .
the distinguishing feature of these illusions is that they relate to misjudgements of geometrical properties of contours and they show up equally for dark configurations on a bright background and viceversa . for the interested reader , a historical survey of the discovery of geometrical - optical illusions
is included in appendix i of @xcite .
our intention here is not to make a classification of these phenomena , which is already widely present in literature ( coren e girgus , 1978 , @xcite ; robinson , 1998 , @xcite ; wade , 1982 , @xcite ) .
the aim of this paper is to propose a mathematical model for gois based on the functional architecture of low level visual cortex ( v1/v2 ) .
this neuro - mathematical model will allow us to interpret at a neural level the origin of gois and to reproduce the arised percept for this class of phenomena .
the main idea is to adopt the model of the functional geometry of v1 provided in @xcite and to consider that the image stimulus will modulate the connectivity .
when projected onto the visual space , the modulated connectivity gives rise to a riemannian metric which is at the origin of the visual space deformation .
the displacement vector field at every point of the stimulus is mathematically computed by solving a poisson problem and the perceived image is finally reproduced .
the considered phenomena consist , as shown in figure [ fig:1ills ] , in straight lines over different backgrounds ( radial lines , concentric circles , etc ) .
the interaction betwen target and context either induces an effect of curvature of the straight lines ( fig .
[ fig:1:4 ] , [ fig:1:5 ] , [ fig:1:1 ] ) , eliminates the bending effect ( fig .
[ fig:1:6 ] ) , or induces an effect of unparallelism ( fig . [ fig:1:2 ] ) .
the paper is organised as follows : in section [ sec:1 ] we review the state of the art concerning the previous mathematical models proposed . in section [ sec:2 ]
we will briefly recall the functional achitecture of the visual cortex and the cortical based model introduced by citti and sarti in @xcite . in section [ sec:3bis ] we will introduce the neuro - mathematical model proposed for gois , taking into account the modulation of the functional architecture induced by the stimulus . in [ sec:3 ] the numerical implementation of the mathematical model
will be explained and applied to a number of examples .
results are finally discussed . 1.8 in 2.1 in 1.8 in 2.1 in 1.8 in
in psychology the distal stimulus is defined as _ the light reflected off a physical object in the external world _ ; when we look at an image ( distal stimulus ) we can not actually experience the image physically with vision , we can only experience it in our mind as proximal stimulus @xcite , @xcite .
geometrical optical illusions arise when the distal stimulus and its percept differ in a perceivable way . as explained by westheimer in @xcite
, we can conveniently divide illusions into those in which spatial deformations are a consequence of the exigencies of the processing in the domain of brightness and the true geometrical - optical illusions , which are misperceptions of geometrical properties of contours in simple figures .
some of the most famous geometric illusions of this last type are shown in figure [ fig:1ills ] .
the importance of this study lies in the possibility , through the analysis of these phenomena combined with physiological recordings , to help to guide neuroscientific research ( eagleman , @xcite ) in understanding the role of lateral inhibition , feedback mechanisms between different layers of the visual process and to lead new experiments and hypothesis on receptive fields of v1 and v2 .
many studies , which relies on neuro - physiological and imaging data , show the evidence that neurons in at least two visual areas , v1 and v2 , carry signals related to illusory contours , and that signals in v2 are more robust than in v1 ( @xcite , @xcite , reviews @xcite , @xcite ) . a more recent study on the tilt illusion , in which the perceived orientation of a grating differs from its physical orientation when surrounded by a tilted context , measured the activated connectivity in and between areas of early visual cortices ( @xcite ) .
these findings suggest that for gois these areas may be involved as well .
neurophysiology can help to provide a physical basis to phenomenological experience of gois opening to the possibility of mathematically modeling them and to integrate subjective and objective experiences . the pioneering work of hoffman @xcite dealt with illusions of angle ( i.e. the ones involving the phenomenon of angular expansion , which is the tendence to perceive under certain conditions acute angles as larger and obtuse ones as smaller ) modeling the generated perceived curves as orbits of a lie transformation group acting on the plane .
the proposed model allows to classify the perceptual invariance of the considered phenomena in terms of lie derivatives , and to predict the slope .
another model mathematically equivalent to the one proposed by hoffman has been proposed by smith , @xcite , who stated that the apparent curve of geometrical optical illusions of angle can be modeled by a first - order differential equation depending on a single parameter . by computing this value an apparent curve
can be corrected and plotted in a way that make the illusion being not perceived anymore ( see for example fig . 8 of @xcite ) .
this permits to introduce a _ quantitative _ analysis of the perceived distortion .
ehm and wackerman in @xcite , started from the assumption that gois depend on the context of the image which plays an active role in altering components of the figure . on this basis they provided a variational approach computing the deformed lines as minima of a functional depending on length of the curve and the deflection from orthogonality along the curve .
this last request is in accordance to the phenomenological property of regression to right angle .
one of the problems pointed out by the authors is that the approach does nt take into account the underlying neurophysiological mechanisms .
an entire branch for modeling neural activity , the bayesian framework , had its basis in helmholtz s theory @xcite : _ our percepts are our best guess as to what is in the world , given both sensory data and prior experience . _
the described idea of unconscious inference is at the basis of the bayesian statistical decision theory , a principled method for determining optimal performance in a given perceptual task ( @xcite ) .
these methods consists in attributing a probability to each possible true state of the environment given the stimulus on the retina and then to establish the way prior experience influences the final guess , the built proximal stimulus ( see @xcite for examples of bayesian models in perception ) .
an application of this theory to motion illusions has been provided by weiss et al in @xcite , and a review in @xcite .
fermller and malm in @xcite attributed the perception of geometric optical illusions to the statistics of visual computations .
noise ( uncertainty of measurements ) is the reason why systematic errors occur in the estimation of the features ( intensity of the image points , of positions of points and orientations of edge elements ) and illusions arise as results of errors due to quantization .
walker ( @xcite ) tried to combine neural theory of receptive field excitation together with mathematical tools to provide an equation able to determine the disparity between the apparent line of an illusion and its corresponding actual line , in order to reproduce the perceptual errors that occur in gois ( the ones involving straight lines ) . in our model we aim to combine psycho - physical evidence and neurophysiological findings , in order to provide a neuro - mathematical model able to interpret and simulate gois .
the visual process is the result of several retinic and cortical mechanisms which act on the visual signal .
the retina is the first part of the visual system responsible for the trasmission of the signal , which passes through the lateral geniculate nucleus , where a pre - processing is performed and arrives in the visual cortex , where it is further processed .
the receptive field ( rf ) of a cortical neuron is the portion of the retina which the neuron reacts to , and the receptive profile ( rp ) @xmath0 is the function that models the activation of a cortical neuron when a stimulus is applied to a point @xmath1 of the retinal plane .
simple cells of visual cortices v1 and v2 are sensitive to position and orientation of the contrast gradient of an image .
their properties have been experimentally described by de angelis in @xcite , see figure [ fig:2e ] . from the neurophysiological point of view the orientation selectivity ,
the spatial and temporal frequency of cells in v2 differs little from the one in v1 ( @xcite ) .
receptive fields in v2 are larger from those in v1 ( @xcite , @xcite ) . considering a basic geometric model , the set of simple cells
rps can be obtained via translations of vector @xmath2 and rotation of angle @xmath3 from a unique mother profile @xmath4 .
daugman @xcite , jones and palmer @xcite showed that gabor filters were a good approximation for receptive profiles of simple cells in the primary visual cortices v1 and v2 .
another approach is to model receptive profiles as gaussian derivatives , as introduced by young in @xcite and koenderink in @xcite , but for our purposes the two approaches are equivalent .
a good expression for the mother gabor filter is : @xmath5 where @xmath6 is the ratio between @xmath7 and the spatial wavelength of the cosine factor .
translations and rotations can be expressed as : @xmath8 hence a general rp can be expressed as : @xmath9 a set of rps generated with equation [ group_law ] is shown in figure [ fig43332 ] .
the retinal plane is identified with the @xmath10-plane , whose local coordinates will be denoted with @xmath11 . when a visual stimulus @xmath12 of intensity @xmath13 activates the retinal layer of photoreceptors , the neurons whose rfs intersect @xmath14 spike and their spike frequencies @xmath15 can be modeled ( taking into account just linear contributions ) as the integral of the signal @xmath16 with the set of gabor filters .
the expression for this output is : @xmath17 in the right hand side of the equation the integral of the signal with the real and imaginary part of the gabor filter is expressed .
the two families of cells have different shapes , hence they detect different features
. in particular odd cells will be responsible for boundary detection .
the term _ functional architecture _ refers to the organisation of cells in the primary visual cortex in structures .
the hypercolumnar structure , discovered by the neuro - physiologists hubel and wiesel in the 60s ( @xcite ) , organizes the cells of v1/v2 in columns ( called hypercolums ) covering a small part of the visual field @xmath18 and corresponding to parameters such as orientation , scale , direction of movement , color , for a fixed retinal position @xmath19 .
we have the set of all possible orientations , title="fig : " ] we have the set of all possible orientations , title="fig : " ] in our framework over each retinal point we will consider a whole hypercolumn of cells , each one sensitive to a specific instance of orientation . hence for each position @xmath19 of the retina @xmath18 we associate a whole set of filters @xmath20 this expression associates to each point of the proximal stimulus in @xmath10 all possible feature orientations into the space of features @xmath21 , and defines a fiber over each point @xmath22 in this way the hypercolumnar structure is described in terms of differential geometry , but we need to explain how the orientation selectivity is performed by the cortical areas in the space of feature @xmath23 ( @xcite ) .
physiologically the orientation selectivity is the action of short range connections between simple cells belonging to the same hypercolumn to select the most probable response from the energy of receptive profiles .
horizontal connections are long ranged and connect cells of approximately _ the same orientation_. since the connectivity between cells is defined on the tangent bundle , we define now the generator of this space .
the change of variable defined through @xmath24 in acts on the basis for the tangent bundle @xmath25 giving as frame in polar coordinates : @xmath26 as presented in @xcite , the whole space of features @xmath27 is described in terms of a 3-dimensional fiber bundle , whose generators are @xmath28 , @xmath29 for the base and @xmath30 for the fiber .
these vector fields generate the tangent bundle of @xmath31 .
since horizontal connectivity is very anysotropic , the three generators are weighted by a strongly anysotropic metric .
we introduce now the sub - riemannian metric with whom citti and sarti in @xcite proposed to endow the @xmath31 group to model the long range connectivity of the primary visual cortex v1 . starting from the vector fields @xmath28 , @xmath32 and @xmath29 we define a metric @xmath33 for which the inverse ( responsible for the connectivity in the cortex ) is : @xmath34 with @xmath35 .
cortical curves in v1 will be a linear combination of vector fields @xmath28 and @xmath32 , the generators of the 2-dimensional horizontal space , while they will have a vanishing component along @xmath36 $ ] .
the functional architectures built in @xmath31 correspond to the neural connectivity measured by angelucci et al . in @xcite and bosking et al . in @xcite . for a qualitatively and quantitative comparison between the kernels and the connectivity patterns
see favali et al . in @xcite . in our work a local formulation of the kernel presented in @xcite will be used .
furthermore , it has been shown by sanguinetti et al . in @xcite
that the geometry of fuctional architecture formally introduced in [ sec:2 ] is naturally encoded in the statistics of natural images .
hence these geometrical structures are compatible with bayesian learning methods .
we consider simple cells at fixed value of @xmath7 depending on position and orientation . for fixed value of @xmath19 , we restrict the connectivity tensor @xmath37 to the @xmath10 plane , subset of the tangent plane to @xmath31 at the point @xmath27 , and obtain the tensor @xmath38 for every value of @xmath3 this tensor has only one non zero eigenvalue .
the corresponding eigenvector has direction @xmath3 .
we will assign to the norm of the output the usual meaning of energy @xmath39 where the output is defined in and is evaluated at the fixed value of @xmath7 .
we will discuss in section [ sec:3 ] the choice of @xmath7 for our experiments .
each point of the hypercolumn is weighted by the energy of simple cells normalized over the whole set of hypercolumn responses : @xmath40 the normalization of the output expresses the probability that a specific cell sensitive to @xmath3 within the hypercolumn over @xmath19 is selected .
the mechanism of intracortical selection attributing a probability to each possible orientation ( state ) given the initial stimulus is connected to the long - range activity : simple cells belonging to different hypercolumns in a neighbourhood of a point @xmath19 sensitive to the same orientation will have a high probability .
the connectivity tensor restricted to the @xmath10 plane and modulated by the output of simple cells will become : @xmath41 this last expression corresponds to a connectivity polarized by the normalized energy of simple cells shown in at points @xmath42 .
the overall cometric ( inverse of the metric tensor ) arising from the action within the hypercolumn over each retinal point @xmath19 is obtained summing up along @xmath3 the previous modulated metric in and will have the following expression : @xmath43 where @xmath44 is a constant factor .
this tensor will have principal eigenvector in the direction @xmath45 , the orientation corresponding to the maximum energy within the hypercolumn .
a visualization of @xmath46 is given in figure [ her_fig ] .
hence this process describes the selection at every point @xmath19 of the most likely direction of propagation of the connectivity , expressed by the values attained by the energy . 1.8 in 1.8 in 1.8 in 2.1 in 2.1 in 2.1 in in the previous section we described the response of the cortex in the presence of a visual stimulus .
* the distal stimulus is projected onto the cortex by means of activity of simple cells . *
the joint action of the short and long range connectivity induces a riemannian tensor @xmath46 on the @xmath10 retinal plane . even though it is not completely clear in which cortical area the perceived image is reconstructed , from a phenomenological point of view it is evident that our visual system recostructs the perceived image .
hence a third mechanism takes place , able to construct the perceived stimulus from the cortical activation . with this mechanism the image distortion which induces the metric tensor @xmath47 ( inverse of @xmath46 )
is estimated . here
we propose to apply infinitesimal strain theory and to identify its inverse @xmath47 with the strain tensor to compute the deformation .
once the displacement vector field is applied to the distal stimulus , we obtain a distorted image which models the proximal one . in this way
we justify the mechanism at the basis of geometrical optical illusions . in this approach
we consider the medium to be subjected only to small displacements , i.e. the geometry of the medium and its constitutive properties at each point of the space are assumed to be unchanged by deformation .
the mathematical question is how to reconstruct the displacement starting from the strain tensor @xmath47 .
we think at the deformation induced by a geometrical optical illusion as a map between the @xmath10 plane equipped with the metric @xmath47 and the @xmath10 plane with the euclidean metric @xmath48 : @xmath49 from the mathematical point of view this means that we look for the change of variable which induces the new metric , i.e. @xmath50 with @xmath51 ( see jost @xcite ) , obtaining the relation : @xmath52 let us notice @xmath46 corresponds to @xmath53 , the map representing the process which builds the modulated connectivity we discussed before . in strain theory
@xmath47 satisfying is called _ right cauchy - green tensor _ associated to the deformation @xmath54 , which from the physical point of view is a map @xmath55 associating the points of the closure of a bounded open set @xmath56 ( initial configuration of a body ) to @xmath57 ( deformed configuration ) . for references
see @xcite , @xcite .
it is possible to introduce the displacement as a map @xmath58 , where @xmath59 .
it follows @xmath60 we can now express the right cauchy - green tensor in terms of displacement : @xmath61 the concept of strain is used to evaluate how much a given displacement differs locally from a rigid body displacement .
one of the strain tensors for large deformations is the so called _ green - lagrangian strain tensor or green - saint venant strain tensor _ defined as : @xmath62 which can be written in terms of the displacement as before : @xmath63 for _ infinitesimal deformations _ of a continuum body , in which the displacement gradient is small ( @xmath64 ) , it is possible to perform a geometric linearization of strain tensors introduced before , in which the non - linear second order terms are neglected .
linearized green - saint venant tensor _ has the following form : @xmath65 which is used in the study of linearized elasticity , i.e. the study of such situations in which the displacements of the material particles of a body are assumed to be small ( as stated at the beginning , infinitesimal strain theory . )
+ here we give the expression in components of @xmath66 : @xmath67 where @xmath68 . expressing @xmath69 in terms of the metric @xmath70 with whom
the initial configuration of the considered body was equipped we obtain : @xmath71 and in its matrix form : @xmath72
starting from we obtain a system of equations with this form : @xmath73 differentiating , substituting and imposing neumann boundary conditions to system we end up with the following differential system : @xmath74 let us explicitly note that tensor @xmath47 is obtained after convolution of gabor filters , so that it is differentiable , allowing to write the system .
hence we solve , recovering the displacement field @xmath75 .
the inverse of tensor expressed in formula is computed discretizing @xmath3 as a vector of 32 values equally spaced in the interval @xmath76 $ ] .
the scale parameter @xmath7 varies in dependence of the image resolution and is set in concordance with the stimulus processed .
it is taken quite large in all examples in such a way to obtain a smooth tensor field covering all points of the image .
this is in accordance with the hypothesis previously introduced that mechanisms in v2 , where the receptive field of simple cells is larger than in v1 , play a role in such phenomena .
the constant @xmath77 has been chosen for all the examples as @xmath78 .
the differential problem in is approximated with a central finite difference scheme and it is solved with a classical pde linear solver .
we now start discussing all results obtained through the presented algorithm .
1.6 in 1.6 in 1.6 in 1.6 in the hering illusion , introduced by hering , a german physiologist , in 1861 @xcite is presented in figure [ fig:1:4 ] . in this illusion two
vertical straight lines are presented in front of radial background , so that the lines appear as if they were bowed outwards . in order to help the reader , in figure [ fig:11:1 ] we superpose to the initial illusion two red vertical lines , which indeed coincide with the ones present in the stimulus .
as described in the previous sections , we first convolve the distal stimulus with the entire bank of gabor filters : we take 32 orientations selected in @xmath79 , @xmath80 pixels . following the process , we compute @xmath46 using equation
, we solve equation obtaining the perceived displacement @xmath81 .
once it is applied to the initial stimulus , the proximal stimulus is recovered .
the result of computation is shown in figure [ her_5_fig1 ] .
the distorted image folds the parallel lines ( in black ) against the straight lines ( in red ) of the original stimulus ( figure [ fig:11:4 ] ) .
a variant of the hering illusion , introduced by wundt in the 19th century , @xcite is presented in figure [ fig:1:5 ] . in this illusion
two straight horizonal lines look as if they were bowed inwards , due to the distortion induced by the crooked lines on the background . for the convolution of the distal stimulus with gabor filters we select 32 orientations in @xmath79 , @xmath82 pixels .
then we apply the previous model , and obtain the result presented in figure [ wundt_sec5 ] .
computed vector fields are concentrated in the central part of the image and point toward the center .
they indicate the direction of the displacement , which bends the parallel lines inwards . in figure [ fig:15:4 ]
the proximal stimulus is computed through the expression : @xmath83 . in black
we indicate displaced dots of the initial details of the distances between the bent curves and the original straight lines are shown .
1.6 in , in order to clarify that the horizontal lines present in the image are indeed straight .
representation of @xmath46 , projection onto the retinal plane of the polarized connectivity in [ ptensor ] .
the first eigenvalue has direction tangent to the level lines of the distal stimulus . in blue the tensor field , in cyan the eigenvector related to the first eigenvalue .
computed displacement field @xmath84 .
displacement applied to the image . in black
we represent the proximal stimulus as displaced points of the distal stimulus : @xmath83 . in red
we give two straight lines as reference , in order to put in evidence the curvature of the target lines , title="fig : " ] 1.6 in , in order to clarify that the horizontal lines present in the image are indeed straight .
representation of @xmath46 , projection onto the retinal plane of the polarized connectivity in [ ptensor ] .
the first eigenvalue has direction tangent to the level lines of the distal stimulus . in blue the tensor field , in cyan the eigenvector related to the first eigenvalue .
computed displacement field @xmath84 .
displacement applied to the image . in black
we represent the proximal stimulus as displaced points of the distal stimulus : @xmath83 . in red
we give two straight lines as reference , in order to put in evidence the curvature of the target lines , title="fig : " ] 1.6 in , in order to clarify that the horizontal lines present in the image are indeed straight .
representation of @xmath46 , projection onto the retinal plane of the polarized connectivity in [ ptensor ] .
the first eigenvalue has direction tangent to the level lines of the distal stimulus . in blue the tensor field , in cyan the eigenvector related to the first eigenvalue .
computed displacement field @xmath84 .
displacement applied to the image .
in black we represent the proximal stimulus as displaced points of the distal stimulus : @xmath83 . in red
we give two straight lines as reference , in order to put in evidence the curvature of the target lines , title="fig : " ] 1.6 in , in order to clarify that the horizontal lines present in the image are indeed straight .
representation of @xmath46 , projection onto the retinal plane of the polarized connectivity in [ ptensor ] .
the first eigenvalue has direction tangent to the level lines of the distal stimulus . in blue the tensor field , in cyan the eigenvector related to the first eigenvalue .
computed displacement field @xmath84 .
displacement applied to the image . in black
we represent the proximal stimulus as displaced points of the distal stimulus : @xmath83 . in red
we give two straight lines as reference , in order to put in evidence the curvature of the target lines , title="fig : " ] this illusion , introduced by ehm and wackermann in @xcite , consists in presenting a square over a background of concentric circles , figure [ fig:1:1 ] .
this context , the same we find in ehrenstein illusion , bends the edges of the square ( red lines in [ fig:19:1 ] ) toward the center of the image .
here we take the same number of orientations , 32 , selected in @xmath79 and @xmath85 pixels .
the resulting distortion is shown in figure [ fig:19:4 ] . 1.6 in .
representation of @xmath46 , projection onto the retinal plane of the polarized connectivity in [ ptensor ] .
the first eigenvalue has direction tangent to the level lines of the distal stimulus . in blue the tensor field , in cyan the eigenvector related to the first eigenvalue .
computed displacement field @xmath84 .
displacement applied to the image . in black
we represent the proximal stimulus as displaced points of the distal stimulus : @xmath83 . in red
we give a square as reference , in order to put in evidence the curvature of the target lines , title="fig : " ] 1.6 in .
representation of @xmath46 , projection onto the retinal plane of the polarized connectivity in [ ptensor ] .
the first eigenvalue has direction tangent to the level lines of the distal stimulus . in blue the tensor field , in cyan the eigenvector related to the first eigenvalue .
computed displacement field @xmath84 .
displacement applied to the image . in black
we represent the proximal stimulus as displaced points of the distal stimulus : @xmath83 . in red
we give a square as reference , in order to put in evidence the curvature of the target lines , title="fig : " ] 1.6 in .
representation of @xmath46 , projection onto the retinal plane of the polarized connectivity in [ ptensor ] .
the first eigenvalue has direction tangent to the level lines of the distal stimulus . in blue the tensor field , in cyan the eigenvector related to the first eigenvalue .
computed displacement field @xmath84 .
displacement applied to the image . in black
we represent the proximal stimulus as displaced points of the distal stimulus : @xmath83 . in red
we give a square as reference , in order to put in evidence the curvature of the target lines , title="fig : " ] 1.6 in .
representation of @xmath46 , projection onto the retinal plane of the polarized connectivity in [ ptensor ] .
the first eigenvalue has direction tangent to the level lines of the distal stimulus . in blue the tensor field , in cyan the eigenvector related to the first eigenvalue .
computed displacement field @xmath84 .
displacement applied to the image . in black
we represent the proximal stimulus as displaced points of the distal stimulus : @xmath83 . in red
we give a square as reference , in order to put in evidence the curvature of the target lines , title="fig : " ] here we present three modified hering illusions ( see figure [ mod_her ] ) : in the first one straight lines are positioned further from the center than in the classical hering illusion . in the second one straight lines
are positioned nearer the center than in the reference hering illusion .
for coherence with the hering example , orientations selected are 32 in @xmath79 and @xmath80 pixels .
all other parameters are fixed during these three experiments . 1.6 in 1.6 in 1.6 in 1.6 in in the proposed modified
hering illusions the vertical lines are straight and parallel as in the hering , but since they are located further / nearer the center of the image the perceived bending results to be less / more intense .
in accordance with the displacement vector fields shown in figure [ fig:11:3 ] , as far as we outstrip / approach the center the magnitude of the computed displacement decreases / increases . in figure [ fig:22:4 ]
two straight lines are put over an incoherent background , composed by random oriented segments .
as we can see from [ fig:23:4 ] , any displacement is perceived nor computed by the present algorithm .
1.6 in 1.6 in 1.6 in 1.6 in the wundt - hering illusion ( figure [ fig:1:6 ] ) combines the effect of the background of the hering and wundt illusions . in this illusion
two straight horizontal lines are presented in front of inducers which bow them outwards and inwards at the same time , inhibiting the bending effect . as a consequence
the horizontal lines are indeed perceived as straight . as
previously explained for the modified hering illusion , also this phenomenon can be interpreted in terms of lateral interaction between cells belonging to the same neighborhood .
here we take 32 orientations selected in the interval @xmath79 , @xmath80 pixels . 1.6 in .
representation of @xmath46 , projection onto the retinal plane of the polarized connectivity in [ ptensor ] .
the first eigenvalue has direction tangent to the level lines of the distal stimulus . in blue the tensor field , in cyan the eigenvector related to the first eigenvalue .
computed displacement field @xmath84 .
displacement applied to the image . in black
we represent the proximal stimulus as displaced points of the distal stimulus : @xmath83 . in red
we give two straight lines as reference , in order to put in evidence the curvature of the target lines , title="fig : " ] 1.6 in .
representation of @xmath46 , projection onto the retinal plane of the polarized connectivity in [ ptensor ] .
the first eigenvalue has direction tangent to the level lines of the distal stimulus . in blue the tensor field , in cyan the eigenvector related to the first eigenvalue .
computed displacement field @xmath84 .
displacement applied to the image . in black
we represent the proximal stimulus as displaced points of the distal stimulus : @xmath83 . in red
we give two straight lines as reference , in order to put in evidence the curvature of the target lines , title="fig : " ] 1.6 in .
representation of @xmath46 , projection onto the retinal plane of the polarized connectivity in [ ptensor ] .
the first eigenvalue has direction tangent to the level lines of the distal stimulus . in blue the tensor field , in cyan the eigenvector related to the first eigenvalue .
computed displacement field @xmath84 .
displacement applied to the image . in black
we represent the proximal stimulus as displaced points of the distal stimulus : @xmath83 . in red
we give two straight lines as reference , in order to put in evidence the curvature of the target lines , title="fig : " ] 1.6 in .
representation of @xmath46 , projection onto the retinal plane of the polarized connectivity in [ ptensor ] .
the first eigenvalue has direction tangent to the level lines of the distal stimulus . in blue the tensor field , in cyan the eigenvector related to the first eigenvalue .
computed displacement field @xmath84 .
displacement applied to the image . in black
we represent the proximal stimulus as displaced points of the distal stimulus : @xmath83 . in red
we give two straight lines as reference , in order to put in evidence the curvature of the target lines , title="fig : " ] the zllner illusion ( figure [ fig:1:2 ] ) consists in a pattern of oblique segments surrounding parallel lines , which creates the effect of unparallelism . as in the previous experiments , in figure [ fig:27:1 ] we superimpose two red lines to identify the straight lines . here
we take 32 orientations selected in the interval @xmath79 , @xmath86 pixels .
representation of @xmath46 , projection onto the retinal plane of the polarized connectivity in [ ptensor ] .
the first eigenvalue has direction tangent to the level lines of the distal stimulus . in blue the tensor field , in cyan the eigenvector related to the first eigenvalue .
computed displacement field @xmath84 .
displacement applied to the image . in black
we represent the proximal stimulus as displaced points of the distal stimulus : @xmath83 . in red
we give two straight lines as reference , in order to put in evidence the unparallelism of the target lines , title="fig : " ] 1.0 in .
representation of @xmath46 , projection onto the retinal plane of the polarized connectivity in [ ptensor ] .
the first eigenvalue has direction tangent to the level lines of the distal stimulus . in blue the tensor field , in cyan the eigenvector related to the first eigenvalue .
computed displacement field @xmath84 .
displacement applied to the image . in black
we represent the proximal stimulus as displaced points of the distal stimulus : @xmath83 . in red
we give two straight lines as reference , in order to put in evidence the unparallelism of the target lines , title="fig : " ] + 1.0 in .
representation of @xmath46 , projection onto the retinal plane of the polarized connectivity in [ ptensor ] .
the first eigenvalue has direction tangent to the level lines of the distal stimulus . in blue the tensor field , in cyan the eigenvector related to the first eigenvalue .
computed displacement field @xmath84 .
displacement applied to the image . in black
we represent the proximal stimulus as displaced points of the distal stimulus : @xmath83 . in red
we give two straight lines as reference , in order to put in evidence the unparallelism of the target lines , title="fig : " ] 1.0 in .
representation of @xmath46 , projection onto the retinal plane of the polarized connectivity in [ ptensor ] .
the first eigenvalue has direction tangent to the level lines of the distal stimulus . in blue the tensor field , in cyan the eigenvector related to the first eigenvalue .
computed displacement field @xmath84 .
displacement applied to the image . in black
we represent the proximal stimulus as displaced points of the distal stimulus : @xmath83 . in red
we give two straight lines as reference , in order to put in evidence the unparallelism of the target lines , title="fig : " ] [ fig:27:4 ]
in this paper we presented a neuro - mathematical model based on the functional architecture of the visual cortex to explain and simulate perceptual distortion due to geometrical - optical illusions and to embed geometrical context . in our model perceptual distortion
is due to the riemannian metric induced on the image plane by the connectivity activated by the image stimulus .
its inverse is interpreted as a strain tensor and we computed the deformation in terms of displacement field which arises as solution of .
this technique has been applied to a number of test cases and results are qualitatively in good agreement with human perception . in the future
this work could be extended to functional architectures involving the feature of scale , starting from models provided by sarti , citti and petitot in @xcite , @xcite .
this will allow to provide a model for scale illusions , such as the delbouf , see @xcite .
indeed , another direction for future works will be to provide a quantitative analysis for the described phenomena , such as the one proposed by smith @xcite and to direct compare the developed theory with observations of gois through neuro - imaging techniques .
this project has received funding from the european union s seventh framework programme , marie curie actions- initial training network , under grant agreement n. 607643 , `` metric analysis for emergent technologies ( manet ) '' .
we would like to thank b. ter haar romeny , university of technology eindhoven , and m. ursino , university of bologna , for their important comments and remarks to the present work .
angelucci , a. , levitt , j.b . , walton , e.j .
, hupe , j.m . , bullier , j. , lund , j.s . : circuits for local and global signal integration in primary visual cortex .
the journal of neuroscience * 22*(19 ) , 86338646 ( 2002 ) bosking , w.h . , zhang , y. , schofield , b. , fitzpatrick , d. : orientation selectivity and the arrangement of horizontal connections in tree shrew striate cortex .
the journal of neuroscience * 17*(6 ) , 21122127 ( 1997 ) murray , m.m . ,
wylie , g.r . ,
higgins , b.a . ,
javitt , d.c . ,
schroeder , c.e . ,
foxe , j.j . : the spatiotemporal dynamics of illusory contour processing : combined high - density electrical mapping , source analysis , and functional magnetic resonance imaging .
the journal of neuroscience * 22*(12 ) , 50555073 ( 2002 ) song , c. , schwarzkopf , d.s . ,
lutti , a. , li , b. , kanai , r. , rees , g. : effective connectivity within human primary visual cortex predicts interindividual diversity in illusory perception .
the journal of neuroscience * 33*(48 ) , 18,78118,791 ( 2013 ) | geometrical optical illusions have been object of many studies due to the possibility they offer to understand the behaviour of low - level visual processing .
they consist in situations in which the perceived geometrical properties of an object differ from those of the object in the visual stimulus . starting from the geometrical model introduced by citti and sarti in @xcite , we provide a mathematical model and a computational algorithm which allows to interpret these phenomena and to qualitatively reproduce the perceived misperception . |
[ sec : intro ] the cauchy problem for a scalar balance law in @xmath3 space dimension @xmath4 is well known to admit a unique weak entropy solution , as proved in the classical result by krukov ( * ? ? ?
* theorem 5 ) .
the same paper also provides the basic stability estimate on the dependence of solutions from the initial data , see ( * ? ? ?
* theorem 1 ) . in the same setting established in @xcite ,
we provide here an estimate on the dependence of the solutions to ( [ eq : probv ] ) from the flow @xmath1 , from the source @xmath2 and recover the known estimate on the dependence from the initial datum @xmath5 .
a key intermediate result is a bound on the total variation of the solution to ( [ eq : probv ] ) , which we provide in theorem [ teo : tv ] . in the case of a conservation law ,
i.e. @xmath6 , and with a flow @xmath1 independent from @xmath7 , the dependence of the solution from @xmath1 was already considered in @xcite , where also other results were presented . in this case , the @xmath8 bound is obvious , since @xmath9 .
the estimate provided by theorem [ teo : tv ] slightly improves the analogous result in ( * ? ? ?
* theorem 3.1 ) ( that was already known , see @xcite ) , which reads ( for a suitable absolute constant @xmath10 ) @xmath11 our result , given by theorem [ teo : estimates ] , reduces to this inequality when @xmath1 and @xmath12 are not dependent on @xmath7 and @xmath13 , but with @xmath14 .
an flow dependent also on @xmath15 was considered in @xcite , though in the special case @xmath16 , but with a source term containing a possibly degenerate parabolic operator . there
, estimates on the @xmath17 distance between solutions in terms of the distance between the flows were obtained , but dependent from an _ a priori _ unknown bound on @xmath18 .
here , with no parabolic operators in the source term , we provide fully explicit bounds both on @xmath19 and on the distance between solutions .
indeed , remark that with no specific assumptions on the flow , @xmath19 may well blow up to @xmath20 at @xmath21 , as in the simple case @xmath22 with zero initial datum . both the total variation and
the stability estimates proved below turn out to be optimal in some simple cases , in which optimal estimates are known .
as an example of a possible application , we consider in section [ sec : application ] a toy model for a radiating gas .
this system was already considered in @xcite .
it consists of a balance law of the type ( [ eq : probv ] ) , but with a source that contains also a non local term , due to the convolution of the unknown with a suitable kernel .
thanks to the present results , we prove the well posedness of the model extending ( * ? ? ? * theorem 2.4 ) to more general flows , sources and convolution kernels .
stability and total variation estimates are also provided .
this paper is organized as follows : in section [ sec : notation ] , we introduce the notation , state the main results and compare them with those found in the literature . section [ sec : application ] is devoted to an application to a radiating gas model .
finally , in sections [ sec : prooftv ] and [ sec : proofcomparison ] the detailed proofs of theorems [ teo : tv ] and [ teo : estimates ] are provided .
[ sec : notation ] denote @xmath23 and @xmath240 , + \infty \right[$ ] .
below , @xmath3 is a positive integer , @xmath25 , @xmath26 denotes the ball in @xmath27 with center @xmath28 and radius @xmath29 .
the volume of the unit ball @xmath30 is @xmath31 . for notational simplicity
, we set @xmath32 .
the following relation can be proved using the expression of @xmath31 in terms of the wallis integral @xmath33 : @xmath34 in the present work , @xmath35 is the characteristic function of the set @xmath36 and @xmath37 is the dirac measure centered at @xmath38 .
besides , for a vector valued function @xmath39 with @xmath40 , @xmath41 stands for the total divergence . on the other hand , @xmath42 , respectively @xmath43 , denotes the partial divergence , respectively gradient , with respect to the space variables .
moreover , @xmath44 and @xmath45 are the usual partial derivatives .
thus , @xmath46 .
recall the definition of weak entropy solution to ( [ eq : probv ] ) , see ( * ? ? ?
* definition 1 ) .
[ def : sol ] a function @xmath47 is a weak entropy solution to ( [ eq : probv ] ) if : 1 . for any constant @xmath48 and any test function @xmath49 @xmath50
\\ \displaystyle \qquad \qquad\qquad \qquad \times \mathrm{sign}(u - k ) \ , \mathrm{d}x \ , \mathrm{d}t \geq 0 ; \end{array } \!\!\!\ ] ] 2 .
there exists a set @xmath51 of zero measure in @xmath52 such that for @xmath53 the function @xmath54 is defined almost everywhere in @xmath27 and for any @xmath29 @xmath55 throughout this paper , we refer to @xcite as general references for the theory of @xmath56 functions .
in particular , recall the following basic definition , see ( * ? ? ? * definition 3.4 and theorem 3.6 ) .
let @xmath57 .
define @xmath58 the following sets of assumptions will be of use below .
@xmath59 { \partial}_u f\in { \mathbf{l^\infty } } ( \omega ; { { \mathbb{r}}}^n ) \\[5pt ] \partial_u ( f - { { \mathrm{div}}}f ) \in { \mathbf{l^\infty } } ( \omega ; { { \mathbb{r } } } ) & f - { { \mathrm{div}}}f \in { \mathbf{l^\infty } } ( \omega ; { { \mathbb{r } } } ) \end{array } \right .
\\[25pt ] \mathbf{(h2)}\!\ ! & \left\ { \begin{array}{ll } f \in { \mathbf{c^{2}}}(\omega ; { { \mathbb{r}}}^n ) & f \in { \mathbf{c^{1}}}(\omega ; { { \mathbb{r } } } ) \\[5pt ] \nabla { \partial}_u f\in { \mathbf{l^\infty}}(\omega ; { { \mathbb{r}}}^{n\times n } ) & \displaystyle \int_{{{{\mathbb{r}}}_+ } } \!\
! \int_{{{\mathbb{r}}}^n } \
! { { \left\|\nabla(f - { { \mathrm{div}}}f)(t , x , \cdot)\right\|}}_{{\mathbf{l^\infty}}({{\mathbb{r}}};{{\mathbb{r}}}^{n } ) } \ , \mathrm{d}x \ , \mathrm{d}t < + \infty \\[10pt ] \partial_t \partial_u f \in { \mathbf{l^\infty } } ( \omega ; { { \mathbb{r}}}^n ) & \partial_t f \in { \mathbf{l^\infty } } ( \omega;{{\mathbb{r } } } ) \\[10pt ] \partial_t { { \mathrm{div}}}f \in { \mathbf{l^\infty } } ( \omega ; { { \mathbb{r } } } ) \end{array } \right .
\\[40pt ] \mathbf{(h3)}\!\ ! & \left\ { \begin{array}{l@{\qquad}l } f \in { \mathbf{c^{1}}}(\omega ; { { \mathbb{r}}}^n ) & f \in { \mathbf{c^{0}}}(\omega ; { { \mathbb{r } } } ) \qquad \partial_u f \in { \mathbf{l^\infty}}(\omega;{{\mathbb{r } } } ) \\[5pt ] { \partial}_u f\in { \mathbf{l^\infty}}(\omega ; { { \mathbb{r}}}^n ) & \displaystyle \int_{{{{\mathbb{r}}}_+ } } \!\
! \int_{{{\mathbb{r}}}^n } \ !
{ { \left\|(f- { { \mathrm{div}}}f)(t , x , \cdot)\right\|}}_{{\mathbf{l^\infty}}({{\mathbb{r}}};{{\mathbb{r } } } ) } \ , \mathrm{d}x\ , \mathrm{d}t < + \infty \end{array } \right .
\end{array}\ ] ] the quantity @xmath60 has a particular role , since it behaves as the _ `` true '' _ source , see ( [ eq : source ] ) .
we note there that the assumptions above can be significantly softened in various specific situations . for instance , the requirement that @xmath1 be lipschitz , which is however a standard hypothesis , see ( * ? ? ?
* paragraph 3 ) , can be relaxed to @xmath1 locally lipschitz in the case @xmath61 and @xmath62 , thanks to the maximum principle ( * ? ? ?
* theorem 3 ) .
furthermore , the assumptions above can be obviously weakened when aiming at estimates on bounded time intervals .
assumptions * ( h1 ) * are those used in the classical results ( * ? ? ?
* theorem 1 and theorem 5 ) .
however , we stress that the proofs below need less regularity . as in @xcite
, we remark that no derivative of @xmath1 or @xmath2 in time is ever needed .
furthermore , @xmath1 needs not be twice differentiable in @xmath63 , for the only second derivatives required are @xmath64 and @xmath65 .
we recall below the classical result by krukov .
[ teo : kruzkov ] let * ( h1 ) * hold .
then , for any @xmath66 , there exists a unique weak entropy solution @xmath63 to ( [ eq : probv ] ) in @xmath67 continuous from the right .
moreover , if a sequence @xmath68 converges to @xmath5 in @xmath69 , then for all @xmath70 the corresponding solutions @xmath71 converge to @xmath72 in @xmath69 .
[ rem : c ] under the conditions * ( h2 ) * and @xmath73 , see * ( h3 ) * , the estimate provided by theorem [ teo : tv ] below , allows to use the technique described in ( * ? ? ?
* theorem 4.3.1 ) , proving the continuity in time of the solution , so that @xmath74 .
recall that ( * ? ? ?
* theorem 1.3 ) and ( * ? ? ?
* theorem 3.2 ) provide stability bounds on ( [ eq : probv ] ) , in the more general case with a degenerate parabolic source , but assuming _ a priori _ bounds on the total variation of solutions .
our first result provides these bounds .
[ teo : tv ] assume that * ( h1 ) * and * ( h2 ) * hold .
let @xmath75 .
then , the weak entropy solution @xmath63 of ( [ eq : probv ] ) satisfies @xmath76 for all @xmath77 .
moreover , let @xmath78 with @xmath33 as in ( [ eq : w ] ) .
then , for all @xmath79 , @xmath80 this estimate is optimal in the following situations : 1 .
if @xmath1 is independent from @xmath15 and @xmath62 , then @xmath81 and the integrand in the right hand side above vanishes .
hence , ( [ result ] ) reduces to the well known optimal bound @xmath82 .
2 . in the 1d case ,
if @xmath1 and @xmath2 are both independent from @xmath38 and @xmath63 , then @xmath81 and ( [ eq : probv ] ) reduces to the ordinary differential equation @xmath83 . in this case , ( [ result ] ) becomes @xmath84 3 . if @xmath85 and @xmath86 then , trivially , @xmath87 and ( [ result ] ) is optimal
. a simpler but slightly weaker form of ( [ result ] ) is @xmath88}\int_{{{\mathbb{r}}}^n } { { \left\|\nabla ( f - { { \mathrm{div}}}f)(t , x , \cdot)\right\|}}_{{\mathbf{l^\infty } } } \mathrm{d}x\ ] ] when the right hand side is bounded . consider now ( [ eq : probv ] ) together with the analogous problem @xmath89 we aim at estimates for the difference @xmath90 between the solutions in terms of @xmath91 , @xmath92 and @xmath93 .
estimates of this type were derived by bouchut & perthame in @xcite when @xmath1 , @xmath12 depend only on @xmath63 and @xmath94 .
here , we generalize their result adding the @xmath95-dependence .
the present technique is essentially based on theorem [ teo : tv ] .
[ teo : estimates ] let @xmath96 , @xmath97 verify * ( h1 ) * , @xmath98 verify * ( h2 ) * and @xmath99 verify * ( h3)*. let @xmath100 .
we denote @xmath101 as in ( [ eq : kappao ] ) and introduce @xmath102 then , for any @xmath103 and @xmath104 , the following estimate holds : @xmath105 the above inequality is undefined for @xmath106 and , in this case , it reduces to ( [ eqkko ] ) .
this bound is optimal in the following situations , where @xmath107 . 1
. in the standard case of a conservation law , i.e. when @xmath108 and @xmath109 are independent of @xmath15 , we have @xmath110 and the result of theorem [ teo : estimates ] becomes , see ( * ? ? ?
* theorem 2.1 ) , @xmath111 2 .
if @xmath112 and @xmath113 , then @xmath114 and theorem [ teo : estimates ] now reads @xmath115(t)\right\|}}_{{\mathbf{l^1}}({{\mathbb{r}}}^n;{{\mathbb{r } } } ) } \mathrm{d}t.\ ] ] 3 .
if @xmath98 and @xmath97 are dependent only on @xmath15 , then theorem [ teo : estimates ] reduces to @xmath116 the estimate obtained in theorem [ teo : estimates ] shows also that , depending on the properties of specific applications , the regularity requirement @xmath117 can be significantly relaxed .
for instance , in the case @xmath118 considered in @xcite , asking @xmath119 of class @xmath120 and @xmath121 of class @xmath122 is sufficient .
see also section [ sec : application ] for a case in which the required regularity in time can be reduced .
in the case of conservations laws , i.e. when @xmath94 , one proves that @xmath123 and the estimate in theorem [ teo : estimates ] takes the somewhat simpler form @xmath124 & + & t \ ,
e^{\kappa_ot } \ , { \mathrm{tv}}(u_o ) \ , { { \left\|{\partial}_u(f - g)\right\|}}_{{\mathbf{l^\infty } } } \\ & + & nw_n \ , t^2 \ , e^{\kappa_ot } \sup_{t \in [ 0,t ] } \left ( \int_{{{\mathbb{r}}}^n } { { \left\|\nabla { { \mathrm{div}}}f(t , x,\cdot)\right\|}}_{{\mathbf{l^\infty } } } \mathrm{d}x \right ) \ ; { { \left\|{\partial}_u(f - g)\right\|}}_{{\mathbf{l^\infty } } } \\ & + & t e^{\kappa_ot } \sup_{t \in [ 0,t ] } \int_{{{\left\|x - x_o\right\| } } \leq r+m(t - t ) } { { \left\|{{\mathrm{div}}}(f - g ) ( t , x,\cdot)\right\|}}_{{\mathbf{l^\infty } } } \ , \mathrm{d}x\end{aligned}\ ] ] when the right hand side is bounded . in the case considered in
* theorem 3.1 ) , @xmath61 , @xmath125 and we obtain ( * ? ? ? * formula ( 3.2 ) ) with @xmath126 instead of the constant @xmath10 therein .
[ sec : application ] the following balance law is a toy model inspired by euler equations for radiating gases : @xmath127 it has been extensively studied in the literature when @xmath61 , see for instance @xcite for the scalar @xmath126d case , @xcite for 1d systems , @xcite for the scalar @xmath3d case .
the estimate provided by theorem [ teo : estimates ] allows us to present an alternative proof of the well posedness of ( [ eq : bl ] ) proved in @xcite .
furthermore , we add stability estimates on the dependence of the solution from @xmath1 and @xmath128 , in the case of @xmath1 dependent also on @xmath7 and with more general source terms .
[ thm : appl ] let @xmath98 satisfy * ( h1 ) * , * ( h2 ) * and * ( h3)*. assume that @xmath129 then , for any @xmath130 , the cauchy problem @xmath131 admits a unique weak entropy solution @xmath132 . moreover , denoting @xmath133 , for all @xmath79 ,
the following estimate holds : @xmath134 if @xmath135 for all @xmath136 $ ] and @xmath137 , then 1 .
2 . let @xmath139 satisfy * ( k ) * and call @xmath140 the solution to ( [ eq : toy ] ) with @xmath128 replaced by @xmath139 . then , @xmath141 fix a positive @xmath142 ( to be specified below ) and consider the banach space @xmath143 ; { \mathbf{l^1}}({{\mathbb{r}}}^n;{{\mathbb{r } } } ) \right)$ ] equipped with the usual norm @xmath144 .
define on @xmath145 the map @xmath146 so that @xmath147 if and only if @xmath63 solves @xmath148 in the sense of definition [ def : sol ] .
note that the source term does not have the regularity required in * ( h1)*. however , by the estimate in theorem [ teo : estimates ] , we can prove that ( [ eq : weak ] ) does indeed have a unique weak entropy solution , see lemma [ lem : weak ] for the details .
the fixed points of @xmath149 are the solutions to ( [ eq : bl ] ) . by theorem [ teo : kruzkov ] and remark [ rem : c ] , @xmath150 for all @xmath151 .
we now show that @xmath146 is a contraction , provided @xmath142 is sufficiently small .
note that @xmath152 moreover , by theorem [ teo : estimates ] @xmath153 } { { \left\|{\mathcal t}w_1 - { \mathcal t}w_2\right\|}}_{{\mathbf{l^1 } } } \\ & \leq & \sup_{t \in [ 0,t ] } \left ( \frac{e^{\kappa t } -1}{\kappa } \sup_{\tau \in [ 0,t ] } { { \left\|k(\tau ) \ast_x ( w_1 - w_2 ) ( \tau)\right\|}}_{{\mathbf{l^1 } } } \right ) \\ & \leq & \frac{e^{\kappa t } -1}{\kappa } \sup_{\tau \in [ 0,t ] } { { \left\|k(\tau)\right\|}}_{{\mathbf{l^1 } } } \ , { { \left\|(w_1 - w_2 ) ( \tau)\right\|}}_{{\mathbf{l^1 } } } \\ & \leq & \frac{e^{\kappa t } -1}{\kappa } \ , k \ , d(w_1 , w_2 ) \ , .
\end{aligned}\ ] ] therefore , @xmath146 is a contraction as soon as @xmath142 is smaller than a threshold that depends only on @xmath154 and on @xmath155 .
therefore , we proved the well posedness of ( [ eq : toy ] ) globally in time .
consider the bound on @xmath18 . by theorem [ teo : tv ] , @xmath156 and an application of gronwall lemma gives the desired bound .
we estimate the @xmath17 norm of the solution to ( [ eq : toy ] ) , comparing it with the solution to @xmath157 by assumption , @xmath158 solves ( [ eq : toy0 ] ) , hence it is its unique solution .
then , evaluating the distance between the solutions of ( [ eq : toy ] ) and ( [ eq : toy0 ] ) by means of theorem [ teo : estimates ] , we get @xmath159 and , thanks to gronwall lemma , we obtain : @xmath160 the final estimate ( [ eq : end ] ) follows from theorem [ teo : estimates ] : @xmath161 and thanks to gronwall lemma , we get the result . the continuity in time is proved as described in remark [ rem : c ] .
[ lem : weak ] let @xmath162 satisfy * ( h1 ) * and @xmath128 satisfy * ( k)*. if @xmath163 , then the estimates in theorem [ teo : tv ] and in theorem [ teo : estimates ] apply also to ( [ eq : weak ] ) .
fix positive @xmath164 and let @xmath165 be a sequence of @xmath166 functions converging to @xmath167 in @xmath168\times { { \mathbb{r}}}^n;{{\mathbb{r}}}\right)$ ]
. apply theorem [ teo : kruzkov ] to the approximate problem @xmath169 to ensure the existence of its weak entropy solution @xmath170 .
apply theorem [ teo : estimates ] to estimate the distance between @xmath170 and @xmath171 : @xmath172;{\mathbf{l^1}}({{\mathbb{r}}}^n;{{\mathbb{r } } } ) ) } & \leq & \int_0^t e^{\kappa(t - t ) } \int_{{{\mathbb{r}}}^n } { { \left|k * ( w_n - w_{n-1 } ) ( t , x)\right| } } \ , \mathrm{d}x \ , \mathrm{d}t \\ & \leq & e^{\kappa t } \ , k \
, { { \left\|w_n - w_{n-1}\right\|}}_{{\mathbf{l^1}}([0,t]\times{{\mathbb{r}}}^n;{{\mathbb{r } } } ) } \end{aligned}\ ] ] showing that the @xmath170 form a cauchy sequence .
their limit @xmath63 solves ( [ eq : toy ] ) , as it follows passing to the limit over @xmath173 in the integral conditions ( [ eq : entropy])([eq : data ] ) and applying the dominated convergence theorem .
the estimates in theorems [ teo : tv ] and [ teo : estimates ] are extended similarly .
[ sec : prooftv ] fix a function @xmath174 with @xmath175 define @xmath176 then , recalling that @xmath32 , @xmath177 the first relation is immediate .
equalities ( [ eq : mu3 ] ) and ( [ eq : mu4 ] ) follow directly from an integration by parts . consider ( [ eq : mu2 ] ) .
the cases @xmath178 follow from direct computations .
let @xmath179 and pass to spherical coordinates @xmath180 , @xmath181 with @xmath182 , @xmath183 and @xmath184 $ ] for @xmath185 . if @xmath179 @xmath186 completing the proof .
recall the following theorem ( see ( * ? ? ?
* theorem 3.9 and remark 3.10 ) ) : [ thm : afp ] let @xmath187 , then @xmath188 if and only if there exists a sequence @xmath170 in @xmath189 converging to @xmath63 in @xmath69 and satisfying @xmath190 moreover , @xmath191 is the least constant @xmath192 for which there exists a sequence as above .
[ prop : normtv ] fix @xmath193 as in ( [ eq : mu ] ) .
let @xmath194 admit a constant @xmath195 such that for all positive @xmath196 , @xmath197 and with @xmath198 as in ( [ eq : mu ] ) @xmath199 then , @xmath200 and @xmath201 , where @xmath202 note that @xmath2030 , 1 \right[$ ] .
if moreover @xmath204 , then @xmath205 we introduce now a regularisation of @xmath63 : @xmath206 , with @xmath207 .
note that @xmath208 and @xmath209 converges to @xmath63 in @xmath69 as @xmath210 .
furthermore , for @xmath197 and @xmath211 positive , we have @xmath212 and @xmath213 thanks to the dominated convergence theorem , at the limit @xmath214 we get @xmath215 remark that for fixed @xmath216 , when @xmath217 , the scalar product @xmath218 is positive ( respectively , negative ) when @xmath219 is in a half - space , say @xmath220 ( respectively , @xmath221 ) .
we can write @xmath222 , with @xmath223 and @xmath167 in the hyperplane @xmath224 .
hence @xmath225 define @xmath226 as in ( [ eq : c1 ] ) and note that @xmath2030 , 1\right[$ ]
. then we obtain , for all @xmath227 , @xmath228 finally when @xmath229 we get @xmath230 and in the limit @xmath231 , by theorem [ thm : afp ] also @xmath232 , concluding the proof of the first statement .
assume now that @xmath233 .
then , using the same computations as above , @xmath234 completing the proof . in the following proof , this property of any function @xmath235 will be of use : @xmath236 for a proof , see ( * ? ? ?
* remark 3.25 ) .
theorem [ teo : tv ] assume first that @xmath237 , the general case will be considered only at the end of this proof .
let @xmath63 be the weak entropy solution to ( [ eq : probv ] ) .
denote @xmath238 and @xmath239 for @xmath240 .
then , for all @xmath241 and for all test functions @xmath242 in @xmath243 , we have @xmath244 \\ \times \mathrm{sign}(u - k ) \ , \mathrm{d}x \ , \mathrm{d}t \geq 0 \end{array}\ ] ] for all @xmath245 , and @xmath246 \\ \times \mathrm{sign}(v - l ) \ , \mathrm{d}y \ , \mathrm{d}s \geq 0 \end{array}\ ] ] for all @xmath247 .
let @xmath248 , @xmath249 and set @xmath250 observe that @xmath251 , @xmath252 , @xmath253 .
choose @xmath254 in ( [ eq : utv ] ) and integrate with respect to @xmath255 .
analogously , take @xmath256 in ( [ eq : vtv ] ) and integrate with respect to @xmath95 . summing the obtained inequalities , we get @xmath257 \mathrm{d}x \ , \mathrm{d}t \ , \mathrm{d}y \ , \mathrm{d}s \geq 0 .
\end{array}\ ] ] introduce a family of functions @xmath258 such that for any @xmath259 : + @xmath2600,1\right [ \\ y ' & \geq & 0 \\ \displaystyle \int_{{\mathbb{r}}}y'(s ) \ , \mathrm{d}s & = & 1 \ , .
\end{array}\ ] ] let @xmath261 and define for @xmath262 , @xmath104 , , left , and of @xmath263 , right .
here @xmath264 and @xmath265 . ] , left , and of @xmath263 , right . here
@xmath264 and @xmath265 . ]
( see figure [ fig : chipsi ] ) : @xmath266 where we also need the compatibility conditions @xmath267 and @xmath268 .
observe that @xmath269}$ ] and @xmath270 as @xmath271 tends to @xmath158 . on @xmath272 and @xmath263
we use the bounds @xmath273 } \quad \mbox { and } \quad \mathbf{1}_{b(x_o , r+m(t_o - t ) ) } \leq \psi \leq \mathbf{1}_{b(x_o , r+m(t_o - t)+\theta ) } \,.\ ] ] in ( [ eq : sumtv ] ) , choose @xmath274 . with this choice , we have @xmath275 setting @xmath276 , the first line in ( [ eq : sumtv ] ) becomes @xmath277\ !
\mathrm{sign}(u - v ) \mathrm{d}x \ , \mathrm{d}t \ , \mathrm{d}y \ , \mathrm{d}s \\ \!\ ! & = & \int_{{{{\mathbb{r}}}_+ } } \int_{{{\mathbb{r}}}^n } \int_{{{{\mathbb{r}}}_+ } } \int_{{{\mathbb{r}}}^n } \left ( { { \left|u - v\right|}}\ , \chi ' \ , \psi - b(t , x , u , v ) \chi \ , y_\theta ' \right ) \ , \psi \ , \mathrm{d}x \ , \mathrm{d}t \ ,
\mathrm{d}y \ , \mathrm{d}s \\ \!\ ! & \leq & \int_{{{{\mathbb{r}}}_+ } } \int_{{{\mathbb{r}}}^n } \int_{{{{\mathbb{r}}}_+ } } \int_{{{\mathbb{r}}}^n } { { \left|u - v\right| } } \ , \chi ' \ , \psi \ , \psi \ , \mathrm{d}x \ , \mathrm{d}t \ , \mathrm{d}y \ , \mathrm{d}s \end{aligned}\ ] ] since @xmath278 is positive for all @xmath279 .
thanks to the above estimate and to ( [ eq : sumtv ] ) , we have @xmath280 \\ & \displaystyle \times \mathrm{sign}(u - v ) \ , \mathrm{d}x\,\mathrm{d}t \ , \mathrm{d}y \ , \mathrm{d}s & \geq & 0 . \end{array}\ ] ] now , we aim at bounds for each term of this sum .
introduce the following notations : @xmath281 then , the above inequality is rewritten as @xmath282 .
choose @xmath283 where , for @xmath284 , @xmath285 satisfies ( [ eq : mu])([eq : mu ] ) and @xmath286 - 1,0\right [ \,.\ ] ] we have @xmath287 and we get @xmath288t , t+{{\eta}}\right [ } \int_{{{\left\|y - x_o\right\|}}\leq r+{{\lambda}}+m(t_o - t)+\theta } { { \left|u(t , y)-u(s , y)\right| } } \ , \mathrm{d}y \ , .
\end{aligned}\ ] ] for @xmath289 , we have that by * ( h1 ) * , @xmath290 and therefore @xmath291 then , using ( [ eq : mu3 ] ) @xmath292 \\ & & \qquad \qquad \qquad \qquad \qquad \qquad \times { { \left\|\nabla \psi\right\| } } \ , \chi \ , \psi \,\mathrm{d}x\,\mathrm{d}t\,\mathrm{d}y\,\mathrm{d}s \\ & \leq & n \ , { { \left\|\nabla{\partial}_u f \right\|}}_{{\mathbf{l^\infty } } } ( t+{\varepsilon } ) \sup_{t\in [ 0,t+{\varepsilon } ] , \atop s \in \left]t , t+{{\eta}}\right [ } \int_{{{\left\|y - x_o\right\|}}\le r+{{\lambda}}+m(t_o - t)+\theta } { { \left|u(t , y)-u(s , y)\right| } } \ , \mathrm{d}y \\ & & + { { \left\|\nabla{\partial}_u f \right\|}}_{{\mathbf{l^\infty } } } \int_0^{t+{\varepsilon } } \int_{{{\mathbb{r}}}^n } \int_{b(x_o , r + m(t_o - t)+\theta ) } { { \left\|x - y\right\| } } \ , { { \left|u(t , x)-u(t , y)\right| } } \ , { { \left\|\nabla \mu\right\| } } \,\mathrm{d}x\,\mathrm{d}y\,\mathrm{d}t\ , , \\ j_{t } & \leq & \int_{{{{\mathbb{r}}}_+ } } \int_{{{\mathbb{r}}}^n } \int_{{{{\mathbb{r}}}_+ } } \int_{{{\mathbb{r}}}^n } { { \left\| \int_s^t \int_v^u { \partial}_{t } { \partial}_{u } f(\tau , y , w ) \ , \mathrm{d}w \ , \mathrm{d}\tau \right\| } } { { \left\|\nabla \psi\right\|}}\ , \phi \,\mathrm{d}x\,\mathrm{d}t\,\mathrm{d}y\,\mathrm{d}s \\ & \leq & { { \eta}}\ , { { \left\|{\partial}_{t } { \partial}_{u } f\right\|}}_{{\mathbf{l^\infty } } } \int_{{{{\mathbb{r}}}_+ } } \int_{{{\mathbb{r}}}^n } \int_{{{{\mathbb{r}}}_+ } } \int_{{{\mathbb{r}}}^n } { { \left|u(t , x)-u(s , y)\right| } } \ , { { \left\|\nabla \psi\right\| } } \,\phi \ , \mathrm{d}x \ , \mathrm{d}t \ , \mathrm{d}y \ , \mathrm{d}s \ , .
\end{aligned}\ ] ] for @xmath293 , we get @xmath294\ ! { \varphi}\ , \mathrm{sign}(u - v ) \,\mathrm{d}x\,\mathrm{d}t\,\mathrm{d}y\,\mathrm{d}s , \\ l_2 & = & \int_{{{{\mathbb{r}}}_+ } } \int_{{{\mathbb{r}}}^n } \ ! \int_{{{{\mathbb{r}}}_+ } } \int_{{{\mathbb{r}}}^n } \left [ \int_0 ^ 1 \
! \nabla ( f - { { \mathrm{div}}}f ) \left(t , rx+(1-r)y , u\right)\cdot ( x - y ) \ , \mathrm{d}r \right ] { \varphi}\\ & & \qquad \qquad \qquad \qquad \times \mathrm{sign}(u - v ) \,\mathrm{d}x\,\mathrm{d}t\,\mathrm{d}y\,\mathrm{d}s .
\end{aligned}\ ] ] then , recalling ( [ eq : phi ] ) , the definitions @xmath295 , @xmath296 , ( [ eq : mu ] ) , ( [ eq : nu ] ) and ( [ eq : chipsi ] ) , we obtain @xmath297 , \atop s \in \left]t , t+{{\eta}}\right [ } \int_{{{\left\|y - x_o\right\|}}\leq r+{{\lambda}}+m(t_o - t)+\theta } { { \left|u(t , y)-u(s , y)\right| } } \ , \mathrm{d}y \\ & & \qquad + \int_0^{t+{\varepsilon } } \int_{{{\mathbb{r}}}^n } \int_{{{\left\|x - x_o\right\|}}\leq r+m(t_o - t)+\theta } { { \left|u(t , x)-u(t , y)\right| } } \ , \mu(x - y ) \,\mathrm{d}x\,\mathrm{d}y\,\mathrm{d}t \bigg]\ , , \\ l_2 & \leq & \int_{{{{\mathbb{r}}}_+ } } \ ! \int_{{{\mathbb{r}}}^n } \
! \int_{{{{\mathbb{r}}}_+ } } \ !
\int_{{{\mathbb{r}}}^n } \!\ !
\int_0 ^ 1 { { \left\|\nabla ( f - { { \mathrm{div}}}f ) \left(t , y+r(x - y),u\right)\right\| } } { { \left\|x - y\right\| } } \chi \ , \psi \ , \mu \ , \nu \mathrm{d}r \,\mathrm{d}x\,\mathrm{d}t\,\mathrm{d}y\,\mathrm{d}s \\ & \leq & \left ( \int_0^{t+\varepsilon } \int_{{{\mathbb{r}}}^n } { { \left\|\nabla ( f - { { \mathrm{div}}}f ) ( t , y,\cdot)\right\|}}_{{\mathbf{l^\infty } } } \mathrm{d}y\ , \mathrm{d}t \right ) \int_{{{\mathbb{r}}}^n } { { \left\|x\right\| } } \ , \mu(x ) \ , \mathrm{d}x \\ & = & { { \lambda}}\ , m_1 \ ; \int_0^{t+\varepsilon } \int_{{{\mathbb{r}}}^n } { { \left\|\nabla ( f - { { \mathrm{div}}}f ) ( t , y,\cdot)\right\|}}_{{\mathbf{l^\infty}}}\ , \mathrm{d}y\ , \mathrm{d}t \end{aligned}\ ] ] where @xmath298 concerning the latter term @xmath299 @xmath300 letting @xmath301 we get @xmath302 collating all the obtained results and using the equality @xmath303
@xmath304 & \leq & \displaystyle \int_{{{\mathbb{r}}}^n } \int_{{{\left\|x - x_o\right\|}}\leq r + m(t_o - t ) } { { \left|u(0,x)-u(0,y)\right| } } \ , \frac{1}{{{\lambda}}^n } \mu_1\left(\frac{{{\left\|x - y\right\|}}}{{{\lambda}}}\right ) \ , \mathrm{d}x \ , \mathrm{d}y \\[10pt ] & & \displaystyle - { { \left\|\nabla{\partial}_u f \right\|}}_{{\mathbf{l^\infty } } } \int_0^{t } \int_{{{\mathbb{r}}}^n } \int_{{{\left\|x - x_o\right\|}}\leq r + m(t_o - t ) } \!\!\ !
{ { \left|u(t , x)-u(t , y)\right| } } \\[10pt ] & & \displaystyle \qquad \qquad \qquad \times \frac{1}{{{\lambda}}^{n+1 } } \ , \mu_1'\left ( \frac{{{\left\|x - y\right\|}}}{{{\lambda } } } \right ) \ , { { \left\|x - y\right\| } } \,\mathrm{d}x\,\mathrm{d}y\,\mathrm{d}t \\[10pt ] & & \displaystyle + \left ( n { { \left\|\nabla{\partial}_u f\right\|}}_{{\mathbf{l^\infty } } } + { { \left\|{\partial}_u f\right\|}}_{{\mathbf{l^\infty } } } \right ) \ !
\int_0^{t } \ !
\int_{{{\mathbb{r}}}^n } \ ! \int_{{{\left\|x - x_o\right\|}}\leq r + m(t_o - t ) } \ !
{ { \left|u(t , x)-u(t , y)\right| } } \\[10pt ] & & \displaystyle \qquad \qquad \qquad \times \frac{1}{{{\lambda}}^n } \mu_1\left(\frac{{{\left\|x - y\right\|}}}{{{\lambda}}}\right ) \,\mathrm{d}x\,\mathrm{d}y\,\mathrm{d}t \\[10pt ] & & \displaystyle + { { \lambda}}m_1 \int_0^{t}\int_{{{\mathbb{r}}}^n } { { \left\|\nabla(f - { { \mathrm{div}}}f)(t , y,\cdot)\right\|}}_{{\mathbf{l^\infty } } } \ , \mathrm{d}y\ , \mathrm{d}t \ , .
\end{array}\ ] ] if @xmath305 and under the present assumption that @xmath306 , using proposition [ prop : normtv ] , ( [ eq : c1 ] ) and ( [ eq : m1 ] ) , we directly obtain that @xmath307 the same procedure at the end of this proof allows to extend ( [ eq : ck ] ) to more general initial data , providing an estimate of @xmath19 in the situation studied in @xcite .
now , it remains to treat the case @xmath308 .
a direct use of gronwall type inequalities is apparently impossible , due to the term with @xmath309 .
however , introduce the function @xmath310 so that @xmath311 denote @xmath312 and integrate ( [ inegalite ] ) on @xmath313 $ ] with respect to @xmath142 for @xmath314 .
it results @xmath315 denote @xmath316 , so that @xmath317 .
the previous inequality reads , using ( [ eq : afp ] ) for @xmath5 , @xmath318 finally , if @xmath319 is such that @xmath320 , then we integrate in @xmath196 on @xmath321 and we get @xmath322 furthermore , by ( [ eq : mu ] ) and ( [ eq : mu ] ) , there exists a constant @xmath323 such that for all @xmath324 @xmath325 divide both sides in ( [ inegalite ] ) by @xmath196 , rewrite them using ( [ eq : fdelta ] ) , ( [ eq : boundmu ] ) , apply ( [ eq : afp ] ) and obtain @xmath326 an application of ( [ eq : fdelta ] ) yields an estimate of the type @xmath327 the positive constant @xmath328 being independent from @xmath197 and @xmath196 . applying proposition [ prop : normtv ]
we obtain that @xmath329 for @xmath330 , where @xmath331 the next step is to obtain a general estimate of the @xmath8 norm .
the starting point is ( [ inegalite ] ) .
recall the definitions ( [ eq : m1 ] ) of @xmath332 and ( [ eq : t1 ] ) of @xmath333 .
moreover , by ( [ eq : mu4 ] ) , @xmath334 divide both terms in ( [ inegalite ] ) by @xmath196 , apply ( [ eq : tv1 ] ) on the first term in the right hand side , apply ( [ eq : afp ] ) on the second and third terms and obtain for all @xmath335 $ ] with @xmath336 @xmath337 an application of gronwall lemma shows that @xmath338 is bounded on @xmath339 $ ] .
indeed , @xmath340 for @xmath341 $ ] , @xmath342 as in ( [ eq : m1 ] ) , ( [ eq : c1 ] ) and @xmath343 m_1/c_1 $ ] .
we now relax the assumption on the regularity of @xmath5 . indeed , let @xmath344 and choose a sequence @xmath345 of @xmath346 functions such that @xmath347 , as in theorem [ thm : afp ] . then , by theorem [ teo : kruzkov ] , the solutions @xmath348 to ( [ eq : probv ] ) with initial datum @xmath349 satisfy @xmath350 where we used also the lower semicontinuity of the total variation . note that ( [ eq : tv ] ) , as well as the relations above , holds for all @xmath341 $ ] , @xmath333 being independent from the initial datum
. therefore , the bound ( [ eq : tv ] ) holds for all @xmath56 initial data .
remark that the bound ( [ eq : tv ] ) is additive in time , in the sense that applying it iteratively for times @xmath333 and @xmath38 yields ( [ eq : tv ] ) for time @xmath351 : @xmath352 \\ & & + \frac{m_1}{c_1 } \int_{t_1}^{t_1+t}e^{\kappa_o(t_1+t - s)}\int_{{{\mathbb{r}}}^n } { { \left\|\nabla(f-{{\mathrm{div}}}f)(s , x,\cdot)\right\|}}_{{\mathbf{l^\infty } } } \ , \mathrm{d}x\ , \mathrm{d}s \\ & = & e^{\kappa_o ( t_1+t ) } \ , { \mathrm{tv}}(u_o ) + \frac{m_1}{c_1}\int_{0}^{t_1+t}e^{\kappa_o(t_1+t - s ) } \int_{{{\mathbb{r}}}^n } { { \left\|\nabla(f-{{\mathrm{div}}}f)(s , x,\cdot)\right\|}}_{{\mathbf{l^\infty } } } \ , \mathrm{d}x\ , \mathrm{d}s\ , . \end{aligned}\ ] ] the bound ( [ eq : tv ] ) can then be applied iteratively , thanks to the fact that @xmath333 is independent from the initial datum .
an iteration argument allows to prove ( [ result ] ) for @xmath353 $ ] . the final bound ( [ result ] ) then follows by the arbitrariness of @xmath354 , thanks to ( [ eq : w ] ) .
[ sec : proofcomparison ] by definition [ def : sol ] , we have @xmath358 , @xmath359 @xmath360 \\
\times \mathrm{sign}(u - l ) \ , \mathrm{d}y \ , \mathrm{d}s \geq 0 \end{array}\ ] ] and @xmath361 , @xmath362 @xmath363 \\ \times \mathrm{sign}(v - k ) \ , \mathrm{d}x \ , \mathrm{d}t \geq 0
. \end{array}\ ] ] choose @xmath364 in ( [ eq : v ] ) and integrate with respect to @xmath255 .
analogously , take @xmath365 in ( [ eq : u ] ) and integrate with respect to @xmath95 . by summing the obtained equations , we get , denoting @xmath366 and @xmath367 : @xmath368 \\ & \times \mathrm{sign}(u - v ) \ , \mathrm{d}x\,\mathrm{d}t\,\mathrm{d}y\,\mathrm{d}s & \geq & 0 \ , .
\end{array}\ ] ] introduce a family of functions @xmath258 as in ( [ eq : y ] ) .
let @xmath369 and define @xmath370 as in ( [ eq : chipsi ] ) , for @xmath262 , @xmath371 , ( see also figure [ fig : chipsi ] ) .
remind that with these choices , equalities ( [ eq : derivatives ] ) still hold .
note that here the definition of the test function @xmath372 is essentially the same as in the preceding proof ; the only change is the definition of the constant @xmath373 , that is now defined with reference to @xmath12 .
we also introduce as above the function @xmath374 that is positive for all @xmath375 , and we have : @xmath376 \psi \ , \mathrm{sign}(u - v ) \ , \mathrm{d}x \ , \mathrm{d}t \ , \mathrm{d}y \ , \mathrm{d}s \\ & \leq & \int_{{{{\mathbb{r}}}_+ } } \ !
\int_{{{\mathbb{r}}}^n } \ !
\int_{{{{\mathbb{r}}}_+ } } \ !
\int_{{{\mathbb{r}}}^n } \left [ { { \left|u - v\right| } } \chi ' \psi - b(t , x , u , v ) \chi y_\theta ' \right ] \psi \ , \mathrm{d}x \ , \mathrm{d}t \
, \mathrm{d}y \ , \mathrm{d}s \\ & \leq & \int_{{{{\mathbb{r}}}_+ } } \ !
\int_{{{\mathbb{r}}}^n } \ !
\int_{{{{\mathbb{r}}}_+ } } \ !
\int_{{{\mathbb{r}}}^n } { { \left|u - v\right| } } \ , \chi ' \ , \psi \ , \psi \ , \mathrm{d}x \ , \mathrm{d}t \ , \mathrm{d}y \ , \mathrm{d}s \ , .
\end{aligned}\ ] ] thanks to the above estimate and ( [ eq : sum ] ) , it results @xmath377 & & \\ & \displaystyle \times \mathrm{sign}(u - v ) \,\mathrm{d}x \,\mathrm{d}t\,\mathrm{d}y\,\mathrm{d}s & \geq & 0 \ , , \end{array}\ ] ] i.e. @xmath378 , where @xmath379 now , we choose @xmath380 as in ( [ eq : nu ] ) , ( [ eq : mu ] ) , ( [ eq : mu ] ) . thanks to lemma [ lem : esti ] , lemma [ lem : estj ] and lemma [ lem : estl ] we obtain @xmath381 besides , we find that : @xmath382 so that @xmath383 in order to estimate @xmath128 as given in ( [ eq : k ] ) , we introduce a regularisation of the @xmath384 dependent functions .
in fact , let @xmath385 and @xmath386 , where @xmath387 and @xmath388 are such that @xmath389 and @xmath390 - 1,1\right[$ ] , @xmath391 .
then , introduce @xmath392 so that we obtain @xmath393 now , we use the relation @xmath394 to obtain @xmath395 } \left ( \rho \left ( \frac{u - v}{\alpha } \right ) \left(p_i(u)-p_i(v ) \right ) \right ) \min \left\{2\alpha , { { \left|u - u_\beta\right| } } \right\ } \ , { \partial}_{y_i } \varphi \,\mathrm{d}y \\ & & + \int_{{{\mathbb{r}}}^n } \int_{u}^{u_\beta } { { \left|p_i'(u)\right| } } { \partial}_{y_i } \varphi \ , \mathrm{d}u \ , \mathrm{d}y \ , .
\end{aligned}\ ] ] when @xmath396 tends to @xmath158 , thanks to the dominated convergence theorem , we obtain @xmath397 applying the dominated convergence theorem again , we see that @xmath398 consequently , it is sufficient to find a bound independent of @xmath396 and @xmath399 on @xmath400 , where @xmath401 integrating by parts , we get @xmath402 we now search a bound for each term of the sum above . now , we collate the estimates obtained in ( [ eq : esti ] ) , ( [ eq : estj ] ) , ( [ eq : estl ] ) , ( [ eq : uffa ] ) and ( [ eq : estk ] ) .
remark the order in which we pass to the various limits : first @xmath413 and , after , @xmath214 .
therefore , we get @xmath414 \\ & & \qquad\quad \times \int_0^{t } \!\ !
\int_{b(x_o , r+m(t_o - t ) ) } { { \left|v(t , x)-u(t , x)\right| } } \ , \mathrm{d}x \ , \mathrm{d}t \\ & & + \biggl [ { { \left\|{\partial}_u(f - g)\right\|}}_{{\mathbf{l^\infty } } } \int_0^t { \mathrm{tv}}(u(t))\ , \mathrm{d}t \\ & & \qquad + \int_0^t \int_{b(x_o , r+m(t_o - t))}\ !
{ { \left\|\left((f - g)-{{\mathrm{div}}}(f - g)\right)(t , y , \cdot)\right\|}}_{{\mathbf{l^\infty } } } \mathrm{d}y\ , \mathrm{d}t \biggr ] \end{aligned}\ ] ] or equivalently @xmath415 where @xmath416 the bound ( [ result ] ) on @xmath18 gives : @xmath417 where @xmath101 is defined in ( [ eq : kappao ] ) and @xmath418 since @xmath419 .
consequently @xmath420 by a gronwall type argument , if @xmath421 , we get @xmath422 yielding @xmath423 while , in the case @xmath424 , we have @xmath425 taking @xmath426 , we finally obtain the result . assuming that also @xmath97 satisfies * ( h2 ) * , allows us to exchange the role of @xmath63 and @xmath121 in ( [ eq : table ] ) .
let @xmath427 and repeating the same computations as above , we obtain @xmath428 so that , finally , @xmath429
+ \int_0^t c(t)\mathrm{d}t .
\end{aligned}\ ] ] [ lem : esti ] let @xmath430 be defined as in ( [ eq : i ] ) .
then , @xmath431t , t+{{\eta}}\right [ } \int_{{{\left\|y - x_o\right\|}}\le r + { { \lambda}}+m(t_o - t)+\theta } { { \left|u(t , y)-u(s , y)\right| } } \ , \mathrm{d}y \ , .
\end{aligned}\ ] ] by the triangle inequality @xmath432 , with @xmath433 then , @xmath434 and by the @xmath17 right continuity of @xmath63 and @xmath121 in time , thanks to theorem [ teo : kruzkov ] @xmath435 for @xmath436 and @xmath437 , we have @xmath438 as @xmath439 , we use on the one hand the @xmath17 right continuity in time of @xmath63 , thanks to theorem [ teo : kruzkov ] , and on the other hand that @xmath76 , thanks to theorem [ teo : tv ] . in particular , we can use ( [ eq : afp ] ) to obtain @xmath440t , t+{{\eta}}\right [ } \int_{{{\left\|y - x_o\right\|}}\le r+{{\lambda}}+m(t_o - t)+\theta } { { \left|u(t , y)-u(s , y)\right| } } \ , \mathrm{d}y \\ & \leq & 2 \sup_{t\in \{0,t\ } \atop s\in\left]t , t+{{\eta}}\right [ } \int_{{{\left\|y - x_o\right\|}}\le r + { { \lambda}}+m(t_o - t)+\theta } { { \left|u(t , y)-u(s , y)\right| } } \ , \mathrm{d}y \ , .
\end{aligned}\ ] ] [ lem : estj ] let @xmath289 be defined as in ( [ eq : jx ] ) . then ,
@xmath441 } { \mathrm{tv}}\left(u(\tau ) \right ) { { \lambda}}\\ & & + n t { { \left\|\nabla{\partial}_u f \right\|}}_{{\mathbf{l^\infty } } } \sup_{t\in [ 0,t ] \atop s\in\left]t , t+{{\eta}}\right [ } \int_{{{\left\|y - x_o\right\|}}\le r + { { \lambda}}+m(t_o - t)+\theta } { { \left|u(t , y)-u(s , y)\right| } } \ , \mathrm{d}y \ , .
\end{aligned}\ ] ] by assumptions * ( h1 ) * , @xmath117 and therefore @xmath442 then , @xmath443 similarly to the proof of lemma [ lem : esti ] , we apply the triangle inequality and obtain @xmath444 where @xmath445 for @xmath446 , we have , thanks to ( [ eq : mu3 ] ) @xmath447 for @xmath448 , we have @xmath449 } { \mathrm{tv}}\left(u(\tau ) \right ) \ , { { \lambda}}\ , , \end{aligned}\ ] ] and for @xmath450 @xmath451 t , t+{{\eta}}\right [ } \int_{{{\left\|x - x_o\right\|}}\le r + { { \lambda}}+m(t_o - t)+\theta } { { \left|u(t , y)-u(s , y)\right| } } \ , \mathrm{d}y \ , \mathrm{d}t \\ & \leq & n { { \left\|\nabla{\partial}_u f \right\|}}_{{\mathbf{l^\infty } } } ( t+{\varepsilon } ) \sup_{t\in [ 0,t+{\varepsilon } ] \atop s\in\left]t , t+{{\eta}}\right [ } \int_{{{\left\|y - x_o\right\|}}\le r + { { \lambda}}+m(t_o - t)+\theta } { { \left|u(t , y)-u(s , y)\right| } } \ , \mathrm{d}y \ , .
\end{aligned}\ ] ] in particular , letting @xmath452 , we prove that @xmath453 and @xmath454 completing the proof .
[ lem : estl ] let @xmath293 be defined as in ( [ eq : lx ] ) and @xmath332 as in ( [ eq : m1 ] )
. then @xmath455}{\mathrm{tv}}(u(\tau)){{\lambda}}\\ & & \qquad \left .
+ t \sup_{t\in [ 0,t ] \atop s\in\left]t , t+{{\eta}}\right [ } \int_{{{\left\|y - x_o\right\|}}\le r + { { \lambda}}+m(t_o - t)+\theta } { { \left|u(t , y)-u(s , y)\right| } } \ , \mathrm{d}y \right ] \\ & & + { { \lambda}}\ , m_1 \int_0^t \int_{{{\mathbb{r}}}^n } { { \left\|\nabla ( f-{{\mathrm{div}}}f)(t , x,\cdot)\right\|}}_{{\mathbf{l^\infty } } } \mathrm{d}x \ , \mathrm{d}t \ , .
\end{aligned}\ ] ] let @xmath456 so that @xmath457 .
clearly , @xmath458 for @xmath459 and @xmath460 , we have @xmath461 proceeding as for @xmath289 , we find the following bound for @xmath462 in @xmath459 , @xmath460 .
@xmath463 } \!\!\!\ ! { \mathrm{tv}}\left(u(\tau )
\right ) { { \lambda}}\\ & & \qquad \quad + ( t+{\varepsilon } ) \sup_{t\in [ 0,t+{\varepsilon } ] \atop s\in\left]t , t+{{\eta}}\right [ } \int_{{{\left\|y - x_o\right\|}}\le r+{{\lambda}}+m(t_o - t)+\theta } { { \left|u(t , y)-u(s , y)\right| } } \ , \mathrm{d}y \biggr]\ , .
\end{aligned}\ ] ] for @xmath464 we have @xmath465 \varphi \\ & & \qquad \qquad \qquad \times \mathrm{sign}(u - v ) \ , \mathrm{d}x \ , \mathrm{d}t \ ,
\mathrm{d}y \ , \mathrm{d}s \\ & \leq & { { \lambda}}\ , m_1 \int_0^{t+{\varepsilon } } \int_{{{\mathbb{r}}}^n } { { \left\|\nabla ( f-{{\mathrm{div}}}f)(t , x,\cdot)\right\|}}_{{\mathbf{l^\infty } } } \mathrm{d}x \ , \mathrm{d}t \ , .
\end{aligned}\ ] ] to complete the proof , it is sufficient to note that @xmath466 . | consider the general scalar balance law @xmath0 in several space dimensions .
the aim of this note is to estimate the dependence of its solutions from the flow @xmath1 and from the source @xmath2 . to this aim ,
a bound on the total variation in the space variables of the solution is obtained .
this result is then applied to obtain well posedness and stability estimates for a balance law with a non local source . _
2000 mathematics subject classification : _
35l65 . _
keywords : _ multi - dimensional scalar conservation laws , krukov entropy solutions . |
correlations between the global x - ray properties of galaxy clusters have proven to be important probes of the intracluster medium ( icm ) .
studies of the relation between the x - ray luminosity ( @xmath0 ) and the emission - weighted temperature ( @xmath1 ) have been particularly powerful .
numerical simulations and analytic models that take into account the effects of gravity and shock heating of the gas only ( i.e. , the so - called `` self - similar '' models ) predict that @xmath2 for massive clusters , yet the observed relation is much steeper ; @xmath3 ( e.g. , markevitch 1998 ; allen & fabian 1998 ; arnaud & evrard 1999 ) .
a number of other observed x - ray scaling relations , for example the total cluster mass ( @xmath4)-@xmath1 and total icm mass ( @xmath5)-@xmath1 relations , have also recently been shown to deviate from their predicted scalings ( e.g. , horner et al .
1999 ; ettori & fabian 1999 ; mohr et al . 1999 ; vikhlinin et al . 1999 ; nevalainen et al .
2000 ; finoguenov et al .
2001 ; mccarthy et al .
2002 , hereafter mbb02 ) .
these discrepancies between theory and observations have motivated a number of authors to examine the potential role of `` additional '' gas physics .
for example , the heating of the icm by galactic winds and/or quasar outflows has been investigated by kaiser ( 1991 ) and also by a whole host of subsequent authors ( e.g. , evrard & henry 1991 ; bower 1997 ; balogh et al .
1999 ; wu et al .
2000 ; loewenstein 2000 ; tozzi & norman 2001 ; borgani et al .
2001 ; babul et al .
2002 , hereafter bblp02 ; mbb02 ; nath & roychowdhury 2002 ; lloyd - davies et al .
recently , the effects of radiative cooling on x - ray scaling relations have also been explored ( e.g. , bryan 2000 ; voit & bryan 2001 ; wu & xue 2002 ; thomas et al .
2002 ; voit et al .
2002 ; dav et al .
these studies find that both heating and cooling can act in a similar manner , by raising the mean entropy of the intracluster gas and , in some cases , establishing a core in the entropy profile .
this , in turn , modifies the x - ray scaling relations of clusters and ameliorates , or possibly eliminates , the discrepancies between theory and observations .
it also potentially explains the emerging observational evidence for an `` entropy floor '' in nearby groups and low mass clusters ( ponman et al .
1999 ; lloyd - davies et al .
2000 ) .
thus far , x - ray observations alone have provided evidence for the entropy floor and it has come almost entirely from low redshift ( @xmath6 ) groups / clusters .
observations of higher redshift clusters are hindered by cosmological dimming [ the bolometric x - ray surface brightness of a cluster scales as @xmath7 .
an additional , _ independent _ probe which could be used to confirm the presence of this excess entropy in low / intermediate redshift clusters and also provide new tests for their high redshift counterparts would be quite useful . here , we show that scaling relations based on the thermal sunyaev - zeldovich effect ( sunyaev & zeldovich 1972 ; 1980 ) - hereafter referred to as the `` sz effect '' - can provide such a probe .
the sz effect is a fractional change in the temperature / intensity of the cosmic microwave background ( cmb ) caused by the inverse - compton scattering of cmb photons off high energy electrons in the icm . on average , the photons gain a small amount of this energy from the scatterings and this results in a slight spectral distortion of the cmb towards clusters . at frequencies of @xmath8 ghz , the sz effect appears as a decrement in the temperature of the cmb , while at higher frequencies it appears as an increment .
the magnitude of the sz effect is determined by the integrated gas pressure along the line - of - sight through the cluster .
since heating / cooling modifies the entropy of the gas , it should be expected that the gas pressure and , thus , the sz effect will also be modified by these processes .
therefore , it can be expected that sz effect scaling relations will tell us something about the entropy history of the gas . unlike the x - ray surface brightness of a cluster ,
the sz effect is not subject to cosmological dimming and can be used to study the effects of excess entropy in the icm out to arbitrarily high redshift .
this makes scaling relations based on the sz effect particularly interesting and attractive tests .
traditionally , detecting and mapping the sz effect has been quite difficult .
this is because the effect is very weak , with typical beam - averaged decrements in the cmb temperature of only a few hundred @xmath9k ( see birkinshaw 1999 for a recent compilation ) . recently , however , extraordinary leaps in detector technology and new observing strategies have allowed observers to make reliable and routine pointed observations of the effect ( for recent data see , e.g. , grego et al .
2001 ; grainge et al . 2002
; reese et al . 2002 ) . with this advance , we feel a detailed study of sz effect scaling relations and what new insights they can give us on the thermal and spatial characteristics of the icm , especially in high redshift clusters , is timely . in this paper
, we construct sz effect scaling relations for massive clusters both with and without excess entropy and focus on how these correlations are modified by the presence of an entropy floor . along the way
, we also discuss the potential for both current and future observations to measure these relations and constrain the entropy floor level . constraining the entropy floor level out to high redshift
could possibly yield information about the source(s ) of the excess entropy , whether it be `` preheating '' by quasar outflows , radiative cooling , or some other non - gravitational process . in a companion paper ( mccarthy et al .
2003 ) , we compare the scaling relations derived here to high redshift sz effect data from the literature . to date , only a few other theoretical studies have investigated the role of excess entropy on the sz effect .
white et al . ( 2002 ) , springel et al .
( 2001 ) , da silva et al .
( 2001 ) , cavaliere & menci ( 2001 ) , and holder & carlstrom ( 2001 ) have all looked at how _ universal _
sz effect properties , such as the sz effect angular power spectrum , sz effect cluster source counts , and/or the mean universal compton parameter , are modified by an entropy floor .
however , measurements of most of these quantities are not feasible with current instrumentation and , therefore , estimates of the entropy floor level of clusters via observations of universal sz effect quantities will have to wait until the next generation of instruments [ e.g. , the sunyaev - zeldovich array ( _ sza _ ) , the arcminute microkelvin imager ( _ ami _ ) , and the array for microwave background anisotropy ( _ amiba _ ) ] come online . in a spirit similar to that of the present study , however ,
holder & carlstrom ( 2001 ) and cavaliere & menci ( 2001 ) have also examined a few sz effect scaling relations for _ individual _ clusters .
we briefly compare our findings to the results of these two studies in 5 .
the present paper is outlined as follows : in 2 , we briefly describe the analytic cluster models developed in bblp02 and employed here ; in 3 , we discuss in a general sense how and why entropy injection is expected to influence the sz effect and how this can be explored with current and future observations through comparisons with predicted scaling relations . in 4 , we derive theoretical scaling relations based on sz effect observables and quantify how excess entropy modifies these relations .
we start by examining scaling relations consisting only of sz effect observables .
these relations are most interesting because they potentially offer a completely ( x - ray-)independent way of probing the intracluster gas .
next , we explore scaling relations between the various sz effect observables and a cluster s mass . these relations , too , can potentially be measured independent of x - ray observations , since clusters can be weighed via gravitational lensing or galaxy velocity dispersion measurements .
finally , we construct and analyse scaling relations between the various sz effect and x - ray observables ( i.e. , x - ray luminosity and emission - weighted temperature ) . in 5 , we discuss and summarize our findings .
the models we consider below were developed in a flat @xmath10-cdm cosmology with @xmath11 , @xmath12 , and @xmath13 ( burles et al .
they are computed for a number of different entropy floor levels spanning the range @xmath14 kev @xmath15 .
this is approximately the range found to match the observed @xmath16 relations of both groups and hot clusters ( e.g. , ponman et al .
1999 ; lloyd - davies et al .
2000 ; tozzi & norman 2001 ; bblp02 ; mbb02 ) .
we work in physical units throughout the paper ( e.g. , @xmath17 rather than @xmath18 ) .
since an in - depth discussion of the cluster models can be found in bblp02 , we present only a brief description of the models here .
we note that the model clusters derived here represent high mass ( @xmath19 m@xmath20 ) , high temperature ( @xmath21 kev ) clusters only .
such a range reflects all of the clusters observed to date through the sz effect and is approximately the range expected to be probed by upcoming interferometric surveys ( e.g. , holder et al . 2000 ) . to mimic the standard self - similar result deduced from numerical simulations
, we implement the `` isothermal '' model of bblp02 . in this model ,
the icm is assumed to be isothermal and is in hydrostatic equilibrium with the dark halo potential .
the distribution of the dark matter in these clusters is assumed to be the same as found in recent high resolution numerical simulations ( moore et al .
1998 ; lewis et al .
this model will serve as our `` baseline '' model for assessing how an entropy floor modifies the sz effect .
the isothermal model has been tested extensively ( bblp02 ; mbb02 ) to make sure it provides a good match to the results of genuine self - similar numerical simulations , such as those done by evrard et al .
( 1996 ) . to test the effects of an entropy floor on the sz effect ,
we make use of the entropy injection ( preheated ) models of bblp02 .
the models can be summarized as follows : the dominant dark matter component , which is unaffected by the energy injection , collapses and virializes to form bound halos .
the distribution of the dark matter in such halos is assumed to be the same as for the self - similar clusters described above .
while the dark component is unaffected by energy injection , the collapse of the baryonic component is hindered by the pressure forces induced by entropy injection .
if the maximum infall velocity due purely to gravity of the dark halo is subsonic , the flow will be strongly affected by the pressure and it will not undergo accretion shocks .
it is assumed that the baryons will accumulate onto the halos _ isentropically _ at the adiabatic bondi accretion rate ( as described in balogh et al .
this treatment , however , is only appropriate for low mass halos . if the gravity of the dark halos is strong enough ( as it is expected to be in the hot clusters being considered here ) that the maximum infall velocity is transonic or supersonic , the gas will experience an additional ( generally dominant ) entropy increase due to accretion shocks . in order to trace the shock history of the gas , a detailed knowledge of the merger history of the cluster / group
is required but is not considered by bblp02 .
instead , it is assumed that at some earlier time the most massive cluster progenitor will have had a mass low enough such that shocks were negligible in its formation , similar to the low mass halos discussed above .
this progenitor forms an isentropic gas core of radius @xmath22 at the cluster center .
the entropy of gas outside of the core , however , will be affected by shocks .
recent high resolution numerical simulations suggest that the `` entropy '' profile for gas outside this core can be adequately represented by a simple analytic expression given by @xmath23 ( lewis et al .
2000 ) , where @xmath24 . for the massive , hot clusters ( @xmath21 kev ) of interest here , @xmath25 ( tozzi & norman 2001 ; bblp02 ) . following this prescription and specifying the parameters @xmath26 , @xmath27 , and @xmath28 ( as discussed in bblp02 ) completely determines the models . under all conditions ,
the gas is assumed to be in hydrostatic equilibrium within the dark halo potential .
the complicated effects of radiative cooling are neglected by these models .
the amplitude of the sz effect is directly proportional to the `` compton parameter '' ( @xmath29 ) which is given by @xmath30 where @xmath31 is the projected position from the cluster center , @xmath32 is the thomson cross - section , and @xmath33 is the electron pressure of the icm at the 3-dimensional position @xmath34 .
the integral is performed over the line - of - sight ( @xmath35 ) through the cluster .
all of the physics of the sz effect is contained within the compton parameter .
it is the sz effect analog of the x - ray surface brightness of a cluster and is a measure of the average fractional energy gain of a photon due to inverse - compton scattering while passing through a cloud of gas ( in this case , the icm ) with an electron pressure profile of @xmath36 . as discussed by bblp02 and mbb02 , the presence of excess entropy will modify both a cluster s density and temperature profiles . in the case where it is preheating 0.1 in
that gives rise to an entropy core , as in the present study , the temperature of the gas near the center of the cluster is increased and , therefore , so is the global emission - weighted temperature of the cluster ( e.g. , fig . 1 of mbb02 ) . at the same time
, the density of the gas at the cluster center is dramatically reduced ( e.g. , fig . 2 of mbb02 ) .
it turns out that , relatively speaking , preheating has a stronger influence on the density than it does on the temperature , at least at the centers of massive clusters .
the result is that the gas pressure in central regions of a cluster is reduced by preheating and , consequently , so is the cluster s compton parameter . to demonstrate this
, we plot cluster pressure profiles ( @xmath37 ) for several values of the entropy floor in figure 1 ( @xmath38 is the radius of the cluster ) .
the addition of an entropy floor leads to a decrease in the gas pressure near the cluster core .
the gas pressure in the outer regions of the clusters , however , remains relatively unchanged as the entropy increase due to gravitational shock heating dominates the non - gravitational entropy injection .
also of note is that the _ relative _ difference between the various models is greatest for the lower mass cluster .
this is expected since the lower mass cluster has a shallower potential well and , thus , is more strongly influenced by the presence of an entropy floor . with an entropy floor significantly affecting the pressure of the icm near the center of a cluster
, the compton parameter will be most strongly modified if it is evaluated within the smallest possible projected radius [ i.e. , the _ central _ compton parameter , @xmath39 . integrating ( or averaging ) the compton parameter within larger projected radii ( for example , @xmath38 , the radius of the cluster ) , on the other hand , will diminish ( though not completely remove ) the effects of entropy injection .
which of these quantities , the central or integrated compton parameters , will be the most useful for constructing scaling relations that are sensitive to the value of @xmath40 will depend upon which other cluster observables are used in the relations .
for example , in 4 , we show that a comparison of @xmath41 with @xmath1 provides a more sensitive test of entropy injection than a comparison of the integrated compton parameter with @xmath1 . on the other hand ,
a scaling relation between the integrated compton parameter and @xmath0 is more sensitive to the entropy floor level than is a scaling relation between @xmath41 and @xmath0 . ultimately , one would like to directly image the entire cluster out to the virial radius with high angular resolution and directly compare theoretical models to the images over the entire cluster profile .
the observability of these quantities will obviously depend on the details of the instrument and observing strategy .
generically , observations of the sz effect filter large - scale emission while finite resolution smears out small - scale structures . fitting to a model ( such as the isothermal @xmath42 model ; cavaliere & fusco - femiano 1976 ; 1978 ) provides a method for effectively deconvolving these effects and estimating the central and integrated compton parameters , but it is important to keep in mind that such quantities are inferred and model - dependent . provided the smallest angular scale resolved is comparable to the typical scale over which the cluster varies , the inferred @xmath41 will be reliable , while inferred integrated compton parameters will not be reliable when extrapolated beyond the filtering scale of the observations . for current interferometric observations [ such as those obtained with the berkeley illinois maryland association ( _ bima _ ) and owens valley radio observatory ( _ ovro _ ) arrays and the ryle telescope ] ,
the highest angular resolution for sz measurements is typically smaller than the core radius of most clusters observed to date ( @xmath43 30 ) while the typical large - scale filtering becomes important on scales larger than about 2. therefore , it can be expected that the inferred values of @xmath41 should be reasonably accurate while any integrated compton parameter / flux density reported on scales larger than 2 will be suspect . in recognition of the above , we construct scaling relations between central compton parameter , the integrated compton parameter within the central 1 ( which is a conservative choice for the filtering scale of current interferometers ) , and various x - ray observables .
however , we also explore relations involving the integrated compton parameter within @xmath38 , as future experiments , such as _ sza _ , will probe angular scales larger than that of current experiments .
we start our discussion of sz effect scaling relations by first examining relations between sz effect quantities only ( i.e. , the central and integrated compton parameters ) . in principle , these relations are completely independent of x - ray results , in that the latter depend on the combination of @xmath45 [ where @xmath46 is the x - ray emissivity ] whereas the sz effect depends on @xmath47 . for this reason , scaling relations which depend only on sz effect quantities have the potential of providing new insights into both the thermal and spatial properties of the icm .
as such , we will invest a little bit more effort discussing and expanding on the nature of these relations than the sz effect - x - ray relations discussed later in 4.2 - 3 .
first , because the sz effect flux density of a cluster is most commonly reported in observational studies and not the integrated compton parameter , we convert the integrated compton parameter ( symbolized as @xmath48 ) , which is given by @xmath49 into flux density ( @xmath50 ) via @xmath51 where @xmath52 describes the shape of the sz effect spectrum and is a function of the dimensionless frequency @xmath53 .
the mean temperature of the present - day cosmic microwave background , @xmath54 , is 2.728 k ( fixsen et al .
1996 ) and @xmath55 , with @xmath56 where @xmath57 at long wavelengths ( the rayleigh - jeans limit ) .
since our aim here is to achieve a _ physical _ understanding of the relationship between @xmath50 and @xmath41 , we remove the frequency dependence of the integrated sz effect flux density by dividing it by the quantity @xmath58 given in ( 4 ) .
then a comparison of the quantity @xmath59 with genuine observations at any observing frequency @xmath60 can be made simply by multiplying by the conversion factor @xmath58 . for reasons discussed in 3
, we investigate the frequency - independent sz effect flux density within a fixed angular radius of @xmath61 ( which we symbolize as @xmath62 ) and within @xmath38 , the radius of the cluster ( which we symbolize as @xmath63 ) . in figure 2 , we plot scaling relations between the ( frequency - independent ) sz effect flux densities and the central compton parameter .
the dotted lines are the self - similar results .
the short - dashed , long - dashed , dot - dashed , and solid lines represent the results of the models of bblp02 with entropy floor constants of @xmath40 = 100 , 200 , 300 , and 430 kev @xmath15 , respectively .
the top panel is for the sz effect flux density within central one arcminute , while the bottom panel is the total sz effect flux density of the cluster ( out to @xmath38 ) .
the thick lines are for @xmath37 and the thin lines are for @xmath64 .
we start by considering the dependence of @xmath65 on @xmath40 and @xmath66 in the top panel of figure 2 .
first , as discussed in 3 , entropy injection leads to a decrease in the gas pressure near the cluster core ( fig .
this , in turn , reduces the magnitude of both @xmath62 and @xmath41 . however , the relative reduction in both of these quantities is not equal .
@xmath62 is derived within larger physical radii than @xmath41 and , therefore , is less affected by the entropy floor .
the result is an increase in the normalization of the @xmath65 relation as the value of @xmath40 is increased .
for example , at @xmath37 , a cluster with a fixed value @xmath67 ( @xmath68 -1060 @xmath69 at 30 ghz ) will have an integrated decrement of @xmath70 mjy ( @xmath71 mjy at 30 ghz ) if no entropy floor is present but will 0.1 in have @xmath72 mjy ( @xmath73 mjy at 30 ghz ) if @xmath74 kev @xmath15 .
furthermore , this trend is amplified at higher redshifts since the sz effect flux density within a fixed angular size probes larger physical regions at higher redshifts ( and , therefore , is less affected by the entropy floor ) .
compare the difference between the models for the @xmath37 lines and the @xmath64 , for example .
the slope of @xmath65 relation is also modified by the presence of an entropy floor .
for example@xmath75 , at @xmath37 and 1.0 ( respectively ) , the self - similar model approximately predicts @xmath76 while the @xmath77 kev @xmath15 model approximately predicts @xmath78 a more precise analytic expression for the @xmath65 relation at any @xmath79 and any entropy floor in the range @xmath80 kev @xmath15 is presented in table 1 . at a fixed redshift , the presence of an entropy floor flattens the relationship between @xmath62 and @xmath41 because clusters with small values of @xmath41 ( i.e. , low mass clusters ) are more strongly influenced ( relatively speaking ) by entropy injection than clusters with large values of @xmath41 .
it is worth noting that , for the most part , the @xmath81 relations ( and , in fact , virtually every relation discussed in the paper ) exhibit almost perfect power - law behavior .
this may seem rather surprising in light of the fact that most x - ray scaling relations show some kind of a break from power - law trends ( see bblp02 , for example ) .
we note , however , that the break in the @xmath16 relation , for example , occurs at the transition between groups and clusters ( roughly @xmath82 kev ) .
this is below the range of temperatures ( and central compton parameters ) studied in the present paper .
we verify that a break does occur for low mass systems ( i.e. , for systems with @xmath83 ) but , because such low mass systems have very weak integrated sz effect signals , there is little hope of observing this feature in the near future . in the bottom panel of figure 2 , we plot the @xmath84 relations . with the quantity @xmath63 only slightly affected by the presence of an entropy floor ( and , therefore , the relative difference in the normalizations of the models are at a maximum ) , these relations are more sensitive to entropy injection than those plotted in the top panel of figure 2 .
for example , for a @xmath37 cluster with a fixed value of @xmath67 , the @xmath85 kev @xmath15 models predicts a flux density that is roughly three times larger than that predicted by the self - similar model ( as opposed to being only about two times larger for the relation presented in the top panel ) .
the steepness of the relation is also altered by an entropy floor . using simple scaling arguments
, one can easily arrive at the predicted self - similar result .
combining @xmath86 and @xmath87 with @xmath88 ( which is just the virial theorem ) yields @xmath89 in excellent agreement with the predictions of the bblp02 isothermal model . however , injection of entropy flattens the relation and at @xmath37 and 1.0 , the @xmath77 kev @xmath15 model approximately predicts @xmath90 an analytic expression for the @xmath84 relation at any @xmath91 and any entropy floor between 100 kev @xmath15 @xmath92 kev @xmath15 is given in table 1 .
since the @xmath84 relation is more sensitive to entropy injection ( relatively speaking ) than the @xmath65 relation , this means that observations obtained with future experiments , such as the _ sza _ , _ amiba _ , and _ ami _ ( which will probe larger angular scales than that of current interferometers ) , will ( at least in theory ) have the best chance of placing tight constraints on the non - gravitational entropy of high redshift clusters .
we note that most of the dependence of the relations plotted in figure 2 ( both the top and bottom panels ) on the entropy floor level , @xmath40 , is due to the central compton parameter , @xmath41 .
a completely resolved central decrement formally requires infinite resolution and there is little to be gained by increasing the resolution beyond the smallest scale on which the cluster structure varies . for diffuse emission ,
higher resolution normally means less signal per resolution element , so pushing to very high resolution will not lead to much new information .
alternatively , the central compton parameter can be estimated by fitting a model ( e.g. , the isothermal @xmath42 model ) to the the observed sz effect surface brightness and extrapolating it to the cluster center ( as discussed briefly in 3 ) .
this method is widely used to estimate @xmath41 ( e.g. , carlstrom et al .
1996 ; holzapfel et al . 1997 ; grego et al . 2000 ; 2001 ; reese et al . 2000 ; 2002 ; pointecouteau et al . 2001 ; 2002 ; jones et al . 2002 ; grainge et al . 2002 ) .
while the _ statistical _ measurement error on @xmath41 for current data is typically only of order @xmath93 @xmath9k ( which is relatively small compared to the differences between the models plotted in figure 2 , assuming @xmath59 is known ) , it has yet to be demonstrated that the _ systematic _ error ( due to assuming an incorrect surface brightness model ) for current data is negligible . if the models employed in the extrapolation provide poor descriptions of the surface brightness profiles , then one would expect the results to vary as a function of instrument characteristics ( e.g. , resolution , field of view ) .
however , a comparison of the results of various studies ( which made use of different instruments ; e.g. , _ ryle _ telescope , _ bima / ovro _ , and _ suzie _ ) of the same clusters reveals that the agreement is quite good , often within one sigma statistical uncertainty ( see holzapfel et al .
1997 , reese et al .
2002 , and jones et al .
2002 , for example ) .
thus , the extrapolation procedure seems to be a viable way estimating the central compton parameter at current sensitivity . by modeling `` mock '' ( future ) _ sza _ observations
, we also demonstrated that this extrapolation procedure is an accurate way of estimating the true underlying values of @xmath41 and @xmath50 of the bblp02 cluster models ( see mccarthy et al .
2003 ) . finally , we point out that part of the motivation for our study of the @xmath44 relations ( and the other scaling relations involving these quantities ) comes from the fact that observational studies often use either @xmath50 or @xmath41 to characterize a cluster s sz effect but not usually both .
for example , integrated sz effect flux densities are often associated with large - beam single - dish experiments , while estimates of the central compton parameter normally come from high resolution interferometric observations . however , if one is able to determine @xmath94 ( i.e. , the sz effect `` surface brightness '' profile ) from a single dataset then , strictly speaking , it is nt necessary to use scaling relations to probe the entropy of the icm .
for example , the sz effect surface brightness profile could be used together with the x - ray surface brightness profile ( if it is known ) to determine the true 3-dimensional entropy distribution of the gas ( with some assumptions about the geometry of the cluster ) .
the disadvantages of this method are : ( 1 ) it requires x - ray observations , and ( 2 ) it probably can not be used for the most distant clusters , since the x - ray signal - to - noise ratio falls sharply with increasing @xmath66 . alternatively
, the entropy distribution could be constrained by comparing the observed sz effect surface brightness profile with theoretically predicted profiles .
this is similar to using scaling relations , since the entropy distribution is being inferred and not measured , but with the important difference that all of the available sz effect information , @xmath94 , is being used in the comparison .
we are in the process of exploring these methods and they will be addressed in detail in a subsequent paper .
for now , we stick with the scaling relations derived above , which are more readily comparable with available observations . 0.1 in in figure 3 , we present scaling relations between @xmath41 and the total cluster dark matter mass within the radius @xmath96 [ i.e. , @xmath97 .
this is the radius within which the mean dark matter mass density is 500 times the critical density at @xmath66 = 0 .
the lines hold the same meaning as in figure 2 . for clarity ,
we plot the @xmath98 predictions in the left - hand panel and the @xmath64 predicts in the right - hand panel .
it is apparent that entropy injection has a substantial effect on the @xmath95 relation .
both the normalization and the steepness of the relation are modified .
first , the normalizations imply that , for a cluster of given mass at a given redshift , entropy injection tends to diminish the strength of @xmath41 .
for example , at @xmath37 , a cluster with @xmath99 will have @xmath100 in the absence of an entropy floor , but the central compton parameter is only half this value for an entropy floor of @xmath101 kev @xmath15 .
this corresponds to a difference of nearly @xmath102 at 30 ghz ( which is large compared to the typical _ bima / ovro _ statistical measurement error of @xmath103 ; reese et al .
2000 ; 2002 ) .
of course , there will also be a measurement error associated with the cluster mass .
typically , x - ray - determined masses have associated statistical uncertainties of about 20% ( nevalainen et al .
2000 ; finoguenov et al .
2001 ) . therefore , based on figure 3
, it should be possible to constrain @xmath40 to within about @xmath104 kev @xmath15 with current data .
physically , the diminution in the normalization @xmath95 relation can be understood since entropy injection greatly reduces the pressure of the cluster gas near the core ( and therefore @xmath41 ; see figure 1 ) but does not affect its dark matter mass .
it is also worth noting that the normalizations of the individual models are a function of redshift .
for example , the @xmath105 kev @xmath15 at @xmath37 predicts a relation somewhat similar to that of the @xmath106 kev @xmath15 at @xmath107 or the @xmath85 kev @xmath15 at @xmath108 .
so , it is extremely important to consider the evolution of galaxy clusters when comparisons are made to observational data ( i.e. , information about the redshift of the clusters is required in order to constrain the level of the entropy floor ) .
another interesting point is that the difference between the normalizations of the models becomes greater at higher redshift .
for example , the difference between @xmath41 for the isothermal and @xmath105 kev @xmath15 models of a @xmath109 cluster at @xmath64 is about 3.7 times larger than at @xmath98 .
this is a result of the fact that the physical size of the isentropic core ( @xmath26 ; discussed in 2 ) is a larger fraction of the cluster virial radius at higher redshifts .
this trend of increasing @xmath110 with redshift can be understood as follows . in bblp02
, it was pointed out that there is a critical mass threshold which determines the importance of gravitational shock heating . for clusters with masses above this threshold , shock heating significantly increases the entropy of the intracluster gas and it begins to dominate the injected non - gravitational entropy ( except for in the very central regions of the clusters ) . for clusters with masses below this threshold ,
shock heating is unimportant .
the critical mass also sets size of the isentropic core , @xmath26 . incorporating the results of recent high resolution hydrodynamic simulations ( lewis et al . 2000 ) , bblp02 defined the critical mass to be that which gives rise to a gas temperature at @xmath38 that is equal to one half of the cluster virial temperature .
using this constraint , it is relatively straightforward to show that the critical mass ( for a fixed entropy floor level ) increases with redshift and , therefore , so does the ratio @xmath110 . for gas in the core of the cluster ( with entropy equal to @xmath40 ) , @xmath111 and , therefore , @xmath112 . at high redshifts ,
the cluster gas is denser and , in the case of an @xmath113 universe ( assumed here for simplicity ) , scales as @xmath114 .
therefore , the temperature of the gas in the core scales as @xmath115 .
the gas temperature at the virial radius , on the other hand , scales as @xmath116 .
thus , the ratio @xmath117 increases with redshift and in order to match the temperature constraint described above , the critical mass threshold and @xmath110 must increase with @xmath66 as well .
the steepness of the @xmath95 relation is also affected by an entropy floor .
if clusters are self - similar , then @xmath87 , @xmath88 , and @xmath118 ( evrard et al .
this leads to @xmath119 while at @xmath37 and 1.0 , the @xmath77 kev @xmath15 model predicts @xmath120 a more precise form the @xmath95 relation as a function of both the entropy floor ( @xmath40 ) and redshift is presented in table 1 [ valid for @xmath121 . the steepening of the relation with the increase in the level of the entropy floor can also be understood in an intuitive sense .
it stems from the fact that the relative decrease in the gas pressure at the cluster center is strongest for low mass clusters ( which have the shallowest potential wells ; see figure 1 ) . again , because of the increasing importance of entropy injection as one goes to low masses , a break from power - law behavior in the @xmath95 relation occurs at the transition between groups and clusters ( not shown ) . for entropy floors of @xmath122 kev @xmath15 ,
the break becomes quite noticeable for systems with @xmath123 , which is below the range of system masses that we are most interested in and far below the range of system masses that can be observed presently .
we have found that , at least theoretically , the best hope for testing galaxy clusters for the presence of excess entropy using the @xmath95 relation is at high redshift , although our results reveal that the relation is sensitive to the entropy floor level even at low redshifts .
the quantity @xmath124 is usually determined through x - ray observations ( e.g. , ettori & fabian 1999 ) and , hence , we have placed it in the `` @xmath41-x - ray scaling relations '' subsection . however , gravitational lensing has also been used recently to trace the mass profiles of clusters out to radii near or exceeding that of @xmath96 with statistical uncertainties similar to that of x - ray - determined masses ( e.g. , clowe & schneider 2001 ; gray et al .
2002 ; athreya et al .
so , like the @xmath44 relation , the @xmath95 relation can also potentially be measured completely independent of x - ray results . in figure 4 , we present scaling relations between @xmath41 and the mean emission - weighted gas temperature ( @xmath1 ) of a cluster .
the lines hold the same meaning as in figure 2 . of the scaling relations presented thus far , the @xmath125 relation is the most sensitive to the entropy floor level .
there is a two - fold effect in that entropy injection diminishes @xmath41 but also _ increases _
@xmath1 for a cluster of mass @xmath126 ( see , e.g. , fig . 1 of mbb02 ) .
for example , at @xmath37 , a cluster that has @xmath127 kev and no entropy floor will have a value of @xmath41 that is about 3.5 times larger than a cluster ( also with @xmath127 kev ) that has an entropy floor of @xmath101 kev @xmath15 . at 30 ghz
, this corresponds to a difference of @xmath128 .
current sz effect and x - ray data can constrain the central decrement ( compton parameter ) and emission - weighted temperature of massive clusters at about the 10% level ( e.g. , reese et al . 2002 ) , making the @xmath125 relation an extremely promising test of non - gravitational entropy injection . furthermore
, the difference in the normalizations becomes progressively larger with increasing redshift .
the slopes of the predicted relations are also very sensitive to the level of entropy injection . combining @xmath87 with @xmath88 , yields the self - similar result @xmath129 while at @xmath37 and @xmath130 , the @xmath77 kev @xmath15 model predicts @xmath131 an analytic expression for the @xmath132 relation as a function of both the entropy floor ( @xmath40 ) and redshift is presented in table 1 ( valid for @xmath133 kev ) .
lastly , we note that a break in slope of the relations occurs at @xmath82 kev , in good qualitative agreement with 0.1 in the position and steepness of the break found in the @xmath134 relation of cavaliere & menci ( 2001 ; see their figure 4 ) .
unfortunately , present sz effect observations are essentially limited to systems with @xmath135 kev . in figure 5 , we plot scaling relations between the central @xmath29 parameter and the total bolometric x - ray luminosity ( @xmath0 ) of a cluster .
the lines hold the same meaning as in figure 2 .
these relations are more complicated than those presented immediately above because both of the coordinates , @xmath41 and @xmath0 , are affected by entropy injection . very interestingly , however , the @xmath132 relation is almost unchanged by the addition of an entropy floor .
both the self - similar and entropy floor models predict very similar slopes and normalizations
( at all redshifts ) . at first sight
, this may seem rather surprising .
however , the predicted @xmath132 relations can be understood as follows .
the luminosity of a cluster of fixed mass is reduced by entropy injection because @xmath136 and entropy injection greatly reduces the gas density near the cluster core , where most of the x - ray emission originates . in the case of the bblp02
entropy floor models , the result is that @xmath137 , roughly ( instead of @xmath138 in the self - similar case ) . on the other hand , we showed in 4.2.1 that an increase in the entropy floor level strongly diminishes @xmath41 for a cluster of fixed mass and this leads to @xmath139 .
this implies that entropy injection gives rise to @xmath140 ( roughly ) , which is very close to the self - similar scaling of @xmath141 a precise analytic expression for the @xmath132 relation as a function of both the entropy floor ( @xmath40 ) and redshift is presented in table 1 ( valid for @xmath142 ergs s@xmath143 ) . despite being insensitive to the entropy floor level , the @xmath132 relation has at least two other potential uses . 0.1 in first , it could be used as a consistency check of other x - ray and sz effect scaling relations .
second , it could allow one to deduce the x - ray luminosity of clusters in cases where only sz effect observations are available ( or vice - versa ) without having to worry about the role of `` additional '' gas physics .
the accuracy of this will be limited by the relatively large amount of intrinsic scatter present in the x - ray luminosities of clusters ( see , e.g. , studies of the @xmath16 relation ; markevitch 1998 ; novicki et al .
2002 ) and by the measurement errors associated with @xmath41 .
apart for the central @xmath29 parameter , we also investigated theoretical scaling relations between the frequency - independent integrated sz effect flux density and the x - ray observables @xmath124 , @xmath0 , and @xmath1 .
again , we evaluated @xmath59 within a fixed angular radius of one arcminute and within @xmath38 .
since these relations can be reconstructed from figures 2 - 5 or table 1 , we summarize the important points here only . for the very same reasons discussed in 4.2.1
, the @xmath144 relation has its normalization decreased by entropy injection and , also , the relation becomes steeper .
how large these effects will be depends upon which radius @xmath59 is evaluated within . because the dark matter mass is unaffected by entropy injection , the @xmath144 relation will _ always _ be most strongly modified if @xmath59 is evaluated within small projected radii ( where entropy injection is most prevalent ) .
thus , evaluating @xmath59 within the central 1 arcminute provides a more sensitive test of the entropy floor than if @xmath59 is measured within @xmath38 .
of course , it follows that @xmath95 relation is the most sensitive to the entropy floor level of any and all of the sz effect-@xmath124 relations .
the @xmath144 relation , however , could offer a useful consistency check .
the @xmath145 relation also becomes steeper and has its normalization decreased with increasing values of @xmath40 ( @xmath59 is diminished while @xmath1 is increased ) .
the maximum ( relative ) difference between the various models occurs when the sz effect flux densities are measured within small projected radii , although the increase in @xmath1 is large enough such that the relation is sensitive to the entropy floor level even when @xmath59 is measured within @xmath38 . again , these relations offer useful consistency checks of the more entropy floor - sensitive @xmath125 relation .
finally , entropy injection has an interesting effect on the @xmath146 relation@xmath147 .
both the sz effect flux density and the luminosity are reduced as the level of entropy injected is increased , similar to the @xmath132 relation . however , in the case of the present relation , the reduction in the luminosity is greater than the reduction in the sz effect flux density ( relatively speaking ) .
this is because @xmath50 is much less sensitive than @xmath41 to the entropy floor level , @xmath40 .
the result is that the normalization of the @xmath146 relation is increased by entropy injection .
the sensitivity of this relation to the entropy floor level depends on what radius the flux density is evaluated within .
it is most sensitive when @xmath50 is integrated out to @xmath38 , since the relative reduction in @xmath50 is at a minimum .
it is least sensitive when the sz effect is evaluated at zero projected radius ( i.e. , @xmath132 ) .
comparisons of observed and predicted cluster x - ray scaling relations has led to suggestions that important ( non - gravitational ) gas physics is being neglected in analytic models and numerical simulations .
it has been found that the presence of a core in the entropy profiles of clusters ( which may be generated through heating and/or cooling processes ) can account for the deviations between observations and theory ( e.g. , bower 1997 ; balogh et al .
1999 ; ponman et al .
1999 ; wu et al .
2000 ; lloyd - davies et al .
2000 ; tozzi & norman 2001 ; bialek et al .
2001 ; borgani et al . 2001 ; voit & bryan 2001 ; bblp02 ; mbb02 ; voit et al .
2002 ; dav et al .
2002 ; lloyd - davies et al .
so far , x - ray observations have provided the only evidence for this excess entropy and it has come from low redshift ( @xmath148 ) clusters alone .
cosmological dimming is partially responsible for the lack of information on the entropy of more distant groups / clusters . in the interest of achieving a better physical understanding of important non - gravitational processes in high redshift clusters ( and
how these processes affect cluster formation and evolution ) , we have derived a number of theoretical scaling relations based on sz effect quantities and have analyzed how these relations are modified by the addition of an entropy floor .
we find that entropy injection reduces the gas pressure in the cores of clusters and , therefore , diminishes the amplitude of both the central and integrated compton parameters ( which are directly proportional to the sz effect ) .
this translates to a steepening of- and a reduction in the normalization of the @xmath95 and @xmath144 relations .
scaling relations between the compton parameter(s ) and x - ray temperature or luminosity are more complicated because both @xmath1 and @xmath0 are also affected by entropy injection ( whereas the cluster mass @xmath126 is not ) .
because @xmath1 is increased by an entropy floor , the @xmath125 and @xmath145 relations have their normalizations severely reduced by entropy injection .
these relations are also considerably steeper than in the case where no entropy floor is present ( i.e. , the isothermal self - similar model ) . on the other hand , because @xmath0 is decreased with entropy injection , the @xmath132 relation is almost unaffected by entropy injection , while the normalization of the @xmath146 relation is increased by entropy injection . finally ,
the @xmath149 , a relation that can ( in principle ) be measured entirely through sz effect observations , has its normalization increased by an entropy floor ( the relative reduction in @xmath41 is much greater than in @xmath59 ) .
the relation is also slightly flattened .
analytic expressions for these relations as a function of entropy floor level and redshift can be found in table 1 .
we are aware of only two other studies that have examined the effects of an entropy floor on cluster sz effect scaling relations : holder & carlstrom ( 2001 ) and cavaliere & menci ( 2001 ) . using simplified cluster models ,
holder & carlstrom ( 2001 ) illustrated the potential power of high-@xmath66 sz effect observations when it comes to studying the physics of icm .
their figure 1 shows the impact of an entropy floor on the central sz effect decrement - cluster core radius relation at @xmath150 .
they find that the addition of an entropy floor tends to decrease the central decrement and increase the cluster core radius for clusters of a fixed total mass .
this is in very good qualitative agreement with our figure 3 , which illustrates that the normalization of the @xmath95 is decreased by entropy injection , and also with our figure 2 , which demonstrates that for clusters of a fixed @xmath41 , the sz effect flux density is increased by an entropy floor and , therefore , the sz effect surface brightness becomes less centrally peaked ( i.e. , the core radius grows ) .
cavaliere & menci ( 2001 ) , on the other hand , implemented a semi - analytic model of galaxy formation and clustering and computed a _
present - day _ @xmath151 scaling relation and analyzed how it was modified by entropy injection from stellar winds and supernovae blasts . in qualitative agreement with the results presented in 4 , these authors found that entropy injection tends to diminish the compton parameter for a cluster of given mean emission - weighted temperature , @xmath1 .
however , our theoretical results are not directly comparable to those of cavaliere & menci ( 2001 ) since their focus was mainly on nearby ( @xmath152 ) groups , whereas we have investigated massive , distant clusters such as those regularly observed by the _ bima / ovro _ arrays and the _ ryle _ telescope .
we also point out that the majority of clusters expected to be found in upcoming `` blind '' surveys will be at high redshift ( @xmath153 ; holder et al .
2000 ) . finally , the cavaliere & menci ( 2001 ) cluster models only invoked low levels of entropy injection ( @xmath154 kev @xmath15 ) consistent with measurements of nearby groups ( ponman et al .
1999 ; lloyd - davies et al .
2000 ) , whereas we have generally focused on higher levels of entropy injection ( which are required to match x - ray observations of nearby massive clusters ; tozzi & norman 2001 ; bblp02 ; mbb02 ) .
one thing we have not addressed in the present study is the effect of radiative cooling on sz effect scaling relations . correctly modeling the effects of radiative cooling in numerical simulations and analytic cluster models
is a daunting task ( see for balogh et al .
2001 for a detailed study of the `` cooling crisis '' ) .
cooling can significantly modify not only the icm density and temperature distributions but it can also modify the underlying dark matter distribution of clusters ( see fig . 12 of lewis et al .
2000 , for example ) .
in addition , cooling gives rise to other `` sub - grid '' processes such as star formation , outflows , and supernovae explosions .
these , too , will also modify the icm s structure and appearance .
while modeling this is quite difficult , we know from both optical and x - ray observations of clusters that cooling must be occurring , at least at some level , and , ultimately , must be modeled correctly in order to attain a deep physical understanding of the observed properties of clusters .
numerical simulations and analytic models which attempt to take this complicated process into account ( e.g. , voit & bryan 2001 ; da silva et al .
2001 ; wu & xue 2002 ; voit et al .
2002 ; dav et al .
2002 ; white et al .
2002 ) are apparently able to match many of the features of the x - ray scaling relations of nearby clusters . in this sense
, cooling has very similar effects on the icm to those induced by `` preheating '' .
cooling has an advantage over preheating in that it does not require some ( as of yet ) unknown source to transfer large amounts of energy into the icm .
however , cooling alone leads to an expected amount of cooled gas in groups and clusters that far exceeds what is observed ( balogh et al .
feedback , be it through preheating or heating after cluster formation , must also be an important process .
we are currently in the process of adding radiative cooling to the bblp02 analytic models ( mccarthy et al . in preparation ) with the intention of exploring how sz effect and x - ray scaling relations are modified by it .
as far as we know , the effects of cooling on sz effect scaling relations have not yet been examined . while , in general , we do not expect the results of the present study to change significantly with the inclusion of cooling ( since cooling affects the icm in a very similar way to that of preheating ) , in detail
, we do anticipate subtle ( or perhaps not so subtle ) differences to be present .
this may also lead to a way of separating out the relative contributions of cooling and heating to cluster scalings .
finally , we note that with the recent release of the reese et al .
( 2002 ) data , there are now enough sz effect observations to begin testing the `` entropy floor '' hypothesis by mapping out the various correlations involving the sz and x - ray observables , and comparing these to their theoretical counterparts derived in this paper
. it would be most interesting to see if these correlations also favor the existence of an entropy floor , and whether the level of this floor is comparable to that required to explain the x - ray trends .
this will be the focus of mccarthy et al .
( 2003 ) , a companion paper to this one .
0.25 in we thank the referee for very useful comments and suggestions .
we also thank peng oh , kathy romer , and mark voit for helpful discussions .
a. b. would like to acknowledge the hospitality extended to him by the canadian institute for theoretical astrophysics during his tenure as cita senior fellow .
i. g. m. is supported by a postgraduate scholarship from the natural sciences and engineering research council of canada ( nserc ) .
a. b. is supported by an nserc operating grant , g. p. h. is supported by the w. m. keck foundation , and m. l. b. is supported by a pparc rolling grant for extragalactic astronomy and cosmology at the university of durham .
allen , s. w. , & fabian , a. c. 1998 , mnras , 297 , l57 arnaud , m. , & evrard , a. e. 1999 , mnras , 305 , 631 athreya , r. m. , et al .
2002 , a&a 384 , 743 babul , a. , balogh , m. l. , lewis , g. f. , & poole , g. b. 2002 , mnras , 330 , 329 balogh , m. l. , babul , a. , & patton , d. r. 1999 , mnras 307 , 463 balogh , m. l. , pearce , f. r. , bower , r. g. , & kay , s. t. 2001 , mnras , 326 , 1228 bialek , j. j. , evrard , a. e. , & mohr , j. j. 2001 , apj , 555 , 597 birkinshaw , m. 1999 , phys .
, 310 , 97 borgani , s. , et al . 2001 , apj , 559 , l71 bower , r. g. 1997 , mnras , 288 , 355 bryan , g. l. 2000 , apj , 544 , l1 burles , s. , nollett , k. m. , & turner , m. s. 2001 , apj , 552 , l1 cavaliere , a. , & fusco - femiano , r. 1976 , a&a , 49 , 137 cavaliere , a. , & fusco - femiano , r. 1978 , a&a , 70 , 677 cavaliere , a. , & menci , n. 2001 , mnras , 327 , 488 clowe , d. , & schneider , p. 2001
, a&a , 379 , 384 da silva , a. c. , et al .
2001 , apj , 561 , l15 dav , r. , katz , n. , & weinberg , d. h. 2002 , apj , 579 , 23 ettori , s. , & fabian , a. c. 1999 , mnras , 305 , 834 evrard , a. e. , & henry , j. p. 1991 , apj , 383 , 95 evrard , a. e. , metzler , c. a. , & navarro , j. f. 1996 , apj , 469 , 494 finoguenov , a. , reiprich , t. h. , & bhringer , h. 2001 , a&a , 368 , 749 fixsen , d. j. , et al .
1996 , apj , 473 , 576 grainge , k. , et al .
2002 , mnras , 329 , 890 gray , m. e. , et al .
2002 , apj , 568 , 141 grego , l. , et al .
2001 , apj , 552 , 2 holder , g. p. , & carlstrom j. e. , 2001 , apj , 558 , 515 holder , g. p. , et al .
2000 , apj , 544 , 629 horner , d. j. , mushotzky , r. f. , & scharf , c. a. 1999 , apj , 520 , 78 jones , m. e. , et al .
2001 , mnras , submitted ( astro - ph/0103046 ) kaiser , n. 1991 , apj , 383 , 104 lewis , g. f. , et al .
2000 , apj , 536 , 623 lloyd - davies , e. j. , bower , r. g. , & ponman , t. j. 2002 , mnras , submitted ( astro - ph/0203502 ) lloyd - davies , e. j. , ponman , t. j. , & cannon , d. b. 2000 , mnras , 315 , 689 loewenstein , m. 2000 , apj , 532 , 17 markevitch , m. 1998 , apj , 504 , 27 mccarthy , i. g. , babul , a. , & balogh , m. l. 2002 , apj , 573 , 515 mccarthy , i. g. , holder , g. p. , babul , a. , & balogh , m. l. 2003 , apj , submitted mohr , j.j . , mathiesen , b. , & evrard , a.e . 1999 ,
apj , 517 , 627 moore , b. , et al .
1998 , apj , 499 , l5 nath , b. b. , & roychowdhury , s. 2002 , mnras , 333 , 145 nevalainen , j. , markevitch , m. , & forman , w. 2000 , apj , 532 , 694 novicki , m. c. , sornig , m. , & henry , j. p. 2002 , apj , submitted ( astro - ph/0208468 ) .
ponman , t. j. , cannon , d. b. , & navarro , j. f. 1999 , nature , 397 , 135 reese , e. d. , et al .
2000 , apj , 533 , 38 reese , e. d. , et al .
2002 , apj , 581 , 53 springel , v. , white , m. , & hernquist , l. 2001 , apj , 549 , 681 sunyaev , r. , & zeldovich , y. 1972 , comments astrophys .
space phys . , 2 , 66 sunyaev , r. , & zeldovich , y. 1980 , mnras , 190 , 413 thomas , p. a. , muanwong , o. , kay , s. t. , & liddle , a. r. 2002 , mnras , 330 , l48 tozzi , p. , & norman , c. 2001 , apj , 546 , 63 vikhlinin , a. , forman , w. , & jones , c. 1999 , apj , 525 , 47 voit , m. g. , & bryan , g. l. 2001 , nature , 414 , 425 voit , m. g. , bryan , g. l. , balogh , m. l. , & bower , r. g. 2002 , apj , 576 , 601 white , m. , hernquist , l. , & springel , v. 2002 , apj , 579 , 16 wu , k. k. s. , fabian , a. c. , & nulsen , p. e. j. 2000 , mnras , 318 , 889 wu , x. , & xue , y. 2002 , apj , 572 , l19 @xmath155 + & @xmath156 & @xmath157 + & @xmath158 & @xmath159 + & @xmath28 & @xmath160 + & @xmath42 & @xmath161 + @xmath162 + & @xmath156 & @xmath163 + & @xmath158 & @xmath164 + & @xmath28 & @xmath165 + & @xmath42 & @xmath166 + @xmath167 + & @xmath156 & @xmath168 + & @xmath158 & @xmath169 + & @xmath28 & @xmath170 + & @xmath42 & @xmath171 + @xmath172 + & @xmath156 & @xmath173 + & @xmath158 & @xmath174 + & @xmath28 & @xmath175 + & @xmath42 & @xmath176 + @xmath177 + & @xmath156 & @xmath178 + & @xmath158 & @xmath179 + & @xmath28 & @xmath180 + & @xmath42 & @xmath181 + | x - ray observations of an `` entropy floor '' in nearby groups and clusters of galaxies offer evidence that important non - gravitational processes , such as radiative cooling and/or `` preheating '' , have strongly influenced the evolution of the intracluster medium ( icm ) .
we examine how the presence of an entropy floor modifies the thermal sunyaev - zeldovich ( sz ) effect .
a detailed analysis of scaling relations between x - ray and sz effect observables and also between the two primary sz effect observables is presented .
we find that relationships between the central compton parameter and the temperature or mass of a cluster are extremely sensitive to the presence of an entropy floor .
the same is true for correlations between the integrated compton parameter and the x - ray luminosity or the central compton parameter .
in fact , if the entropy floor is as high as inferred in recent analyses of x - ray data , a comparison of these correlations with both _
current _ and future sz effect observations should show a clear signature of this excess entropy .
moreover , because the sz effect is redshift - independent , the relations can potentially be used to track the evolution of the cluster gas and possibly discriminate between the possible sources of the excess entropy . to facilitate comparisons with observations , we provide analytic fits to these scaling relations . |
we report our analysis of the web traffic of approximately one thousand residential users over a two - month period .
this data set preserves the distinctions between individual users , making possible detailed per - user analysis .
we believe this is the largest study to date to examine the complete click streams of so many users in their place of residence for an extended period of time , allowing us to observe how actual users navigate a hyperlinked information space while not under direct observation .
the first contributions of this work include the discoveries that the popularity of web sites as measured by distinct visitors is unbounded ; that many of the power - law distributions previously observed in web traffic are aggregates of log - normal distributions at the user level ; and that there exist two populations of users who are distinguished by whether or not their web activity is largely mediated by portal sites .
a second set of contributions concerns our analysis of browsing sessions within the click streams of individual users .
the concept of a web session is critical to modeling real - world navigation of hypertext , understanding the impact of search engines , developing techniques to identify automated navigation and retrieval , and creating means of anonymizing ( and de - anonymizing ) user activity on the web .
we show that a simple timeout - based approach is inadequate for identifying sessions and present an algorithm for segmenting a click stream into _ logical sessions _ based on referrer information .
we use the properties of these logical sessions to show that actual users navigate hypertext in ways that violate a stateless random surfer model and require the addition of backtracking or branching .
finally , we emphasize which aspects of this data present possible opportunities for anomaly detection in web traffic .
robust anomaly detection using these properties makes it possible to uncover `` bots '' masquerading as legitimate user agents .
it may also undermine the effectiveness of anonymization tools , making it necessary to obscure additional properties of a user s web surfing to avoid betraying their identity .
in the remainder of this paper , after some background and related work , we describe the source and collection procedures of our web traffic data .
the raw data set includes over 400 million http requests generated by over a thousand residential users over the course of two months , and we believe it to provide the most accurate picture to date of the hypertext browsing behavior of individual users as observed directly from the network .
our main contributions are organized into three sections : * we confirm earlier findings of scale - free distributions for various per - site traffic properties aggregated across users .
we show this also holds for site popularity as measured by the number of unique vistors .
( [ section - host ] ) * we offer the first characterization of individual traffic patterns involving continuous collection from a large population .
we find that properties such as jump frequency , browsing rates , and the use of portals are not scale - free , but rather log - normally distributed .
only when aggregated across users do these properties exhibit scale - free behavior .
( [ section - user ] ) * we investigate the notion of a web `` session , '' showing that neither a simple timeout nor a rolling average provide a robust definition .
we propose an alternative notion of _ logical _ session and provide an algorithm for its construction . while logical sessions have no inherent temporal scale , they are amenable to the addition of a timeout with little net effect on their statistical properties .
( [ section - session ] ) we conclude with a discussion of the limitations of our data , the implications of this work for modeling and anomaly detection , and potential future work in the area .
internet researchers have been quick to recognize that structural analysis of the web becomes far more useful when combined with actual _ behavioral _ data .
the link structure of the web can differ greatly from the set of paths that are actually navigated , and it tells us little about the behavior of individual users .
a variety of behavioral data sources exist that can allow researchers to identify these paths and improve web models accordingly .
the earliest efforts have used browser logs to characterize user navigation patterns @xcite , time spent on pages , bookmark usage , page revisit frequencies , and overlap among user paths @xcite .
the most direct source of behavioral data comes from the logs of web servers , which have been used for applications such as personalization @xcite and improving caching behavior @xcite .
more recent efforts involving server logs have met with notable success in describing typical user behavior @xcite . because search engines serve a central role in users navigation , their log data is particularly useful in improving search results based on user behavior @xcite .
other researchers have turned to the internet itself as a source of data on web behavior .
network flow data generated by routers , which incorporates high - level details of internet connections without revealing the contents of individual packets , has been used to identify statistical properties of web user behavior and discriminate peer - to - peer traffic from genuine web activity @xcite .
the most detailed source of behavioral data consists of actual web traffic captured from a running network , as we do here .
the present study most closely relates to the work of qiu _ et al . _
@xcite , who used captured http packet traces to investigate a variety of statistical properties of users browsing behavior , especially the extent on which they appear to rely on search engines in their navigation of the web .
we have also used captured http requests in our previous work to describe ways in which pagerank s random - surfer model fails to approximate actual user behavior , which calls into question its use for ranking search results @xcite .
one way of overcoming these shortcomings is to substitute actual traffic data for ranking pages @xcite .
however , this may create a feedback cycle in which traffic grows super - linearly with popularity , leading to a situation ( sometimes called `` googlearchy '' ) in which a few popular sites dominate the web and lesser known sites are difficult to discover @xcite .
more importantly for the present work , simply accepting traffic data as a given does not further our understanding of user behavior .
we can also overcome the deficiencies of the random - surfer model by improving the model itself . this paper offers analysis of key features of observed behavior to support the development of improved agent - based models of web traffic @xcite
. the present study also relates to work in anomaly detection and anonymization software for the web .
the web tap project , for example , attempted to discover anomalous traffic requests using metrics such as request regularity and interrequest delay time , quantities which we discuss in the present work @xcite .
the success of systems that aim to preserve the anonymity of web users is known to be dependent on a variety of empirical properties of behavioral data , some of which we directly address here @xcite .
the click data we use in this study was gathered from a dedicated freebsd server located in the central routing facility of the bloomington campus of indiana university ( figure [ fig : architecture ] ) .
this system had a 1 gbps ethernet port that received a mirror of all outbound network traffic from one of the undergraduate dormitories .
this dormitory consists of four wings of five floors each and is home to just over a thousand undergraduates .
its population is split roughly evenly between men and women , and its location causes it to have a somewhat greater proportion of music and education students than other campus housing . to obtain information on individual http requests passing over this interface
, we first use a berkeley packet filter to capture only packets destined for tcp port 80 . while this eliminates from consideration all web traffic running on non - standard ports
, it does give us access to the great majority of it .
we make no attempt to capture or analyze encrypted ( https ) traffic using tcp port 443 .
once we have obtained a packet destined for port 80 , we use a regular expression search against the payload of the packet to determine whether it contains an http get request . if we do find an http get request in the packet
, we analyze the packet further to determine the virtual host contacted , the path requested , the referring url , and the advertised identity of the user agent .
we then write a record to our raw data files that contains the mac address of the client system , a timestamp , the virtual host , the path requested , the referring url , and a flag indicating whether the user agent matches a mainstream browser ( internet explorer , mozilla / firefox , safari , or opera ) .
we maintain record of the mac address only in order to distinguish the traffic of individual users .
we thus assume that most computers in the building have a single primary user , which is reasonable in light of the connectedness of the student population ( only a small number of public workstations are available in the dormitory ) . furthermore
, as long as the users do not replace the network interface in their computer , this information remains constant .
the aggregate traffic of the dormitory was sufficiently low so that our sniffing system could maintain a full rate of collection without dropping packets .
while our collection system offers a rare opportunity to capture the complete browsing activity of a large user population , we do recognize some potential disadvantages of our data source . because we do not perform tcp stream reassembly
, we can only analyze http requests that fit in a single 1,500 byte ethernet frame . while the vast majority of requests do so , some get - based web services generate extremely long urls . without stream
reassembly , we can not log the web server s response to each request : some requests will result in redirections or server errors , and we are unable to determine which ones . finally , a user can spoof the http referrer field ; we assume that few students do so , and those who do generate a small portion of the overall traffic .
the click data was collected over a period of about two months , from march 5 , 2008 through may 3 , 2008 .
this period included a week - long vacation during which no students were present in the building . during the full data collection period
, we logged nearly 408 million http requests from a total of 1,083 unique mac addresses .
not every http request from a client is indicative of an actual human being trying to fetch a web page ; in fact , such requests actually constitute a minority of all http requests . for this reason ,
we retain only those urls that are likely to be requests for actual web pages , as opposed to media files , style sheets , javascript code , images , and so forth .
this determination is based on the extension of the url requested , which is imprecise but functions well as a heuristic in the absence of access to the http _ content - type _ header in the server responses .
we also filtered out a small subset of users with negligible activity ; their traffic consisted largely of automated windows update requests and did not provide meaningful data about user activity .
finally , we also discovered the presence of a poorly - written anonymization service that was attempting to obscure traffic to a particular adult chat site by spoofing requests from hundreds of uninvolved clients .
these requests were also removed from the data set .
we found that some web clients issue duplicate http requests ( same referring url and same target url ) in nearly simultaneous bursts .
these bursts occur independently of the type of url being requested and are less than a single second wide .
we conjecture that they may involve checking for updated content , but we are unable to confirm this without access to the original http headers .
because this behavior is so rapid that it can not reflect deliberate activity of individual users , we also removed the duplicate requests from the data set .
privacy concerns and our agreement with the human subjects committee of our institution also obliged us to try to remove all identifying information from the referring and target urls .
one means of doing so is to strip off all identifiable query parameters from the urls . applying this anonymization procedure affects roughly one - third of the remaining requests .
the resulting data set ( summarized in table [ table : data ] ) is the basis for all of the description and analysis that follows .
.approximate dimensions of the filtered and anonymized data set . [ cols="<,^ " , ] even though we have defined sessions logically , they can still be considered from the perspective of time .
if we calculate the difference between the timestamp of the request that first created the session and the timestamp of the most recent request to add a leaf node , we obtain the duration of the logical session .
when we examine the distribution of the durations of the sessions of a user , we encounter the same situation as for the case of interclick times : power - law distributions @xmath0 for every user .
furthermore , when we consider the exponent of the best power - law fit of each user s data , we find the values are normally distributed with a mean value @xmath1 , as shown in figure [ fig : bracket_gamma ] .
no user has a well - defined mean duration for their logical sessions ; as also suggested by the statistics of interclick times , the presence of strong regularity in a user s session behavior would be anomalous . for the best power - law approximation to the distribution of logical session duration for each user .
the fit is a normal distribution with mean @xmath2 and standard deviation @xmath3 .
these low values of @xmath4 indicate unbounded variance and the lack of any central tendency in the duration of a logical session . ]
it is natural to speculate that we can get the best of both worlds by extending the definition of a logical session to include a timeout , as was done in previous work on referrer trees @xcite .
such a change is quite straightforward to implement : we simply modify the algorithm so that a request can not attach to an existing session unless the attachment point was itself added within the timeout .
this allows us to have one branch of the browsing tree time out while still allowing attachment on a more active branch . while the idea is reasonable
, we unfortunately find that the addition of such a timeout mechanism once again makes the statistics of the sessions strongly dependent on the particular timeout selected .
as shown in figure [ fig : logical_stat ] , the number of sessions per user , mean node count , mean depth , and ratio of nodes to tree depth are all dependent on the timeout . on the other hand , in contrast to sessions defined purely by timeout , this dependence becomes smaller as the timeout increases , suggesting that logical sessions with a timeout of around 15 minutes may be a reasonable compromise for modeling and further analysis .
in this paper we have built on the network - sniffing approach to gathering web traffic that we first explored in @xcite , extending it to track the behavior of individual users .
the resulting data set provides an unprecedented and accurate picture of human browsing behavior in a hypertext information space as manifested by over a thousand undergraduate students in their residences .
the data confirm previous findings about long - tailed distributions in site traffic and reveal that the popularity of sites is likewise unbounded and without any central tendency .
they also show that while many aspects of web traffic have been shown to obey power laws , these power - law distributions often represent the aggregate of distributions that are actually log - normal at the user level .
the lack of any regularity in interclick times for web users leads to the conclusion that sessions can not be meaningfully defined with a simple timeout , leading to our presentation of logical sessions and an algorithm for deriving them from a click stream .
these logical sessions illustrate further drawbacks of the random surfer model and can be modified to incorporate timeouts in a relatively robust way .
these findings have direct bearing on future work in modeling user behavior in hypertext navigation .
the stability of the proportion of empty - referrer requests across all users implies that although not every page is equally likely to be the cause of a jump , the overall chance of a jump occurring is constant in the long run .
the finding that the branching factor of the logical sessions is definitely greater than one means that plausible agent - based models for random walks must incorporate state , either through backtracking or branching @xcite .
our indications as to which distributions show central tendencies and which do not are of critical importance for anomaly detection and anonymization . to appear plausibly human ,
an agent must not stray too far from the expected rate of requests , proportion of empty - referrer requests , referrer - to - host ratio , and node count and tree depth values for logical sessions . because these are log - normal distributions
, agents can not deviate more than a multiplicative factor away from their means . at the same time , a clever agent must mimic the heavy - tailed distributions of the spacing between requests and duration of sessions ; too _ much _ regularity appears artificial .
although our method of collection does afford us with a large volume of data , it suffers from several disadvantages which we are working to overcome in future studies .
first , our use of the file extension ( if any ) in requested urls is a noisy indicator of whether a request truly represent a page fetch .
we are also unable to detect whether any request is actually satisfied or not ; many of the requests may actually result in server errors or redirects . both of these problems could be largely mitigated without much overhead by capturing the first packet of the server s response , which should indicate an http response code and a content type in the case of successful requests .
this data set is inspiring the development of an agent - based model that replaces the uniform distributions of pagerank with more realistic distributions and incorporates bookmarking behavior to capture the branching behavior observed in logical sessions @xcite .
the authors would like to thank the advanced network management laboratory at indiana university and dr
. jean camp of the iu school of informatics for support and infrastructure .
we also thank the network engineers of indiana university for their support in deploying and managing the data collection system .
special thanks are due to alessandro flammini for his insight and support during the analysis of this data .
this work was produced in part with support from the institute for information infrastructure protection research program .
the i3p is managed by dartmouth college and supported under award number 2003-tk - tx-0003 from the u.s .
dhs , science and technology directorate .
this material is based upon work supported by the national science foundation under award number 0705676 .
this work was supported in part by a gift from google .
opinions , findings , conclusions , recommendations or points of view in this document are those of the authors and do not necessarily represent the official position of the u.s .
department of homeland security , science and technology directorate , i3p , national science foundation , indiana university , google , or dartmouth college .
m. bouklit and f. mathieu .
backrank : an alternative for pagerank ? in _
www 05 : special interest tracks and posters of the 14th international conference on world wide web _ , pages 11221123 , new york , ny , usa , 2005 .
acm .
s. pandey , s. roy , c. olston , j. cho , and s. chakrabarti .
shuffling a stacked deck : the case for partially randomized ranking of search engine results . in k.
bhm , c. s. jensen , l. m. haas , m. l. kersten , p .- .
larson , and b. c. ooi , editors , _ proc .
31st international conference on very large databases ( vldb ) _ , pages 781792 , 2005 .
f. qiu , z. liu , and j. cho .
analysis of user web traffic with a focus on search activities . in a.
doan , f. neven , r. mccann , and g. j. bex , editors , _ proc .
8th international workshop on the web and databases ( webdb ) _
, pages 103108 , 2005 . | we examine the properties of all http requests generated by a thousand undergraduates over a span of two months .
preserving user identity in the data set allows us to discover novel properties of web traffic that directly affect models of hypertext navigation .
we find that the popularity of web sites the number of users who contribute to their traffic lacks any intrinsic mean and may be unbounded .
further , many aspects of the browsing behavior of individual users can be approximated by log - normal distributions even though their aggregate behavior is scale - free .
finally , we show that users click streams can not be cleanly segmented into sessions using timeouts , affecting any attempt to model hypertext navigation using statistics of individual sessions .
we propose a strictly logical definition of sessions based on browsing activity as revealed by referrer urls ; a user may have several active sessions in their click stream at any one time .
we demonstrate that applying a timeout to these logical sessions affects their statistics to a lesser extent than a purely timeout - based mechanism .
[ http ] [ information networks ] [ navigation , user issues ] |
we derive information on the physical properties of the solar atmosphere by interpreting the polarization profiles of spectral lines .
extracting this information is not easy , because the observed profiles depend on the atmospheric parameters in a highly non - linear manner through the absorption matrix and the source function vector . to solve the problem ,
least - squares inversion techniques ( its ) based on analytical or numerical solutions of the radiative transfer equation were developed in the past .
these methods compare the observed stokes profiles with synthetic profiles emerging from an initial guess model atmosphere .
the misfit is used to modify the atmospheric parameters until the synthetic profiles match the observed ones .
this yields a model atmosphere capable of explaining the measurements , within the assumptions and limitations of the model .
the first its were proposed by @xcite and @xcite .
many other codes have been developed since then ( cf .
table 1 in * ? ? ?
* ; * ? ? ?
today we have an it for almost any application we may be interested in : lte or non - lte line formation , one - component or multi - component model atmospheres , photospheric or chromospheric lines , etc .
these codes have been used intensively to study the magnetism of the solar atmosphere , and now they are essential tools for the analysis of spectro - polarimetric measurements .
this paper concentrates on recent achievements and future challenges of its .
additional information on its can be found in the reviews by del toro iniesta & ruiz cobo ( 1996 ) , socas - navarro ( 2001 ) , and del toro iniesta ( 2003 ) .
there have been no significant improvements of classical least - squares its in recent years : no new algorithms have appeared , and existing codes have not been optimized for speed .
however , the experience accumulated has been used to develop new codes for specific purposes .
the complexity of both atmospheric and line - formation models has also increased .
for example , a few years ago the most sophisticated inversions of stokes profiles from sunspots were based on one - component models with gradients of the physical parameters ( westendorp plaza et al .
2001 ) , while two - component inversions are now performed on a routine basis @xcite .
even more complex models are necessary to explain the net circular polarization ( ncp ) of spectral lines emerging from sunspot penumbrae . @xcite used micro - structured magnetic atmospheres , assuming that the penumbra is formed by optically thin magnetic fibrils .
adopted an uncombed penumbral model with two different components representing inclined flux tubes embedded in a more vertical ambient field . both models
successfully reproduce the anomalous stokes @xmath0 profiles observed near the neutral line , for visible and infrared ( ir ) lines considered separately .
the synthetic ncps , however , are a bit smaller than the observed ones , implying that there is still room for improvement . an important advance in the field has been the development of alternative methods for real - time analyses of large data sets .
other significant achievements have come from the application of classical its to state - of - the - art observations .
these issues are examined in more detail in the next subsections .
conceptually , the simplest inversion is one that uses a look - up table .
the idea is to create a database of synthetic stokes profiles from a large number of known model atmospheres , and look for the profile in the database which is closest to the observed spectrum .
the corresponding model is adopted as representative of the physical conditions of the atmosphere from which the profile emerged . despite its simplicity , this method had seldom been put into practice until @xcite drew attention to principal component analysis ( pca ) as a means to accelerate the search in the look - up table . by virtue of pca , the stokes profiles can be expressed in terms of a few coefficients only .
the comparison between observed and synthetic profiles is then performed very quickly , because the calculation does not involve the many wavelength points describing the full line profiles .
pca lies at the heart of several codes developed in the last years .
the database is the most critical component of any it based on look - up tables , as its size increases dramatically with the number of free parameters . to keep this number to a minimum , only milne - eddington ( me ) atmospheres
have been used for pca inversions of photospheric lines ( e.g. , socas - navarro , lpez ariste , & lites 2001 ) . even under me
conditions , the parameter space can not be sampled very densely .
the discrete nature of the database introduces numerical errors , and so pca analyses are less accurate than least - squares me inversions ( e.g. , * ? ? ?
* ; * ? ? ?
however , the method gives an idea of the quality of the fit in terms of the so - called pca distance .
when this distance is large , the observed profile can not be associated with any profile in the database , making it possible to identify pixels that deserve closer attention .
a nice feature of pca inversions is that the search algorithm is independent of the database .
the synthesis of stokes profiles may be very time - consuming , but the inversion will always be fast .
this opens the door to the analysis of lines for which atomic polarization effects are important .
have developed a pca inversion code to exploit the diagnostic potential of the hanle effect in the d@xmath1 line at @xmath2 nm .
the database is created using a line formation code which solves the statistical equilibrium of a he atom with 5 terms , in the presence of magnetic fields .
coherences between fine - structure levels within each atomic term are accounted for to treat the zeeman and hanle regimes , including level crossing ( incomplete paschen - back effect ) .
this code has been applied to prominences @xcite and spicules @xcite .
the speed of pca inversions makes it possible to handle large amounts of data in real time . since 2004
, pca is used at the french - italian themis telescope to derive vector magnetic fields from mtr measurements of the @xmath3 nm lines @xcite .
full maps are inverted in about 10min , which is more or less the time needed to take the observations . at the telescope
, pca is very useful for quick - look analyses , allowing one to select interesting targets or to continue the observation of interesting regions . while real - time analyses are appealing , it is important to keep in mind their limitations . in most cases ,
a proper interpretation of the observations will require more sophisticated its , which however can use the results of pca methods as initial guesses .
another promising technique explored in recent years is stokes inversion based on artificial neural networks ( anns ) .
the idea was introduced by carroll & staude ( 2001 ) and developed by @xcite . essentially , an ann is an interpolation algorithm .
one starts by setting up the structure of the network , i.e. , the number of layers and the number of neurons in each layer .
the input layer receives the observations ( the stokes profiles or a suitable combination thereof ) and the last layer outputs the unknown atmospheric parameters .
the ann must be trained before applying it to real data . to this end , stokes profiles synthesized using different me atmospheres are presented to the ann in order to find the synaptic weights and biases of the neurons that return the model parameters used to compute the profiles .
the training process is very slow but , once accomplished , the ann will invert a full map in a matter of seconds .
indeed , anns are the fastest its available nowadays . for the moment
, however , they have not been used in any scientific application .
several strategies have been explored to optimize the performance of anns .
it seems that the best choice is to invert one parameter at a time with a dedicated ann .
finding all atmospheric parameters with a single ann is possible , but requires a larger number of neurons and the training process becomes very complicated .
@xcite has shown that the mean field strengths inferred with the help of anns are reasonably accurate from a statistical point of view .
on average they resemble those provided by me inversions .
however , the errors can be very large for individual pixels , as evidenced by the large r.m.s .
uncertainties ( fig . 8 in that paper indicates an uncertainty of 0.3 kg for fields of @xmath4 kg , i.e. , a relative error of 25% ; the error is even larger for weaker fields ) .
this means that anns may be appropriate for quick - look analyses and other applications where high precision is not required . for detailed studies of physical processes ,
current anns seem not to be accurate enough .
a limitation of both pca - based inversions and anns is the use of me atmospheres , which precludes the determination of gradients of physical parameters from the observed profiles .
it remains to be seen whether these methods can be modified in order to reliably recover gradients along the line of sight ( los ) .
as discussed in sect .
3.3 , gradients appear to be essential for the analysis of observations at very high spatial resolution .
major breakthroughs have come from the application of its to state - of - the - art observations , including spectro - polarimetry in the near ir , simultaneous observations of visible and ir lines , spectro - polarimetry of molecular lines , and observations at very high spatial resolution .
the development of polarimeters for the ir , most notably the tenerife infrared polarimeter ( tip ; * ? ? ?
* ; * ? ? ?
* ) , has opened a new window with lines that offer excellent magnetic sensitivity and chromospheric coverage .
one example is the triplet at @xmath5 nm , which has become an essential tool to investigate the upper chromosphere .
the formation of the triplet is complex and not really well understood .
however , since the triplet lines are nearly optically thin , me atmospheres provide a good description of their shapes .
an inversion code specifically designed for @xmath5 nm , helix , was presented by @xcite .
it is based on the unno - rachkovsky solution of the radiative transfer equation and includes an empirical treatment of the hanle effect . outside active regions ,
the linear polarization profiles of @xmath6 nm show the signatures of the hanle effect , which needs to be taken into account for correct retrievals of vector magnetic fields @xcite .
helix implements the pikaia genetic algorithm , rather than the more common marquardt algorithm employed by other least - squares its .
using the magnetic information obtained with helix , @xcite were able to trace individual coronal loops in an emerging flux region .
they found upflows at the apex of the loops and downdrafts near the footpoints , which is what one expects for magnetic loops rising from deeper layers .
other applications of the code have been presented by @xcite , @xcite , and @xcite .
routine observations of the ir triplet of at @xmath7 nm are now possible with the mtr mode of themis and the spectro - polarimeter for infrared and optical regions ( spinor ; * ? ? ?
* ; * ? ? ?
* ) mounted on the dunn solar telescope ( dst ) of the national solar observatory at sacramento peak .
the ca ir triplet lines are excellent diagnostics of the chromosphere , with the advantage that their interpretation is much simpler than that of other chromospheric lines such as h , k , and h@xmath8 .
however , they still require non - lte computations .
non - lte inversions of the ca ir triplet lines observed with spinor have been presented by @xcite and @xcite .
although these analyses are very demanding in terms of computational resources , they hold great promise for quantitative diagnostics of the thermal and magnetic structure of the solar chromosphere
. profiles of the pair of lines at 630 nm _
( left ) _ and the lines at 1565 nm _
( right ) _ observed simultaneously with polis and tip in the limb - side penumbra of ar10425 , near the neutral line .
dots indicate the observations , solid the best - fit profiles resulting from a simultaneous inversion of the data . _
bottom : _ uncombed penumbral model inferred from the inversion . from left to right : temperature , los velocity , field strength , and field inclination in the background ( solid ) and the flux - tube ( dashed ) components ( from beck , bellot rubio , & schlichenmaier , in preparation ) .
[ fig1],title="fig:",height=130 ] profiles of the pair of lines at 630 nm _ ( left ) _ and the lines at 1565 nm _ ( right ) _ observed simultaneously with polis and tip in the limb - side penumbra of ar10425 , near the neutral line .
dots indicate the observations , solid the best - fit profiles resulting from a simultaneous inversion of the data . _
bottom : _ uncombed penumbral model inferred from the inversion . from left to right : temperature , los velocity , field strength , and field inclination in the background ( solid ) and the flux - tube ( dashed ) components ( from beck , bellot rubio , & schlichenmaier , in preparation ) .
[ fig1],title="fig:",height=102 ] profiles of the pair of lines at 630 nm _
( left ) _ and the lines at 1565 nm _ ( right ) _ observed simultaneously with polis and tip in the limb - side penumbra of ar10425 , near the neutral line .
dots indicate the observations , solid the best - fit profiles resulting from a simultaneous inversion of the data . _
bottom : _ uncombed penumbral model inferred from the inversion . from left to right : temperature , los velocity , field strength , and field inclination in the background ( solid ) and the flux - tube ( dashed ) components ( from beck , bellot rubio , & schlichenmaier , in preparation ) .
[ fig1],title="fig:",height=102 ] profiles of the pair of lines at 630 nm _ ( left ) _ and the lines at 1565 nm _ ( right ) _ observed simultaneously with polis and tip in the limb - side penumbra of ar10425 , near the neutral line .
dots indicate the observations , solid the best - fit profiles resulting from a simultaneous inversion of the data . _
bottom : _ uncombed penumbral model inferred from the inversion . from left to right : temperature , los velocity , field strength , and field inclination in the background ( solid ) and the flux - tube ( dashed ) components ( from beck , bellot rubio , & schlichenmaier , in preparation ) .
[ fig1],title="fig:",height=130 ] profiles of the pair of lines at 630 nm _ ( left ) _ and the lines at 1565 nm _ ( right )
_ observed simultaneously with polis and tip in the limb - side penumbra of ar10425 , near the neutral line .
dots indicate the observations , solid the best - fit profiles resulting from a simultaneous inversion of the data . _
bottom : _ uncombed penumbral model inferred from the inversion . from left to right : temperature , los velocity , field strength , and field inclination in the background ( solid ) and the flux - tube ( dashed ) components ( from beck , bellot rubio , & schlichenmaier , in preparation ) .
[ fig1],title="fig:",height=102 ] profiles of the pair of lines at 630 nm _ ( left ) _ and the lines at 1565 nm _ ( right ) _ observed simultaneously with polis and tip in the limb - side penumbra of ar10425 , near the neutral line .
dots indicate the observations , solid the best - fit profiles resulting from a simultaneous inversion of the data .
_ bottom : _ uncombed penumbral model inferred from the inversion . from left to right : temperature , los velocity , field strength , and field inclination in the background ( solid ) and the flux - tube ( dashed ) components ( from beck , bellot rubio , & schlichenmaier , in preparation ) .
[ fig1],title="fig:",height=102 ] simultaneous observations of visible and ir lines improve the accuracy of inversion results due to their different sensitivities to the various atmospheric parameters @xcite .
several instruments are capable of such observations , including tip and polis ( polarimetric littrow spectrograph ; * ? ?
* ; * ? ? ?
* ) at the german vtt in tenerife , and spinor at the dst .
figure 1 shows examples of stokes @xmath0 profiles of the lines at @xmath9 nm , @xmath10 nm , @xmath11 nm , and @xmath12 nm observed in the penumbra of ar10425 near the neutral line . these profiles were taken strictly simultaneously with tip and polis on august 9 , 2003 .
both the visible and ir lines exhibit large asymmetries and even pathological shapes , suggesting the presence of two magnetic components in the resolution element .
we have carried out an analysis of these profiles in terms of an uncombed penumbral model using the code described by @xcite .
the best - fit profiles from the simultaneous inversion are represented by the solid lines .
the quality of the fits is certainly remarkable , but it should be stressed that even better fits are achieved from the inversion of only the visible or the ir lines . indeed , fitting both sets of lines simultaneously is much more difficult , because a model atmosphere appropriate for the visible lines may not be appropriate for the ir lines , and vice versa .
measurements in two or more spectral regions thus constrain the range of acceptable solutions .
the lower panels of fig . 1
show the uncombed model resulting from the inversion ( beck , bellot rubio , & schlichenmaier , in preparation ) .
the dashed lines represent a penumbral flux tube with larger los velocities than the background atmosphere , indicated by the solid lines .
the field is weaker and more horizontal in the tube .
these inversions confirm that the uncombed model is able to explain the observed shapes of both visible and ir lines . at the same time , they allow us to determine the position and width of the penumbral tubes , which is not easy from visible or ir lines alone . other examples of the analysis of simultaneous measurements in the visible and ir are given by @xcite and @xcite .
the new polarimeters have also extended our capabilities to observe molecular lines , providing increased thermal sensitivity .
molecular lines are mostly seen in sunspot umbrae , because higher temperatures dissociate the parent molecules .
usually , they show smaller polarization signals than atomic lines .
two codes capable of inverting molecular lines are spinor @xcite and the one developed by @xcite .
spinor has been used to analyze the oh lines at @xmath12 nm and @xmath13 nm , improving the determination of umbral temperatures @xcite .
the second code has been employed to invert the cn lines at @xmath14 nm @xcite .
these lines are very interesting because in the umbra they show large linear polarization signals but very small stokes @xmath0 signals , just the opposite behavior of atomic lines . finally
, major progress has come from the application of its to high spatial resolution spectroscopic and spectro - polarimetric measurements .
the main advantage of high spatial resolution is that the results are less dependent on filling factor issues .
an example of the inversion of high spatial resolution stokes profiles is the work of @xcite , who derived the magnetic and thermal properties of umbral dots from observations taken with the la palma stokes polarimeter @xcite at a resolution of about 07 .
another promising type of measurements are those provided by fabry - prot interferometers such as the interferometric bidimensional spectrometer ( ibis ; * ? ? ?
* ; * ? ? ?
* ) and the telecentric solar spectrometer ( tesos ; * ? ? ?
* ; * ? ? ?
* ) , which has recently been equipped with the kis / iaa visible imaging polarimeter ( vip ; * ? ? ?
* ; * ? ? ?
combined with adaptive optics systems , these instruments perform 2d vector spectro - polarimetry at high spatial , spectral , and temporal resolutions , which is necessary to investigate fast processes in large fields of view .
an example of the inversion of 2d spectroscopic measurements with tesos is the derivation of the thermal and kinematic properties of a sunspot penumbra at different heights in the atmosphere ( @xcite ; see also @xcite ) .
the angular resolution of these observations is 05 . in the future ,
significant progress may come from the routine inversion of lines showing hyperfine structure , such as @xmath15 nm and @xmath16 nm ( lpez ariste , tomczyk , & casini 2002 ) .
these lines exhibit sign reversals in the core of stokes @xmath0 and multiple peaks in stokes @xmath17 and @xmath18 for weak fields .
interestingly , the shape of the anomalies depends on the magnetic field strength , rather than on the magnetic flux . such an unusual behavior can be used to investigate the magnetism of the quiet sun .
in fact , from the shape of the observed profiles it would be possible to determine directly the strength of the magnetic field , even in the weak field regime . to exploit the diagnostic potential of these lines , however , it is necessary to implement the appropriate zeeman patterns in existing its and to lower the noise level of current observations , which is barely enough to detect the subtle signatures induced by hyperfine structure .
its have proven to be essential tools to characterize the properties of the solar atmosphere .
their application to high precision spectro - polarimetric measurements , however , has started to raise concerns on the limitations of some spectral lines .
this is an important problem that deserves further investigation .
other challenges facing its in the near future include the implementation of more realistic model atmospheres and the development of strategies for the analysis of the large amounts of data to be delivered by upcoming instruments .
@xcite have shown that it is possible to fit a given set of stokes profiles of the pair of lines at @xmath3 nm with very different field strengths by slightly changing the temperature stratification and the microturbulent velocity , for typical conditions of quiet sun internetwork regions .
the reason is the different formation height of the two lines .
this quite unexpected result suggests that one can not determine reliable internetwork field strengths from @xmath9 nm and @xmath10 nm without a prior knowledge of the actual temperature stratification .
other lines such as @xmath19 nm and @xmath20 nm could provide the necessary information . we have found a similar problem with the @xmath3 nm lines even in the umbra , where the strong field regime applies . more specifically , we have detected a cross - talk problem between the stray light coefficient , the temperature , and the magnetic field strength and inclination ( see * ? ? ?
* ; * ? ? ?
the results of one - component inversions of umbral profiles with the stray light contamination as a free parameter do differ from those in which the stray light factor is fixed to the value inferred from a simultaneous inversion of visible and ir lines . with a mere difference of 7% in the stray light coefficient
, the temperatures at @xmath21 from the two inversions may differ by up to @xmath22k , and the field strength by about @xmath23 g .
the fits are equally good in both cases , so it is not possible to decide which inversion is better .
it appears that the relatively small zeeman splitting of the @xmath3 nm lines does not allow to clearly distinguish between larger stray light factors and weaker fields , which produces cross - talk among the various atmospheric parameters .
given these concerns , more detailed studies of the limitations of visible and ir lines seem warranted , for a better understanding of the results obtained from them . to minimize the risk of cross - talk problems in the inversion , it is desirable to use simultaneous measurements in different spectral ranges
. this will require modifications of current inversion codes to account for different stray light levels and different instrumental profiles in the different spectral ranges .
the new observational capabilities , in particular the availability of simultaneous observations of visible and ir lines , offer us a unique opportunity to increase the realism of the atmospheric models implemented in existing its .
the need for better models is indicated by the small ( sometimes systematic ) residuals observed in inversions of profiles emerging from complex magnetic structures .
as an example , consider the uncombed penumbral model .
right now we use two different lines of sight to represent the background and flux - tube atmospheres ( cf .
@xcite ) , but this is a very simplistic approximation .
the ambient field lines have to wrap around the flux tube , hence the properties of the background can not be the same far from the tube and close to it .
this may have important consequences for the generation of asymmetrical stokes profiles . in much the same way ,
the flux tube is not always at the same height within the resolution element , because the magnetic field is not exactly horizontal .
therefore , different rays will find the tube at different heights .
finally , lines of sight crossing the center of the tube sense the properties of the tube over a larger optical - depth range than lines of sight crossing the tube at a distance from its axis .
neither of these effects are modeled by current inversion codes . probably , the subtle differences between observed and best - fit profiles ( cf .
[ fig1 ] ) would disappear with a more complex treatment of the magnetic topology of sunspot penumbrae .
stokes polarimetry at the diffraction limit is needed to study the physical processes occurring in the solar atmosphere at their intrinsic spatial scales .
we are pushing our technological capabilities to the limit by building grating spectro - polarimeters and filter magnetographs for diffraction - limited observations . on the ground ,
examples of already operational or upcoming state - of - the - art instruments include tip , polis , dlsp , spinor , ibis and tesos+vip . among space - borne instruments we have the spectro - polarimeter and filter polarimeter onboard solar - b , imax onboard sunrise , hmi onboard sdo , and vim onboard solar orbiter .
these instruments will deliver data of unprecedented quality in terms of spatial and spectral resolution .
we hope to further our understanding of the solar magnetism with them . however , the success of this endeavor will critically depend on our ability to extract in an appropriate way the information contained in the observations .
we do not only want to investigate the morphology and temporal evolution of the various solar structures from diffraction - limited images , but also to derive their magnetic and kinematic properties accurately using polarization measurements .
reliable determinations of vector magnetic fields call for least - squares inversions .
the problem is the enormous data flows expected : classical least - squares its are considered to be too slow for real - time analyses of the observations .
this is the reason why it is taken for granted that the most sophisticated inversions of the data will be based on me models .
the question naturally arises as to whether or not me atmospheres are appropriate for the interpretation of stokes measurements at very high spatial resolution . and @xmath0 profiles of 525.06 nm emerging from an intergranular lane as computed from mhd simulations ( solid ) . dotted and
dashed lines represent the best - fit profiles from a me inversion and a sir inversion with gradients , respectively . _ middle and right : _ stratifications of atmospheric parameters used for the spectral synthesis ( solid ) .
the results of the me inversion and the sir inversion are given by the dotted and dashed lines , respectively .
[ fig2],title="fig : " ] and @xmath0 profiles of 525.06 nm emerging from an intergranular lane as computed from mhd simulations ( solid ) . dotted and
dashed lines represent the best - fit profiles from a me inversion and a sir inversion with gradients , respectively . _ middle and right : _ stratifications of atmospheric parameters used for the spectral synthesis ( solid ) .
the results of the me inversion and the sir inversion are given by the dotted and dashed lines , respectively .
[ fig2],title="fig : " ] and @xmath0 profiles of 525.06 nm emerging from an intergranular lane as computed from mhd simulations ( solid ) . dotted and
dashed lines represent the best - fit profiles from a me inversion and a sir inversion with gradients , respectively . _ middle and right : _ stratifications of atmospheric parameters used for the spectral synthesis ( solid ) .
the results of the me inversion and the sir inversion are given by the dotted and dashed lines , respectively .
[ fig2],title="fig : " ] and @xmath0 profiles of 525.06 nm emerging from an intergranular lane as computed from mhd simulations ( solid ) .
dotted and dashed lines represent the best - fit profiles from a me inversion and a sir inversion with gradients , respectively . _ middle and right : _ stratifications of atmospheric parameters used for the spectral synthesis ( solid ) .
the results of the me inversion and the sir inversion are given by the dotted and dashed lines , respectively .
[ fig2],title="fig : " ] and @xmath0 profiles of 525.06 nm emerging from an intergranular lane as computed from mhd simulations ( solid ) . dotted and
dashed lines represent the best - fit profiles from a me inversion and a sir inversion with gradients , respectively . _ middle and right : _ stratifications of atmospheric parameters used for the spectral synthesis ( solid ) .
the results of the me inversion and the sir inversion are given by the dotted and dashed lines , respectively .
[ fig2],title="fig : " ] and @xmath0 profiles of 525.06 nm emerging from an intergranular lane as computed from mhd simulations ( solid ) . dotted and
dashed lines represent the best - fit profiles from a me inversion and a sir inversion with gradients , respectively . _
middle and right : _ stratifications of atmospheric parameters used for the spectral synthesis ( solid ) .
the results of the me inversion and the sir inversion are given by the dotted and dashed lines , respectively .
[ fig2],title="fig : " ] to shed some light on this issue , i have used the mhd simulations of @xcite to synthesize the stokes profiles of the imax line ( @xmath24 nm ) emerging from a typical magnetic concentration in an intergranular lane .
the atmospheric parameters needed for the calculation have been taken from a simulation run with average magnetic flux density of @xmath25 g .
figure [ fig2 ] displays the atmospheric stratifications and the corresponding stokes profiles at 01 resolution ( solid lines ) .
the dotted lines show the results of a me inversion of the synthetic profiles . as can be seen , the fits to stokes @xmath26 and @xmath0 are not very successful , due to the extreme asymmetries of the profiles .
the atmospheric parameters inferred from the me inversion are some kind of average of the real stratifications , but the significance and usefulness of these average values are questionable when the parameters feature such strong variations along the los .
an analysis of the same profiles with sir @xcite allowing for vertical gradients of field strength and velocity yields much better fits to stokes @xmath26 and @xmath0 ( dashed lines ) .
although the fits can still be improved , the important point is that this simple sir inversion is able to recover the gradients of field strength and velocity with fewer free parameters than the me inversion ( 8 as opposed to 9 ) .
this additional information could be essential to understand many physical processes , so it is important to have it .
present - day computing resources are sufficient to determine gradients from the high spatial resolution observations delivered by _ grating instruments _ such as polis and the solar - b spectro - polarimeter ( solar - b / sp ; lites , elmore , & streander 2001 ) .
a sir inversion of the four stokes profiles of 2 spectral lines ( 10 free parameters , 135 wavelength points , model atmosphere discretized in 41 grid points ) takes @xmath27s on a dual xeon workstation running at @xmath28ghz .
optimizing the code , a cadence of @xmath29s may easily be reached .
the real - time analysis of a full polis slit ( 450 pixels in @xmath30s ) would then require 20 such workstations .
the analysis of solar - b / sp data ( 1000 pixels every @xmath30s ) would require 50 workstations . in both cases ,
the total cost would be a minor fraction of the cost of the instruments themselves .
the situation is rather different for vector magnetographs .
these instruments measure only a few wavelength points , i.e. , line profiles are not available .
probably , the most we can do with this kind of data is a full least - squares me inversion .
an additional complication is that the data rates will be huge , much larger than those expected from grating instruments .
for example , hmi will observe about @xmath31 pixels every 80@xmath32s . to cope with such data flows , pca methods and anns
are being proposed as the only option to invert the observations in real time .
we have already mentioned that on average the results of these methods coincide with those from me inversions .
however , since large errors occur for many individual pixels , it is clear that me inversions would be preferable over pca or ann analyses .
but , how to perform me inversions of vector magnetograph data at the required speed ?
the solution could be hardware inversion on field programmable gate arrays ( fpgas ) , which is about 10@xmath33 times faster than software inversion depending on the frequency of the processor and the implementation of the algorithm . at the iaa
, we are studying the feasibility of such an electronic inversion for the analysis of vim data @xcite .
the first working prototype is expected to be ready by the end of 2007 .
inversion techniques ( its ) have become essential tools to investigate the magnetism of the solar atmosphere .
nowadays , they represent the best option to extract the information contained in high precision polarimetric measurements .
the reliability and robustness of least - squares stokes inversions have been confirmed many times with the help of numerical tests .
part of the community , however , is still concerned with uniqueness issues .
these concerns will hopefully disappear with the implementation of more realistic model atmospheres . during the last years
, major progress in the field has resulted from the application of its to state - of - the - art observations .
the advent of spectro - polarimeters for the near ir has represented a breakthrough , allowing the observation of atomic and molecular lines that provide increased magnetic and thermal sensitivity , and extended chromospheric coverage .
the potential of simultaneous observations of visible and ir lines for precise diagnostics of solar magnetic fields has just started to be exploited .
visible and ir lines constrain the range of acceptable solutions , which is especially useful for the investigation of complex structures with different magnetic components and/or discontinuities along the los .
finally , we have begun to invert spectro - polarimetric observations at very high spatial resolution , with the aim of reaching the diffraction limit of current solar telescopes ( 0102 ) .
high spatial resolution allows to separate different magnetic components that might coexist side by side , thus facilitating the determination of their properties .
the application of its to these observations is casting doubts on the capabilities of certain lines for investigating particular aspects of solar magnetism
. a detailed study of the limitations of spectral lines , in particular the often used pair at @xmath3 nm , seems necessary to clarify their range of usability .
an obvious cure for any problem that might affect the observables is to invert visible and ir lines simultaneously .
this will require modifications of current its to account for different instrumental effects in the different spectral ranges .
perhaps the most important challenge facing its in the next years is the analysis of the enormous data sets expected from upcoming space - borne polarimeters .
so far , the efforts have concentrated on the development of fast pca - methods and anns for real - time inversions of the data .
however , the unprecedented quality of these observations in terms of spectral and spatial resolution makes it necessary to explore the feasibility of more complex inversions capable of determining gradients of field strength and velocity along the los .
tests with numerical simulations demonstrate the importance of gradients to reproduce the very large asymmetries of the stokes profiles expected at a resolution of 0102 .
current computational resources allow us to determine gradients from full line profiles observed with grating spectro - polarimeters such as the one onboard solar - b , at a very reasonable cost .
gradients might also be recovered from high - resolution filtergraph observations if sufficient wavelength points are available .
interestingly , real - time me inversions of data with limited wavelength sampling seem possible using fpgas .
the feasibility of such electronic me inversions needs to be assessed . at the same time
, it is important to continue the development of pca and ann methods to provide the more complex me inversions with good initial guesses .
bellot rubio , l. r. , schlichenmaier , r. , & tritschler , a. 2006 , , 453 , 1117 bellot rubio , l. r. , tritschler , a. , kentischer , t. , beck , c. , & del toro iniesta , j. c.2006 , 26th meeting of the iau , 16 - 17 august , 2006 , prague , jd03 , # 58 martnez pillet , v. , collados , m. , snchez almeida , j. , et al . in asp conf .
183 , high resolution solar physics : theory , observations , and techniques , ed .
t. r. rimmele , k. s. balasubramaniam & r. r. radick ( san francisco : asp ) , 264 orozco surez , d. , lagg , a. , & solanki , s. k. 2005 , in esa sp-596 , proc.intl.sci.conf . chromospheric and coronal magnetic fields , ed .
d. e. innes , a. lagg , s. k. solanki & d. danesy ( noordwijk : esa ) , 59 | inversion techniques ( its ) allow us to infer the magnetic , dynamic , and thermal properties of the solar atmosphere from polarization line profiles . in recent years , major progress has come from the application of its to state - of - the - art observations .
this paper summarizes the main results achieved both in the photosphere and in the chromosphere .
it also discusses the challenges facing its in the near future . understanding the limitations of spectral lines , implementing more complex atmospheric models , and devising efficient strategies of data analysis for upcoming ground - based and space - borne instruments , are among the most important issues that need to be addressed .
it is argued that proper interpretations of diffraction - limited stokes profiles will not be possible without accounting for gradients of the atmospheric parameters along the line of sight .
the feasibility of determining gradients in real time from space - borne observations is examined . |
there is currently great interest in experiments which obtain useful information about a quantum system or state in single runs of an experiment .
very recently it has been possible to distinguish the quantum mechanical and classical models of the interaction of an atom with a mode of a high finesse optical cavity as the result of continuous monitoring of the output light while a single atom passes through the cavity @xcite . as a result of this continuous monitoring the quantum mechanical backaction of the measurement process may be expected to have a significant effect on the evolution of individual runs of the experiment .
moreover , it may be possible in the future to modify the evolution of the system through feedback based on this continuous measurement @xcite .
current experimental technology , such as that described in @xcite , is reaching the point at which determining the state of the system and observing the effects of backaction and feedback in a single run of an experiment is a real possibility .
with this situation in mind , we discuss the identification of the state of the system following a period of continuous observation and the extent to which this state can be tracked , taking into account factors such as imperfect initial knowledge of the state and imperfect detection efficiency .
we focus on a model of continuous position measurement of a mechanical oscillator which is relevant to the experiment of mabuchi _ et al . _ but which also has relevance to other instances of interferometric position monitoring such as gravitational wave detection .
the problem of describing quantum systems undergoing continuous measurement has attracted much theoretical interest in recent years .
as discussed by wiseman @xcite these theories admit a variety of interpretations ; as tools for efficient stochastic calculation of ensemble averages in lieu of solving master equations @xcite , as equations describing the evolution of systems conditioned on measurements @xcite and as a description of the evolution of a system coupled to an environment , in which collapse of the wavefunction is supposed to be associated with the coupling to the environment @xcite . here , we take the second viewpoint namely that the conditioned state represents the observer s best description of the system state given the results of the continuous measurement process . adopting the first or third viewpoints one
is led to describe the system by a pure state vector throughout the evolution although the reasons for doing so are somewhat different in each case . by contrast , a description of one s conditioned state of knowledge necessarily requires mixed states in order to account for incomplete knowledge of the system . from this viewpoint
the fundamental equation for the conditioned evolution is the stochastic master equation ( sme ) @xcite .
this is able to account for the effects of mixed initial states , imperfect detection efficiencies and the existence of unmeasured couplings to the environment .
however , to date , relatively little work has attempted to address the evolution of the conditioned state in any of these situations @xcite . in this paper
we consider a system which is simple enough that almost all the work can be done analytically and which admits a treatment of all of these imperfections .
this helps in developing intuition about the role of sme s and their possible relevance to experiments .
a projective measurement has the property that if the result of the measurement is known , the state after the measurement is pure , and depends only on the measurement result .
it would be hoped that in a continuous measurement there would be some finite interval of time after which the measurement has effectively given rise to a projection , so that the system is placed in a particular state which depends only on the sequence of measurement results and which can be calculated without knowledge of the initial state .
if the resulting state is pure then a stochastic schrdinger equation ( sse ) would be a perfectly adequate tool for describing the subsequent system evolution . in this paper
we investigate the conditions which lead to such an effective collapse and over what timescale it takes place .
this is made possible by considering density matrices and the sme rather tha wave vectors and the sse . in a real experiment there
will also be uncontrolled , unmeasured couplings of the system to the environment and in this case the effects of the measurement will compete not only with the coherent internal dynamics of the system but also with the randomizing effects of the coupling to the bath .
this may lead to mixed conditioned states even after long periods of continuous measurement and limit the observer s ability to make inferences about the system state .
understanding the process by which the conditioned state may collapse onto a pure state and the effects of noise as described by the sme allows us to define conditions under which continuous measurements in real experiments are approximations to ideal measurements .
this paper is organized as follows . in sec .
[ ssme ] we establish our simplified model of the continuous position measurement of an oscillator and solve the sme for gaussian initial states .
we find the time over which the second - order moments approach their steady - state values and calculate the entropy of the conditioned state as it becomes pure . in sec .
[ class ] we discuss the classical problem of state identification for the noisy measurement of the the position of an oscillator and derive a kind of uncertainty principle relating the observation and process noises if the classical model is to reproduce the sme . in sec .
[ mrd ] we show that the time - scale over which the second - order moments of the conditioned state reach their steady state is the same as that over which the conditioned state is completely determined by the measurement record .
section [ heat ] discusses the effect of heating , noise and detection inefficiency on these conclusions . finally in sec .
[ discuss ] we summarize and make some comments about future extensions of this work .
in this paper we shall consider the abstract model of continuous position measurement discussed by caves and milburn @xcite with a harmonically bound rather than a free measured particle .
projective position measurements are imagined to be made on a sequence of meters coupled briefly to the system and the limit of very frequent meter interactions and very broad initial meter position distributions is taken .
this leads to a continuous evolution for the system of interest .
although this model should correspond in some limit to any continuous position measurement of a single oscillator at the standard quantum limit , one system which does realize it at least approximately is the dispersive regime of single atom cavity quantum electrodynamics ( qed ) . in this system
the position of an atom inside a high - finesse optical cavity causes a phase shift of the field driving the cavity which can be monitored by homodyne detection @xcite .
the sme for the conditioned evolution of the system in its it form is @xcite @xmath0dt+2\alpha { \cal d}% [ x]\rho _ { \text{c}}dt+\sqrt{2\alpha } { \cal h}\left [ e^{-i\phi } x\right ] \rho_{\text{c}}dw . \label{sme1}\ ] ] the superoperators @xmath1 $ ] and @xmath2 $ ] acting on a density matrix @xmath3 are @xmath1\rho = c\rho c^{\dagger } -\frac{1}{2}% c^{\dagger } c\rho -\frac{1}{2}\rho c^{\dagger } c$ ] and @xmath2\rho = c\rho + \rho c^{\dagger } -$]tr@xmath4 where @xmath5 is an arbitary operator .
we will imagine that the atom is harmonically bound , @xmath6 the constant @xmath7 describes the strength of the measurement interaction and in the cavity qed example depends on the strength of the coherent driving and of the damping of the cavity . for the moment we will consider a one - sided cavity and perfect detection so that all the output light is detected .
this assumption will be relaxed in sec.([heat ] ) . in a generalization of the caves and milburn model
we will allow projective measurements of any quadrature of the meters , not just position , since this can be realized in the cavity qed experiments by varying the phase @xmath8 of the local oscillator in the homodyne detection of the output light .
the resulting measurement current @xmath9 ( suitably scaled ) is @xmath10 this stochastic master equation with the full dependence on @xmath11 was discussed by diosi @xcite in the context of a phenomenological model of position measurement through photon scattering where the kind of measurement made on the scattered photon determines the value of @xmath11 .
clearly if we choose @xmath12 the homodyne detection is an effective measurement of the atomic position , whereas if @xmath13 the measurement results are independent of the system state and only contain information about the noisy potential seen by the atom . for @xmath14
the conditioned evolution is linear in the system state @xcite .
this is somewhat like a continuous quantum eraser where the continuous measurement of one quadrature of the probe system destroys the position information written into the other quadrature of the probe . for @xmath15
the sse corresponding to this model in the case of pure states of the system has been considered in @xcite . in this work
we will use a simpler means of solving the evolution equations which straightforwardly applies to mixed states .
it has been shown by jacobs and knight @xcite that the sse corresponding to eq .
( [ sme1 ] ) is one for which gaussian pure states
@xcite remain gaussian and pure under the evolution . thus ,
if the system is initially in a mixture of gaussian pure states , the conditioned state will remain gaussian under the sme ( [ sme1 ] ) .
this property holds true for single mode systems where the hamiltonian is at most quadratic in @xmath16 and @xmath17 and the operator @xmath5 appearing in the lindblad term @xmath1 $ ] is linear in @xmath16 and @xmath17 and in all likelihood for multimode linear systems also .
if we restrict ourselves to gaussian initial states for example to thermal states of the oscillator then there are only five quantities which completely define the state @xmath18,@xmath19,@xmath20,@xmath21 and @xmath22 . from now on we will use @xmath23 to indicate the conditioned expectation value tr@xmath24 .
the requirement that the initial states be gaussian is not unduly restrictive since these are typically the states that are most stable in linear systems and would therefore be a reasonable assumption for an initial state .
moreover , there is numerical evidence that non - gaussian pure initial states become approximately gaussian under the stochastic schrdinger equation evolution corresponding to the sme ( [ sme1 ] ) on timescales fast compared to those considered here @xcite .
it has also been shown that in at least one such linear system an arbitary density matrix can eventually be written as a probabilistic sum over gaussian pure states after sufficient evolution subject to the unconditioned master equation @xcite . in considering such a linear system we are in effect specifying a semiclassical evolution , since the equations of motion for the wigner function of the state in the unconditioned evolution are similar to the classical liouville equations for a phase space density , see for example @xcite
. however we here include the important quantum feature of the measurement backaction , which is represented by the fact that the momentum diffusion is determined by the accuracy with which the particle s position is monitored @xcite .
so , as the noise in the position measurement is decreased by increasing @xmath25 the momentum diffusion expressed by the lindblad term in the sme ( [ sme1 ] ) increases . following breslin and milburn @xcite we can derive a system of differential equations for the first and second - order moments from the sme ( [ sme1 ] ) since @xmath26tr@xmath27 .
a similar calculation is performed in @xcite .
we define the dimensionless quantities @xmath28 @xmath29 , the second - order moments @xmath30 , @xmath31 , @xmath32 and a dimensionless parameter describing the relative strengths of the measurement and harmonic dynamics , @xmath33 .
the heisenberg uncertainty principle now requires that @xmath34 .
a pure state has the property that @xmath35 representing the fact that a gaussian pure state is a minimum uncertaintly state for some pair of conjugate quadrature variables @xcite . in terms of these new quantities ,
the it stochastic differential equations for the first and second - order moments and the measured photocurrent are [ momev ] @xmath36 as in @xcite , the it rules for stochastic differential equations @xcite and the properties of gaussian states @xcite result in deterministic equations for the conditioned second - order moments which are decoupled from the equations for the means .
the constant term in the equation for @xmath37 refers to the momentum diffusion due to the position measurement and remains in the master equation for the unconditioned evolution .
the non - linear terms describe the conditioning of the state on the measurement .
the noisy contribution to the equation for @xmath38 seems a little like a stochastic impulsive force , however it is perhaps better to think of this term as updating the expected position given the measurement result @xmath39 in analogy with classical bayesian state estimation .
equations like those above for the second - order moments of the conditioned state arise very frequently in classical continuous - time observation and control problems .
they can be collected into a riccati matrix differential equation @xcite for the covariance matrix , @xmath40 a single variable riccati equation which arose from the stochastic schrdinger equation for this system was found and solved in @xcite . in practice
it may not be homodyne but rather heterodyne detection which can be experimentally achieved with noise at the quantum limit @xcite . in this case , the local oscillator is detuned from the cavity frequency by a frequency @xmath41 which is large compared to all system frequencies , with the result that the phase @xmath8 changes very rapidly .
the quantum theory of heterodyne detection is described by wiseman and milburn @xcite .
the appropriate conditioned evolution can be described by averaging all the trigonometric functions of @xmath8 in the evolution equations ( [ momev ] ) except where they are multiplied by it increments .
thus the equations for the second order moments are exactly those for homodyne detection with @xmath15 where @xmath42 is replaced by @xmath43 corresponding to halving the signal to noise ratio of the measurement . considering the stochastic integrals
@xmath44 in the limit firstly of infinite @xmath41 then of infinitesimal @xmath45 leads to equations for the first order moments under heterodyne detection @xmath46 where @xmath47 and @xmath48 are independent wiener increments .
again this is formally identical to the evolution with homodyne detection and @xmath15 where @xmath49 is replaced by @xmath50 and in which a second independent noise process @xmath48 is present .
scaled versions of the two quadrature components of the experimental photocurrent are given by @xmath51 again note the replacement of @xmath49 by @xmath50 in the equation for @xmath52 as compared to the equation for @xmath39 with @xmath15 .
so if the quadrature - phase @xmath53 current is collected and used to account for the noisy potential , or alternatively fed back in order to compensate this evolution , then heterodyne detection is equivalent to homodyne detection with half the signal to noise ratio as far as the motional state is concerned . for all phases of the local oscillator
@xmath54 the second - order moments possess a steady state .
for example if @xmath15 @xmath55 which defines a pure state that agrees with the solution given in @xcite .
the steady states are found to have exactly the same second order moments regardless of the initial purity of the system . assuming ideal detection
, the observer is eventually able to ascribe a pure state to the system .
when the harmonic oscillator dynamics dominates over the measurement ( @xmath56)@xmath57the steady conditioned state is approximately a coherent state with @xmath58 if the measurement dynamics dominate ( @xmath59 ) then @xmath60and the conditioned state is strongly squeezed in position , as one would expect for a position measurement which is rapid enough to overcome the internal dynamics of the system .
the product of the position and momentum variances is greater than that required by the heisenberg uncertainty principle as a result of the internal hamiltonian which gives a non - zero correlation @xmath61 .
finally , the scaling we have chosen for the variances makes the limit of the free particle appear singular , however this limit exists and the results agree with belavkin s @xcite .
pure steady states also exist for other values of @xmath8 but the full expressions are rather complicated so we will just consider two special cases .
if the oscillator dynamics dominate ( @xmath62 ) the steady conditioned states are insensitive to the local oscillator phase leaving the conditioned state nearly in a coherent state @xmath63 if @xmath59 the conditioned state is strongly dependent on the choice of local oscillator phase , @xmath64 rather surprisingly it is possible for some phases of the local oscillator that the momentum variance is in fact smaller than the position variance .
this is because the measurement for non - trivial @xmath8 is a simultaneous measurement of the position of the oscillator and the momentum kicks to which it is being subjected and this can result in a more sharply defined momentum than position .
this is really only a possibility for phases of the local oscillator where there is very little position information in the record .
in the cases @xmath13 the differential equations for the second order moments of the state are simply those that result from the unitary evolution of a simple harmonic oscillator and the sme ( [ sme1 ] ) describes a stochastic unitary evolution . as a result there is no steady state for the moments , which grow according to the conventional schrdinger equation . in this case
the measurement current is white noise . in @xcite
it was noted that where a lindblad operator is hermitian there exists an unravelling of the master equation which does not localize the conditioned state .
the reason for this is clear in this context , such an unravelling corresponds to a measurement in which the observer obtains no information about the system state .
the matrix riccati equation ( [ riccati ] ) has an analytic solution given by reid @xcite .
where @xmath65 and @xmath66 obey the linear coupled matrix equations @xmath67 and for times where @xmath65 is non - singular , the solution for the covariance matrix is @xmath68 the full solution for all the second - order moments and arbitary initial conditions is complicated and not particularly illuminating . in order to expose the general form of the solution we will just consider the position variance in the case of an initial state of the oscillator with @xmath69 @xmath70
we have scaled time by the harmonic oscillation frequency @xmath71 .
when @xmath72 the system is close to its steady state value regardless of @xmath73 .
the non - linearity of the terms describing the conditioning of the system state cause the time over which the conditioned state becomes pure to be independent of the initial temperature of the state .
for definiteness we will define a collapse time @xmath74 as being the time at which the state has become effectively pure . when @xmath75 this collapse time is @xmath76 since in this regime @xmath77 . in this regime @xmath78 and there are many oscillations of the particle before the state is determined .
that the time to determine a particular pure state of the system should increase with the frequency seems reasonable since unexpected values of the measurement current could be due to a mistaken idea of the position of the particle , the white noise in the measurement record or to motion due to the oscillation and these possibilities will be more difficult to distinguish if the atomic motion is fast
. for smaller @xmath49 , the measurement is becoming very good and this estimate for the collapse time is optimistic since it is hard to determine the state of the system in less than one period of the mechanical oscillation .
reductions in the conditioned momentum variance will only occur as the hamiltonian evolution creates a correlation between the position and the momentum of the state .
even if the harmonic potential were absent a continuous measurement of position will give some information about the momentum of the particle .
when @xmath59 the particle is essentially free as far as the measurement is concerned and the time for the state reduction to occur turns out to be @xmath79 which is determined solely by the strength of the measurement and the mass of the particle . in this situation @xmath80 .
increasing the measurement coupling means that the time for the measurement to take place is reduced and in the limit of infinite @xmath7 the model essentially describes a projective measurement of the position .
if heterodyne detection is used rather than homodyne detection then @xmath49 is replaced by @xmath50 in the above equations with obvious implications for the timescale of the system collapse .
a pure state describes a situation in which an observer has maximal information about the system .
a mixed state describes less than maximal knowledge of system and it turns out that the amount of missing information , in an information theoretic sense , required to complete the specification of the state may be measured by the von neuman entropy @xmath81tr@xmath82 of the density operator , see for example @xcite .
thus the entropy allows us to quantify the extent to which the measurement has determined the state of the system at a given time and also the extent to which other environmental couplings limit what an experimenter can say about the system state .
another commonly used measure of the `` mixedness '' of a given density matrix is the linear entropy or purity @xmath83tr@xmath84 . for a single mode gaussian state
these quantities are simple functions of the unitless `` area '' of the state in phase space @xmath85 @xcite @xmath86 note that @xmath87 is just the determinant of the covariance matrix of the position and momentum probability distributions for the conditioned state . for a pure state @xmath88 and
as the state becomes increasingly mixed it occupies a larger phase space area such that as @xmath89 . as we would expect
if the state is widely spread in phase space then our knowledge of the system is poor and the information needed to complete the description of the state is large .
the time evolution of the variances and the linear and von neumann entropies of the conditioned state with @xmath90 is plotted in fig .
( [ fig1 ] ) .
these parameters are chosen since the measurement dynamics are not fast enough to obscure the harmonic oscillation totally and because achieving very small values of @xmath49 will probably be difficult in practice .
several features of fig .
( [ fig1 ] ) are relevant .
firstly the initial very rapid reduction of the position variance is associated with the the first part of the measurement record making a reasonably accurate determination of the position . then over the timescale of the harmonic oscillation
the momentum variance also reduces as the dynamics correlate the position and momentum .
note that the reduction in momentum variance occurs only when there is a strong correlation between the position and the momentum .
as @xmath61 becomes small the position variance decays more rapidly and the reduction of the momentum variance slows
. eventually all the second - order moments decay to the steady state values predicted above .
this initial fast reduction of the position variance is accompanied by a fast reduction of the von neumann entropy which damps to zero as the system approaches steady state .
although we have been considering an abstract model for continuous position measurement this work is motivated by emerging experimental possibilities in areas such as cavity quantum electrodynamics ( qed ) .
the position dependent phase shift induced by an atom strongly coupled to a high finesse optical cavity mode in the dispersive limit of cavity qed @xcite realizes the abstract position measurement coupling considered here .
it is currently possible to detect the presence of a single cold atom in the cavity through measurements of the output field @xcite and great progress has been made in observing single atom events with broad bandwidth close to the dispersive regime @xcite .
the phase shift changes most quickly with position where the gradient of the field is greatest and if the atom is harmonically confined in this region then the model discussed above would be approximately realized .
there is also the far off - resonant optical potential which will lead to large forces on the atom in this regime which would move the atom quickly away from this region of the standing wave .
however an optical standing wave at a nearby frequency could in principle be tuned to cancel the ac - stark shift of the ground state in the region of the harmonic confinement .
in fact the dipole force could straightforwardly be included in our simplified model resulting in only minor changes for sufficiently strong harmonic confinement . for the moment
we will not specify the exact source of the potential confining the atom , but the use of light forces from a far off - resonant optical standing wave or a standing wave in another mode of the cavity are possibilities .
there is also experimental work aimed at confining ions in high finesse optical cavities which could realize such a system @xcite .
if we imagine a harmonic potential confining the atom to this region of the standing wave with some constant restoring force to overcome the dipole force then in the far - detuned and lamb - dicke limit the resulting sme would be exactly eq .
( [ sme1 ] ) above @xcite .
the constant @xmath7 would then be equal to @xmath91 , where @xmath92 is the maximal single photon rabi coupling in the cavity , @xmath93 is the number of photons present in the driven cavity in the steady state , @xmath94 is the detuning between the atomic and cavity resonances the external laser driving is on the cavity resonance , @xmath95 is the cavity field decay rate and @xmath96is the wave number for the light resonant inside the cavity . using the parameters of @xcite which are based on cavity parameters achieved by hood _
@xcite gives @xmath97 s@xmath98m@xmath99 and this determines the rate of decay of the off - diagonal terms in the position representation of the density matrix under the unconditioned evolution .
this rather large number means that the density matrix elements @xmath100where @xmath16 and @xmath101 are separated by nine nanometers or around 1% of a wavelength will decay in the unconditioned evolution at the rate @xmath102hz .
the decay of off - diagonal elements of the density matrix in a particular basis is often associated with decoherence and the emergence of classicality @xcite . in this case
the decoherence is due in to the measurement coupling . in @xcite
this decay of off - diagonal density matrix elements is described as state reduction , in this work we are interested in state reduction onto a pure state and the rate of this process is determined not only by @xmath7 but also by the length scale of a typical pure state of the uncoupled system .
thus we found above that reduction onto a pure state took place in a time @xmath103 and the dependence on the length scale of the harmonic oscillator is clear . in order to find the rate of collapse onto a pure state in the conditioned evolution it is therefore also necessary to know the oscillation frequency of the atom due to its harmonic confinement . assuming this is achieved optically , the value for @xmath104 could be in the range of tens to hundreds of kilohertz .
so for example in @xcite the potential for a cesium atom resulting from the same cavity and driving parameters gives @xmath105 khz , while @xmath106 khz has been achieved for cesium in optical lattices by haycock _
for such a hypothetical experiment with cesium we now have @xmath107 and @xmath108 respectively .
estimates , as outlined above , for the time for an experimenter to determine a pure state of the system through heterodyne detection are then @xmath109s and @xmath110s .
both these times are reasonably close to the minimum collapse time for this accuracy of detection which corresponds to the free particle limit with @xmath111 in current experiments detection efficiencies and bandwidths will have a significant effect on the information that can be gathered from the record . for trapped ions harmonic frequencies
would be around an order of magnitude larger and state reduction times would then also be around an order of magnitude longer .
however the cavity finesse used here for such experiments may be more difficult to achieve at typical frequencies of ion transitions and the size of the cavity will be limited by the ion trap electrodes . what is dramatic about this time is that it is so short , single cold atoms have been observed close to the centre of the cavity mode in the experiment at caltech for times of the order of hundreds of microseconds .
this is another confirmation of the extent to which this experiment operates near the standard quantum limit of position measurement ( @xcite and references therein ) and an indication of the importance of continuous quantum measurement theory to its interpretation .
note that it will be difficult to achieve very small values of @xmath49 for which the measurement dynamics dominate over the oscillation in the well .
this is the result of the two main constraints on the applicability of the model .
firstly it is necessary that the harmonic potential confine the atom to well within a wavelength , thus justifying the lamb - dicke approximation in the master equation ( [ sme1 ] ) .
this requires that the recoil frequency for the cavity transition @xmath112 be much smaller than the harmonic oscillation . moreover attaining the dispersive regime
requires that the saturation parameter @xmath113 be much smaller than one .
it is possible to express @xmath49 in terms of these quantities : @xmath114where @xmath115 is the cavity mediated spontaneous emission rate @xcite . assuming that values of @xmath116and @xmath117 are chosen to satisfy a particular level of approximation , the rate at which the conditioned states become pure is only increased by changing the cavity parameters through increasing @xmath118 relative to the oscillation frequency .
moreover , since both the first two factors in this expression for @xmath49 must be large for the sme ( [ sme1 ] ) to correspond to the cavity system , extremely large values of @xmath119are necessary to achieve rapid measurement of the system state .
if we are to interpret the conditioned state as the best description of the observer s knowledge of the quantum mechanical state given the results of a series of measurements , we would expect a similarity between these equations and classical bayesian state observation .
the analogy between the sme ( [ sme1 ] ) and kalman filtering for a classical position measurement was discussed in @xcite but only the equations for the position probability distributions were considered . here
we formulate the continuous time position measurement state observer for a classical harmonic oscillator and find that there is a close analogy between the sme and the classical theory for all moments of the conditioned probability distribution as long as we restrict ourselves to gaussian states and allow for noisy driving of the classical oscillator .
the problem of noisy , classical , continuous time position measurement of a harmonic oscillator can be formulated @xmath120 & = & e[\eta ( t)\eta ( t^{\prime } ) ] = \delta ( t - t^{\prime } ) , \\
e[\epsilon ( t)\eta ( t^{\prime } ) ] & = & f\delta ( t - t^{\prime } ) \end{aligned}\ ] ] we have used the same scaling of the variables as in the quantum problem .
we imagine that as well as having an imperfect measurement of the system the oscillator is subject to a white noise force
. there may be some correlation between the oscillator ( plant or process ) noise and the measurement noise and so @xmath121 and @xmath122 are correlated wiener processes . as in the quantum mechanical case the limit of small
@xmath49 is the limit of good position measurement .
the continuous time theory of kalman filtering then provides the best estimate of the system state @xmath123 and the second - order moments of the posterior probability distribution @xmath124 @xcite , @xmath125 where @xmath126 is an independent wiener increment with @xmath127 proportional to the innovation process @xmath128 .
note that the circumflex employed here indicates that the quantity is an estimate of the classical variable and not that it is a quantum operator .
the moments of the posterior probability distribution have been given the same notation as the moments of the conditioned quantum mechanical state .
if we make the identifications @xmath129 then this system of equations is formally identical to the system which determines the evolution of the quantum mechanical conditioned state eqns ( [ momev ] ) .
we see that there is a classical model of noisy position measurement for which the equations of motion for the posterior probability distribution of the classical state given by the kalman filter reproduce the stochastic master equation .
what is specifically quantum mechanical in the sme is that we can not , even in principle , specify the process noise and the measurement noise separately .
classically one could imagine isolating the system sufficiently that @xmath117 is as large as we like .
with @xmath130 there would be no momentum disturbance on the atom and after a sufficiently long observation time the state of the system would be determined exactly so that @xmath131 clearly this does not correspond to a quantum state .
however the quantum theory of the problem guarantees that any coupling to the system which gives position information about the state of the system must also disturb the momentum .
this momentum disturbance must be sufficient that the conditioned state always obeys the heisenberg uncertainty principle .
thus only some classically possible models of position measurement are allowed by quantum mechanics . for a given level of measurement noise
, the process noise must be at least sufficient to ensure that the observer can never infer probability distributions for the position and momentum which do not constitute valid quantum states .
this is a measurement - disturbance uncertainty relation @xcite for continuous measurement ; reducing the noise in the measurement must increase the noise in the evolution .
this backaction noise does not however behave like classical process noise , its properties are entirely determined by the measurement .
if we vary the basis of the measurement on the meter vary @xmath8 then we are able to alter the correlation between the measurement noise and the apparent process noise in the classical model .
we have found here that the symmetric moments of the conditioned state always obey a system of equations which also describes kalman filtration for a classical problem , but that altering the specific quantum measurement alters both the observation _ and _ the process noise of the relevant classical problem . since the symmetric moments are the moments of the wigner function , for example @xcite , the conditioned wigner function can be interpreted as the direct analogue of a classical posterior probability distribution for the system .
this relationship is not going to be as straightforward for more complicated quantum systems where the conditioned wigner function can be negative and can not be interpreted as a probability distribution .
we have established above that the conditioned state of the system can in priniciple become pure after a finite observation time .
what might not be clear is that the state which results is uniquely determined by the measurement record . in an ideal projective measurement
the probabilities of obtaining the various measurement results depend upon the initial state .
however , once a result has been obtained the conditional state after the measurement depends on that result alone . in a similar way , for continuous measurements we find that , while the initial state affects the probability of obtaining particular measurement records , the conditional state following the measurement is determined by the particular measurement record which was obtained , provided that record sufficiently long . in particular if the initial state of the system is very poorly known to an observer then we might hope that there is effectively a maximum likelihood estimate of the system state which depends only on the measurement current and which converges to the actual system state within @xmath132 .
if propagating the stochastic master equation with the actual initial state of the system provides the _ a posteriori _ estimate of the system state given the measurement results , then such a maximum likelihood estimate would result from propagating an initial state with very large position and momentum variances in the sme ( [ sme1 ] ) , say @xmath133 which gives a nearly flat prior probability distribution for the initial state . in this section
we demonstrate that such a strategy does indeed work .
thus the purity of the conditioned density matrix indicates , as would be hoped , that there is only one pure state of the system at time @xmath134 which is consistent with the known sequence of measurement results .
suppose that two observers , alice and bob say , postulate different initial states of the system @xmath135 which we will continue to assume to be gaussian .
for example , alice may have more information than bob about the system in which case bob would start with a more mixed initial density matrix reflecting his initial lack of knowledge .
given that they both recieve the same measurement record @xmath136 and propagate their conditioned states according to the sme ( [ sme1 ] ) , it should be the case that at some time @xmath137 , around @xmath138 , alice and bob agree on the system state , so that @xmath139 . from the previous section
we know that after the time @xmath140 the second order moments of both conditioned states will be equal @xmath141 and so we focus on the equations for the first order moments of the conditioned state for each observer @xmath142 in this section we will omit the tildes used above to indicate that we have scaled the position and momentum to the natural units for the harmonic oscillator .
the stochastic increment @xmath143 is the infinitesimal increment of the measured current which both alice and bob have access to .
we can express the increment @xmath144 in terms of the other quantities @xmath145 and find stochastic differential equations for the differences between the means @xmath146 , [ errors ] @xmath147 the deterministic part of this system of equations describes a damped harmonic oscillation for @xmath148 where the damping and oscillation rates depend on the second - order moments of bob s conditioned state .
the damping in these equations is not present in an analogous equation for @xmath149 given by mabuchi @xcite for a free particle .
the equation adopted in @xcite is obtained from the continuous limit of the repeated position measurement model of caves and milburn @xcite and does not contain a noise term in the stochastic differential equation for @xmath16 . in fact the omission of this term is in error and if the continuous limit of the repeated position measurement model of caves and milburn is taken correctly then the noisy contribution to @xmath150 which we obtained from the sme ( [ sme1 ] ) , is in fact present .
it is the damping which results from this term which leads to all observers agreeing about the conditioned state after a sufficiently long observation time .
note that after the time @xmath151 eqs ( [ errors ] ) are in fact a system of ordinary differential equations since at that point the covariances of the two conditioned states are equal .
the differences in the means then damp to the steady state values @xmath152 indicating that alice and bob do eventually agree about the state of the system regardless of their initial states .
thus we have shown that the conditioned state eventually depends only on the measurement record but not the time - scale over which this occurs .
it is straightforward to use the it chain rule @xcite to find differential equations for the expectation values of the covariance matrix for the difference between the conditioned means of alice and bob [ covev ] @xmath153\right ) & = & 2e[e_{x}e_{p}]-% \frac{4}{r}v_{xx}^{\text{b}}e[e_{x}^{2 } ] \nonumber \\ & & + \frac{2}{r}\left ( v_{xx}^{\text{b}}-v_{xx}^{\text{a}}\right ) ^{2 } , \\
\frac{1}{\omega } \frac{d}{dt}\left ( e[e_{p}^{2}]\right ) & = & -2e[e_{x}e_{p}]-% \frac{4}{r}v_{xp}^{\text{b}}e[e_{x}e_{p } ] \nonumber \\ & & + \frac{2}{r}\left ( v_{xp}^{\text{b}}-v_{xp}^{\text{a}}\right ) ^{2 } , \\
\frac{1}{\omega } \frac{d}{dt}\left ( e[e_{x}e_{p}]\right ) & = & e[e_{p}^{2}]-e[e_{x}^{2 } ] \nonumber \\ & & -\frac{2}{r}v_{xp}^{\text{b}}e[e_{x}^{2}]-\frac{2}{r}v_{xx}^% { \text{b}}e[e_{x}e_{p } ] \nonumber \\ & & + \frac{2}{r}\left ( v_{xx}^{\text{b}}-v_{xx}^{\text{a}}\right ) % \left ( v_{xp}^{\text{b}}-v_{xp}^{\text{a}}\right ) .\end{aligned}\ ] ] although we have taken the expectation value for this system of equations , the noise terms all become zero after @xmath138 so these ordinary differential equations eventually describe the whole dynamics .
we are now interested in the time - scale over which the determinant of this covariance matrix becomes zero .
unfortunately the time dependence of the conditioned state variances prevents a closed form solution of this system of equations and we have not found a matrix riccati form for the overall system .
it is however straightforward to investigate the problem numerically .
we wish to show that all observers will agree about the conditioned state of the system in roughly @xmath154 . to do this it is sufficent to show that an arbitary observer will agree with some preferred observer in that time . for this reason we will assume that alice has sufficient information to describe the state of the system as a pure state and that she has access to sufficient earlier measurement records that @xmath155 .
an experimenter is going to be in bob s position of not having precise knowledge of or control over the preparation of the state .
we would expect that if bob makes an accurate assessment of his initial knowledge of the state then @xmath156 $ ] is at first of the same order as @xmath157 or smaller . since @xmath158 may be large initially , the stochastic term in the equations ( [ errors ] ) for bob s errors dominates for very short times and essentially determines a random initial condition for @xmath159 such that @xmath160\lesssim v_{xx } $ ] .
this reflects the fact that for short times the measurement record is dominated by noise . from this point on we find numerically that @xmath156\simeq v_{xx}$ ] and that all the elements of the covariance matrix damp to zero within @xmath138 .
thus we have found numerically that the conditioned state depends only on the measurement record after a time equal to the time over which the conditioned state becomes pure .
most importantly we found this to be the case even where all the variances were set to very large initial values , the largest we tried being @xmath161(0)= e[e_{p}^{2}](0)=10^{10}$ ] .
thus as we anticipated , bob can make an accurate estimate of the state within the collapse time even in the absence of accurate information about the initial state . in this case
bob s conditioned state corresponds to the maximum likelihood estimate discussed above .
interestingly , even with the pessimistic initial condition where @xmath156(0)$ ] is significantly larger than @xmath162 this corresponds to bob overestimating the accuracy of his estimate of the particle s position and the difference between his estimate and the actual value being very large the determinant of the covariance matrix damps on the characteristic timescale @xmath163 .
this feature is more pronounced for larger values of @xmath164 again this is because for large initial values of @xmath165 , bob s estimate of the system state is essentially dependent on the measurement current alone .
these behaviours are demonstrated for the parameters used in the previous section in figure ( [ fig2 ] ) . from this
we can conclude that after the time @xmath138 , any experimenter knows the state of the system regardless of how the system is initially prepared and of how much control the experimenter has over this process . in the next section when we introduce detection efficiency and thermal couplings , these
will just modify the second - order moments of the conditioned state such that the steady conditioned state is no longer pure . since the equations for
the first order moments will be unmodified all that is necessary to reproduce the results of this section in this more general case is to use the collapse time @xmath138 which is appropriate for the new system .
although this will mean that the experimenter is left with broader position and momentum probability distributions , it will not mean that different experimenters disagree about the means of these distributions .
so regardless of the detection efficiency the conditioned state is eventually uniquely determined by the measurement record and can be regarded as known by the experimenter .
in order to discuss the effects of detection efficiency and other uncontrolled coupling to a bath in the model we will add an extra momentum diffusion term to the sme to obtain for @xmath166 @xmath167dt+2\alpha { \cal d}% [ x]\rho _ { \text{c}}dt+2\beta { \cal d}[x]\rho _ { \text{c}}dt \nonumber \\ & & + \sqrt{2\alpha } { \cal h}[x]\rho _ { \text{c}}dw .
\label{sme2}\end{aligned}\ ] ] this simple modification to the master equation is intended to model several possible imperfections in a real experiment .
one contribution to @xmath168 is the effect of detection efficiency @xcite so that @xmath168 is at least @xmath169 where the overall detection efficiency is @xmath170 .
the effect of cavity loss through the unmonitored mirror in a cavity qed experiment is an effective detection inefficiency .
where the loss rates out of the two mirrors are @xmath171 , and if only the light passing through the second mirror is detected , then we get @xmath172 where @xmath173 .
the best situation is if the mirror in front of the detection apparatus has significantly higher transmission than the mirror used to drive the cavity .
scattering losses of the mirror will also be an effective detection inefficiency but these are typically much smaller than transmission losses . in experiments with single atoms there
will also be heating due to spontaneous emission into free space which will lead to a contribution to @xmath168 equal to @xmath174 where @xmath175 is the free space decay rate of the excited state of the atomic transition .
other than the restriction to the lamb - dicke regime this is the largest correction to the adiabatically eliminated master equation given above when a moderate detuning from the atomic transition is employed as in @xcite .
the spontaneous emission contribution to the heating is also proportional to the measurement coupling @xmath7 such that @xmath176 and to minimize the effect of spontaneous emission we must use cavities with the largest possible value of @xmath177 recall that if this rate is large the signal to noise of the position measurement also improves .
contributions to @xmath168 in other systems , for example the interferometric detection of the position of a moving mirror @xcite , will also come from any coupling of the oscillator to a thermal bath . we found that the standard quantum brownian motion master equation @xcite led to steady conditioned states that did not obey the heisenberg uncertainty principle for small values of @xmath178 .
this a result of the non - lindblad terms in this master equation . the master equation
we adopt here solves this problem by considering only coupling to very high temperature thermal baths for which the thermal contribution to @xmath168 is@xmath179 where @xmath180 is the coupling rate to the thermal reservoir and @xmath181 is the temperature of the bath . this should be an adequate description of heating in the experiment as long as the bath to which the system is coupled is of sufficiently high temperature that only the diffusive evolution is significant for the timescales of interest .
there is now no sse equivalent to this sme ( [ sme2 ] ) and pure states become mixed during the evolution .
however the sme continues to preserve gaussian states so the previous calculation for the evolution of state can be straightforwardly modified .
the second - order moments of the conditioned state still reach a steady state and we can easily find an expression for the steady conditioned state phase space area @xmath182is given by @xmath183 the linear and von neumann entropies of the steady conditioned states are plotted for a range of detection efficiencies in fig .
( [ fig3 ] ) .
even though we have effectively coupled the oscillator to an infinite temperature bath the conditioned steady states have in some sense a finite temperature but stochastically varying mean values of position and momentum .
we will consider some limiting cases here for the steady state variances in this more general situation . where the measurement is strong ( @xmath59 ) the position variance is insensitive to these imperfections , @xmath184 .
if the dynamics dominate ( @xmath56 ) then the position and momentum variances have the same dependence on @xmath168 as @xmath87 , @xmath185 .
finally the whole time evolution of the second - order moments of the conditioned state can be determined by solving the matrix riccati equation .
again we will just consider the position variance as a function of time where @xmath186 @xmath187 and we have scaled time by the harmonic oscillation frequency @xmath188 . as before when @xmath189the system reaches its steady state value and the non - linearity of the terms describing the conditioning of the system state mean that the time for this to occur is independent of the initial state .
interestingly this time is in fact shorter than was required to purify the conditioned state in the case of ideal detection .
this is essentially because the extra noise means that past observations become irrelevant more quickly , not leaving enough time to determine a pure state completely.while the steady state is reached increasingly fast it corresponds to an increasingly high effective temperature .
when @xmath75@xmath57this time to reach the steady state is @xmath190 since in this regime @xmath191 .
as was noted in the previous section the time for the conditioned state variances to reach their steady state is also the time that is necessary for different observers to agree about the system state .
the time evolution of some of these quantities is plotted in figure ( [ fig4 ] ) .
in this paper we have established that for a simple class of systems quantum trajectory theories allow the determination of a unique post - measurement state that depends only on the measurement results over a finite time .
we have discussed the effects of experimental imperfections on this state determination and the analogy to classical state observation .
the systems to which this analysis is applicable have the property that the conditioned density matrix is at all times gaussian and its evolution is exactly that of the posterior probability distribution for an appropriate classical state observer .
while we have considered only the case of position measurement the same treatment will be applicable to these other linear systems .
clearly in more complicated systems , such as the resonant interaction of an atom with the single mode of an optical cavity , this will not always be the case and the conditioned wigner function will sometimes be non - positive .
we would however expect that the central result we have shown here , the purity of the conditioned state after sufficiently long continuous observation and the dependence of this state on the initial state only through the measurement results , will still hold for these more interesting and complicated systems .
this will however require numerical simulation of the stochastic master equation for such systems .
another feature that should generalize is the interpretation of the sme as a state observer presumably an optimal one for the quantum system .
a. d. would like to thank kurt jacobs for continuing stimulating discussions on continuous measurement theory and in particular for suggesting that it would be interesting to calculate the purity of conditioned states with and without thermal couplings .
it is also a pleasure to acknowledge discussions with hideo mabuchi .
this work is supported by the marsden fund of the royal society of new zealand .
a. n. korotkov , ( unpublished ) ; k. mlmer , j. mod
. opt . * 44 * , 1937 ( 1997 ) ; j. ruostekoski and d. f. walls , phys .
a * 58 * , r50 ( 1998 ) ; h. mabuchi and h. m. wiseman , ( unpublished ) ; k. jacobs , private communication ( unpublished ) . | the possibility of determining the state of a quantum system after a continuous measurement of position is discussed in the framework of quantum trajectory theory .
initial lack of knowledge of the system and external noises are accounted for by considering the evolution of conditioned density matrices under a stochastic master equation .
it is shown that after a finite time the state of the system is a pure state and can be inferred from the measurement record alone .
the relation to emerging possibilities for the continuous experimental observation of single quanta , as for example in cavity quantum electrodynamics , is discussed . |
over the past decade , x - ray observations have made it increasingly clear that black holes are capable of accreting gas under a variety of flow configurations . in particular ,
there is convincing evidence that a distinct form of hot accretion exists at sub - eddington rates , which contrasts with the classical `` cold and thin '' accretion disk scenario ( shakura & sunyaev 1973 ) . hot accretion appears to be common in the population of supermassive black holes in galactic nuclei and during the quiescent phases of accretion onto stellar - mass black holes in x - ray transients ( e.g. narayan et al .
1998a ; lasota et al . 1996 ; di
matteo et al .
2000 ; esin et al .
1997 , 2001 ; menou et al .
1999 ; see narayan , mahadevan & quataert , 1998b ; melia & falcke 2001 ; narayan 2003 ; narayan & quataert 2005 for reviews ) .
theories for the structure and properties of hot accretion flows remain controversial , however . inspired by the work of shapiro , lightman & eardley ( 1976 ) , narayan & yi ( 1994 ; 1995a , b ) derived self - similar adaf solutions which emphasized the important stabilizing role of radial heat advection ( see also ichimaru 1977 ; rees et al .
1982 ; abramowicz et al . 1988 ) .
subsequent analytical work on hot accretion flows has emphasized the possibility of outflows , motivated by a positive bernoulli constant ( narayan & yi 1994 ; adios : blandford & begelman 1999 ) , and the potential role of convection ( cdafs ; narayan et al .
2000 , quataert & gruzinov 2000a ) .
hydrodynamical and mhd numerical simulations , on the other hand , have greatly contributed to the subject by highlighting important dynamical aspects of the problem which are not captured by idealized analytical models ( e.g. stone et al .
1999 ; stone & pringle 2001 ; hawley et al .
2001 ; devilliers et al . 2003 ; proga & begelman 2003 ; mckinney & gammie 2004 ) .
the purpose of this work is to consider the possibility that thermal conduction , which has been a largely neglected ingredient , could affect the global properties of hot accretion flows substantially . in 2 , we use existing observational constraints on a few nearby galactic nuclei to argue that hot accretion is likely to proceed under weakly - collisional conditions in these systems , thus implying that thermal conduction could be important . in 3 , we extend the original 1d self - similar solutions of narayan & yi ( 1994 ) to include a saturated form of thermal conduction and we study the effects of conduction on the flow structure and properties . in 4 , we generalize the 2d solutions of narayan & yi ( 1995a ) in the presence of thermal conduction and find that this extra degree of freedom allows the emergence of rotating thermal outflows . in 5 , we discuss our results in the context of recently established observational trends for hot accretion flows and related outflows . in 6
, we comment on several limitations of our work , potential additional consequences of strong thermal conduction in hot accretion flows as well as directions for future work .
_ chandra _ observations provide tight constraints on the density and temperature of gas at or near the bondi capture radius in sgr a * and several other nearby galactic nuclei .
these observational constraints have been used before to estimate the rate at which the gas is captured by the black hole in these systems , following bondi theory ( e.g. loewenstein et al . 2001 ; di matteo et al .
2003 ; narayan 2003 ) .
here , we use these same constraints ( taken from loewenstein et al . 2001 ; baganoff et al . 2003 ; di matteo et al . 2003 ; ho , terashima & ulvestad 2003 ) to calculate mean free paths for the observed gas .
it should be noted that these constraints are not all equally good , in the sense that _ chandra _ has probably resolved the gas capture radius in sgr a * and m87 but not in the other nuclei .
galactic nuclei and their associated gas properties on @xmath4 scales are listed in table [ tab : one ] : @xmath5 is the gas number density , @xmath6 is the temperature , @xmath7 is the @xmath4 size - equivalent at the nucleus distance , @xmath8 is the mean free path for the observed @xmath4 conditions and @xmath9 is the capture radius , inferred from the gas temperature and the black hole mass in each nucleus . from spitzer s ( 1962 )
expressions for collision times , a standard expression for the thermal speed and a coulomb logarithm @xmath10 , one gets the simple mean free path scaling @xmath11 cm ( cowie & mckee 1997 ; valid for both electrons and protons in a 1-temperature gas ) .
we have used this relation to calculate @xmath8 values in table [ tab : one ] .
the inferred mean free paths are in the few hundredths to few tenths of the observed @xmath4 scales ( @xmath12 values in table [ tab : one ] ) . assuming no change in the gas properties down to the inferred capture radius , values of @xmath13
are deduced in four of the six nuclei .
one could also reformulate these scalings by stating that the @xmath4 mean free paths are systematically @xmath14 , where @xmath15 is the schwarzschild radius ( the typical length - scale associated with the accreting black holes ) .
these numbers suggest that accretion will proceed under weakly - collisional conditions in these systems .
the weakly - collisional nature of adafs has been noted before ( e.g. mahadevan & quataert 1997 ) , in the sense of collision times longer than the gas inflow time , but these are model - dependent statements .
direct observational constraints on the gas properties near the capture radius favor a weakly - collisional regime of hot accretion , more or less independently of the exact hot flow structure .
whether the gas adopts an adaf , cdaf or adios type configuration once it crosses the capture radius , it is expected to become even more weakly - collisional as it approaches the black hole , since the relative mean free path scales as @xmath16 in these flow solutions with a virial temperature profile and a density profile @xmath17 .
tangled magnetic fields in the turbulent accretion flow would likely reduce the effective mean free paths of particles .
the magnitude of this reduction , which will depend on the field geometry , is currently unknown .
we return to this important issue in our discussion section .
even if the effective mean free paths of electrons are reduced to values @xmath18 , thermal conduction may be important .
it is therefore of interest to investigate its effects on the structure and properties of hot accretion flows in general .
we begin with a study of 1d adaf models with conduction in 3 and then move on to study a 2d version of these hot accretion flows in 4 .
the standard spitzer formula for thermal conduction applies only to gas well into the collisional ( `` fluid '' ) regime , with a mean free path @xmath19 , for any relevant flow scale @xmath20 .
we expect radial temperature variations on scales @xmath21 from self - similarity . in a weakly - collisional regime with @xmath22 or even @xmath23 , a saturated ( or equiv .
`` flux - limited '' ) thermal conduction formalism should be adopted if one is to avoid unphysically large heat fluxes .
we adopt the formulation of cowie and mckee ( 1977 ) and write the saturated conduction flux as @xmath24 , where @xmath25 is the saturation constant ( presumably @xmath26 ) , @xmath27 is the gas mass density and @xmath28 is its isothermal sound speed . with this prescription ,
the largest achievable conduction flux is approximated as the product of the thermal energy density in electrons times their characteristic thermal velocity ( assuming a thermal distribution and equal electron and ion temperatures ; see cowie and mckee 1977 for details ) .
because it is a saturated flux , it no longer explicitly depends on the magnitude of the temperature gradient but only on the direction of this gradient .
heat will flow outward in a 1d hot accretion flow with a near - virial temperature profile , hence the positive sign adopted for @xmath29 . with this formulation for conduction ,
the steady - state 1-temperature self - similar adaf solution of narayan & yi ( 1994 ) can be generalized to include the divergence of the saturated conduction flux in the entropy - energy equation ( see also honma 1996 and manmoto et al .
2000 for investigations of the role of `` turbulent '' heat transport in one - dimensional adaf - like flows ) . adopting the same cylindrical geometry and notation as narayan & yi ( 1994 )
, we consider a `` visco - turbulent '' , differentially rotating hot flow which satisfies the following height - integrated equations for the conservation of momentum and energy @xmath30 supplemented by the continuity equation , which is rewritten as @xmath31 .
the last term in equation ( [ eq : ent ] ) is the additional divergence of the saturated conduction flux . in these equations
, @xmath32 is the cylindrical radius , @xmath33 is the keplerian angular velocity around the central accreting mass , @xmath34 is the gas radial inflow speed , @xmath35 is its angular velocity , @xmath27 is the mass density , @xmath28 is the isothermal sound speed , @xmath36 is the gas specific entropy , @xmath37 is its temperature , @xmath38 is the flow vertical thickness and @xmath39 is the advection parameter ( @xmath40 for negligible gas cooling ) .
a shakura - sunyaev prescription has been used to capture the `` visco - turbulent '' nature of the flow , with an equivalent kinematic viscosity coefficient @xmath41 , where @xmath42 is the traditional viscosity parameter .
we look for solutions to these equations of the form @xmath43 and use the notation @xmath44 .
we find that the three unknowns @xmath45 , @xmath46 and @xmath47 satisfy the relations @xmath48 while the density scale , @xmath49 , is fixed by the accretion rate , @xmath50 .
contrary to the original adaf solution , the energy relation ( eq . [ [ eq : enrel ] ] ) is not a quadratic in @xmath51 because of the extra saturated conduction term .
we solve this fourth order polynomial with a standard numerical technique ( press et al .
1992 ) and discard the two imaginary roots as well as the real root with negative @xmath47 , which is unphysical because of a conduction flux going up the temperature gradient .
the main features of the new self - similar solutions with saturated conduction are illustrated in figure [ fig : one ] , as a function of the value of the saturation constant , @xmath52 .
we are showing results for two distinct solutions here , with @xmath53 and @xmath54 , while we have fixed @xmath40 and @xmath55 in both cases . even though a value of @xmath56 may not be physically meaningful , it is useful to illustrate more clearly how solutions depend on @xmath57 .
original 1d adaf solutions are recovered at small @xmath25 values .
as @xmath25 approaches unity , however , the solutions deviate substantially from the standard adaf . the sound speed , @xmath28 , and the radial inflow speed , @xmath58 , both increase with the magnitude of conduction , while the squared angular velocity , @xmath59 , decreases .
formal solutions still exist when @xmath59 becomes negative but they are obviously physically unacceptable .
the solution with @xmath53 reaches the non - rotating limit at @xmath60 , while the solution with @xmath61 reaches this limit for @xmath62 .
both solutions have sub - virial temperatures ( @xmath63 ) and subsonic radial inflow speeds ( @xmath64 ) , which appear to be rather general properties .
supersonic solutions may exist for large @xmath42 values when @xmath65 .
as the level of saturated conduction is increased , more and more heat flows from the hotter , inner regions , resulting in a local increase of the gas temperature relative to the original adaf solution .
simultaneously , the gas adjusts its angular velocity ( which reduces the level of viscous dissipation ) and increases its inflow speed to conserve its momentum balance ( increasing at the same time the level of advection ) .
one can show from the energy equation that @xmath66 , where @xmath67 and @xmath68 are the magnitudes of the conduction and advection terms : solutions cease to exist once advection balances exactly the heating due to saturated conduction .
we have found that variations in the advection parameter , @xmath69 , have relatively minor effects on the solutions , as long as @xmath69 is not @xmath70 .
the breakdown of the solutions ( when @xmath71 ) occurs at lower values of the saturation constant @xmath25 for smaller @xmath42 and larger @xmath57 values .
solutions with @xmath72 can be viewed as bondi - like solutions with saturated conduction .
clearly , conduction can significantly affect the structure and properties of hot accretion flows , even at a level well below the saturated flux value ( i.e. for @xmath73 ) .
by analogy with the reasoning of narayan & yi ( 1995a ) on convection in 2d self - similar adaf solutions , one might expect conduction to be more important in the polar regions of hot accretion flows than at their equator . to better understand the role of conduction , it is therefore important to study it in a more realistic two - dimensional setting , without height integration .
we now extend the analysis of the previous section to two - dimensional hot accretion flows , following closely the methods of narayan & yi ( 1995a ) .
we also discuss in appendix c the relation between our work and that of anderson ( 1987 ) . adopting spherical polar coordinates ( @xmath74 )
, we consider axisymmetric , steady - state flows ( @xmath75 ) .
we define keplerian angular and linear velocities as , respectively , @xmath76 and we use the same @xmath42-prescription as before to describe the visco - turbulent nature of the flow : @xmath77 where @xmath78 is the isothermal sound speed ( @xmath79 ) .
the fluid equations obeyed by the visco - turbulent accretion flow are written as follows : @xmath80+\frac{1}{\rho}\nabla \cdot ( \lambda \nabla t),\end{aligned}\ ] ] where @xmath81 is the viscous stress tensor ( divided by the dynamic viscosity , @xmath82 ) .
the notation for the other flow variables is the same as in 3 .
it should become clear , later on , that our main results are not expected to depend critically on the exact form of the viscous stress tensor adopted . to allow a direct comparison with our results in 3 and for consistency with narayan & yi ( 1995a )
, we use the same advection parameter @xmath69 to quantify the cooling efficiency of the flow .
we will see , however , that this parameter is in fact ill - defined for solutions with conduction .
fully developed forms of these equations ( with a few typos and no thermal conduction term ) can be found in narayan & yi ( 1995a ) .
contrary to the formulation adopted in 3 , we do not write the conduction flux in a saturated form here , but return instead to a traditional form where the flux depends linearly on the local temperature gradient . justifications for this choice are provided in appendix a. we adopt a form of the thermal conductivity coefficient which preserves the radial self - similarity of the solutions : @xmath83 following narayan & yi ( 1995a ) , we consider separable solutions satisfying @xmath84 in narayan & yi ( 1995a ) , the @xmath85 in the equivalent of eq .
( [ sin ] ) is missing , in what appears to be a typo .
note that we make the same @xmath86 assumption as narayan & yi ( 1995a ) . with these forms ,
solutions automatically satisfy the continuity equation , eq .
( [ cont ] ) .
the problem is reduced to a one - dimensional differential system along the polar coordinate , @xmath87 .
we are left to solve the following four coupled differential equations for the dimensionless variables @xmath88 and @xmath89 : @xmath90 , \label{momc}\\\nonumber\\ -\frac{3}{2}\epsilon ^{\prime } v_r & \:=\:&\alpha\left\{3v_r^{2}+ \left(\frac{dv_r}{d\theta}\right)^{2 } + \sin^{2}\theta\left[\frac{9}{4}\omega^{2 } + \left(\frac{d\omega}{d\theta}\right)^{2}\right ] \right\}\nonumber\\ & & \qquad\qquad+\lambda_{0}^{\prime}\frac{1}{p } \left[c_{s}^{2}+\frac{1}{\sin\theta}\frac{d}{d\theta } \left(\sin\theta \frac{d c_{s}^{2}}{d\theta}\right)\right ] . \label{energy}\end{aligned}\ ] ] for convenience , the parameters @xmath91 were introduced in the energy equation ( eq . [ [ energy ] ] ) , where @xmath92 is the adiabatic index of the gas .
this is a seventh - order differential system of four variables .
the three momentum equations ( eqs [ [ moma][momc ] ] ) are identical to those presented in narayan & yi ( 1995a ) .
the energy equation ( eq . [ [ energy ] ] ) differs in that extra terms corresponding to the radial and polar contributions to heat conduction are now present .
this qualitative difference turns out to be crucial in providing the extra degree of energetic freedom necessary for inflow or outflow at the pole . while the original energy equation seemed to require that rotating solutions be static ( @xmath93 ) along the polar axis because of the positivity of all the viscous terms ( no rotating solutions with @xmath94 at @xmath95 were found by narayan & yi 1995a ) , the addition of thermal conduction terms allows for radial velocities at the pole which are generally non - zero .
furthermore , for outflows to be energetically favored , the polar contribution to heat conduction ( second term in the last square bracket in eq .
[ [ energy ] ] ) must be negative enough to overcome the positive sum of the radial conduction term and all the viscous terms on the rhs of eq .
( [ energy ] ) . when that happens , advection can act as an effective heat source , with a positive sign for the radial velocity , @xmath34 ( on the lhs of eq .
[ [ energy ] ] ) .
as we shall see below , this sign change for @xmath34 preferentially happens in polar regions of the flow and only for certain combinations of model parameters .
boundary conditions for the system of differential equations are obtained by requiring that all flow variables be symmetric and continuously differentiable across the pole and the equator .
this corresponds to the following conditions at @xmath96 and @xmath97 : @xmath98 as noticed by narayan & yi ( 1995a ) , these conditions are not all independent from each other , given eqs .
( [ moma][energy ] ) , so that the system ends up not being over - constrained . the net mass accretion rate , which one obtains from the volume integral of the continuity equation ( eq . [ [ cont ] ] ) , is @xmath99 where the radial dependence has been omitted because of self - similarity .
this integral supplies two additional boundary conditions to our system because we enforce it by solving the corresponding differential equation for @xmath100 and imposing the values of @xmath101 at @xmath95 and @xmath102 .
the @xmath101 integral then uniquely determines the scaling of @xmath103 in our solutions . in what follows ,
we have used @xmath104 ( from @xmath105 to @xmath106 ) consistently in all the solutions .
we find numerical solutions with an implementation of the relaxation code described in hameury et al .
we do not implement the adaptive mesh feature but simply solve the set of equations on a grid of @xmath107@xmath108 points equally spaced in @xmath87 .
after recovering the original 2d adaf solutions of narayan & yi ( 1995a ) , we explored the parameter space of solutions with conduction in detail , in terms of the model parameters @xmath42 , @xmath109 and @xmath110 .
figure [ fig1 ] shows latitudinal profiles ( @xmath95 at the pole ) of the four dynamical variables in our model , for solutions with @xmath111 , @xmath112 and several values of @xmath110 . except for the density , @xmath27 ( which is normalized to @xmath113 ) , the radial velocity , @xmath34 , the angular velocity , @xmath35 , and the isothermal sound speed squared , @xmath114 , are all scaled to the local keplerian values . for small enough values of @xmath110 ( = @xmath115 ) ,
solutions remain very close to the original 2d adaf solutions of narayan & yi ( 1995a ) , with @xmath34 ( nearly ) zero at the pole .
note that this solution , with @xmath112 , is denser and cooler at the equator than at the pole and it corresponds to something like a `` thin disk '' surrounded by a hot coronal atmosphere . as the value of @xmath110
is increased , the solutions start deviating substantially from the original adaf , with faster radial inflow at the equator and significant radial outflow appearing in polar regions . the solution with @xmath116 in fig .
[ fig1 ] illustrates this transition clearly .
the flow also becomes hotter ( except around the poles ) , is rotating more slowly ( being more pressure supported ) and has a more uniform density profile , with a reduced density at the equator : it is gradually approaching spherical symmetry .
as we shall see below , some of these variations in properties are shared by the original adaf solutions if the flow is forced to be more advective ( by reducing the value of @xmath109 ) . as the value of @xmath110
is increased further , another qualitative change in the solutions occurs , with inflow at the pole , outflowing regions away from the pole and still faster inflow at the equator . in other words , the outflow adopts a conical geometry .
the solution with @xmath117 in fig .
[ fig1 ] illustrates this regime . for even larger values of @xmath110 ,
the solution switches to a global inflow pattern ( @xmath118 ) , with increasingly faster inflow in the polar regions , which ultimately overcomes the inflow speed at the equator ( @xmath119 ) . at the end of this sequence ( @xmath120 )
, the flow approaches the non - rotating , uniform density and uniform temperature limit : the 2d solutions with conduction become bondi - like , in a close analogy with what we have seen for 1d solutions with conduction in 3 . in this limit , the value of @xmath121 ( see appendix b ) .
we note that along this entire sequence of solutions , maximum inflow and outflow speed remained everywhere subsonic .
as we shall see below in more detail , most of these solution properties can be understood in terms of the relative contributions of radial and polar conduction terms to the flow energetic balance .
figure [ fig2 ] shows a comparable sequence of latitudinal profiles in solutions with a reduced value of @xmath109 ( @xmath122 ) and the same @xmath111 as before . for small enough values of @xmath110 ( @xmath123 ) , the solutions are again close to the original 2d adaf solutions of narayan & yi ( 1995a ) , with a more spherically symmetric profile than in the @xmath112 case , but still a @xmath34 value which is ( nearly ) zero at the pole . as the value of @xmath110
is increased , a sequence similar to that in fig .
[ fig1 ] obtains , with a few important differences : inflow and outflow velocities , outflow opening angles and the critical @xmath110 value at which the non - rotating limit is reached are all smaller than in the @xmath112 case .
in addition , there is no conical outflow regime along the @xmath110 sequence in this subclass of solutions .
figure [ fig3 ] illustrates in a different way the flow geometry and the relative importance of the rates of mass inflow and outflow , for four of the solutions shown in fig .
[ fig1 ] ( @xmath111 , @xmath112 ) .
isodensity contours are shown by solid lines , while color contours trace the flux of mass per unit solid angle ( @xmath124 ) .
the scale of the cylindrical coordinates ( @xmath125 ) used to label the axes , as well as the isodensity contour levels , are arbitrary because of the radially self - similar nature of the solutions considered .
figure [ fig4 ] summarizes the most significant results for the existence and the properties of outflows that emerged from our systematic survey of the parameter space of hot accretion flow solutions with conduction .
focusing on the specific @xmath126 case , the various curves in fig .
[ fig4 ] cover the @xmath127 space .
the four panels show ( a ) the flow radial velocity at the pole , @xmath128 , in units of the keplerian value , ( b ) the angular velocity at the equator , @xmath129 , in units of the keplerian value , ( c ) the opening angle of outflowing regions , including conical outflow cases , and ( d ) the ratio of outflowing to net inflowing mass rate ( which is also the accretion rate defined in eq .
[ [ eq : mdot ] ] ) . a number of clear trends emerge from this parameter space survey .
solutions with small @xmath109 values ( @xmath130 ) do not exhibit outflows , but only pure inflow regimes .
conversely , solutions with the largest values of @xmath109 are the most efficient at producing outflows .
furthermore , even among the solutions with @xmath131 , there is only a finite range of @xmath110 values for which an outflow regime exists .
the larger @xmath109 is , the larger are the velocity , opening angle and mass flux of the outflow , and so is the maximum value of @xmath132 beyond which the outflow ceases to exist .
we also find that solutions with conical outflows exist only for sufficiently high @xmath109 values .
finally , for large @xmath109 solutions , the outflowing regime persists even to very small values of @xmath133 but outflows become increasingly weaker , slower and narrower .
it should not be a surprise that these various flow properties scale with the ratio @xmath133 , given the form of the energy equation ( eq . [ [ energy ] ] ) .
all these trends are consistent with the detailed results shown in figs .
[ fig1 ] and [ fig2 ] .
the increased efficiency of large @xmath109 solutions at producing outflows can be interpreted as resulting from the larger latitudinal temperature gradients present in these solutions .
this makes the contribution of the polar conduction term more important , relative to the radial conduction term ( which is tied to a fixed radial temperature gradient set by self - similarity ) .
mass outflow rates @xmath134 of the net inflow rate , outflow velocities @xmath135 of the local keplerian velocity and outflow opening angles @xmath136 degs are reached in solutions with @xmath137 and @xmath138 .
for even larger values of @xmath109 , outflow properties continue to follow the trends shown in fig .
[ fig4 ] . for instance
, we have found @xmath139 solutions with mass outflow rates @xmath140 of the net inflow rate , outflow velocities @xmath141 of the local keplerian velocity and outflow opening angles @xmath142 degs .
numerical convergence becomes increasingly difficult at large @xmath109 values , however , and we have not explored this regime as systematically as the low @xmath109 one . two additional features in fig
. [ fig4 ] are worth emphasizing .
all solutions reach the non - rotating limit for values of @xmath143 and @xmath144 .
a derivation of these asymptotic values for the non - rotating limit is provided in appendix b. we also find that there is a wedge in the @xmath132 vs. @xmath109 space , such that solutions do not exist for a growing range of @xmath132 as the value of @xmath109 is increased . despite careful efforts to reach numerical convergence in these parts of the parameter space ,
we have failed to find solutions of sufficient numerical accuracy ( or at all ) in these regions .
we have not clearly identified the origin of this difficulty .
let us stress that all the solutions shown in figures , and otherwise discussed in this work , have reached satisfactory numerical convergence .
we find that the value of the viscosity parameter , @xmath42 , affects quantitatively ( but not qualitatively ) the properties of hot accretion flow solutions with conduction . for larger values of @xmath42 , the radial inflow and outflow speeds
are increased and the overall density scale is reduced , as expected from the original adaf @xmath42 scalings .
figure [ fig23 ] shows a specific comparison of solutions with @xmath137 , @xmath145 and @xmath146 , @xmath147 and @xmath148 .
the outflow speed is much larger in the @xmath149 solution , reaching a value @xmath150 at the pole .
yet , the mass outflow rate ( @xmath151 ) is not much greater in the @xmath152 solution than in the other two solutions ( @xmath153-@xmath154 ) because of the reduced density scale associated with faster radial velocities .
the outflow opening angles are comparable in all three solutions . for consistency and conciseness , we focus on solutions with @xmath111 below . to understand
better the various properties of these hot accretion flows with conduction , let us now focus on the detailed energy and momentum balance that they satisfy .
figure [ fig5 ] shows latitudinal profiles of the advection term ( a ) , on the lhs of the energy equation ( eq . [ [ energy ] ] ) , and of the conduction ( b ) and viscous ( c ) terms , on the rhs of eq .
( [ energy ] ) , for the same specific sequence of solutions as shown in fig .
[ fig1 ] ( @xmath111 , @xmath137 ) . in the solution with negligibly small conduction ( @xmath155 ) ,
the advection and the viscous terms balance each other exactly and heat advection acts to cool the flow ( which corresponds to positive values in panel a ) . in the solution with @xmath156 , however , the magnitude of the conduction term is already very important , being comparable to or larger than the magnitude of the viscous term in most of the flow .
in addition , conduction both cools the flow in the polar regions ( which corresponds to negative values in panel b ) and heats it up in the equatorial regions . as a result of this cooling of the polar regions ( where the contribution from the viscous term is comparatively small , simply by geometry )
, the balance between these various terms permits a negative advection term in the polar regions , which corresponds to advective heating and outflow .
as the value of @xmath110 reaches @xmath157 , the latitudinal profile of temperature starts to flatten significantly ( see fig .
[ fig1 ] ) , thus reducing the contribution of the polar component of conduction .
the radial component , on the other hand , is strictly positive and can only increase in magnitude with the value of @xmath110 ( because of the radial self - similarity of the solutions ) . as a result , the net conduction term switches sign again in the polar regions ( which are now heated by conduction ) , but it is able to remain somewhat negative at intermediate values of the polar angle , @xmath87 .
this situation leads to the conical outflow regime previously identified
. finally , for even larger values of @xmath158 , the radial component of conduction dominates everywhere over the polar component ( which is further reduced by the continued flattening of the temperature profile ; fig .
[ fig1 ] ) , thus forcing the net conduction term to act strictly as a heat source ( as in the 1d solutions of 3 ) .
the energy equation is then satisfied only if advection acts to cool the flow everywhere , which must then be globally inflowing . for the largest values of @xmath110 considered ,
advection very nearly balances conduction as the viscous term drops precipitously and the non - rotating limit is reached . in good analogy with what we have seen for 1d solutions with conduction in 3 , the non - rotating bondi - like limit corresponds to strict equality between conduction and advection . from this examination of energy balance in our radially self - similar solutions ,
it is clear that the key to thermal outflows is the existence of strong enough cooling in the flow for advection to be allowed to act as a heat source .
this is preferentially achieved in polar regions , simply because these are regions where the rate of viscous dissipation must drop by geometrical constraint .
let us comment here on the fact that the advection parameter , @xmath69 , is actually ill - defined in our solutions with conduction . since @xmath69 is associated with the viscous dissipation term in eq .
( [ energy ] ) , it represents a form of cooling which is proportional to the rate of viscous dissipation . as a result , in solutions with an energy balance dominated by conduction ( e.g. solutions with large @xmath110 in fig .
[ fig5 ] ) , a small value of @xmath69 ( @xmath159 ) does not correspond to a globally efficient cooling anymore , as it used to in the original adaf solutions .
this undesirable situation would disappear in 2-temperature solutions with more realistic descriptions of cooling , such as the ones described by narayan & yi ( 1995b ) .
note also that this is not a critical flaw for our analysis : the energetics of all the solutions discussed here depends only on the value of @xmath160 , so that large @xmath109 values can equally well be interpreted as corresponding to low @xmath57 values . for completeness
, we now examine in detail the momentum balance in these same solutions ( @xmath111 , @xmath161 ) .
figure [ fig6 ] shows latitudinal profiles of various terms in the radial and polar components of the momentum equation ( eqs .
[ [ moma ] ] and [ [ momb ] ] , respectively ) .
the magnitude of the various terms on the lhs and the rhs of these equations are labeled with different colors in the panels of fig .
[ fig6 ] , for the same sequence of increasing @xmath110 values as before .
note that we ignore the azimuthal momentum equation , which simply expresses angular momentum conservation . for a negligibly small level of conduction ( @xmath155 ) ,
the @xmath162-momentum balance is dominated by pressure support against gravity in the polar regions , while rotational support becomes gradually dominant in the equatorial regions .
as the value of @xmath110 is increased , the flow becomes hotter , more spherically symmetric and the @xmath162-momentum balance tends to be more dominated by pressure support at all values of the polar angle , @xmath87 .
the inertial and viscous terms are never very important for @xmath162-momentum balance along this @xmath110 sequence . as for the @xmath87-momentum equation , for all values of @xmath110 considered
, it is satisfied through a balance between latitudinal pressure gradients and projected centrifugal force , with little contribution from the viscous terms .
perhaps one of the most interesting property of the profiles shown in fig .
[ fig6 ] is that they show explicitly that there is no distinction , from the point of view of momentum balance , between inflowing and outflowing regions of the flow .
this confirms that outflows , when they exist , are purely determined by energetic considerations .
this further justifies us classifying them as `` thermal '' in origin .
much of the discussion in the literature on the likelihood of outflows from hot accretion flows is articulated around the value of the bernoulli parameter , @xmath163 ( or , more generally , the bernoulli function ) .
narayan & yi ( 1994 ; 1995a ) first emphasized the positivity of @xmath163 in their solutions and raised the possibility that this would lead to bipolar outflows .
furthermore , blandford & begelman ( 1999 ) s motivation and method to develop adios is rooted in the notion that hot accretion flows with positive @xmath163 must have outflows and that their structure and properties should be accordingly modified , leading to a mass accretion rate varying with radius .
the relevance of the value of @xmath163 for the generation of outflows has also been discussed in the context of numerical simulations ( e.g. stone et al .
1999 ) and it has been put into question ( e.g. abramowicz et al . 2000 ) . since our hot accretion flow solutions with conduction produce outflows spontaneously , it is of considerable interest to study the relation between these outflows and the value of @xmath163 in the solutions .
figure [ fig7 ] shows latitudinal profiles of the dimensionless bernoulli parameter , @xmath164 ( same notation as narayan & yi 1995a ) in a set of solutions with @xmath126 , @xmath53 and different values of @xmath69 and @xmath165 .
in addition to @xmath166 , scaled latitudinal profiles of the radial velocity , @xmath34 , are shown , for an easy identification of inflowing vs. outflowing regions .
the nine panels in fig .
[ fig7 ] are organized in such a way that the three rows correspond to @xmath167 , @xmath168 and @xmath169 , from top to bottom , with increasing values of @xmath165 from left to right . from this ensemble of solutions ,
one concludes that the positivity of @xmath166 does not in general discriminate between inflowing and outflowing regions in our solutions .
although the solution corresponding to @xmath170 and @xmath171 would seem to indicate it does , this is in fact a coincidence since we find many counterexamples .
notice in particular in fig .
[ fig7 ] how solutions with @xmath166 changing from positive at the pole to negative at the equator can be alternatively ( nearly ) static , outflowing or inflowing at the pole .
similarly , solutions with globally positive values of @xmath166 can be globally inflowing without any difficulty .
the additional panels in figure [ fig8 ] show two particularly striking examples of the loose relation between the sign of @xmath166 and the inflow / outflow properties of our solutions . in one solution , with @xmath172 ,
the outflow has a conical geometry .
while the value of @xmath166 decreases monotically with @xmath87 , the radial velocity is clearly not monotonic , with two sign changes . in the other solution , with @xmath173 ,
the outflow is polar but there is a small regions at mid - latitudes where outflowing gas has negative @xmath166 value .
although this type of `` violation '' is generally marginal , it clearly demonstrates that there is no strict relation between the sign of @xmath166 and the inflow / outflow properties of gas in our solutions .
it should be noted that for large enough values of @xmath110 , one would expect the bernoulli parameter ( which is an adiabatic , inviscid quantity ) to lose its physical significance , since the flow must then become strongly diabatic .
we suspect that the @xmath174 outflow region shown in the left panel of fig .
[ fig8 ] and the globally inflowing solutions with globally positive @xmath166 values shown in fig .
[ fig7 ] are all manifestations of this loss of significance of @xmath166 in flows with significant conduction .
it is difficult to evaluate , from self - similar solutions alone , what the radiative efficiency of hot accretion flows with conduction may be . in our solutions ,
the flow structure adjusts so that the divergence of the conduction flux generated in the innermost regions is balanced by an increased level of entropy advection .
there will be no extra source of energy available in the presence of conduction ( in fact less is available in our solutions , since the viscous dissipation rate drops with @xmath35 ) , but the dissipated energy is redistributed differently in flows with conduction than in flows without .
this redistribution affects the local flow properties ( density , temperature ) , which in turn affect its cooling properties . using our 1d solution scalings ( 3 ) for simplicity , it is possible to estimate crudely the modified level of emission expected relative to a standard adaf .
free - free emission scales as @xmath175 for a given value of the accretion rate , @xmath176 .
we find that this emission is effectively reduced , e.g. by as much as @xmath177 in the 1d solution of fig .
[ fig : one ] with @xmath178 ( and much more in the @xmath61 solution ) .
even though conduction heats up the gas locally , the reduced density resulting from the larger inflow speed dominates , leading to a net decrease in the expected level of free - free emission .
on the other hand , the very steep dependence of synchrotron emission on the electron temperature ( e.g. mahadevan 1997 ) suggests that hotter solutions ( with conduction ) may be radiatively more efficient from the point of view of that process .
these are only rough scalings , however . in light of several complications expected in global solutions with conduction ( see discussion in 6 ) , it is unlikely that we can determine , from self - similar solutions alone , how the global radiative efficiency of a hot accretion flow may be affected by conduction .
we postpone a thorough discussion of this important issue to future work and focus here instead on the possible relevance of the outflows found in our solutions to a variety of observationally established trends for accreting black holes . in recent years
, much progress has been made in understanding the outflow / jet properties of accretion flows around black holes in x - ray binaries and active galactic nuclei .
various observational studies suggest that , in both classes of systems , there is a regime of steady outflow when black holes accrete well below the eddington rate ( for luminosities @xmath179 ) , while the outflow regime becomes flaring / unsteady at higher accretion rates ( e.g. ho 2002 ; gallo , fender , pooley 2003 ; merloni , heinz & di matteo 2003 ; fender , belloni & gallo 2004 ; nipoti , blundel & binney 2005 ; sikora , stawarz & lasota 2006 ) .
some of the latest results support the notion that black holes accreting at sub - eddington rates do so through an advective form of accretion ( e.g. greene , ho & ulvestad 2005 ; koerding , fender & migliari 2006 ) , but there is some evidence in the `` low / hard '' state of at least two galactic black hole systems ( miller et al .
2006 ) that a standard thin disk remains embedded in the hot flow ( a situation which may favor outflows according to our large @xmath109 solutions ) .
it is tempting to relate these various observational trends with our results on thermal outflows from steady , hot accretion flows with conduction .
the association would be strengthened if one could show that thermal conduction is expected to be equally important in hot accretion flows around stellar - mass and supermassive black holes .
after all , we made the case for weakly - collisional accretion conditions only for systems harboring supermassive black holes in 2 .
while the temperature profile does not depend on black hole mass in self - similar solutions scaled in units of the schwarzschild radius , density profiles do . to address this question , we compare the scalings for the coefficient of turbulent viscosity ( @xmath180 ) and the spitzer coefficient of thermal diffusivity ( @xmath181 ) , using for simplicity the adaf scalings in schwarzschild units of narayan et al .
( 1998b ) .
the comparison of these two diffusivity coefficients is the relevant one since it captures the relative magnitude of `` viscous '' and conduction effects on any scale of interest .
we find that the two coefficients scale identically ( linearly ) with the mass of the central black hole , which suggests that conduction would indeed be equally important in hot accretion flows around stellar - mass and supermassive black holes .
this conclusion is also consistent with the similar scalings with black hole mass of the accretion time and collision times discussed by mahadevan & quataert ( 1997 ) .
it would be interesting to strengthen this comparison by using , as much as possible , observational constraints on hot accretion flows in x - ray binaries , as we have done for agn in 2 . establishing on a firmer ground the relevance of the thermal outflows found in our solutions to observational trends for accreting black holes will clearly require additional work .
in the meantime , we note that our solutions allow a direct scaling of the expected outflow power , per unit accreted mass , which is a quantity of obvious interest .
recalling that the bernoulli constant , @xmath166 , represents the asymptotic value of the energy of gas escaping from the central mass gravity , we can use it as a measure of the outflowing energy rate and compare that to the rate of energy release due to accretion .
the outflow power is simply @xmath182 .
the accretion power is @xmath183 , if we pick a radius of maximum energy release of @xmath184 schwarzschild radii to fix the accretion efficiency to @xmath185 as is usually done . at @xmath186 , @xmath187 . for a ratio of @xmath188 ( solutions with @xmath189 in fig .
[ fig4 ] ) and a corresponding value of @xmath190 for outflowing regions ( fig .
[ fig7 ] ) , we deduce a ratio @xmath191 that reaches up to @xmath192 in our solutions .
therefore , provided that solutions with realistic cooling properties can be constructed with radiative efficiencies @xmath193 ( i.e. less than a hundredth of the fiducial @xmath147 value ) , it may be possible to explain the `` jet - power dominance '' of radiatively - inefficient accretion flows which has been recently emphasized by several authors ( e.g. nagar , falcke & wilson 2005 ; koerding et al .
2006 ; allen et al .
interestingly , values of @xmath194 as we just derived are comparable to the values of @xmath195 of a few percents inferred by allen et al .
( 2006 ) . in this context ,
understanding how thermal outflows may relate to the relativistic jets of accreting black holes will likely be important .
after all , outflow velocities decrease with radius as @xmath196 in our solutions ( by virtue of self - similarity ) , which is a rather unappealing feature .
thermal outflows could be collimated and accelerated by their environment or , alternatively , they could contribute to the collimation and acceleration of narrower and faster jets , such as the ones emerging in relativistic mhd numerical simulations of adiabatic tori ( e.g. devilliers et al . 2005 ; hawley & krolik 2006 ) .
one of the best systems available to us for the study of radiatively - inefficient accretion flows ( and related outflows ) is sgr a*. in recent years , the focus of theoretical work to interpret the rich set of data has in fact shifted to include a jet or outflow component ( e.g. falcke & biermann 1999 ; falcke & markoff 2000 ; melia & falcke 2001 ; markoff et al .
2001 ; yuan , markoff & falcke 2002 ) .
observations of the size ( e.g. bower et al .
2004 ; shen et al .
2005 ; yuan , shen , huang 2006 ) , variability ( e.g. yuan , quataert & narayan 2003 ; yusef - zadeh et al .
2006 ) and polarization ( e.g. agol 2000 ; quataert & gruzinov 2000b ; bower et al . 2003
; marrone et al .
2006 ) of this source all have tremendous potential to further constrain the properties of the hot accretion flow and its outflow , and they may be able to help us discriminate between the thermal outflows seen in our solutions and other outflow alternatives .
direct constraints on outflow geometries for accreting black holes are generally scarce , but it is interesting to note that junor , biretta & livio ( 1999 ) observe an opening angle @xmath197 degs at the base of the radio jet in m87 , in line with the wide outflow geometries found in our solutions . on the other hand ,
other geometries have been proposed for other classes of systems ( e.g. elvis 2000 ) .
the motivation to understand the nature and properties of outflows from accreting black holes goes beyond what we have just outlined , as outflows may be an efficient process by which black holes influence their environments .
recently , this general point has been raised in the specific context of radiatively - inefficient accretion flows by several groups .
soria et al . ( 2006a , b ; see also di matteo , carilli & fabian 2001 ) invoked slow outflows to effect a feedback on the rate of bondi accretion in several low - luminosity agn .
the wide outflow geometries in our solutions may suggest that the conceptual use of a simple bondi spherical inflow geometry to estimate gas capture rates is inappropriate if indeed a large fraction of the black hole capture sphere is the site of gas outflow .
the work of nagar et al .
( 2005 ) and allen et al .
( 2006 ) further suggests that feedback due to outflows associated with radiatively - inefficient accretion flows can reach far beyond the black hole immediate vicinity and affect star formation in the host galaxy , or even the surrounding hot gas when the host is located at the center of a group or a cluster of galaxies
. it will be interesting to develop these arguments , and more detailed feedback scenarios ( e.g. vernaleo & reynolds 2006 ) , in light of what we have learned in our study of thermal outflows .
the possibility that tangled magnetic fields strongly limit the efficiency of thermal conduction in a hot flow constitutes a major theoretical uncertainty for our models . in recent years
, a similar issue has been discussed in the context of galaxy cluster cooling flows .
several studies have concluded that magnetic fields may not limit too strongly the efficiency of thermal conduction ( e.g. narayan & medvedev 2001 ; gruzinov 2002 ; see also chandran & cowley 1998 ; chandran et al .
1999 ) , but this remains somewhat of an open issue . if tangled magnetic fields were to limit the effective mean free paths in the hot flow to @xmath198 , conduction would likely still have important consequences for the flow structure and properties .
if , on the other hand , tangled magnetic fields act to make @xmath199 , a much weaker role for conduction ( if any ) is to be expected .
explicit numerical simulations with anisotropic conduction along field lines in an mhd turbulent medium could greatly help in settling this issue ( see , e.g. , cho & lazarian 2004 ; sharma et al .
it should be emphasized that the energy transport mechanism which permits the existence of outflows in our solutions does not have to be thermal conduction .
any form of heat transport ( e.g. turbulent , as emphasized by balbus 2004 ) acting like thermal conduction would have similar consequences for the flow .
one of several limitations of our self - similar hot accretion flow solutions with conduction is their 1-temperature structure .
narayan & yi ( 1995b ) have shown how crucial it is to account for the 2-temperature structure of the hot flow to obtain solutions with realistic cooling properties . in such solutions ,
the expected decoupling of electron and ion temperatures in the inner regions of the hot flow would modify the role of conduction .
for one thing , ions may be partially insulated from electron conduction .
in addition , self - similarity is broken in the 2-temperature regime and the electron temperature profile flattens to a sub - virial slope ( e.g. narayan et al .
in global solutions , conduction is expected to flatten both the polar and the radial temperature profiles ( while it only affects the polar profile in the radially self - similar solutions discussed here ) . furthermore ,
if a saturated form of conduction is appropriate , the saturated flux should be reduced in regions of the flow where the electrons become relativistic , since their `` free streaming '' velocity then becomes limited by the speed of light .
finally , the break - down of self - similarity implies that the radial contribution to conduction , which acts only to heat up the flow in our solutions , will have to be done at the expense of inner regions that will now be cooling by conduction .
clearly , it will be interesting to see how all these effects combine with each other in global models of hot accretion flows with conduction .
a potentially more fundamental limitation of our work may be the use of simple fluid equations to describe a weakly - collisional plasma obeying a more complex set of equations .
for instance , balbus ( 2001 ) has argued that the dynamical structure of hot flows could be affected by the anisotropic character of conduction in the presence of magnetic fields ( see also parrish & stone 2005 ) , while quataert et al .
( 2002 ) have emphasized additional effects due to the anisotropic pressure tensor present in the weakly - collisional regime ( see also sharma et al .
2006 ) . on the other hand , it is possible that a scale analysis could justify the use of simpler fluid equations to describe the dynamics of the weakly - collisional magnetized plasma on the largest scales of interest in the context of hot accretion flows .
such an analysis would clearly be valuable since essentially all the literature on hot accretion flows relies on collisional fluid equations ( magnetized or not ) . in recent years
, much emphasis has been put on reducing the rate at which gas is accreted by a central black hole , within a hot flow , either by arguing that a negative bernoulli constant is required for an accretion solution with imposed winds to be physically viable ( blandford & begelman 1999 ) or by invoking a gradual reduction in the rate of mass accretion in the flow due to the effects of convection ( narayan et al .
2000 ; quataert & gruzinov 2000a ) .
our results , combined with the possibility that conduction may be important in hot accretion flows , could pose a challenge to both of these scenarios .
we have already seen that the sign of the bernoulli constant does not determine the existence of thermal outflows , or even discriminate between inflowing and outflowing regions , in our solutions with conduction .
our solutions also show that a radially varying @xmath101 is not necessarily associated with the existence of such outflows .
in addition , if conduction is important on all scales of interest in the hot flow , convection may be suppressed altogether when a displaced fluid element efficiently reaches thermal equilibrium with its environment through conduction .
obviously , our results do not exclude the existence of other classes of spontaneously outflowing solutions with radially varying @xmath101 , density profiles deviating from the canonical @xmath200 , more realistic flow patterns allowing for non - zero latitudinal velocities , polar funnels , or a role for other radial and latitudinal energy / momentum transport mechanisms ( see , e.g. , the issues of adiabatic stability discussed in blandford & begelman 2004 , but also balbus 2001 for the diabatic case with anisotropic conduction and magnetic fields ) .
it would be particularly interesting to contrast the study of these generalized solutions with the results for averaged flow properties emerging from simple hydrodynamical and mhd numerical simulations ( e.g. mckinney & gammie 2002 ; proga 2005 ) , in an effort to make the best use of these two complementarity approaches to the black hole accretion problem .
if conduction is generally important in hot accretion flows around black holes , it could have additional interesting consequences besides the ones illustrated in our specific solutions .
for instance , if the hot flow residing inside the capture radius around a black hole is able to heat up the ambient gas outside of that capture radius , the bondi rate of gas capture ( @xmath201 ) may be effectively reduced . while gruzinov ( 1998 ) invoked
`` turbulent '' heat flows to achieve this , the weakly - collisional conditions discussed in 2 for several galactic nuclei suggest that this could be done by `` microscopic '' conduction itself .
more detailed models are therefore needed to address various interesting issues related to conduction , to understand better how conduction may affect the structure and properties of hot accretion flows and to determine how these effects may relate to the observational characteristics of dim accreting black holes .
we are grateful to s. balbus , a. beloborodov , w. dorland , j .-
lasota , r. narayan , e. quataert and e. spiegel for useful discussions and comments .
we also thank m. a. morse for assistance in producing the mass flux color figure .
this research was supported in part by the national science foundation under grant no .
phy99 - 07949 ( at kitp ) .
* appendices *
in 3 , we have seen that a saturated form of conduction scales in such a way that it permits 1d radially self - similar solutions . consequently , we first attempted to find 2d radially self - similar solutions with a saturated conduction flux written as @xmath202 which depends only on the direction of the local temperature gradient , not its magnitude . from taking minus the divergence
, we obtained the following conduction term in the energy equation : @xmath203 \big\},\end{aligned}\ ] ] within the framework of our self - similar solutions .
we have found this prescription to be numerically unstable , or at least difficult to work with .
this may be due to the explicit coupling between the radial and polar conduction terms that appears within this prescription ( which does not seem well suited to our separable , radially self - similar solutions ) .
it may also be caused by the structure of the mathematical operator resulting from this prescription : it no longer acts as a diffusion process .
consequently , in our 2d treatment ( in 4 ) , we returned to a standard diffusive operator for conduction , with a conductivity coefficient scaled with radius to preserve self - similarity .
this results in the following conduction term in the energy equation ( see also eq . [ [ energy ] ] ) : @xmath204.\ ] ] from a comparison of these two prescriptions , it is clear that , in the limit of strong conduction , when latitudinal profiles of most flow quantities become nearly flat , the two conduction parameters , @xmath25 and @xmath205 , are simply related to each by @xmath206 . using the simple scalings of appendix b ,
the relation becomes @xmath207 where we have used @xmath208 , @xmath111 for the last equality .
this correspondence is helpful in showing that , in all our 2d solutions , the non - rotating limit is reached for @xmath209 equivalent values which are @xmath70 , in good agreement with our 1d results ( in 3 ) . in other words , even in 2d solutions , a level of conduction well below that of the ( non - relativistic ) saturated conduction flux is sufficient to alter the structure and properties of hot accretion flows substantially .
as we have seen in 4 , for a given value of the flow parameter @xmath109 , there is a maximum value of the conduction parameter , @xmath110 , at which the solutions reach the non - rotating , bondi - like limit .
no solutions of physical significance exist beyond this maximum value since they would require imaginary values of @xmath35 , as we shall now establish .
let us evaluate eqs .
( [ moma][energy ] ) , at the flow equator , using adequate boundary conditions , in the limit of low viscosity ( @xmath210 ) , small radial velocity ( @xmath211 ) and under the assumption that second order derivatives with respect to @xmath87 become small ( which is generally accurate for solutions approaching the non - rotating limit ) .
we obtain : @xmath212 the first equation establishes that , in the strong conduction limit , @xmath213 .
the second equation yields , in combination with the first , @xmath214 , in the same limit . to be exact ,
this is not the limit @xmath215 , but rather @xmath216 .
the actual non - rotating limit occurs at a slightly lower value of @xmath217 .
the third equation places the explicit constraint on @xmath218 for the non - rotating limit , since the left - hand side must be positive for the value of @xmath35 to be real .
this fixes the maximum value of the conduction parameter @xmath205 to @xmath219 where we have used the crude approximation that momentum ( @xmath220 ) profiles are uniform along @xmath87 , which results in @xmath221 .
this estimate of the maximum value of @xmath205 agrees well with our numerical results ( e.g. in fig
. [ fig4 ] ) .
our results on 2d hot accretion flows with conduction share strong similarities with those of anderson ( 1987 ) .
first , let us mention that we have succeeded in reproducing many of the results reported by this author , using the same prescription for conduction ( i.e. with a conduction coefficient which scales not only with radius , but also with the local flow viscosity and density ) . with this prescription ,
the conduction term in the energy equation becomes : @xmath222.\ ] ] returning to our analysis , there are still differences between our work and that of anderson ( 1987 ) , both in the formalism and in the results .
anderson works in the limit of `` weak shear viscosity , '' with several viscous terms dropped from the dynamical equations ( as compared to ours ) .
he adopts a polytropic equation of state , a prescription for viscosity which differs from shakura - sunyaev and there is no inertial term in his radial momentum equation . as a result
, his set of differential equations is reduced to a third order system , as opposed to a seventh order one in our case . despite these technical differences ,
our results are qualitatively consistent with many of his conclusions .
differences appear mostly at the quantitative level , for instance in the rates of mass outflow ( which appear a few times larger in our solutions ) .
one qualitative feature which has apparently been missed by anderson ( 1987 ) is the existence of solutions with conical outflow geometries in sub - regions of the parameter space . from this comparison
, we can safely conclude that many of the qualitative features exhibited by our solutions should be relatively robust to changes in the various prescriptions adopted ( e.g. conduction and viscosity ) .
abramowicz , m.a .
1988 , apj , 332 , 646 abramowicz , m.a . , lasota , j .- p . & igumenshchev , i.v .
2000 , mnras , 314 , 775 agol , e. 2000 , apj , 538 , l121 anderson , m. 1987 , mnras , 227 , 623 allen , s.w . , dunn , r.j.h . ,
fabian , a.c . ,
taylor , g.b . & reynolds , c.s .
2006 , mnras submitted , astro - ph/0602549 baganoff , f.k .
2003 , apj , 591,891 balbus , s. a. 2001 , apj , 562 , 909 balbus , s. a. 2004 , apj , 600 , 865 blandford , r.d . & begelman , m.c . 1999 ,
mnras , 303 , l1 blandford , r.d . & begelman , m.c .
2004 , mnras , 349 , 68 bower , g. c. et al .
2004 , science , 304 , 704 bower , g. c. , wright , m. c. h. , falcke , h. & backer , d. c. 2003 , apj , 588 , 331 chandran , b. & cowley , s.c . 1998 , phys .
, 80 , 3077 chandran , b. et al . 1999 ,
apj , 525 , 638 cho , j. & lazarian , a. 2004 , j. kor .
soc . , 37 , 557 cowie , l.l . & mckee , c.f .
1977 , apj , 211 , 135 devilliers , j .-
, hawley , j.f . & krolik , j.h .
2003 , apj , 599 , 1238 devilliers , j .-
p . , hawley , j.f . ,
krolik , j.h .
& hirose , s. 2005 , apj , 620 , 878 di matteo , t. et al .
2000 , mnras , 311 , 507 di matteo , t. et al .
2003 , apj , 582 , 133 di matteo , t. , carilli , c. l. & fabian , a. c. 2001 , apj , 547 , 731 elvis , m. 2000 , apj , 545 , 63 esin , a.a .
, mcclintock , j.e .
& narayan , r. 1997 , apj , 489 , 865 esin , a.a .
2001 , apj , 555 , 483 falcke , h. & biermann , p. l. 1999 , a&a , 342 , 49 falcke , h. & markoff , s. 2000 , a&a , 362 , 113 fender , r. p. , belloni , t. m. & gallo , e. 2004 , mnras , 355 , 1105 gallo , e. , fender , r. p. & pooley , g. g. 2003 , mnras , 344 , 60 greene , j. e. , ho , l. c. & ulvestad , j. s. 2006 , apj , 636 , 56 gruzinov , a. 1998 , astro - ph/9809265 gruzinov , a. 2002 , astro - ph/0203031 hawley , j.f . ,
balbus , s.a . & stone , j.m .
2001 , apj , 554 , l49 hawley , j.f .
& krolik , j.h .
2006 , apj , 641 , 103 hameury , j .- m . , menou , k. , dubus , g. , lasota , j .-
p . , & hur , j .- m .
1998 , mnras , 298 , 1048 ho , l.c .
2002 , apj , 564 , 120 ho , l.c . ,
terashima , y. & ulvestad , j.s .
2003 , apj , 589 , 783 honma , f. 1996 , pasj , 48 , 77 ichimaru , s. 1977 , apj , 214 , 840 igumenshchev , i.v . , narayan , r. & abramowicz , m.a .
2003 , apj , 592 , 1042 junor , w. , biretta , j. a. & livio , m. 1999 , nature , 401 , 891 koerding , e. , fender , r. , migliari , s. 2006 , mnras in press , astro - ph/0603731 lasota , j .-
1996 , apj , 462 , 142 loewenstein , m. et al .
2001 , apj , 555 , l21 mahadevan , r. 1997 , apj , 477 , 585 mahadevan , r. & quataert , e. 1997 , apj , 490 , 605 manmoto , t. et al .
2000 , apj , 529 , 127 markoff , s. , falcke , h. , yuan , f. & biermann , p. l. 2001 , a&a , 379 , l13 marrone , d.p . ,
moran , j.m . ,
zhao , j .- h .
& rao , r. 2006 , apj in press , astro - ph/0511653 mckinney , j. c. & gammie , c. f. 2002 , apj , 573 , 728 mckinney , j. c. & gammie , c. f. 2004 , apj , 611 , 977 melia , f. & falcke , h. 2001 , ara&a , 39 , 309 menou , k. et al .
1999 , apj , 520 , 276 merloni , a. , heinz , s. & di matteo , t. 2003 , mnras , 345 , 1057 miller , j. m. et al .
2006 , apj submitted , astro - ph/0602633 nagar , n. m. , falcke , h. & wilson , a. s. 2005 , a&a , 435 , 521 narayan , r. et al .
1998a , apj , 492 , 554 narayan , r. 2003 , in lighthouses of the universe , eds .
m gilfanov , r. sunyaev et al .
, astro - ph/0201260 narayan , r. , igumenshchev , i.v . &
abramowicz , m.a .
2000 , apj , 539 , 798 narayan , r. , mahadevan , r. & quataert , e. 1998b , in theory of black hole accretion disks , eds .
m. abramowicz , g. bjornsson & j. pringle , p. 148
, astro - ph/9803141 narayan , r. & medvedev , m.v .
2001 , apj , 562 , l129 narayan , r. & quataert , e. 2005 , science , 307 , 77 narayan , r. & yi , i. 1994 , apj , 428 , l13 narayan , r. & yi , i. 1995a , apj , 444 , 231 narayan , r. & yi , i. 1995b , apj , 452 , 710 nipoti , c. , blundell , k. m. & binney , j. 2005 , mnras , 361 , 633 press , w.h .
1992 , numerical recipes in fortran : the art of scientific computing ( cambridge : cup ) proga , d. 2005 , apj , 629 , 397 proga , d. & begelman , m.c .
2003 , apj , 594 , 767 quataert , e. & gruzinov , a. 2000a , apj , 539 , 809 quataert , e. & gruzinov , a. 2000b , apj , 545 , 842 rees , m.j .
1982 , nature , 295 , 17 shakura , n.i . & sunyaev , r.a . 1973 ,
a&a , 24 , 377 shapiro , s.l . ,
lightman , a.p . & eardley , d.m .
1976 , apj , 204 , 187 sharma , p. , hammett , g.w . ,
quataert , e. & stone , j.m .
2006 , apj , 637 , 952 shen , z .- q . et al . 2005 , nature , 438 , 62 sikora , m. , stawarz , l. , & lasota j .-
2006 , mnras submitted , astro - ph/0604095 soria , r. et al . 2006a , apj in press , astro - ph/0511293 soria , r. et al .
2006b , apj in press , astro - ph/0511341 stone , j.m . ,
pringle , j.e . & begelman , m.c .
1999 , mnras , 310 , 1002 spitzer , l. , 1962 , physics of fully ionized gases vernaleo , j. c. & reynolds , c. s. 2006 , apj accepted , astro - ph/0511501 yuan , f. , markoff , s. & falcke , h. 2002 , a&a , 383 , 854 yuan , f. , quataert , e. & narayan , r. 2003 , apj , 598 , 301 yuan , f. , shen , z.q . & huang , l. 2006 , apjl in press , astro - ph/0603683 yusef - zadeh , f. et al .
2006 , apj in press , astro - ph/0510787 lccccc + nucleus & @xmath5 & @xmath6 & @xmath223 & @xmath12 & @xmath224 + & ( @xmath225 ) & ( @xmath226 k ) & ( cm ) & & + + + sgr a * & @xmath227 & @xmath228 & @xmath229 & @xmath230 & @xmath230 + ngc 1399 & @xmath231 & @xmath232 & @xmath233 & @xmath234 & @xmath235 + ngc 4472 & @xmath148 & @xmath232 & @xmath236 & @xmath237 & @xmath238 + ngc 4636 & @xmath238 & @xmath239 & @xmath240 & @xmath241 & @xmath242 + m87 & @xmath243 & @xmath232 & @xmath244 & @xmath245 & @xmath235 + m32 & @xmath238 & @xmath230 & @xmath246 & @xmath148 & @xmath247 + | motivated by the low - collisionality of gas accreted onto black holes in sgr a * and other nearby galactic nuclei , we study a family of 2d advective accretion solutions with thermal conduction .
while we only impose global inflow , the accretion flow spontaneously develops bipolar outflows .
the role of conduction is key in providing the extra degree of freedom ( latitudinal energy transport ) necessary to launch these rotating thermal outflows .
the sign of the bernoulli constant does not discriminate between inflowing and outflowing regions .
our parameter survey covers mass outflow rates from @xmath0 to @xmath1 of the net inflow rate , outflow velocities from @xmath0 to @xmath2 of the local keplerian velocity and outflow opening angles from @xmath0 to @xmath3 degs . as the magnitude of conduction
is increased , outflows can adopt a conical geometry , pure inflow solutions emerge , and the limit of 2d non - rotating bondi - like solutions is eventually reached .
these results confirm that radiatively - inefficient , hot accretion flows have a hydrodynamical propensity to generate bipolar thermal outflows . |
dark matter accounts for roughly 80% of the matter in the universe .
dark matter models which attempt to naturally explain this dark matter density generally rely on one of two coincidences . " wimp models ( or wimpless variants @xcite )
rely on the fact that , for certain motivated choices of dark matter particle mass and coupling , the relic density @xcite which one would calculate from thermal freeze - out is approximately that required by observation .
this coincidence is sometimes referred to as the wimp miracle . "
asymmetric dark matter relies on the second coincidence , namely , that the dark matter and baryonic matter densities are similar ( they differ by a factor of @xmath0 ) .
the basic idea of asymmetric dark matter is that the dark matter particle is distinct from the anti - particle , and the current abundance arises almost entirely from one species ( conventionally taken as the particle ) . in this way ,
dark matter is similar to baryonic matter .
a variety of mechanisms have been suggested for generating the asymmetry between baryonic matter and anti - matter .
if the same mechanism also generates the asymmetry between dark matter and dark anti - matter , then the number density of dark matter should be simply related to the baryonic number density .
if the mass of the dark matter particle is also similar to mass of the nucleon ( @xmath1 ) , then the dark matter and baryon energy densities will be similar . for recent reviews of asymmetric dark matter ,
see @xcite .
the two generic features of asymmetric dark matter models are * dark matter annihilation is suppressed because only the particle is abundant in nature , while the anti - particle is not , * the dark matter particle is light , with a mass similar to that of the nucleon ( though there are exceptions @xcite to this result )
. in particular , many asymmetric dark matter models may thus be able address recent hints for low - mass dark matter arising from the dama @xcite , cogent @xcite , cresst @xcite and cdms @xcite experiments . in these proceedings
, we will discuss the theoretical motivations for these features , the classes of models which satisfy the needed criteria , and the impact of these features on asymmetric dark matter constraints and detection possibilities .
[ [ symmetries ] ] symmetries + + + + + + + + + + an asymmetric dark matter candidate must be a particle excitation of a complex field , for example , a complex scalar or a dirac fermion . only in this case is the particle distinct from the anti - particle .
generically , one expects that this should only be the case if the field is charged under an unbroken @xmath2 symmetry ( or a subgroup of @xmath2 which rotates the field by a nontrivial complex phase ) .
the charge of the field under this complexifying " symmetry is what distinguishes the particle from the anti - particle .
another way of seeing this is to note that a complex field will have two degenerate mass eigenstates which correspond the particle and the anti - particle .
the complexifying symmetry forbids any terms in the lagrangian ( such as a majorana mass term ) which could split these eigenstates . in the absence of this symmetry
, one would generically expect the presence of terms in the lagrangian which break the degeneracy , resulting in two non - degenerate mass eigenstates , each of which is its own anti - particle .
if dark matter is truly stable , then generically it should be the lightest particle charged under some unbroken symmetry .
although this stabilizing " symmetry may be the same as the complexifying symmetry described above , it does not have to be .
importantly , the complexifying symmetry can not be a @xmath3 parity , and dark matter need not be the lightest particle charged under it .
dark matter must be the lightest particle charged under the stabilizing symmetry , and that symmetry may be a parity .
[ [ self - annihilation ] ] self - annihilation + + + + + + + + + + + + + + + + + the expectation for asymmetric dark matter is that there is no self - annihilation .
particles can thus only annihilate against anti - particles , which are not abundant in nature .
it is easy to see the origin of this expectation ; a particle / anti - particle system has no conserved charge which could obstruct the annihilation process , while a particle / particle system has a conserved charge provided the particle is charged under a symmetry which is not a parity . since asymmetric dark matter is charged under the complexifying symmetry , which is necessarily not a parity , two asymmetric dark matter particles can not annihilate except to a lighter state with the same charge .
but there is a caveat to this argument . since dark matter
need not be the lightest particle charged under the complexifying symmetry , there is no a priori reason why there should not be a lighter charged state to which two asymmetric dark matter particles could annihilate .
asymmetric dark matter self - annihilation is not forbidden unless the stabilizing symmetry is the same as the complexifying symmetry .
if the stabilizing symmetry is instead a parity , then charge under the stabilizing symmetry is only conserved modulo 2 ; although dark matter decay is forbidden , annihilation would be allowed .
a simple example of this issue , familiar from the standard model , is the proton . if we ignore the internal structure of the proton and treat it as a fundamental particle , then the proton is distinct from the anti - proton because of its charge ( @xmath4 ) under @xmath5 . but this complexifying symmetry is not the stabilizing symmetry , since the proton is not the lightest particle charged under @xmath5 ( this is @xmath6 ) .
the proton is the lightest particle with baryon number , and its annihilation is forbidden if this stabilizing symmetry is also a continuous complexifying symmetry , @xmath7 .
but if baryon and lepton number were just parity symmetries , then the proton would still be stable , but the process @xmath8 would be allowed .
if asymmetric dark matter can self - annihilate with a large enough cross section , then the asymmetry can be washed out . in order for asymmetric dark matter to remain asymmetric ,
its self - annihilation cross section must be small enough to have frozen out in the early universe , resulting in an asymmetry which persists to the current epoch .
such a small self - annihilation cross section could arise if the stabilizing symmetry were a continuous symmetry which was very weakly broken to a parity .
an alternative approach would be for the complexifying symmetry to be very weakly broken , thus implying that the dark matter was almost " complex @xcite .
[ [ particle - mass ] ] particle mass + + + + + + + + + + + + + although asymmetric dark matter does not in principle require a connection between the dark matter and baryon asymmetries , much of the motivation is lost in the absence of this connection
. we will thus focus on models for which these asymmetries are related . for such models ,
it is worth noting that asymmetric dark matter really requires two coincidences .
in addition to a mechanism connecting the net dark matter number density to the net baryon number density , one requires a mechanism for relating the dark matter particle mass to the mass of the lightest baryon ( @xmath1 ) .
for a model of asymmetric dark matter , it is not enough for the particle and anti - particle to be distinguishable ; one must also have mechanism for generating the dark matter asymmetry .
as with the baryon asymmetry , this amounts to satisfying the sakharov conditions .
there are a vast array of asymmetric dark matter models @xcite which satisfy these conditions in a variety of ways .
one loose way of classifying asymmetric dark matter models is by the method for relating the dark matter asymmetry to the baryon asymmetry .
either the dark matter asymmetry could be generated first and then transferred to baryons in some way , or vice versa , or both asymmetries could be generated at the same time by the same mechanism .
these classifications are of course somewhat ambiguous , since the definition of dark sector vs. visible sector is sometimes just a matter of taste .
one can write a general template for creating a model of asymmetric dark matter as follows : * _ pick a class : _ decide if an asymmetry is first generated in the dark sector , in the standard model sector , or in both simultaneously ; * _ pick a generation mechanism : _ choose a mechanism for satisfying the sakharov conditions and generating the initial asymmetry in the chosen sector ; * _ pick a transfer mechanism : _ choose a mechanism for transferring the asymmetry , if needed , from the sector where it was generated to the other sector ; * _ pick a mass : _ find a mechanism for setting the dark matter particle mass so as to generate the correct dark matter density . for each of these steps ,
a variety of possibilities have been studied .
regardless of which sector contains the initial asymmetry , the generation of this asymmetry requires that the sakharov conditions be satisfied .
mechanisms for generating the initial asymmetry are thus typically generalizations of standard mechanisms for baryogenesis / leptogenesis , and include strongly first - order phase transitions , out - of - equilibrium annihilation or decay , the affleck - dine mechanism @xcite , etc .
the asymmetry can then be transferred from one sector to another by sphalerons , annihilations , co - annihilations , decays , etc .
if the asymmetry is being transferred to the standard model sector , then this mechanism must violate @xmath9 or @xmath10 .
if it is transferred to the dark sector , then it must violate @xmath11 , where @xmath11 is a dark sector quantum number which counts the asymmetry between dark particles and dark anti - particles .
but this transfer mechanism need not violate @xmath12 or include an intrinsic departure from thermal equilibrium , since these conditions have already been achieved with the generation of the initial asymmetry .
finally , there are some models for which the dark matter mass can naturally be tied to the gev scale , thus explaining the relic density . for example , for models of mirror matter @xcite , the dark sector is a mirror copy of the standard model sector . in this case , the dark matter candidate is a mirror baryon , and its mass is automatically of the same scale as the standard model baryons . in models
where asymmetric dark matter is generated from hidden sector baryogenesis @xcite , the dark matter candidate is chirally charged under a symmetry @xmath13 , under which the right - handed @xmath14- and @xmath15-quarks and @xmath16-lepton are also charged .
since all of these fields are chiral under @xmath13 , all of their masses are proportional to the symmetry - breaking scale of @xmath13 .
the mass of @xmath14 , @xmath15 and @xmath16 are all @xmath17 , implying that the symmetry - breaking scale of @xmath13 , and thus the mass of the dark matter , are similar .
from the point of view of detection , the main constraints on asymmetric dark matter arise from its relatively low mass and small self - annihilation cross section .
we can summarize the difference between asymmetric dark matter and wimps , in this context , as follows : * _ direct detection : _ similar to low - mass wimp searches .
current sensitivity is greatly degraded for the lighter asymmetric dark matter candidates ; * _ indirect detection : _ no signals expected .
a self - annihilation cross section large enough to be detected would erase the dark matter asymmetry ; * _ collider searches : _ similar to low - mass wimp searches .
sensitivity improves as the dark matter mass decreases ; * _ astrophysical constraints : _ asymmetric dark matter which is captured in large astrophysical objects can collapse to form a black hole .
constraints can be placed on the dark matter capture rate by the observation of old neutron stars which have not collapsed .
an interesting new feature arising for asymmetric dark matter is the possibility of constraining dark matter - nucleon interactions based on the absence of black hole formation within large astrophysical bodies .
the basic idea is a variant of the dark matter search strategy used by neutrino detectors .
when a dark matter particle scatters off a nucleus in any dense astrophysical body ( such as the sun ) , it will lose kinetic energy to the nucleus elastic recoil .
if the dark matter particle velocity falls below the escape velocity of the astrophysical body , it will be gravitationally captured .
after many orbits the dark matter particle will have scattered enough to thermalize , and the dark matter will collect in a dense region near the core of the astrophysical body .
neutrino detectors search for the neutrinos which are emitted from the sun or earth when dark matter in this dense region annihilates .
but what if dark matter annihilation were highly suppressed , as in the case of asymmetric dark matter ? in this case , dark matter would keep collecting within the astrophysical body , with no mechanism for depleting the dark matter .
if enough dark matter is captured , the dark matter would become gravitationally unstable to collapse into a black hole .
if this black hole grows , it could then destroy the astrophysical body .
observations of astrophysical bodies which have not been destroyed thus place a bound on the dark matter capture rate , and in turn on the dark matter - nucleon scattering cross section .
several authors have considered this class of constraints @xcite for neutron stars in globular clusters .
the tightest constraints arise for bosonic asymmetric dark matter , for which there is no fermi degeneracy pressure to obstruct dark matter collapse .
we can briefly summarize the path of this type of analysis : 1 .
_ dark matter accumulates : _ the accumulation rate depends on @xmath18 and @xmath19 , the dark matter - neutron scattering cross section and the thermally - averaged dark matter self - annihilation cross section , respectively .
dark matter thermalizes : _ if the lifetime of the neutron star is not sufficient for thermalization , then there is no constraint .
dark matter forms a bose - einstein condensate : _ for the relevant parameter space , when sufficient dark matter collects at the neutron star core , it will form a bose - einstein condensate ( bec ) .
dark matter in the bec phase crosses the bosonic chandrasekhar bound : _ this bound depends on the dark matter mass , @xmath20 , and the strength of self - interactions .
once dark matter crosses this bound , a black hole will form .
the black hole evolves : _ a large black hole will grow by accreting baryonic and dark matter .
a small black hole will quickly evaporate away .
if the black hole will grow large enough to consume the astrophysical body within its lifetime , the point in parameter - space would be ruled out by observation .
we consider an asymmetric dark matter field @xmath21 with self - annihilation cross section @xmath22 and a repulsive @xmath23 self - interaction .
this interaction is not generically forbidden by any symmetry of the theory , so one does not expect @xmath24 to be particularly small @xcite .
we will take the dark matter density within a globular cluster , @xmath25 , to be @xmath26 , as a benchmark
. the actual density may be much smaller , but this would just result in a rescaling of the constraint on @xmath18 .
we can express the chandrasekhar bound for self - interacting bosonic matter as @xcite n_chand & & 2m_pl^2 m_x^2 ( 1 + 32 m_pl^2 m_x^2 ) ^12 .
a black hole will form when the number of particles in the bec phase exceeds @xmath27 .
if @xmath24 is extremely small , then @xmath28 , but for natural values of @xmath24 we instead find @xmath29 , similar to the fermionic chandrasekhar bound .
@xmath30 , the number of dark matter particles accumulated within the core of the neutron star , is given by n_acc & ~ & where @xmath31 is the volume of the thermalization region , @xmath32 is the neutron star lifetime and @xmath33 is the capture rate .
the capture rate is proportional to @xmath34 @xcite provided @xmath35 .
a bound on @xmath30 can thus be rephrased as a bound on @xmath18 , which can be compared to results from direct detection experiments .
but it is important to note that , for @xmath36 , all dark matter particles which reach the neutron star scatter against it ; as a result , increasing @xmath18 can not increase the capture rate any further .
thus , if dark matter is not excluded for @xmath37 , then there exists no value of @xmath18 for which it is excluded . in figure
[ fig : neutronstarbounds ] @xcite , we plot bounds on @xmath18 arising from observations of old neutron stars in globular clusters for a variety of choices for @xmath19 and @xmath24 .
one can see that as @xmath20 increases , constraints on @xmath18 initially become tighter because fewer dark matter particles are needed for black hole formation .
but for large enough @xmath20 , constraints on @xmath18 become very weak because the black hole which is formed is so small that it evaporates away unless the black hole can capture dark matter rapidly enough to feed " the black hole and keep it growing . for this reason
, an extremely small self - interaction coupling @xmath24 can cause these constraints to be slightly tighter ; self - interaction causes a the formation of a larger black hole , which can then grow to consume the neutron star .
but for even a moderate value of @xmath24 , a black hole will never form . similarly , for @xmath38 , dark matter depletion will be rapid enough that a black hole will never form .
thus , even a self - annihilation cross section which is so small as to be unobservable at any current or anticipated experiments would completely eliminate any constraints from neutron star observations .
plane , arising from observations of old neutron stars in globular clusters .
we assume a globular cluster density @xmath39 .
the red , green , blue , and purple contours ( from left to right ) denote @xmath40 , respectively .
solid , dotted , dashed , and dot - dashed contours denote self - annihilation cross sections @xmath41 , respectively . in the gray region ,
dark matter does not thermalize within the lifetime of the neutron star . ]
we see that these bounds can be tightly constraining for the case of bosonic dark matter , especially in the low - mass region which would be relevant for the data of dama , cogent , cresst and cdms .
but for this to be the case , the stabilizing symmetry must also be a complexifying symmetry ( thus forbidding self - annihilation ) ; if it is broken to a parity , it must be broken extremely weakly .
likewise , these constraints are only relevant if self - interactions are either non - existent or attractive .
if we had instead chosen @xmath42 , the constraint on @xmath18 would be weakened by a factor @xmath43 .
interestingly , asymmetric dark matter can also be constrained by its effect on stellar evolution @xcite .
in particular , a large density of asymmetric dark matter in the core of the sun can lead to a change in the expected solar neutrino flux which differs from observation .
since the ambient dark matter density near the solar system is much less uncertain than in globular clusters , this type of analysis of solar evolution provides a nice complement to bounds based on old neutron stars in globular clusters . for @xmath44 ,
current direct detection experiments tend to rapidly lose sensitivity .
for such models collider - based searches may provide the best possibility for detection .
colliders can search for direct dark matter pair production , through the same contact operator which mediates dark matter - nucleon scattering interactions @xcite .
these mono - anything " searches will produce missing transverse momentum from the dark matter pair , along with some radiated standard model particles which recoil against the dark matter pair .
alternatively , colliders can produce heavy exotic particles which are charged both under @xmath45 and under the dark matter stabilizing symmetry .
the cascade decay of these heavy exotic particles will produce standard model jets , plus missing transverse momentum .
an advantage of direct collider searches is that they probe the same effective contact operators which mediate direct detection , thus potentially allowing one to correlate data from the two detection strategies .
an advantage of cascade searches , however , is that they are based on the production of qcd - coupled particles , a task at which hadron colliders excel .
but in the context of asymmetric dark matter , an advantage of both search strategies is that they increase in sensitivity for low - mass dark matter , since it is easier to produce low - mass dark matter at a collider . for asymmetric dark matter at the lower end of the expected mass range
, colliders may thus provide the best sensitivity .
future direct detection experiments may also begin to become sensitive to this region of parameter space , but only for detectors using lighter target materials ( to maximize the recoil energy arising from the scattering of a low - mass dark matter particle ) and for detectors with lower recoil energy thresholds .
asymmetric dark matter in the @xmath46 range could potentially explain the data from dama , cogent , cresst and cdms .
these low - mass signals are in some tension with bounds from xenon100 @xcite , and there has much recent work on potential resolutions of this tension arising from deviations from typical assumptions about dark matter interactions and astrophysical distributions @xcite .
most of these results also hold for asymmetric dark matter models .
asymmetric dark matter models which are relevant for the low - mass direct detection signals are tightly constrained by lhc monojet searches @xcite .
however , the sensitivity of such searches depends in detail on the spin of the dark matter particle , the choice of effective operator , as well as the flavor structure of the quark couplings .
different choices can lead to a dramatic weakening of these constraints , allowing consistency between the low - mass direct detection signals and lhc bounds @xcite .
moreover , if the particle mediating the interaction has a relatively small mass ( @xmath47 ) , then the scattering interaction may still be short - ranged , while the effective operator approximation will not be valid for dark matter production at the lhc . for such models , monojet signals may be significantly suppressed , though it may be possible to directly produce the mediating particle at colliders @xcite .
although asymmetric dark matter may potentially explain the low - mass direct detection signals , it can not explain the potential gamma - ray excess from the galactic center @xcite .
although this excess could be consistent with low - mass dark matter , an annihilation cross section large enough to produce these signals would erase any dark matter asymmetry .
it is interesting to consider how one may distinguish asymmetric dark matter from a more standard wimp candidate . as we have seen , the main features one would expect from asymmetric dark matter are a low - mass particle with no indirect detection signatures .
however , such a signature could also be reproduced by a particle with @xmath48-wave suppressed annihilation . in this case
, a distinguishing signature may arise from dark matter searches at neutrino detectors .
neutrino detectors search for the neutrinos which arise when dark matter annihilates in the core of the sun .
if the sun is in equilibrium , then the rate at which dark matter is captured is the same as the rate at which it is annihilated .
the key point here is that , if the sun is in equilibrium , then the annihilation rate is independent of the annihilation cross section because the annihilation rate is equal to the capture rate , which is determined by the scattering cross section . in particular
, a very small annihilation cross section would imply a very large equilibrium density . as a result ,
even dark matter with a very small annihilation cross section can still yield a detectable neutrino rate , provided the dark matter is in equilibrium
. if dark matter has a small mass , the ambient number density will be very large . as a result ,
dark matter with a scattering cross section large enough to explain the low mass direct detection data could be in equilibrium even if @xmath49 @xcite . as a result ,
low - mass dark matter with a largely @xmath48-wave annihilation cross section and very small @xmath50-wave contribution ( perhaps arising from chirality - suppressed annihilation to @xmath14-quarks or @xmath16-leptons ) can still yield a large event rate at neutrino detectors . on the other hand ,
if low - mass dark matter is found using direct and collider search strategies , but low - energy events are not seen at neutrino detectors , this implies that the annihilation cross section is indeed very small .
while by no means determinative , this may perhaps provide evidence suggesting that dark matter is asymmetric , with an annihilation cross section which is more heavily suppressed than @xmath48-wave / chirality suppression .
we are grateful to j. bramante , k. fukushima and e. stopnitzky for collaboration .
we are also grateful to the organizers of ppc 2013 .
our attendance at ppc 2013 was supported in part by department of energy grant de - fg02 - 13er41913 .
j. l. feng and j. kumar , phys .
lett . * 101 * , 231301 ( 2008 ) [ arxiv:0803.4196 [ hep - ph ] ] ; j. l. feng , j. kumar and l. e. strigari , phys .
b * 670 * , 37 ( 2008 ) [ arxiv:0806.3746 [ hep - ph ] ] ; j. l. feng , h. tu and h. -b .
* 0810 * , 043 ( 2008 ) [ arxiv:0808.2318 [ hep - ph ] ] . ya . b. zeldovich , adv .
astrophys . * 3 * , 241 ( 1965 ) ; h.y .
chiu , phys .
* 17 * , 712 ( 1966 ) ; g. steigman , ann .
nucl . part .
* 29 * , 313 ( 1979 ) ; r.j .
scherrer and m.s .
turner , phys .
d * 33 * , 1585 ( 1986 ) .
k. petraki and r. r. volkas , int .
j. mod .
a * 28 * , 1330028 ( 2013 ) [ arxiv:1305.4939 [ hep - ph ] ] .
k. m. zurek , arxiv:1308.0338 [ hep - ph ] . m. r. buckley and l. randall , jhep * 1109
* , 009 ( 2011 ) [ arxiv:1009.0270 [ hep - ph ] ] .
r. bernabei _ et al . _
[ dama and libra collaborations ] , eur . phys .
j. c * 67 * , 39 ( 2010 ) [ arxiv:1002.1028 [ astro-ph.ga ] ] .
c. e. aalseth _
[ cogent collaboration ] , phys .
lett . * 106 * , 131301 ( 2011 ) [ arxiv:1002.4703 [ astro-ph.co ] ] . c. e. aalseth _
_ [ cogent collaboration ] , phys .
d * 88 * , 012002 ( 2013 ) [ arxiv:1208.5737 [ astro-ph.co ] ] . g. angloher , m. bauer , i. bavykina , a. bento , c. bucci , c. ciemniak , g. deuter and f. von feilitzsch _
et al . _ ,
j. c * 72 * , 1971 ( 2012 ) [ arxiv:1109.0702 [ astro-ph.co ] ] .
r. agnese _ et al . _
[ cdms collaboration ] , [ arxiv:1304.4279 [ hep - ex ] ] .
s. tulin , h. -b .
yu and k. m. zurek , jcap * 1205 * , 013 ( 2012 ) [ arxiv:1202.0283 [ hep - ph ] ] .
see , for example , s. nussinov , phys .
b * 165 * , 55 ( 1985 ) ; d. hooper , j. march - russell and s. m. west , phys .
b * 605 * , 228 ( 2005 ) [ hep - ph/0410114 ] ; s. b. gudnason , c. kouvaris and f. sannino , phys .
d * 73 * , 115003 ( 2006 ) [ hep - ph/0603014 ] ; m. y. .khlopov and c. kouvaris , phys .
d * 77 * , 065002 ( 2008 ) [ arxiv:0710.2189 [ astro - ph ] ] ; r. foadi , m. t. frandsen and f. sannino , phys .
d * 80 * , 037702 ( 2009 ) [ arxiv:0812.3406 [ hep - ph ] ] ; d. e. kaplan , m. a. luty and k. m. zurek , phys .
d * 79 * , 115016 ( 2009 ) [ arxiv:0901.4117 [ hep - ph ] ] ; t. cohen , d. j. phalen , a. pierce and k. m. zurek , phys .
d * 82 * , 056001 ( 2010 ) [ arxiv:1005.1655 [ hep - ph ] ] ; j. shelton and k. m. zurek , phys .
d * 82 * , 123512 ( 2010 ) [ arxiv:1008.1997 [ hep - ph ] ] ; h. davoudiasl , d. e. morrissey , k. sigurdson and s. tulin , phys .
lett . * 105 * , 211304 ( 2010 ) [ arxiv:1008.2399 [ hep - ph ] ] ; n. haba and s. matsumoto , prog .
phys . *
125 * , 1311 ( 2011 ) [ arxiv:1008.2487 [ hep - ph ] ] ; p. -h .
gu , m. lindner , u. sarkar and x. zhang , phys .
d * 83 * , 055008 ( 2011 ) [ arxiv:1009.2690 [ hep - ph ] ] ; m. blennow , b. dasgupta , e. fernandez - martinez and n. rius , jhep * 1103 * , 014 ( 2011 ) [ arxiv:1009.3159 [ hep - ph ] ] ; l. j. hall , j. march - russell and s. m. west , arxiv:1010.0245 [ hep - ph ] ; a. falkowski , j. t. ruderman and t. volansky , jhep * 1105 * , 106 ( 2011 ) [ arxiv:1101.4936 [ hep - ph ] ] ; m. l. graesser , i. m. shoemaker and l. vecchi , jhep * 1110 * , 110 ( 2011 ) [ arxiv:1103.2771 [ hep - ph ] ] ; m. t. frandsen , s. sarkar and k. schmidt - hoberg , phys .
d * 84 * , 051703 ( 2011 ) [ arxiv:1103.4350 [ hep - ph ] ] ; m. r. buckley , phys . rev .
d * 84 * , 043510 ( 2011 ) [ arxiv:1104.1429 [ hep - ph ] ] ; h. iminniyaz , m. drees and x. chen , jcap * 1107 * , 003 ( 2011 ) [ arxiv:1104.5548 [ hep - ph ] ] ; d. e. kaplan , g. z. krnjaic , k. r. rehermann and c. m. wells , jcap * 1110 * , 011 ( 2011 ) [ arxiv:1105.2073 [ hep - ph ] ] ; n. f. bell , k. petraki , i. m. shoemaker and r. r. volkas , phys .
d * 84 * , 123505 ( 2011 ) [ arxiv:1105.3730 [ hep - ph ] ] ; c. cheung and k. m. zurek , phys .
d * 84 * , 035007 ( 2011 ) [ arxiv:1105.4612 [ hep - ph ] ] ; e. del nobile , c. kouvaris and f. sannino , phys .
d * 84 * , 027301 ( 2011 ) [ arxiv:1105.5431 [ hep - ph ] ] ; j. march - russell and m. mccullough , jcap * 1203 * , 019 ( 2012 ) [ arxiv:1106.4319 [ hep - ph ] ] ; h. davoudiasl , d. e. morrissey , k. sigurdson and s. tulin , phys .
d * 84 * , 096008 ( 2011 ) [ arxiv:1106.4320 [ hep - ph ] ] ; y. cui , l. randall and b. shuve , jhep * 1108 * , 073 ( 2011 ) [ arxiv:1106.4834 [ hep - ph ] ] ; p. kumar and e. ponton , jhep * 1111 * , 037 ( 2011 ) [ arxiv:1107.1719 [ hep - ph ] ] ; nucl .
b * 854 * , 666 ( 2012 ) [ arxiv:1108.3967 [ hep - ph ] ] ; m. r. buckley and s. profumo , phys . rev .
lett . * 108 * , 011301 ( 2012 ) [ arxiv:1109.2164 [ hep - ph ] ] ; m. cirelli , p. panci , g. servant and g. zaharijas , jcap * 1203 * ( 2012 ) 015 [ arxiv:1110.3809 [ hep - ph ] ] ; t. lin , h. -b .
yu and k. m. zurek , phys .
d * 85 * , 063503 ( 2012 ) [ arxiv:1111.0293 [ hep - ph ] ] ; b. von harling , k. petraki and r. r. volkas , jcap * 1205 * , 021 ( 2012 ) [ arxiv:1201.2200 [ hep - ph ] ] ; s. tulin , h. -b .
yu and k. m. zurek , jcap * 1205 * , 013 ( 2012 ) [ arxiv:1202.0283 [ hep - ph ] ] ; j. march - russell , j. unwin and s. m. west , jhep * 1208 * , 029 ( 2012 ) [ arxiv:1203.4854 [ hep - ph ] ] ; c. arina , j. -o .
gong and n. sahu , nucl .
b * 865 * , 430 ( 2012 ) [ arxiv:1206.0009 [ hep - ph ] ] ; c. arina , r. n. mohapatra and n. sahu , phys .
b * 720 * , 130 ( 2013 ) [ arxiv:1211.0435 [ hep - ph ] ] .
see , for example , t. d. lee and c. -n .
yang , phys .
rev . * 104 * , 254 ( 1956 ) ; i. kobzarev , l. okun , and i. pomeranchuk , sov.j.nucl.phys .
* 3 * , 837 ( 1966 ) ; s. i. blinnikov and m. y. .khlopov , sov .
j. nucl .
phys . * 36 * , 472 ( 1982 ) [ yad .
fiz . * 36 * , 809 ( 1982 ) ] ; s. i. blinnikov and m. khlopov , sov .
astron .
* 27 * , 371 ( 1983 ) [ astron .
60 * , 632 ( 1983 ) ] ; r. foot , h. lew and r. r. volkas , phys .
b * 272 * , 67 ( 1991 ) ; r. foot , h. lew and r. r. volkas , mod .
phys . lett .
a * 7 * , 2567 ( 1992 ) ; r. foot and r. r. volkas , phys .
d * 52 * , 6595 ( 1995 ) [ hep - ph/9505359 ] ; z. berezhiani , d. comelli and f. l. villante , phys .
b * 503 * , 362 ( 2001 ) [ hep - ph/0008105 ] ; m. pavsic , int .
j. theor .
phys . * 9 * , 229 ( 1974 ) [ hep - ph/0105344 ] ; a. y. .ignatiev and r. r. volkas , phys .
d * 68 * , 023518 ( 2003 ) [ hep - ph/0304260 ] ; r. foot and r. r. volkas , phys .
d * 68 * , 021304 ( 2003 ) [ hep - ph/0304261 ] ; z. berezhiani , p. ciarcelluti , d. comelli and f. l. villante , int .
j. mod .
d * 14 * , 107 ( 2005 ) [ astro - ph/0312605 ] ; r. foot and r. r. volkas , phys . rev .
d * 69 * , 123510 ( 2004 ) [ hep - ph/0402267 ] ; p. ciarcelluti , int .
j. mod .
d * 14 * , 187 ( 2005 ) [ astro - ph/0409630 ] ; p. ciarcelluti , int . j. mod .
d * 14 * , 223 ( 2005 ) [ astro - ph/0409633 ] ; z. berezhiani , s. cassisi , p. ciarcelluti and a. pietrinferni , astropart .
* 24 * , 495 ( 2006 ) [ astro - ph/0507153 ] ; z. berezhiani , eur .
j. st * 163 * , 271 ( 2008 ) ; h. an , s. -l .
chen , r. n. mohapatra and y. zhang , jhep * 1003 * , 124 ( 2010 ) [ arxiv:0911.4463 [ hep - ph ] ] ; r. foot , arxiv:1305.4316 [ astro-ph.co ] . b. dutta and j. kumar , phys . lett .
b * 699 * , 364 ( 2011 ) [ arxiv:1012.1341 [ hep - ph ] ] .
i. affleck and m. dine , nucl .
b * 249 * , 361 ( 1985 ) .
b. dutta and j. kumar , phys .
b * 643 * , 284 ( 2006 ) [ hep - th/0608188 ] .
w. h. press and d. n. spergel , astrophys .
j. * 296 * , 679 ( 1985 ) ; i. goldman and s. nussinov , phys .
d * 40 * , 3221 ( 1989 ) ; a. de lavallaz and m. fairbairn , phys .
d * 81 * , 123521 ( 2010 ) [ arxiv:1004.0629 [ astro-ph.ga ] ] ; c. kouvaris and p. tinyakov , phys .
lett . * 107 * , 091301 ( 2011 ) [ arxiv:1104.0382 [ astro-ph.co ] ] ; c. kouvaris and p. tinyakov , phys . rev .
d * 83 * , 083512 ( 2011 ) [ arxiv:1012.2039 [ astro-ph.he ] ] ; s. d. mcdermott , h. -b .
yu and k. m. zurek , phys .
d * 85 * , 023519 ( 2012 ) [ arxiv:1103.5472 [ hep - ph ] ] ; g. bertone and m. fairbairn , phys .
d * 77 * , 043515 ( 2008 ) [ arxiv:0709.1485 [ astro - ph ] ] ; m. mccullough and m. fairbairn , phys .
d * 81 * , 083520 ( 2010 ) [ arxiv:1001.2737 [ hep - ph ] ] .
j. bramante , k. fukushima and j. kumar , phys .
d * 87 * , 055012 ( 2013 ) [ arxiv:1301.0036 [ hep - ph ] ] .
n. f. bell , a. melatos and k. petraki , phys .
d * 87 * , 123507 ( 2013 ) [ arxiv:1301.6811 [ hep - ph ] ] .
p. jetzer , phys .
rept . * 220 * , 163 ( 1992 ) .
e. w. mielke and f. e. schunck , gr - qc/9801063 .
e. w. mielke and f. e. schunck , nucl .
b * 564 * , 185 ( 2000 ) [ gr - qc/0001061 ] .
a. gould , astrophys .
j. * 321 * , 571 ( 1987 ) .
p. scott , m. fairbairn and j. edsjo , monthly notices of the royal astronomical society , 394 , 82 ( 2009 ) ; d. t. cumberbatch , j. a. guzik , j. silk , l. s. watson and s. m. west , phys .
d * 82 * , 103503 ( 2010 ) ; m. taoso , f. iocco , g. meynet , g. bertone and p. eggenberger , phys . rev .
d * 82 * , 83509 ( 2010 ) ; j. casanellas and i. lopes , astrophys . j. * 705 * , 135 ( 2009 ) ; i. lopes , j. casanellas and d. eugenio , phys .
d * 83 * , 63521 ( 2011 ) ; i. lopes and j. silk , astrophys .
j. * 757 * , 130 ( 2012 ) [ arxiv:1209.3631 [ astro-ph.sr ] ] . j. l. feng , s. su and f. takayama , phys .
lett . * 96 * , 151802 ( 2006 ) [ hep - ph/0503117 ] ; m. beltran , d. hooper , e. w. kolb , z. a. c. krusberg and t. m. p. tait , jhep * 1009 * , 037 ( 2010 ) [ arxiv:1002.4137 [ hep - ph ] ] ; j. goodman , m. ibe , a. rajaraman , w. shepherd , t. m. p. tait and h. -b .
yu , phys .
b * 695 * , 185 ( 2011 ) [ arxiv:1005.1286 [ hep - ph ] ] ; j. goodman , m. ibe , a. rajaraman , w. shepherd , t. m. p. tait and h. -b .
yu , phys .
d * 82 * , 116010 ( 2010 ) [ arxiv:1008.1783 [ hep - ph ] ] ; a. rajaraman , w. shepherd , t. m. p. tait and a. m. wijangco , phys .
d * 84 * , 095013 ( 2011 ) [ arxiv:1108.1196 [ hep - ph ] ] ; p. j. fox , r. harnik , j. kopp and y. tsai , phys .
d * 85 * , 056011 ( 2012 ) [ arxiv:1109.4398 [ hep - ph ] ] ; p. j. fox , r. harnik , r. primulando and c. -t .
yu , phys .
d * 86 * , 015010 ( 2012 ) [ arxiv:1203.1662 [ hep - ph ] ] ; y. bai and t. m. p. tait , phys .
b * 723 * , 384 ( 2013 ) [ arxiv:1208.4361 [ hep - ph ] ] ; u. haisch , f. kahlhoefer and j. unwin , jhep * 1307 * , 125 ( 2013 ) [ arxiv:1208.4605 [ hep - ph ] ] ; p. j. fox and c. williams , phys .
d * 87 * , 054030 ( 2013 ) [ arxiv:1211.6390 [ hep - ph ] ] ; l. m. carpenter , a. nelson , c. shimmin , t. m. p. tait and d. whiteson , arxiv:1212.3352 [ hep - ex ] ; t. lin , e. w. kolb and l. -t .
wang , arxiv:1303.6638 [ hep - ph ] .
j. kumar , d. sanford and l. e. strigari , phys .
d * 85 * , 081301 ( 2012 ) [ arxiv:1112.4849 [ astro-ph.co ] ] ; s. chatrchyan _ et al .
_ [ cms collaboration ] , jhep * 1209 * , 094 ( 2012 ) [ arxiv:1206.5663 [ hep - ex ] ] ; g. aad _ et al .
_ [ atlas collaboration ] , jhep * 1304 * , 075 ( 2013 ) [ arxiv:1210.4491 [ hep - ex ] ] .
e. aprile _ et al . _ [ xenon100 collaboration ] , phys .
lett . * 107 * , 131302 ( 2011 ) [ arxiv:1104.2549 [ astro-ph.co ] ] ; e. aprile _ et al . _
[ xenon100 collaboration ] , phys .
lett . * 109 * , 181301 ( 2012 ) [ arxiv:1207.5988 [ astro-ph.co ] ] .
m. t. frandsen , f. kahlhoefer , c. mccabe , s. sarkar and k. schmidt - hoberg , jcap * 1307 * , 023 ( 2013 ) [ arxiv:1304.6066 [ hep - ph ] ] ; y. -y .
mao , l. e. strigari and r. h. wechsler , arxiv:1304.6401 [ astro-ph.co ] ; r. c. cotta , a. rajaraman , t. m. p. tait and a. m. wijangco , arxiv:1305.6609 [ hep - ph ] .
a. kurylov and m. kamionkowski , phys .
d * 69 * , 063503 ( 2004 ) [ hep - ph/0307185 ] ; f. giuliani , phys .
lett . * 95 * , 101301 ( 2005 ) [ hep - ph/0504157 ] ; s. chang , j. liu , a. pierce , n. weiner and i. yavin , jcap * 1008 * , 018 ( 2010 ) [ arxiv:1004.0697 [ hep - ph ] ] ; z. kang , t. li , t. liu , c. tong and j. m. yang , jcap * 1101 * , 028 ( 2011 ) [ arxiv:1008.5243 [ hep - ph ] ] ; j. l. feng , j. kumar , d. marfatia and d. sanford , phys .
b * 703 * , 124 ( 2011 ) [ arxiv:1102.4331 [ hep - ph ] ] .
j. l. feng , j. kumar and d. sanford , phys .
d * 88 * , 015021 ( 2013 ) [ arxiv:1306.2315 [ hep - ph ] ] .
i. m. shoemaker and l. vecchi , phys .
d * 86 * , 015023 ( 2012 ) [ arxiv:1112.5457 [ hep - ph ] ] .
l. goodenough and d. hooper , arxiv:0910.2998 [ hep - ph ] ; d. hooper and l. goodenough , phys .
b * 697 * , 412 ( 2011 ) [ arxiv:1010.2752 [ hep - ph ] ] ; d. hooper and t. linden , phys .
d * 84 * , 123005 ( 2011 ) [ arxiv:1110.0006 [ astro-ph.he ] ] ; k. n. abazajian and m. kaplinghat , phys .
d * 86 * , 083511 ( 2012 ) [ arxiv:1207.6047 [ astro-ph.he ] ] ; d. hooper and t. r. slatyer , arxiv:1302.6589 [ astro-ph.he ] ; c. gordon and o. macas , arxiv:1306.5725 [ astro-ph.he ] ; w. -c . huang , a. urbano and w. xue , arxiv:1307.6862 [ hep - ph ] . k. griest and d. seckel , nucl . phys .
b * 283 * , 681 ( 1987 ) [ erratum - ibid .
b * 296 * , 1034 ( 1988 ) ] ; g. jungman , m. kamionkowski , k. griest , phys .
rept . * 267 * , 195 - 373 ( 1996 ) .
[ hep - ph/9506380 ] . | we review the theoretical framework underlying models of asymmetric dark matter , describe astrophysical constraints which arise from observations of neutron stars , and discuss the prospects for detecting asymmetric dark matter .
address = department of physics and astronomy , university of hawaii , honolulu , hi 96822 , usa |
the broad history of the universe , according to the standard cosmological model , is a succession of expansion epochs from the earliest moments of the big bang to the present day . in reverse chronological order ,
these are : dark - energy domination , during which something like a cosmological constant drives the accelerated cosmic expansion ; matter - domination , during which cold dark matter and baryons shape the large - scale structure of the universe ; radiation - domination , during which the properties of the cosmic fluid are described by the relativistic degrees of freedom of the standard model ( sm ) of particle physics and beyond ; and inflation , during which the potential energy of a scalar degree of freedom dominates the cosmic energy density , producing an exponential burst of cosmic expansion . despite fundamental uncertainties the nature of dark energy and dark matter , and the physics beyond the standard model remain open questions , whilst the search continues for direct evidence of the inflationary birth of the universe this sequence of epochs forms the architecture of a widely - accepted , concordant model of the universe .
the aim of this article is to give a more quantitatively accurate account of cosmic history than one obtains from patching together inflationary , radiation- , matter- , and dark energy - dominated epochs .
it may often suffice to describe the cosmic energy density @xmath0 in terms of the scale factor @xmath1 and the equation of state @xmath2 for inflation , radiation , matter , and dark energy .
however , it is rare to find a detailed history of the trace of the stress - energy tensor of the cosmological fluid , or equivalently the ricci scalar curvature , @xmath3 where @xmath4 , and @xmath5 .
, the einstein field equations are @xmath6 , with @xmath7 and @xmath8 .
the variable @xmath9 is introduced to avoid confusing the trace with temperature @xmath10 . ]
the final output of our calculations is a determination of @xmath11 and @xmath9 , as functions of cosmic time or redshift : see fig .
[ fig : tracefullhist ] . to make these calculations ,
we must properly account for the thermal history of the cosmological fluid , the dynamics of the phase transition of quantum chromodynamics ( qcd ) , and we must give some account of the electroweak phase transition .
we do not solve the full boltzmann equations , as necessary for the standard calculations of the light element abundances in big bang nucleosynthesis and the relic dark matter abundance in extensions of the standard model ( e.g. @xcite ) , but instead use free field thermodynamics where applicable .
we also make some assumptions about inflation and reheating in the early universe in order to follow @xmath11 and @xmath9 all the way back to the inflationary epoch .
it is often supposed that the trace of the stress - energy tensor is negligibly small during the radiation - dominated epoch .
it is also supposed that the dominant contribution to the trace comes from baryons and dark matter .
both of these suppositions are incorrect : massive particle species contribute non - negligibly to the trace when the temperature is comparable to their mass ( see e.g. refs .
@xcite , wherein the history of @xmath9 is required to study cosmic evolution in a scalar - tensor theory of gravitation ) , and the effect of phase transitions on the trace of the stress - energy tensor is in fact much larger than the small contribution from baryons and dark matter . our motivation is to chart a detailed history of @xmath9 as comprehensively as we can .
two advances outside of cosmology make this possible .
first , progress in lattice qcd shows that the self - interactions of the quark - gluon plasma produce a trace that is very different than what one would obtain using free field theory @xcite .
second , the higgs mass is now known or at least , the mass of a particle whose most straightforward interpretation is the standard model higgs @xcite . because we do nt know what lies beyond the standard model , we offer two scenarios . in the first scenario ( despite all arguments about naturalness ) nothing lies beyond the standard model until one reaches the energy scale of inflation , only somewhat below the typical scale of grand unified theories . in the second scenario ,
there is a supersymmetric extension of the standard model plausibly within the reach of discovery at the lhc , and then inflationary physics at a similar energy scale to before .
our final figure covers the entire history of the universe from inflation to the present day for both scenarios .
a good ( and , by now , entirely conventional ) account of the expanding universe for @xmath12 can be achieved by including only non - relativistic baryons , pressureless cold dark matter , and dark energy with @xmath13 . of course photons and neutrinos
are present , but their contributions to the stress - energy tensor are negligible at this stage . as yet , there is no significant evidence for time - evolving dark energy ( e.g. ref .
@xcite ) ; however , the inferred , model - independent evolution at @xmath14 remains the subject of intense investigation as a measure of the dark energy equation - of - state .
thus @xmath13 for dark energy is only a default assumption .
the trace of the stress - energy tensor , and therefore the ricci scalar curvature , during this epoch is therefore @xmath15 as illustrated in fig .
[ fig : tracemattera ] .
the central values of the wmap - derived cosmological parameters@xmath16 for baryons , @xmath17 for cold dark matter , and @xmath18 @xcite for dark energy are used to set @xmath19 , @xmath20 , @xmath21 and @xmath22 where @xmath1 is the expansion scale factor .
the hubble constant is @xmath23 km / s / mpc with @xmath24 @xcite . for the moment we neglect any non - relativistic neutrino species , whose contribution is small relative to the baryons and dark matter .
we also ignore all fluctuations and effects of structure formation : our purpose is only to find the robertson - walker ( rw ) background that best approximates the expanding universe at very large scales .
the free field treatment of the radiation - dominated epoch is based on the following standard formulas for a collection of non - interacting fermions and bosons in thermal equilibrium : @xmath25 where the sum is over all species of particle , @xmath26 is the mass of the @xmath27-th species , @xmath28 is its multiplicity , and @xmath29 for bosons and @xmath30 for fermions @xcite .
we have allowed for different species to have different temperatures @xmath31 . for each collection of mutually thermally equilibrated particles , the common temperature must evolve in such a way that the collection s contribution to the stress - energy tensor is conserved .
it is not necessary to generalize ( [ freee ] ) and ( [ freep ] ) to include chemical potentials , because all the relevant chemical potentials ( for example , for baryon number ) are small compared to the temperature at times when explicit expressions for non - zero pressure are needed .
the contribution to the trace @xmath9 by photons is identically zero , whereas a non - zero trace is obtained for any free , massive species . in the relativistic limit @xmath32
the leading contribution to the trace is @xmath33 this means that as long as the relativistic fluid includes a massive species , then the trace , and therefore the curvature , is non - vanishing .
the energy density and pressure decay at slightly different rates as a species cools , so that as the temperature drops past about half the mass , a bump occurs in the contribution to the trace . as an illustrative example , fig .
[ fig : curvaturespike ] shows the ratio @xmath34 for a single boson or fermion species cooling below its rest mass .
we expect this process to be repeated for each relativistic species of the standard model , briefly raising the trace above a background level set by any abundant , non - relativistic particles for which @xmath35 , or else the most massive , relativistic species in equilibrium for which @xmath36 . by a species ( @xmath37 ) cooling below its rest mass
the solid ( dashed ) line is for fermions ( bosons ) . according to a free field treatment , the freeze - out of each species of the standard model results in a bump to the trace . ]
the most recent example of this phenomena is due to electrons as the temperature of the cosmic fluid drops below the electron mass .
turn ahead to fig .
[ fig : traceradera ] to see this illustrated . whilst standard model neutrinos become non
- relativistic more recently , their contribution to the trace is negligible .
ev at @xmath38 cl , where the lower bound is due to the mass splittings seen in neutrino oscillation experiments @xcite and the upper bound is obtained from cosmological observations @xcite . for these masses ,
neutrinos become non - relativistic at a time when the radiation energy density is smaller than the contribution due to baryons and cold dark matter .
consequently , the peak ratio of the trace contributed by neutrinos to non - relativistic matter is found to be @xmath39 occuring at a redshift @xmath40 .
hence , this is at most a @xmath41 change in the cosmic trace .
this is comparable to other phenomena which shift the effective number of relativistic neutrino species upwards from @xmath42 to @xmath43 which we also neglect .
see ref .
@xcite for a discussion of these corrections . ]
consequently , all one needs to model the cosmic trace at temperatures @xmath44gev or for @xmath45 , long before the onset of the matter - dominated era , is baryons , cold dark matter , and dark energy as given by eq .
( [ eqn : mattercurv ] ) . at temperatures @xmath46gev or
@xmath47 , a virtual parade of particle species , each cooling past their individual rest mass energies , keeps the cosmic trace well above the level of the energy density in cold dark matter . proceeding to higher temperatures ,
it is convenient to describe the relativistic gas in terms of an effective number of degrees of freedom @xmath48 for the energy density and pressure @xmath49 the stress - energy trace is @xmath50 . for the free , massive particles of the standard model composing the cosmological fluid it is straightforward to use equilibrium thermodynamics to calculate @xmath51 . as we pass to larger temperatures and redshifts , however , the main questions are : ( 1 ) when can we use a free field treatment of thermodynamics ? ( 2 )
when free field thermodynamics fails , what should we use instead ?
we will consider four cases in the following sections : the qcd phase transition , where lattice results are available and are significantly different from free field results ; the standard model electroweak transition , where ring - improved one - loop calculations indicate that departures from free field expectations are not too great ; a supersymmetric extension of the standard model , where we _ assume _ that the effect of interactions on the thermodynamics is fairly modest ; and finally inflation and reheating .
mesons and baryons melt into constituent quarks and gluons at the qcd phase transition .
the best source of information about thermodynamics in the vicinity of the transition is lattice qcd : see for example refs .
while the results of these works are qualitatively similar , quantitatively they differ enough so that it is best to summarize each separately . the authors of ref .
@xcite provide the following parametrized form : @xmath52 \left ( { d_2 \over t^2 } + { d_4 \over t^4 } \right ) \ , .
\label{bazavovform}\ ] ] we will use the values @xmath53 , @xmath54 , @xmath55 , and @xmath56 given in @xcite , which include a @xmath57 shift downward in temperature of the lattice data , which slightly reduces the discrepancy with @xcite ( to be described in the next paragraph ) .
the pressure as a function of @xmath10 can be extracted from the general result @xmath58 together with some information about @xmath59 at low temperatures . in order to find @xmath60 and @xmath61 at low temperatures
, we used the free field expressions ( [ freee ] ) and ( [ freep ] ) with the sum running over pions , kaons , and the @xmath62 , @xmath63 , and @xmath64 mesons , together with protons , neutrons , and their anti - particles .
the free field value of @xmath34 matches the one given in ( [ bazavovform ] ) at @xmath65 .
( it is coincidental that this is almost exactly the mass of the neutral pion . ) for @xmath66 we extracted @xmath59 from ( [ pfromtheta ] ) and then @xmath63 from ( [ bazavovform ] ) .
the authors of @xcite provide the following parametrized form as a fit to their results : @xmath67 \ , , \label{fodorform}\ ] ] where @xmath68 and @xmath69 , @xmath70 , @xmath71 , @xmath72 , @xmath73 , @xmath74 , and @xmath75 .
the relation ( [ pfromtheta ] ) can again be used to extract @xmath61 .
the form ( [ fodorform ] ) describes the lattice data from @xmath76 to @xmath77 , and it also is a good fit to the free field thermodynamics from @xmath78 to @xmath76 .
thus we can use ( [ fodorform ] ) essentially without additional low - temperature information or match to free field thermodynamics at @xmath79 .
we note that a fairly good fit to the lattice data of @xcite , which covers the range @xmath80 , can be achieved with a simpler functional form : @xmath52 { d_{3/2 } \over t^{3/2 } } \ , , \label{threehalvesform}\ ] ] where @xmath81 , @xmath82 , and @xmath83 .
the @xmath84 scaling is curious because it suggests a leading dimension @xmath85 for the operator which breaks conformal invariance at higher temperatures .
the approximate @xmath84 scaling persists over less than a factor of @xmath86 on the temperature axis , so we should not take this scaling dimension too seriously .
no free field construction will give a scalar operator with @xmath87 , but below @xmath77 we are far from asymptotic freedom .
, @xmath88 , and @xmath89 ) have been included .
the curve labeled `` hotqcd '' is based on ref .
@xcite , while the curve labeled `` wuppertal '' is from ref .
@xcite ; see the main text for details .
the free field theory curve was obtained by setting the masses of all the quarks and gluons artificially to @xmath90 .
this counter - factual assumption is made in order to obtain the narrowest possible peak in @xmath34 in the vicinity of the qcd transition , illustrating that interactions have a significant effect beyond introducing dynamically generated masses . ]
the results from @xcite and @xcite both support the conclusion that the peak in @xmath34 is narrower and higher than one can get from free field expressions : see fig .
[ fig : qcdcomparison ] .
perhaps this is not too surprising given that the qcd transition is almost a second - order phase transition , blunted only weakly into a cross - over , whereas free field expressions based on a finite particle spectrum are necessarily far from describing sharp phase transitions .
as is evident from fig .
[ fig : qcdcomparison ] , there is still some question about just how high and narrow the peak in @xmath34 is .
the main difficulty is that there are a number of different lattice actions , all of which have in principle the same continuum limit , but which can in practice give significantly different results at finite lattice spacing . extrapolating to the continuum based on finite lattice studies can never be fully systematic . the hotqcd results shown in fig .
[ fig : qcdcomparison ] are representative of p4 lattice data with lattice spacing @xmath1 chosen so that @xmath91 in the euclidean time direction . using instead an asqtad action results in a peak roughly @xmath92 lower .
the wuppertal results shown in fig . [ fig : qcdcomparison ] are based on a stout - improved staggered fermion action with @xmath93 in the range of @xmath94 to @xmath95 .
( for details on the p4 , asqtad , and stout - improved staggered fermion actions , see @xcite and references therein . ) going forward , we will use the p4 results of @xcite ; not that we endorse them as more accurate , but instead because they are the most different from free field results , providing us with an upper limit of how sharp the qcd transition might plausibly be as a result of interactions . in order to obtain a complete account of the visible sector degrees of freedom in the vicinity of the qcd transition ,
we combine the energy density and pressure obtained from ( [ bazavovform ] ) and ( [ pfromtheta ] ) with free field treatments of all the leptons , and also the @xmath96 and @xmath97 quark , where all particles are constrained to have the same temperature . a free field treatment is obviously not perfect ( particularly for the @xmath96 quark ) , but improved approximations would be complicated .
it is far from obvious that the electroweak transition can be approximated by free field thermodynamics .
the masses of all observed standard model particles owe their existence to a non - zero higgs expectation value , @xmath98 . but
this expectation value is eventually driven to zero at high temperatures .
an electroweak scale contribution to the cosmological constant accompanies this change in the higgs expectation value .
meanwhile , electroweak interactions introduce thermal corrections to particle masses . without accounting properly for these thermal corrections and other loop effects
, the higgs mass itself , now known to be approximately @xmath99 in vacuum @xcite , would become imaginary once the higgs expectation value falls below the point where the tree - level potential is concave up .
one of the main conclusions of this section , illustrated in fig .
[ fig : electroweakcomparison ] , is that , close to the peak of @xmath34 , free field thermodynamics based on the vacuum particle spectrum nevertheless provides a decent approximation to the ring - improved one - loop treatment of @xcite , which is the simplest account of the electroweak transition that avoids obvious inconsistencies such as imaginary masses . at substantially higher temperatures
, we will show that the ring - improved one - loop treatment predicts a negative value of @xmath34 .
a more modern understanding of the electroweak transition @xcite , based in part on lattice simulations , is that the transition is not weakly first order , as predicted by the ring - improved one - loop treatment , but is instead a cross - over .
if anything , we expect the full non - perturbative results for the trace @xmath9 of the stress tensor to be closer to the free field results than the ring - improved one - loop results are , though it is likely that the ring - improved one - loop treatment is still a good guide well above and well below the cross - over . the treatment of @xcite proceeds in three steps : 1 .
first one produces thermally improved formulas for all the fields using self - energy diagrams .
the schematic form of these masses is @xmath100 , where @xmath101 is a gauge coupling and @xmath102 is the higgs field expectation value .
the precise forms of all the masses are listed in table [ table : thermalmasses ] .
next one assembles an effective potential , correct through one - loop order , as follows : + @xmath103 \nonumber \\ v_{\rm therm}(\phi , t ) & = & \sum_j { g_j s_j \over 2\pi^2 } t^4 \int_0^\infty dx \ , x^2 \log(1 - s_j e^{-\sqrt{x^2+m_j(\phi , t)^2/t^2 } } ) \ , , \nonumber \end{aligned}\ ] ] + where the sums over @xmath27 run over the entries in table [ table : thermalmasses ] , and @xmath104 for bosons and fermions , as in ( [ freee ] ) and ( [ freep ] ) .
@xmath105 is the one - loop vacuum correction to the effective potential , and its dependence on @xmath10 arises only because , by prescription , one replaces the vacuum masses by the thermally improved mass formulas .
@xmath106 is the one - loop contribution from thermal occupation numbers for all the fields .
3 . at any fixed temperature @xmath10 ,
let @xmath107 be the value of the higgs field which minimizes @xmath108 .
then the pressure is @xmath109 .
the energy density can be extracted from the first law formula @xmath110 .
.thermally corrected masses of standard model particles as a function of the higgs expectation value @xmath102 and the temperature @xmath10 , following @xcite .
a few entries require explanation . at @xmath111
, @xmath112 becomes a physical , massless photon , but at @xmath113 it mixes with the higgs scalars along with the longitudinal @xmath114 and @xmath115 .
the @xmath116 entry corresponds to quanta of the magnitude @xmath102 of the higgs doublet .
we ignore the masses of all the fermions except the top quark , and we also ignore strong interactions.[table : thermalmasses ] [ cols="^,^,^ " , ] @xcite .
the sharp feature at @xmath117 , corresponding to the temperature at which the symmetry of the higgs potential is broken , is a shortcoming of the ring - improved one - loop treatment . ] in fig .
[ fig : electroweakcomparison ] , we compare the results of the ring - improved one - loop approach to the case of free field thermodynamics based on the masses and multiplicities of the standard model particle spectrum at zero temperature .
the kink in @xmath34 at @xmath117 shows the location of the first order transition predicted by the ring - improved one - loop approach .
this kink is actually a very slight discontinuity in @xmath34 as a function of @xmath10 .
the kink is smoothed out by the non - perturbative effects that make the entire transition a crossover .
it is clear from fig .
[ fig : electroweakcomparison ] that electroweak interactions and symmetry restoration have essentially no effect on the peak value of @xmath34 , and that they are mainly significant on the high - temperature side of this peak . in the next two paragraphs
we continue our investigation of the high - temperature behavior analytically . at temperatures high above the electroweak transition
, we expect that the leading contribution to the trace is due to the most massive relativistic species in the thermal bath . in the electroweak symmetric phase , however , the masses of most particles vanish .
although the longitudinal gauge bosons @xmath118 have large , thermally - improved masses , these temperature - dependent corrections drop out when we evaluate @xmath119 .
this is because their thermally - improved masses , in the high - temperature , symmetry - restored phase , are dimensionless constants times the temperature , resulting in a contribution to @xmath59 that scales precisely as @xmath120 .
the only particle in the spectrum whose mass in the symmetry - restored phase is _ not _ linear in the temperature is the higgs .
so its thermal excitations dominate @xmath9 at high temperatures .
an important sub - dominant contribution is the classical higgs potential .
altogether , at @xmath121 , @xmath122 where @xmath123 the contribution to @xmath124 shown explicitly is from the tree - level higgs potential .
the corrections , indicated by @xmath125 , are numerically small given the measured value of the higgs mass .
the @xmath126 term is from the thermal excitations of the higgs .
numerically , @xmath127 , valid for @xmath128 . in the absence of contributions beyond the standard model ,
the trace @xmath9 is negative at high temperatures , starting at about the temperature @xmath129 where the negative @xmath126 term in ( [ thetaexpand ] ) overtakes the positive @xmath124 term .
the most negative value attained by @xmath34 occurs at @xmath130 at which point @xmath131 . because this negative trace is a result found using the ring - improved one - loop approximation , it is not to be regarded as iron - clad .
we leave for future work the interesting question of whether it persists beyond this approximation .
( heavy curves ) , and pressure @xmath132 ( light curves ) , are shown for the standard model ( solid curves ) and a representative supersymmetric extension ( dashed curves ) . in the lower panel ,
the difference @xmath133 is shown .
note that adding a single dark matter species to the standard model curves does not produce a discernible change in these figures . ]
we may reasonably expect the lhc or direct dark matter detection experiments to discover new particle species at higher masses .
consequently , we supplement the standard model particle spectrum with a dark matter ( dm ) particle . until it freezes out , a single , relativistic , bosonic species contributes @xmath134 to the trace .
if a bosonic dark matter particle has a mass @xmath135 or @xmath136 then there will be a window in which the collective stress - energy trace is negative due to the standard model higgs . such a possibility is reflected in fig . [ fig : traceradera ] , where we include a @xmath137 dark matter particle which freezes out after it becomes non - relativistic , in our extrapolation of @xmath9 back to times before and temperatures above the standard model electroweak transition .
in hindsight , it is perhaps not too surprising that a weakly coupled theory like the standard model manages to have thermodynamic properties not too far from free field theory .
we therefore feel justified to include here a similar free field account of the constrained minimal supersymmetric standard model ( cmssm ) , in which we use darksusy @xcite with input parameters corresponding to the best fit to the recent lhc data ( table 4 of ref .
@xcite : @xmath138 , @xmath139 , @xmath140 , @xmath141 , and @xmath142 ) to generate a zero - temperature mass spectrum , which we then plug into ( [ freee ] ) and ( [ freep ] ) without regard to the modifications that will arise from interactions and symmetry restoration .
though these modifications are probably significant on the high - temperature side of the peak in @xmath34 , the main uncertainty lies not in their precise determination , but our ignorance of what the superpartner spectrum really looks like ( assuming it is there at all ) . as seen in figs .
[ fig : degreesoffreedom ] and [ fig : traceradera ] , an anomalous trace arises as the spectrum of sparticles with masses in the range @xmath143 tev become non - relativistic .
more specifically : the best - fit model we used as a benchmark has a neutralino lsp at @xmath144 , additional color singlet sparticles below @xmath145 , and colored sparticles above @xmath146 except for a stop at approximately @xmath147 .
proceeding to ever higher energies requires a theory to guide our calculation of the trace . in the approach to gut - scale energies
, we may pass through a desert , or there may be new particle species littering the landscape .
we should also expect couplings to run , although this is a small effect in simple extensions of the standard model . pausing to reflect on fig .
[ fig : traceradera ] , we may assess the degree of error commonly made in the literature by assuming the curvature to be negligible , or else solely determined by the non - relativistic matter content extrapolated back from the matter - dominated era .
instead , we see that at various epochs , when the number of degrees of freedom changes sharply , the trace rises to an appreciable fraction of the total energy density .
the history of curvature at earlier times requires a theory of the early universe , which we take to be the inflationary dynamics of a scalar field . whereas other theories of the primordial universe will yield different predictions for the trace and curvature , the merit of inflation is the consistency with observations and ubiquity of examples .
the generic course of inflation is exponential expansion proceeding with @xmath148 caused by the slow roll of the inflaton @xmath149 down its potential until some late stage when it breaks from slow roll and plummets down the potential towards the vacuum . in broad terms
, the inflaton undergoes damped evolution as it decays into a turbulent mixture of relativistic particles and massless radiation .
it is only after some transient period that these decay products equilibrate and can be appropriately described as a thermal bath with temperature @xmath10 .
the description of the cosmological fluid , trace and curvature , can then merge with our description of the high temperature limit beyond the standard model . a detailed analysis of the post - inflationary , preheating epoch including the approach to equilibrium requires a numerical simulation of the nonlinear , quantum fields ( _ e.g. _ refs .
nonetheless , we can capture the basic behavior with a semi - analytic model . to be definite
, we consider an inflaton with potential @xmath150 , which decays into relativistic particles and massless radiation via a damping term . following refs .
@xcite , we evolve the system @xmath151 with @xmath152 , @xmath153 .
the field is started in a slow - roll state , with @xmath154 gev chosen to yield satisfactory inflation @xcite . as slow roll ends , the inflaton begins to oscillate about the potential minimum , such that @xmath155 swings between @xmath156 and @xmath157 ; despite the fact that the time - averaged inflaton pressure is small compared to the energy density , the trace and curvature become negative .
this pressureless phase is shortlived , as an assumed inflaton decay width @xmath158 gev causes the field to decay exponentially with cosmic time ; the trace @xmath159 plummets until it meets the anomalous trace @xmath160 due to any massive species among the inflaton decay products .
this scenario , beginning from the end stages of inflation , is illustrated in fig .
[ fig : traceinflera ] .
the variety of inflationary models , and the challenges of accurately simulating the nonlinear dynamics of reheating , translates into a wide degree of uncertainty in the behavior of the trace and curvature in the approach to equilibration . in the above model
we have assumed an idealized description of the radiation produced by the decay of the inflaton .
if the inflaton decay products do not achieve the relativistic equation of state @xmath161 until after a brief equilibration period , then the trace may be much larger .
if one of the decay products includes a kinetic - energy dominated scalar field with equation of state @xmath162 then the trace contains a residual , negative term that decays as @xmath163 . this possibility is also illustrated in fig .
[ fig : traceinflera ] .
the gap between the trace and the energy density is so wide in the interval between the end of inflation and the electroweak epoch , assuming no new physics , that we may question whether there are any other sources of curvature or trace .
for example , long - wavelength fluctuations in the energy density and pressure might lead to an rms contribution .
if the pressure and energy density perturbations imparted by inflation unto the cosmological fluid are characterized by a propagation speed @xmath164 where @xmath165 , then our question becomes : is @xmath166 as closely tuned to @xmath167 as is @xmath168 , the background equation of state ? in the case of perfectly adiabatic perturbations , with @xmath169 ,
then the answer is yes . at some level , however , we expect a departure from adiabaticity .
recall that the linearized perturbation to the ricci scalar in a coordinate system @xmath170 is @xmath171 , and @xmath172 , where @xmath173 is a familiar variable from studies of inflation .
( see ref .
@xcite . ) for perturbations of wavelength much greater than the radiation - era hubble radius , specifically those perturbations responsible for the observed large - scale structure , then @xmath173 is a constant in time and @xmath174 becomes negligibly small so that @xmath175 . in this case , the homogeneous quantities we have calculated provide the leading contribution . to conclude this brief history of curvature , we show all three epochs in a single display , in fig . [
fig : tracefullhist ] .
the history of the curvature and trace are of interest , beyond their intrinsic value as a record of the significant events in the large - scale evolution of our universe .
for example , an accounting of the thermal history of the universe is necessary to accurately account for the spectrum of inflation - produced gravitational waves @xcite .
however , the most dramatic effect may arise when considering new gravitational phenomena beyond general relativity .
gravitational theories that introduce new degrees of freedom may require an accurate model of the trace of the stress - energy tensor .
the canonical example is the nonminimally - coupled scalar field , for which the lagrangian is @xmath176 and where @xmath177 includes the standard model .
the @xmath178 equation of motion is @xmath179 so the curvature introduces a novel , time- and space - dependent effective mass such that when @xmath180 then the field is susceptible to a tachyonic instability ( _ e.g. _ @xcite ) .
the curvature is set by the dynamics of the field , which are in turn sourced by the trace of the stress - energy of any other forms of matter .
specifically , @xmath181 in the case of a strongly coupled field with @xmath182 , for which @xmath183 so that the @xmath178 field does not dominate the stress - energy , then we have approximately @xmath184 . in the case of a conformally - coupled field , @xmath185 and @xmath186 , then @xmath187 , which means that the trace calculated in this paper is directly applicable to the evolution of @xmath178 : every feature in our @xmath9 curve results in a feature in @xmath178 .
similarly , nonlocal quantum gravitational effects that depend upon quantities such as @xmath188 or @xmath189 , where @xmath190 is a fourth - order conformally - invariant operator @xcite , are sensitive to the history of curvature .
consequently , the results presented in this paper for the evolution of the trace may prove useful in detailed studies of new gravitational physics , such as gravitationally - coupled inflation @xcite or dark - energy motivated modifications of gravity ( _ e.g. _ ref .
@xcite , or @xcite and references therein ) .
we thank p. langacker , m. lisanti , and r. mawhinney for useful discussions .
the work of r.r.c . was supported in part by nsf phy-1068027 .
the work of s.s.g . was supported in part by the department of energy under grant no .
de - fg02 - 91er40671 . g. -b .
zhao , r. g. crittenden , l. pogosian and x. zhang , phys .
lett . * 109 * , 171301 ( 2012 ) [ arxiv:1207.3804 [ astro-ph.co ] ] .
d. larson , j. dunkley , g. hinshaw , e. komatsu , m. r. nolta , c. l. bennett , b. gold and m. halpern _ et al .
_ , astrophys .
j. suppl .
* 192 * , 16 ( 2011 ) [ arxiv:1001.4635 [ astro-ph.co ] ] .
a. g. riess , l. macri , s. casertano , h. lampeitl , h. c. ferguson , a. v. filippenko , s. w. jha , w. li _ et al .
_ , astrophys . j. * 730 * , 119 ( 2011 ) .
[ arxiv:1103.2976 [ astro-ph.co ] ] .
e. komatsu _ et al .
_ [ wmap collaboration ] , astrophys .
j. suppl .
* 192 * , 18 ( 2011 ) [ arxiv:1001.4538 [ astro-ph.co ] ] .
k. n. abazajian , e. calabrese , a. cooray , f. de bernardis , s. dodelson , a. friedland , g. m. fuller and s. hannestad _ et al .
_ , astropart .
* 35 * , 177 ( 2011 ) [ arxiv:1103.5083 [ astro-ph.co ] ] .
j. lesgourgues , s. pastor , phys .
rept . * 429 * , 307 - 379 ( 2006 ) .
[ astro - ph/0603494 ] .
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 ] ] .
p. b. arnold and o. espinosa , phys .
d * 47 * , 3546 ( 1993 ) [ erratum - ibid .
d * 50 * , 6662 ( 1994 ) ] [ arxiv : hep - ph/9212235 ] e. mottola and r. vaulin , phys .
d * 74 * , 064004 ( 2006 ) [ arxiv : gr - qc/0604051 ] .
s. deser and r. p. woodard , phys .
lett . * 99 * , 111301 ( 2007 ) [ arxiv:0706.2151 [ astro - ph ] ] .
f. l. bezrukov , m. shaposhnikov , phys .
lett . * b659 * , 703 - 706 ( 2008 ) .
[ arxiv:0710.3755 [ hep - th ] ] . | ( 0,0)(0,0)(397,120)pupt-2439 the trace of the stress - energy tensor of the cosmological fluid , proportional to the ricci scalar curvature in general relativity , is determined on cosmic scales for times ranging from the inflationary epoch to the present day in the expanding universe .
the post - inflationary epoch and the thermal history of the relativistic fluid , in particular the qcd transition from asymptotic freedom to confinement and the electroweak phase transition , leave significant imprints on the scalar curvature .
these imprints can be of either sign and are orders of magnitude larger than the values that would be obtained by naively extrapolating the pressureless matter of the present epoch back into the radiation - dominated epoch . |
extended - emission line regions ( eelr ) are a common feature of powerful radio galaxies . in the high - redshift objects , these regions display a particular range of phenomena which are of especial interest for studies of the properties of agn and the formation and evolution of their host galaxies .
extended line and rest - frame ultraviolet continuum emission appears , above reshifts of around 0.8 , to be aligned with the extended double radio structures ( mccarthy et al .
1987 , chambers , miley & van breugel 1987 ) although the radio and optical emitting regions are not necessarily co - extensive . in some cases
, however , there is a close correlation between the radio emission and the eelr over distances of tens of kiloparsecs and extreme kinematic components are observed associated with the radio emission ( e.g. , mccarthy et al .
1987 , mccarthy 1993 ) , suggesting a continuing interaction between the radio plasma and the ism . in an effort to understand the high redshift phenomena in more detail
, we are studying selected low redshift objects which exhibit some of these characteristics .
the correlation between eelr and radio emission is not so common at low @xmath1 .
nevertheless , a number of cases of regions of ionized gas found coincident or close to a radio feature either jet or lobe are known and , in some cases , well studied ( e.g. , ngc 7385 , simkin & ekers 1979 ; 3c171 , heckman , van breugel & miley 1984 ; 3c277.3 , van breugel et al .
1985a ; minkowski s object , van breugel et al . 1985b
; 4c 29.30 , van breugel et al.1986 ; centaurus a , morganti et al .
1990 ; 3c285 , van breugel & dey 1993 ; pks 2250 - 41 , clark et al .
1997 and pks 1932 - 43 , villar - martin et al . in prep .
some of them are believed to represent an interaction between the radio plasma and the ism : a ` jet / cloud ' interaction .
pks 2152 - 699 is a powerful , nearby radio galaxy ( @xmath2 , 1 arcsec = 0.8 kpc for h@xmath3=50 km s@xmath4 mpc@xmath4 ) with a cloud of highly ionized gas observed at about 10 arcsec ( @xmath58 kpc ) from the nucleus .
the very highly ionized cloud ( hic ) , with ions up to fe@xmath6 and a blue optical continuum , was first studied in detail by tadhunter et al .
( 1987 , 1988 ) .
although the hic has a velocity very similar to systemic , broad , blue wings in the [ o iii ] emission lines ( up to 3000 km / s from the systemic velocity of the hic and galaxy ) have been observed .
the presence of some gas at high velocities led tadhunter et al .
( 1988 ) to conclude that the cloud could possibly be related to interaction between the radio plasma and the ism .
however , later work ( di serego alighieri et al . 1988 ; fosbury et al.1990 ) showed that the hic has a very blue optical to near - uv continuum with 10% linear optical polarization and the e - vector perpendicular to the position angle of the nucleus - hic axis .
the existence of the blue , polarized continuum led di serego alighieri et al .
( 1988 ) to suggest that this resulted from the scattering by dust in the cloud of beamed radiation from the nucleus . in this picture , the source of ionization for the cloud is uv radiation from the nucleus .
the presence of transverse ionization gradients also lent support to the beamed illumination model ( tadhunter et al .
1988 ) . the quality of the radio maps available until now ( christiansen et al.1977 , tadhunter et al .
1988 , jones & mcadam 1992 , fosbury et al .
1990 ) does not allow a detailed investigation of the radio emission in the region of the hic . to search for clear evidence of a jet / cloud interaction in pks 2152 - 699 ,
we have made new radio observations with the australia telescope compact array ( atca ) at both 3 and 6 cm wavelength . in presenting the radio data and comparing with the optical morphology
, we use a broadband red image of the galaxy taken with the hst pc 2 available from the public archive .
vlbi observations of the nuclear region were presented by tingay et al .
( 1996 ) .
subsequent papers will report a series of infrared , optical , ultraviolet and x - ray observations which allow us to study the broad - band spectral energy distribution of the nucleus and separate components within the interaction site .
the observations reported in this paper were made on the 19 and 23 january , 21 march and 1 april 1992 using the four 6-km configurations available with atca and the standard continuum correlator setup with a bandwidth of 128 mhz and 32 channels .
the two simultanous frequencies were set to 4.74 and 8.64 ghz .
the source was observed for 12h in each configuration .
the data were calibrated by using the miriad package ( sault , teuben & wright 1995 ) , which is necessary for the calibration of the polarization of atca data .
the flux scale is based on the recent compilation of mesurements of the primary calibrator pks 1934 - 638 by reynolds ( 1996 ) which corresponds to 5.91 jy at 4.7 ghz and 2.84 at 8.6 ghz .
this differs from the previous calibration in use at the compact array by ( new - old ) -8.1% at 4.7 ghz and + 9.8% at 8.6 ghz . at 4.7 ghz
we have made images with both uniform weight and using the `` robust '' parameter equal to 0.5 .
the former gives the full resolution of @xmath7 ( pa = @xmath8 ) while the latter gives a beam of @xmath9 arcsec ( pa = @xmath10 ) .
using the `` robust '' parameter we were able to give more weight to the short baselines and therefore the extended low - brightness emission is better imaged . the total intensity map at 4.7 ghz at lower resolution is shown in fig . 1 .
at 8.6 ghz , the data have a full resolution of @xmath11 arcsec ( pa = @xmath12 ) .
the rms noise of the total intensity maps is about 0.80 mjy beam@xmath4 in the 4.7 ghz map and 0.65 mjy beam@xmath4 in the 8.6 ghz map
. * figure 1 .
* atca 4.8 ghz image of pks 2152 - 699 .
the contour levels are 1 , 1 , 2 , 3 , 4 , 5 , 6 , 8 , 10 , 12 , 16 , 20 , 30 , 40 , 80 , 160 @xmath13 3 mjy beam@xmath4 .
the total fluxes derived from these synthesis observations are given in table 1 . at 8.6 ghz
, it appears to be lower than single dish observations ( 6.22 jy at 8.4 ghz from wright & otrupcek 1990 ) , most likely due to missing short spacings .
.radio properties of pks 2152 - 699 [ cols="<,^,^,^,^ " , ] \a ) beam size [email protected] arcsec + b ) beam size [email protected] arcsec
+ we have also obtained the images of the stokes parameters ( @xmath14 ) , the polarized intensity image ( @xmath15 ) and position - angle image ( @xmath16 ) . here
we present these images only for the data at 4.7 ghz because at 8.6 ghz the polarization maps are not completely reliable , probably due to instrumental polarization affecting the data in the earliest days of atca when the data presented here were collected .
the rms noise of the @xmath17 and @xmath18 maps at 4.7 ghz is about 0.17 mjy beam@xmath4 .
the polarized intensity and the fractional polarization ( @xmath19 ) were estimated only for the pixels for which @xmath20 .
2 shows a greyscale image of the polarized intensity while fig .
3 shows contours of the total intensity with superimposed vectors whose length is proportional to the fractional polarization and whose position angle is that of the electric field .
the mean fractional polarization of the different regions is given in table 1 . * figure 2 . * greyscale image of the polarized intensity of pks2152 - 69 at 6 cm .
the range is between 0 and 10 mjy .
* figure 3 .
* contours ( as in fig . 1 ) of the total intensity at 4.8 ghz image of pks 2152 - 699 with superimposed vectors
whose length is proportional to the fractional polarization and whose position angle is that of the electric field . in order to estimate the spectral index in some regions we have produced an 8.6 ghz map degrading the resolution to match that of the 4.7 ghz ( uniform weight ) .
the spectral index @xmath21 ( defined as @xmath22 ) was estimated only from the regions with signal above @xmath23 in both maps .
the values of the spectral index must be interpreted with caution because the observations were not made with matched arrays : a more complete study of both polarization and spectral index is in progress and will be presented in a forthcoming paper .
compared to the previous observations from the molonglo synthesis telescope ( tadhunter et al .
1988 ) and from atca ( fosbury et al.1990 ) , the radio structure is now shown more clearly .
an important feature is the bright source @xmath24 arcsec ne of the nucleus , i.e. , at the same distance and similar but not identical position angle from the nucleus as the hic .
this component , which we call rc , was not visible in the previous low resolution radio images available for pks 2152 - 699 although it was seen with a lower significance in the early atca image by fosbury et al .
( 1990 ) .
the new radio images show that pks 2152 - 699 , although strictly classified as a fanaroff & riley type ii source on the basis of the ratio of hot - spot separation to total source length , it is unusual in showing the hot - spots near the centres of the lobes .
this behaviour , which is intermediate between fr i and ii types , has been observed before ( e.g. capetti , fanti & parma 1995 ) in objects with radio power on the border between the two classes ( @xmath25 w hz@xmath4 at 4.7 ghz ) .
the total radio power of pks 2152 - 699 ( @xmath26 w hz@xmath4 at 4.7 ghz ) is indeed typical of the transition region between fr i and ii galaxies . for the objects studied by capetti et al .
( 1995 ) , the peculiar morphology could be due to precession of the central engine or perhaps by an externally - induced change in jet direction .
the southern lobe of pks 2152 - 699 shows an edge - brightened , arc - like structure which suggests jet precession .
the bright spot then represents the primary hot - spot and the current point of impact of the jet with the ism while the arc - like bright emission represents the previous impact points during the change of direction of the jet ( see also the simulations in cox , gull & scheuer 1991 ) . something similar could also occur in the northern lobe with the radio plasma far away from the hot - spot could represent the backflow from previous , more distant , fading hot spots . a similar scenario , with the lobe considered as a relic from earlier jet activity , has been suggested for 3c111 ( linfield & perley 1984 ) .
hot - spots in the middle of the lobe can also be due to projection effects or the result of an interaction moving the jet direction .
the spectral indices observed in the hot - spots are unusually steep ( @xmath27 ) which could be due to the result of a spectral break between these two frequencies .
this has been observed in the knots of 3c277.3 , an object with many characteristics in common with pks 2152 - 69 .
the radio emission in pks 2152 - 699 is asymmetric both in flux and core / lobe distance .
the southern lobe extends for approximately 30 arcsec ( 24 kpc ) while the northern extends to 45 arcsec ( 36 kpc ) .
we do not find clear evidence for a jet , but a thin , low brightness emission bridge between the nucleus and the northern lobe is visible at 4.7 ghz .
although very weak , the total emission from the northern lobe shows substructure .
this substructure can be seen very clearly in polarized emission ( fig . 2 ) .
two arm - like structures are visible along its southern and northern edges .
the northern lobe shows a somewhat higher fractional polarization while the hotspots appear to be less polarized than the rest of the lobes . in the southern lobe ,
the polarization vectors follow the edge - brightened arc - like structure ( see fig .
3 ) , typical of such sources .
the fractional polarization in the lobes is quite high although not exceptional .
pks 2152 - 699 has a prominent core ( see table 1 ) .
high resolution observations made in 1988 - 89 with the parkes - tidbinbilla interferometer by jones et al .
( 1994 ) give a core flux of 784 mjy at 2.3ghz while more recent observations obtained in 1994 with the same instrument give 583 mjy at the same frequency ( morganti et al .
this may indicate some variability of the flux density of the nucleus although new observations will be necessary to confirm this .
the core has an inverted spectral index , @xmath28 .
the ratio @xmath29 between the core flux density and the extended flux density derived at 4.7 ghz is @xmath300.06 .
this is a typical value for broad - line galaxies as pks 2152 - 699 is classified by tadhunter et al.(1988 ) while it is large compared to typical fr ii narrow - line radio galaxies as discussed in morganti et al .
( 1997 ) for data collected at 2.3 ghz .
pks 2152 - 699 is known to be situated in a galaxy - poor environment ( tadhunter et al .
however , the presence of a nearby ( @xmath31 arcmin east ) strong radio source , pks 2153 - 69 , was already known from the model fitting to low resolution data carried out by ekers ( 1969 ) .
this source is unresolved in the maps of christiansen et al .
( 1977 ) and tadhunter et al .
( 1988 ) and has been identified by jones & mcadam ( 1992 ) with a 20th magnitude galaxy with unknown redshift but likely to be in the background given its magnitude . with the new data we have mapped the region corresponding to pks 2153 - 699 and resolved the source into two lobes or tails . some trail of radio emission to the west is also observed .
more faint structure may be missed because of the attenuation of the primary beam . in fig .
4 the contours of the radio emission are shown .
the cross indicates the position of the optical identification , situated between the two lobes .
* figure 4 . *
contours of the total intensity at 4.8 ghz image of the nearby galaxy pks 2153 - 699 .
the contour levels are -1 , 1 , 2 , 4 , 6 , 8 , 10 , 12 , 14 , 16 , 20 , 30 @xmath13 2 mjy beam@xmath4 a broad - band hst pc 2 image ( f606w filter , single 500s exposure ) of pks 2152 - 699 was retrieved from the public archive .
after interactively cleaning cosmic - ray events , we have used this image to produce an overlay with the new radio image at 4.7ghz .
the registration between the two images was carried out by aligning the radio nucleus ( ra(j2000 ) = 21 57 06.0 , dec(j2000 ) = -69 41 24.0 ) with the peak of the optical image ( @xmath32 0.2 arcsec ) .
the overlay is shown in fig . 5 while fig .
6 presents a cartoon which shows more clearly the spatial relationship between the radio , optical emission and dust absorption components .
the hst image , which includes line ( predominantly [ o iii ] , h@xmath33 and [ n ii ] ) emission and continuum , clearly shows structural details of the hic and several additional components which fall along the same axis . the position angle between the nucleus and the observed radio component , rc , near the hic is 34@xmath34
this is intermediate between that of the large - scale radio structure ( pa@xmath35 defined from the low resolution most radio image , tadhunter et al .
1988 ) and that from vlbi observations presented by tingay et al .
the vlbi observations reveal a core - jet morphology on the parsec scale and the pa of this structure is @xmath36 , close to that of the centroid of the optical hic .
the overlay shows that the hic is situated to the east of rc which itself is coincident with a prominent part of the substructure of the cloud .
this latter component was identified as a red stellar object ( c ) in the line - free continuum observations reported by di serego alighieri et al .
( 1988 ) and is therefore predominantly continuum rather than line emission with a colour very different from the rest of the hic .
the dust band crossing the galaxy in pa@xmath37 was inferred from the groundbased images from the colour gradient measured across the galaxy ( di serego alighieri et al .
1988 ) .
* figure 5 * hst f606w image with radio 6 cm contours superimposed .
the hst image has been manually cleaned of cosmic ray events and is shown on a logarithmic scale .
the radio contours are also logarithmic , ranging from 3 to 700 mjy beam@xmath4 .
the radio and optical nuclei have been forced to coincide .
the regions containing some of the faint features are shown with a stretched intensity scale .
* figure 6 * cartoon showing the different absorption and emission components in pks2152 - 699 .
the vlbi image is represented to a different scale at the lower left .
in addition to the emission components to the ne of the nucleus , which extend to greater distance than shown in fig . 5 ( tadhunter et al . 1988 ) , the hst image shows two faint regions to the sw at radial distances of approximately 11 and 16 arcsec in pa 237@xmath34 .
these were not detected in the groundbased images .
there is also a bright ` bow ' structure , convex to the nucleus , approximately 3 arcsec along pa@xmath38 to the ne .
this is diametrically opposite the faint sw structures .
this bow was seen as a nuclear extension in the groudbased emission line imaging ( see fig . 5 ) .
the combination of radio and optical images of pks 2152 - 699 allow us to make the following statements : 1 .
there is a radio component associated with the hic although it is not situated at the brightest optical location .
the optical emission coincident with this is red in colour and probably continuum rather than line emission .
2 . there is a progression in pa from the vlbi axis ( = the nucleus hic axis ) to the nucleus
inner radio component to the large - scale radio double axis ( 44 34 23@xmath34 ) .
there are a number of distinct optical structures ( of which the hic is the most prominent ) lying close to the 44@xmath34 pa axis .
these exist on both sides of the nucleus although they are more numerous and brighter to the ne .
a dust lane crosses the nucleus in pa@xmath39 with emission along its northern edge
. this could be the low ionization line emission reported by tadhunter et al.(1988 ) but we can not exclude a continuum contribution .
the northern and southern lobes have a very different morphologies the northern being diffuse with a central hotspot , the southern brighter with a ` trailed ' hotspot .
the discovery of the bright radio component next to the hic supports the hypothesis of a jet / cloud interaction in pks 2152 - 699 and it reinforces the similarity between this object and 3c277.3 , one of the best examples of jet / cloud interaction ( van breugel et al .
1985a ) . as pointed out by tadhunter et al .
( 1988 ) , comparison of the two objects shows a similarity in the emission - line spectra , the presence of the blue optical continuum and the offset between the cloud / nucleus axis and the radio axis .
they also pointed out , however , that the cloud of ionized gas in 3c277.3 shows a large velocity gradient while the hic in pks 2152 - 699 shows broad , low intensity wings on the [ o iii ] lines extending to the blue .
the presence of extensive , chaotic dust lanes in the inner regions of the galaxy suggests that the the radio jet is interacting with a fragment of a merging galaxy .
we have estimated for pks 2152 - 699 the minimum pressure associated with the bright radio component to be @xmath40 dynes @xmath41 .
we do not have a good measure of the density of the warm line - emitting gas but if we assume t@xmath4215000 k ( tadhunter et al.1988 ) and @xmath43 @xmath44 ( a typical value for a bright extranuclear cloud ) we find that , as in the case of other jet / cloud interactions , the radio and line - emitting gas pressures are quite comparable ( clark & tadhunter 1996 ) suggesting that the region of high ionization has been compressed by the interaction with the radio plasma .
the presence of eelr in high and intermediate redshift radio galaxies is often associated with asymmetries in the radio morphology and polarization ( pedelty et al . 1989 , liu & pooley 1991 , clark et al . 1997 ) . in pks 2152 - 699
, however , there is no strong polarization asymmetry and the stronger , closer ( in the sky plane ) radio lobe is on the side opposite to the jet / cloud interaction .
in contrast with the more powerful radio galaxies at higher redshifts , where the polarization and structural asymmetries appear to be due to large - scale ism density variations across the source , the interaction seen here may be a transitory phenomenon without a major influence on the main radio lobe structure .
we have presented new radio maps of the radio galaxy pks 2152 - 699 obtained with the atca .
these reveal a bright radio component next to the cloud of very highly ionized gas , situated about 8 kpc from the nucleus .
the radio component is separated from the brightest region of the optical cloud but coincides with source of red continuum emission .
this structure , together with the unusual morphology of the northern radio lobe , suggests that a jet / cloud interaction has taken place .
such interactions are relatively rare at low redshift but , when they occur , do offer the opportunity for quantitative investigation of the properties of the jets themselves .
we are currently persuing such studies using these and other new observational data on this source .
this work is based on observations with the australia telescope compact array ( atca ) , which is operated by the csiro australia telescope national facility .
it also uses observations made with the nasa / esa hubble space telescope obtained from the data archive at the st - ecf .
raef is affiliated to the astrophysics division , space science department , european space agency .
rm acknowledges support from the ditac international science & technology program .
sault , r.j . ,
teuben , p.j . , wright , m.c.h .
1995 in _ `` astronomical data analysis software and systems iv '' _ , eds .
r. shaw , h.e .
payne and j.j.e .
hayes , astronomical society of the pacific conference series , 77 , 433 | we present radio observations of the radio galaxy pks 2152 - 699 obtained with the australia telescope compact array ( atca ) .
the much higher resolution and s / n of the new radio maps reveals the presence of a bright radio component about 10 arcsec ne of the nucleus .
this lies close to the highly ionized cloud previously studied in the optical and here shown in a broadband red snapshot image with the hst pc 2 .
it suggests that pks 2152 - 699 may be a jet / cloud interaction similar to 3c277.3 .
this could cause the change in the position angle ( of @xmath0 ) of the radio emission from the inner to the outer regions . on the large scale ,
the source has fanaroff & riley type ii morphology although the presence of the two hot - spots in the centres of the lobes is unusual .
the northern lobe shows a particularly relaxed structure while the southern one has an edge - brightened , arc - like structure .
galaxies : active galaxies : interactions galaxies : radio continuum |
for certain areas of current astronomical research , such as the detection and mass determination of black holes in galactic nuclei , spatial resolution is the main limitation . for others ,
e.g. the case of extrasolar planets , the challenge is to detect very faint signals in the presence of an overwhelmingly bright star .
interferometry from space has the potential to overcome both of these obstacles : it has no fundamental limit to spatial resolution , and interferometric nulling is by far the most powerful method to suppress unwanted light from a bright point source
. the space interferometer mission ( sim ) will be the first mission to demonstrate the feasibility of this new and exciting technology .
scheduled for launch in 2005 , sim is the first new mission of nasa s origins program after sirtf .
it is designed primarily for astrometry at the -level .
however , sim may also have significant capabilities for synthesis imaging . in terms of spatial resolution ,
sim with a maximum baseline of 12 m promises at least a factor of six improvement over the hubble space telescope ( hst ) . in this paper
, we discuss the potential of sim to significantly improve upon current observational limits in two specific astronomical cases , namely line - emitting gas disks rotating around black holes in galactic nuclei , and continuum zodiacal dust disks around nearby stars . in the next section , we will describe how sim performs synthesis imaging , and discuss some aspects and limitations that will affect the observations .
section [ simsim ] gives an overview of the capabilities of the code which we use to simulate various classes of model sources .
input models and first results of simulations for the two examples mentioned above are described in sections [ bh ] and [ exo ] .
we discuss and summarize our findings in section [ sum ] .
appendices describe the basics of interferometric imaging , and the simulation of measurement errors in our models , respectively .
as of november 1998 , certain aspects of sim design which are important for synthesis imaging have not yet been finalized .
the most important of these is to have a sufficiently complete range of of baseline spacings available . for the purpose of this paper , we assume that sim will have baselines between 0.5 and 12 m in increments of 0.5 m , similar to the original design ( see @xcite for an overview ) . with these capabilities , sim will be well suited to perform reliable synthesis imaging for a wide range of complicated sources .
the basic `` measurement '' for synthesis imaging consists of determining the amplitude and phase of the interference fringes for a given baseline and spacecraft orientation .
these two quantities together form the `` complex visibility '' . in short ,
the signal of the two interfering beams on the detector depends on the path difference or `` delay '' between them .
the interference pattern is a cosine wave whose amplitude and phase are the desired observables . if the path delay is accurately known and controllable , the complex visibilty can be determined by `` scanning '' the delay . repeating this measurement for a large number of baselines then allows a more or less complete reconstruction of the sky brightness distribution , depending on its complexity . in appendix
a we describe in more detail the operation of the interferometer in terms of fourier optics . here , we concentrate on the performance of sim for astronomical targets . the relative length of the baseline vector is monitored and controlled with an internal laser metrology system , while its exact length and orientation are determined by observing nearby bright stars . pointing stability
is achieved with two additional guide interferometers .
the science interferometer is aligned such that the phase center ( zero delay , or white light fringe ) coincides with the position of a reference star .
the delay lines are adjusted such that the phase center moves to the science target which has to lie in the 15 field of regard of sim . in the case of astrometric measurements of a point source _ position _ ,
the only interesting observable is the delay needed to maximize the white light fringe .
this determines the position of the source with respect to the reference star .
identification of the white light fringe is unambiguous since the light is spectrally dispersed : only the central fringe will show no phase shift for all spectral channels . in the case of an extended source that is to be imaged ,
the complex visibility is measured with the delay modulation method ( @xcite ) or some variation thereof , which we describe in appendix b. since the signal is spectrally dispersed , the complex visibilities are measured simultaneously for all spectral channels .
this is the basis for the spectral synthesis technique that we describe in section [ specsyn ] .
baseline length , orientation , and wavelength determine the -coordinate to which the complex visibility is assigned .
the wavenumbers @xmath1 and @xmath2 are defined as @xmath3 and @xmath4 , where @xmath5 and @xmath6 are the baseline projections onto the @xmath7 and @xmath8 axes of the reference frame .
the measurement is repeated for as many baselines as necessary to sufficiently fill the -plane .
the image can then be recovered from the -data by fourier - inversion , and spurious responses removed by computer restoration .
the spatial resolution of a synthesis imaging observation is not a priori defined .
a diffraction - limited filled aperture with a diameter of @xmath9 operating at a wavelength of @xmath10 has a central response fwhm of 1.22 @xmath11 .
in contrast , the final resolution of an image reconstructed from a set of interferometer measurements is not uniquely defined . by varying the weighting of the individual -coordinates , or even removing some baselines completely
, one can put more or less emphasis on certain spatial frequencies in the reconstruction , and thus change the effective resolution , even _ after _
the data are taken .
for example , using a gaussian weighting which falls to 0.25 at the longest ( 12 m ) baseline results in a fwhm of 7 mas with very low sidelobes .
another option is to use only a narrow ring of long baselines which yields a resolution almost twice as high as that of the full telescope .
this fact is regularly exploited in `` aperture masking '' observations of bright double stars where collecting area is not an issue , but the highest possible resolution is required .
therefore , sim will yield a resolution for synthesis imaging anywhere between 5 and 10 mas , depending on the actual distribution of measurements in the -plane , and the weighting algorithm used .
the field of view ( fov ) is limited by the coherence length of the light , @xmath12 , where @xmath13 is the coherence time of the light , @xmath14 is the speed of light , and @xmath15 is the bandwidth of the filter used for the observations .
it can be easily calculated as the product of the spectral and spatial resolutions , which can be seen as follows : while the path lengths from the two interferometer elements are assumed to be equal in the center of the field , an angular separation @xmath16 from the center along the baseline vector @xmath17 results in a path difference @xmath18 . as long as this difference is small compared to @xmath19 , the fringe contrast will not be affected .
thus , @xmath20 , or @xmath21 .
for example , a 10 nm wide channel at @xmath22 would result in a fov of 0.6 .
any single baseline has a fov that forms a 0.6 wide , long strip on the sky that is positioned perpendicular to the baseline .
however , as the baseline rotates , sources along the strip more than 0.6 away from the center will appear and disappear in the fov , thus affecting the measurements in an unpredictable way . in order to avoid this confusion
, sim will have a circular aperture stop in its optical path .
there is yet another constraint that sets even tighter limits on the fov , namely , the need to avoid aliasing . especially for a complicated source structure or crowded field
this can be a major limitation on any deconvolution algorithm , since the sum of the grating rings of all the sources in the field can and will mimic fake sources , as will be demonstrated in section [ clumps ] .
unless a priori knowledge about the source structure exists , aliasing must be avoided for reliable source reconstruction .
this means that in practice the sim fov is limited to a circle with a diameter of 0.2 , which can be seen as follows .
the position of the first grating ring of a point source in the field is determined by the spacing of the baselines . in the case of 0.5 m increments ,
the grating ring has a _ radius _ of 10/0.5=20 times @xmath23 or 0.2 . to make sure that no aliasing occurs from any source inside the fov , it has to be limited to a _ diameter _ of 0.2 .
all of our simulations in this paper , however , have been carried out with a fov of 0.3 in order to demonstrate the above effects .
figure [ fig1 ] shows a possible -coverage obtained with 170 measurements and the resulting point spread function ( psf ) or `` dirty beam '' . here , and
in all the simulations that are described in this paper , we have set the longest baseline to 10 m , which corresponds to the requirement of the original sim design . the current design offers a maximum baseline of 12 m. the grating rings in the psf described above are clearly visible . the number of samples in the -plane in fig
. [ fig1 ] is twice the number of measurements made .
this is because any astronomical object has a purely real brightness distribution , so the mutual coherence function of the incoming wavefront is hermitian
. therefore , once the complex visibility of point has been measured , its conjugate can be entered at point ( @xmath24,@xmath25 ) .
in addition , different from the case of radio interferometry , the total flux of the source is determined during every visibility measurement .
this gives the value at the origin of the -plane ( the `` zero spacing '' ) to high accuracy .
the `` noisy '' residuals around the central peak of the psf are due to the incomplete -coverage .
since , however , the -coordinates of the sim measurements are defined with extremely high precision , the shape of the psf can be _ calculated _ very accurately and without the uncertainties and noise of an actual measurement .
it is this precise knowledge of the psf that makes sim such a powerful imager , since the success of deconvolution algorithms like clean or the maximum entropy method ( mem ) is limited by the knowledge of the psf .
the accuracy of the `` raw '' synthesized image before deconvolution depends in a complicated way on several parameters , including the noise properties and flux distribution of the source , the stability of the phase measurements , and the actual -coverage . in order to investigate these issues in detail ,
we have developed a software toolkit which we will describe in section [ simsim ] .
jumping ahead , we list here the results of the simulations described in sections [ bh ] and [ exo ] in order to provide an estimate of typical integration times needed to perform synthesis imaging of extended astronomical objects .
the surface brightness threshold for making scientifically useful images with sim in the continuum band from 500 - 800 nm is @xmath26 @xmath27 0.3 mjy/ which corresponds to @xmath28 = 17.7/. under the same conditions
an emission line such as can be imaged in a narrow ( 4 nm was assumed in the simulations ) band with a threshold surface brightness of @xmath29 ergs/@xmath30/sec/. a full synthesis with sim will take typically 4 6 hours of on - target integration time for sources at these brightness levels . during this time , 150 200 different baselines and baseline orientations will be observed .
observations can of course be repeated to increase the sensitivity .
these sensitivities can be achieved even in the presence of point sources up to v=12 in the field of view using the nulling imaging mode discussed in the next section . because of its unprecedented baseline stability , sim has great potential for interferometric nulling of point sources .
one of the sim beam combiners will introduce an achromatic 180 phase shift in one of the inteferometer arms by polarization inversion , thus eliminating the light from a point source that is located at the phase center .
this capability is a programmatic technology demonstration requirement .
the experience gained is crucial for the design of future space missions that are aimed at the detection of extra - solar planets , like e.g. the terrestrial planet finder ( tpf ) .
the requirement for sim is to reach a nulling efficiency of @xmath0 .
many interesting science programs intend to use this capability to determine the spatial extent of objects like stars or supernova shells by measuring the light leakage around the phase center . in this paper
, we will describe another mode of observation , namely synthesis imaging in nulling mode . in section 5
, we will discuss the specific features of this technique for the case of exo - zodiacal dust disks . since sim is designed to perform astrometric measurements in a dispersed fringe mode
, it has a built - in spectroscopic capability .
that is to say , the amplitude and phase of an object s wavefront are monitored simultaneously at a number of different wavelengths .
since the same physical baseline translates into different -coordinates at different wavelengths , this information can , in principle , be used to increase the -coverage .
this method is usually referred to as spectral synthesis imaging .
spectral synthesis has one major caveat :
the implicit assumption in the reconstruction of the source is that its _ structure does not change with wavelength_. in the optical regime this assumption is far from safe .
in fact , it is often plain wrong , especially when strong emission lines are present in the bandpass considered .
that does not mean that one can not use spectral synthesis with sim at all .
there are classes of objects for which the above assumption is acceptable .
for example , in the case of zodiacal disks around solar - type stars , it may be reasonable to assume that the structure of the source does not vary strongly over the sim passband , since it is observed only in reflected starlight .
we will present the results of spectral synthesis simulations for exo - zodiacal disks in section [ exo ] .
in order to simulate the imaging capabilities of sim , we have developed , a software package written in the interactive data language ( idl ) environment .
the motivation was to obtain realistic performance estimates for a variety of target sources , and to develop optimal observing strategies .
the principal steps carried out during a simulation are illustrated in fig .
[ fig2 ] . in brief , a model source distribution is put on a finely - sampled grid ( typically @xmath31 pixel ) . for simplicity
, the field center is assumed to be the center of rotation of the spacecraft during the synthesis imaging , and no effort has been made so far to account for possible pointing instabilities .
the mutual coherence function of the source wavefront is calculated via a fast fourier transform ( fft ) , and sampled on the -coordinates observed by sim .
noise is added , and the actual phase and amplitude retrieval process is simulated .
the details of this algorithm are described in appendix b. so far , we have incorporated photon noise according to poisson statistics , and random phase calibration errors with an amplitude of up to 0.5 .
a final fourier inversion of this sampled and corrupted data then gives the `` dirty '' image .
the psf is calculated by transforming a mutual coherence function of unit amplitude and zero phase at each -sample point of the observation .
deconvolution algorithms like clean or the maximum entropy method ( mem ) can then be applied to minimize the effects of the incomplete -coverage .
the nulling capability of sim is modeled as follows . before the noise
is added to the source wavefront , its amplitude @xmath32 ( at each -coordinate ) is reduced by most of the amplitude @xmath33 of the source at the phase center @xmath34 : [ null_eq ] a_nulled = a - a_c(1- ) where @xmath35 denotes the accuracy to which nulling can be achieved . for sim , @xmath35 is expected to be around @xmath0 .
it is limited by stray light due to imperfections in the optical elements and by the stability of the pathlengths .
the phase of the wavefront is set to zero , resulting in an intrinsically symmetric source reconstruction .
the reason for the loss of phase information is that the delay can not be modulated without losing the null : once the light is minimized , the delay lines have to be kept as stable as possible .
the reduced amplitude of the signal results in lower photon counts on the detector which , in turn , reduces the poisson noise .
equation [ null_eq ] ensures that the total flux in the reconstructed image is equal to the flux in the whole field - of - view * except for the central pixel * which , in the simulations , contains the flux of the nulled star .
this is , of course , known for the models .
however , in a real observation , such _ a priori _ information on the exact value of @xmath33 does not exist .
it is only possible to measure the total flux including the central source by doing a standard observation in the direct beam combiner .
an iterative approach will be necessary , which adds an appropriate * constant * to each amplitude measurement ; this constant is an estimate of the total flux in the field , minus the flux from the central source .
if this value is chosen correctly , the reconstructed image will be zero at the phase center .
since the model sources are intrinsically constructed on a regular grid , as are the -coordinates of the `` measurements '' , there is no need for gridding procedures in our code .
however , real sources and real observations will not be so conveniently organized , so that in order to fully model the sim observations , weighting and gridding algorithms will have to be implemented .
this work is currently underway .
figure [ fig3 ] illustrates the graphical user interface ( gui ) of .
the top panel allows the user to select between a number of model sources and to choose various observational parameters of sim such as -coverage , spectral resolution , etc .
the intermediate steps of the simulation are displayed in the four graphics windows .
all results can be saved in postscript format for printing or in fits format for further work with idl or any other data analysis environment . in the following sections we use to model the capabilities of sim to image two interesting classes of astronomical objects .
black holes have been detected convincingly in at least a dozen galaxies , both active and quiescent .
the techniques for such detections are multifold ( see @xcite for a review ) , but have concentrated on the dynamics of the gas and stars in the neighborhood of the black hole .
stellar kinematical studies are often hampered by uncertainties regarding the stellar velocity dispersion anisotropy , and the need for high spatial resolution in connection with intrinsically low surface brightness makes such measurements very challenging even with hst .
however , some galaxies have circumnuclear disks of ionized gas which provide a more favorable tool for black hole detection .
the emission lines from the ionized gas are bright and have large equivalent width , and the dynamical state of the gas is relatively simple ; its rotation curve directly yields the black hole mass . as a starting point for our discussion , we have chosen m 87 as the best known example for such a line - emitting gaseous disk .
the next section briefly describes our input model for m 87 .
m 87 , a dominant giant elliptical galaxy in the virgo cluster , is the prototype for galaxies that harbour a central black hole , the presence of which is revealed by a line - emitting gaseous disk . the nucleus of m 87 has been imaged with hst in both continuum ( @xcite ) and line emission ( @xcite , and the ionized gas has been studied spectroscopically in great detail ( @xcite ) we thus have a fairly clear picture of the properties of the nuclear gas disk .
table [ tab1 ] summarizes the parameters we used to construct a realistic model flux distribution for the gaseous disk around the central black hole in m 87 .
the surface brightness over the central 0.25 of our model is @xmath36 ergs/@xmath30/ , as derived from the observed [ sii ] flux ( @xcite ) and an assumed line ratio of [ sii]/(+[nii ] ) = 4 .
the gas is assumed to show keplerian rotation according to [ rotvel_eq ] ( ) ^2=0.08869 where denotes the black hole mass , @xmath37 the angular distance from the center , and @xmath38 the distance to the galaxy .
the numbers listed in table [ tab1 ] then result in the rotation curve shown in fig .
[ fig4 ] .
the emission from the rotating disk would appear in a range of spectral channels , according to the doppler shift of the gas velocity projected along the line of sight .
for the above parameters , a ( hypothetical ) diffraction - free 10m - telescope in space , equipped with an integral field spectrometer with 4 nm wide channels , would see the source as in fig .
[ fig5 ] . here
, we have convolved the `` true '' flux distribution of the m 87 model with the `` clean beam '' , a gaussian fitted to the central part of the psf .
while the central , unresolved continuum source appears in all channels , the appearance of the emission depends on parameters like channel width , black hole mass , inclination of the disk , etc . to demonstrate the power of our analysis tools , we present in the next section what can be expected from sim observations of m 87 and similar sources . in fig .
[ fig6 ] , we present a few examples of our simulations that are intended to give an impression of what imaging with sim can do in this context .
the first row of the left panel again shows those channels of our m 87 model containing emission .
the channels again are 4 nm wide and centered at 554 , 558 , and 662 nm , respectively .
the second row shows the result of the direct fourier inversion of the complex visibilities measured by sim .
the -coverage was the same as in fig . 1 , and the integration time was 100s per measurement , for a total on - source integration time of about 5 hours .
the nominal sim throughput of 0.3 was assumed .
we added photon noise to the measurements according to poisson statistics .
the third row shows the results after 200 iterations with the mem algorithm .
the emission structure of the high velocity gas is well recovered .
the same is true for the intensity profile , as shown in fig .
[ fig7 ] .
the spatial extent of the emission in each channel provides valuable information on the velocity gradient , and thus the black hole mass .
this is demonstrated in fig .
[ fig8 ] , where we have run the same simulation as described above , but for various black hole masses .
we show only the blueshifted wing of the emission for values of = 2 , 3 , and 4@xmath39 .
the sim observations are very sensitive to the value of , a direct consequence of the linear relationship between black hole mass and radial distance in a keplerian rotation curve , cf .
. [ rotvel_eq ] .
however , it should be stressed that in practice fitting procedures are required to solve for all the free model parameters including black hole mass , disc inclination , and the actual distribution of disk surface brightness .
we emphasize that no attempt has been made to optimize the deconvolution process .
for example , the stability and convergence of mem can be greatly improved by using `` a priori '' knowledge about the source structure . however , in order to better be able to compare the results for different source models , we have for all the simulations in this paper started the mem deconvolution with an empty array , and stopped after 200 iterations .
for the real sim observations , one would make every effort to optimize the deconvolution strategy for each individual dataset .
this is , in fact , another area where future simulation work with will be useful .
however , it is our experience that if no trace of the source structure is evident in the direct fourier inversion , no improvement can be expected from deconvolution , regardless of the algorithm employed .
thus , the quality of the data can be most directly judged from the direct fourier inversion itself , before any restaoration .
an important goal of is to define requirements on various aspects of the sim spacecraft .
as an example , what is the minimum spectral resolution of the sim beam combiner that allows observations such as the one discussed here ? certainly , the channel width must not be smaller than the wavelength shift of the emission line under observation at a distance of at least one resolution element from the center .
this , of course , depends on the black hole mass and the distance to the host galaxy . in fig .
[ fig9 ] , we have plotted the limiting black hole mass ( i.e. the minimum mass that causes a shift of the emission of at least one spectral resolution element at an angular separation of two spatial resolution elements ( 0.02 ) from the center ) as a function of the distance to the host galaxy for a number of possible sim spectral resolutions and disk inclinations . for example , with 1 nm wide spectral channels ( dotted lines ) , sim can potentially detect @xmath40 black holes up to a distance of @xmath27 80 mpc , if the disk has an inclination @xmath41 .
whether the -flux in cases different from m 87 is sufficient to acquire a high signal - to - noise ratio in a reasonable amount of time is more uncertain .
circumstellar dust disks are commonly thought to provide the material for planet formation .
the structure of such disks is therefore important for our understanding of the processes that lead to the formation of planetary systems .
many searches have been started for such disks , most of which are based on coronographic observations and concentrate on stars with a high ir - excess as measured by the iras satellite ( e.g. @xcite ) . until most recently , however , only one unambiguous detection of a circumstellar dust disk outside the solar system had been reported , namely the system . from modeling the disk
, @xcite found that it is the uncommonly large scattering cross - section of the dust particles that makes a special case .
if there were other nearby systems like , they would have been detected regardless of the disk inclination .
@xcite conclude that at least an order of magnitude improvement in the suppression of the stellar light is needed to detect more `` normal '' circumstellar disks , i.e. with scattering cross section less than a tenth of that in .
sim , with its interferometric nulling capability , differs from groundbased coronagraphic observations in a number of ways : the starlight suppression in sim is not limited by atmospheric seeing or the efficiency of an adaptive optics system , but only by the quality of the optical elements and the pathlength stability .
the high spatial resolution of sim allows the detection of disks much smaller than that in , and sim can therefore image such systems to greater distances @xmath38 .
in fact , sim s ability to detect exozodiacal disks increases with @xmath38 , as long as the disk fills the 0.2 fov of sim .
this somewhat surprising behavior is due to the fact that the surface brightness of the disk varies only slightly with @xmath38 , so its total signal on the sim detector will stay more or less constant .
the signal from the star , on the other hand , decreases as @xmath42 , so that the noise contribution from the nulling residuals of @xmath0 become less important .
in addition , the further away the star is , the more it resembles a true point source .
this will reduce light leakage around the null , which is also a source of photon noise even for a perfect nulling beam combiner .
since all phase information in the nulling process is lost , the source after reconstruction _ always _ appears symmetric .
that means that eventual asymmetries in the disk can not be recovered .
aliasing due to the incomplete sampling of the mutual coherence function can sometimes mimic companion stars or planets . in what follows
, we investigate the sim potential for the detection of zodiacal disks and specifically address the points listed above .
we use a simple model of a circumstellar dust disk , based on the zodiacal emission of the solar system ( @xcite ) .
the particle number density of the solar system zodiacal dust cloud can be well modeled by the expression = _ 0 r^- e^-(z / r)^ where @xmath37 is the in - plane radial distance from the center , @xmath43 is the modulus of the perpendicular - to - plane distance ( in au ) , @xmath44 , @xmath45 , and @xmath46 .
we treat the scattering efficiency of the dust particles as a flux scaling parameter , and we assume isotropic scattering . in what follows ,
we investigate the two limiting cases of a face - on ( @xmath47=90 ) and edge - on ( @xmath47=0 ) orientation of the disk .
the radial surface brightness distribution of the face - on disk can be derived from integrating the volume density .
for @xmath48 it is given by the expression _
( i=90^ ) = c_1 r^-+ 1 . here
, @xmath49 is a scaling factor that characterizes the scattering efficiency of the dust particles in an exo - solar system relative to the solar system .
the normalization constant @xmath50 is the surface brightness of the face - on disk , as seen from outside the solar system , at a distance of 1 au from the sun .
assuming isotropic scattering , this number is ( at @xmath51=550 nm ) _ 1 = 0.13 mjy / sr
. an approximate ratio of the surface densities of the edge - on and face - on directions at a given value for @xmath37 can be obtained by ratioing the characteristic scale heights of the @xmath43-direction and along a tangent to @xmath37 in the plane : e / f = ^1/ 5 .
an approximation to the edge - on surface brightness distribution is therefore [ eq1 ] _ ( i=0^ ) = ce / f_1 r^-+ 1e^ . where @xmath52 is used .
figure [ fig10 ] gives a visual impression of the model in both the face - on and edge - on case .
the spectrum of the dust in the wavelength range discussed here ( 500 1000 nm ) is assumed to be that of a black body with the temperature of the central star . in this section ,
we investigate the question whether sim is capable of finding other systems like , should they exist . in this context , we define `` -like '' as having about 1000 times the dust content and surface brightness of the solar system , i.e. @xmath49=1000 , without any detailed modeling of the specific properties of itself . as mentioned above
, the sim performance in terms of imaging of circumstellar disks is very limited for nearby systems , but improves with increasing distance , as long as the disk fills the fov .
for example , the disk itself is impossible to image with sim , because even with @xmath53 nulling , the photon noise of the remaining light from would make the detection of the faint disk emission impossible .
figure [ fig11 ] demonstrates the increasing sensitivity of sim up to distances of 1 kpc .
again , we note that no attempt has been made to optimize the restoration procedure in order to avoid additional degrees of freedom . as can be seen in figure [ fig11 ] , sim can easily detect -like systems up to a distance of about 1 kpc , beyond which our model disk no longer fills the sim fov and its total flux drops rapidly .
this distance limit depends linearly on the extent of the disk , for which we have assumed a value of 100 au .
this is a somewhat arbitrary number , chosen such that the angular diameter at a distance of 1 kpc is 0.2 , equal to the sim fov .
the central star is assumed to be of solar type ( @xmath54=5 ) .
the above results change dramatically if systems with lower dust contents than are considered .
the dust content of our model disks is parametrized by the constant @xmath49 in eq .
figure [ fig12 ] summarizes our results on studying various values for @xmath49 .
disks with more than a tenth of the dust content of can be detected with sim .
however , it is evident that zodiacal disks similar to that of the solar system will not be found with sim , regardless of their distance or inclination .
it is interesting that in all of our reconstructions , the disk structure after deconvolution appears much more clumpy than the smooth input model .
in fact , when looking at the results , one might in many cases think to have found a number of planets .
these `` clumps '' are artefacts of the restoration process on our incomplete -data .
while the sidelobe pattern itself is known to high accuracy , the poisson noise associated with the signal does not have the shape of the psf .
spikes in the noise distribution create residuals that are mistaken as real source structure by any deconvolution algorithm .
the fact that the clumps are indeed randomly distributed can be seen when comparing three independent simulations of the face - on zodiacal disk of a -like system at 1 kpc and a nulling efficiency of @xmath55 in figures [ fig11],[fig12 ] , and [ fig13 ] , which only differ in their random photon noise .
a repetition of the observation would allow identification of such spurious objects . in order to investigate the effects of a degradation of the sim nulling capability
, we have run a series of simulations with different values for @xmath35 , the nulling efficiency in eq .
[ null_eq ] .
figure [ fig13 ] demonstrates that for a -like system at 1 kpc , the most favorable case in all of the simulations we have run so far , sim could detect the disk even if it had only @xmath56 .
however , anything worse than that would not suppress the starlight sufficiently even at such high distances .
we conclude that in order to be a powerful search engine for exo - zodiacal disks , sim needs to achieve at least @xmath57 , and anything better would improve the performance considerably . a massive planet
if present is believed to sweep up any inward - spiraling debris and thus cause a dust - free region in the circumstellar disk inward of the planet s orbit .
such an `` inner gap '' is believed to exist in the -system , as inferred from its spectral energy distribution ( @xcite ) , but the gap has never been directly imaged owing to the large radius which is masked out by coronographic instruments . with its high spatial resolution and sensitive nulling ,
sim is able to directly image possible inner clear regions in exo - zodiacal systems .
this is demonstrated in fig.[fig14 ] where we have investigated the case of a -like disk at a distance of 500 pc with and without a 20 au inner clear region .
clearly , the signature of the disk depends on its size and the distance to the system , but a 20 au radius gap would be detectable in systems as distant as 1 kpc . here , we just want to illustrate the intriguing prospect of directly imaging such gaps ( and later , possibly the planets that cause them ) .
we have built an interactive software package that enables us to perform numerical simulations of imaging observations with sim .
the results obtained so far can be summarized as follows : sim , with a spectral resolution of 1 nm , can resolve the velocity structure of line - emitting gaseous disks around black holes in the centers of nearby galaxies down to black hole masses of @xmath40 at distances up to 100 mpc . with this resolution , sim is in principle capable of determining the mass of the black hole to 10% accuracy .
the necessary integration time depends on the -flux from the disk ; for cases like m 87 , about 5 hours is sufficient .
a coarser spectral resolution of sim would constrain the accessible parameter space for quantities such as distance , black hole mass , or disk inclination . given a nulling efficiency of @xmath0 , sim can image zodiacal disks with dust contents similar to that of the -system with high signal - to - noise ratio in about 5 hours of integration time .
the limiting distance depends only on the angular size of the disk , which should be no smaller than 0.1 .
inner clear regions in the disk can be directly detected and imaged with sim for systems at a distance of up to 1 kpc provided that the dust - free region has a radius of at least 20 au .
we are grateful to r. van der marel for his advice on the black hole simulations , and to w. traub for his expertise on the zodiacal disk model .
we also would like to thank s. unwin , m. colavita and m. shao for their interest in and continued support of this project . the work described in this paper
is carried out at stsci with financial support from the sim project at the jet propulsion laboratory ( jpl ) .
the idea that an interferometer could be constructed to measure the angular diameters of stars is usually credited to @xcite , although no details were provided in that particular reference .
twenty years later , @xcite took it up and , another 30 years later with pease , constructed the now - famous stellar interferometer on the 100-inch telescope ( @xcite ) .
figure [ fig15 ] shows a sketch of the original drawing from that paper , to which we have added symbols indicating the spacing @xmath58 of the outrigger mirrors and the spacing @xmath59 of the mirrors at the telescope aperture which direct the light to the primary mirror .
variations of this sketch appear in many introductory textbooks on physics and astronomy .
the image of a star in the focal plane of the stellar interferometer is crossed by interference fringes , as shown in cross section in fig.[fig16 ] , where some additional parameters are defined : the _ fringe amplitude _ @xmath32 , the _ fringe period _ @xmath60 and the _ fringe phase _ @xmath61 . in this picture the average signal from the star has been normalized to 1 .
the overall size of the airy pattern is set by the aperture formed by the cascade of small reflecting mirrors which direct the light to the telescope primary mirror . as is well known from the general theory of wave diffraction ,
this fraunhofer diffraction pattern in the focal plane of the telescope is the square modulus of the fourier transform of the distribution of illumination ( in amplitude and in relative phase ) over the aperture .
the expression for the stellar interferometer can be derived in a straightforward way , at least in one dimension , by starting from the result for the single - slit diffraction problem .
we then use standard theorems from the theory of fourier transforms , in particular the shift theorem and the convolution theorem , and apply them to a plane wavefront over the entrance aperture ( pupil ) . in this way it is straightforward to show that the final fraunhofer pattern for an unresolved star is the square modulus of @xmath62 with @xmath63 , and @xmath64 , @xmath61 being an angle as defined in fig.[fig16 ] .
this is the product of the pattern @xmath65 for a single slit of width @xmath66 offset to @xmath67 , and the pattern @xmath68 for a very thin double slit offset to @xmath69 , where @xmath70 is the angular distance of the star from the direction normal to the baseline @xmath58 .
we note that : * as the source position @xmath70 changes , the fringe phase moves at an amplified rate @xmath71 .
the positional accuracy is therefore _ magnified _ by @xmath72 . *
the fringe period remains set by the separation @xmath59 of the secondary apertures no matter what is the separation @xmath58 of the primary apertures on the outriggers .
sim can be regarded as a michelson stellar interferometer with two specific features : * the separation @xmath59 of the secondary apertures is reduced to zero which we ignore here . ] .
the fringe period therefore becomes infinitely large , and the entire image of the star is either light or dark depending on the fringe phase . * a delay @xmath73 is added to one side of the interferometer light path ( e.g. between m@xmath74 and m@xmath75 in order to shift the double - slit pattern from @xmath76 back to within a small offset @xmath77 from the position of the star @xmath78 .
this is necessary because of the finite bandwidth of the signal , which results in a limited coherence length , as discussed in section [ fov ] .
we will discuss the bandwidth effects on the fringe contrast later in this section .
the fraunhofer pattern for a monochromatic point source in our one - dimensional model of sim is therefore the square modulus of : [ eqa2 ] m(u ) = 2 f(u - u_0 ) ( ( l_0 - ) / ) , which is [ pointsource ] p(u ) = |m(u)|^2 = p_0(u)(1 + ( 2/ ) ) where @xmath79 .
viewed as a function of the delay offset @xmath77 , this is a co - sinusoidal pattern with an average value set by the intensity @xmath80 of the star is ignored further here .
it comes in because we have doubled the collecting area by using two outrigger mirrors , and it will be subsumed into @xmath80 by the overall interferometer calibration procedure . ] , and varying from zero to @xmath81 .
the fringe amplitude here is 1 , appropriate for a point source .
consider now an extended source in the sky , which we integrate in a direction perpendicular to the interferometer baseline in order to produce a 1-d image @xmath82 where @xmath61 is an angle measured in the plane of the baseline ( fig .
[ fig17 ] ) .
let us further assume that the position @xmath70 to which we have initially offset the fringe pattern from our calibration grid star ( by adding a delay @xmath83 ) is somewhere in the middle of @xmath82 .
each part @xmath84 of the spatially - extended source then contributes a fringe pattern : [ fringecontrib ] dp ( ) = b()d(1 + e^-2i(l- ) / ) ) ; where again @xmath85 and @xmath86 denotes the real part of the exponential .
let us further simplify the notation by assuming that the fov is confined to a small region on the sky , so that @xmath87 , and measure @xmath58 in wavelengths , @xmath88 .
the total fringe pattern obtained by integrating equation [ fringecontrib ] is then : p = b_0 + _ -^+b()e^-2i ( l - l_0-/)d .
we can write the integral in this equation as : b_0 e^2i(l_0+/ ) ( l ) , where we recognize @xmath89 as the fourier transform of the normalized @xmath82 : ( l ) = .
since @xmath82 will in general not be symmetric about @xmath70 , @xmath89 will be complex , so we can write @xmath90 , leading to : [ spacefringe ] p = p_0 \ { 1 + a(l ) ( 2(l_0 - ( l ) + /))}.
there are 3 unknowns here : the average power @xmath80 , the fringe amplitude @xmath91 , and the fringe phase @xmath92 ( measured here in turns ) .
they can all be determined by measuring @xmath93 as a function of delay @xmath77 . in practice , this can be achieved by continously scanning @xmath77 over @xmath94 , and recording @xmath93 over a number of time intervals .
this is the basis for our simulation of the measurement process and the introduction of photon noise , as will be described in appendix b. let us now consider the effects of the finite bandwidth of the interfering light . in order to keep the algebra manageable , we consider only the case of a plane wave from a point source .
the fringe pattern we have derived in equation [ pointsource ] will be different for different values of the wavelength @xmath51 .
the total response is the sum of the fringe patterns for all wavelengths over the spectral range of the passband .
suppose the passband has a spectral shape @xmath95 .
then , in analogy to equation [ fringecontrib ] , that part of the spectrum at frequency @xmath96 contributes a fringe : [ spectrafringe ] dp ( ) = s()d(1 + e^2i(l_0 - ) / ) . where @xmath97 .
we define a delay time @xmath98 where @xmath14 is the speed of light .
then we have : dp ( ) = s()d\{1 + e^-2i}. the total fringe pattern is obtained by integrating : p = s_0 + _ -^+s()e^-2id .
we recognize the integral as the fourier transform @xmath99 of the filter function @xmath95 .
as an example , let s take a rectangular filter of height @xmath100 and width @xmath15 centered at @xmath101 .
the transform of this function is related by the shift theorem to the transform of a function @xmath102 having the same shape but centered at the origin .
the transform of @xmath102 is @xmath103 , so the transform of @xmath104 is @xmath105 , and the fringe pattern becomes : p ( ) = s_0 \{1 + ( ) ( 2 _ 0 ) } , so that if we scan @xmath106 and record the intensity of the stellar image in the focal plane of sim we will obtain the pattern shown in figure [ fig18 ] .
in order to provide an estimate of sim sensitivity and image quality , various noise components associated with the measurements need to be modeled as realistically as possible .
the systematic ( calibration ) errors of the spacecraft are hard to predict at this time , because the design has not been finalized and a detailed hardware model is not yet available . as of now
, we can only make assumptions on parameters such as fringe phase stability . on the other hand , the statistical errors due to photon statistics are straightforward to model .
they are implemented in as follows .
the fundamental shape of the signal as a function of path difference ( delay ) between the two interfering beams has been derived in equation [ spacefringe ] . here
, the term `` signal '' means the total number of photons falling on the detector over the integration time .
if the delay @xmath77 is continously modulated over @xmath94 , then one can integrate the detector signal over four time `` bins '' , during each of which the delay moves over a quarter wave .
this method has first been described by @xcite and was suggested for use in astronomical interferometers by @xcite .
it is usually referred to as the `` quadrature method '' and has been used in many ground - based interferometers .
the usual implementation uses a triangular modulation of the delay with piezo - electric elements ( e.g. @xcite ) . in ground - based interferometers
the delay is scanned over the range of one wavelength in about 10 ms owing to the short coherence time of the atmospheric seeing , which otherwise moves the fringe phase @xmath61 in a random way . in principle , there is no need for high modulation frequencies in space , so that instead of a continous modulation , one could envisage a `` stop - and - stare '' mode for the delay @xmath77 .
however , since the continous modulation technique is well established and simple to handle mathematically , we have used it to simulate the measurements in . following @xcite , the signal in the four bins @xmath107 ... @xmath108 which , for mathematical simplicity have the limits @xmath109$],@xmath110 $ ] , @xmath111 $ ] , and @xmath112 $ ] can calculated by integrating equation [ spacefringe ] over these intervals , yielding : @xmath113 @xmath80 here is the total signal ( in photons ) in the fov over the spectral passband during the integration time . in the case of a simulation , the values for the fringe amplitude and phase , @xmath114 and @xmath115 , are known from the fourier transform of the model source structure .
therefore , the values @xmath116 can be calculated , and poisson noise can be added to them , before using the `` measurements '' to recover the source .
these values are re - entered in an empty array ( the -plane ) , and the ( sparsely - filled ) array is fourier back - transformed to yield the reconstructed image of the source . as an option for the user , an additional noise component can be added that tries to model the expected phase calibration errors .
our simulations have shown , however , that up to the specified maximum error of 0.5 , this has negligible effects on the restoration quality .
since the psf can also be computed , the recovered image can further be improved by applying restoration algorithms like clean or mem .
allen , r. j. , peterson , d. , & shao , m. , 1997 , spie proc .
2871 , 504 crane , p. 1993
, , 106 , 1371 fizeau , a. h. 1868 , c. r. acad .
( paris ) , 66 , 943 ford , h. c. 1994 , , 435 , l27 harms , r.j .
1994 , , 435 , l35 lagage , p.o .
& pantin , e. 1994 , nature , 369 , 628 lauer , t. r. , 1992 , , 103 , 703 macchetto , f. , marconi , a. , axon , d. j. , capetti , a. , sparks , w. b. , & crane , p. 1997
, , 489 , 579 marconi , a. , axon , d. j. , macchetto , f. d. , capetti , a. , sparks , w. b. , & crane , p. , 1997
, , 289 , l21 michelson , a. a. 1890 , phil . mag .
, 30 , 1 michelson , a. a. & pease , f.g .
1921 , , 53 , 249 kalas , p. & jewitt , d. 1996 , , 111 , 1347 kormendy , j. & richstone , d. 1995 , , 33 , 581 reach , w. t. , franz , b. a. , kelsall , t. , & weiland , j. l. 1996 , in aip conf .
348 , `` unveiling the cosmic infrared background '' , ed .
e. dwek ( new york : aip ) , 37 shao , m. & staelin , d.h .
1977 , j. opt .
am . , 67 , 81 shao , m. 1988 , a&a , 193 , 357 smith , b.a . , fountain , j.w . , &
terrile , r.j .
1992 , a&a , 261 , 499 traub , w. a. 1998 , in `` proc . of conference on exo - zodiacal
dust '' , ed .
backman , d.e .
nasa / cp-1998 - 10155 , 129 wyant , j. c. 1975 , appl .
11 , 2622 lcccl stellar
background & @xmath118 = 15.0 mag/ & 1 + central point source & i = 16.1 & 1 + h@xmath119(+ [ nii ] ) disk flux & @xmath36 ergs/@xmath30/s/ & 2 + radial intensity profile & @xmath120 & + black hole mass & @xmath121 & 3 + distance & 16 mpc & + inclination & 45 & + position angle & 0 & + grid resolution & 0.005 & | we present numerical simulations for a possible synthesis imaging mode of the space interferometer mission ( sim ) .
we summarize the general techniques that sim offers to perform imaging of high surface brightness sources , and discuss their strengths and weaknesses .
we describe an interactive software package that is used to provide realistic , photometrically correct estimates of sim performance for various classes of astronomical objects .
in particular , we simulate the cases of gaseous disks around black holes in the nuclei of galaxies , and zodiacal dust disks around young stellar objects .
regarding the first , we show that a keplerian velocity gradient of the line - emitting gaseous disk and thus the mass of the putative black hole can be determined with sim to unprecedented accuracy in about 5 hours of integration time for objects with surface brigthness comparable to the prototype m 87 .
detections and observations of exo - zodiacal dust disks depend critically on the disk properties and the nulling capabilities of sim
. systems with similar disk size and at least one tenth of the dust content of can be detected by sim at distances between 100 pc and a few kpc , if a nulling efficiency of @xmath0 is achieved .
possible inner clear regions indicative of the presence of massive planets can also be detected and imaged .
on the other hand , exo - zodiacal disks with properties more similar to the solar system will not be found in reasonable integration times with sim .
simsim # 1=0.6 # 1=0.8 # 1= # 1=1.2 # 1#2=.45 = .45 |
one of the oldest problems in physics is that of determining the motion of @xmath0 bodies under a specified mutual force . commonly referred to as the @xmath0-body problem , it occurs frequently in many distinct subfields and remains an active area of research . when the specified interaction is gravitation the problem is particularly interesting , partly because of obvious astrophysical applications and partly because some basic issues in the statistical behaviour of such systems are still not well - understood .
one - dimensional self - gravitating systems ( ogs s ) continue to play an important role in this regard . even in the simplified setting of one spatial dimension ,
there are still many open questions about the ogs concerning its ergodic behaviour , the conditions ( if any ) under which equipartition of energy is attained , and whether or not it can reach a true equilibrium configuration from arbitrary initial conditions .
furthermore , even for non - relativistic ( newtonian ) gravity , ogs s have proven to be very useful in modeling many diverse physical systems .
stable core - halo structures have been shown to exist in the ogs phase - space that are reminiscent of those found in globular clusters @xcite , in which a dense core of particles near equilibrium are surrounded by a cloud of particles with high kinetic energy that interact very weakly with the core .
the ogs also models the motion of stars interacting with a highly flattened galaxy @xcite and the dynamics of flat , parallel sheets colliding along a perpendicular axis @xcite . a preliminary study of the relativistic case yielded a complete derivation of the partition and single - particle distribution functions in both the canonical and microcanonical ensembles to leading order in a post - newtonian expansion @xcite .
recently non - relativistic ogs s have been shown to exhibit a new phase of evolution in which fractal spatial structure emerges from non - fractal initial conditions @xcite . even for small values of @xmath0 , ogs s exhibit interesting novel behaviour and model interesting physical systems .
the 3-body ogs is equivalent to a system of two elastically colliding billiard balls in a uniform , gravitational field @xcite , as well as to a bound state of three quarks to form a `` linear baryon '' @xcite .
it can be extended to fully include relativistic gravitational interactions @xcite , and investigations have been carried out for both the equal mass @xcite and unequal mass @xcite cases .
furthermore , it is isomorphic to a system in which a billiard elastically collides with a wedge of fixed angle in a uniform , gravitational field @xcite . as such , one can study the 3-body ogs ( non - relativistically and relativistically ) by studying the motion of a single particle moving in two spatial dimensions in a specified potential .
the motion in this case is readily visualizable , and the different types of motion can be classified into three categories : annulus , where each particle always crosses the other two in succession ; pretzel , in which two particles cross each other at least twice in a row before either crosses the third ; and chaotic , where the sequence of particle crossings does not progress in a discernible pattern . in this paper
we carry out an investigation of the 4-body ogs .
analogous to its 3-body counterpart , this system is isomorphic to the motion of a single particle moving in three spatial dimensions in a specified potential
. consequently the 4-body case is of particular interest in that it is the largest value of @xmath0 for which the motion of the system can be directly visualized @xcite ) .
we consider only the non - relativistic case ( which to our knowledge has never been studied ) , leaving the relativistic case for future work .
the outline of our paper is as follows .
we begin with an overview of the problem of 4-body motion in newtonian gravity , and describe a general classification we then close our paper with some concluding remarks .
in ( 1 + 1 ) dimensional newtonian gravity , the hamiltonian of our system of particles is given by : @xmath2 which is simply the sum of the kinetic energies of the particles and the potential interactions between them .
the gravitational potential is determined from @xmath3 where @xmath4 is the mass density of the system , here modeled as a set of four point particles .
in one spatial dimension the solution to this equation is @xmath5 and the potential is given by @xmath6 .
_ _ _ _ the equations of motion are given by : @xmath7 since the momentum is conserved and since the potential depends only on the separations between the particles , there are actually only six _ independent _ degrees of freedom in the 4-body system : the three separations between the particles and their conjugate momenta .
this is easily seen by making the following convenient change of coordinates @xmath8 where @xmath9 .
the conjugate momenta are given by @xmath10 where we have set @xmath11 by conservation of momentum .
without loss of generality the centre of mass of the system can be fixed at the origin .
@xcite in the equal mass case the hamiltonian ( [ newtham ] ) becomes @xmath12\label{hamnew}\end{aligned}\ ] ] which is the hamiltonian of a single particle moving in three spatial dimensions in a linear potential whose shape is that of a 3-simplex .
in general a the @xmath0 particle ogs can be mapped to a single particle moving in @xmath13 dimensions in a linear potential whose equipotential surfaces are that of an @xmath13 simplex .
as such , the 4-body ogs is the largest system for which the motion can be directly visualized
. we shall refer to the form ( [ hamnew ] ) as the hamiltonian of the box - particle . using this result , plot the potential in the equal mass case from equation ( [ hamnew ] ) , defined as @xmath14 which can be drawn in @xmath15 space as shown in figure [ fig : potential ] .
an equipotential surface is that of a cube of pyramid - shaped sides . a cross - section of this surface through any of the edges of one of these pyramids yields a hexagon whose sides are not all of equal length .
this is reflective of the fact that on such cross - sections the problem reduces to that of the three - body problem with unequal masses @xcite since two particles occupy the same position . .
larger values of @xmath16 will produce larger scaled `` boxes '' . ]
we turn now to the problem of solving the equations of motion ( [ newtmot ] ) . as there are no singularities in the potential whenever two particles cross ,
we assume that the particles pass through each other freely upon collision . prior to any collision , solving the equations is trivial : since the acceleration of each particle is constant at any given instant , the trajectory of each particle is a quadratic function of time .
however after each subsequent crossing the acceleration of a given particle changes its magnitude , since the number of bodies to the right and left of it have changed .
hence an analytic closed - form solution to the equations of motion ( [ newtmot ] ) is completely impractical .
we therefore solve the equations of motion numerically , using a matlab ode(ordinary differential equation ) routine to integrate the equations of motion ( [ newtmot ] ) . for the most part the standard `` ode45 '' routine proved to be completely sufficient for our needs .
in some cases we also used the `` ode113 '' routine because of ou to perform the numerical analysis we rescale the variables to have dimensionless values @xmath17 where @xmath18 is the total mass of the system and @xmath19 and @xmath20 are the dimensionless momenta and positions respectively .
the dimensionless mass and hamiltonian are : @xmath21 using the dimensionless variables in ( [ unitrescale1 ] ) and ( [ unitrescale2 ] ) , the hamiltonian ( [ newtham ] ) becomes : @xmath22 as for the equations of motion ( [ newtmot ] ) one gets : @xmath23 where @xmath24 .
@xmath1 a time step in the numerical code has a value @xmath25 .
all the numerical calculations were carried out using the rescaled variables ( [ unitrescale1 ] ) .
henceforth we drop all of the `` hats '' of the dimensionless variables for convenience . note
that since the energy is a constant of the motion , we could further rescale all quantities in terms of @xmath26 , thereby fixing this final redundant scale
. however we shall find it convenient to employ the above rescalings , as it affords us more freedom in choosing initial conditions .
prior to any collision between the particles the evolution of the system is straightforward : each particle moves with a constant acceleration that is proportional to the difference between the total mass on its right and left sides .
however after a collision , where we assume that the particles pass through each other , the mass difference changes , and with it the accelerations of the particles .
it is these repeated changes in the accelerations of the particles that yield the interesting dynamics of the system . from the perspective of the box particle ,
such crossings correspond to the box particle crossing any plane that bisects the 3-simplex through its vertices and edges , yielding a discontinuous change in the box particle s acceleration .
these planes occur in pairs whose line of intersection is along each of the three principal axes , for a total of six such planes .
their equations are given by setting any one of the six quantities in eq .
( [ changevar1 ] ) to zero , and each plane corresponds to the crossing of a pair of particles . for example @xmath27
corresponds to the crossing of particles 1 and 2 , whereas @xmath28 corresponds to the crossing of particles 2 and 4 .
initially we have all of the crossings listed in order as a string of vanishing quantities .
the direction of crossing is irrelevant ; for example 1 crossing 2 is equivalent to 2 crossing 1 . at any given instant we can fix the positions of particles in a certain order from left to right , i.e. @xmath30 , in which case we only have 3 possible crossings : @xmath31 .
we denote these using the braid operators @xmath32 , with @xmath33 corresponding to an interchange between the right - most pair of particles , @xmath34 for the middle pair and @xmath35 for the left - most pair .
more generally , instead of defining the actual particles as @xmath36 , we define the positions as that sequence : the left - most particle is at position @xmath37 , next @xmath38 , and so on with the right - most particle being at position @xmath0 . given any sequence of particle crossings
, we can employ braid group notation @xcite to classify the motion , denoting pair crossings with the set @xmath39 .
for example the sequence @xmath40 is described by @xmath41 , and the initial configuration @xmath30 becomes the final configuration @xmath42 . since crossing direction
is irrelevant , the reciprocal notation for the braid group is ignored ( i.e. @xmath43 in this context ) .
we are now ready to classify the distinct kinds of motion that can occur .
consider first the 3-body case .
this problem can be mapped to that of a single particle moving in a hexagonal - shaped well , with the bisectors of the hexagon denoting pair - crossings of the particles @xcite . in this case
we have @xmath44 as the braid operators .
for any string of these we can classify the motion in the equal mass case as follows : @xmath45 by comparing subsequent items in the string ; we have included the @xmath46 descriptors employed previously in 3-body studies @xcite . in other words , since crossing direction is irrelevant , the only interesting types of motion are when the same pair of particles crosses twice in a row ( @xmath47-motion ) corresponding to crossing a single bisector of the hexagon twice in succession , or when one particle crosses each of its compatriots in succession ( @xmath48-motion ) corresponding to the crossing of two successive bisectors . in the unequal mass case
the hexagon is no longer symmetric , and so @xmath49 , say , is now distinguishable from @xmath50 .
however one could impose an equivalence relation between these two motions and continue to employ the above notation .
any given motion in the system can be characterized by a sequence of letters @xmath47 and @xmath48 ( called a symbol sequence ) , with a finite exponent @xmath51 denoting @xmath51-repeats and an over - bar denoting an infinite repeated sequence .
this idea extends naturally to the 4-body system .
here the braid operators are @xmath32 .
we can construct the following definitions : @xmath52 whose results can be visualized in figure [ fig:4body_motion_class ] .
the @xmath47 and @xmath48 motions still represent the same physical situations as in the 3-body case .
however there is now a new type of motion : @xmath53 motion , which is when two particles cross one another and then the other two cross one another .
we can proceed further , generalizing our arguments and definitions to the @xmath0-body case and make a more formal definition of our motion classes whose specific cases for @xmath54 and @xmath55 will be equivalent to what we have described above . to describe crossings of @xmath0 particles for a specific trajectory we have the braid operators @xmath56 .
a sequence of @xmath57 pair crossings will be described by @xmath58 where @xmath59 for all @xmath60 is a discrete integer function .
here @xmath61 means that the particles currently in positions @xmath62 and @xmath63 cross .
any given sequence of @xmath57 braid operators forms a unique ordered list of crossings for the given trajectory .
as stated previously , crossing directions are irrelevant .
now we define a new function @xmath64 using the finite forward difference function of @xmath65 @xcite .
we are interested in the absolute difference between subsequent terms , and thus we have @xmath66 for all @xmath67 now @xmath68defines a metric that describes the relative `` distance '' between any pair of crossings , and we classify the motion according to this distance @xmath69 denoting each type by increasing letters of the alphabet . in other words , @xmath47-motion corresponds to any 2 crossings in nearest proximity two particles cross each other twice in succession
. @xmath48-motion corresponds to any 2 crossings in next - nearest proximity two particles cross each other , and then one of them crosses its other nearest neighbour .
@xmath53-motion corresponds to any 2 crossings in next - to - next - nearest proximity : two particles cross each other and then a neighbouring pair cross each other .
we can continue on in this fashion until we reach the extreme case in which the right - most pair of particles cross one another followed by the crossing of the left - most pair ( or vice - versa ) .
we illustrate this classification with some examples . for @xmath54
, we have @xmath70 .
suppose we have a crossing sequence @xmath71 , yielding from ( [ 3bodymotions ] ) the symbol sequence @xmath72 . by our definition
above we have @xmath73 and from ( [ braiddiffmotions ] ) we get @xmath72 as we expected . for @xmath55 , we have @xmath74 .
consider the previous example @xmath41 , which from ( [ 4bodymotions ] ) yields the symbol sequence @xmath75 , the same result we would obtain from computing successive values of @xmath68 ( which are @xmath76 for this example ) .
the preceding classification system is limited to pair - wise crossings and does not cover situations in which more than one pair of particles crosses at the exact same time step .
while the braid group notation does allow for multiple collisions by writing them `` left - to - right '' , it is important in our system whether or not these collisions occur at the same time step . for the @xmath0-body problem ,
a simultaneous collision of @xmath57 particles corresponds to the crossing of a single particle through an @xmath77-dimensional surface in the interior of the @xmath78 simplex .
this surface is obtained by continually bisecting the simplex along its ( higher - dimesional ) edges and vertices until the surface of appropriate dimensionality is obtained .
we can denote such collisions by extending the braid group notation with the set @xmath79 , where the subscript denotes which set of particles is involved , beginning with the left - most , and the superscript ( on these subscripts ) denotes the number of particles in the collision .
for example @xmath80 denotes a 7-particle collision that involves particles 9 - 15 .
we shall drop the superscript `` 2 '' when pair - wise collisions are involved .
all collisions yield crossings except for the situation in which the initial conditions cause @xmath57 particles to occupy the same point throughout the motion . in this latter case
the system reduces to that of an ( unequal mass ) @xmath81-body problem . of course for the 3 and 4-body cases the numbers of multiple collisions are simple .
there is a single type of multiple collision in the 3-body case , which occurs when the hex particle crosses the origin . in the 4-body case
we can have two kinds of 3-body collisions ( described by @xmath82 ) that occur when the box particle crosses the line of intersection of any two bisecting planes of the 3-simplexes .
we also have two kinds of 4-body collisions ( described by @xmath83 ) .
the former occurs when two pairs of particles cross each other at the same time , and corresponds to the box particle crossing one of the three lines connecting opposite vertices of the pyramids in the simplex ( see figure [ fig : potential ] ) .
the latter is when all four particles cross at once , equivalent to the box particle crossing the origin .
in this section we consider the equal - mass case .
we study the behaviour of the four - body system using a variety of initial conditions , and compare to the 3-body case @xcite where relevant .
many patterns of motion in the 3-body system have natural counterparts in the 4-body case .
a necessary check on the code is to see if we can reproduce results in the 3-body case . indeed , if _ one _ of the box - particle s position _ and _ momentum coordinates is initially zero it will remain zero throughout the motion , and the motion of the box - particle will be restricted to a plane .
this corresponds to the situation mentioned near the end of the previous section , as is easily shown . from the hamiltonian
( [ hamnew ] ) , the equations of motion for the @xmath84 coordinate is completely arbitrary ; we could have chosen @xmath85 or @xmath4 equivalently . ]
are : @xmath86\label{padot}\end{aligned}\ ] ] @xmath87 so if the initial conditions are @xmath88 and @xmath89 , then we have clearly from eq.([adot ] ) that @xmath90for all @xmath91 , which implies that eq.([padot ] ) becomes @xmath92\notag\\ & = & 0\label{padot2}\end{aligned}\ ] ] and so the motion is restricted to the @xmath93 plane . similarly , it is not difficult to convince oneself that the box - particle is also restricted to the @xmath94 plane when the initial momentum and position @xmath95 and @xmath4 are both zero , and to the @xmath96 plane when @xmath97 and @xmath85 are initially both zero .
note , however , that even if there is no momentum along @xmath84 initially , it does not mean that the particles will never move in the @xmath84 direction . indeed if the initial position @xmath98
, then the particle will acquire momentum according to eq.([padot ] ) .
consequently we should recover all of the patterns of motion that the hex particle exhibits in the 3-body case @xcite for the subset of initial conditions in which @xmath88 and @xmath89 .
we found this to be the case , and recovered the annulus , pretzel and chaotic motions referred to in the introduction .
figure [ fig : beg_2d ] illustrates some examples belonging to the first and second classes . * * * * note that this is not an equal - mass 3-body system , but rather one with unequal masses , because the two particles that are moving together are like one single particle with a mass twice as large .
indeed , two trajectories ( out of four ) are exactly identical , which means that two particles are moving together . .
the initial conditions for the plots on the left are : @xmath99 @xmath100 @xmath101 @xmath102 @xmath89 .
@xmath85 will be calculated so that eq .
( [ hamnew ] ) is satisfied initially .
the initial conditions for the plots on the right are : @xmath103 @xmath100 @xmath104 @xmath102 @xmath89 . ]
similarly , it is not surprising that we recover the two - body case if we choose _ two _ momentum and position coordinates to be zero initially .
for example , if we set @xmath105 , @xmath106 and @xmath107 , @xmath108 , then the box - particle will move on a line parallel to the @xmath85-axis . in this case , the @xmath109 plots show only two distinguishable trajectories , because we have actually two pairs of particles moving . unlike the reduction to the three - body case from the equal mass four - body problem
, we have here a reduction to an equal mass two - body problem .
of course the more interesting situation is when there is motion in all spatial directions , generating three - dimensional patterns .
changing the initial conditions a little bit from those used for the plots shown in figure [ fig : beg_2d ] by giving , for example , a small momentum along the @xmath84 direction gives us the three - dimensional trajectories ( figure [ fig : beg_3d ] ) .
not surprisingly , these patterns are direct generalizations of what we observed in figure [ fig : beg_2d ] , since the motion in the @xmath84 direction simply perturbs the patterns previously obtained in the @xmath110 plane when we set @xmath88 and @xmath111 , initially .
essentially the original hex - particle patterns develop a `` thickness '' in the @xmath84 direction . .
the initial conditions for the plots on the left are : @xmath99 @xmath112 @xmath101 @xmath102 @xmath89 . ] for the three - dimensional annulus , we chose initial conditions to show that a non - zero @xmath113 induces a non - zero @xmath114 , even if @xmath115 .
it is also interesting to compare the peaks of figure [ fig : beg_2d ] with those of figure [ fig : beg_3d ] .
we see that the three - dimensional trajectories experience a small deviation ( near the peak ) in the trajectories of the two particles that were exactly the same for the two - dimensional box - particle trajectories .
the more we increase the value of the initial conditions @xmath113 or @xmath116 the bigger the deviation , and in the @xmath15 space , the particle will have a larger amplitude in the @xmath84 direction .
to further investigate the trajectories that can be obtained in the full three - dimensional case , we follow the method in @xcite - that is using initial condition constraints of _ fixed - energy _
( fe ) and of _ fixed - momentum _ ( fm).initial values for @xmath117 and @xmath97 , adjusting @xmath118 using the hamiltonian constraint ( [ hamnew ] ) , and check that @xmath119 remains at its initial value throughout the motion ( figures [ fig : beg_2d ] , [ fig : beg_3d ] , [ fig : ann_pretz ] , [ fig:2d_pretz_poinc ] , [ fig : odd_shape ] , [ fig : nice_poinc ] ) .
we have chosen to vary @xmath118 , to more easily facilitate comparison with the previous results in the equal - mass 3-body case @xcite .
similarly for the fixed - momenta case , we will choose @xmath120 and @xmath118 while allowing @xmath119 to vary as the initial conditions vary .
we imposed error tolerances on the matlab ode routine and checked that the total energy of the system remained constant throughout the motion for any given set of initial conditions .
since we were using an improved version of the code , we were able to get the error down to @xmath121 in most cases , and always less than @xmath122 .
as noted above , small perturbations of the three - body case yield three - dimensional trajectories that usually have a nice shape similar to that of the three - body case if projected onto one of the planes @xmath123 or @xmath124 . however , the third axis is generally just a periodic oscillation with no real pattern relative to the other axes , as in figure [ fig : ann_mess ] .
this trajectory s motion is characterized by the symbol sequence @xmath125 . *
* plane , the plot is much more random in the @xmath124 plane , although one can still observe a semi - periodic pattern .
the bottom - left isometric plot of the trajectory shows that the periods of oscillation in the various axes do not line up nicely for these initial conditions .
the bottom - right plot show the four particle positions vs. time * * ; * * the symbol sequence is * * * * @xmath125 . ]
plane , the plot is much more random in the @xmath124 plane , although one can still observe a semi - periodic pattern .
the bottom - left isometric plot of the trajectory shows that the periods of oscillation in the various axes do not line up nicely for these initial conditions .
the bottom - right plot show the four particle positions vs. time * * ; * * the symbol sequence is * * * * @xmath125 . ]
however , cases where these periods do line up can be obtained with carefully chosen initial conditions .
for example , figure [ fig : ann_pretz ] shows a trajectory that has a pretzel form when projected onto two of the planes ( @xmath126 and @xmath127 ) , and an annulus when projected onto the third ( @xmath124 ) . the three - dimensional isometric ( in which there is no perspective so that objects further away do _ not _ appear smaller ) view shows a periodic three - dimensional path - a qualitatively new feature that has no analogue in the 3-body system .
the motion here is described by @xmath128 , which is very similar . in this case
however , we see that from every direction the trajectory forms a recognizable two - dimensional plot ( third direction is not shown , but it is a pretzel similar to the top - left ) . since the periods of the motions in various axes are compatible in this case , we see from the isometric plot ( middle - left ) that a fully three - dimensional periodic trajectory exists .
note that fe ( @xmath129 ) constraints were used in this case for conve ] . in this case
however , we see that from every direction the trajectory forms a recognizable two - dimensional plot ( third direction is not shown , but it is a pretzel similar to the top - left ) .
since the periods of the motions in various axes are compatible in this case , we see from the isometric plot ( middle - left ) that a fully three - dimensional periodic trajectory exists .
note that fe ( @xmath129 ) constraints were used in this case for conve ] to give an even better idea of the types of paths that can be obtained with compatible periods , another example is shown in figure [ fig : ann_pretz_2 ] .
the symbol sequence for this trajectory is similar to the other example in that it contains mostly sets of @xmath130 and @xmath131 ; however @xmath132also shows up occasionally .
suffice it to say that many interesting trajectories can be obtained , although most are generalizations of the three - body case , as seen in the previous section , or are simply chaotic .
as expected we find that the latter case produces the same type of dense orbits in three - dimensions as its 3-body counterpart did in two .
we will now focus our attention on the various special cases that we can attain by starting particles very close to one another .
for example , we can put two sets of two particles together , and we expect those two to repeatedly cross one another , while on average the two pairs behave as a equal - mass two body system .
similarly we could put three particles very close together and one separate . from this and the fact that we have equal masses we will consider six initial configurations : 1 + 1 + 1 + 1 , 2 + 2 , 3 + 1 , 2 + 1 + 1 , 1 + 2 + 1 and 1 + 1 + 2 , where `` a+b+c '' notation indicates the relative size of the initial @xmath133 values , with largest at the left and smallest at the right ; a number larger than unity indicates that the difference between the @xmath133-values of these particles is small relative to all other spacings .
for example , `` 2 + 1 + 1 '' denotes that we have a pair of particles starting close together ( relative to the other spacings ) with the greatest @xmath133 values , and two independent ( far apart ) particles with the lower two @xmath133 values . strictly speaking the 1 + 1 + 2 and 2 + 1 + 1 cases
are symmetric ( equal masses ) and indeed they produce similar plots as we will see . for each of these possible cases , we construct plots of the box - particle in @xmath15 space and the associated plot of particle positions on the line .
the easiest way to find suitable initial conditions is to use ( [ changevar1 ] ) to choose @xmath134 and @xmath84 so that we get suitable particle starting positions , and set the momenta to zero .
thus we use the fixed - momenta conditions and allow @xmath119 to vary .
only one angle of the @xmath15 plots is shown , though the captions describe them in more detail .
figure [ fig:1_1_1_1 ] shows the case where the four particles begin spaced at equal unit spacing with no momentum .
predictably they are all drawn together and cross at the same point , continuing this pattern indefinitely . since all four particles always ` collide ' at the same time step
, the symbol sequence in terms of @xmath47 , @xmath48 , and @xmath53 is undefined .
instead the motion reduces to that of a single body of mass @xmath135 the box particle travels back and forth along a line in @xmath15space , with all particles simultaneously crossing at the origin .
this motion is unstable : a slight change in any of the initial conditions ( either via a small displacement or small momentum ) throws the system into chaos , as shown in the bottom two images of figure [ fig:1_1_1_1 ] .
plot is a single line diagonal through the three axes , as shown here .
the bottom two images show what happens if we change the initial conditions very slightly ( here we set @xmath136 ) .
a ] plot is a single line diagonal through the three axes , as shown here .
the bottom two images show what happens if we change the initial conditions very slightly ( here we set @xmath136 ) .
a ] in figure [ fig:2_2 ] , we see the case with two pairs of particles , denoted as 2 + 2 .
each of the two tightly bound states undergo mildly irregular motion ( where each particle randomly attains the maximal separation from the origin ) , but that the orbit of each bound pair about the other is quite regular , with fairly constant frequency and amplitude .
numerically we estimate the frequency of the oscillation of the pairs at @xmath137 cycles per time step , with an amplitude of @xmath138 .
the motion is a string of @xmath47 s , @xmath48 s , and @xmath53 s without any discernible pattern , the only noticeable feature being that the @xmath48 s always occur in pairs ( @xmath139 ) .
plot is a very dense pretzel if proj ] figure [ fig:1_3 ] shows the 3 + 1 case with three particles closely bound together .
these undergo a mildly chaotic orbit that itself regularly oscillates with an amplitude of about @xmath140 and a frequency of @xmath141 cycles per time step about the 4th particle in a loosely bound state . _ _ _ _ the symbol sequence is mostly @xmath48 s , with @xmath47 or @xmath53occasionally appearing .
note that while we still do not see any @xmath142s here , we do see some odd numbered groups of @xmath48 s .
plot is an almost two - dimensional pretzel rotated sidewa ] plot is an almost two - dimensional pretzel rotated sidewa ] plot is an almost two - dimensional pretzel rotated sidewa ] figures [ fig:2_1_1]-[fig:1_1_2 ] show the cases where only one pair of particles is closely bound , namely 2 + 1 + 1 , 1 + 2 + 1 and 1 + 1 + 2 respectively . fig .
[ fig:2_1_1 ] generates a thick ` fish ' pattern : the tightly bound state of two particles executes a @xmath143 motion with respect to the other two .
however the full 4-body sequence does not exhibit any clear pattern .
plot is a `` fish '' type shape in one plane , and a mess in the others ( as seen in the previo ] fig .
[ fig:1_2_1 ] illustrates a similar kind of motion , in which the tightly bound state undergoes a @xmath144 motion with respect to the other two ; the time period shown on the graph is too short to see this explicitly in the figure .
again the 4-body crossing sequence has no discernible pattern .
plot is an alm ] in fig .
[ fig:1_1_2 ] we see another similar trajectory , but with subtly distinct features .
the two particles are not as tightly bound as in the preceding cases , but ( over the time scales we observed ) execute a highly regular oscillatory interaction with the other two particles .
the four particles repeatedly come very close together before executing near parabolic motion about the centre of mass . however after about 80 time steps the trajectory degenerates into chaos .
while the paired particles remain bound , they eventually fall out of the parabolic motion and begin to move chaotically with respect to the other particles .
no pattern is obvious in the symbol sequence .
we will now examine the poincare plots of these trajectories in the newtonian system .
unfortunately , this turns out to be much more difficult in the four - body case than it was in the three - body .
the reason is that in the three - body system , the poincare plots were a collection of points in two - dimensions @xcite , while in the four - body system they are a collection of points in _ three _ dimensions .
the latter collection is much more difficult to visualize since it is impossible for the human eye to determine the depth of a single point on these three - dimensional plots with sufficient precision .
furthermore , the outer layers of points tend to obscure inner ones exists .
thus we will examine the poincare plots of several independent three - dimensional trajectories , and then look at ways of visualizing the `` total '' poincare plot , with a broad range of initial conditions combined .
poincare plots are plots of sections of phase space , affording a comparison of classes of trajectories over a broad range of initial conditions .
we shall set @xmath129 throughout . in the equal mass case
all of the bisectors of the 3-simplex represent equivalent crossings of particles .
thus we can plot the crossing of any two particles ( equivalent to the @xmath15 particle crossing one of the bisectors in the three - dimensional potential ) all on the same graph .
in the three - body case the poincare plot was that of the radial momentum of the simplex particle plotted against its angular momentum every time the particle crossed a simplex bisector @xcite .
the most natural extension of the construction of these plots to three dimensions is to use spherical coordinates , and plot the radial momentum , denoted here as @xmath145 against the two angular momenta squared : @xmath146 and @xmath147 .
more concisely , we define @xmath148 to be the distance from the origin to our point of crossing in @xmath15 space , @xmath149 to be the _ azimuthal _ angle in the @xmath126 plane ( with @xmath150 ) , and @xmath151 to be the _ polar _ angle from the @xmath84 axis ( with @xmath152 ) . from this , we denote the associated momenta of @xmath153 and @xmath151 as @xmath154 and @xmath155 respectively . specifically , from simple geometry relating our spherical coordinates @xmath156 to our cartesian coordinates @xmath15 , we have that @xmath157 and the unit vectors for these spherical coordinates are : @xmath158,\quad\hat{\phi}=\left[\begin{array}{c } -\sin\phi\\ \cos\phi\\ 0 \end{array}\right],\quad\hat{\theta}=\left[\begin{array}{c } \cos\phi\cos\theta\\ \sin\phi\cos\theta\\ -\sin\theta \end{array}\right]\label{sphereuv}\ ] ] the desired momenta are : @xmath159 where @xmath160 is the momentum vector in @xmath15 space .
then by substituting ( [ spheretrig ] ) and ( [ sphereuv ] ) into ( [ spheremom1 ] ) , we can find an expression for the momenta in terms of @xmath120 and @xmath118 : @xmath161 thus we can use ( [ spheremom2 ] ) to plot @xmath162 , @xmath146 and @xmath147 whenever two of the four particles cross one another . as we will discover further on ,
it becomes very difficult to visualize a complete poincare plot in three dimensions .
we therefore will build up to this case by first examining individual poincare plots generated from single t to keep things simple , we will first look at the case where our trajectory is in two dimensions . this should result in a two dimensional poincare plot as well , as in the three - body case . since @xmath118 is varied to keep the initial conditions consistent
, we will arbitrarily choose @xmath85 and @xmath97 as the variables that we set to zero , thus making it into a two dimensional trajectory as seen in previous sections .
our intuition is indeed correct , as can be seen in figure [ fig:2d_pretz_poinc ] .
there a two - dimensional pretzel shape produces a two - dimensional poincare plot that coincides with those seen in the three - body case , consistent with the quasiperiodic motion of this trajectory .
as expected , the value of the @xmath163 ( the momenta of the azimuthal angle ) remains zero throughout .
this can also be seen from ( [ spheremom2 ] ) , with @xmath164 and @xmath165 .
more importantly , this confirms that the poincare plots of the four - body system will be generalizations of the three - body system , as was seen with the trajectories , validating our approach for generating these plots .
now , we will look at the poincare plots for some of the `` nicer '' three - dimensional trajectories . from figure [ fig : ann_pretz ]
we generate a poincare plot that is given in figure [ fig : ann_pretz_poinc ] , shown from a few angles . , shown here from all three directions , and from an isometric perspective .
the broad circular bands are indicative of the quasiperiodic character of this trajectory . ]
this plot is extremely interesting in that it does not seem to exhibit the same patterns as the three - body case . remembering that this shape is actually a combination of annuli and pretzels
, we can begin to see some patterns . first , the elliptical shells ( they are actually extruded ellipses ) in this plot are very similar to the ellipses in the pretzel section of the three - body poincare plots , although no geometric pattern is immediately obvious .
second , if we compare the scale on this plot to that of the previous and final poincare plots ( see next section ) , we discover that indeed it is only occurring in a smaller region of the larger plot - potentially the three - dimensional equivalent of the `` pretzel region '' seen in the three - body case ( albeit more complex , since this trajectory is not exclusively a pretzel ) .
another example is shown in figure [ fig : odd_shape ] .
we can see here that because of the skew of the shape in the trajectory plot , the poincare plot is not nearly as structured as the previous ones . still
, we can note some pattern , although it is much more difficult to see in three dimensions .
each trajectory is quasiperiodic , but spread over a three dimensional space , with the pretzel trajectories forming a non - overlapping knotted structure . as with the 3-body case ,
the poincare plots consist of circles with some width .
these occupy a three dimensional volume , and the more loosely dispersed dots form a `` shell '' over the top of the poincare plot .
examining other poincare plots of three - dimensional trajectories , we reach the limits of this particular approach .
as we saw earlier , very few initial conditions actually produce nice three - dimensional paths , and the ones that do not produce similarly uninformative poincare plots .
it is possible that if run for a significant number of time steps these plots would materialize into more than just a random collection of points in space , but the amount of computation required to produce the hundreds of thousands of time steps that would be required to check this is daunting , and probably not the best use of resources .
moreover , it becomes increasingly difficult to visualize these plots as the number of time steps increases . however , there is some support for the conjecture that more time steps will reveal better patterns .
more time steps yield more particle crossings , and thus more points on the poincare plot . instead of extending the length of the trajectory though
, we can instead choose initial conditions such that two particles are extremely close together and cross frequently .
we illustrate this with an example in figure [ fig : nice_poinc ] .
_ _ the motion of the box - particle is characterized by a huge number of @xmath47 s with an occasional @xmath48 and @xmath53 ( as we would expect due to the large number of collisions between the two tightly bound particles ) . _ _ _ _ this trajectory produces a very nice three dimensional shape , demonstrating that three - dimensional poincare plots can be considerably more complicated than their two - dimensional counterparts . to conclude this section
, we plot a graph ( figure [ fig : poinc_comb ] ) that contains all of the previous poincare plots combined , just to give an idea of scale .
now we move on to the problem of visualizing the complete poincare plot , with a variety of initial conditions included .
the traditional approach generally involves choosing a range of initial conditions that fill in the important regions , and plotting all of the points on a combined two - dimensional plot .
this approach has two immediate problems in three dimensions .
first , the amount of space to fill in three - as opposed to two - dimensions makes the task of manually choosing initial conditions formidable .
there is no reasonable way of ensuring that all of the possibilities have been covered , especially since there are a large number of variations and combinations of three - dimensional trajectories .
we attempted to address this problem by automating the generation of data over a specific range of initial conditions .
with five independent variables the number of possible plots is very large , necessitating a significant reduction in the number of time steps for each initial condition .
we generated poincare data for each trajectory for five hundred time steps , giving about 1,000 poincare points per trajectory .
the second problem is that of effectively visualizing all of these discrete points in three - dimensional space .
we approached this problem by converting the data into volumetric density data , separating out the space into millions of tiny three - dimensional boxes , assigning a position corresponding to the location of the box and a value corresponding to the number of poincare points that fall inside .
this yields a large three - dimensional grid of values , each representing the density of points in the poincare plot at that approximate location .
the advantage of this approach is that we can now easily take slices and contours , colouring them according to the volumetric value at the corresponding locations .
this effectively decomposes the three dimensional data into a series of two dimensional plots .
however there are disadvantages , the most significant being that we are more limited in our ability to zoom in to see self - similar patterns and structures at the resolution the number of chosen grid points .
although we can always use more grid points , we quickly arrive at a situation in which there are too few points in a small region to satisfactorily `` fill up '' the density data .
however we expect that some of the other important features , like the highly saturated regions of chaos , will show up well in our new plots
. since those regions have a large number of points , they should get a very high density value and thus stand out from the surrounding region . with enough points and a small enough grid , it should also still be possible to see the general shape of the poincare plot in three dimensions .
we will first examine two special slices : that of @xmath166 ( the `` bottom '' slice ) and @xmath167 ( the `` side '' slice ) .
for the bottom slice , we can use the condition @xmath166 to select the applicable points .
this yields 400,000 points out of the 12 million that we generated .
as we see in figure [ fig : poinc_tot_bottom ] , this two - dimensional slice compares quite well to the 3-body case @xcite .
the figure suggests the presence of mixed regions of chaos and integrability , as further implied by figures 20 and 21 . .
notice the complex fractal - like patterns that show up as we zoom closer . ] . ] for the side slice , we could not use the condition of @xmath155 being _ exactly _ zero .
@xmath166 will be true for any of the two - dimensional trajectories in planes passing through the @xmath84 axis .
since @xmath118 is calculated from @xmath129 , most - if not all - of the two - dimensional trajectories in our parameter scan will be on planes passing through the @xmath84 axis ( initial conditions with @xmath89 are extremely unlikely ) .
however , @xmath167 would only be true of a trajectory that remained in a cone rooted at the origin .
hence we use the constraint of @xmath155 being `` close '' to zero , instead of exactly zero to get an equivalent slice .
we used @xmath168 which produced about five - hundred thousand points . ) . ] the first thing that we notice is that the side slice does not display the same patterns and fractal - like properties as the bottom slice .
this could be due either to too few time steps or to a failure to cover a sufficient range of initial conditions .
it could also possibly mean that spherical coordinates are not the best choice for a generalization of poincare plots ; perhaps cylindrical coordinates or another transformation would have produced more revealing results .
however it is also possible that the system simply does not demonstrate any clearly definably pattern when sliced along the @xmath167 plane .
we can get an idea of the true three - dimensional shape of this poincare graph by slicing the volume into multiple @xmath169 planes , each coloured according to density with brighter colours being higher densities .
unfortunately this approach allows us to see little more than the general `` shape '' of the plot . as the diagrammatic slices are not enlightening
we shall not present them here .
argest lyapunov exponent ( @xmath170 in our notation ) is then obtained from @xmath171\label{largelyap}\ ] ] below is a table o ll|ll|ll + & & & & & + & & & & & + & & & & & + & & & & & + & & & & & + the value of the lyapunov exponent is for the perturbed plot in the bottom left . * initial conditions @xmath172 , @xmath173 , @xmath174 , @xmath175 , @xmath176 and @xmath177 . looking at
the lyapunov exponents , a trend of values emerge .
first , trajectories that were considered to be mostly chaotic in nature ( based on the lissajous figures accompanying them ) have lyapunov exponents on the order of @xmath178 for trajectories that are periodic , the lyapunov exponents are on the order of @xmath179 for these small lyapunov exponents , the exact numerical values for the exponents are more uncertain because the small fluctuations in these graphs are on the order of the lyapunov exponent .
the mean value of the graphs over the last hundred collisions are used to obtain a numerical estimate .
there also seems to be a range of intermediate values for lyapunov exponents for quasiperiodic trajectories in the range of @xmath180to @xmath181 one noteworthy feature of some lyapunov graphs is orbital bifurcation , in which the apparent stabilization of the largest lyapunov exponent is punctuated by a sudden increase toward a new value .
we found this to occur in a small number of graphs we analyzed
we have carried out the first investigation of the non - relativistic 4-body problem for a one - dimensional self - gravitating system in the equal mass case .
this system is the largest value of @xmath0 in which the motion can be directly visualized in terms of the motion of a single particle , referred to as the box particle , which is inside a linear potential whose equipotential surfaces are a simplex that has the shape of six square pyramids with their bottoms joined in the shape of a cube .
we showed how to classify the motions of this system in terms of braid group operators , and were able to generalize this classification to arbitrary values of @xmath0 .
we find that the trajectories of the box - particle form natural generalizations of the 3-body case , thickening in one direction for small departures from planar motion , in which the box - particle s position _ and _ momentum coordinates are initially zero .
a common ( and somewhat unexpected ) feature is that quasi - periodic motion can appear in certain two - dimensional projections of the box - particle s trajectory whilst other projections yield an apparently chaotic motion .
we generalized the poincare plots of these trajectories to be three - dimensional , where the radial momentum of the box - particle is plotted against the two components of its angular momenta in spherical coordinates whenever the box particle crosses a bisector of the equipotential simplex .
while plots for specific trajectories can be straightforwardly constructed , a 3-dimensional poincare plot over a large variety of initial conditions proves extremely difficult to visualize in any practical sense .
we found that slices for constant azimuthal momenta yielded a fractal pattern , but slices of constant polar momenta yielded no discernible pattern .
this work was supported in part by the natural sciences and engineering research council of canada .
we are grateful to j. emerson for helpful discussions .
we are grateful to steve blanchet for his contributions to the early development of this work .
we are also grateful to marius oltean for his technical assistance while working on numerics . | we report on the results of a study of the motion of a four particle non - relativistic one - dimensional self - gravitating system .
we show that the system can be visualized in terms of a single particle moving within a potential whose equipotential surfaces are shaped like a box of pyramid - shaped sides .
as such this is the largest @xmath0-body system that can be visualized in this way .
we describe how to classify possible states of motion in terms of braid group operators , generalizing this to @xmath0 bodies .
we find that the structure of the phase * the 4-body problem in a ( 1 + 1)-dimensional self - gravitating system * * * + * @xmath1 * @xmath1 * a. lauritzen , p. gustainis , and r.b .
mann * + * dept . of physics & astronomy , university of waterloo waterloo , ont n2l 3g1 , canada
* + * pacs numbers : 04.40.-b , 04.25.-g , 05.45.ac , 04.20.jb * + * * + * * |
relativistic laser - plasma interactions have been a topical issue for the past decade @xcite , in the context of ignitor physics of inertial confinement fusion , aiming at additional fast heating and subsequent ignition of highly compressed targets by means of an external intense laser pulse @xcite .
the laser pulse drives relativistic currents , compensating return currents , and creates a pattern of counterpropagating currents which are subject to current filamentation instabilities ( cfi ) including the weibel mode @xcite . in the theoretical arena
, numerous versions of analytical and numerical methods have been developed in the past , to explore this type of instabilities @xcite . in general ,
the physical mechanism of the electromagnetic cfi is explained as follows : when the compensation of the counterpropagating electron currents is disturbed in the transverse direction , magnetic repulsion between the two currents reinforces the initial disturbance . as a consequence , a larger and larger magnetic field
is produced as time increases , degrading the transport properties .
many efforts have been devoted to this crucial problem , both related to laboratory electron beams @xcite as well as to astrophysics @xcite . concerning laser interaction with plasmas , for the
low ( nonrelativistic ) intensity regimes the inverse bremsstrahlung absorption is known to predominate around the cutoff region , where the weibel - type instability associated with temperature anisotropy can take place @xcite . in the ablative plasmas ,
the collisional and related nonlocal effects were investigated @xcite .
the original motivation for this work was triggered by more recent publications that have been quantitatively treated with the counterstreaming relativistic cfi in the collisionless limit @xcite .
a series of works could be linked with the ignitor physics by irradiating a relativistic laser pulse : fast ignition of the compressed fuel requires at least @xmath0 of external energy to be deposited within @xmath1 into the precompressed core @xcite .
if carried by @xmath2 electrons , it implies a current of @xmath3 which exceeds the transport limit of about @xmath4 @xcite , by more than a factor @xmath5 . the essential feature is the breakup of the relativistic electron beam into many filaments when propagating in dense plasma @xcite .
the physics underlying this phenomenon is just the cfi as mentioned above .
this type of instability leads to nonlinear filamentation and coalescence of the relativistic electron beam @xcite and the formation of strong magnetic fields @xcite . in the astronomical point of view , one often encounters such morphology in a variety of celestial objects , particularly , in the astrophysical jets @xcite .
state - of - the - art observations by utilizing very long baseline interferometry have revealed the filamentary structure of the jets @xcite , involving transverse magnetic fields @xcite .
more recently , large - scale toroidal magnetic fields have been discovered in the galactic center @xcite , accompanied with splendid filamentary radio arcs @xcite .
regarding the ignitor physics of laboratory plasmas , the effects of collisions and beam thermal spread play a significant role in the cfi caused by the ultrahigh relativistic currents through ablative coronal plasma @xcite , whose density rises from @xmath6 ( cutoff density ) to @xmath7 ( a thousand times solid density ) , over a radial distance of @xmath8 @xcite .
the role of collisions in a laser - produced plasma has been discussed in the early literature by motz @xcite , though the description was restricted to the nonrelativistic fashion .
it is important to note that the collision frequency invokes additional parameter disturbing the universal density scaling , in terms of the time scale of a plasma oscillation period @xmath9 and the spatioscale of a skin depth @xmath10 , which is valid only for collisionless regimes . in fully relativistic regimes ,
the collision cross section should be evaluated by using the mott scattering formula @xcite . presuming a small angle scattering and averaging over the angle yield the electron - ion collision frequency that can be defined as @xmath11 , where @xmath12 is the number density of ions , @xmath13 , @xmath14 , @xmath15 , and @xmath16 , @xmath17 , @xmath18 , and @xmath19 are the electron momentum , the electron rest mass , the averaged charge number , and the coulomb logarithm , respectively @xcite , and the electron - electron collision frequency is given as @xmath20 .
introducing the current neutral condition of @xmath21 , where @xmath22 , @xmath23 , and @xmath24 are the beam electron density , the plasma electron density , and its velocity , respectively , we find the ratio of @xmath25 for the plasma return current . when assuming @xmath26 to be almost constant , the ratio can be estimated as @xmath27 equation ( [ eq:0 ] ) indicates that indeed , collisional effects are important in the supersolid density regions of @xmath28 , and particularly , for the lorentz plasma with @xmath29 .
in addition , the beam thermal spread will occur there , since the electrons penetrating through the ablative corona are expected to be thermalized via the collisional and collisionless dissipative processes @xcite .
the thermal effects involved with the ratio of the transverse temperature to the total energy of beam electron , i.e. , @xmath30 with expected values in the range less than unity @xcite , which also disturbs the aforementioned scaling .
in this paper , i present fundamental eigenmode properties and growth rates in the linear stage of the cfi including the effects that violate the universal density scaling .
the theory is expanded in fully relativistic regime on the basis of the self - consistent beam - maxwell equations . to the best of my knowledge ,
a comprehensive treatment of electron - electron and electron - ion collisional effects on relativistic counterstreaming electron currents with thermal spread has not been carried out , so far .
for the symmetrically counterstreaming currents , i found that collisional and thermal effects are likely to lower the growth rate of the relativistic weibel instability .
the most significant result is that the finite collisional coupling between electron and ion creates the growing oscillatory and the growing wave modes , which stand out for long - wavelength perturbations , and in the moderate to strong collisional regime , even for short - wavelength perturbations , the growth rate of the oscillatory mode exceeds that of the suppressed weibel instability . in this aspect ,
the present work goes beyond the framework of the well - established theory of electromagnetic instabilities .
the asymmetric configuration effects of the counterstreaming currents are also investigated by using a slow return current approximation .
furthermore , i argue that the collisional coupling between electron and electron creates a growing wave mode , but its growth rate is lower than that of the suppressed weibel instability .
it is also shown that thermal effects participate in lowering the growth rate of the weibel instability .
although , in the case that includes thermal corrections , the present calculation is valid for the smaller wave number @xmath31 as shown later , the most interesting range @xmath32 can be fairly covered . in order to spell out these subjects ,
the present paper is organized as follows : in sec .
[ sec:2 ] , the linear theoretical analysis of the relativistic cfi is expanded systematically .
the basic equations introduced in sec
. [ sec:2a ] are linearized so as to obtain a dispersion relation of the cfi along the manner outlined in the appendix .
the generic dispersion relation is presented in sec .
[ sec:2b ] , and its approximate expressions are derived in sec .
[ sec:2c ] . in sec .
[ sec:3 ] , for an application , the newly derived equations are solved for typical parameters of counterstreaming relativistic currents , and the properties of complex eigenmodes are investigated for the cases including the effects of electron - electron collision ( sec .
[ sec:3a ] ) , electron - ion collision ( sec .
[ sec:3b ] ) , thermal corrections ( sec .
[ sec:3c ] ) , and for the special case without collisional and thermal effects ( sec .
[ sec:3d ] ) , as well as , for the case with an asymmetrical configuration of counterstreaming currents ( sec .
[ sec:3e ] ) , closely relevant to ignitor physics . as a matter of convenience , formulas of the growth rates
are explicitly written down for some interesting cases .
finally , sec .
[ sec:4 ] is devoted to concluding remarks .
begin with the nonlinear beam - maxwell equations that include both friction and pressure terms . assuming the ions to be at rest and to provide a uniform charge - neutralizing background , we study the relativistic dynamics of two uniform , counterstreaming electron currents by employing the following set of equations in the dimensionless form , @xmath33 @xmath34 @xmath35 @xmath36 @xmath37 where @xmath38 , @xmath39 , and the subscript @xmath40 , @xmath41 labels the two electron components and @xmath42 labels the countercomponent of @xmath43 , viz . , @xmath44 , @xmath45 for @xmath46 , @xmath41 , respectively .
in particular , @xmath47 and @xmath48 describe the electron - ion and electron - electron momentum exchanges , and the other notations are standard . for normalization ,
i have used the initially uniform density @xmath49 , the speed of light @xmath50 , and the electron plasma frequency @xmath51 . note that the poisson eq .
( [ eq:5 ] ) is equivalent to a combination of the continuity eq .
( [ eq:1 ] ) and the ampere - maxwell eq .
( [ eq:4 ] ) . according to a procedure similar to that developed by califano _
@xcite , i investigate the behavior of small amplitude perturbations by linearizing eqs .
( [ eq:1])@xmath52([eq:4 ] ) .
i impose current neutrality of @xmath53 , where @xmath54 are the initial velocities in @xmath55 direction . under the current neutrality
, there exists no magnetic field initially .
the cfi is studied in the @xmath55-@xmath56 plane . in order to derive the dispersion relation ,
all perturbed quantities are assumed to be in the form of @xmath57 $ ] . as a result , a magnetic field @xmath58 and a corresponding electric field @xmath59
are generated .
the purely transverse mode is investigated in detail throughout this paper , i.e. , @xmath60 and @xmath61 , whereas the longitudinal mode such as two - stream instability with @xmath62 and @xmath63 is not taken into account at the moment .
the pressure of each electron components is connected with its density by a polytropic relation , which depends on the characteristic frequency @xmath64 and wave number @xmath31 of the mode being considered . as it is well known , in the case that the ratio of @xmath65 is much larger than the electron thermal speed , the adiabatic exponent of @xmath66 is adequate for the polytrope @xcite .
henceforth , i assumed @xmath67 , so that @xmath68 . then , we get a closed form of the linearized eqs .
( [ eq:1])@xmath52([eq:4 ] ) .
the background ions are supposed to be fixed on the short time scale of @xmath69 . along these assumptions , in the nonrelativistic limit eqs .
( [ eq:1])@xmath52([eq:5 ] ) involve the resistive transverse wave modes and longitudinal wave mode with thermal correction which were discussed in ref .
@xcite .
the extended dispersion relation of the relativistic cfi including collisional and thermal effects is then found self - consistently . in collisionless cases
the dispersion relation that can be expressed as functions of @xmath70 and @xmath71 does not include the imaginary unit @xmath72 explicitly and contains the purely real and purely imaginary solutions of @xmath73 @xcite .
the purely real solutions consist of the pairs of the positive and negative solutions , corresponding to purely oscillatory and/or purely oscillatory wave modes , while the purely imaginary solutions are concomitant with the complex conjugate solutions , to yield purely growing and purely decaying ( damped ) modes . in the collisional case considered here
, the dispersion relation includes the imaginary unit explicitly .
hence , the solutions of @xmath73 may depart from the real and imaginary axis in the complex plane . in this sense , hereafter we refer to such solutions , i.e. , complex eigenmodes with real and imaginary part , as _
dephasing modes_. below , i explicitly write down the generalized dispersion relation containing the dephasing modes . after some manipulations outlined in the appendix ,
the dispersion relation can be obtained in the complex form of @xmath74+i{\rm im}[\omega , ( k/\omega)^2]=0 $ ] , where [ eq:6 ] @xmath75 = \left [ \left ( 1 + \tau \omega_k^{-2 } \right ) - \left ( \omega_{11}^{-2 } + \omega_{11,t}^{-2 } \omega_k^{-2 } \right ) \right]\ ] ] @xmath76\ ] ] @xmath77\ ] ] @xmath78 \omega_k^{-2 } , \label{subeq:6a}\ ] ] @xmath79 = - \left ( \omega_{12}^{-2 } + \omega_{12,t}^{-2 } \omega_k^{-2 } \right)\ ] ] @xmath80\ ] ] @xmath81 \left [ \left ( \omega_{22}^{-2 } + \omega_{22,t}^{-2 } \omega_k^{-2 } \right ) + \left ( \omega_{32}^{-2 } + \omega_{32,t}^{-2 } \omega_k^{-2 } \right ) \omega_k^{-2 } \right]\ ] ] @xmath82 \omega_k^{-2}. \label{subeq:6b}\ ] ] here , @xmath83 is defined in eq .
( [ subeq:9b ] ) later , @xmath84 , and furthermore , @xmath85 and @xmath86 are defined by [ eq:7 ] @xmath87 @xmath88 @xmath89 @xmath90
@xmath91 @xmath92 @xmath93 @xmath94 @xmath95 @xmath96 where @xmath97 is the lorentz factor , and @xmath98 and @xmath99 stand for the dephasing factors which can be expressed as @xmath100 @xmath101 @xmath102 @xmath103 @xmath104 t } \over \phi } , \ ] ] @xmath105 @xmath106 and the abbreviations are [ eq:9 ] @xmath107 @xmath108 @xmath109 @xmath110 @xmath111 @xmath112 and @xmath113 , @xmath114 , and @xmath115 . in collisionless limits , it follows that in eq .
( [ subeq:9a ] ) @xmath116 , @xmath117 , and @xmath118 , @xmath119 , @xmath120 , @xmath121 ; for infinitesimal thermal spread , in eq .
( [ subeq:9b ] ) @xmath83 , @xmath122 , @xmath123 ; and for symmetrically counterstreaming currents of @xmath124 , in eq .
( [ subeq:9c ] ) @xmath125 .
it is noted that the second order terms for thermal correction of the form of @xmath126 have been neglected , as explained in the appendix .
the corresponding condition @xmath127 turns out to be consistent with the results obtained later , as well as , with the aforementioned adiabatic condition of @xmath128 , where @xmath129 is the electron thermal speed . when assuming the normalized frequencies @xmath130 and @xmath131 to be constants , the dispersion eq .
( [ eq:6 ] ) can be expressed as @xmath132 , which contains , in general , ten complex solutions of @xmath73 , consisting of five pairs of positive and negative solutions . in a special case
, they may include purely real and/or purely imaginary solutions .
in this section , i investigate some specific cases contained in the general result : first for @xmath133 with ( 1 ) only electron - electron collisions and ( 2 ) only electron - ion collisions , and then for ( 3 ) @xmath134 and ( 4 ) @xmath133 , without collisions . + for the case of @xmath135 , @xmath136 , and @xmath137 , the dephasing factors of eq .
( [ eq:8 ] ) asymptotically lead to @xmath138 @xmath139 therefore , in eq .
( [ eq:7 ] ) we read @xmath140 , @xmath141 , @xmath142 , @xmath143 , and eq .
( [ eq:6 ] ) then reduces to [ eq:11 ] @xmath144 - \omega_{41}^{-2 } \omega_{43}^{-2 } \omega_k^{-2 } , \label{subeq:11a}\ ] ] @xmath145 instead of eq . ( [ eq:7 ] )
, i introduce the definitions of [ eq:12 ] @xmath146 @xmath147 where @xmath148 , which may be rewritten as @xmath149 the definitions of @xmath150 , @xmath151 , and @xmath152 in eq .
( [ subeq:12a ] ) are recalled later .
it is noted that for a trivial case of copropagating currents with @xmath153 , i.e. , @xmath154 , eq .
( [ eq:13 ] ) reduces to @xmath155 , indicating that indeed , dephasing effects vanish .
making use of eqs .
( [ eq:12 ] ) and ( [ eq:13 ] ) , the approximate dispersion eq .
( [ eq:11 ] ) can be written in the form of @xmath156 \approx 0 . \label{eq:14}\ ] ] equation ( [ eq:14 ] ) contains six solutions of @xmath64 .
one should note that the factor of @xmath157 , regardless of the transformation of eq .
( [ eq:13 ] ) , involves a simple eigenmode of the relativistic plasma oscillation .
more details are discussed in sec .
[ sec:3a ] . for the case of @xmath158 , @xmath159 , and @xmath137 , the dephasing factors of eq .
( [ eq:8 ] ) asymptotically lead to @xmath160 @xmath161 therefore , in eq .
( [ eq:7 ] ) @xmath143 , and eq .
( [ eq:6 ] ) then reduces to [ eq:16 ] @xmath144\ ] ] @xmath162 @xmath163\ ] ] @xmath164 here , we note the relations of @xmath165 @xmath166 for heuristic ways , i properly choose the dephasing boost - frame defined as @xmath167 note that in contrast with eq .
( [ eq:13 ] ) , the wave number is also transformed by the operator in eq .
( [ eq:18 ] ) . taking account of the transformation ,
i give the definitions of @xmath168 @xmath169 instead of eq . ( [ eq:7 ] ) . by using eqs .
( [ eq:18 ] ) and ( [ eq:19 ] ) , the approximate dispersion eq .
( [ eq:16 ] ) can be expressed as @xmath170 \approx 0 , \label{eq:20}\ ] ] where @xmath171 .
equation ( [ eq:20 ] ) contains six solutions .
it is noted that the form of eq .
( [ eq:20 ] ) is quite similar to that of eq .
( [ eq:14 ] ) , though there exist some differences in dephasing property between the two .
for example , the eigenmode relevant to the plasma oscillation , which is contained in the factor @xmath172 , is now undergoing the transformation of eq .
( [ eq:18 ] ) .
more on these is given later in sec .
[ sec:3b ] and [ sec:3e ] .
for the case of @xmath130 , @xmath136 , and @xmath134 , the dephasing factors of eq .
( [ eq:8 ] ) asymptotically lead to @xmath173 @xmath174 therefore , in eq .
( [ eq:7 ] ) , @xmath175 , @xmath176 , @xmath177 , @xmath178 , and eq .
( [ eq:6 ] ) then reduces to [ eq:22 ] @xmath179\ ] ] @xmath76\ ] ] @xmath180 @xmath181 and we have the relations of @xmath182 in this case , dephasing effects disappear , and the dispersion relation yields purely real and purely imaginary solutions .
concerning eq .
( [ eq:23 ] ) , i give the definitions of [ eq:24 ] @xmath183 @xmath184 @xmath185 @xmath186 instead of eq . ( [ eq:7 ] ) .
note that the definitions of @xmath150 , @xmath151 , and @xmath152 have already appeared in eq .
( [ subeq:12a ] ) . using the definitions of eq .
( [ eq:24 ] ) , the approximate dispersion eq .
( [ eq:22 ] ) can be written as @xmath187\ ] ] @xmath188\ ] ] @xmath189 in general , eq . ( [ eq:25 ] ) contains ten solutions . at first glance , the form of eq .
( [ eq:25 ] ) seems to be different from that of eqs .
( [ eq:14 ] ) and ( [ eq:20 ] ) .
indeed , there are some differences in dispersive property , for instance , the plasma oscillation mode contained in the factor of @xmath190 $ ] involves the thermal dispersion . in the special case without thermal corrections , however , eq .
( [ eq:25 ] ) recovers the same form with eq .
( [ eq:20 ] ) , as shown below . for the case of @xmath130 , @xmath136 , and @xmath137 , the dephasing factors of eq .
( [ eq:8 ] ) asymptotically lead to @xmath191 @xmath192 therefore , in eq .
( [ eq:7 ] ) , @xmath175 , @xmath177 , @xmath143 and @xmath193 , and then eq . ( [ eq:6 ] ) reduces to @xmath194 \approx 0 , \label{eq:27}\ ] ] where @xmath195 have been defined in eq .
( [ eq:24 ] ) , and @xmath196 .
it is found that eq .
( [ eq:27 ] ) maintains the form similar to eqs .
( [ eq:14 ] ) and ( [ eq:20 ] ) .
this dispersion eq .
( [ eq:27 ] ) exactly coincides with that obtained by califano _
in the following , we find the solutions contained in the approximate dispersion eqs .
( [ eq:14 ] ) , ( [ eq:20 ] ) , ( [ eq:25 ] ) , and ( [ eq:27 ] ) .
the complex eigenmodes are explicitly written down , and surveyed for wide parameter ranges of counterstreaming relativistic currents . at first , i seek the solutions of eq .
( [ eq:14 ] ) in terms of @xmath64 .
let us consider the symmetrical configuration of counterstreaming currents such as @xmath197 @xcite , having @xmath124 .
the choice of the parameters may be instructive for making a direct comparison between the present results and the previous ones @xcite .
equation ( [ eq:14 ] ) can be then cast to @xmath198 = 0 , \label{eq:28}\ ] ] where @xmath199 and @xmath200 , and the transformation eq .
( [ eq:13 ] ) reduces to @xmath201 the first factor of ( left - hand side ) ( lhs ) of eq .
( [ eq:28 ] ) yields a simple electrostatic mode , corresponding to the relativistic plasma oscillation : @xmath202 and @xmath203 , where @xmath204 below , we refer to this eigenmode as oscillatory mode ( _ o _ mode ) .
it turns out that momentum exchange between electrons and electrons does not disturb the plasma oscillation .
the second factor of lhs of eq .
( [ eq:28 ] ) contains four solutions .
the values of @xmath205 are connected with @xmath206 through the complex operator in eq .
( [ eq:29 ] ) .
therefore , in contrast to collisionless cases , @xmath206 values are of complex , to give @xmath207 \right ] , \label{eq:31}\ ] ] where @xmath208 , and @xmath209 @xmath210 @xmath211 @xmath212 . \label{eq:32}\ ] ] note the relation of @xmath213 and @xmath214 .
moreover , i rewrite eq .
( [ eq:31 ] ) in the form of the polar coordinate : @xmath215 , where @xmath216 and @xmath217 are purely real numbers . for the signs ( @xmath218 ) each
, we have two solutions with @xmath219 and @xmath220 , which are referred to as positive mode ( _ p _ mode ) and negative mode ( _ n _ mode ) , respectively .
the squared magnitude and the polar angle of the complex eigenmodes are given by [ eq:33 ] @xmath221 ^ 2 + \beta^2(k^2 , \tilde{\nu}_{ee } , \gamma_0 ) } , \label{subeq:33a}\ ] ] @xmath222 , \label{subeq:33b}\ ] ] respectively , where @xmath223 @xmath224 one should note that eq .
( [ subeq:33b ] ) restricts its parameter range to @xmath225 , because the argument of right - hand side ( rhs ) of eq .
( [ subeq:33b ] ) is negative definite , i.e. , the ratio is @xmath226 for the _ p _ mode and @xmath227 for the _
n _ mode . in order to reconstruct the polar angles consistent with the argument
, i appropriately define @xmath228 and @xmath229 for the _ p _ mode , and @xmath230 and @xmath231 for the _
note the allowable parameter ranges of @xmath232 and @xmath233 . for the _ p _ mode with @xmath234 and @xmath235 , the complex eigenmodes of @xmath236 and @xmath237
$ ] reflect decaying and growing waves , respectively .
thus , the phase of the decaying and growing waves propagates toward @xmath56 and @xmath238 direction , respectively . on the other hand , for the _
n _ mode with @xmath239 and @xmath240 , the complex eigenmodes of @xmath241 and @xmath242 $ ] reflect growing and decaying waves , respectively .
in contrast to the _ p _ mode , the phase of the growing and decaying waves propagates toward @xmath56 and @xmath238 direction , respectively . in fig . [ fig:1 ] for the typical current speed of @xmath243 ( @xmath244 ) ,
i show the polar coordinate plots of the complex eigenmodes @xmath64 for given wave numbers @xmath31 , varying electron - electron collision parameter @xmath130 .
the trajectories of @xmath64 can be compared to those of the arrowhead of vectors . in the collisionless limit of @xmath158
, we read @xmath245 and @xmath246 in eq .
( [ eq:32 ] ) , and @xmath247 in eq .
( [ eq:34 ] ) , so that @xmath248 in eq . ( [ subeq:33b ] ) . hence , the vectors of the _ p _ mode direct the real axis , reflecting the purely oscillatory wave mode ( @xmath249 ; @xmath250 ) , whereas the vectors of the
_ n _ mode direct the imaginary axis , reflecting the purely growing ( @xmath251 ) and purely decaying ( @xmath252 ) mode .
the unstable mode just represents the electromagnetic weibel instability in a collisionless plasma , which is discussed in sec .
[ sec:3d ] later . as shown in fig .
[ fig:1](a ) for @xmath253 , for the small but finite value of @xmath130 the vectors of the _
n _ mode depart clockwise from the imaginary axis . as @xmath130 increases , the real components of the vectors increase , while the imaginary components decrease .
in this aspect , such a growing wave mode is considered to be the dephasing weibel mode with reduced growth rate . in the strong collisional regime , the polar angles of the vectors of the growing and decaying mode approach @xmath254 and @xmath255 , respectively , and the vectors shrink , reducing both their real and imaginary components .
the vectors of the _ p _ mode do not largely depart from the real axis for the small value of @xmath31 . as seen in fig .
[ fig:1](b ) , for the moderate value of @xmath256 , the trajectories of the _
n _ mode are similar to those for the smaller @xmath31 value . as for the vectors of the _ p _ mode
, i now find the clockwise deviation from the real axis .
this represents the decaying ( @xmath257 ) and growing ( @xmath258 ) electromagnetic mode . in the moderate collisional regime
, the magnitude of the deviation angles tends to be large , though the maximum value is found to be small , compared with that for the _
n _ mode . in the strong collisional regime ,
the vectors of the _ p _ mode are likely to return to the real axis , reducing their magnitude . for comparison ,
the fixed vectors of the _ o _ mode are also plotted in the fig .
[ fig:1 ] ( a , b ) . in these cases ,
the magnitude of the vectors is larger than that of the _ p _ and _ n _ mode . for the _ p _ and
n _ mode , the angular frequency of the oscillations can be defined by @xmath259 and @xmath260 , respectively ; and the linear growth rate can be defined by @xmath261 and @xmath262 , respectively .
they are summarized as follows : [ eq:35 ] @xmath263 } , \label{subeq:35a}\ ] ] @xmath264 } .
\label{subeq:35b}\ ] ] here , note the relation of @xmath265 and @xmath266 . in fig .
[ fig:2 ] for @xmath243 , i show the growth rate @xmath267 of eq .
( [ subeq:35b ] ) as a function of the wave number @xmath31 for given collision parameters @xmath268 , @xmath45 , and @xmath269 . for the _
n _ mode , it is found that the collisional effects lower the growth rate for all @xmath31 , maintaining the dependence of @xmath270 for small value of @xmath31 , as well as , the saturation property for large value of @xmath31 .
such asymptotic properties of the growth rate can be also seen in the purely growing weibel mode in a collisionless plasma @xcite .
in addition , the collisional effects create the _ p _ mode due to the dephasing mechanism mentioned above .
the growth rate has the dependence of @xmath271 and @xmath272 in the small- and large-@xmath31 region , respectively , taking a peak around the moderate value of @xmath31 .
such a peak tends to be prominent for @xmath273 , and then decreases as @xmath130 further increases , as consistent with fig .
[ fig:1](b ) . as a result
, the growth rate can not exceed that of the _
n _ mode corresponding to the dephasing weibel mode with reduced growth rate . within the present framework
, it seems that both modes do not definitely cut off the growth of short wavelength perturbations . in fig .
[ fig:3 ] , for @xmath243 and @xmath274 , i show the growth rate @xmath267 as a function of the angular frequency @xmath275 of eq .
( [ subeq:35a ] ) , varying the wave number @xmath31 as a parameter . the growth rate of the _
n _ mode turns out to be larger than that of the _ p _ mode for all @xmath31 , which is in contrast with the case including only electron - ion collisional effects , as shown in fig .
[ fig:8 ] later .
moreover , i found that the angular frequencies of the _ p _ and _ n _ mode are separated at @xmath276 , and the oscillation frequency of the _ p _ mode is always higher than that of the _
the separation frequency , where @xmath277 , is lower than the frequency of eq .
( [ eq:30 ] ) for the _ o _ mode , i.e. the plasma cutoff frequency .
the relation of @xmath278 for both the _ p _ and _ n _ mode ensures that the direction of the group velocities coincides with that of the phase velocities of carrier wave . for large value of @xmath31
, the oscillation frequency of the _ p _ mode goes far beyond the relativistic plasma frequency , while the growth rate decreases .
these properties again appear in the case including electron - ion collisional effects ( see also fig .
[ fig:8 ] ) . in fig .
[ fig:4 ] for @xmath243 , i show the growth rate @xmath267 as a function of the collision parameter @xmath130 for given values of @xmath279 and @xmath45 .
it is found that for moderate value of @xmath130 , the growth rate of the _ p _ mode takes the peak value , which tends to be well pronounced especially for @xmath280 .
for example , for @xmath281 the growth rate takes the peak of @xmath282 at @xmath283 . in the weak collisional regime , it has dependence of @xmath284 , but on the other hand that of the _
n _ mode is almost constant , to give , e.g. , @xmath285 for @xmath281 . in the strong collisional regime ,
the growth rates of both the _ p _ and _ n _ mode decrease , to exhibit the asymptotic behaviors of @xmath286 and @xmath287 , respectively . in fig .
[ fig:5 ] for @xmath274 , i show the growth rate @xmath267 as a function of @xmath31 for given values of @xmath288 , @xmath289 , and @xmath290 , corresponding to @xmath291 ( @xmath292 ) , @xmath293 ( @xmath294 ) , and @xmath295 ( @xmath296 ) , respectively . for small value of @xmath31 ,
both the _ p _ and _ n _ mode seem to increase their growth rates as @xmath297 increases . for large value of @xmath31 , however , such properties appear merely in weak to mild relativistic regime .
namely , in strong relativistic regime , the growth for large @xmath31 tends to be suppressed , for example , as for the _
n _ mode , the saturation level of the growth rate decreases as @xmath297 increases .
the peak of the growth rate of the _ p _ mode is likely to shift to the smaller-@xmath31 region , reducing its value . although , the energy dependence of the _
n _ mode appears again in the case including electron - ion collisional effects , the _ p _
mode significantly changes its property , as shown below .
i seek the solutions of eq .
( [ eq:20 ] ) in terms of @xmath64 . in the case of the symmetrically counterstreaming currents with @xmath197 and @xmath124 , eq .
( [ eq:20 ] ) can be cast to @xmath298 = 0 , \label{eq:36}\ ] ] where @xmath199 and @xmath200 .
note that eq .
( [ eq:36 ] ) has the same form as eq .
( [ eq:28 ] ) for the previous case .
however , now all @xmath70 and @xmath299 values are being dephased by the transformation eq .
( [ eq:18 ] ) , to provide the eigenmodes significantly different from those derived from eq .
( [ eq:28 ] ) .
the first factor of lhs of eq .
( [ eq:36 ] ) contains a modified electrostatic mode of the relativistic plasma oscillation .
the eigenmode may be written in the form of @xmath300 , and the growth rate can be then defined by @xmath301 , that is , @xmath302 it is found that the dissipative effects owing to the electron - ion collisions create the growing and decaying oscillatory mode , which _ seemingly _ carries its phase toward @xmath238 and @xmath56 direction , respectively . as far as ignoring thermal and asymmetrical effects of counterstreaming currents
is concerned , this mode is independent of wave number , so that the group velocity is null . note that in the collisionless limit of @xmath136 , eq . ( [ eq:37 ] ) reduces to eq .
( [ eq:30 ] ) which denotes the relativistic plasma oscillation . in this sense
, we refer to the eigenmode identified by eq .
( [ eq:37 ] ) , for convenience , as _ o _ mode .
the second factor of lhs of eq . ( [ eq:36 ] ) contains four complex solutions . along
the manner explained in sec .
[ sec:3a ] , i express the solutions in the polar coordinate form of @xmath215 , where @xmath216 and @xmath217 are purely real numbers . for the signs ( @xmath218 ) each , we get two solutions .
the squared magnitude and the polar angle of the complex solutions are given by [ eq:38 ] @xmath303 ^ 2 + \left [ { \tilde{\nu}_{ei } \over \gamma_0 ^ 3 } + { \rm sgn}(\mu ) \left ( \tilde{\nu}_{ei } \alpha - \beta \right ) \right]^2 } , \label{subeq:38a}\ ] ] @xmath304 , \label{subeq:38b}\ ] ] respectively , where @xmath305 @xmath306 and , @xmath307 @xmath308 @xmath309 note the relation of @xmath213 and @xmath214 .
below , i refer to the solutions with @xmath219 and @xmath220 , as _ p _ mode and _ n _ mode , respectively . as is the case with eq .
( [ subeq:33b ] ) , eq .
( [ subeq:38b ] ) holds a parameter range of @xmath225 , because the argument of rhs of eq .
( [ subeq:38b ] ) is negative definite , i.e. , the ratio is @xmath226 for the _ p _ mode and @xmath227 for the _
thus , recalling the definitions of @xmath228 and @xmath229 for the _ p _ mode , and @xmath230 and @xmath231 for the _
n _ mode , the complex eigenmodes @xmath310 denote the decaying ( @xmath311 ) and growing ( @xmath52 ) wave that carry their phases toward @xmath56 and @xmath238 direction , respectively ; and @xmath312 denote the growing ( @xmath311 ) and decaying ( @xmath52 ) wave that carry their phases toward @xmath56 and @xmath238 direction , respectively . in fig .
[ fig:6 ] for the current speed of @xmath243 , i show the polar coordinate plots of the complex eigenmodes @xmath64 for given wave numbers @xmath31 , varying electron - ion collision parameter @xmath131 .
we compare the trajectories of @xmath64 to those of the arrowhead of vectors . in the collisionless limit of @xmath136
, we read @xmath245 and @xmath246 in eq .
( [ eq:40 ] ) , and @xmath247 in eq .
( [ eq:39 ] ) , so that @xmath248 in eq . ( [ subeq:38b ] ) .
hence , the vectors of the _ p _ mode direct the real axis , reflecting the purely oscillatory wave mode ( @xmath249 ; @xmath250 ) , while the vectors of the _
n _ mode direct the imaginary axis , reflecting the purely growing ( @xmath251 ) and purely decaying ( @xmath252 ) mode .
the unstable mode represents the electromagnetic weibel instability in a collisionless plasma . as shown in fig .
[ fig:6](a ) for @xmath253 for small but finite value of @xmath131 the vectors of the _ p _ mode depart clockwise from the real axis . as @xmath131 increases , the real components of the vectors decrease , while the imaginary components increase .
the magnitude of the deviation angles is larger than that in the case including only electron - electron collisional effects [ compare fig . [ fig:1](a ) ] . in the strong collisional regime ,
the vectors shrink , reducing both their real and imaginary components .
this is the decaying and growing electromagnetic mode , which possesses the allowed range of polar angle of @xmath232 and @xmath313 , respectively . for the parameter range of
@xmath131 being considered , the vectors of the _
n _ mode do not largely depart from the imaginary axis , that is , the purely growing weibel mode is not so dephased .
the most remarkable property can be seen in the _ o _ mode .
for @xmath159 , the vectors leave the real axis , and as @xmath131 increases the magnitude of the deviation angles increases . as
consistent with eq .
( [ eq:37 ] ) , at @xmath314 the vectors of the growing and decaying oscillatory mode have the angles of @xmath315 and @xmath316 , respectively ; and in the strong collisional limit , asymptotically approach @xmath317 and @xmath318 , respectively . in fig .
[ fig:6](b ) , for the moderate value of @xmath256 , i now clearly find , for @xmath159 , the clockwise deviation of vector pairs of _ p _ , _ n _ , and _ o _ modes .
this essentially means that all modes are in phase lag , because of the frictional nature of collisions .
the vectors of the _ p _ and _ n _ mode exhibit the behavior similar to that displayed in fig .
[ fig:1](b ) .
in contrast with the case for small @xmath31 , the polar angles of the vectors of the _ p _ mode can not reach to @xmath315 and @xmath316 , and in the strong collisional regime the vectors are likely to return to the real axis , reducing their magnitude . on the other hand , the vectors of the _
n _ mode largely deviate from the imaginary axis until the angles reach to @xmath319 and @xmath320 . in the weak collisional regime ,
the vectors increase the real components and decrease the imaginary components , whereas in the strong collisional regime , there is decrease both in the real and imaginary components . as a result ,
the electron - ion collisional effects lower the growth rate of the _ n _ mode , namely , the dephasing weibel mode .
note that the trajectories of the vectors of the _ o _ mode are the same as those displayed in fig .
[ fig:6](a ) . for the moderate value of @xmath31 , the magnitude of the vectors of the _
n _ mode is found to be comparable to that of the _ o _ mode . for the _ p _ and _ n _ mode , the angular frequency of the oscillations can be defined by @xmath259 and @xmath260 , respectively ; and the linear growth rate can be defined by @xmath261 and @xmath262 , respectively .
they are summarized as follows : [ eq:41 ] @xmath321 } \over { \zeta_2(\tilde{\nu}_{ei}^2 ) } } \right ] } , \label{subeq:41a}\ ] ] @xmath322 } \over { \zeta_2(\tilde{\nu}_{ei}^2 ) } } \right ] } .
\label{subeq:41b}\ ] ] here , note the relation of @xmath265 and @xmath266 . in fig .
[ fig:7 ] for @xmath243 , i show the growth rate @xmath267 of eqs .
( [ eq:37 ] ) and ( [ subeq:41b ] ) as a function of the wave number @xmath31 for given collision parameters @xmath323 , @xmath45 , and @xmath269 . for the _ n _ mode , the collisional effects lower the growth rate , especially in the large-@xmath31 region where the growth rate tends to saturate .
in contrast to the case including only electron - electron collisions , the growth rate for small @xmath31 , which has the dependence of @xmath270 , is found to be not so depressed due to the electron - ion collisional effects ( compare fig . [ fig:2 ] ) .
p _ and _ o _ mode , which appear owing to the collisional effects , are prominent for the parameter of @xmath324 , and sufficiently surpass the _
n _ mode in growing long - wavelength ( small-@xmath31 ) perturbations .
this feature is in major contrast to the case including only electron - electron collisional effects . for the larger collision parameter ,
it is noteworthy that even in the large-@xmath31 region , the growth rate of the _ o _ mode exceeds that of the _
n _ mode , i.e. , the suppressed weibel mode .
this is one of the most important results in the present paper . in the large-@xmath31 region ,
the growth rate of the _ p _ mode decreases , showing the dependence of @xmath272 , and is far below that of the _
n _ and _ o _ mode . in the small-@xmath31 region ,
the growth rate of the _ p _ mode is almost independent of @xmath31 , and is always smaller than that of the _ o _ mode .
that is , @xmath325^{1/2 } < \tilde{\nu}_{ei } / ( \sqrt{\gamma_0 } \zeta_2^{3/4 } ) $ ] for @xmath326 and @xmath327 . in fig .
[ fig:8 ] , for @xmath243 and @xmath314 , i show @xmath267 of eq .
( [ subeq:41b ] ) as a function of the angular frequency @xmath275 of eq .
( [ subeq:41a ] ) , varying @xmath31 as a parameter . for comparison , in the @xmath328 plane
, i also plot a fixed point given by eq .
( [ eq:37 ] ) for the _ o _ mode .
it is found that for large value of @xmath31 , the growth rate of the _
n _ mode is larger than that of the _ o _ mode , which is always larger than that of the _ p _ mode . note that , for the _ p _ and _ n _ mode , @xmath278 , and the angular frequencies of both modes are clearly separated at the frequency of @xmath329 , where @xmath277 . as a result ,
the oscillation frequency of the _ p _ mode is always higher than that of the _ n _ mode .
these features could be also seen in fig .
[ fig:3 ] for the case including electron - electron collisional effects . in fig .
[ fig:9 ] , for @xmath243 , i show the growth rate @xmath267 of eq .
( [ eq:37 ] ) , and ( [ subeq:41b ] ) as a function of the collision parameter @xmath131 for given values of @xmath279 and @xmath45 . in the weak collisional regime , the growth rates of the _
o _ and _ p _ mode are both proportional to @xmath131 , while the growth rate of the _
n _ mode is almost constant . for smaller @xmath31 , the growth rates of the _ o _ and _ p _ mode can more readily exceed the growth rate of the _
n _ mode , and for the moderate value of @xmath131 , they take the peak values . the growth rate of the _ o _ mode can
, even for large @xmath31 , exceed that of the _
n _ mode , and takes the peak of @xmath330 at @xmath331 .
the growth rate decreases as @xmath131 further increases , showing the dependence of @xmath332 .
it is also noted that in the strong collisional regime , the growth rates of the _ p _ and _ n _ mode have the dependencies of @xmath333 and @xmath332 , respectively . as would be expected , anyhow ,
all modes are suppressed in the strong collisional regime . in fig .
[ fig:10 ] for @xmath314 , i show the growth rate @xmath267 of eqs .
( [ eq:37 ] ) and ( [ subeq:41b ] ) as a function of @xmath31 for given values of @xmath288 ( @xmath291 ) , @xmath289 ( @xmath244 ) , and @xmath290 ( @xmath334 ) .
it should be remarked that for the lower current speed , the _ o _ mode becomes most dominant .
in contrast to the case including only electron - electron collisional effects , the _ o _ and _ p _ mode monotonically reduce their growth rates as @xmath297 increases , and soon the _ o _ mode is overcome by the _
n _ mode from large-@xmath31 region .
the energy dependence of the _
n _ mode is similar to that shown in fig .
[ fig:5 ] , namely , in weak to mild relativistic regime , the saturation level of the growth rate increases as @xmath297 increases , while in strong relativistic regime , it tends to decrease . in small-@xmath31 region
, the growth rate seems to increase as @xmath297 increases , since its curve , roughly proportional to @xmath31 , shifts to the smaller-@xmath31 region .
such an apparent redshift can be also seen in the large-@xmath31 region for the _ p _ mode . in eq .
( [ eq:25 ] ) , i seek the solutions in terms of @xmath64 . regarding the symmetrically counterstreaming currents of @xmath197 , @xmath124 , and @xmath335 , eq .
( [ eq:25 ] ) can be cast to @xmath336\ ] ] @xmath337 = 0 , \label{eq:42}\ ] ] where @xmath199 , @xmath338 , and @xmath339 .
the first factor of lhs of eq .
( [ eq:42 ] ) contains four purely real solutions , but two of them are found to be inconsistent with the assumption of @xmath340 ( not shown ) .
the other two solutions can be expressed as @xmath202 , and for @xmath341 , we obtain @xmath342 where @xmath343 .
note that eq .
( [ eq:43 ] ) is of the order of @xmath344 , and therefore , the assumption @xmath340 requires @xmath345 , which is consistent with the aforementioned relation of @xmath346 .
equation ( [ eq:43 ] ) just corresponds to the relativistically extended dispersion of the bohm - gross wave with nonzero group velocity ( e.g. , see ref .
@xcite for the nonrelativistic limit ) .
in contrast to eqs .
( [ eq:30 ] ) and ( [ eq:37 ] ) , we find the thermal dispersion terms characterized by @xmath347 in eq .
( [ eq:43 ] ) .
physically , this reflects the debye screening by electrons @xcite . in the limit of @xmath348 ,
( [ eq:43 ] ) reduces to eq .
( [ eq:30 ] ) for oscillatory mode .
moreover , the second factor of lhs of eq .
( [ eq:42 ] ) contains six solutions .
numerical calculation indicates that these consist of four purely real solutions and a purely imaginary solution concomitant with its complex conjugate : @xmath349 . for the purely growing mode that is of interest here
, we can define the linear growth rate as @xmath301 . in fig .
[ fig:11 ] ( solid curve ) , i show @xmath267 as a function of the wave number @xmath31 for @xmath243 and @xmath350 , as an example .
it is found that the thermal effects simply lower the growth rate of the purely growing weibel instability , at least , in the range of @xmath280 .
note that in contrast to collisional cases , the thermal effects do not take part in dephasing the purely oscillatory and purely growing mode , but merely give rise to the mode - dispersion which can reduce the growth rate . in the collisionless limits without thermal corrections , the dephasing operators in eqs .
( [ eq:18 ] ) and ( [ eq:29 ] ) asymptotically approach unity , vanishing their imaginary parts .
then , eqs .
( [ eq:28 ] ) , ( [ eq:36 ] ) , and ( [ eq:42 ] ) degenerate into a unique equation , @xmath351 = 0 , \label{eq:44}\ ] ] where @xmath199 and @xmath200 . as expected , eq . ( [ eq:44 ] ) has the same form with eqs .
( [ eq:28 ] ) and ( [ eq:36 ] ) .
the first factor of lhs of eq .
( [ eq:44 ] ) yields the plasma oscillation mode given by eq .
( [ eq:30 ] ) .
the second factor of lhs of eq . ( [ eq:44 ] ) contains four solutions .
in contrast to collisional cases , the values of @xmath70 are obtained as purely real numbers . concerning the polar form @xmath215 , eqs .
( [ eq:33 ] ) and ( [ eq:38 ] ) reduce to [ eq:45 ] @xmath352 @xmath353 where , @xmath354 recalling the definitions of @xmath228 and @xmath355 for the _ p _ mode , and @xmath230 and @xmath356 for the _
n _ mode , eq .
( [ subeq:45b ] ) leads to @xmath249 and @xmath251 .
it is , therefore , found that for the _ p _ mode , the real solutions of @xmath357 describe the purely oscillatory wave mode , while for the _
n _ mode , the imaginary solutions of @xmath358 describe the purely growing ( @xmath311 ) and purely decaying ( @xmath52 ) mode .
the phase properties are shown in figs . [ fig:1 ] and [ fig:6 ] .
now , the angular frequency of the oscillation and the linear growth rate can be defined by @xmath359 and @xmath360 , respectively .
taking the inequality @xmath361 into consideration , the angular frequency and the growth rate can be expressed as [ eq:47 ] @xmath362^{1/2 } , \label{subeq:47a}\ ] ] @xmath363^{1/2 } , \label{subeq:47b}\ ] ] respectively .
note the relation of @xmath364 . in fig .
[ fig:11 ] ( dotted curve ) , i show the linear growth rate @xmath267 of eq .
( [ subeq:47b ] ) for @xmath243 , as a function of the wave number @xmath31 .
this growth rate is just of the relativistically extended electromagnetic weibel instability in a collisionless plasma @xcite .
it is found that for @xmath365 and @xmath366 , eq . ( [ subeq:47b ] ) simply exhibits the asymptotic property of @xmath367 and @xmath368 , respectively .
it might be instructive to compare fig .
[ fig:11 ] with figs .
[ fig:2 ] and [ fig:7 ] for the case including collisional effects .
it is confirmed that this mode is in a special case of the modes presented in sec .
[ sec:3a]@xmath52[sec:3c ] . for another comparison ,
we are concerned with an asymmetrical configuration of counterstreaming currents .
first , in eq .
( [ eq:27 ] ) for the collisionless case , i change the parameters to an asymmetrical , though still current - neutral initial beam configuration with @xmath369 ( @xmath370 ) and @xmath371 ( @xmath372 ) .
note that the corresponding beam - to - plasma density ration , @xmath373 , is of relevance to electron transport in the context of ignitor physics . here
the electron beam density could be estimated as @xmath374 and the plasma density as @xmath375 @xcite .
it is around the low - density foot of steeply rising density profile of laser - ablative plasma @xcite , where the filamentation dynamics may be most prominent . as shown in fig .
[ fig:12 ] ( crosses ) , it is found that for such parameters the growth rate reduces by a factor @xmath269 in the small-@xmath31 region and by factor @xmath376 in the saturation region .
the strong reduction is , more or less , favorable for energetic electrons to propagate through the ablative corona surrounding a highly compressed ignitor plasma . in the denser region
, electron - ion collisions might play an important role in attenuating _
slow _ return currents , owing to the relation of @xmath377 , where @xmath378 and @xmath379 denote the beam electron - ion collision frequency and beam electron plasma frequency , respectively . at this juncture , in order to take account of the collisional effects more plausibly , one may replace the dispersion eq .
( [ eq:20 ] ) with @xmath380 \approx 0 , \label{eq:48}\ ] ] where , instead of eqs .
( [ eq:18 ] ) and ( [ eq:19 ] ) , use @xmath381 and @xmath382 @xmath383 respectively
. for simplicity , we set the current parameters to the same as the previous ones : @xmath369 ( @xmath370 ) and @xmath371 ( @xmath372 ) , but now providing the slow return current with @xmath384 to be resistive such as @xmath385 ( @xmath386 ) .
numerical calculations indicate that eq .
( [ eq:48 ] ) contains six complex solutions , consisting of three pairs of the positive and negative solutions .
for the three unstable modes each , one can define the growth rates as @xmath387 . in fig .
[ fig:12 ] , i show @xmath267 as a function of @xmath31 , for the collision parameters of @xmath388 and @xmath389 , as an example . it is found that as a whole the wave number dependence itself is similar to that for the symmetrically counterstreaming cases ( compare figs .
[ fig:7 ] and [ fig:10 ] ) . the purely growing mode ( @xmath390 : crosses )
is disturbed due to the collisional effects on the slow return current to yield the _
n _ mode , namely , the dephasing weibel mode . however , the dephasing effects result in only a slight increase of the growth rate . as shown in fig .
[ fig:12 ] ( solid curve ) for @xmath389 the increasing rate is about @xmath391 at most around @xmath392 .
i mention that for the larger collision parameter , the growth rate of the _
n _ mode is depressed below that for the collisionless ( @xmath390 ) case .
in addition , the dephasing effects create the unstable modes analogous to the _ o _ and _ p _ mode that were introduced in sec .
[ sec:3b ] .
the growth rate of the corresponding _ o _ mode now quite weakly depends on the wave number .
both the modes are found to surpass the _
n _ mode in growing long - wavelength perturbations , as shown in the figure .
in particular , for the parameter region of @xmath393 , the growth rate of the _ o _ mode exceeds , even in the short - wavelength region , that of the _
n _ mode , that is , along the mechanism elucidated in sec .
[ sec:3b ] , the electron - ion collision affects the slow return current , even if the forward beam current is in the collisionless regime .
this consequence is consistent with the previous results obtained by carrying out the fully relativistic and electromagnetic particle simulation for asymmetrically counterstreaming electron currents in plasma @xcite .
in conclusion , i have systematically investigated the details of collisional and thermal effects on the relativistic current filamentation instability , generalizing dispersion relation of the beam - maxwell system . for specific cases ,
the approximate dispersions have been derived and applied to the instability analysis of typical counterstreaming relativistic electron currents , relevant to ignitor physics . for the symmetrically
counterstreaming cases , the particular results are summarized as follows : + ( i ) effects of electron - electron collision suppress the relativistic weibel instability for all wavelengths .
the effects newly create a growing wave mode , but its growth rate is always lower than that of the suppressed weibel instability .
( ii ) effects of electron - ion collision suppress the relativistic weibel instability , especially for short - wavelength perturbations .
the effects create growing oscillatory and wave mode . for long - wavelength perturbations ,
the growth rates of both modes tend to exceed the growth rate of the suppressed weibel instability . for the stronger collisional coupling , and lower current speed as well , the growth rate of the oscillatory mode can , for all wavelengths , exceed that of the suppressed weibel instability .
( iii ) effects of thermal spread simply suppress the relativistic weibel instability , at least , in a moderate wavelength range .
+ for the asymmetrically counterstreaming case , the relativistic weibel instability for the symmetrical case is strongly suppressed , though the electron - ion collision still affects the slow return current , creating the unstable modes , as mentioned above in result ( ii ) .
the important point is that , in general , the growing oscillatory mode is created by dephasing a purely oscillatory mode , and the growing wave mode is created by dephasing either a purely oscillatory wave mode or a purely growing mode . while the collisional effects invoke phase lag , reflecting the inverse transformation of eqs .
( [ eq:13 ] ) and ( [ eq:18 ] ) for electron - electron and electron - ion collision , respectively , the thermal effects do not dephase the purely oscillatory , oscillatory wave , and growing mode , but involve mode dispersion . hopefully
, intense laser - plasma interaction experiment will be able to reproduce these fundamental consequences , though they were derived by leaving out longitudinal modes , complexities of mode - coupling , and so forth .
i am grateful to j. meyer - ter - vehn for a useful discussion .
in this appendix , i briefly explain the derivation of eq .
( [ eq:6 ] ) . by linearizing the continuity
( [ eq:1 ] ) , the density perturbation of electron component @xmath43 is described as @xmath394 , where @xmath395 , and @xmath396 stand for the first order quantities of @xmath397-directional momentum of the component @xmath43 . making use of the relations of @xmath398 , which are obtained by linearizing @xmath55 component of the vector eq .
( [ eq:2 ] ) , the first order momenta @xmath399 can be expressed as functions of the first order electric fields @xmath400 . substituting these expressions into the linearized continuity equation mentioned above ,
i obtain the first order equations for density perturbation in the form of @xmath401 e_{1,x } + \left ( 1 - \tilde{t}_{0,\bar{a } } \omega_k^{-2 } + i \nu_1
\right ) e_{1,y } } \over { \left ( 1 - \tilde{t}_{0,a } \omega_k^{-2 } + i \nu
\right ) \left ( 1 - \tilde{t}_{0,\bar{a } } \omega_k^{-2 } + i \nu \right ) + \tilde{\nu}_{ee}^2 } } \label{eq : a1}\ ] ] for @xmath402 . here ,
the abbreviations of eq .
( [ eq:9 ] ) have been used . substituting the expressions of @xmath396 , eq .
( [ eq : a1 ] ) , and @xmath403 derived from eq .
( [ eq:3 ] ) , into the linearized eq .
( [ eq:4 ] ) , yields the first order equations of @xmath400 . for the manipulations , the second order terms for thermal correction of the form of @xmath126 in the products of , e.g. ,
@xmath404 are neglected , such that @xmath405 .
finally , we arrive at the equation of the form of @xmath406 ( @xmath407 ) , where the determinant of the dielectric tensor is as it is well known , if the imaginary part of the complex eigenvalues @xmath64 is much smaller than the real part , one can calculate them by the taylor expansion method @xcite
. however , this is not the case being considered , as seen in figs .
[ fig:1 ] , [ fig:3 ] , [ fig:6 ] , and [ fig:8 ] .
that is why , i have attempted to directly solve the complex eq .
( [ eq:6 ] ) to extract the eigenvalues .
m. tatarakis , f.n .
beg , e.l .
clark , a.e .
dangor , r.d .
edwards , r.g .
evans , t.j .
goldsack , k.w.d .
ledingham , p.a .
norreys , m.a .
sinclair , m .- s.wei , m. zepf , and k. krushelnick , phys .
lett . * 90 * , 175001 ( 2003 ) .
k. asada , s. kameno , m. inoue , z .- q .
shen , s. horiuchi , and d. c. gabuzda , _ astrophysical phenomena revealed by space vlbi _ , edited by h. hirabayashi , p. g. edwards , and d. w. murphy ( isas , sagamihara , 2000 ) . | i theoretically found eigenmodes and growth rates of relativistic current filamentation instability in collisional regimes , deriving a generalized dispersion relation from self - consistent beam - maxwell equations .
for symmetrically counterstreaming , fully relativistic electron currents , the collisional coupling between electrons and ions creates the unstable modes of growing oscillation and wave , which stand out for long - wavelength perturbations . in the stronger collisional regime
, the growing oscillatory mode tends to be dominant for all wavelengths . in the collisionless limit ,
those modes vanish , while maintaining another purely growing mode that exactly coincides with a standard relativistic weibel mode .
it is also shown that the effects of electron - electron collisions and thermal spread lower the growth rate of the relativistic weibel instability .
the present mechanisms of filamentation dynamics are essential for transport of homogeneous electron beam produced by the interaction of high power laser pulses with plasma . |
compact white dwarf ( wd ) binaries ( with orbital periods in the range from minutes to hours ) are important for several areas of astrophysics . the orbits of these systems decay via the emission of gravitational waves , constituting the largest signals for the next generation space - based gravitational wave interferometers .
systems of sufficiently short orbital period will merge within a hubble time , the result of which may produce a variety of exotic objects , such as helium - rich sd0 stars , r crb stars and am cvn binaries .
most importantly , when the total binary mass is near the chandrasekhar limit , the merged wds may collapse into a neutron star or explode as a type ia supernova ( e.g. , webbink 1984 ; iben & tutukov 1984 ) .
recent studies have provided support for such `` double degenerate '' progenitors of sne ia .
( e.g. , gilfanov & bogdan 2010 ; di stefano 2010 ; maoz et al .
2010 ; li et al . 2011 ; bloom et al . 2012 ; schaefer & pagnotta 2012 ) .
the outcome of a wd binary merger depends on the masses of the wds and their pre - merger conditions ( e.g. , segretain et al . 1997 ; yoon et al . 2007 ; loren - aguilar et al .
2009 ; van kerkwijk et al . 2010
; dan et al . 2012 ; raskin et al .
most previous studies of pre - merger binary wds have focused equilibrium tides and considered tidal dissipation in a perameterized way ( e.g. , mochkovitch & livio 1989 ; iben et al . 1998 ; willems et al . 2010 ; piro 2011 ) .
none of these studies have sought to predict the magnitude and location of tidal heating due to dynamical tides , which dominate the tidal responses of the binary wds . in two recent papers ( fuller & lai 2011 , 2012 , hereafter paper i and paper ii , respectively )
, we presented the first _ ab initio _ calculations of dynamical tides in realistic wd models . in paper
i , we considered resonant excitations of wd g - modes during binary decay and showed that the modes reach non - linear amplitudes near the surface of the star .
this implies that , rather than exciting discrete g - modes , the binary companion will excite a continuous train of gravity waves , which propagate outward and dissipate in the outer envelope of the wd .
we studied such continuous tidally excited waves in paper ii . for a canonical carbon / oxygen wd ( consisting of co core with a he - h envelope )
, we showed that the outgoing waves are primarily launched at the co / he transition region , and propagate toward the wd surface , where they are likely dissipated through a combination of non - linear processes and radiative damping .
we computed the energy and angular momentum flux carried by the waves in order to predict the orbital and spin evolution of wds in compact binaries .
we found that such dynamical tides cause the binary wds to be nearly synchronized prior to merger .
furthermore , the tidal heating rate can be quite large at short orbital periods ( exceeding tens of solar luminosities just before merger , depending on the system parameters ) , potentially leading to significant observable signatures . in this _
, we show that tidal heating may trigger a thermonuclear runaway hydrogen fusion event in a co wd
. the observational consequence of such an event would likely be an outburst that resembles a classical nova .
we call this new phenomenon a `` tidal nova '' ( tn ) . unlike all other types of novae or supernovae ,
a tn does not rely on mass accretion or collapse .
we present a simple two - zone model for the angular momentum evolution of a differentially rotating wd , which we use to calculate the radial tidal heating profile within the wd .
we then evolve the wd model including tidal heating to calculate changes in its temperature , luminosity , and internal structure . for a wide range of physically plausible parameters , we demonstrate that tidal heating induces a thermonuclear runaway event .
finally , we discuss the observational signatures of such an event , and compare our predictions to observations of short - period wd binaries .
using the method described in paper ii , we calculate the amplitude of tidally excited gravity waves inside a wd .
we consider a circular orbit with angular frequency @xmath2 .
the wd spins at an angular frequency @xmath3 , and the spin is aligned with the orbit . in the corotating frame ,
the frequency of the dominant @xmath4 tidal potential is @xmath5 . for a wd of mass @xmath6 and radius @xmath7 ( and given internal structure ) with a companion of mass @xmath8 ,
the energy and angular momentum fluxes carried by the gravity waves can be written as & & _ z ( , ) = t_0 ( ) f(),[jdot ] + & & ( , ) = t_0 f(),[edot ] where [ t0 ] t_0 ( ) = ( ) ^5 , with @xmath9 the orbital angular frequency ( @xmath10 is the total mass and @xmath11 is the orbital semi - major axis ) .
the dimensionless function @xmath12 ( similar to the tidal lag angle in the language of equilibrium tide theory ) determines the magnitude of wave excitation , and is strongly dependent on the internal structure of the wd and the tidal frequency @xmath13 . in paper
ii we have calculated @xmath12 for @xmath14 co wd models of various surface temperatures and slow rotation .
we found that @xmath12 is an erratic function of @xmath13 because of the `` quasi - resonance cavity '' formed by the co core inside the he / h shell .
however , because of the strong dependence of @xmath12 on @xmath13 [ the envelope of @xmath12 approximately scales as @xmath15 , at sufficiently short orbital periods , tidal spin - up combined with orbital decay via gravitational radiation ensure that @xmath16 .
the orbital period at which this transition occurs is @xmath17 minutes , depending on the wd masses and temperatures [ see eq .
( 79 ) of paper ii ] . at periods
@xmath18 , the tidal energy transfer rate is [ edot2 ] , where @xmath19 is the moment of inertia of the wd , and @xmath20 is the binary inspiral time due to gravitational radiation , t_gw= 4.210 ^ 5yr ( ) ( ) ^1/3 ( ) ^ 8/3 .
[ tgw ] when the outgoing gravity waves damp in the wd envelope and locally deposit their angular momentum , some of the wave energy is converted into rotational kinetic energy , while the rest is converted to heat .
the heating rate is [ eheat ] _ heat = e ( 1 - ) .
if the wd maintains some differential rotation , @xmath3 in the above equation should be the rotation rate of the layer in which the waves damp , and heat will also be generated through viscous angular momentum transport .
our calculations indicate that the gravity waves reach non - linear amplitudes and break in the outer layers of the wd .
the location of wave breaking depends on various parameters ( e.g. , orbital and tidal frequencies ) , but is always at @xmath21 and the exterior mass @xmath22 ( paper ii ) .
since a small fraction of the stellar mass absorbs the entire angular momentum flux , the outer layer spins up rapidly .
if it spins up faster than angular momentum can be transported to the core , the outer layer will rotate synchronously with the orbit .
outgoing waves approaching the synchronized envelope will be absorbed near corotation and deposit their angular momentum , causing the synchronized envelope to move to larger depths ( see goldreich & nicholson 1989 ) .
we consider a simple two - zone model for the spin evolution of the wd . in this model ,
the envelope of the star rotates synchronously with the orbit ( @xmath23 ) , while the core rotates sub - synchronously ( @xmath24 ) .
the envelope and core are coupled , with angular momentum being transferred to the core according to a parameterized coupling time , @xmath25 .
the angular momentum of the core - envelope system evolves according to & & [ jedot ] ( i_env_env ) = _ z(,_core ) - ( _ env - _ core ) , + & & [ jcdot ] ( i_core_core ) = ( _ env - _ core ) , where @xmath26 is the moment of inertia of the envelope . here
, @xmath27 is the angular momentum flux which can be calculated from equation ( [ jdot ] ) .
we have assumed that the gravity waves are excited in the core and absorbed in the envelope ) , which is the region where the outgoing gravity waves are excited . ] .
consequently , the angular momentum source term @xmath27 is only present in the envelope evolution equation , although it is dependent on the tidal frequency in the core , @xmath28 . using @xmath29 , equations ( [ jedot ] ) and ( [ jcdot ] )
can be integrated to find @xmath30 and @xmath31 as a function of time or orbital period .
the mass @xmath32 of the envelope corresponds to @xmath33 .
the thickness ( or @xmath34 ) of the envelope is dependent on the parameter @xmath25 . in
stably stratified stars like wds , angular momentum can be transported by magnetic fields . in the presence of a poloidal field
@xmath35 connecting the core and envelope , @xmath25 can be estimated from the alfven wave crossing time , [ ta ] t_a = 10 ^ 2 yr ( ) for our co wd model . for wds without an intrinsic magnetic field , angular momentum may be transported via the tayler - spruit dynamo ( spruit 2002 ) . to estimate @xmath25
, we calculate the effective viscosity for angular momentum transport via the tayler - spruit dynamo , @xmath36 , as outlined in spruit 2002 . without including the effects of composition gradients in the wd [ see equation ( 32 ) in spruit 2002 ] .
a more realistic estimate of the rotational profile of the wd should take composition gradients into account .
] we find @xmath37 .
thus we expect the coupling time to lie in the range @xmath38 for the short orbital periods of interest .
the mass @xmath34 of the synchronized envelope as a function of orbital period for a @xmath39 co wd model with a @xmath40 companion . the solid ( black ) line has @xmath41 , the dot - dot - dashed ( green ) line @xmath42 , the dot - dashed ( orange ) line @xmath43 , and the dashed ( red ) line @xmath44 . ]
figure [ wdcrit ] plots the value of @xmath34 as a function of orbital period for our @xmath14 wd model with a @xmath45 companion , using values of @xmath25 ranging from @xmath46 to @xmath47 .
we begin our calculation at @xmath48 and use @xmath49 and @xmath50 , as appropriate at long orbital periods where tidal effects are negligible .
we see that for the range of @xmath25 considered , @xmath34 remains small ( @xmath51 ) at all orbital periods of interest .
thus , the synchronized envelope most likely does not extend down to the c / he transition layer where the gravity waves are excited , justifying our assumption that @xmath27 is a function of @xmath31 .
however , the envelope does extend to very large optical depths , suggesting that binary wds may be observed to be synchronized at large orbital periods even if their cores are not synchronized .
note that since @xmath52 , the core of the star contains most of the angular momentum , and its spin evolves in the same manner as discussed in paper ii .
in the two - zone model discussed in 3 , the total tidal heating rate @xmath53 may be calculated from equation ( [ eheat ] ) with @xmath54 , and the tidal heat is deposited entirely at the base of the synchronized envelope where @xmath55 . in a real wd , the heat deposition will occur over a range of depths that depends on the details of wave breaking and viscous angular momentum transport . for simplicity ,
here we choose to deposit the tidal heat uniformly per unit mass in the synchronized envelope .
the heating rate per unit mass , @xmath56 , is then @xmath57 although the radial dependence of this heating function is unlikely to be realistic , we find that the results below are not strongly dependent on the form of @xmath58 . to understand the effect of tidal heating on the wd properties , we evolve wd models using the extra heating term calculated via equation ( [ epsheat1 ] ) .
we use the one - dimensional stellar evolution code mesa ( paxton et al .
2010 ) to evolve our wd models , starting from an initial orbital period of one hour .
we present results for a @xmath14 co wd model with a @xmath59 hydrogen shell and a @xmath40 companion .
the surface temperature of the @xmath14 co wd model with a @xmath40 companion as a function of orbital period , for initial temperatures of @xmath60 k ( top ) and @xmath61 k ( bottom ) .
the solid black lines are calculated with @xmath62 while the dashed ( red ) lines are calculated with @xmath63 .
the dotted lines are calculated for a wd with no tidal heating and the same initial temperature .
the ( blue ) dot - dashed lines correspond to equation ( [ ttide ] ) .
the ( red ) stars mark the points at which tidal novae occur .
the asterisks mark the position of the secondary of the 12.75 minute binary wd system sdss j065133 + 284423 ( brown et al .
2011 ) . ] figure [ 6temp ] displays the surface temperature as a function of orbital period for our tidally heated wd . for comparison
, we also show the temperature of a non - tidally heated wd and the `` tidal heating temperature '' , defined as [ ttide ] t_eff , heat= ( ) ^1/4 . at long orbital periods ( @xmath64 minutes ) ,
the tidal heating has little effect on the surface temperature of the wd . at shorter periods ( @xmath65 minutes ) ,
the temperature becomes substantially larger due to tidal heating .
several of the curves end abruptly due to the ignition of a thermonuclear runaway event , at which point we terminate our evolution calculations . for small values of @xmath25 ,
the tidal heat is deposited at shallow depths and quickly diffuses to the surface such that the luminosity of the wd is @xmath66 , where @xmath67 is the luminosity of a non - tidally heated wd .
however , for larger values of @xmath25 , most of the tidal heat is deposited deeper in the wd where it can not quickly diffuse outward .
this leads to lower surface temperatures , although the internal temperature may increase substantially .
figure [ 63temp ] shows the interior temperature profile of our wd at three different orbital periods , using @xmath63 . at long orbital periods ,
the temperature profile is similar to that of a non - tidally heated wd .
as the orbital period decreases , the interior heats up , with the local temperature maximum at @xmath68 . if the base of the hydrogen layer reaches a temperature of @xmath69 k , hydrogen burning will be ignited .
temperature profile of the wd ( as a function of exterior mass @xmath70 ) at orbital periods of 45 minutes ( black ) , 20 minutes ( green ) , and 12 minutes ( red ) .
these temperatures are calculated for the @xmath14 wd model with an initial surface temperature of @xmath71 k , a @xmath40 companion and @xmath63 . ] in the depicted model , the layer just above the he / h transition ( at @xmath72 ) is composed of largely degenerate hydrogen gas .
the ignition of fusion in this layer can thus spark a thermonuclear runaway .
in general , our calculations show that these _ tidal novae _
occur only in initially cool wds ( @xmath73 k in the absence of tidal heating ) .
they do not occur in hotter wds because the hydrogen is not degenerate and can burn stably .
also , tidal novae require that the waves deposit some of the heat near the base of the hydrogen layer , i.e. , @xmath74 . in our two zone model ,
such heating occurs for coupling times @xmath75 .
overall , we find that the tidal novae occur at orbital periods @xmath76 , depending on the location of heat deposition , initial temperature of the wd , and companion mass .
we have shown that under rather general conditions ( see the last paragraph of section 4 ) , tidal dissipation in compact wd binaries can lead to nova outbursts prior to binary merger or mass transfer . while we do not attempt to predict the detailed observational signal of a tidal nova ( tn ) , we speculate that it may be very similar to a classic nova
. however , in contrast to classical novae in cvs , a tn would occur in a compact system with no evidence for mass transfer .
our results indicate that a tn would precede the beginning of mass transfer or merger by about @xmath77 yrs [ see eq .
( [ tgw ] ) ] , provided the conditions outlined in the previous paragraph are satisfied . in most classical novae
, the initial outburst is followed by a period of stable hydrogen burning at near the eddington luminosity , in which the hydrogen shell of the wd inflates to a radius of order @xmath78 .
however , the ultracompact nature of the wd system involved in a tn ( where @xmath79 ) may preclude such a phase because the stably burning hydrogen shell would inflate beyond the wd s roche lobe .
this shell may then accrete on to the companion star or be ejected from the system .
therefore , we expect most of the hydrogen to be burned or ejected during in a tn . in the absence of mass transfer to supply a fresh hydrogen , recurrent novae would be unlikely
. thus , the occurrence rate of these tn may be comparable to that of wd mergers involving a co wd
. our theory can be constrained by comparing the prediction of our tidal heating calculations to observed compact wd binaries .
the 12.75 minute system sdss jj065133 + 284423 provides the best opportunity ( brown et al .
this system is composed of a primary with @xmath80 k and mass @xmath81 , and a secondary with @xmath82 k and mass @xmath83 .
comparison with figure [ 6temp ] indicates that the luminosity of the secondary is likely dominated by tidal heating .
our result for a co wd with an initial temperature of @xmath60 k and a value of @xmath84 is most consistent with the observed temperature of the secondary .
these results indicate that a tn may occur in this system in the future . in principle ,
tidal heating may change the structure of the wd enough to alter the dynamics of gravity wave propagation .
however , we find that this is not the case ( i.e. , no interior convection zone forms ) , with the exception of a thermonuclear runaway event .
our simple two - zone model for the wd obviously needs improvement , and we have neglected the effects of mixing induced by the breaking gravity waves and viscous angular momentum transport . if the mixing is strong enough to smooth out the wd composition gradients , the dynamics of gravity wave excitation and tidal heat deposition may be altered . furthermore , if the surface hydrogen mixes into the wd interior where it burns , the surface hydrogen layer will be gradually depleted and a tn will not occur .
observations of the ejecta of classical novae indicate substantial enrichment with core elements , although the mixing mechanism is not well understood ( truran 2002 ) .
these and other aspects of tn in compact wd binaries warrant further study .
future observations may be able to test whether tn occur and in turn provide information about the tidal processes at work in wd binaries .
the observation of a nova - like event in a system with no evidence for mass transfer would be strong evidence for the existence of tn and for the tidal heating mechanism studied in this paper .
measurements of hydrogen surface abundances in compact wd systems could also constrain our theory .
the observation of a wd with a thick hydrogen envelope in a very tight ( @xmath85 minutes ) detached binary would indicate tn do not usually occur .
if wds in tight binaries are observed to have little to no hydrogen on their surface , this may indicate that tn have stripped the surface hydrogen , or that the hydrogen has been destroyed due to efficient mixing processes .
observations of compact binary wds detected in future surveys may provide opportunities to test these theories .
we thank bill paxton , lars bildsten , and eliot quataert for useful discussions .
jf acknowledges the hospitality ( fall 2011 ) of the kavli institute for theoretical physics at ucsb ( funded by the nsf through grant 11-astro11f-0016 ) where part of the work was carried out .
this work has been supported in part by nsf grant ast-1008245 , nasa grants nnx12af85 g and nnx10ap19 g . | compact binary white dwarfs ( wds ) undergoing orbital decay due to gravitational radiation can experience significant tidal heating prior to merger . in these wds ,
the dominant tidal effect involves the excitation of outgoing gravity waves in the inner stellar envelope and the dissipation of these waves in the outer envelope .
as the binary orbit decays , the wds are synchronized from outside in ( with the envelope synchronized first , followed by the core ) .
we examine the deposition of tidal heat in the envelope of a carbon - oxygen wd and study how such tidal heating affects the structure and evolution of the wd .
we show that significant tidal heating can occur in the star s degenerate hydrogen layer .
this layer heats up faster than it cools , triggering runaway nuclear fusion .
such `` tidal novae '' may occur in all wd binaries containing a co wd , at orbital periods between 5 min and 20 min , and precede the final merger by @xmath0-@xmath1 years . |
the detection of wind profiles in the @xmath21032 , 1038 resonance doublet in _ copernicus _ spectra of stars with spectral types between o4 and b1 @xcite provided the first evidence for the existence of high - energy , non - radiative processes in the outflows from hot stars .
the presence of this ion was surprising , since only the hottest o stars are expected to produce it directly through photoionization . in a pivotal paper , @xcite demonstrated that x - rays can produce sufficient by auger ionization of , which is the dominant ionization stage of oxygen in the winds of stars in the temperature range where is observed . the x - ray flux required by the auger mechanism
was subsequently detected in spectra of ob - type stars obtained by the _
observatory @xcite .
since the existence of is linked directly to the presence of x - rays , the shape and strength of the lines can be used to trace the distribution of hot , x - ray emitting gas in the winds of early - type stars @xcite .
x - rays are believed to be caused by the formation of strong shocks in the wind , which convert mechanical energy of the flow into localized sources of heat .
these shocks could arise from various processes , such as : ( a ) the `` line de - shadowing instability '' intrinsic to the line - driving mechanism responsible for hot - star winds @xcite ; ( b ) the interfaces of large - scale co - rotating interactions regions ( cirs ) , which might be responsible for recurrent variability in hot - star winds @xcite ; or ( c ) collisions ( channeling ) of material along closed ( open ) magnetic fields emanating from the surface of the star @xcite .
since these mechanisms involve the evolution of complicated , non - stationary flows , their tracers might be expected to exhibit substantial variations .
for example , if the bulk of the x - rays or ions are produced by a few strong shocks ( as in the case of cirs and simple magnetic field configurations ) , large variations might be expected as the distribution or strength of the shocks evolves .
conversely , little or no variability would be expected if the global distribution of the shocks ( e.g. , resulting from many ensembles of shocks generated by the line - deshadowing instability ) do not change much with time . in either case ,
fluctuations in the tracers of the high energy processes provide information about the origin and distribution of the hottest gas in these outflows . until recently ,
x - rays were the predominant tracers of shock phenomena in the winds of hot stars .
the limited x - ray data available suggest little short - term variability , which is generally interpreted as implying that many shock ensembles are present in the wind ( see , e.g. , * ? ? ?
* ) , so that the evolution of any one of them is of little consequence . however , a variety of selection effects might also bias this interpretation ; see , e.g. , @xcite . at a fundamental level , spatially localized variations in the x - ray flux
may be difficult to detect against the background contributions collected from nearly the entire volume of the wind .
this situation changed dramatically with the launch of the _ far ultraviolet spectroscopic explorer ( fuse ) _
satellite , which provides routine access to the resonance lines of . observations of these wind features permit more sensitive searches for variability , primarily because they are formed by the resonant scattering of photons in localized regions of the wind ( i.e. , the column of material projected against the disk of the star ) .
such observations also permit wind features to be identified , even though they are otherwise difficult to detect because of blends with strong stellar and interstellar lines . moreover , the _ fuse _ spectrographs are themselves very sensitive , so that ob - type stars in the magellanic clouds are easily accessible . in order to exploit this new capability
, we designed several principal investigator ( pi ) team programs to obtain sparse time series observations for a large sample of ob - type stars in the milky way and magellanic clouds .
these time series were obtained by breaking long integrations into two or three observations separated by intervals that typically amounted to a few days .
although these time series provide the bare minimum of information required to detect variations , they permitted a broad range of stellar parameters to be surveyed .
the aim was to provide a rudimentary assessment of the frequency of variability as a function of spectral type and luminosity class ; to determine whether was as variable as the stellar wind features in and ; and to glean whatever information possible about the nature of the variations .
although this approach is statistical in nature , it complements other studies undertaken with _ fuse _ designed to characterize stellar wind variability in and other far ultraviolet wind lines for specific targets ( e.g. , * ? ? ?
it also provides the information required to assess the effect of stellar wind variability on measurements of the interstellar lines ( see , e.g. , * ? ? ?
in tables [ t1 ] and [ t2 ] , we summarize the milky way ( mw ) and magellanic clouds ( mc ) observations , respectively
. basic properties ( galactic coordinates , spectral - type , magnitude and reddening , terminal and projected rotational velocities ) are indicated .
columns 7 and 8 show the separation in days between two successive observations and the dates when they were taken , respectively .
most of these sparse time series consist of 23 observations separated by intervals ranging from a few days to several months .
the ninth column identifies the rootname of the _ fuse _ data set , and is followed by remarks concerning the presence of radial velocity variations in the object .
the time - series observations were obtained mainly through two large programs conducted by the _ fuse _ pi team , namely the ism " ( p101 , p102 , p122 ) and hot stars " ( p117 ) working group investigations .
the purposes of these multiple observations were to investigate the time - variability in the winds of early - type stars from a stellar point of view ( this paper ) and to assess its impact on the interstellar absorption measurements of @xcite .
other ( non - proprietary ) data from guest investigator programs were also included ( observation rootnames starting with a and b ) , as were targets obtained as early - release observations ( rootnames starting with x ) .
table [ t1 ] indicates that the mw sample is strongly biased toward later type stars ( o9-b0-b0.5 ) , while the mc sample ( table [ t2 ] ) is slightly more uniformly distributed but of smaller size .
these selection effects limit the conclusions that can be drawn : the sample is biased in spectral type ; it is not large enough for a meaningful statistical analysis to be performed for each spectral and luminosity class ; and the number of multiple observations is very small .
it is nonetheless the largest sample available to study variability in the winds of early - type stars .
a total of 66 stars ( 44 galactic stars , 20 lmc stars , and 2 smc stars ) were observed with typically 2 or 3 observations , except for a few cases with 4 exposures .
a future complementary study will involve the analysis of the high time - resolution observations of two lmc stars @xcite .
_ fuse _ consists of four co - aligned telescopes and spectrographs , termed `` channels '' , two of which have sic coatings to maximize the sensitivity in the wavelength range @xmath39051105 , while the other two have al - lif coatings for sensitivity in the @xmath310001187 range .
this wavelength region provides access to a rich suite of resonance and excited lines , in particular for the ions considered here : , , and .
the channels are referred to as sic1 and sic2 and lif1 and lif2 , respectively .
this design includes overlap at certain wavelengths , which helps to distinguish between real features in the spectra and instrumental artifacts , particularly fixed - pattern noise .
since we are interested in time - variability in the spectra , this redundant coverage is essential to make sure the differences are in the spectrum of the source and not due to detector defects .
more complete descriptions of the design and performance of the _ fuse _ spectrographs are given by @xcite and @xcite , respectively .
most of the observations were obtained through the @xmath4 ( lwrs ) apertures , except for a few cases where they were obtained through the @xmath5 ( mdrs ) and @xmath6 ( hirs ) apertures .
since flux can be lost through the narrower apertures , systematic multiplicative corrections were applied to adjust the flux levels to a common value .
these corrections do not affect the detection of variations .
in a few cases , however , and especially for lif1b or lif2a ( which contain the lines ) , the adjustments are compromised by contamination from the `` worm , '' a feature that deforms the continuum non - linearly .
the worm is understood to be due to shadowing from grid wires in front of the detector @xcite .
standard processing with version 2.0.5 of the calfuse calibration pipeline software was used to extract and calibrate the spectra .
the software screened the data for valid photon events , removed burst events , corrected for geometrical distortions , spectral motions , satellite orbital motions , and detector background , and finally applied flux and wavelength calibrations .
the extracted spectra associated with the separate exposures of a given observation were aligned by cross - correlating the positions of strong interstellar lines , co - added , and rebinned to a spectral resolving power of @xmath7 ( @xmath8 @xmath0 ) . to compare to wind variability in , we chose to examine the lines of and based on two main criteria : ( i ) they are present over a wide range of spectral types ( i.e. , and are dominant ionization stages or near dominant ionization stages , see below ) .
( ii ) the lines fall in sections of the spectrum where they are minimally contaminated by blending with interstellar features . at wavelengths below 1000 ,
the signal - to - noise level becomes significantly weaker because the effective areas of the sic channels are smaller .
below , we briefly recall some properties of the ions under consideration and their behavior through the h - r diagram of ob - type stars .
this is mostly based on the two recent _ fuse _ atlases of ob - type stars in the mw @xcite and the mc @xcite owing to the high ionization required to form ( i.p .
: 113.9138.1 ev ) , the @[email protected] , 1037.617 resonance doublet is the most useful tracer of high - energy processes in the optical / uv region of the spectrum .
its prevalence is believed to be directly related to the distribution of x - rays in the winds of the hot stars . in this sense , is often referred to as a `` super - ion '' ( see , e.g. , * ? ? ?
other super - ions exist in fuv spectra ( e.g. , and ) , but remains unique because it is two stages above the dominant ionization stage of . because of this , auger ionization of is believed to be the dominant process for producing @xcite . throughout the early o - type dwarf stars and from o2 to b0 supergiants , the lines exhibit typical p cygni profiles @xcite , which consist of an extended blue - shifted absorption trough and red - shifted emission peak . because the separation between the lines is only 1650 @xmath0 , the absorption trough of @xmath91038 is often blended with the emission part of @xmath91032 .
similarly , the absorption trough of @xmath91032 can be lost in the saturated absorption line of ly@xmath10 . moreover , several other interstellar absorption lines ( , , and h@xmath11 ) can complicate the appearance of stellar line profiles .
we discuss the different contaminations of the wind profiles in more detail in section [ tv ] .
one should refer to the normalized spectra in the mc _ fuse _ atlas of ob - type stars @xcite , and to the unnormalized spectra of the mw _ fuse _ atlas of ob - type stars @xcite for illustrations of these spectra as a function of spectral type .
@[email protected] , 1128.008 ( i.p . : 51.465.0 ev , similar to ) is a resonance doublet expected to be near the dominant ionization stage in the winds of o - type stars @xcite .
it weakens substantially in spectra of b0 and later type stars , for which becomes the dominant ion .
even though it is a dominant ion , it remains unsaturated because of its very low cosmic abundance ( @xmath12 = 5.57 $ ] compared to 8.87 for o and and 7.27 for s ) .
this doublet has a luminosity dependence . in mid - o spectra
, it appears mainly as photospheric absorption in dwarfs , a weak p cygni profiles in giants , and strong p cygni profiles in supergiants @xcite .
@[email protected] , 1072.973 ( i.p .
: 34.847.3 ev ) is one stage below the expected dominant ion stage of s in the winds of early o - type stars , but starts to dominate in late o and early b - type stars .
@xmath91063 is a resonance line , while @xmath91073 arises from low - lying excited levels that produce two transitions at wavelengths 1072.973 and 1073.516 . since the transition at 1072.973 is
about 9 times stronger than the transition at 1073.516 , we consider only the transition at 1072.973 .
these transitions have a similar luminosity effect to the doublet for mid - o spectral types .
however , in the spectra of early b - type stars , the lines are nearly independent of luminosity @xcite .
we note that even though the ion fraction of both and could be substantially larger than the ion fraction of , their low abundance ( 0.05% and 2.5% as abundant as o , respectively ) could make their detection difficult , especially in the mc where the metallicity is down by about @xmath13% compared to the mw .
several careful steps were taken to uncover time - dependent changes in the profiles of these sparse time series observations : + ( i ) for each species , we first transformed the spectra from wavelength space to velocity space in order to facilitate direct comparisons between the components of the doublet and lines from different species .
this velocity is in the heliocentric frame , but , unfortunately , the velocity zero point in _ fuse _ spectra remains uncertain by several tens of @xmath0 .
the relative velocity can be accurate to a few @xmath0 within one channel , but can vary by more than 10 @xmath0 between channels . using the velocity scale , we compared both the calibrated fluxes and the ratio of the fluxes .
+ ( ii ) the blue component of the doublet is affected by the airglow emission and the interstellar absorption of ly@xmath10 , which lies 1805 @xmath0 blueward of @xmath91032 . in tables
[ t1 ] and [ t2 ] , we give the terminal velocity , @xmath14 , which should give roughly the extent of the wind profiles . is implied . ]
when @xmath14 is greater than 1805 @xmath0 , the blue component of the doublet is completely absorbed by the interstellar ly@xmath10 line .
therefore , the @xmath91038 component was first checked for possible variations .
if the ratio of the fluxes in the time series is a horizontal line , we conclude that the wind profiles were the same at the times when the observations occurred .
note that for the red component of the doublet there is generally no information at velocities below 400500 @xmath0 because of blending with the saturated interstellar @xmath91036 line .
+ ( iii ) if there were significant changes in the wind profiles between the different observations , we cross - checked that the profiles ( and the ratio of the fluxes ) were the same in the redundant spectra extracted from different channels of the observations . the @xmath9@xmath91032 , 1038 resonance doublet appears in lif1a , lif2b , sic1a , and sic2b spectra . because the mirrors and gratings move slightly and can be misaligned , information in one channel is occasionally lost , but usually there is more than one channel available .
this step is especially important with _ fuse _ observations because the aging of detector components can introduce shifts and distortions in the spectral features ( known as the walk " ) that could mimic the variations we are looking for .
however , this effect is larger near the edge of detector segments sic1a and sic2b and rather small near the center of detector segments lif1a and lif2b .
the walk problem is also not expected to occur at the same position with the same amplitude in the different detector segments .
thus , by carefully comparing the wind profiles from different segments , we can determine if the observed variation is intrinsic to the stellar wind .
+ ( iv ) further information comes from the @xmath91032 transition when this line is not too contaminated by the ly@xmath10 airglow or absorption or airglow .
we required that when @xmath91032 is available , it exhibits similar variations at the same velocities as the @xmath91038 component .
if the wind profiles are unsaturated , we can also check that the line strength of @xmath91032 is approximately twice the value of @xmath91038 , as expected for optically thin absorption . since the velocity separation between the doublet is only 1650 @xmath0 , there are cases where the wind profiles can overlap . + ( v ) similar procedures applied for and , but because the doublet separations of these species are much larger than for ( 2900 and 2480 @xmath0 , respectively ) , there is generally no problem of overlapping wind profiles .
there is also less blending with saturated interstellar lines .
when strong h@xmath11 lines are present , they can contaminate @xmath91073 at velocities greater than 2000 @xmath0 and @xmath91063 at velocities less than 200 - 500 @xmath0 .
for , there are generally no saturated interstellar lines , but possible blends with photospheric @[email protected] , 1124.9 and @[email protected] , 1128.3 can complicate the interpretation of these profiles .
we note that the variations of @xmath91073 generally appear larger than for @xmath91063 , contrary to the expectation based on their rather similar oscillator strengths . a similar problem with the relative oscillator strengths of these transitions
was noted by @xcite .
these procedures ensured that the observed variations occur in the stellar wind profile , and also that weak variations could be detected in spectra characterized by signal - to - noise ratios of @xmath8 .
figure [ fig1 ] shows 3 examples from the galactic sample , for which we overplotted the spectra taken at two or three different times .
the differences of these fluxes are also plotted for @xmath9@xmath91032 , 1038 , @xmath9@xmath91062 , 1073 , and @xmath9@xmath91118 , 1128 versus velocity . in the following figures , we instead plot @xmath15 $ ] as a function of @xmath16 , which has more physical meaning than a flux difference because it is related to the optical depth of a wind absorption .
the left - hand panel of figure [ fig1 ] shows an example where no variation is detected .
we note a slight change in the difference of the profiles around @xmath17 @xmath0 , probably near the terminal velocity of this star . however , no such change is observed in spectra from other segments , and we therefore rule out wind variability in this case .
the right - hand panel shows a large amplitude variation that extends over a large fraction of all profiles .
the middle panel shows a very small variation that extends over @xmath3500 @xmath0 only in the profile . in this case the redundant information provided by both components of the doublet permits small variations in comparatively noisy spectra to be detected unambiguously .
note that in both the middle and right - hand panels , the effect of the worm can be observed in @xmath18 for velocities @xmath19 @xmath0 . yet by following the above procedure , we can distinguish between wind variability and instrumental artifacts .
see also @xcite for other examples .
wholesale shifts in the positions of photospheric lines indicate systematic motion of the star , possibly in a binary system .
such motion produces a characteristic `` s '' shape in ratios or differences of spectra at a location that depends on the shift in radial velocity between the two spectra , and over a wavelength region that depends on the width of the photospheric line .
we typically noted variations of this sort near the systemic velocity of the star , which indicates only small radial velocity variations .
figure [ fig2 ] ( upper - left panel ) shows an example of such an `` s''-shape near the systemic velocity ( 0 @xmath0 ) in the profiles of and , which are largely photospheric .
these variations are not seen in the line , which instead shows substantial variations between @xmath20 and @xmath21 @xmath0 ; i.e. , in the outflowing wind .
an entry of `` rv var ''
( rv var ? ) in the last column of tables [ t1 ] and [ t2 ] indicates the presence ( suspected presence ) of photospheric radial velocity variations in the _ fuse _ time series .
in figure [ fig2 ] , the logarithm of the ratio of the fluxes observed at different times is presented as a function of the velocity for 4 examples in the mw and in the lmc , for @xmath21032 , 1038 , @xmath21062 , 1073 , and @xmath21118 , 1128 .
based on analogous figures constructed for the whole sample , we summarize in table [ t3 ] our assessment of whether the wind profiles for these ions were variable at the time of these observations .
we also indicate the velocity range over which the variation is observed . in some cases
, the high velocity value for was derived using the other species as there is overlap of the wind profiles at velocities larger than 1650 @xmath0 .
for some cases particularly in the mc sample no information is given concerning the velocity extent of the wind variations due to uncertainties resulting from the lower signal - to - noise of the data and the greater degree of contamination by and airglow emission lines .
( n ) " in table [ t3 ] indicates that the spectra suffer either from low signal - to - noise level or some problem in the fluxes , or that the wind is contaminated by other stellar lines ( in the case of radial velocity variables , and particularly for the lines ) .
thus when
( n ) " is listed , detections of variability are more uncertain .
several inferences can be drawn from these measurements : \(1 ) for the whole sample , 56% of the wind profiles vary with time . for the mw @xmath22 ( 55% ) and @xmath23 ( 59% ) for the mc
show variability .
the frequency of wind variability is therefore similar in both samples .
if we remove 8 stars with spectral type later than b1 for which is unlikely to be present @xcite , 64% of the combined sample is variable .
in contrast , only 15% and 5% of the whole sample vary in the wind profiles of and .
these latter percentages should be treated cautiously because the strengths of these ions vary with spectral type and luminosity class ( see [ ions ] ) .
they are expected to be strong in the winds of supergiants , and for those stars , the profiles are often variable ( see tables [ t1 ] and [ t2 ] ) .
\(2 ) figure [ fig3 ] shows the fraction of stars for which we observed wind changes as a function of spectral types and luminosity classes , where the error bars are @xmath24 , with @xmath25 the total number of stars in a given bin .
the sample is not large enough to draw any definite conclusions , but suggests that the incidence of variability increases for earlier spectral types .
this , however , could also be a selection effect caused by the stronger appearance of the p cygni profiles in the earlier - type stars .
when all the spectral types are considered , all luminosity classes have a similar percentage of wind variability .
however , if we take a more homogeneous sample and consider only spectral types between o9 and b0.5 , we find that the wind profiles vary in 7/10 ( 70% ) cases for supergiant stars ( luminosity class i ) , 8/12 ( 67% ) cases for giant stars ( luminosity class ii - iii ) , and 6/13 ( 46% ) cases for the dwarf stars ( luminosity class iv - v ) .
the results are , however , not statistically significant at the @xmath26 level , and are insufficient to claim that wind variability occurs more frequently in supergiant and giant stars compared to dwarfs
. we note that for and , variability is observed only in supergiants ( except one case observed in a class ii star ) .
this is not surprising because in less luminous stars these lines are predominantly photospheric .
\(3 ) from table [ t3 ] , the full - width of the variation in the profiles ranges from approximately 225 @xmath0 to 2200 @xmath0 , but most fluctuations occur between @xmath3400 @xmath0 and 1100 @xmath0 .
the sample is not large enough to study spectral type or luminosity dependence ; but in any case no dependence is indicated .
\(4 ) the strength and velocity of the variation can change over times as short as @xmath27 day . however , detailed observations with high - time resolution are required to characterize the time scales rigorously .
\(5 ) in figure [ fig4 ] , the full - width of the observed variations ( @xmath28 ) , and the maximum absolute radial velocity where significant variability occurs ( @xmath29 ) are plotted as a function of @xmath1 . to investigate possible correlations
, we used the spearman rank - order correlation test , which has the advantage that the statistical significance of a non - zero rank correlation can be quantified reliably without any assumption concerning the distribution of uncertainties .
we denote the rank - order correlation coefficient by @xmath30 , and its statistical significance by @xmath31 .
the statistical significance is defined in the interval [ 0,1 ] , where small values indicate greater significance .
+ although the top panel of figure [ fig4 ] suggests some correlation between @xmath28 and @xmath1 , the spearman s correlation test gives @xmath32 with @xmath33 ( for the mw sample ; @xmath34 with @xmath35 for the whole sample )
. the width over which the wind profile of the varies is only weakly correlated with the terminal velocity .
a least - squares fit to the mw sample gives @xmath36 @xmath0with a minimum @xmath37 and a goodness of fit of 0.09 .
+ in contrast , there is a significant correlation between @xmath29 and @xmath1 as @xmath38 with @xmath39 ( for the mw sample ; @xmath40 with @xmath41 for the whole sample ) .
a least - squares fit to the data ( mw and mc ) gives @xmath42 @xmath0with a minimum @xmath43 and a goodness of fit of 0.31 .
the trend is close to the 1:1 relationship as illustrated in figure [ fig4 ] .
\(6 ) the absolute optical depth variation can be characterized approximately by defining @xmath44 , where @xmath45 is the maximum positive logarithmic ratio of the fluxes observed at a specific velocity , and @xmath46 is the maximum absolute negative logarithmic ratio of the fluxes associated with the variation .
we find a wide range of values for @xmath91032 : @xmath47 .
the detection of @xmath480.2 is precluded by the noise level in these data .
\(7 ) @xmath49 does not seem to depend on the luminosity class or spectral type of the star ( see figure [ fig2 ] ) .
the largest values of @xmath49 are observed for stars in the lmc .
( note that the @xmath50-scale in figure [ fig2 ] is larger for the lmc compared to the mw . )
in the previous section , we assessed the frequency of variability in the wind line for many ob - type stars .
extensive studies of the winds through ultraviolet resonance lines in other species show that narrow or discrete absorption components ( dacs ) are associated with the extended absorption troughs ( for a recent review , see * ? ? ?
these dacs become particularly strong near @xmath1 . the typical p cygni profiles observed for lower ionization stages ( e.g. , , )
are also observed for the doublet @xcite .
figure [ fig5 ] shows that strong , high velocity absorption components are also observed in .
several pieces of evidence confirm that these absorptions occur in the wind : ( i ) no strong photospheric lines were identified in this region of the spectrum ( we note , however , potential photospheric and lines that can contaminate the spectrum for the later type stars ; see zsarg et al . 2003 ) ; ( ii ) the velocity of the features changes from one observation to the next ( for the best case see figure 3 in * ? ? ?
* ) ; ( iii ) both components of the doublet can be observed ( see figure [ fig5 ] and lehner et al .
2001 ) ; and ( iv ) in cases of a binary , aligning the photospheric lines do not align these features .
the projected rotational velocity also broadens the photospheric lines , and there is no correlation between the width of these absorption features and the @xmath51 values listed in table [ t1 ] .
finally , these high - velocity features do not have an interstellar origin because in some cases they are too broad .
furthermore , transient features are not commonly observed in the interstellar gas .
these features are therefore due to absorption in the stellar wind . in table
[ t4 ] , we present the stars for which we observed unambiguously the wind absorption features and their blue - shifted velocities , @xmath52 , which correspond to the deepest part of these features .
( when the feature has a flat bottom , its centroid is used ) .
the last column also presents the strength of these features , classified as weak , medium and strong .
a weak " ( small width and small optical depth ) absorption corresponds to features as observed in the spectra of hd161807 ; a medium " ( intermediate ) absorption corresponds to features as observed in the spectra of cpd@xmath53721184 ; and a strong " ( large width and large optical depth ) absorption corresponds to features as observed in the spectra of hd92554 ( see figure [ fig5 ] ) . sometimes weak and strong features are observed in the same spectrum .
this is the case , e.g. , for hd168941 , for which the @xmath54 wind feature is blended with the @xmath55 absorption interstellar line .
we take this qualitative approach rather than a quantitative approach because the continuum is very uncertain for these generally complex features ( but see zsarg et al .
2003 for a more quantitative approach ) . below we
summarize the properties of these wind absorption features : \(1 ) they are observed in 33% of our sample ( 34% in the mw , 32% in the mc ) , in the spectra of stars with spectral types from o7.5 to b0.5 and in every luminosity class . since strong absorption components are preferentially found near @xmath1 ( see point 4 below ) they are especially difficult to detect for stars with @xmath56 @xmath0 owing to blending with the strongly saturated interstellar ly@xmath10 line .
stars with such terminal velocities range from o3 to about o7-o8 @xcite
. furthermore , these stars typically exhibit strongly saturated p cygni profiles in the resonance lines , which makes the detection of localized absorption enhancements all the more difficult .
a similar problem arises when @xmath57 @xmath0 because of the contamination of the wing of interstellar ly@xmath10 and absorption lines .
stars with such terminal velocities typically have spectral types later than b0.5 @xcite .
\(2 ) their shapes are usually not gaussian and can be asymmetric .
\(3 ) they appear with different strengths , and their full - widths at half maximum vary approximately from less than 100 @xmath0 up to 400500 @xmath0 , the latter generally being observed near @xmath1 ( see point 4 ) .
\(4 ) for the strong features , @xmath52 is similar to @xmath1 ( @xmath58 , except for sk@xmath536705 where @xmath59 ) , while for weak and medium features , @xmath52 is generally red - shifted by a few hundred @xmath0 with respect to @xmath1 ( @xmath60 ; except for sk@xmath5367101 where @xmath61 ) .
\(5 ) some features appear to vary on time scales of a day , but other features are at the same velocity after several days or months have passed .
\(6 ) the sparse time - series observations suggest that these features accelerate slowly with an acceleration of @xmath62 to less than @xmath3@xmath63 kms@xmath64 ( see figure [ fig5 ] and figure 3 in lehner et al .
2001 ) .
\(7 ) there is no direct relation between the widths of these features and the @xmath51 of star .
neither their width nor strength correlates with luminosity class or spectral type .
\(8 ) there is no systematic difference in the frequency of occurence between the mw and mc sample , though the size of the sample is certainly not large enough to draw any definitive conclusions .
we have described a survey for stellar wind variability based on sparse time - series observations of a large sample of stars observed with _
fuse_. this survey demonstrates that the winds in ob - type stars are variable in the doublet on various scales both in time and velocity .
the typical signatures of wind variability are observed via ( i ) variation in the profiles for 56% of the spectral types from o3 to b3 and 64% of the spectral types from o3 to b1 , and ( ii ) wind absorption features in 33% of the cases for the whole sample , or 46% of the cases for spectral types from o3 to b0.5 .
these percentages have to be considered lower limits .
owing to the nature of this survey , only a few exposures separated by a few days or a few months were obtained with no regard to the time scales believed to be relevant to the variability of hot - star winds .
furthermore , the quality of _ fuse _ data precludes the detection of small - amplitude variations . in view of the high incidence of variability detected under these circumstances ,
variability in wind profiles is likely ubiquitous . at spectral types later than b1 , no wind variability is observed .
this could be due to limitations in the detectability of wind activity caused by the weakness or absence of this feature in these stars , as well as the smaller number of stars observed in those spectral types . in the former case ,
this might be explained by the fact that for later spectral types , becomes the dominant ion , so that can not be produced via auger ionization @xcite .
the sample is not large enough to relate the frequency of the temporal variations or the absorption features to specific spectral types or luminosity classes .
but it _ suggests _ that stars with spectral types earlier than o8 may show variations more often .
the width over which the wind profile of the varies is poorly correlated with the terminal velocity for the mw sample .
there is , however , a significant correlation ( close to a 1:1 relationship ) between the blue - edge velocity of the variation ( i.e. , at the maximum ( negative ) velocity at which no variation is observed ) and @xmath1 ( see figure [ fig4 ] ) .
this might imply that systematically stronger shocks occur at larger velocities ( i.e. , near @xmath1 ) .
although the high velocity absorption features seen in directly indicate the outflow of hot gas , it is not clear from this survey exactly how these absorption features evolve with time .
figure 3 in @xcite and figure [ fig5 ] give nice examples of velocity - varying absorption wind features , but more intensive time - series observations are required to characterize this evolution ( see * ? ? ?
we note , however , that the stronger ( both in optical depth and width ) components are near the terminal velocity , while weaker components appear and evolve at lower velocities . by comparing the frequency of variability exhibited by to that seen in and , we found that only a small percentage of the whole sample shows wind variations for the lower ions , and those variations are mainly observed in supergiant stars .
this may be largely due to two selection effects .
one is that s and p are much less abundant than o , thereby making it more difficult to detect variations in these ions .
the other is that our sample is biased toward late o - type and early b - type stars , where the wind becomes weaker for ;
i.e. , the line becomes principally a photospheric line .
the low frequency of observed variability in is harder to understand , because is expected to be the dominant ion in the winds of late o- and early b - type stars .
if it is not merely an abundance selection effect , then this result may imply that the highly ionized component of these cooler winds is enhanced compared with hotter winds .
suppose , e.g. , that the variability exhibited by a wind line is proportional to its optical depth ; i.e. , @xmath65 .
the ionization fraction is @xmath66 ( see , e.g. , massa et al .
2002 ) , where @xmath67 is the abundance of the element . since variability is detected more frequently than variability , @xmath68@xmath69@xmath70 , i.e. @xmath71@xmath72@xmath70 , and @xmath73@xmath74@xmath70 .
if is dominant , @xmath73@xmath75 , and hence @xmath73@xmath76 , which is at least an order of magnitude larger than typically observed for early- and mid - o type stars @xcite . this evidence for an excessive abundance of in the winds of comparatively cool stars where it is least expected is consistent with the predictions of @xcite , who concluded that the presence of x - rays can substantially alter the ionization balance of lower - density winds .
this work is based on data obtained for the guaranteed time team by the nasa - cnes - csa fuse mission operated by the johns hopkins university .
financial support to u. s. participants has been provided by nasa contract nas5 - 32985 .
this research has made use of the nasa astrophysics data system abstract service and the simbad database , operated at cds , strasbourg , france .
we thank the referee , doug gies , for helpful comments that improved the presentation of these results .
bianchi , l. , et al .
2000 , , 538 , 57 cassinelli , j. p. , & olson , g. d. 1979 , , 239 , 304 crampton , d. , bernard , d. , harris , b. l. , & thackeray , a. d. 1976 , , 176 , 683 cranmer , s. r. & owocki , s. p. 1996
, , 462 , 469 conti , p. s. , garmany , c. d. , & massey , p. 1986
, , 92 , 48 feldmeier , a. , puls , j. , & pauldrach , a. w. a. 1997 , , 322 , 878 fitzpatrick , e. l. 1988 , , 335 , 703 fitzpatrick , e. l. 1991 , , 103 , 1123 fullerton , a. w. , massa , d. l. , prinja , r. k. , howarth , i. d. , willis , a. j. , & owocki , s. p. , in prep garrison , r. f. , hiltner , w. a. , & schild , r. e. 1977 , , 35 , 111 guetter , h. h. 1968 , , 80 , 197 harnden , f. r. et al .
1979 , , 234 , l51 hill , p. w. 1970 , , 150 , 23 hill , p. w. , kilkenny , d. , & van breda , i. g. 1974 , , 168 , 451 hiltner , w. a. 1956 , 2 , 389 houck , n. , & cowley , a. p. 1975 , michigan spectral survey , ann arbor , dep .
michigan , 1 houck , n. , & smith - moore , m. 1988 , michigan spectral survey , ann arbor , dep .
michigan , 4 howarth , i. d. , siebert , k. j. , hussain , g. a. j. , & prinja , r. k. 1997 , , 284 , 265 jaxon , e. g. , guerrero , m. a. , howk , j. c. , walborn , n. r. , chu , y .- h . , & wakker , b. p. 2001
, , 113 , 1130 karlsson , b. , 1969 , arkiv for astron .
, 5 , 241 lehner , n. , fullerton , a. w. , sembach , k. r. , massa , d. l. , & jenkins , e. b. 2001 , , 556 , l103 lennon , d. j. 1997 , , 317 , 871 macfarlane , j. j. , cohen , d. h. , & wang , p. , 1994
, , 437 , 351 macfarlane , j. j. , waldron , w. l. , corcoran , m. f. , wolff , m. j. , wang , p. , & cassinelli , j. p. 1993
, , 419 , 813 maz - apellniz , j. & walborn , n. r. 2002 , in proc .
212 , a massive star odyssey , from main sequence to supernova " , ed . k. a. van der hucht , a. herrero , & c. esteban ( san francisco : asp ) , 560 massa , d. l. , fullerton , a. w. , sonneborn , g. , hutchings , j. b. 2003 , , 586 , in press moos , h. w. , et al .
2000 , , 538 , l1 morgan , w. w. , code , a. d. , & whitford , a. e. 1955 , , 2 , 41 morton , d. c. 1979 , , 189 , 57 mullan , d. j. 1984 , , 283 , 303 oskinova , l. m. , ignace , r. , brown , j. c. , & cassinelli , j. p. 2001
, , 373 , 1009 owocki , s. p. , castor , j. i , & rybicki , g. b. 1988 , , 335 , 914 patriarchi , p. , & perinotto , m. 1992 , , 258 , 285 pellerin , a. et al .
2002 , , 143 , 159 penny , l. 1996 , , 463 , 737 prinja , r. k. 1998 , in _ cyclical variability in stellar winds _ , ed .
l. kaper & a. w. fullerton ( berlin : springer ) , 92 prinja , r. k. , barlow , m. j. , & howarth , i. d. 1990 , , 361 , 607 prinja , r. k. , & crowther , p. a. 1998 , 300 , 828 sahnow , d. j. , 2002 , in _ the fuse instrument and data handbook _ ,
version 2.0 sahnow , d. j. , et al .
2000 , , 538 , l7 schild , r. e. 1970 , , 161 , 855 seward , f. d. , forman , w. r. , giacconi , r. , griffiths , r. e. , harnden , f. r. , jones , c. , & pye , j. p. 1979
, , 234 , l55 snow , t. p. & morton , d. c. 1976 , , 32 , 429 ud - doula , a. & owocki , s. p. 2002
, , 576 , 413 uesugi , a. , & fukuda , i. 1982 , revised catalogue of stellar rotational velocities , department of physics & astronomy , kyoto university walborn , n. r. 1971 , , 23 , 257 walborn , n. r. 1972 , , 77 , 312 walborn , n. r. 1973 , , 78 , 1067 walborn , n. r. 1976 , , 205 , 419 walborn , n. r. 1977 , , 215 , 53 walborn , n. r. 1982 , , 87 , 1300 walborn , n. r. , fullerton , a. w. , crowther , p. a. , bianchi , l. , hutchings , j. b. , pellerin , a. , sonneborn , g. , & willis , a. j. 2002 , , 141 , 443 walborn , n. r. , lennon , d. j. , haser , s. m. , kudritzki , r .-
voels , s. a. 1995 , , 107 , 104 zsarg , j. , fullerton , a. w. , lehner , n. , & massa , d. 2003 , , submitted figure [ fig1 ] : examples of _ fuse _ spectra for three galactic stars for the @xmath21032 , 1038 , @xmath21062 , 1073 , and @xmath21118 , 1128 doublets . the spectra and difference spectra are shown in units of @xmath77 ergcm@xmath64s@xmath79@xmath79 . for hd64568 ( left panel ) ,
the observations were separated by 2.5 days ; for hd47417 ( middle panel ) , the observations were separated by 0.8 day , and for hd210509 ( right panel ) , the observations were separated by 1.4 and 1.3 days .
the black , red , and blue profiles denote the first , second , and third observations , respectively .
the dotted line in the right panel indicates the terminal velocity of hd210509 .
the position of the lyman @xmath80 airglow line is indicated in the uppermost panels .
the range of variability in the line of hd47417 is emphasized by vertical red lines in difference flux panel .
figure [ fig2 ] : the logarithm of the ratio of the fluxes ( taken at different times indicated by the @xmath81 in days ) vs. velocity for the @xmath21032 , 1038 , @xmath21062 , 1073 , and @xmath21118 , 1128 for 4 stars in the galaxy and the lmc , showing the variety of variations in the profiles .
note that the @xmath50-scale of a panel is the same for a given species but can change from species to species .
vertical dotted lines indicate @xmath1 , while the extent of the variation is shown by the solid thick vertical lines .
the variation in the profile ratio of and in the spectra of hd187459 at @xmath82 @xmath0 indicates radial velocity variations , perhaps due to motion in a binary system .
figure [ fig3 ] : fraction of stars in the combined galactic and magellanic clouds sample exhibiting wind variability for different temperature ( left panel ) or luminosity classes ( right panel ) .
the dotted line represents the average fraction of observed wind variability .
figure [ fig4 ] : the full - width ( @xmath28 , top panel ) and blue edge ( @xmath29 , bottom panel ) of the variability observed in the wind with respect to the terminal velocity ( @xmath14 ) .
triangles correspond to the mw stars , and squares denote stars in the lmc .
typical measurement errors are about 10% for @xmath1 and 100200 @xmath0 for @xmath28 and @xmath29 .
the dotted lines indicate a 1:1 relationship .
the dashed lines are a least - squares fit weighted by the inverse of the quadratic sum of the errors to the data ( only mw sample in top panel , both samples in bottom panel ) . figure [ fig5 ] : absorption components in the wind profiles of early - type stars in the galaxy .
the profiles are plotted in flux units versus velocity with respect to @xmath91032 ( panel a ) or @xmath91038 ( panel b ) .
@xmath81 indicates the time in days between successive observations .
black , red , and blue profiles indicate the first , second , and third observations , respectively .
the vertical dotted lines indicate the `` centroid '' of the dac .
see @xcite for other examples in the mw and the lmc .
llrrccccccc hde303308 & o3v((f ) ) & 287.60 & @xmath83 & 8.17 & 111@xmath84 & 1.7 & 2000 - 05 - 25 & p1221601 & + & w73 & & & 0.45 & 3035@xmath84 & & 2000 - 05 - 27 & p1221602 & + hd64568 & o3v((f@xmath85 ) ) & 243.14 & @xmath86 & 9.38 & & 2.5 & 2000 - 03 - 25 & p1221104 & + & w82 & & & 0.37 & & 5.7 & 2000 - 04 - 03 & p1221102 & + & & & & & & & 2000 - 04 - 05 & p1221103 & + cpd-592600 & o6v((f ) ) & 287.60 & @xmath87 & 8.61 & 142@xmath84 & 1.9 & 2000 - 03 - 21 & p1221401 & + & w73 & & & 0.53 & 3120@xmath84 & 1.9 & 2000 - 03 - 23 & p1221402 & + & & & & & & & 2000 - 03 - 25 & p1221403 & + cpd-592603 & o7v & 202.18 & @xmath88 & 8.77 & 164@xmath89 & 67.3 & 2000 - 01 - 30 & s3040501 & + & w73 & & & 0.46 & 1840@xmath89 & 17.0 & 2000 - 03 - 21 & p1221501 & + & & & & & & & 2000 - 04 - 06 & s3040502 & + hd152590 & o7.5v & 344.84 & @xmath90 & 8.48 & 60@xmath84 & 35.1 & 2001 - 07 - 08 & b0710601 & rv var + & w72 & & & 0.38 & 1785@xmath84 & & 2001 - 07 - 12 & b0710602 & + hd61347 & o9ib & 230.60 & @xmath91 & 8.43 & 116@xmath84 & 1.1 & 2000 - 04 - 15 & p1022001 & + & m55 & & & 0.45 & 1775@xmath84 & & 2000 - 04 - 16 & p1022002 & + hd210809 & o9iab & 99.85 & @xmath92 & 7.54 & 117@xmath84 & 1.4 & 2000 - 08 - 05 & p1223101 & rv var + & w76 & & & 0.33 & 2135@xmath84 & 1.3 & 2000 - 08 - 07 & p1223102 & + & & & & & & & 2000 - 08 - 08 & p1223103 & + hd153426 & o9ii & 258.07 & @xmath93 & 7.47 & 108@xmath84 & 105 & 2000 - 03 - 31 & p1027201 & rv var + & w73 & & & 0.45 & 2200@xmath84 & & 2000 - 07 - 14 & p1027202 & + hd91651 & o9v : n & 286.55 & @xmath94 & 8.84 & 292@xmath84 & 1.3 & 2000 - 05 - 25 & p1023101 & + & w73 & & & 0.29 & 1705@xmath84 & & 2000 - 05 - 27 & p1023102 & + hd92554 & o9.5iin & 287.60 & @xmath95 & 9.47 & 298@xmath84 & 1.5 & 2000 - 05 - 26 & p1023201 & rv var ? + & g77 & & & 0.39 & 1260@xmath84 & & 2000 - 05 - 27 & p1023202 & + hd168941 & o9.5ii - iii & 275.23 & @xmath96 & 9.34 & 116@xmath84 & 2.2 & 2000 - 08 - 30 & p1016501 & rv var + & w82 & & & 0.37 & 1795@xmath84 & & 2000 - 09 - 01 & p1016502 & + hd156292 & o9.5iii & 345.35 & @xmath97 & 7.49 & 102@xmath84 & 1.9 & 2000 - 04 - 03 & p1027402 & rv var + & w73 & & & 0.56 & 1355@xmath84 & & 2000 - 04 - 05 & p1027403 & + hd163892 & o9iv ( ( n ) ) & 7.15 & @xmath98 & 7.44 & 201@xmath84 & 1.8 & 2001 - 04 - 27 & p1027601 & rv var + & w73 & & & 0.47 & 1405@xmath84 & & 2001 - 04 - 29 & p1027602 & + hde308813 & o9.5v & 294.80 & @xmath99 & 9.38 & & 2.0 & 2000 - 03 - 23 & p1221901 & rv var + & s70 & & & 0.37 & & 1.3 & 2000 - 03 - 25 & p1221902 & + & & & & & & & 2000 - 03 - 27 & p1221903 & + cpd@xmath53721184 & b0iii & 233.12 & @xmath100 & 10.68 & & 0.8 & 2000 - 03 - 27 & s5140101 & + & h74 & & & 0.23 & & 1.1 & 2000 - 03 - 28 & s5140102 & + & & & & & & & 2000 - 03 - 29 & s5140103 & + hd192035 & b0iii - iv(n ) & 68.81 & @xmath101 & 6.44 & 155@xmath102 & 2.5 & 2000 - 06 - 17 & p1028601 & + & w71 & & & 0.42 & & 2.0 & 2000 - 06 - 19 & p1028602 & + & & & & & & & 2000 - 06 - 21 & p1028603 & + hd47417 & b0iv & 205.35 & @xmath103 & 6.97 & & 0.8 & 2000 - 03 - 15 & p1021602 & + & m55 & & & 0.31 & & & 2000 - 03 - 16 & p1021601 & + hd203374a & b0ivpe & 100.26 & @xmath104 & 6.69 & 350@xmath102 & 0.4 & 2001 - 08 - 02 & b0300101 & + & m55 & & & 0.60 & & & 2001 - 08 - 02 & b0300102 & + bd+532820 & b0ivn & 9.26 & @xmath105 & 9.95 & & 1.5 & 2000 - 08 - 06 & p1223201 & + & h56 & & & 0.40 & & 1.3 & 2000 - 08 - 07 & p1223202 & + & & & & & & & 2000 - 08 - 08 & p1223203 & + hd191495 & b0iv - v & 72.74 & @xmath106 & 8.26 & 235@xmath102 & 1.0 & 2000 - 08 - 10 & p1222901 & rv var + & w71 & & & 0.40 & & & 2000 - 08 - 11 & p1222902 & + hd195965 & b0v & 83.33 & @xmath107 & 8.20 & & 141.2 & 2000 - 06 - 17 & p1028801 & + & m55 & & & 0.36 & & 0.9 & 2000 - 11 - 08 & p1028802 & + & & & & & & & 2000 - 11 - 09 & p1028803 & + hd186994 & b0.2iv & 78.62 & @xmath108 & 7.51 & 125@xmath102 & 63.4 & 2001 - 07 - 02 & p2160801 & rv var + & w71 & & & 0.20 & & & 2001 - 09 - 07 & p2160802 & + hd207538 & b0.2v & 12.12 & @xmath109 & 7.31 & 51@xmath84 & 226 & 1999 - 12 - 08 & p1162902 & + & ma02 & & & 0.63 & & & 2000 - 07 - 21 & p1162903 & + hd187459 & b0.5ib & 68.81 & @xmath101 & 6.44 & 139@xmath84 & 0.8 & 2000 - 08 - 10 & p1028201 & rv var ? + & m55 & & & 0.42 & 1185@xmath84 & & 2000 - 08 - 11 & p1028202 & + hd224151 & b0.5ii & 115.44 & @xmath110 & 6.00 & 115@xmath84 & 2.1 & 2000 - 08 - 11 & p1224101 & rv var + & m55 & & & 0.48 & 1280@xmath84 & 0.6 & 2000 - 08 - 13 & p1224102 & + & & & & & & & 2000 - 08 - 14 & p1224103 & + hde332407 & b0.5iii : & 64.28 & @xmath111 & 8.50 & 140@xmath102 & 1.8 & 2000 - 06 - 10 & p1222801 & + & w71 & & & 0.48 & & & 2000 - 06 - 12 & p1222802 & + hd161807 & b0.5iii : n & 351.78 & @xmath112 & 6.99 & 350@xmath102 & 21.9 & 2000 - 08 - 17 & p1222301 & + & g77 & & & 0.23 & & & 2000 - 09 - 08 & p1222302 & + hd172140 & b0.5iii & 5.28 & @xmath113 & 9.93 & & 1.5 & 2000 - 05 - 18 & p1016601 & + & h70 & & & 0.25 & & 2.1 & 2000 - 05 - 20 & p1016602 & + & & & & & & & 2000 - 05 - 22 & p1016603 & + hd62866 & b0.5iiin & 237.48 & @xmath114 & 9.01 & & 8.9 & 2000 - 04 - 02 & p1221002 & + & g77 & & & 0.35 & & & 2000 - 04 - 11 & p1221004 & + hd97913 & b0.5ivn & 290.84 & @xmath106 & 8.80 & 310@xmath102 & 1.1 & 2000 - 05 - 26 & p1221701 & + & g77 & & & 0.32 & & & 2000 - 05 - 27 & p1221702 & + bd+354258 & b0.5vn & 77.19 & @xmath115 & 9.41 & & 3.0 & 2000 - 06 - 14 & p1017901 & rv var + & m55 & & & 0.29 & & & 2000 - 06 - 17 & p1017902 & + hd148422 & b1ia & 329.92 & @xmath116 & 8.60 & 81@xmath84 & 2.1 & 2000 - 04 - 08 & p1015001 & + & hc75 & & & 0.28 & 1335@xmath84 & 1.9 & 2000 - 04 - 10 & p1015002 & + & & & & & & & 2000 - 04 - 12 & p1015003 & + hd191877 & b1ib & 61.57 & @xmath117 & 6.28 & 152@xmath84 & 420 & 2000 - 06 - 05 & p1028701 & rv var ? + & m55 & & & 0.18 & 1160@xmath84 & & 2001 - 07 - 30 & p2051101 & + hde235783 & b1ib & 101.69 & @xmath118 & 8.68 & 93@xmath84 & 1.5 & 2000 - 08 - 05 & p1223301 & + & m55 & & & 0.36 & 1070@xmath84 & 1.3 & 2000 - 08 - 07 & p1223302 & + & & & & & & & 2000 - 08 - 08 & p1223303 & + bd+483437 & b1iab & 354.70 & @xmath119 & 8.69 & 54@xmath84 & 1.3 & 2000 - 08 - 06 & p1018401 & + & m55 & & & 0.35 & & 1.4 & 2000 - 08 - 07 & p1018402 & + & & & & & & & 2000 - 08 - 09 & p1018403 & + hd225757 & b1iiin & 309.48 & @xmath120 & 10.59 & & 2.0 & 2000 - 08 - 06 & p1017701 & + & c76 & & & 0.22 & & 0.6 & 2000 - 08 - 08 & p1017702 & + & & & & & & & 2000 - 08 - 09 & p1017703 & + hd91597 & b1iiine & 286.86 & @xmath121 & 9.84 & & 1.4 & 2000 - 02 - 04 & p1023002 & + & g77 & & & 0.30 & & 2.3 & 2000 - 02 - 06 & p1023003 & + & & & & & & & 2000 - 02 - 08 & p1023004 & + bd+523210 & b1v & 10.9608 & @xmath122 & 10.69 & & 1.3 & 2000 - 08 - 06 & p1223501 & + & m55 & & & 0.24 & & 1.3 & 2000 - 08 - 07 & p1223502 & + & & & & & & & 2000 - 08 - 08 & p1223503 & + hd73 & b1.5iv & 114.17 & @xmath123 & 8.48 & 105@xmath102 & 49.9 & 2000 - 08 - 01 & p1010101 & + & w71 & & & 0.07 & & & 2000 - 09 - 20 & p1010102 & + hd202347 & b1.5v & 44.85 & @xmath124 & 7.50 & 125@xmath102 & 84.7 & 2000 - 06 - 20 & p1028901 & + & w71 & & & 0.17 & & & 2000 - 09 - 13 & p1028902 & + bd+532885 & b2iii & 102.75 & @xmath125 & 10.46 & & 2.7 & 2000 - 08 - 02 & p1223601 & + & m55 & & & 0.28 & & & 2000 - 08 - 05 & p1223602 & + hd47961 & b2v & 203.02 & @xmath126 & 7.50 & & 11.7 & 2001 - 02 - 20 & p1310401 & + & k69 & & & 0.08 & & & 2001 - 03 - 04 & p1310402 & + hd51013 & b3v & 235.11 & @xmath127 & 8.81 & & 438 & 2000 - 01 - 24 & a0630901 & + & hs88 & & & 0.05 & & & 2001 - 04 - 05 & a0630902 & + hd47777 & b3v & 203.12 & @xmath128 & 7.90 & 135@xmath102 & 11.7 & 2001 - 02 - 20 & p1310201 & + & g68 & & & 0.07 & & & 2001 - 03 - 04 & p1310202 & + llrrccccccc sk@xmath5367167 & o4inf@xmath129 & 277.87 & @xmath130 & 12.54 & 2005@xmath84 & 285 & 1999 - 12 - 17 & p1171902 & + & w95 & & & 0.14 & & & 2000 - 09 - 27 & p1171901 & + sk@xmath537145 & o4 - 5iiif & 281.86 & @xmath131 & 11.47 & 2500@xmath102 & 2.7 & 2000 - 10 - 02 & p1031501 & + & w77 & & & 0.20 & & 1.0 & 2000 - 10 - 05 & p1031503 & + & & & & & & 0.5 & 2000 - 10 - 06 & p1031502 & + & & & & & & & 2000 - 10 - 06 & p1031504 & + sk@xmath5367111 & o6ia(n)fp var & 277.75 & @xmath132 & 12.57 & 2090@xmath133 & 138 & 1999 - 09 - 26 & x0200101 & rv var ? + & w02 & & & 0.11 & & & 2000 - 10 - 20 & p1173001 & + bi208 & o7v@xmath134 & 277.54 & @xmath135 & 14.02 & & 0.5 & 2000 - 09 - 30 & p1172704 & + & wpc & & & 0.03 & & 0.1 & 2000 - 10 - 01 & p1172703 & + & & & & & & 1.2 & 2000 - 10 - 01 & p1172702 & + & & & & & & & 2000 - 10 - 01 & p1172705 & + bi272 & o7iii - ii : & 277.24 & @xmath136 & 13.20 & 3400@xmath102 & 287 & 1999 - 12 - 18 & p1172902 & + & wpc & & & 0.17 & & & 2000 - 09 - 29 & p1172901 & + sk@xmath5367101 & o8ii((f ) ) & 277.78 & @xmath137 & 12.63 & 2005@xmath84 & 284 & 1999 - 12 - 20 & p1173403 & + & w02 & & & & & & 2000 - 09 - 29 & p1173401 & + bi173 & o8ii : & 279.67 & @xmath138 & 13.00 & 2850@xmath102 & 61.9 & 2000 - 10 - 03 & p1173201 & rv var + & w02 & & & 0.17 & & & 2000 - 12 - 04 & p1173202 & + sk@xmath5367191 & o8v & 277.66 & @xmath139 & 13.46 & 1750@xmath84 & 287 & 1999 - 12 - 17 & p1173102 & + & c86 & & & 0.10 & & & 2000 - 09 - 29 & p1173101 & + sk@xmath536803 & o9i & 279.70 & @xmath140 & 13.13 & & 1.3 & 2000 - 10 - 04 & a0490402 & + & c86 & & & 0.48 & & & 2000 - 10 - 05 & a0490401 & + sk@xmath5369124 & o9ib : & 279.61 & @xmath141 & 12.66 & 1430@xmath84 & 62.1 & 2000 - 10 - 03 & p1173601 & + & wpc & & & 0.12 & & & 2000 - 12 - 04 & p1173602 & + sk@xmath536544 & o9v & 277.78 & @xmath137 & 12.63 & & 140 & 2001 - 10 - 27 & p1173401 & + & c86 & & & 0.14 & & & 2002 - 03 - 16 & p1173401 & + sk@xmath536705 & o9.7ib & 278.89 & @xmath142 & 11.34 & 1665@xmath133 & 2.8 & 2000 - 10 - 04 & p1030704 & + & f88 & & & 0.15 & & & 2000 - 10 - 07 & p1030703 & + sk@xmath536521 & o9.7iab & 276.19 & @xmath143 & 12.02 & 1330@xmath133 & 0.6 & 2000 - 10 - 05 & p1030901 & + & w95 & & & 0.20 & & 0.1 & 2000 - 10 - 05 & p1030904 & + & & & & & & 0.4 & 2000 - 10 - 05 & p1030903 & + & & & & & & & 2000 - 10 - 05 & p1030902 & + sk@xmath537108 & o9ii & 282.52 & @xmath144 & 13.25 & & 636 & 1999 - 12 - 21 & a0491401 & + & c86 & & & 0.08 & & & 2001 - 09 - 17 & a0491402 & + sk@xmath537085 & b0i & 281.26 & @xmath145 & 12.30 & & 636 & 1999 - 12 - 21 & a0491301 & + & j01 & & & 0.15 & & & 2001 - 09 - 17 & a0491302 & + sk@xmath536841 & b0.5ia & 279.02 & @xmath146 & 12.01 & 865@xmath84 & 290 & 1999 - 12 - 18 & p1174102 & + & f91 & & & 0.16 & & & 2000 - 10 - 03 & p1174101 & + av488 & b0.5iaw & 300.51 & @xmath147 & 11.90 & 1040@xmath84 & 0.1 & 2000 - 10 - 08 & p1176803 & + & w95 & & & 0.14 & & 2.0 & 2000 - 10 - 08 & p1176802 & + & & & & & & & 2000 - 10 - 10 & p1176801 & + sk@xmath536728 & b0.7ia & 278.06 & @xmath148 & 12.28 & & 4.1 & 1999 - 12 - 16 & a0490201 & + & f88 & & & 0.10 & & & 1999 - 12 - 20 & a0490202 & + sk@xmath536875 & b1i & 278.65 & @xmath149 & 12.03 & & 6.5 & 2000 - 10 - 03 & a0490501 & + & j01 & & & 0.19 & & & 2000 - 10 - 10 & a0490502 & + sk@xmath5370120 & b1ia & 280.72 & @xmath150 & 11.59 & & 0.1 & 2000 - 09 - 28 & a0491002 & + & f88 & & & 0.14 & & & 2000 - 09 - 28 & a0491001 & + sk@xmath536714 & b1.5ia & 278.27 & @xmath151 & 11.52 & 610@xmath84 & 4.3 & 2000 - 09 - 27 & p1174201 & + & f91 & & & 0.10 & & 0.6 & 2000 - 10 - 01 & p1174203 & + & & & & & & & 2000 - 10 - 02 & p1174202 & + av18 & b2ia & 303.36 & @xmath152 & 12.48 & & 381 & 2000 - 05 - 29 & a1180101 & + & l97 & & & 0.21 & & & 2001 - 06 - 13 & b0890101 & + & & & & & & & & & + lllll hde303308 & o3v((f ) ) & v. @xmath153 $ ] & n.v . & n.v .
+ hd64568 & o3v((f@xmath85 ) ) & n.v . & n.v . & n.v .
+ cpd-592600 & o6v((f ) ) & v.(n ) & n.v . & n.v .
+ cpd-592603 & o7v & v. @xmath154 $ ] & n.v . & n.v .
+ hd152590 & o7.5v & v. @xmath155 $ ] & n.v . & n.v .
+ hd61347 & o9ib & v. @xmath156 $ ] & v. @xmath157 $ ] & n.v .
( n ) + hd210809 & o9iab & v. @xmath158 $ ] & v. @xmath159 $ ] & v. @xmath159 $ ] + hd153426 & o9ii & n.v.(n ) & n.v . & n.v .
+ hd91651 & o9v : n & n.v . & n.v . & n.v .
+ hd92554 & o9.5iin & v. @xmath160 $ ] & v. @xmath161 $ ] & n.v .
( n ) + hd168941 & o9.5ii - iii & v. @xmath162 $ ] & n.v . & n.v .
+ hd156292 & o9.5iii & n.v . & n.v . & n.v .
+ hd163892 & o9iv ( ( n ) ) & v. @xmath163 $ ] & n.v . & n.v
. + hde308813 & o9.5v & n.v . & n.v . & n.v
. + hd186994 & b0.2iv & n.v . & n.v . & n.v
. + cpd@xmath53721184 & b0iii & n.v . & n.v . & n.v .
+ hd192035 & b0iii - iv(n ) & v. @xmath164 $ ] & n.v . & n.v .
+ hd47417 & b0iv & v. @xmath165 $ ] & n.v . & n.v .
+ hd203374a & b0ivpe & n.v . & n.v . & n.v
. + bd+532820 & b0ivn & n.v . & n.v . & n.v
. + hd191495 & b0iv - v & v.(n ) & n.v . & n.v
. + hd195965 & b0v & n.v . & n.v . & n.v
. + hd207538 & b0.2v & n.v . & n.v . & n.v .
+ hd187459 & b0.5ib & v. @xmath166 $ ] & v. @xmath166 $ ] & v. @xmath166 $ ] + hd224151 & b0.5ii & v. @xmath167 $ ] & n.v . & n.v . + hde332407 & b0.5iii : & v.(n ) & n.v . & n.v .
+ hd161807 & b0.5iii : n & v. @xmath168 $ ] & n.v . & n.v .
+ hd172140 & b0.5iii & v. @xmath169 $ ] & n.v . & n.v .
+ hd62866 & b0.5iiin & v. @xmath170 $ ] & n.v . & n.v .
+ hd97913 & b0.5ivn & v. @xmath171 $ ] & n.v . & n.v .
+ bd+354258 & b0.5vn & v. @xmath172 $ ] & n.v . & n.v .
+ hd148422 & b1ia & v.(n ) & n.v . & n.v
. + hd191877 & b1ib & v.(n ) & n.v . & n.v
. + hde235783 & b1ib & v.(n ) & n.v . & n.v .
+ bd+483437 & b1iab & n.v . & n.v . & n.v
. + hd225757 & b1iiin & n.v . & n.v . & n.v .
+ hd91597 & b1iiine & v. @xmath156 $ ] & n.v . & n.v .
+ bd+523210 & b1v & n.v . & n.v . & n.v
. + hd73 & b1.5iv & n.v.(n ) & n.v . & n.v
. + hd202347 & b1.5v & n.v . & n.v . & n.v .
+ bd+532885 & b2iii & n.v . & n.v . & n.v
. + hd47961 & b2v & n.v . & n.v . & n.v
. + hd51013 & b3v & n.v . & n.v . & n.v
. + hd47777 & b3v & n.v . & n.v . & n.v
. + sk@xmath5367167 & o4inf@xmath129 & n.v . & n.v . & n.v
. + sk@xmath537145 & o4 - 5iiif & v.(n ) & n.v . & n.v .
+ sk@xmath5367111 & o6:iafpe & v. & v. @xmath173 $ ] & v. @xmath173 $ ] + bi208 & o7v@xmath134 & v. @xmath174 $ ] & n.v . & n.v .
+ bi272 & o7:iii - ii : & v. @xmath175 $ ] & n.v . & n.v .
+ sk@xmath5367101 & o8ii((f ) ) & v. @xmath176 $ ] & n.v . & n.v
. + bi173 & o8ii : & v. & n.v . & n.v .
+ sk@xmath5367191 & o8v & v. @xmath177 $ ] & n.v . & n.v .
+ sk@xmath536803 & o9i & n.v . & n.v . & n.v .
+ sk@xmath5369124 & o9ib : & v. @xmath178 $ ] & v.(n ) & n.v .
+ sk@xmath536544 & o9v & v. @xmath179 $ ] & n.v . & n.v .
+ sk@xmath536705 & o9.7ib & v. & v. & n.v . + sk@xmath536521 & o9.7iab & v. & v. & n.v
. + sk@xmath537108 & o9ii & n.v . & n.v . & n.v .
+ sk@xmath536841 & b0.5ia & v. @xmath180 $ ] & v. @xmath180 $ ] & n.v .
+ av488 & b0.5iaw & n.v . & n.v . & n.v
. + sk@xmath536728 & b0.7ia & n.v . & n.v . & n.v .
+ sk@xmath537085 & b0i & v. & v. @xmath180 $ ] & n.v
. + sk@xmath536875 & b1i & n.v . & n.v . & n.v
. + sk@xmath5370120 & b1ia & n.v . & n.v . & n.v
. + sk@xmath536714 & b1.5ia & n.v . & n.v . & n.v
. + av18 & b2ia & n.v . & n.v . & n.v . + llccc hd152590 & o7.5v & 1785 & 1700@xmath181 & strong + & & & 1300 & weak + hd61347 & o9ib & 1775 & 1250 & weak + hd153426 & o9ii & 2200 & 2150 & strong + hd92554 & o9.5iin & 1260 & 1125 & strong + hd168941 & o9.5ii - iii & 1795 & 1625@xmath181 & strong + & & & 1250 & weak + hd156292 & o9.5iii & 1355 & 1300 & strong + & & & 1100 & weak + hd163892 & o9iv ( ( n ) ) & 1405 & 1320 & medium + & & & 1175 & medium + cpd@xmath53721184 & b0iii & & 1075 & medium + bd+532820 & b0ivn & & 1200 & medium + hd187459 & b0.5ib & 1185 & 1125 & strong + hd224151 & b0.5ii & 1280 & 1250 & strong + hde332407 & b0.5iii : & & 1275 & medium + hd161807 & b0.5iii : n & & 1120 & weak + hd62866 & b0.5iiin & & 980 & weak + & & & 1130 & weak + hd191877 & b1ib & & 1000 & strong + hd91597 & b1iiine & & 1170 & weak + & & & 1370 & weak + sk@xmath5367101 & o8ii((f))&2005 & 950 & weak
+ sk@xmath5367191 & o8v & 1750 & 1700@xmath181 & strong + sk@xmath5369124 & o9ib : & 1430 & 1150 & medium + & & & 1350 & medium + sk@xmath536705 & o9.7ib & 1665 & 1000 & strong + sk@xmath536521 & o9.7iab & 1330 & 1000 & medium + sk@xmath537085 & b0i & & 850 & strong + sk@xmath536841 & b0.5ia & 865 & 800 & strong + | we have used the _ far ultraviolet spectroscopic explorer ( fuse ) _ to conduct a snap - shot survey of variability in the winds of 66 ob - type stars in the galaxy and the magellanic clouds .
these time series consist of two or three observations separated by intervals ranging from a few days to several months .
although these time series provide the bare minimum of information required to detect variations , this survey demonstrates that the doublet in the winds of ob - type stars is variable on various scales both in time and velocity . for spectral types from o3 to b1 ,
64% vary in time . at spectral types later than b1 , no wind variability is observed . in view of the limitations of this survey , this fraction represents a lower limit on the true incidence of variability in the wind lines , which is very common and probably ubiquitous . in contrast , for and , only a small percentage of the whole sample shows wind variations , although this may be principally due to selection effects .
the observed variations extend over several hundreds of @xmath0 of the wind profile and can be strong .
the width over which the wind profile varies is only weakly correlated with the terminal velocity ( @xmath1 ) , but a significant correlation ( close to a 1:1 relationship ) is derived between the maximum velocity of the variation and @xmath1 .
high velocity wind absorption features ( possibly related to the discrete absorption components seen in other wind lines ) are also observed in 46% of the cases for spectral types from o3 to b0.5 .
these features are variable , but the nature of their propagation can not be determined from this survey . if x - rays can produce sufficient by auger ionization of , and the x - rays originate from strong shocks in the wind , this study suggests that stronger shocks occur more frequently near @xmath1 , causing an enhancement of near @xmath1 . |
this is a revised version of the article that we submit before .
there was a problem in the construction of graphical ends .
we are presently working to fix it ( replace the previous boundary with a planar boundary curve and use perron method ) .
the main geometric constructions will be mantained . here
we present the halfspace type theorem , that correspond to section 4 of the previous article .
d. hofmann e w. meeks proved a beautiful theorem on minimal surfaces , the so - called `` halfspace theorem '' in @xcite : there is no non planar , complete , minimal surface properly immersed in a halfspace of @xmath6 we focus in this paper complete surfaces with constant mean curvature @xmath4 in the product space @xmath1 where @xmath2 is the hyperbolic plane and @xmath3 is the set of real numbers . in the context of @xmath7-surfaces in @xmath1
it is natural to investigate about halfspace type results . before stating our result we would like to emphasize that , in last years there has been work in @xmath7-surfaces in homogeneous 3-manifolds , in particular in the product space @xmath8 new examples were produced and many theoretical results as well .
halfspace theorem for minimal surfaces in @xmath9 is false , in fact there are many vertically bounded complete minimal surfaces in @xmath9 @xcite . on the contrary , we are able to prove the following result for @xmath4 surfaces .
[ halfspacetheorem ] let @xmath10 be a simply connected rotational surface with constant mean curvature @xmath11 let @xmath12 be a complete surface with constant mean curvature @xmath0 different from a rotational simply connected one .
then , @xmath12 can not be properly immersed in the mean convex side of @xmath13 in @xcite , l. hauswirth , h. rosenberg and j. spruck prove a halfspace type theorem for surfaces on one side of a horocylinder .
the result in @xcite is different in nature from our result because in @xcite , the `` halfspace '' is one side of a horocylinder , while for us , the `` halfspace '' is the mean convex side of the rotational simply connected surface .
the proof of our result is a geometric application of the classical maximum principle to surfaces with constant mean curvature @xmath4 in @xmath5 * maximum principle . *
_ let @xmath14 and @xmath15 be two connected surfaces of constant mean curvature @xmath11 let @xmath16 be a point such that @xmath14 and @xmath15 are tangent at @xmath17 the mean curvature vectors of @xmath14 and @xmath15 at @xmath18 point towards the same side and @xmath14 is on one side of @xmath15 in a neighborhood of @xmath19 then @xmath14 coincide with @xmath15 around @xmath19 by analytic continuation , they coincide everywhere . _
the proof of the maximum principle is based on the fact that a constant mean curvature surface in @xmath9 locally satisfies a second order elliptic pde ( cf .
@xcite , @xcite where the author prove the maximum principle in @xmath20 the proof generalizes to space forms and to @xmath9 as well ) .
we notice that our surfaces are not compact , while the classical maximum principle applies at a finite point
. it will be clear in the proof of theorem [ halfspacetheorem ] that we that we are able to reduce the analysis to finite tangent points , because of the geometry of rotational surfaces of constant mean curvature @xmath11 our halfspace theorem leads to the following conjecture ( strong halfspace theorem ) .
* conjecture . * _ let @xmath21 @xmath22 be two complete properly embedded surfaces with constant mean curvature @xmath0 different from the rotational simply connected one .
then @xmath23 can not lie in the mean convex side of @xmath24 @xmath25 _ for @xmath26
the conjecture is true and it is known as maximum principle at infinity ( cf .
r. sa earp and e. toubiana find explicit integral formulas for rotational surfaces of constant mean curvature @xmath27 $ ] in @xcite . a careful description of the geometry of these surfaces is contained in the appendix of @xcite . for @xmath30 the surface
@xmath29 has two vertical ends ( where a vertical end is a topological annulus , with no asymptotic point at finite height ) that are graphs over the exterior of a disk @xmath31 of hyperbolic radius @xmath32 by graph we mean the following : the graph of a function @xmath33 defined on a subset @xmath34 of @xmath2 is @xmath35 when the graph has constant mean curvature @xmath36 @xmath33 satisfies the following second order elliptic pde furthermore , up to vertical translation , one can assume that @xmath29 is symmetric with respect to the horizontal plane @xmath43 for @xmath44 the surface @xmath45 has only one end and it is a graph over @xmath2 and it is denoted by @xmath13 for any @xmath28 each end of the surface @xmath29 is the vertical graph of a function @xmath48 over the exterior of a disk @xmath31 of radius @xmath49 the asymptotic behavior of @xmath48 has the following form : @xmath50 @xmath51 where @xmath52 is the hyperbolic distance from the origin . the positive number @xmath53 is called the _ growth _ of the end .
the function @xmath48 is vertical along the boundary of @xmath55 furthermore the radius @xmath56 is always greater or equal to zero , it is zero if and only if @xmath57 and tends to infinity as @xmath58 or @xmath59 as we pointed out before , the function @xmath60 is entire and its graph corresponds to the unique simply connected example @xmath13 notice that , any end of an immersed rotational surface ( @xmath46 ) has growth smaller than the growth of @xmath61 while any end of an embedded rotational surface ( @xmath47 ) has growth greater than the growth of @xmath13 * proof of theorem [ halfspacetheorem ] .
* one can assume that the surface @xmath10 is tangent to the slice @xmath62 at the origin and it is contained in @xmath63 suppose , by contradiction , that @xmath12 is contained in the mean convex side of @xmath13 lift vertically @xmath13 if there is an interior contact point between @xmath12 and the translation of @xmath61 one has a contradiction by the maximum principle .
as @xmath12 is properly immersed , @xmath12 is asymptotic at infinity to a vertical translation of @xmath13 one can assume that the surface @xmath12 is asymptotic to the @xmath10 tangent to the slice @xmath62 at the origin and contained in @xmath63 * @xmath66 and @xmath12 has a first finite contact point @xmath68 this means that @xmath69 does not meet @xmath12 at a finite point , for @xmath70 and then @xmath66 and @xmath12 are tangent at @xmath18 with mean curvature vector pointing in the same direction . in this case , by the maximum principle @xmath66 and @xmath12 should coincide .
* @xmath66 and @xmath12 meet at a point @xmath17 but @xmath18 is not a first contact point.then , for @xmath71 small enough , @xmath69 intersect @xmath12 transversally .
it follows from the maximum principle that there are no compact component of @xmath12 contained in @xmath74 denote by @xmath75 a non compact connected component of @xmath12 contained in @xmath74 note that the boundary of @xmath75 is contained in @xmath73 consider the family of rotational non embedded surfaces @xmath76 @xmath77 translate each @xmath29 vertically in order to have the waist on the plane @xmath78 by abuse of notation , we continue to call the translation , @xmath79 the surface @xmath29 intersects the plane @xmath80 in two circles .
denote by @xmath81 the radius of the larger circle .
denote by @xmath82 the part of the surface outside the cylinder of radius @xmath83 notice that @xmath84 is embedded . by the geometry of the @xmath85 when @xmath86 is great enough ,
say @xmath87 @xmath88 is outside the mean convex side of @xmath13 then , @xmath88 does not intersect @xmath65 furthermore , when @xmath89 @xmath84 converge to @xmath73 now , start to decrease @xmath86 from @xmath90 to one . before reaching @xmath44 the surface @xmath91 first meets @xmath10 and then touches @xmath75 tangentially at an interior finite point , with @xmath75 above @xmath92 this depends on the following two facts .
* the boundary of @xmath75 lies on @xmath69 and the boundary of any of the @xmath84 lies on the horizontal plane @xmath78 * the growth of any of the @xmath84 is strictly smaller than the growth of @xmath13 thus the end of @xmath93 is outside the end of @xmath10 .
b. nelli , h. rosenberg : _ minimal surfaces in @xmath94 _ bull .
33 ( 2002 ) 263 - 292 ; errata minimal surfaces in @xmath1 [ bull .
new series 33 ( 2002 ) , 263 - 292 ] bull .
, new series 38(4 ) ( 2007 ) 1 - 4 .
b. nelli , r. sa earp , w. santos , e. toubiana : _ existence and uniqueness of @xmath7-surfaces with one or two parallel convex curves as boundary in @xmath1 _ annals of global analysis and geometry , doi 10.1007/s10455 - 007 - 9087 - 3 , * 33 * , n. 4 ( 2008 ) 307 - 321 .
r. sa earp , e. toubiana : _ an asymptotic theorem for minimal surfaces and existence results for minimal graphs in @xmath96 _ published on line may 2008 , mathematische annalen , doi 10.1007/s00208 - 008 - 0237 - 0 , 342 , n. 2 ( 2008 ) 309 - 331 | _ we prove a vertical halfspace theorem for surfaces with constant mean curvature @xmath0 properly immersed in the product space @xmath1 where @xmath2 is the hyperbolic plane and @xmath3 is the set of real numbers .
the proof is a geometric application of the classical maximum principle for second order elliptic pde , using the family of non compact rotational @xmath4 surfaces in @xmath5 _ |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.